76 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 1, NO.

3, SEPTEMBER 2015

Congestion Management of Power System with

Interline Power Flow Controller Using Disparity
Line Utilization Factor and Multi-objective
Differential Evolution
Akanksha Mishra and G.V. Nagesh Kumar, Member, IEEE

Abstract—The restructuring of the electric power market Zsein Series transformer impedance of line i–n.
has led to complex power transmission congestion problems. gin Series transformer conductance of line i–n.
Additionally, scheduled power flows in the transmission line, as bin Series transformer susceptance of line i–n.
well as spontaneous power exchanges have also risen sharply
in recent years. The proper placement of IPFC can improve the I. I NTRODUCTION
transmission line congestion problem to a great extent. This paper
proposes a disparity line utilization factor (DLUF) for the optimal
placement of IPFC to control the congestion in transmission lines.
DLUF determines the difference between the percentages of Mega
P OWER system operation problems increase with size,
loading, and the complexity of the network. Restructur-
ing in the electric power industry has further enhanced the
Volt Ampere utilization of each line connected to the same bus.
The IPFC is placed in the lines with maximum DLUF. A multi-
problems of power systems related to power delivery and
objective function consisting of reduction of active power loss, power quality. The deregulated electric power industry, for
minimization of total voltage deviations, minimization of security example, has changed its operations, structure, ownership, and
margin and minimization of installed IPFC capacity is considered management models. The issue of transmission congestion is
for the optimal tuning of IPFC using differential evolution particularly prominent in deregulated and competitive markets,
algorithm. The proposed method is implemented for IEEE-30
bus test system under different loading conditions and the results
thus requiring an appropriate management strategy [1]. In
are presented and analyzed to establish the effectiveness on the the new competitive electric market, it is now mandatory
reduction of congestion. for the electric utilities to operate in ways that make better
use of existing transmission facilities, and in conjunction
Index Terms—Congestion, differential evolution algorithm, with maintaining the security, stability, and reliability of the
interline power flow controller, line utilization factor, optimal supplied power.
placement, optimal tuning.
In the literature, FACTS devices have been used for several
purposes including congestion management. It is a well-
recognized fact that the performance of FACTS devices in
N OMENCLATURE a power system mainly depends on its placement and tuning.
n Bus j, k. Gitizadeh and Kalantar [2] investigated a simulated annealing
Vn Complex voltage at bus (j, k). based optimization method for placement of flexible AC
V n , θn Magnitude and angle of Vn respectively. transmission systems (FACTS) devices in order to relieve
Vsein Complex controllable series injected voltage congestion in the transmission lines while increasing static
source. security margin and voltage profile of a given power system.
Vsein , θsein Magnitude and angle of Vsein respectively. Qian et al. [3] used sensitivity analysis and extended equal
Vi Complex voltage at bus i. area criterion to find the optimal location and capability of
V i , θi Magnitude and angle of V i respectively. FACTS in a power system for enhancing static voltage and
Pi ,Qi Sum of active and reactive power leaving bus i. transient stability. P. Ye et al. [4] proposed an algorithm for
Pni , Qni IPFC branch active and reactive powers leaving optimal congestion dispatch calculation with UPFC control.
bus n. A decomposition control method was introduced to solve
Iji , Iki Current in line j–i and k–i respectively. this optimal power flow problem. Mandala and Gupta [5]
Manuscript received April 13, 2015; revised May 31 and July 19, 2015; proposed a method to determine the optimal location of thyris-
accepted July 24, 2015. Date of publication September 30, 2015; date of tor controlled series compensators (TCSCs) for congestion
current version August 28, 2015. management. The optimal location is determined based on
A. Mishra is with Department of Electrical and Electronics Engineering,
GITAM Institute of Technology, GITAM University, Visakhapatnam, India. real power performance index and also on reduction in total
(e-mail: [email protected]). system active power and reactive power losses. Reddy et
G.V. Nagesh Kumar is working as Associate Professor in Department al. [6] has presented the optimal location of FACTS controllers
of Electrical and Electronics Engineering, GITAM Institute of Technology,
GITAM University, Visakhapatnam, India. (e-mail: [email protected]). considering branch loading (BL), voltage stability (VS), and
Digital Object Identifier 10.17775/CSEEJPES.2015.00038 loss minimization (LM) as objectives using genetic algorithm
2096-0042
c 2015 CSEE
MISHRA et al.: CONGESTION MANAGEMENT OF POWER SYSTEM WITH INTERLINE POWER FLOW CONTROLLER USING DISPARITY LINE UTILIZATION FACTOR AND MULTI-OBJECTIVE DIFFERENTIAL EVOLUTION 77

