Sei sulla pagina 1di 14

Published in IET Generation, Transmission & Distribution

Received on 24th February 2010


Revised on 9th May 2010
doi: 10.1049/iet-gtd.2010.0143
ISSN 1751-8687
Usage-based transmission loss allocation
under open access in deregulated
power systems
P.V. Satyaramesh
1
C. RadhaKrishna
2
1
APTRANSCO, Vidyut Soudha, Hyderabad, AP 500062, India
2
UGC ASC, J.N.T. University, Hyderabad, AP, India
E-mail: satya_pasu@yahoo.com
Abstract: This study presents a usage-based methodology of transmission loss allocation in deregulated power
systems under open access. This new approach calculates the portion of real power transmission loss
contributions from the generators and simultaneously the portion of real power transmission loss allocated to
the loads using their contract obligations with the generators in the open access environment. A power ow
procedure is used to calculate power loss in the system. It is desirable to take network loss effect of injection
power at each node for calculating contribution of transmission loss by each generator and loss allocated to
loads based on its contractual obligations with consumer. In this study, a methodology is proposed to
calculate the total real power loss in transmission network into components to be allocated to the generators
and loads. This study focuses on development and application of the equivalent loss compensation concept
for open access environment. Effectiveness of the proposed algorithms has been demonstrated on sample
5-bus and IEEE-118 bus power networks.
1 Introduction
In the open access system, consumers/Discom companies
require a fair and equitable pricing structure that reects
both the share of power consumed in the network and the
cost of active transmission loss, based on loss that they
cause. Loss is always present in transmission lines and
transformers because of resistances. Total loss in
transmission system typically amounts only to the extent of
35% of the total generation. Even though this gure is
small, it is signicant in terms of accumulated effect on
revenue. Any proposal for restructuring without a solution
to the problems because of loss is incomplete and
unacceptable. The loss allocations have inuence on
decision making of the electricity-market participants for
their nancial commitments and their prots. There is a
need to nd the contribution of loss by each generator and
distribute the same among market participants. In general,
each trade should include its share of transmission loss. In
essence, the net generation should equal the sum of the
demands and the transmission loss caused by the trade.
The total transmission loss caused by all the trades on the
network can be either measured or calculated.
Many different loss allocation schemes exist in this area,
but no single method gained universal acceptance. The
problem of allocating the transmission-active power loss
among the power system users has become more important
with the increase in the competition level in electricity
markets. Commercial issues related to charges for power
loss are taken into consideration, subject to negotiations
between consumers/distribution utilities and generating
companies.
In open access system, a non-prot organisation known as
independent grid operator (IGO) usually is responsible for
the operation of the system. In addition to the operation
and control of a system, typical tasks of an IGO may also
include accepting schedules from generators, providing
access to the successful consumers and allocating
transmission loss among the generators and consumers. An
IGO also plays the role of a supervisor for system planning
IET Gener. Transm. Distrib., 2010, Vol. 4, Iss. 11, pp. 12611274 1261
doi: 10.1049/iet-gtd.2010.0143 & The Institution of Engineering and Technology 2010
www.ietdl.org
and security. It would keep track of all information related to
the trading and calculate the transmission usage for each
generator. In this model, suppliers and consumers
independently arrange trades, setting by themselves under
concurrence of IGO the amount of generation and
consumption and the corresponding nancial terms.
Coordination is necessary because physical laws dictate the
ow of power from the generators to the consumers in a
transmission network. The IGO coordinates among the
independent trades, which would not lead to a violation of
transmission network constraints. In addition, power ows
must be balanced throughout the network and transmission
loss must be included in power balance. The energy loss in
the transmission network is a function of the aggregate
trades, and therefore the trades have to account for their
own loss.
The main other concern of the IGO, with respect to
transmission loss, is to allocate loss either to the generators
or consumers. Based on the allocation of transmission loss
and previously agreed terms, a generator or consumer is
required either to compensate or to pay its share of loss. In
either case, the allocation of transmission loss is a debatable
issue for the simple fact that transmission loss is a complex
non-linear function of power provided by the generators
and sensitive to state of system. This methodology is
computationally efcient and can be utilised to calculate the
loss contributions from the generators and its allocation to
consumers. The objective is to recover as much of the
active power loss as possible subject to the terms and
conditions of agreement. IGO employs loss-compensation
approaches to balance the difference between the total
system loss and the recovered loss.
The evaluation of loss allocations by proportional sharing
procedures has been widely discussed in many papers [111].
Cost-based transmission loss allocation methods are
presented in [12]. These methods are suggested to allocate
the system loss to generators and loads in a pool market or
to individual transactions in a bi-lateral contracts market.
They are mainly classied into four categories: (i) Pro-rata
methods allocate the transmission system loss to the
generators and loads proportional to their active power
generation and load consumption. The main disadvantage
of this approach is not taking topology of network into
account. It is not fair as it allocates the same amount of
loss when two identical loads are considered, in which one
may be located near to generators and the other is far away
from generators. (ii) Incremental transmission loss (ITL)
methods utilise the sensitivities with respect to nodal
injections to allocate the loss to generators and loads. The
ITL methods depend on the selection of the slack bus and
there is no allocation of loss to the slack bus. (iii)
Proportional sharing procedures allocate the system loss by
using the tracing techniques. The main drawback is that
there is no possibility of allocation of loss to generators and
loads at the same time. (iv) Transactional loss allocation
methods are formulated according to bilateral contracts in
competitive markets. This paper focuses on the allocation
of loss where generators jointly maintain the contractual
obligations with consumer/Discom companies in an open
access system.
The transmission loss compensation schemes considered
as ancillary services are presented in [1316]. A non-prot
organisation known as independent system operator usually
is responsible for the operation of these ancillary services.
The negative loss allocations are due to counter-ows. It is
due to the power ow that opposes the initial ow in a
particular transmission line [17, 18]. The net effects of
several transactions simultaneously reduce the net real
system loss. It is argued in [15] that negative loss
allocations are unreasonable, opting for heuristic allocation
in which for each transaction is assigned positive loss
allocation.
The content of present paper is organised as follows.
Section 2 presents mathematical formation of proposed loss
allocation methodology under open access environment.
Algorithm for the proposed approach is indicated in
Section 3. Sample 5 bus power systems is used as an
example to illustrate the proposed method in Section 4,
followed by conclusions in Section 5.
2 Mathematical formulation
of loss allocation in open access
system
A method to allocate transmission loss for simultaneous
bilateral transactions in open access is proposed. The paper
presents an approach to allocate total loss of the system
based on
(i) the exact loss formula, using bus injected powers and
(ii) a set of coefcients (the amounts of generation utilised by
consumers) that measure the participation of each bus
generator in relation to located-bus-contracted demands
and vice versa. These coefcients are named as usage
coefcients. There are no additional approximations
assumed in the proposed approach, resulting in avoidance
of inaccuracies induced during the formulation stages of the
method.
The standard loss formula for power network [19] is
expressed in terms of amount of power utilised by Discom
companies/consumers from individual generators. The
proposed method illustrates the splitting of total loss as
allocations to each generator and each load by using the
usage coefcients.
The methodology starts from a converged load ow
solution, which gives the entire information pertaining to
the network such as bus voltages, complex line ows, slack
1262 IET Gener. Transm. Distrib., 2010, Vol. 4, Iss. 11, pp. 12611274
& The Institution of Engineering and Technology 2010 doi: 10.1049/iet-gtd.2010.0143
www.ietdl.org
bus power generation and total transmission loss etc. Once a
load ow solution is obtained, this result is then adopted in
the proposed methodology to allocate transmission loss to
each load and generator using their usage coefcients.
Consider a power system with NGEN generators and NB
loads (no. of buses) connected through a transmission
network. It is assumed that all bus loads are supplied
through contractual obligations. An attempt is made here
to separate the non-linear system loss into the sum of NB
terms (one transaction per allocation of loss to loads) and
similarly the sum of NGEN terms (one transaction per
contribution of loss from generators). The main difculty
in allocating a component of system loss to a generator/
consumer arises because of the highly non-linear nature of
loss equation in which the combined set of all usage
coefcients interact through the load ow terms. Other
important factors are consideration of mutual interactions
between different transactions and interaction of reactive
power ows in the real power loss allocation. Thus, the loss
allocation procedure depends on path and the usage
coefcients of generators and loads.
Let the generators set G {G1, G2, G3, . . ., GNGEN}
and the load set L {L1, L2, L3, . . ., LNB}.
A transmission loss formula using bus injected powers and
the system parameters is given [19] as follows
P
L
=

