ISSSN:2229-6093

Satyavir Singh et al ,Int.J.Computer Technology & Applications,Vol 3 (3), 887-894

Power Tracing in a Deregulated Power System : IEEE 14-Bus Case

Satyavir Singh
Indian Institute of Technology, Rookee,India
E-mail: [email protected]

Abstract

In this present era , a fair transmission pricing scheme is sensitivity analysis cannot consider accurately, the
an important issue due to incresed deregulation and reactive power transfer allocation and system non
restructuring of power sector. In this view, issue of linearity. A novel electricity tracing method has been
tracing the flow of electricity has been gain importance proposed [4] which, under the assumption that nodal
as its solution helps in evaluating a fair and transparent inflows are shared proportionally between the nodal
tariff.An electricity tracing method would make it outflows, allows one to trace the flow of electricity in a
possible to charge the generators and/or consumers on meshed network. This method was proposed by J.Bialek
the basis of actual transmission facility used. This paper [5] .The complete description of this method for the
focuses on tracing of electricity using Bialek’s tracing tracing of real power is given in section 2.
algorithm .Case study is carried out using an IEEE 14- The characteristics of nodal method are given below:
bus system with three simultaneous bilateral transactions (i) The transmission losses must be removed from the
simulated in Power world simulator. lines before the application of the method.
(ii) Matrix calculation is more complex and the speed is a
1. Introduction problem for a big network.
(iii) The method can handle cyclic flows in the system, so
Power system operation in many electricity supply the method is suitable for systems with loop flows.
systems worldwide, has been experiencing dramatic (iv) If MVAR tracing is required, the method becomes
changes due to the on going restructuring of the industry. messy with the introduction of artificial nodes.
The vertically integrated structure of power industry is
being replaced by market structure which led to a In [6] a new method for determining the generator’s
significant increase in power wheeling transactions. In contribution to a particular load is presented. The method
such a structure a transmission system is being used by uses the nodal generation distribution factors (NGDF-s).
multiple generation and load entities that do not own the It features a search algorithm, capable of handling the
transmission system. In view of market operation it active and reactive powers. Paper [7] provides new
becomes more important to know the role of individual insights into the electricity tracing methodology, by
generators and loads to transmission lines and power representing the inverted tracing upstream and
transfer between individual generators to loads. Basically downstream distribution matrices in the form of matrix
there are three methods for power tracing which are power series and by applying linear algebra analysis. A
described below [1]: rigorous mathematical proof of the invertibility of the
tracing distribution matrices is given, along with a proof
of convergence for the matrix power series
1.1. Node method
1.2. Graph method
In a meshed transmission network there are number of
possible routes by which electrical power can flow from This method assumes that a generator has the priority to
sources to sinks. It is possible to determine relation provide power to the load on the same bus and is based
between the generators/loads and the flows in on the following lemmas of graph theory [8].
transmission lines by means of sensitivity analysis, that is Lemma 1: A lossless, finite-nodes power system without
by determining how a change in nodal loop flow has at least one pure source, i.e. a generator bus
generation/demand influences the flow in a particular line with all incident lines carrying outflows.
[2, 3]. Although this method based on dc load flow and Lemma 2: A lossless, finite-nodes power system without

IJCTA | MAY-JUNE 2012 887

Available [email protected]
 ISSSN:2229-6093
Satyavir Singh et al ,Int.J.Computer Technology & Applications,Vol 3 (3), 887-894

loop flow has at least one pure sink, i.e. a load bus with
all incident lines carrying inflows. The characteristics of 10MW 40MW
the graph method are: i
(i) The transmission losses must be removed from the j m
lines before the application of the method.
(ii) Extraction factor, Contribution factor matrices are
sparse and easy to calculate.
(iii) The method cannot handle loop flows, but is able to
n k
detect loop flows.
(iv) A generator has the priority to provide power to the 60MW 90MW
load on the same bus
Figure 1. proportinal sharing principle
1.3. Method of commons