for management of congestion. Acharya and Mithulanan- is formulated for optimal tuning of IPFC using differential
than [7] propose two new methodologies for the placement evolution algorithm. The multi-objective function comprises
of series FACTS devices for congestion management. The reduction of active power loss, minimization of total voltage
overall objective of FACTS device placement can be either to deviations, and minimization of security margin with the usage
minimize the total congestion rent or to maximize the social of minimum value of installed IPFC. Tuning of IPFC for
welfare. Zhang et al. [8] presented an optimal power flow reduction of loss further reduces line congestion. Reduction of
(OPF) control in electric power systems incorporating IPFC voltage deviation and security margin ensures power quality
with the minimum total capacity of the converters of IPFC and system security. The proposed method is implemented
and minimizing the total active power loss of the system and tested on an IEEE 30 bus system with different loading
for reducing congestion in the lines. Mohamed et al. [9] conditions.
has compared three variants of PSO, namely, basic PSO,
inertia weight approach PSO, and constriction factor approach II. IPFC M ODELING
PSO considering a single objective, i.e., to minimize the IPFC consists of at least two back-to-back DC-AC convert-
transmission line loss. ers connected via a common DC link. Vi , Vj , and Vk are
FACTS devices are preferred in modern power systems complex voltages at buses i, j, and k, respectively. V l =
based on the requirement, and are found to deliver good Vl ∠θl (l = i, j, k) and Vl , θl are the magnitude and angle of
solutions [10]. Out of all FACTS devices, IPFC is considered V l . Vsein is the complex controllable series injected voltage
to be the most flexible, powerful, and versatile as it employs at source, which represents the series compensation of the series
least two VSCs with a common DC link. Hence, IPFC has the converter. Vsein is defined as Vsein = Vsein ∠θ sein (n = j,
capability of compensating multi-transmission lines. FACTS k). Vsein and θsein are the magnitude and angle of Vsein .
devices, such as TCSC and SSSC are also placed on the The basic model of IPFC, as shown in Fig. 1, consists of
most congested line. However, IPFC is a device connected to three buses i, j, and k [17], [18]. Two transmission lines are
multiple transmission lines. In its simplest form, it has at least connected with the bus i in common. The equivalent circuit of
two converters placed on two transmission lines connected to the IPFC with two converters is represented with two series
a common bus [11]. Proper placement of IPFC is, therefore, injected voltage sources, as shown in Fig. 2. Zsein is the series
a subject to be analyzed. transformer impedance. Pi and, Qi as given in (1) and (2) are
Location and tuning of FACTS devices in the power system the sum of the active and reactive power flows leaving the bus
is an important issue, and hence optimal placement and tuning i. The IPFC branch active and reactive power flows leaving
of IPFC has been proposed based on a previous study [10], bus n are Pni and Qni and the expressions are given in (3)
[12]. Differential evolution is a heuristic approach for mini- and (4). Iji , Iki are the IPFC branch currents of branch j–i
mizing nonlinear and non-differentiable continuous functions. and k–i leaving buses j and k, respectively [19].
It has very fast convergence, requires few control variables, X
and is robust and easy to use [14], [15]. It is also considered Pi = Vn2 gii − Vi Vn [gin cos(θi − θn ) + bin sin(θi − θn )]
as good alternative evolutionary algorithms for power system X
+ Vi V sein [gin sin(θi − θsein ) − bin cos(θi − θsein )]
applications [16]. Several authors have reported placement
and tuning of IPFC and other FACTS devices using various (1)
X
2
conventional and heuristic methods. Qi = −Vi bii − Vi Vn [gin sin(θi − θn )
Line utilization factor (LUF) is used for determining con- n=j,k
gestion of a single transmission line. Single-line FACTS
X
− bin cos(θi − θn )] − Vi V sein [gin sin(θi − θsein )
devices can be placed on the transmission line with maximum n=j,k
LUF value. However, IPFC is a multiline series FACTS device.
− bin cos(θi − θsein )] (2)
In its simplest form it consists of at least two converters
placed on two transmission lines with a common bus. The first Pni = Vn2 gnn − Vi Vn [gin cos(θn − θi ) + bin sin(θn − θi )]
converter of IPFC can be placed on the line with maximum + Vn V sein [gin sin(θn − θsein ) − bin cos(θn − θsein )] (3)
LUF. However, the placement of the other converter is an Qni = −Vn2 bnn − Vi Vn [gin sin(θn − θi ) − bin cos(θn − θi )]
issue that becomes more and more complex with increases
+ Vn V sein [gin sin(θn − θsein ) − bin cos(θn − θsein )] (4)
in system size, number of IPFCs, and the complexity of the
IPFC. Hence, LUF is not a sufficient index for obtaining IPFC where n = j, k
location placement.
gin + jbin = 1/Zsein = ysein , gnn + jbnn = 1/Zsein
In this paper, the difference of line utilization factors be-
tween two lines has been used for determining the optimal = ysein
X X .
location of IPFC. LUF gives an estimate of the difference gii = gin , bii = bin
of the percentage of line being used for the power flow. n=j,k n=j,k
First, all lines are ranked in terms of line congestion. Then,
Assuming a lossless converter, the active power supplied by
DLUF is calculated for all the lines that share a common
one converter equals the active power demanded by the other,
bus with the most congested line. The IPFC is placed in the
if there are no underlying storage systems
lines with maximum value of DLUF to reduce congestion
∗ ∗
and power loss in the system. A multi-objective optimization Re(V seij Iji + V seik Iki )=0 (5)
78 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 1, NO. 3, SEPTEMBER 2015