NB
i=1

NB
j=1
[A
ij
(P
i
P
j
+ Q
i
Q
j
) + B
ij
(Q
i
P
j
P
i
Q
j
)] (1)
where
A
ij
=
R
ij
|V
i
||V
J
|
cos(d
i
d
j
)
B
ij
=
R
ij
|V
i
||V
J
|
sin(d
i
d
j
)
(2)
P
L
is the real power loss of the power system, V
i
is the voltage
magnitude of bus i, d
i
is the voltage phase angle of bus i, S
i
is
the injected power at bus i (S
i
P
i
+ jQ
i
), Z
ij
R
ij
+ jX
ij
and Z
ij
is the (i, j )th element of Zbus.
2.1 Active power loss allocation to
generators
Let a
ij
be the usage coefcient, that is fraction of power
generated at jth bus received by the load at ith bus.
The load at bus i can be expressed as the sum of usage
amounts from different generators that is
P
Loadi
=

NGEN
j=1
a
ij
P
Gj
, where i = 1, 2, . . . , NB (3)
The injected real power at bus i is given as
P
i
= P
Gi
P
Loadi
(4)
Substituting the above equation (3) in (4), the injected
powers can be written as
P
i
= P
Gi

NGEN
j=1
a
ij
P
Gj
, where i = 1, 2, . . . , NB (5)
The above equations can be rewritten as, by introducing x
ij,
P
i
=

NGEN
j=1
x
ij
P
Gj
, where i = 1, 2, . . . , NB (6)
where x
ij
2a
ij
, for non-generation buses; x
ij
1 2a
ij
,
for generator buses.
Rearrange the above equation (1) as components of self-
power (active, reactive) and mutual-power components
P
L
=

NB
i=1

NB
j=1
[A
ij
(P
i
P
j
) + B
ij
(Q
i
P
j
P
i
Q
j
)] + [A
ij
Q
i
Q
j
]
(7)
The injected active powers at ith and jth bus are written in the
form of (8) as shown below
P
i
=

NGEN
k=1
x
ik
P
Gk
, where i = 1, 2, . . . , NB (8)
P
j
=

NGEN
m=1
x
jm
P
Gm
, where j = 1, 2, . . . , NB (9)
The above equation (7) can be rewritten as
P
L
= P
L1
+ P
L2
+ P
L3
(10)
where the terms P
L1
, P
L3
can be treated as the loss
components because of solely the active and the reactive
power injections, respectively, and P
L2
represents as the loss
component created by interaction between the active and
reactive power injections.
Substituting active injected powers expressed in (8) and (9)
in the equation
P
L1
=

NB
i=1

NB
j=1
A
ij
P
i
P
j
(11)
and can be written as
P
L1
=

NGEN
k=1

NGEN
m=1

NB
i=1

NB
j=1
A
ij
x
ki
x
mj
P
Gk
P
Gm
_ _ _ _
(12)
IET Gener. Transm. Distrib., 2010, Vol. 4, Iss. 11, pp. 12611274 1263
doi: 10.1049/iet-gtd.2010.0143 & The Institution of Engineering and Technology 2010
www.ietdl.org
Substituting injected powers expressed in (8) and (9) in
P
L2
=

NB
i=1

NB
j=1
B
ij
(P
i
Q
j
+ Q
i
P
j
) (13)
and can be written as
P
L2
=

NGEN
k=1

NB
i=1

NB
j=1
B
ij
(x
ki
Q
j
+ Q
i
x
kj
)P
Gk
_ _
(14)
The active power loss because of purely reactive power
injection at each bus also may be taken as P
L3
, where
P
L3
=

NB
i=1

NB
j=1
[A
ij
Q
i
Q
j
] (15)
It is observed that the active power loss caused because of
interaction of reactive power injections is around 2% of
total active power loss. Hence, it is assumed that
reallocation of the active power loss because of interaction
of reactive power injections to kth generator is
P
Gk

NGEN
k=1
P
Gk
P
L3
(16)
Substitute the above equations (12), (14), (16) in (10) and
rearrange to decompose into self-terms and interactive terms.
The loss contribution component because of individual
generator kth alone is expressed as
P
(k,k)
L
=
P
Gk

NGEN
k=1
P
Gk
P
L3
+

NB
i=1

NB
j=1
A
ij
x
ki
x
mj
P
Gk
P
Gk
+

NB
i=1

NB
j=1
B
ij
(x
ki
Q
j
+ Q
i
x
kj
)P
Gk
(17)
P
L
(k,k)
is a part of total loss that completely depends on power
generation of kth generator.
The loss component because of interaction of generator k
and generator m is expressed as
P
(k,m)
L
=

NB
i=1

NB
j=1
A
ij
(x
ki
x
mj
+x
mi
x
kj
)P
Gk
P
Gm
, k =m (18)
P
L
(k,m)
is a part of total loss that arises from interaction
between kth generator and other generators of network.
It is common practice to allocate the above term as half of
the absolute value of P
L
(k,m)
to each generator of pair (k, m)
rather than the total amount to the individual generators.
P
(k)
L
= P
(k,k)
L
+

NGEN
m=1,m=k
1
2
P
(k,m)
L
(19)
and the same computation can be applied to the other terms
of the generators.
Total active power loss is
P
L
=

