JOURNAL OF SUSTAINABLE ENERGY, VOL. 1, NO.

4, DECEMBER, 2010

POWER FLOW TRACING METHOD FOR ELECTRICITY

TRANSMISSION AND WHEELING PRICING
OANA POP*, STEFAN KILYENI**, PETRU ANDEA**,
CONSTANTIN BARBULESCU**, CRISTIAN CRACIUN**
*Academy of Romanian Scientists, Timisoara Branch
Bd. V. Parvan, Nr. 2, 300223, Timisoara; Tel: 0256-403406; E-mail: [email protected]
**Politehnica University of Timisoara, Electrical and Power Engineering Faculty,
Power Systems Department,
Bd. V. Parvan, Nr. 2, 300223, Timisoara; Tel: 0256-403416;
E-mail: [email protected]; [email protected]

Abstract: The power transmission costs, which are

charged to the market participants, are a central
issue of the new deregulated electricity markets. The
paper presents a methodology for transmission cost
allocation based on the principle of proportional
sharing states that any power flow leaving a bus is
made up of the flows entering that bus in a
proportional manner, thus satisfying Kirchhoff first
law. A properly established cost provides economical
signals for short-term and long-term recovery current
expenses and for a fair cost allocation to market
participants. An original method of transmission
cost allocation considering the power losses is
proposed. The appropriate software tool has been
developed in Mathematica environment. The results
are presented and analyzed for the Test 25 test
power system.

Keywords: transmission system, tracing method,

cost allocation, proportional sharing principle,
power flow
1. INTRODUCTION
Electricity transmission and wheeling service
pricing becomes a more complex and more important task
within the ongoing deregulation of electric power
industry. It has significant impacts on the market
efficiency, the development of transmission systems,
the sitting of power plant, the demand growth and its
geographical distribution. Transmission pricing
discusses how to allocate the entire cost of a transmission
system among all the system users [1], [2], [3]. Because
monetary flows are related to the nodal prices, the impact of
generators and loads on operation constraints and the
interactive correlation between active and reactive
power can be considered.
Total transmission service cost is separated into
more practical line-related costs and system wide cost,
and can be flexibly distributed between generators and loads.
It is necessary to find out the contributions of each
generator and load to transmission system power flows

before a reasonable distribution off line-related costs

and system-wide cost can be obtained [4], [5]. This
paper uses the Kirschen tracing method to calculate the
contributions of generators and loads to power flows
and active losses [6], [7]. This method is based on the
proportional sharing principle. The novelty of the
proposed approach refers to AC power flow based
version.
The buses and the network elements are organized
in homogeneous groups based on the following concepts:
the domain of the generator, the commons and the links
[8]. The domain of a generator represents a subset of buses,
which are supplied by certain generator. The commons of
a generator are defined as a subset of neighboring
buses supplied by the same generators. Having the
buses divided into commons, each network element can be
either internal to a common (for example, it connects two
buses which are part of the same com-mon) or external
(for example, it connects two buses belongs to different
commons). A link is represented by one or more external
branches connecting the same commons.
In order to illustrate this method, the 25-bus test power
system is used, designed within the Power Systems
Department, from Politehnica University of Timisoara,
Romania [12].
Concerning the structure of the paper, following
the Introduction presented within the 1st section, the 2 nd
one is focusing on describing the method. The results
are presented and discussed within the 3 rd section. The
case study power system is presented within the same
section. The last section presents the conclusion of the
paper.

2. METHOD DESCRIPTION
Kirschen method organizes the network's buses
and branches in homogeneous groups according to the
following concepts: the domain of generator, the commons and the links [9]-[10]. These concepts are used to
obtain the state graph of generators contribution to
consumers in a domain and for determining the generator
contributions to individual consumers and power flows.