where the superscript ∗ denotes the complex conjugate. is the maximum MVA rating of the line, and MVAik is the
actual MVA rating of the line between bus i and bus k.
The flow chart for placement of IPFC is given in Fig. 3.

Start

Read the line data and bus data

Calculate the power flow and LUF of each

Fig. 1. Basic model of IPFC. line

Vseij Rank the lines in descending order on the

Iij Zseij basis of LUF
Vj
Vi
Pi+jQi Pji+jQji
Calculate the DLUF values for lines having
common bus with respect to the line ranking
Re(VseijIij−+VseijIij−) = 0 highest in congestion.

Vseik Pki+jQki
Iki Zseik Place IPFC in the lines having highest value
Vk
of DLUF

Fig. 2. Equivalent circuit of IPFC.

YES
More
IPFC
III. D ISPARITY L INE U TILIZATION FACTOR
NO
Line utilization factor (LUF) is an index used for determin-
ing the congestion of the transmission lines. It is given by Perform the load flow analysis and calculate
the LUF of lines
MVAij
LUFij = (6)
MVAmax
ij
Compare the congestion in lines before and
where LUF ij is line utilization factor of the line connected to after IPFC placement
bus i and bus j, MVAmaxij is maximum MVA rating of the line
between bus i and bus j, and MVAij is actual MVA rating of
the line between bus i and bus j. Stop

LUF gives an estimate of the percentage of line being

utilized and is an efficient method to estimate the congestion Fig. 3. Flow chart for placement of IPFC.
in a line. For placement of IPFC, there should be at least
two lines connected to a common bus. Therefore, LUF is not
sufficient for placement of IPFC. Hence, a new index disparity IV. O PTIMAL T UNING OF IPFC
line utilization factor is proposed for the optimal placement An objective function is formulated to find the optimal size
of an IPFC. The index provides an estimate of the difference of IPFC, which minimizes the active power loss, total voltage
of the percentage of line being used for the power flow. All deviations, and security margin with usage of minimum value
the lines are first ranked in descending order of their line of installed IPFC.
utilization factors. The line that is the first rank is considered
as the most congested line. DLUF is calculated for the lines
A. Objective Function
connected to the line with highest congestion. All the line
pairs connected to the same bus are ranked based on DLUF. A multi-objective function formulated is given in (8)
The line set that has highest value of DLUF is considered to X
minF = min wi fi (8)
be the optimal location for IPFC for congestion management. i=1 to 4
Assuming both lines of same rating
  where w1 , w2 , w3 , w4 are the weighting factors.
 MVAij − MVAik 
DLUF(ij)−(ik) =   (7) w1 + w2 + w3 + w4 = 1 (9)
MVAmax 
w1 = w2 = w3 = w4 = 0.25.
where, DLUF (ij)−(ik) is the disparity line utilization factor of
the line set i–j and i–k connected to bus i and bus j, MVAij is Weighting factors are used to reflect the relative importance
the MVA rating of the line between bus i and bus j, MVAmax of the objective functions. In this study, equal preference is
MISHRA et al.: CONGESTION MANAGEMENT OF POWER SYSTEM WITH INTERLINE POWER FLOW CONTROLLER USING DISPARITY LINE UTILIZATION FACTOR AND MULTI-OBJECTIVE DIFFERENTIAL EVOLUTION 79