NGEN
k=1
P
k
L
(20)
2.2 Active power loss allocation to loads
Alternately, in similar lines to that of Section 2.1 above for
generators, the active power loss allocations to loads are
derived.
Let b
ij
be usage coefcient, that is the fraction of load
power at jth bus supplied by the generation at the ith bus.
The generation at bus i can be expressed as the sum of
usage amounts from different loads, that is
P
Gi
=

NB
j=1
b
ij
P
Loadj
, where i = 1, 2, . . . , NGEN (21)
Substituting the above equations (21) in (4), the injected
powers can be written as
P
i
=

NB
j=1
b
ij
P
Loadj
P
Loadi
, where i = 1, 2, . . . , NGEN
(22)
The above equations can be rewritten as, by introducing h
ij
P
i
=

NB
j=1
h
ij
P
Loadj
, where i = 1, 2, . . . , NGEN (23)
where h
ij
b
ij
where j 1, . . . ,NB and j =i or j distinct i
h
ij
= b
ij
1 for j = i
The injected active powers at ith and jth bus are written in the
form of (24) and (25) as shown below
P
i
=

NB
k=1
h
ik
P
Loadk
, where i = 1, 2, . . . , NGEN (24)
P
j
=

NB
m=1
h
jm
P
Loadm
, where j = 1, 2, . . . , NGEN (25)
1264 IET Gener. Transm. Distrib., 2010, Vol. 4, Iss. 11, pp. 12611274
& The Institution of Engineering and Technology 2010 doi: 10.1049/iet-gtd.2010.0143
www.ietdl.org
Equation (1) is split into
P
L
= P

L1
+ P

L2
+ P

L3
(26)
where the terms P

L1
, P

L3
can be treated as the loss
components because of purely active and reactive power
injections, respectively, and P

L2
represents the loss
components created by interaction between the active and
reactive power injections.
Rearrange and decompose into self-terms and interactive
terms in similar lines of procedure described for generators.
The active loss component because of individual load kth
alone is expressed as
P
(k,k)
L
=
P
Loadk

NGEN
k=1
P
Loadk
P
L3
+

NB
i=1

NB
j=1
A
ij
h
ki
h
mj
P
Loadk
P
Loadk
+

NB
i=1

NB
j=1
B
ij
(h
ki
Q
j
+ Q
i
h
kj
)P
Loadk
(27)
The loss component because of interaction of load k with
load m is expressed as
P
(k,m)
L
=

NB
i=1

NB
j=1
A
ij
(h
ki
h
mj
+h
mi
h
kj
)P
Loadk
P
Loadm
, k =m
(28)
The loss contribution to the kth load is given by reallocating
half of the amount of loss component because of interaction
of load k with load m
P
(k)
L
= P
(k,k)
L
+

NB
m=1,m=k
1
2
P
(k,m)
L
(29)
Total active power loss is
P
L
=

NB
k=1
P
k
L
(30)
2.3 Modication of slack bus elements
The injected power at slack bus is embedded with the total
loss of the system, which are not evident at the outset. The
injected power given in (6) and (23) at the slack bus does
not tally with the corresponding elements of a
ij
, b
ij
for
slack bus for this reason. Therefore there is a need to
modify these elements corresponding to the slack bus in
order to make injected power at the slack bus power tally
with (6) and (23).
Actually, the injected power at slack bus, s is
P
s
= P
Gs
+ P
L
P
Loads
(31)
and (20) can be written as P
L
=

NGEN
k=1
(P
k
L
/P
Gk
)P
Gk
.
Substituting the above equation in (31) and the modied
vectors related to slack bus are
P
s
= P
Gs
+

NGEN
k=1
P
k
L
P
Gk
P
Gk

NGEN
k=1
a
sk
P
Gk
(32)
for slack bus
x
sk
= a
sk
+
P
k
L
P
Gk
if slack bus is a non-generator bus
(33)
x
sk
= 1 a
sk
+
P
k
L
P
Gk
if slack bus is a generator bus (34)
In similar lines, elements of h for slack bus can be modied as
h
sk
= b
sk
+
P
k
L
P
Loadk
if slack bus is a non-generator bus
(35)
h
sk
= b
sk
1 +
P
k
L
P
Loadk
if slack bus is a generator bus
(36)
2.4 Presence of counter-ows
In a deregulated electrical power system network, every
transaction of electricity between a generator and a customer
results in some transmission loss. However, in some cases, a
generator while supplying its load may actually decrease the
transmission loss by opposing an initial ow. This ow is
commonly called as counter-ow. A generator can cause
decrease in transmission loss only when there exists an initial
ow in the opposite direction by another generator. The
counter-ow in the system will be dependent on relative
position of the generators and loads in the system.
As counter-ows cannot exist with a singe source in the
network and requires at least two sources in network, it is
obvious to divide the benet of counter-ow among the
generators. As mentioned in the above section, it is
common practice to allocate the cross terms as half of the
values rather than the total amount to the individual
generators.
Counter-ows indicate only the relative magnitudes of
ow contributions in a line by the generators in a system.
The concept of counter-ow stems from the relative
position of suppliers (generating utilities) and buyers
IET Gener. Transm. Distrib., 2010, Vol. 4, Iss. 11, pp. 12611274 1265
doi: 10.1049/iet-gtd.2010.0143 & The Institution of Engineering and Technology 2010
www.ietdl.org
(loads) with respect to each other. This relative position
makes a difference in the overall transmission loss allocation.
Thus a user who causes more network loss must be charged
more whereas a user who helps to reduce the loss, because of
counter-ow, must be rewarded.
3 Computational procedure for
allocation of loss
The computational procedure of the proposed concept is
briey described here. First, an IGO acquires all the
required contractual information and quantity of power
agreement from the participants. The IGO develops
preferred values of x
ij
, h
ij
considering the utilisation of
power by all participants. Then, the IGO performs
feasibility studies using power ow tools. In case of no
congestion, the preferred trading with generators is feasible
and the IGO has to evaluate the loss caused by each
generator or load because of its utilisation and allocate the
loss to all the participants. In presence of congestion,
however, the preferred trading with generators is infeasible
and the IGO has to nd other solutions, for example
generation rescheduling to relieve the congestion. Since loss
management for the congestion case, which may occur for
a short period compared to normal operation period, is not
in the scope of this paper, only the loss allocation method
for the feasible trading is reported. The proposed method
can be utilised to allocate transmission loss among the
generators and loads in a deregulated network. The
algorithm presented for loss allocation for the generators is
described in detail below. Alternately, in similar lines an
algorithm for loss allocation for loads is also developed.
In the rst step of iteration, the loss has been assumed to
be zero. This makes the required generation equal to the load
at the beginning.
Then the program computes loss allocation to the
generators by using (19) and total loss by (20) and updates
the usage coefcients of slack bus by using (33) and (34)
with this computed loss.
In second iteration, the programcomputes the loss allocation
of generators with updated usage coefcients, total loss and
compares the total loss with the loss in rst iteration. If
difference is not within the specied tolerance, the usage
coefcients of generators are updated again and the program
proceeds for next iteration. The process is repeated until
specied tolerance is obtained. The detailed algorithm for
computing loss allocation to generators is given below.
Algorithm overview
1. Read power system data and bus voltages, complex line
ows and slack bus power generation and total transmission
loss of converged load ow.
2. Calculate the injected powers from solution of the power
ow, by using (4).
3. Form the matrices A, B as per (2).
4. Read the amount of specied contracted generations for
the loads, that is a
ij
, b
ij
, respectively.
5. Find the loss because of own interaction of reactive
powers, P
L3
, according to (15). Allocate the P
L3
to the
respective generators according to the amount of generation
by (16).
6. Set the iteration count i 1.
7. Set dp loss 0, [P
L
]
(i)
0.0 and set d 10
28
.
8. Calculate P
L
(k,k)
, P
L
(k,m)
where k and m 1, 2, . . . , NGEN
as per (17) and (18), respectively.
9. Calculate loss shared by all generators by P
L
k
where k 1,
2, . . . , NGEN as per (19).
10. Calculate the total loss P
L
, as per (20).
11. Set [P
L
]
(i+1)
P
L
.
12. d
p loss
[P
L
]
i+1
2[P
L
]
i
.
13. Update the contributions of loss for P
L
k
where k slack
bus according to (33) and (34).
14. Check d
ploss
d.
15. If No increment i i + 1 and go to 8.
16. Print allocation of loss to the generators.
17. Stop.
Similarly, the same algorithm is used for allocation of loss
to loads by replacing NGEN with NB and applying the
relevant equations.
4 Loss compensation scheme for
active power
The proposed approach for loss compensation to be utilised
by IGO is presented in this section. The section presents
an approach that every trading arrangement is required to
have, with each individual generator/load for taking care of
its own loss or required to compensate the loss in open
access. To perform this task, a loss compensation scheme is
developed in this section. The advantage of this approach is
that once an agreement in the open access is so dened,
results derived in the previous sections become valid. The
loss allocation formulae dened in previous section are used
1266 IET Gener. Transm. Distrib., 2010, Vol. 4, Iss. 11, pp. 12611274
& The Institution of Engineering and Technology 2010 doi: 10.1049/iet-gtd.2010.0143
www.ietdl.org
to determine the allocation of loss for each generator/load.
Having identied the allocation of loss, the approach
calculates the equivalent loss compensation required to
generate at the slack bus because of the specied injection
of real power at other buses. The IGO then suggests to
compensate specied amount of loss by either increasing
the generation or reducing the load. The methodology is
presented in the following.
The power balance equation including loss before injection
can be shown as
P
s
+