ISSN 2067 5534 2010 JSE

63

JOURNAL OF SUSTAINABLE ENERGY, VOL. 1, NO. 4, DECEMBER, 2010

The method is applied independently both for active and

reactive power, based on complete AC power flow [11].
The domain of a generator represents a subset N of
n buses, which are supplied by this generator. A particular
bus belongs to a generator domain if there is a trace
through the transmission network for which the direction of
power flow is from the generator to the bus. In the
manner of domain definition results that there is an
overlap between the domains of the various generators. It
is noted that the active domain of a generators does
not usually cover the same set of buses as its reactive
domain.
The generator common is a subset of neighbored
buses from the subset N, which are supplied by the
same generators. The subsets of buses which are not
connected and not supplied by the same generators are
treated as separate commons. A bus therefore belongs to
only one common. The rank of a common is defined as
the number of generators supplying the buses within this
common.
Once the buses have been divided into commons, each
network element within the common connects two buses
which belong to the same common. If a network branch is
inside of a common, then it will connect two buses which
are belonging to different commons. A link consists by one
or more external network branches connecting the same
commons. It is very important to note that the actual flows
in all branches of a link are all in the same direction.
Furthermore, the flow direction in a link is always from a
common of rank i to a common of rank j, where j is always
greater than i.
Knowing the direction of the power flow through the
network elements and being defined the commons and the
links, the graph theory can be applied. The state graph for
the problem studied can be obtained using the following
representation conventions: the graph vertices correspond to
commons, graph arcs correspond to links between
commons and the arc orientation is given by the power
flows. The initial buses of arcs correspond to commons of
lower rank, while the terminal buses correspond to
commons of higher rank.
The state graph provides only a qualitative view of
power system. The inflow of a common is defined as the
sum of the injected power by the sources connected to
buses located in this common and the power imported in
this common from other common by links. This inflow
is always positive.
The calculus of generator contribution to various
commons (D is the set of commons) is made in the
following manner [11]:

Pjki w ij Pjk
Pk

(1)

Pjk , k D \ j

(2)

jD

Pjki
wki

jD

Pk

, kD

(3)

where: w j the contribution of the i generator to the

i

load and the outflow of the j common; wk the contribution of the i generator to the load and the outflow of the
k common; Pjk the flow on the link between the j and k
commons; Pjki the flow on the link between the j and k
commons due to the i generator; Pk the inflow of the k
common.
The inflow of the root buses of state graph is produced
entirely by the generators embedded in these commons.
The proportion of the outflow traceable to each of these
generators can therefore be readily computed and
propagated to commons of higher rank.
Knowing the common a bus belongs to and the
con-tributions of each generator to each common,
there-fore gives the ability to compute how much
power each generator contributes to each load. It also
makes it possible to compute what proportion of the
use of each branch can be apportioned to each
generator. For network elements linking buses in
separate com-mons, the proportion of usage should
based on the contribution of the generator to the lower
ranked common.
Within the methodology the hypothesis regarding the
contribution of the generators to the power losses on a
branch in proportion to the rate of use of this branch is
adopted. So the power losses allocation can be performed.

3. CASE STUDY
A 25 buses test power system has been used for the
analyses [12]. It has 29 branches. It was developed
based on the south-west side of the Romanian Power
System. 6 P-U buses (the slack bus is bus number 1) and 19
P-Q buses, the voltage level for 2 buses is 400 kV, 8 buses
are at 220 kV, 10 buses at 110 kV, one bus at 24 kV, 2
buses at 15 kV and 2 buses at 10 kV. In this particular
operating condition, 4 P-Q buses and 3 P-U buses have
zero consume power and the source from bus 6 is a
synchronous compensator (Fig. 1).
17 network elements are electrical overhead lines
(one of 400 kV, 8 of 220 kV and 8 of 110 kV) and one
is under-ground cable; 5 are transformers and 6 autotransformers. The generated and consumed active
power, for the 25 buses test system is synthesized in
Table 1. The active power flow through the network
elements is presented in Table 2.