given to all the objective functions. Hence, the value of each the overload on the transmission lines is formulated as in (15):
weight is taken as 0.25, such that the sum is equal to unity.
f4 (x) = min(P Q21 + P Q22 ) (15)
These weights will not change with the optimization process.
1) Reduction of Loss: The expression for reduction of active where P Q denotes capacity of each VSCs of IPFC
power loss [9] is given in (10) and (11),   2
2 2 Vi − V seij − Vj
lk P Q1 + P Q2 = V seij
X Zij
minf1 (x) = Ploss (10)   2
n=j,k Vi − V seik − Vk
+ V se ik . (16)
|Vi |2 Gin − |Vi | |Vn | [Gin cos θin + Bin sin θin ]
 
Zik
Ploss =
− |Vi | |Vsin | [Gin cos θsin + Bin sin θsin ]
 2
|Vn | Gin − |Vi | |Vn | [Gin cos θni + Bin sin θni ]
 B. Equality Constraints
+ (11)
− |Vn | |Vsin | [Gik cos θsin + Bik sin θsin ] Pgi + Pi − Pdi
where lk is the number of transmission lines, V i = Vi ∠θi Xn

and V n = Vn ∠θn are the voltages at the end buses i and n = Vi Vj Yij cos(θij + δj − δi ) ∀i (17)
(n = j, k); V sin = Vsin ∠θsin (n = j, k) is the series injected j=1

voltage source of nth line, s stands for series; Gin and Bin Qgi + Qi − Qdi
are the transfer conductance and susceptance between bus i Xn

and n (n = j, k), respectively. = Vi Vj Yij sin(θij + δj − δi ) ∀i. (18)

The magnitude and phase angle of the series injected voltage j=1

of V sij and V sik are determined optimally.

1) Minimization of Voltage Deviation: To have a good C. Inequality Constraints
voltage performance, the voltage deviation at each bus must Vimin ≤ Vi ≤ Vimax ∀i ∈ load bus (19)
be made as small as possible. The voltage deviation (VD) [20] max
can be expressed by (12): Sij (V, δ)| ≤ Sij ∀ij. (20)
N bus
f2 (x) = min(V D) = min(
X
|Vk − Vkref |2 ) (12) D. IPFC Constraints
k=1 V semin ≤ V se ≤ V semax (21)
where Vk is the voltage magnitude at bus k. θse min
≤ θse ≤ θse max
(22)
2) Minimization of Security Margin: This function depends
on the static voltage stability and shows whether the chance where Vslk , θslk are the injected voltage magnitude and angle,
of voltage collapse is reduced. Voltage collapse is a situation Ilk is line current magnitude through the series converter, and
when a system is unable to provide the load demand and Sslk is power injected by VSC.
is considered to be a critical state. By knowing this critical
state, the system can be protected against voltage collapse. V. R ESULTS AND D ISCUSSION
The security rate of a system according to the critical state An IEEE 30 bus test system is considered as shown in Fig.
can be expressed [21], [22] as follows in (13). 4, in which bus no. 1 is considered as a slack bus and bus
P lim P initial nos. 2, 5, 8, 11, 13 are considered as PV buses while all
Sj − Sj
j∈JL j∈J other buses are load bus. This system has 41 interconnected
SM = P limL (13) lines. The IEEE 30 bus test system load flow is obtained using
Sj
j∈JL MATLAB software and the results have been presented. Load
flow analysis including IPFC is then performed. Only load
where JL = A set contains all load buses.
buses are considered for IPFC placement. Equal weights of
SM has a value between zero and one for a system with
0.25 have been considered for all objectives. The results have
stable operating condition. SM = 0 at the voltage stability
been analyzed for normal loading, 110% loading and 125%
limit. A negative value of SM means the system cannot
loading condition.
provide the initial load. Thus, nearer the value of SM is to
LUF values of all the lines, without and with optimal
one, the system is considered to be more stable. Hence, (13)
placement and tuning of IPFC, are presented in Table I. It is
is a maximization function. Since minimization is the aim
established that line 3–4 is the most congested line connected
of the multi-objective function rather than maximization, the
between the load buses. All possible DLUF index calculation
objective function in (13) is rewritten in (14).
P initial for line 3–4 has been shown in Table II as test cases. The
Sj parameters of differential evolution algorithm used for tuning
j∈JL the IPFC have been mentioned in Table III. The results for
f3 (x, u, z) = 1 − SM = P lim (14)
Sj 110% load and 125% load have also been presented and
j∈JL
analyzed.
3) Minimization of Total Capacity of Installed IPFC: The It is observed from Table I that the line connected between
total capacity of the installed IPFC [13] required for solving buses 3–4 is the most congested line connected to the load bus.
80 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 1, NO. 3, SEPTEMBER 2015