NB
i=1,i=s
P
i
= P
L
(37)
where P
s
is the active power injection at slack bus, P
i
is the
active power injection at ith bus and P
L
is the active power
loss of system.
The total transmission loss is expressed as a function of
initial power loss P
0
L
and the change in total transmission
loss as DP
L
because of small injection of active power at
buses, that is
P
L
= P
0
L
+DP
L
(38)
The equivalent loss compensation for active power with the
injection of real power, DP
i
at ith bus generally results in
decreasing the injection at the slack bus with all contracts
on the system remaining unchanged, that is with the
additional amount of power DP
i
injected at bus i into the
system to off load the loss compensation at the slack bus,
the power balance equation becomes
P
s
DP
s
+

NB
i=1,i=s
(P
i
+DP
i
) = P
L
+DP
L
(39)
DP
s
+

NB
i=1,i=s
DP
i
= DP
L
(40)
The exact transmission loss using bus injected powers and
system parameters are given in (1). Now the active power
loss of system can be expressed as Taylor series about initial
active power loss P
L
0
, before injection, where the terms in
Taylor series expansion are the injected power DP
i
at each
bus with only the linear terms retained
P
L
= P
0
L
+

NB
i=1,i=s
P
L
P
i
DP
i
(41)
The term (P
L
/P
i
) represents the loss sensitivity factor,
which is dened as an incremental change of real power
loss by an incremental change of scheduled power injected
into bus. This loss sensitivity factor can be obtained by
differentiating (1) w.r.t. P
i
, that is
P
L
P
i
= 2A
ii
P
i
+

NB
j=1,j=i
[(A
ij
+ A
ji
)P
j
+ (B
ij
B
ji
)Q
j
] (42)
By rearranging (4446)
P
L
P
0
L
=

NB
i=1,i=s
P
L
P
i
DP
i
= DP
L
= DP
s
+

NB
i=1,i=s
DP
i
(43)
DP
s
=

NB
i=1,i=s
1
P
L
P
i
_ _
DP
i
(44)
Let
g
i
= 1
P
L
P
i
_ _
(45)
g
i
is called as the loss compensation index at bus i. The
physical interpretation of g
i
is as follows. g
i
MW injected
at ith bus is equivalent to reducing loss compensation by
1 MW at the slack bus
DP
s
=

NB
i=1,i=s
g
i
DP
i
(46)
The equivalent loss compensation at ith bus is given by
DP
i
(DP
s
/g
i
).
The proposed approach is classied into three categories:
1. Total system loss is compensated by slack bus as explained
in the computational procedure above (loss allocation
methodology).
2. Loss compensation scheme
(a) Each generator uses its own generation to compensate for
its allocated loss i.e. the generators are self-compensating
with their specied compensation fractions (self-loss
compensating scheme).
(b) All consumers buy IGO compensation service with least
system loss optimisation or least loss cost optimisation.
3. Combination of both (a) and (b) mentioned in (2) above.
4.1 Self-loss compensating scheme
Slight increase in the generation at generator bus to share loss
from the slack bus results in a small change of loss. The small
change of loss in the system assuming no changes in loads
after increasing DPGi of each generator bus to compensate
IET Gener. Transm. Distrib., 2010, Vol. 4, Iss. 11, pp. 12611274 1267
doi: 10.1049/iet-gtd.2010.0143 & The Institution of Engineering and Technology 2010
www.ietdl.org
the loss, becomes
DP
s
=

NGEN
Gi=1,Gi=s
g
Gi
DP
Gi
(47)
where g
Gi
= (1 (P
L
/P
Gi
)) and
P
L
P
Gi
= 2A
ii
P
Gi
+

NGEN
j=1,j=i
[(A
ij
+ A
ji
)P
Gj
+ (B
ij
B
ji
)Q
Gj
]
(48)
4.2 IGO-acquired loss compensation
The IGO corrects the generation dispatch for loss
compensation at designated buses using the proposed
methodology. The corrective generation dispatch with loss
compensation must also take into consideration the
transmission line overload and voltage violations. The OPF
program has been executed by changing the generator
limits corresponding to the loss compensation along with
other constraints such as voltage, overloads of transmission
line etc. Then OPF model for the optimal loss allocation
minimises the cost or price or total loss, while satisfying all
the necessary network limits
min