This method was proposed by Kirschen and additional 60𝑀𝑊

details of the method can be found in [8, 9]. The × 10𝑀𝑊 = 6𝑀𝑊
100𝑀𝑊
technique can be applied to both active and reactive Supplied by j-i, and
power. The principal ideas of the method are presented
below. 60𝑀𝑊
× 90𝑀𝑊 = 54𝑀𝑊
100𝑀𝑊
(i) The transmission losses can be directly accounted for Supplied by line k-i.
and the contributions to them calculated. Bialek’s tracing algo is based on the proportional sharing
(ii) The contribution factors to all the loads inside a principle with two tracing algorithms viz upstream and
Common as well as the outflows are constant. downstream looking algorithm the description of these
(iii) More simple and intuitive for large systems. methods are given in section 2.1 and 2.2 respectively.
(iv) The size and shape of a Common is subjected to
change radically even with a small change in the 2.1. Tracing Electricity using upstream looking
direction of line flows. algo

The total flow 𝑃𝑖 , the inflow to the 𝑖 𝑡ℎ bus, is the sum of

2. Proportional sharing principle all the inflows through the lines connected to the bus and
the local bus injection.
The proportional sharing principle is based on kirchhoff’s
current law and is topological in nature. It deals with a 𝑃𝑖 = 𝑃𝑖−𝑗 + 𝑃𝐺𝑖 𝑓𝑜𝑟 𝑖 = 1,2, … … … . 𝑛 ...(1)
𝑗ɛℜ
general transportation problem and assumes that the
network node is a perfect mixer of incoming flows.
Practically the only requirement for the input data is that where ℜ is the set of nodes directly supplying node i,
Kirchhoff’s current law must be satisfied for all the nodes implying power flow towards 𝑖 𝑡ℎ node. If the line losses
in the network.In this respect the method is equally are neglected, then 𝑃𝑗 −𝑖 = 𝑃𝑖−𝑗 . Equation (1) can be
applicable to ac as well as dc power flow. Figure1 further expanded to become:
illustrates the basic principal to trace the flow of
electricity where four lines are connected to bus i two 𝑃 𝑗 −𝑖
with inflows and two with out flows.The nodal sum i.e. 𝑃𝑖 = 𝑗ɛℜ 𝑃𝑗 + 𝑃𝐺𝑖 𝑓𝑜𝑟 𝑖 = 1,2, … … … . 𝑛 ...(2)
𝑃𝑗
total incoming or total outgoing power at node i is equal
to 100 MW. According to proportional sharing principle 𝑃 𝑗 −𝑖
By defining 𝑐𝑗𝑖 = to express relationship between
𝑗
The 40MW out flowing in line i-m consists of line flow and the nodal flow at the 𝑗𝑡ℎ node, using
40𝑀𝑊 proportional sharing principle 𝑃𝑗 −𝑖 = 𝑐𝑗𝑖 𝑃𝑗 , substituting
× 10𝑀𝑊 = 4𝑀𝑊
100𝑀𝑊 this in (2) yields:
Supplied by line j-i, and

40𝑀𝑊 𝑃𝑖 − 𝑗ɛℜ 𝐶𝑗𝑖 𝑃𝑗 = 𝑃𝐺𝑖 𝑜𝑟 𝐴𝑢 𝑃 = 𝑃𝐺 ...(3)

× 90𝑀𝑊 = 36𝑀𝑊
100𝑀𝑊
Supplied by line k-i. In the same way , the 60MW out P is the vector of gross nodal flows; 𝑃𝐺 is the vector of
flowing in line i-n consists of nodal generations, while 𝐴𝑢 is called the Upstream matrix,
which elements can be generalized as follow:

IJCTA | MAY-JUNE 2012 888

Available [email protected]
 ISSSN:2229-6093
Satyavir Singh et al ,Int.J.Computer Technology & Applications,Vol 3 (3), 887-894

𝑃𝑖 − 𝑙ɛ𝜇 𝐶𝑙𝑖 𝑃𝑙 = 𝑃𝐿𝑖 𝑜𝑟 𝐴𝑑 𝑃 = 𝑃𝐿 ...(10)

1 𝑓𝑜𝑟 𝑖 = 𝑗
𝑃 𝑗 −𝑖
𝐴𝑢 𝑖𝑗 = −𝐶𝑗𝑖 = − 𝑓𝑜𝑟 𝑗ɛℜ ...(4) 𝑃 is the vector of net nodal powers; 𝑃𝐿 is the vector of
𝑃𝑗
nodal load demands, while 𝐴𝑑 is called the Downstream
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 matrix, which elements can be generalized as follow:

The 𝑖 𝑡ℎ element of 𝑃 = 𝐴𝑢 −1 𝑃𝐺 shows the participation 1 𝑓𝑜𝑟 𝑖 = 𝑙

of the 𝑘 𝑡ℎ generation to the 𝑖 𝑡ℎ nodal flow and determines 𝑃 𝑙−𝑖
the relative participation of the nodal generations in 𝐴𝑑 𝑖𝑙 = −𝐶𝑙𝑖 = − 𝑓𝑜𝑟 𝑙ɛ𝜇 ...(11)
𝑃𝑙
meeting a retailer’s demand, given as: 0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

𝑃𝑖 = 𝑛
𝑘 =1 𝐴𝑢 −1 𝑖𝑘
𝑃𝐺𝑘 𝑓𝑜𝑟 𝑖 = 1,2, … … … . 𝑛 ...(5) The 𝑖 𝑡ℎ element of 𝑃 = 𝐴𝑑 −1 𝑃𝐿 shows the distribution of
the 𝑖 𝑡ℎ nodal power between all the loads in the system.
In summation form,
A line out flow in line j-i from node i can be therefore
calculated using proportional sharing principle ,as
𝑃𝑖 = 𝑛
𝑘 =1 𝐴𝑑 −1 𝑖𝑘
𝑃𝐿𝑘 𝑓𝑜𝑟 𝑖 = 1,2, … … … . 𝑛 ...(12)
𝑃 𝑗 −𝑖
𝑃𝑗 −𝑖 = 𝑛
𝑘 =1 𝐴𝑢 −1 𝑖𝑘
𝑃𝐺𝑘 𝑓𝑜𝑟 𝑖 = 1,2, … . 𝑛 ...(6)
𝑃𝑖
The inflow to node i from line i-l can be calculated using
the proportional sharing principle as
Finally, load demand at the 𝑖 𝑡ℎ bus, applying the
proportional methodology is given by: 𝑃 𝑖−𝑙
𝑃𝑖−𝑙 = 𝑛
𝑘=1 𝐴𝐷 −1 𝑖𝑘
𝑃𝐿𝑘 𝑓𝑜𝑟 𝑖 = 1,2, … . 𝑛 ...(13)
𝑃𝑖
𝑃 𝐿𝑖
𝑃𝐿𝑖 = 𝑃𝑖
𝑃𝑖
this equation allows to determine how the line flows
supply individual loads.
𝑃 𝐿𝑖
𝑃𝐿𝑖 = 𝑛
𝑘 =1 𝐴𝑢 −1 𝑖𝑘
𝑃𝐺𝑘 𝑓𝑜𝑟 𝑖 = 1,2, … … . 𝑛 ...(7)
𝑃𝑖
The generation at a node is also an inflow and can be
calculated using the proportional sharing principle as
This equation shows the contribution of the 𝑖 𝑡ℎ system
generator to the 𝑘 𝑡ℎ load demand and can be used to trace 𝑃 𝐺𝑖
where the power of a particular load comes from. 𝑃𝐺𝑖 = 𝑃𝑖
𝑃𝑖

2.2. Tracing Electricity using downstream 𝑃𝐺𝑖 =

𝑃 𝐺𝑖 𝑛
𝐴𝑑 −1 𝑃𝐿𝑘 𝑓𝑜𝑟 𝑖 = 1,2, … … . 𝑛 ...(14)
𝑃𝑖 𝑘=1 𝑖𝑘
looking algo

The total flow 𝑃𝑖 , the outflow to the 𝑖 𝑡ℎ bus, is the sum of This equation again shows that the share of the output of
all the outflows through the lines connected to the bus the 𝑖 𝑡ℎ generator used to supply the 𝑘 𝑡ℎ load demand. The
and the local bus load results obtained in case of equation (7) and equation (14)
are same i.e in case of equation (14) a transpose of table 2
results.
𝑃𝑖 = 𝑙ɛ𝜇 𝑃𝑖−𝑙 + 𝑃𝐿𝑖 𝑓𝑜𝑟 𝑖 = 1,2, … … … . 𝑛 ...(8)