Table 1. Configuration of the P-U and P-Q buses

Bus
64

U [kV]

PC [MW]

Pg [MW]

Bus

U [kV]

ISSN 2067 5534 2010 JSE

PC [MW]

Pg [MW]

JOURNAL OF SUSTAINABLE ENERGY, VOL. 1, NO. 4, DECEMBER, 2010

Bus
1
2
3
4
5
6
7
8
9
10
11
12
13

U [kV]
24
15.8
15.5
10.6
114.334
10.815
405.053
400.23
236.179
236.791
238.393
235.084
222.341

PC [MW]
80
8
80
0
0
0
350
530
156
175
400
0
170

Pg [MW]
738.75
1042.68
680.68
50
20
4
0
0
0
0
0
0
0

Bus
14
15
16
17
18
19
20
21
22
23
23
25

U [kV]
221.854
224.237
223.315
118.353
118.535
116.18
117.199
113.287
113.52
113.538
113.197
112.132

PC [MW]
237
0
0
0
120
32
22
20
35
12
58
24

Pg [MW]
0
0
0
0
0
0
0
0
0
0
0
0

P [MW]
0.06
1.38
0.63
0.09
0.12
11.51
0.11
0.1
0.03
0.03
0.05
0.03
0.11
0.08

Lij [km]
42
3
122
72
32
15
72
54
86
12
37
42
3
122

Fig. 1. The case study power system

Table 2. Active power flows on the system branches
Bus i
1
7
2
10
7
17
17
3
12
11
4
9
12
11
15

Bus j
7
9
10
8
8
19
20
11
11
17
18
12
18
13
13

Pij [MW]
657.5
249.2
1032.1
471.8
58.1
23.8
23
598.6
14
46.9
49.8
92.1
78
60.5
240.6

P [MW]
1.26
0.39
2.61
0.7
0.12
0.22
0.22
2.1
0.09
0.03
0.24
0.73
0.08
1.22
2.77

Lij [km]
153
173
51
31
29
94
72
68
86
72
136
38
153
173
51

Bus i
6
11
13
23
23
10
15
15
22
16
20
18
5
21

The active power losses for the entire power system are
27.11 MW.

Bus j
13
14
14
25
24
15
23
16
21
22
19
20
21
25

Pij [MW]
3.9
102.6
134.4
11.5
58.1
373
81.7
47.9
12.8
47.8
8.4
7.6
19.9
12.6

This paper proposes a simple method of

transmission cost allocation related to power losses, based

ISSN 2067 5534 2010 JSE

65

JOURNAL OF SUSTAINABLE ENERGY, VOL. 1, NO. 4, DECEMBER, 2010

on results obtained with Kirschen method for operating

condition without losses. A software tool in Mathematica
environment has been developed.
The system contains 6 generators, so will result 6
commons (Fig. 2). The common 5, corresponding to source
of bus 5, will contain the largest number of buses, even the
generated power is 20 MW and the consumed power is
149 MW. The generator of common 3 produces 680.85 MW

for a load of 534 MW. It is clear that this generator will

supply other areas through the link 3-6 (the lines 11-13
and 11-14), for a total amount of 163.1 MW. Also,
commons 3 and 4 are linked by the 4-3 line, with the
normal sense of the power flow from bus 4 to bus 3. This
link consists of lines 12-11 and 18-20, the active power
transfer being 21.6 MW.

Fig. 2. The choice of commons

Table. 3. Defining the commons in the system
Common
1
2
3
4
5
6

Buses
1, 7, 9
2, 8, 10
3, 11, 19, 20
4, 18, 12
5, 15, 16, 21, 22, 23, 24, 25
6, 14, 13

Pgenerated [MW]
738.75
1042.68
680.68
50
20
4

Pconsumed [MW]
586
713
534
120
149
407

Table. 4. Defining the links between the commons

Common i
1
2
5
1
4
3

Common j
2
5
6
4
3
6

Network elements
7-8
10-15
15-13
9-12
12-11,18-20
11-13, 11-14

The graph obtained is presented in Fig. 3. The root bus is

selected to be the slack bus 1. Beginning from this bus, his
66

Ptransfer [MW]
58.1
373
240.6
92.1
21.6
163.1

participation on the links 1-4 and 1-2 is obvious. The source

1 will participate on link 4-3 with 6.51 MW; source 4 with

ISSN 2067 5534 2010 JSE

JOURNAL OF SUSTAINABLE ENERGY, VOL. 1, NO. 4, DECEMBER, 2010

15.08 MW. Through the link 2-5 the contributions are:

19.69 MW (generator of bus 1) and 353.31 MW
(generator of bus 2). The computing algorithm ends with
common 6. To meet the load power in common 6, all the
system sources are going to participate as follows:
generator 1 - 13.57 MW, generator 2 - 216.3 MW;

generator 3 - 158.08 MW; generator 4 - 3.5 MW;

generator 5 - 12.24 MW; generator 6 - 4 MW.
The generator contribution to power flow through the
network elements is computed (Table 5).