NGEN
i
c
i
DP
i
(49)
subject to the constraints
S
ij
S
max
ij
DP
i
P
max
i
P
0
i
DP
i
P
0
i
P
min
i
V
min
i
V
i
V
max
i
(50)
4.3 Combination method
The combination method involves the use of both the
algorithms described above.
5 Case studies
5.1 Loss allocation methodology
To demonstrate the effectiveness of the proposed
transmission loss allocation methodology for bilateral
energy transactions in open access, the proposed
methodology is applied to three cases such as a sample
5-bus system, IEEE-118-bus system and Indian practical
power system.
The sample system with ve nodes, seven branches and
three generators is shown in Fig. 1. The total trading of
system is 330 MW and three generators have bilateral
trading agreements with individual loads. The
corresponding usage coefcients are shown in Table 1. The
total trading operates with these three suppliers G1, G2
and G3 to supply the ve loads according to the contracted
amount of MW using the whole network, that is there
exists ve simultaneous trades corresponding to three
generators and ve loads.
Test calculations have been performed on sample 5-bus
system and the results are tabulated in Table 1. A
computer program is developed on Matlab to implement
the proposed approach. Table 1 shows the allocation of loss
to generators/loads and the corresponding values of loss
compensation indices g
i
, obtained through the application
of proposed approach. The total network loss
(7.5135 MW) is successfully allocated to all users and the
results are presented in Table 1. The generation at bus-1
(slack bus) supplies the system loss of 7.5135 MW. Using
the algorithm presented in the previous sections, the
process of contribution of loss from generators converges
within ve iterations. It is observed that active power loss
because of purely reactive power injection is
P
L3
0.0137 MW. It is around 0.0182% of the total loss.
It is observed that the developed methodology gives rise to
loss allocations to both generators/loads, which are always
positive in absence of counter-ows.
5.1.1 Impact of the proposed usage coefcients
in loss allocation: The impact of proposed usage
coefcients, alpha and beta used in formulae is studied
from the point of view of the fairness to all users and from
the economic signal to the market. This is because, for the
same scenario of generation, demand, network and
corresponding loss, the loss allocation between users may
be very different, depending on usage between generators
and loads.
Five different cases are carried out in order to show the
impact of alpha. The same power ow solution is utilised
for all cases and hence the total system loss is the same for
all ve cases.
Case 1: From Table 2, one observes that generator 1 trades an
amount of 90 MW to supply the customer loads at buses 3
and 4, whereas generator 2 trades an amount of 120 MW
to supply the customer load at bus 5 and the generator 3
supplies to buses 2 and 3.
Figure 1 Sample 5-buses test system
1268 IET Gener. Transm. Distrib., 2010, Vol. 4, Iss. 11, pp. 12611274
& The Institution of Engineering and Technology 2010 doi: 10.1049/iet-gtd.2010.0143
www.ietdl.org
Under these conditions the loss allocation can be obtained
as follows:
P
(1,1)
L
= 2.98 MW, P
(2,2)
L
= 5.30 MW, P
(3,3)
L
= 0.35 MW
P
(1,2)
L
= 0.14 MW, P
(1,3)
L
= 0.26 MW, P
(2,3)
L
= 1.24 MW
P
1
L
= 3.05 MW, P
2
L
= 4.61 MW, P
3
L
= 0.14 MW
It is observed from the above results that the cross term
between G1 and G2 is negative. This means that loss
allocated to G1 and G2 decreases when the trading of both
generators G1 and G2 coexist.
The negative sign further signies that generator 1 causes a
ow in the same direction as the net whereas generator 2
causes a ow in the opposite direction. The same analogy
will be applicable to the cross terms between G2 and G3. It
is also observed that the calculated loss contribution from
G3 is negative. This means that the trading amount of G3
causes only counter-ow, which helps generally to reduce
the system loss. Hence, generator 3 will receive a negative
loss allocation, which has to be rewarded.
It is observed that depending on usage coefcients of
suppliers (generators) and consumers (loads), the allocation
of transmission loss gets changed abruptly. Since counter-
ow would increase the allocated loss of some generators
and decrease the allocated loss of other generators, the
choice of usage coefcients becomes a subject of
importance. This fact will be illustrated under ve different
cases assuming the following conditions:
1. System state condition remains unchanged.
2. The delivered amounts for utilisation from the generator 3
are varied.
3. The corresponding changes are made in generator 1.
4. Keeping all bilateral contracts and usage from generator 2
xed,
5. Case 1 is taken as base.
Case 2: As an illustration, the amount of trading by generator
1 with load 3 is increased and load 4 is decreased in steps of
10 MW (to maintain generation at bus constant). This
means that the relative positions of load 3 and load 4 are
changed w.r.t. generator 1.
Simultaneously the amount of trading by G3 with load 3 is
increased by 10 MW (to maintain load at bus 3 constant) and
corresponding changes made for the amount of trading by
G3. The load 4 is entered into utilisation of power from
G3 newly (to maintain load at bus 4 constant). It is
understood that the relative distance for load 4 is some
what decreased with respect to G1. The revised usage
coefcients of generators and loads are shown in Table 2.
The loss allocation obtained is as follows
P
(1,1)
L
= 2.75 MW, P
(2,2)
L
= 5.30 MW, P
(3,3)
L
= 0.46 MW
P
(1,2)
L
= 0.03 MW, P
(1,3)
L
= 0.37 MW,
P
(2,3)
L
= 1.34 MW
P
1
L
= 2.92 MW, P
2
L
= 4.61 MW, P
3
L
= 0.02 MW
The loss allocated to the generator 1 has slightly decreased
because of decrease of relative distances of its contracts.
Case 3: With further increase of amount of trading by load 3
and decrease of load 4 with generator 1 and corresponding
changes made in the amount of trading by generator 3 for
load 3 and load 4, the relative distances of loads 3 and load