where μ is the set of nodes directly supplied from node i, 3. Results and discussion
implying power flowing from the 𝑖 𝑡ℎ node. If the line
IEEE 14-bus system is simulated in power world
losses are neglected, then 𝑃𝑙−𝑖 = 𝑃𝑖−𝑙 .Equation (8) can
simulator with additional three bilateral transactions
be further expanded into:
which involve different transaction locations. The detail
of the transactions are as follows:
𝑃 𝑙−𝑖
𝑃𝑖 = 𝐿ɛ𝜇 𝑃𝑙 + 𝑃𝐿𝑖 𝑓𝑜𝑟 𝑖 = 1,2, … … … . 𝑛 ...(9)
𝑃𝑙
T1: Injection of 20 MW at bus1 and removal at bus 5;
𝑃 𝑙−𝑖 T2: Injection of 20 MW at bus2 and removal at bus 14;
Defining 𝑐𝑙𝑖 = 𝑃𝑙
expressing relationship between line T3: Injection of 20 MW at bus3 and removal at bus 1;
flow and the nodal flow at the 𝑙 𝑡ℎ node and using The transmission network data and load flow results are
proportional sharing principle, 𝑃𝑙−𝑖 = 𝑐𝑙𝑖 𝑃𝑙 . Substituting given in table 4-7. The proportional sharing
this in (9) yields

IJCTA | MAY-JUNE 2012 889

Available [email protected]
 ISSSN:2229-6093
Satyavir Singh et al ,Int.J.Computer Technology & Applications,Vol 3 (3), 887-894

approach has been tested on IEEE 14-bus system using

Matlab simulation program.
Table 1 depicts the contribution of generators to line
flows , 𝑃𝑖𝑗 ,𝑡𝑟𝑎𝑛𝑠 is the power flow in line i-j , 𝑃𝐺𝑖,𝐿𝐼𝑁𝐸 is
the contribution of generator Gi(connected to bus i) to the
line flows. The graphical reprentation of table 1 is shown
in figure 3. It can be seen that generator G3 contribute
zero power to the transmission system . The reason
behind it is that the total power of generator G3 is goes to
local load i.e. L3 which can be seen in table 2 and figure
4.Table 3 depicts the contribution of line flows to loads.
Again it can be seen that load L1 is free from
transmission system use. This is due to fact that power is
available locally for this load by generator G1 which can
be seen in figure 4. Results are also verified in table 1,2
and 3 and a slight mismatch is there because of losses in
the system. Proportional sharing of losses can also be
done for more accurate results. .
Figure 2. IEEE 14-bus system simulated in power
world simulator

3.1 Contribution of generators to line flow

Table 1. Contribution of generators to line flows

BUS(i-j) 𝑷𝒊𝒋,𝒕𝒓𝒂𝒏𝒔 𝑷𝑮𝟏,𝑳𝑰𝑵𝑬 𝑷𝑮𝟐,𝑳𝑰𝑵𝑬 𝑷𝑮𝟑,𝑳𝑰𝑵𝑬

1-2 151.0 151.0 0 0

1-5 82.9 82.9 0 0
2-3 67.4 48.2341 19.1659 0
2-4 64.9 46.4450 18.4550 0
2-5 53.0 37.9289 15.0711 0
3-4 -8.7 6.8857 1.8143 0
4-5 -50.4 44.8107 5.5893 0
4-7 35.9 28.4135 7.4865 0
4-9 20.5 16.2250 4.2750 0
5-6 52.7 46.8557 5.8442 0
6-11 7.8 6.9350 0.8650 0
6-12 9.5 8.4465 1.0535 0
6-13 24.3 21.6052 2.6948 0
7-8 0 0 0 0
7-9 35.9 28.4135 7.4865 0
9-10 4.8 3.7990 1.0010 0
9-14 22.2 17.5705 4.6295 0
10-11 -4.2 3.7342 0.4658 0
12-13 3.3 2.9340 0.3660 0
13-14 13.6 12.0918 1.5082 0

IJCTA | MAY-JUNE 2012 890

Available [email protected]
 ISSSN:2229-6093
Satyavir Singh et al ,Int.J.Computer Technology & Applications,Vol 3 (3), 887-894