Fig. 3. The state graph

Table. 5. The contribution of the generators to power
(the operating condition considering the active power losses)
Bus i

Bus j

Pij [MW]

1
7
2
10
7
17
17
3
12
11
4
9
12
11
15
6
11
13
23
23
10
15
15
22
16
20
18
5
21

7
9
10
8
8
19
20
11
11
17
18
12
18
13
13
13
14
14
25
24
15
23
16
21
22
19
20
21
25

657.5
249.2
1032.1
471.8
58.1
23.8
23
598.6
14
46.9
49.8
92.1
78
60.5
240.6
3.9
102.6
134.4
11.5
58.1
373
81.7
47.9
12.8
47.8
8.4
7.6
19.9
12.6

flow

through

the

network

Pijg1

Pijg2

Pijg3

Pijg4

Pijg5

Pijg6

[MW]
657.5
249.2
0
0
58.1
23.3
22.5
0
14.0
45.8
0
92.1
78.0
59.1
0
0
100.3
39.1
0
0
0
0
0
0
0
7.5
4.6
0
0.1

[MW]
0
0
1032.1
471.8
0
0
0
0
0
0
0
0
0
0
240.6
0
0
0
11.5
58.1
373.0
81.7
47.9
12.8
47.8
0
0
0
4.9

[MW]
0
0
0
0
0
0.5
0.5
598.6
0
1.1
0
0
0
1.4
0
0
2.3
93.8
0
0
0
0
0
0
0
0.1
0
0
0

[MW]
0
0
0
0
0
0
0
0
0
0
49.8
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.8
3.0
0
0

[MW]
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
19.9
7.6

[MW]
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3.9
0.0
1.5
0
0
0
0
0
0
0
0
0
0
0

ISSN 2067 5534 2010 JSE

elements

67

JOURNAL OF SUSTAINABLE ENERGY, VOL. 1, NO. 4, DECEMBER, 2010

Finally, the transmission cost allocated to generators is

computed, taking into account the active power losses.

The unit transmission cost on lines of 2 $/MWkm is

considered. The results are synthesized in Table 6.

Table. 6. The allocation of transmission costs to the generators (including the active power losses)
Bus
i

Bus
j

1
7
2
10
7
17
17
3
12
11
4
9
12
11
15
6
11
13
23
23
10
15
15
22
16
20
18
5
21

7
9
10
8
8
19
20
11
11
17
18
12
18
13
13
13
14
14
25
24
15
23
16
21
22
19
20
21
25

Line cost ij
[$]

Pij
[MW]

360
657.5
612
249.2
24
1032.1
306
471.8
346
58.1
102
23.8
62
23
58
598.6
188
14
144
46.9
136
49.8
172
92.1
144
78
272
60.5
76
240.6
550
3.9
232
102.6
58
134.4
84
11.5
6
58.1
244
373
144
81.7
64
47.9
30
12.8
144
47.8
108
8.4
172
7.6
24
19.9
74
12.6
Total [$]
Total transmission cost [$]

Gen. 1 [$]
236700
152510.4
0
0
20102.6
2376.6
1395
0
2632
6595.2
0
15841.2
11232
16075.2
0
0
23269.6
2267.8
0
0
0
0
0
0
0
810
791.2
0
7.4
492606.2