4 are reduced. The revised usage coefcients of generators
and loads are shown in Table 2.
Table 1 Usage coefcients of sample 5-bus test system and loss allocations to the
generators and loads
Node no. Percentage MVA
contracted agreement of
consumers with the
generators
Loss allocated
to loads
Loss allocated
to generator
Gamma value, g
i
G1, % G2, % G3, %
1 0 0 0 0 2.0717 1
2 0 33.3333 0 0.1676 2.7662 0.989233
3 100 0 0 2.0717 2.6756 0.982985
4 0 66.667 0 2.5986 1.018831
5 100 2.6756 1.089155
total 100 100 100 7.5135 7.5135
IET Gener. Transm. Distrib., 2010, Vol. 4, Iss. 11, pp. 12611274 1269
doi: 10.1049/iet-gtd.2010.0143 & The Institution of Engineering and Technology 2010
www.ietdl.org
Loss allocation obtained is as follows
P
(1,1)
L
= 2.56 MW, P
(2,2)
L
= 5.30 MW, P
(3,3)
L
= 0.62 MW
P
(1,2)
L
= 0.08 MW, P
(1,3)
L
= 0.41 MW,
P
(2,3)
L
= 1.45 MW
P
1
L
= 2.80 MW, P
2
L
= 4.61 MW, P
3
L
= 0.1 MW
This in turn results in decrease of loss allocation of
generator 1. The counter-ow because of coexistence
of G1 and G3 coincides with dominant ow (note
the value of P
L
3
) and results in increase of loss allocation
to G3.
Case 4: The inuence in loss allocation because of
further change of amount of trading by G1 and G3 is shown
in below:
P
(1,1)
L
= 2.40MW, P
(2,2)
L
= 5.30MW, P
(3,3)
L
= 0.81MW
P
(1,2)
L
= 0.19MW, P
(1,3)
L
= 0.38MW, P
(2,3)
L
= 1.56MW
P
1
L
= 2.6805MW, P
2
L
= 4.6067MW, P
3
L
= 0.2233MW
Note the revised value of P
L
3
.
Case 5: The inuence in loss allocation because of further
change is given below
P
(1,1)
L
= 2.27MW, P
(2,2)
L
= 5.30MW, P
(3,3)
L
= 1.04MW
P
(1,2)
L
= 0.29MW, P
(1,3)
L
= 0.28MW, P
(2,3)
L
= 1.67MW
P
1
L
= 2.5587MW, P
2
L
= 4.6097MW, P
3
L
= 0.3451MW
The loss allocation to G3 became positive which is a
signicant change.
Table 3 shows consolidated loss allocations to generators
for all the ve cases. It is observed that loss allocation to
generator 1 is decreased from base case to case 5 whereas
for generator 3 it is increased. This will be true until a
turning point is reached. Such a result is due to the fact
Table 2 Usage coefcients data for Cases 15 and allocation results
Case no. Generator Consumer 2 Consumer 3 Consumer 4 Consumer 5 Allocation Total loss, MW
Case 1 No MW (a
ij
) MW (a
ij
) MW (a
ij
) MW (a
ij
)
G1 0 10(0.11111) 80(0.88889) 0 3.0435 7.5135
G2 0 0 0 120 (1) 4.61
G3 40(0.333) 80(0.666) 0 0 20.14
Case 2 G1 0 20(0.2222) 70(0.7778) 0 2.9235 7.5135
G2 0 0 0 120 (1) 4.61
G3 40(0.333) 70(0.5833) 10(0.08333) 0 20.02
Case 3 G1 0 30(0.3333) 60(0.667) 0 2.8 7.5135
G2 0 0 0 120 (1) 4.61
G3 40(0.333) 60(0.5) 20(0.16667) 0 0.125
Case 4 G1 0 40(0.44444) 50(0.55556) 0 2.6805 7.5135
G2 0 0 0 120 (1) 4.61
G3 40(0.33333) 50(0.41667) 30(0.25) 0 0.223
Case 5 G1 0 50(0.55556) 40(0.44444) 0 2.5587 7.5135
G2 0 0 0 120 (1) 4.6097
G3 40(0.33333) 40(0.33333) 40(0.33333) 0 0.3451
Table 3 Comparison of loss allocations to generators for ve
cases
Generator Case 1 Case 2 Case 3 Case 4 Case 5
1 3.0435 2.9235 2.8 2.6805 2.5587
2 4.61 4.61 4.61 4.61 4.61
3 20.14 20.02 0.125 0.223 0.3448
7.5135 7.5135 7.535 7.5135 7.5135
1270 IET Gener. Transm. Distrib., 2010, Vol. 4, Iss. 11, pp. 12611274
& The Institution of Engineering and Technology 2010 doi: 10.1049/iet-gtd.2010.0143
www.ietdl.org
that the transactions of G3 initially are on the counter-ow
direction helping to reduce the loss allocation to generator
3. At some point, the transactions of G3 will become
aligned with the dominant ow causing the loss allocations
to increase.
5.1.2 Case study on IEEE-118 bus system: Another
case study is also carried out on a larger system by considering
IEEE-118 bus system with 118 nodes, 186 branches and 19
generators for testing the proposed approach. The total
trading of system is 4242 MW. In order to simplify the
contents of IEEE-118 bus system, the bilateral trading
agreements with individual load nodes and corresponding
generator nodes are only shown in Table 4. It becomes
voluminous to mention Table 4 in the form of Table 1.
For example, take second row of Table 4. The generation
at bus 10 is given as 450 MW and the load at bus 1 as
51 MW in IEEE-118 bus system. It is assumed that the
load bus 1 has a bilateral trading agreement with generator
at bus 10. The coefcients of a, b are computed as follows:
a
1,10
51/450 0.1133 and b10, 1 1.0, that is 100%
of its load is met from generator 10. In similar lines, the
coefcients of a, b are computed for remaining generator
and load nodes, respectively.
The results are tabulated in Tables 4 and 5. Tables 4 and 5
show the loss allocation to generators and loads, respectively,
corresponding values of loss compensation indices g
i
,
obtained through the application of proposed approach for
IEEE-118 bus system. The total trading operates with
these 19 generators to supply 99 consumers according to
the contracted amount of MW using the whole network. It
is assumed that the generation at bus-69 (slack-bus)
supplies the system loss of 132.863 MW.
It is observed from the above case study that there is no
counter-ow in the system.
5.2 Loss compensation
The proposed categories of loss compensation allocation
methodologies presented in the previous section are tested
on IEEE-118 bus system in this paper. They are mainly
(a) self-loss compensating scheme, (b) IGO-acquired loss
compensation scheme and (c) both (a) and (b) together.
The results are tabulated subsequently.
The compensating capacities and the prices of the players
that bid to provide loss compensation services are shown in
Table 6.The test results show that the self-acquisition and
Table 4 Contracted load data from generators of IEEE-118 bus test system and loss allocation to generators
S. no. Bus no. Participation of buying nodes in the transaction with the generator Allocated loss g
i
1 10 1 2 3 4 6 7 8 11 12 13 14 41 0.2533 0.922632
2 12 16 18 0.02513 1.00146
3 25 15 17 19 20 21 23 43 50 0.01428 0.99268
4 26 24 27 28 29 31 32 33 35 36 0.08406 0.961766
5 31 102 108 0.0002 0.99842
6 46 84 109 0.00074 0.99985