160

140

120
Line flow(MW)

100

80
G3
60 G2
40 G1

20

Line 6-11
Line 6-12
Line 6-13

Line 9-10
Line 9-14
Line 4-7
Line 4-9
Line 1-2
Line 1-5
Line 2-3
Line 2-4
Line 2-5
Line 3-4

Line 4-5
Line 5-6

Line 7-8
Line 7-9

Line 12-13
Line 13-14
Line 10-11
Figure 3. Contribution of generators to line flows

3.2 Contribution of generators to load:

Table 2. Contribution of generators to loads

BUS LOAD(MW) 𝑷𝑮𝟏,𝑳𝑶𝑨𝑫 (MW) 𝑷𝑮𝟐,𝑳𝑶𝑨𝑫 (MW) 𝑷𝑮𝟑,𝑳𝑶𝑨𝑫 (MW)

1 20 20 0 0
2 21.7 15.5294 6.1706 0
3 94.2 54.0301 20.5653 19.6046
4 47.8 37.8320 9.9680 0
5 27.6 24.5392 3.0608 0
6 11.2 9.9579 1.2421 0
7 0 0 0 0
8 0 0 0 0
9 29.5 23.3482 6.1518 0
10 9 7.5333 1.4667 0
11 3.5 3.1119 0.3881 0
12 6.1 5.4235 0.6765 0
13 13.5 12.0029 1.4971 0
14 34.9 28.9166 5.9834 0
TOTAL 319 242.2250 57.1704 19.6046

IJCTA | MAY-JUNE 2012 891

Available [email protected]
 ISSSN:2229-6093
Satyavir Singh et al ,Int.J.Computer Technology & Applications,Vol 3 (3), 887-894

100
Generator contribution to load(MW) 90
80
70
60
G3
50
G2
40
30 G1
20
10
0
L1 L2 L3 L4 L5 L6 L7 L8 L9 L10 L11 L12 L13 L14

Figure 4. Contribution of generators to loads

3.3. Contribution of line flows to loads

Table 3. Contribution of line flows to loads

LOAD

L1 L2 L3 L4 L5 L6 L7 L8 L9 L10 L11 L12 L13 L14 TOTAL

LINE
1-2 0 15.9032 52.9128 25.5469 7.8533 3.1869 0 0 15.7664 3.8696 0.9959 1.7357 3.8413 15.3660 146.978
1-5 0 0 2.5670 13.3996 17.4823 7.0943 0 0 8.2697 3.8517 2.2170 3.8638 8.5511 14.4267 81.7237
2-3 0 0 67.1300 0 0 0 0 0 0 0 0 0 0 0 67.1300
2-4 0 0 5.1730 27.0028 0 0 0 0 16.6649 2.9583 0 0 0 12.2866 64.0856
2-5 0 0 1.6292 8.5045 11.0957 4.5026 0 0 5.2486 2.4446 1.4071 2.4523 5.4272 9.1564 51.8682
3-4 0 0 9.3820 0 0 0 0 0 0 0 0 0 0 0 9.3820
4-7 0 0 0 0 0 0 0 0 18.8677 3.3493 0 0 0 13.9106 36.1276
4-9 0 0 0 0 0 0 0 0 10.6320 1.8874 0 0 0 7.8387 20.3581
4-5 0 0 4.1070 21.0028 0 0 0 0 13.2307 2.3487 0 0 0 9.7547 50.4439
5-6 0 0 0 0 0 11.2000 0 0 0 3.7632 3.5000 6.1000 13.5000 13.1504 51.2136
6-11 0 0 0 0 0 0 0 0 0 3.8212 3.5539 0 0 0 7.3751
6-12 0 0 0 0 0 0 0 0 0 0 0 6.1753 1.6400 1.5976 9.4129
6-13 0 0 0 0 0 0 0 0 0 0 0 0 12.0800 11.7672 23.8472
7-8 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
7-9 0 0 0 0 0 0 0 0 18.8685 3.3495 0 0 0 13.9112 36.1292
9-10 0 0 0 0 0 0 0 0 0 5.2420 0 0 0 0 5.2420
9-14 0 0 0 0 0 0 0 0 0 0 0 0 0 22.3740 22.3740
10-11 0 0 0 0 0 0 0 0 0 3.7670 0 0 0 0 3.7670
12-13 0 0 0 0 0 0 0 0 0 0 0 0 1.6473 1.6046 3.2519
13-14 0 0 0 0 0 0 0 0 0 0 0 0 0 13.5280 13.5280