The allocation of transmission costs

Gen. 2 [$]
Gen. 3 [$]
Gen. 4 [$]
Gen. 5 [$]
0
0
0
0
0
0
0
0
24770.4
0
0
0
144370.8
0
0
0
0
0
0
0
0
51
0
0
0
31
0
0
0
34718.8
0
0
0
0
0
0
0
158.4
0
0
0
0
6772.8
0
0
0
0
0
0
0
0
0
0
380.8
0
0
18285.6
0
0
0
0
0
0
0
0
533.6
0
0
0
5440.4
0
0
966
0
0
0
348.6
0
0
0
91012
0
0
0
11764.8
0
0
0
3065.6
0
0
0
384
0
0
0
6883.2
0
0
0
0
10.8
86.4
0
0
0
516
0
0
0
0
477.6
362.6
0
0
562.4
302213.6
41324.8
7375.2
1040
846791.8

In case of consumers, working in a similar manner, the

final results presented in Table 7 are obtained (transmis-

Gen. 6 [$]
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2145
0
87
0
0
0
0
0
0
0
0
0
0
0
2232

sion costs allocated to consumers in the presence of

active power losses).

Table. 7. The transmission cost allocation to consumers (including the active power losses)
Bus Bus
Pij
i
j
[MW] Load 7 [$]
1
7
657.5
126000
7
9
249.2
0
2
10 1032.1
0
10
8
471.8
0
7
8
58.1
0
17 19
23.8
0
17 20
23
0
3
11 598.6
0
12
11
14
0
11
17
46.9
0
4
18
49.8
0
9
12
92.1
0
12 18
78
0
11
13
60.5
0
15 13 240.6
0
68

Load 8 [$]
20916
0
11500.5
144370.8
20102.6
0
0
0
0
0
0
0
0
0
0

The allocation of transmission costs

Load 9 [$] Load 10 [$] Load 11 [$] Load 13 [$] Load 14 [$] Load 18 [$]
56160
0
3290.9
277.4
441.5
26366.2
95472
0
5594.4
471.8
750.6
44822.5
0
4265.8
0
3268.9
2584.3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
22669.8
0
7325.7
0
0
0
1718.5
144.9
230.6
0
0
0
0
0
0
0
0
0
0
0
0
6359.4
0
0
1572.3
132.6
210.9
12597.2
0
0
0
0
0
10546.5
0
0
0
9190.3
7265.7
0
0
0
0
10212
8073.6
0
ISSN 2067 5534 2010 JSE

JOURNAL OF SUSTAINABLE ENERGY, VOL. 1, NO. 4, DECEMBER, 2010

Bus Bus
Pij
i
j
[MW] Load 7 [$]
6
13
3.9
0
11
14 102.6
0
13 14 134.4
0
23 25
11.5
0
23 24
58.1
0
10 15
373
0
15 23
81.7
0
15 16
47.9
0
22 21
12.8
0
16 22
47.8
0
20 19
8.4
0
18 20
7.6
0
5
21
19.9
0
21 25
12.6
0
Total [$]
126000
Bus
i

Bus
j

1
7
2
10
7
17
17
3
12
11
4
9
12
11
15
6
11
13
23
23
10
15
15
22
16
20
18
5
21

Pij
[MW]

7
657.5
9
249.2
10
1032.1
8
471.8
8
58.1
19
23.8
20
23
11
598.6
11
14
17
46.9
18
49.8
12
92.1
18
78
13
60.5
13
240.6
13
3.9
14
102.6
14
134.4
25
11.5
24
58.1
15
373
23
81.7
16
47.9
21
12.8
22
47.8
19
8.4
20
7.6
21
19.9
25
12.6
Total [$]
Total transmission cost [$]

Load 8 [$]
0
0
0
0
0
0
0
0
0
0
0
0
0
0
196889.9

The allocation of transmission costs

Load 9 [$] Load 10 [$] Load 11 [$] Load 13 [$] Load 14 [$] Load 18 [$]
0
0
0
1197.9
947.1
0
0
0
0
0
23803.2
0
0
0
0
0
7795.2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
33041.6
26122.3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
151632
4265.8
34845.9
57937.4
85550.7
100691.8