7 49 34 39 104 106 118 0.02426 1.003071
8 54 46 48 0.00039 1.00717
9 59 42 45 73 0.00288 1.01835
10 61 44 47 49 53 0.00851 0.989152
11 65 59 60 72 85 0.03696 1.00065
12 66 51 52 54 55 56 62 83 0.05708 0.99164
13 69 66 67 70 74 80 107 0.09504 0.938287
14 80 75 76 77 78 79 90 114 117 0.19346 0.94378
15 89 22 40 57 58 82 88 91 92 93 94 95
96 99 100 101 103 105 110 115 0.48657 0.91149
16 100 112 116 0.08289 0.971498
17 103 98 113 0.00252 0.99883
18 111 86 97 0.00155 0.99708
IET Gener. Transm. Distrib., 2010, Vol. 4, Iss. 11, pp. 12611274 1271
doi: 10.1049/iet-gtd.2010.0143 & The Institution of Engineering and Technology 2010
www.ietdl.org
IGO provided compensation procedures developed in this
paper are effective and provide good exibility in arranging
for loss compensation service in usage-based schemes of
open access.
5.2.1 Self compensation by generators: The loss
contributions of generators and loss compensation indices
are calculated by using (47) and (48) for IEEE-118 bus
system is shown in Table 6. The self-acquisition option is
exercised in a straightforward manner by a generator as
shown in Table 6. The equivalent generation required at
generator bus will be calculated by using the loss
compensation indices to compensate its share of loss. The
loss compensation capacities along with their costs
determined by this procedure for self-loss compensation are
summarised in Table 6. The cost incurred because of the
self-loss compensation scheme is 2556.8 $/h.
5.2.2 Least-price loss compensation methodology:
The generators make use of least-price loss compensation
service acquired by IGO. The IGO-acquired least-price
loss compensation scheme for generators is the solution of
the modied OPF problem formulated with the specied
data. Since at the optimal solution, the IGO calculates loss
compensation for each of the generators, the total costs will
be distributed accordingly to generator. The results for
least-price loss compensation are shown in Table 6. The
least cost price for the loss compensation is 3897.5 $/h.
The major portion of loss compensation is taken place from
89th generator bus, as it is a nearby generator for the major
loads. It is observed that the schemes depend on the
location of generators and loads, which participated in
least-price compensation of the loss. It is observed from
Table 6, even if the compensation capacities available at
25th, 49th and 61st generator buses, IGO could not utilise
their compensation capacities because of the process of
optimal prices for loss compensation. This is mainly
because of consideration of the limits imposed on the
loading of transmission lines interconnected to the said
generators. Thus, the limits on the transmission loadings
have an impact on evaluation of the least-price loss
compensation.
5.2.3 Self compensation and least-price loss
compensation methodologies: In this case generators
10, 12, 31, 46, 89, 100, 103 and 111 undertake self-
acquisition for loss compensation and the remaining
generators make use of least-price loss compensation service
acquired by the IGO, shown in Table 6. Since at the optimal
solution, the IGO calculates loss compensation for each of
the generator, the total costs will be distributed accordingly
to generators. The generators are able to obtain their loss
compensation at the designated slack bus, bus 69; the cost is
3897.5/MWh. It is observed that even if there is provision
Table 5 Loss allocation to the loads for IEEE-118 bus system
Bus
no.
Allocated
loss
g
i
Bus
no.
Allocated
loss
g
i
Bus
no.
Allocated
loss
g
i
Bus
no.
Allocated
loss
g
i
Bus
no
Allocated
loss
g
i
1 0.02829 0.96857 2 0.01108 0.968724 3 0.0216 0.97546 4 0.0219 1.0021 6 0.02919 1.05584
7 0.01162 1.02974 8 0.01679 1.042086 11 0.03915 1.03397 12 0.02616 1.0015 13 0.01885 0.97994
14 0.00787 0.97958 41 0.0208 1.084766 16 0.00739 0.99671 18 0.01774 0.9961 15 0.00569 0.9859
17 0.00078 1.04556 19 0.00286 0.984473 20 0.00117 1.00325 21 0.00098 1.0405 23 0.00048 1.04367
43 0.00117 0.9826 50 0.00115 1.006053 24 0.00352 1.06583 27 0.01901 1.0222 28 0.00459 1.0348
29 0.00645 1.00481 31 0.01153 0.998416 32 0.01589 1.04435 33 0.0061 0.9771 35 0.00875 0.9698
36 0.00822 0.97027 102 0.00014 0.995756 108 0.00006 1.00877 84 0.00046 1.0713 109 0.00028 1.00438
34 0.00707 0.97235 39 0.00335 1.053988 104 0.0046 0.992 106 0.00528 1.0141 118 0.00396 1.05618
46 0.00023 0.99985 48 0.00016 1.010088 42 0.00176 1.05438 45 0.00101 1.0165 73 0.00011 1.03781
44 0.00088 1.02022 47 0.00181 1.017666 49 0.00458 1.00307 53 0.00124 1.0215 59 0.02622 1.01835
60 0.00734 0.98781 72 0.00117 1.082888 85 0.00223 0.98216 51 0.00243 1.0078 52 0.00258 1.02727
54 0.01722 1.00717 55 0.00904 1.016109 56 0.01205 1.01312 62 0.01101 0.9928 83 0.00275 0.95551
66 0.00945 0.99164 67 0.00685 1.018839 70 0.01582 0.98914 74 0.01853 1.0858 80 0.03216 0.94378
107 0.01223 1.0262 75 0.01873 0.955515 76 0.02722 1.06518 77 0.02571 1.0734 78 0.02874 1.08255
79 0.01573 1.07308 90 0.06615 0.960219 114 0.00323 1.05716 117 0.00795 0.9631 22 0.00785 0.97323
40 0.05314 1.09777 57 0.00945 1.002757 58 0.00945 1.02116 82 0.04247 1.0208 88 0.03793 1.0452
91 0.0078 0.95991 92 0.05174 0.942535 93 0.00951 1.04043 94 0.0235 0.9679 95 0.03389 0.99497
96 0.02979 1.00789 99 0.03302 1.01685 100 0.02902 0.9715 101 0.01736 1.0167 103 0.0181 0.99883
105 0.02439 1.00795 110 0.0307 0.993747 115 0.01746 1.05776 112 0.02235 1.0509 116 0.06054 1.04006
98 0.00211 1.01363 113 0.00041 1.047313 86 0.00101 1.1013 97 0.00054 0.9918
1272 IET Gener. Transm. Distrib., 2010, Vol. 4, Iss. 11, pp. 12611274
& The Institution of Engineering and Technology 2010 doi: 10.1049/iet-gtd.2010.0143
www.ietdl.org
for compensation of loss by generators 49 and 61, they could
not be participated because of the optimal solution.
6 Conclusions
The paper presents a new method for loss allocations to
consumers/Discoms in power systems under open access
markets environment. The proposed approach has been
demonstrated on a sample 5-bus and IEEE-118 bus
systems. The approach proposed here enables the
calculation of actual active power loss contributions from