IJCTA | MAY-JUNE 2012 892

Available [email protected]
 ISSSN:2229-6093
Satyavir Singh et al ,Int.J.Computer Technology & Applications,Vol 3 (3), 887-894

80 Line 1-2
Line 1-5
70 Line 2-3
Line 2-4
Line 2-5
60 Line 3-4
Line 4-7
50 Line 4-9
Line 4-5
40 Line 5-6
Line 6-11
30 Line 6-12
Line 6-13
Line 7-8
20 Line 7-9
Line 9-10
10 Line 9-14
Line 10-11
0 Line 12-13
Line 13-14
L1 L2 L3 L4 L5 L6 L7 L8 L9 L10 L11 L12 L13 L14

Figure 5. Contribution of line flows to loads

4. Bus data and load flow results:

Table 4. Line data for the test system

Table5. Load flow result

IJCTA | MAY-JUNE 2012 893

Available [email protected]
 ISSSN:2229-6093
Satyavir Singh et al ,Int.J.Computer Technology & Applications,Vol 3 (3), 887-894

Table 6: Transformer data record

Table 7: Power flow results with generator and load data record

Conclusion
[2] Rudnick. H., Palma. R. and Fernandez. J. E., "Marginal
pricing and supplement cost allocation in transmission open
In a deregulated environment, generation, transmission
access', IEEE Trans., PWRS-10, (2), 1995, pp. 1125-1142.
and distribution are independent activities. This
unbundling of the transmission services has resulted in [3] NG. W.Y., "Generalized generation distribution factors for
need to trace the flow of power i.e. to assess the impact of power system security evaluations", IEEE Trans., PAS-100, (3),
a particular generator or the load on the power system. 1981, pp.1001-1005.
This paper is based on Bialek’s tracing method which
may be applied to both real and reactive power flows. [4] Bialek, J., “Identification of source-sink connections in
The results corresponding to the upstream looking transmission networks”. Proceedings of fourth IEE conference
algorithm tabulated in table 1 and 2 allows one to assess: on Power system control and management, London, April 1996,
pp. 200-204.
 What amount of power from a generator goes to [5] J. Bialek, "Tracing the flow of electricity," IEE Proc. Gener.
a particular transmission line (Table 1). Transm. Distrib. vol. 143. no. 4. Jul 1996, pp. 313- 320.
 What amount of power from a generator goes to
a particular load (Table 2). [6] Ferdinand Gubina, David G., and Ivo B., “A Method for
Determining the Generators’ Share in a Consumer Load” IEEE
The results corresponding to the downstream looking Trans Power Sys. Vol. 15, no. 4, November 2000, pp. 1376-
algo tabulated in table 2 and 3 allows one to assess: 1381.

[7] Chira Achayuthakan, Chris J. Dent, Janusz W. Bialek, and

 What amount of power from a generator goes to Weerakorn Ongsakul, “Electricity Tracing in Systems With and
a particular load (Table 2). Without Circulating Flows: Physical Insights and Mathematical
 The contribution of a particular transmission line Proofs”. IEEE Trans Power Systems, vol. 25, No. 2, May 2010,
flow to a particular load (Table 3). pp. 1078-1087

[8] F.F. Wu, Y. Ni., and P. Wei., "Power transfer allocation for
open access using graph theory - Fundamentals and applications
References in systems without loop flows," IEEE Trans. Power Systems,
vol. 15, no. 3, Aug 2000 , pp. 923-929.
[1] M. W. Mustafa and H. Shareef, “A Comparison of Electric
Power Tracing Methods Used in Deregulated Power Systems” [9] D. Kirschen, R. AlIan, G. Strbac “Contributions of
First International Power and Energy Conference PECon 2006, Individual Generators to Loads and Flows”, IEEE Trans Power
Putrajaya, Malaysia, November 28-29, 2006, pp. 156-160 Systems, vol. 12, Feb 1997, pp. 52-60.

IJCTA | MAY-JUNE 2012 894

Available [email protected]