The allocation of transmission costs

Load 19 [$] Load 20 [$] Load 21 [$] Load 22 [$] Load 23 [$] Load 24 [$] Load 25 [$]
2047.4
1200.6
0
0
0
0
0
3358.2
2040.9
0
0
0
0
0
0
0
190.8
853.2
292.5
1413.8
400.6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2427.6
0
0
0
0
0
0
62
1364
0
0
0
0
0
3419.8
1303.5
0
0
0
0
0
439.2
98.8
0
0
0
0
0
3585.6
3168
0
0
0
0
0
123.8
289.6
0
0
0
0
0
754.6
573.6
0
0
0
0
0
205.3
480.2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
966
0
0
0
0
0
348.6
0
0
0
1928.9
8623.5
2956.6
14290.4
4048.6
0
0
0
0
1732.2
8372.5
1660.1
0
0
502.5
2246.5
0
0
316.6
0
0
235.6
0
0
0
148.4
0
0
1128.2
5044.2
0
0
710.8
907.2
0
0
0
0
0
0
733.6
573.6
0
0
0
0
0
0
0
293
0
0
0
184.6
0
0
0
0
0
0
932.4
18064.3
11092.8
4279
16767.4
4981.4
24425.3
9368.1
846791.8

4. Conclusion
The authors are proposing a new method of assessing
the transmission cost related to active power losses.
The Kirschen method is suitable to estimate the performance of the system. This method is simple, intuitive and
is based on complete AC power flow. The cost

allocation has been performed separately for sources and

consumers, tracing the active power flow. The two
components of the transmission cost allocation can be
weighted differently (in the range 01, their sum is
obviously 1). Due to obvious influence of the active
power losses, it is recommended not to be neglected.

ISSN 2067 5534 2010 JSE

69

JOURNAL OF SUSTAINABLE ENERGY, VOL. 1, NO. 4, DECEMBER, 2010

References

[7]

[1] M. Shahidepour, H. Yahim, Z. Li, Market Operations in electric

power systems. Forecasting, scheduling and risk
management, John Wile y &Sons, 2002.
[2] M. Ilic, F. Galiana and L. Fink, Power system
restructuring: Engineering and Economics, Norwell, MA:
Kluwer, 1998.
[3] H.H. Happ, Cost of wheeling methodologies, IEEE
Transactions on Power Systems, Vol. 9, No. 1, February
1995, pp. 147-156.
[4] J. Bialek, Topological generation and load distribution
factors for supplement charge allocation in transmission open
access, IEEE Transactions on Power Systems, Vol. 12, No.
3, August 1997, pp. 1185-1193.
[5] O. Pop, M. Nemes, Power system allocation with network
matrices, Proceeding of the 7th International Power Systems
Conference, November 22-23, 2007, Timisoara,
Romania, pp. 527-534.
[6] O. Pop, C. Barbulescu, M. Nemes, St. Kilyeni, The influence
of the common structure modification on the allocation,
Proceedings of World Academy of Science, Engineering

70

[8]

[9]

[10]

[11]

[12]

and Technology, Paris, 04-06 July, 2008, Vol. 30, pp. 126132.
O. Pop, C. Barbulescu, M. Nemes, St. Kilyeni, The influence
of the common structure modification on the active power
losses allocation, Proceedings of World Academy of
Science, Engineering and Technology, Vienna, 13-15
August, 2008, vol. 31, July 2008, pp. 660-665.
G. Strbac, D. Kirschen, S. Almed, Allocating transmission
system usage on the basis of traceable contributions of
generators and loads to flow, IEEE Transactions on Power
Systems, Vol. 13, No. 2, May 1998, pp. 527-534.
D. Kirschen, G. Strbac, Tracing active and reactive power
between generators and loads using real and imaginary
currents, IEEE Transactions on Power Systems, Vol. 14, No.
4, November 1999, pp. 1212-1319.
D. Kischen, R. Allan, G. Strbac, Contributions of
individual generators to loads and flow, IEEE
Transactions on Power Systems, Vol. 12, No. 1,
February 1997, pp. 52-66.
O. Pop, Contributions on the tariff access assess-ment to
the transport system, PhD Thesis, Electrical and Power
Engineering Faculty, Politehnica University of
Timisoara, 2009.
St. Kilyeni, Optimization techniques in power engineering,
Orizonturi Universitare, Timisoara, 2009 (in Romanian).

ISSN 2067 5534 2010 JSE