generators and allocation of loss to consumers. This useful
tool is expected to be effectively suitable for planning the
loss contributions from generators under open access
environment. Once market has dened a set of proposed
contractual agreements among generators and loads, the
approach systematically calculates the loss components
allocated to each individual generator and loads by using
exact loss equation. The loss compensation techniques
presented in the paper are very effective and expected to
be highly useful in open access environment. The loss
compensation by loads is under future work. The proposed
approach for loss allocation is tested for large practical
power system that is an Indian state power system and
similar results can be expected for large practical systems.
Then proposed methodology is tested on Indian state
system, which comprises of 481 buses with 29 generating
plants, 399 load buses and 854 branches. The total
generation and total load of the system are 7576.938 and
7237.239 MW, respectively, and total loss obtained
through load ow studies is 342.818 MW. The usage
coefcients are calculated according to their utilisations
based on their bilateral contracts.
7 Acknowledgments
The authors acknowledge the support and encouragement
extended to them by their respective organisations.
However, the views expressed in this paper are of the
authors in their individual capacity and not necessarily that
of APTRANSCO of Andhra Pradesh.
Table 6 Costs of compensation data in the IEEE-118 bus system and results for compensation schemes
Bus compensation data A. Self-compensation B. Least
compensation
C. Self and least compensation
Compensation
node
Ck,
$/MWH
Max Dpk,
MW
Dpk,
MW
Total cost,
$/h
Dpk,
MW
Total cost,
$/h
Mode Dpk,
MW
Total cost,
$/h
10 12.5 30 27.454 343.18 30 375 self 27.45 343.125
12 15 5 2.509 37.64 5 75 self 2.5093 37.6395
25 30 30 1.44 43.2 0 0 self 1.439 43.17
26 14 10 8.7402 122.36 10 140 self 8.74 122.36
31 15 2 0.02 0.3 2 30 self 0 0
46 19 5 0.0742 1.41 5 95 self 0.074 1.406
49 27 20 2.4186 65.3 0 0 IGO 0 0
54 13 20 0.04 0.52 20 260 IGO 20 260
59 12 10 0.2828 3.39 10 120 IGO 10 120
61 30 50 0.86 25.8 0 0 IGO 0 0
65 20 10 3.7 74 10 200 IGO 10 200
66 22.5 35 5.8 130.5 35 787.5 IGO 35 787.5
69 15 15 9.5 142.5 15 225 IGO 25 375
80 20 10 20.5 410 10 200 IGO 10 200
89 20 60 53.382 1067.64 60 1200 IGO 50 1000
100 10 10 8.5 85 10 100 self 8.53 85.3
103 10 4 0.3 3 4 40 self 0.252 2.52
111 15 4 0.155 2.33 4 60 self 0.1555 2.3325
2558.06 3907.5 3580.353
IET Gener. Transm. Distrib., 2010, Vol. 4, Iss. 11, pp. 12611274 1273
doi: 10.1049/iet-gtd.2010.0143 & The Institution of Engineering and Technology 2010
www.ietdl.org
8 References
[1] BIALEK J.: Tracing the ow of electricity, IEE Proc.
Gener. Transm. Distrib., 1996, 143, (4), pp. 313320
[2] DING Q., ABUR A.: Transmission loss allocation in a
multiple-transaction framework, IEEE Trans. Power Syst.,
2004, 19, (1), pp. 5220
[3] KIRSCHEN D., ALLAN R., STRBAC G.: Contributions of
individual generators to loads and ows, IEEE Trans.
Power Syst., 1997, 12, (1), pp. 5260
[4] KIRSCHEN D., STRBAC G.: Tracing active and reactive power
between generators and loads using real and imaginary
currents, IEEE Trans. Power Syst., 1999, 14, (4),
pp. 13121318
[5] GUBINA F., GRGIC D., BANIC I.: A method for determining
the generators share in a consumer load, IEEE Trans.
Power Syst., 2000, 15, (4), pp. 13761381
[6] WU F.F., NI Y., WEI P.: Power transfer allocation for open
access using graph theory-fundamentals and applications
in systems without loop ow, IEEE Trans. Power Syst.,
2000, 15, (3), pp. 923929
[7] BIALEK J.: Allocation of transmission supplementary
charge to real and reactive loads, IEEE Trans. Power Syst.,
1998, 13, (3), pp. 749754
[8] WU Z.Q., CHEN G.Z.: MVA power ow and loss analysis for
electricity market, IEE Proc. Gener. Transm. Distrib., 2001,
148, (2), pp. 153158
[9] PENG J.C., JIANG H.: Contributions of individual
generators to complex power losses and ows part 1:
fundamental theory, IEE Proc. Gener. Transm. Distrib.,
2002, 149, (2), pp. 182185
[10] PENG J.C., JIANG H.: Contributions of individual
generators to complex power losses and ows part 2:
algorithm and simulations, IEE Proc. Gener. Transm.
Distrib., 2002, 149, (2), pp. 186190
[11] RETA R., VARGAS A.: Electricity tracing and loss allocation
methods based on electricity concepts, IEE Proc. Gener.
Transm. Distrib., 2001, 148, (6), pp. 518522
[12] ZHAOXIA J., FUSHUAN W.: Review of cost based
transmission losses allocation methods, IEEE Trans. PSCE,
2006, 6, pp. 15871591
[13] SINGH H., PARALEXOPOULOS A.: Competitive procurement of
ancillary services by an independent system operator, IEEE
Trans. Power Syst., 1999, 14, (2), pp. 498504
[14] BERIZZI A., BOVO C., MARANNINO P.: Allocation of
transmission losses in presence of both bilateral and pool
market models. Second IEEE Power Engineering Society
Int. Conf. on Power Industry Computer Applications
Innovative Computing for Power Electric Energy Meets
the Market., pp. 336341
[15] TAO S., GROSS G.: Transmission loss compensation in
multiple transaction networks. IEEE Transduction on
Power Systems Conf., Trondheim, 28 June2 July 1999,
vol. 15, issue (3), pp. 238243
[16] HUANG G., ZHANG H.: Transaction based power ow
analysis for transmission utilization allocation. Proc. IEEE
Power Engineering Society Summer Meeting, 2001, vol. 2
[17] MUTALE E.J., STRBAC G., CURCIC S., JENKINS N.: Allocation of
losses in distribution systems v, ith embedded
generation, IEE Proc. Gener. Transm. Distrib., 2000, 147,
(1), pp. 714
[18] GALIANA F.D., CONEJO A.J., KOCKAR I.: Incremental
transmission loss allocation under pool dispatch, IEEE
Trans. Power Syst., 2002, 17, (1), p. 2633
[19] KOTARI D.P., DHILLON J.S.: Power system optimization
(Prentice Hall of India Private Limited, New Delhi, 2004)
1274 IET Gener. Transm. Distrib., 2010, Vol. 4, Iss. 11, pp. 12611274
& The Institution of Engineering and Technology 2010 doi: 10.1049/iet-gtd.2010.0143
www.ietdl.org

Potrebbero piacerti anche