TABLE I
LUF VALUES OF IEEE 30 B US T EST S YSTEM W ITHOUT IPFC, WITH
U NTUNED IPFC AND WITH DE T UNED IPFC

From LUF LUF with LUF with

Line To Bus
Bus Without Untuned DE Tuned
No. (RB No.)
(SB No) IPFC IPFC IPFC
1 1 2 1.8029 1.7979 1.7767
2 1 3 0.9483 0.9039 0.8940
3 2 4 0.4939 0.4768 0.4667
4 3 4 0.8415 0.8334 0.8309
5 2 5 0.8532 0.8451 0.8408
6 2 6 0.6473 0.6279 0.6201
7 4 6 0.7173 0.7439 0.7190
8 5 7 0.2539 0.2360 0.2324
9 6 7 0.3866 0.3835 0.3848
10 6 8 0.3970 0.4899 0.4570
11 6 9 0.3273 0.3042 0.3050
12 6 10 0.2479 0.2330 0.2333
13 9 11 0.4704 0.5146 0.5097
14 9 10 0.6789 0.6806 0.6781
15 4 12 0.5284 0.5028 0.5034
16 12 13 0.6929 0.7521 0.7421
17 12 14 0.1645 0.1642 0.1642
18 12 15 0.3858 0.3861 0.3862
19 12 16 0.2122 0.2133 0.2136
20 14 15 0.0317 0.0323 0.0324
21 16 17 0.0868 0.0889 0.0892
22 15 18 0.1481 0.1485 0.1486
23 18 19 0.0370 0.0378 0.0378
24 19 20 0.0651 0.0644 0.0644
Fig. 4. IEEE 30 Bus Test System with IPFC installed at line connected 25 10 20 0.1622 0.1600 0.1597
between buses 3–4 and 4–12. 26 10 17 0.1002 0.0978 0.0974
27 10 21 0.2525 0.2509 0.2505
28 10 22 0.1296 0.1308 0.1306
29 21 23 0.0571 0.0564 0.0559
In the 30 bus system, two lines are connected to line 3–4. So, 30 15 23 0.0924 0.0954 0.0957
two test cases for IPFC placement are considered, as shown 31 22 24 0.0545 0.0555 0.0554
in Table II. For each test case, DLUF is calculated, and it is 32 23 24 0.0305 0.0328 0.0328
33 24 25 0.0356 0.0318 0.0318
observed that DLUF is maximum between lines connected to 34 25 26 0.0434 0.0433 0.0433
buses 3–4 and 4–12. Hence, lines between buses 3–4 and buses 35 25 27 0.0780 0.0740 0.0740
4–12 are selected for optimal placement of IPFC. It is observed 36 28 27 0.3122 0.3013 0.3009
37 27 29 0.1158 0.1147 0.1146
that placement of IPFC at the location reduces the congestion 38 27 30 0.0890 0.0882 0.0881
in line 3–4 from 0.8415 to 0.8334. The LUF values before 39 29 30 0.0426 0.0427 0.0426
and after placement of IPFC are compared in Fig. 5. After 40 8 28 0.0799 0.0969 0.0901
41 6 28 0.2166 0.2021 0.2030
optimal tuning of IPFC using differential evolution algorithm
the congestion in the line reduces to 0.8240.
TABLE II
IPFC P LACEMENT ON THE BASIS OF DLUF
1 LUF without IPFC
0.9 Line 1 Line 2 LUF of
0.8 LUF with op!mal tuning of
LUF LUF
Sl. No. SB No– SB No– DLUF Line 1
0.7 Line 1 Line 2
IPFC using DE RB No. RB No. with IPFC
LUF (p.u.)

0.6 Case 1 3–4 4–6 0.8415 0.7173 0.1242 0.8365

0.5 Case 2 3–4 4–12 0.8415 0.5284 0.3131 0.8330
0.4
0.3
0.2 TABLE III
D IFFERENTIAL E VOLUTION PARAMETERS FOR IPFC T UNING
0.1
0 Sl. No. Parameter Value
2 5 8 11 14 17 20 23 26 29 32 35 38 41 1 Cross over probability (CR) 0.3
Line Number 2 Step size (F) 0.1

Fig. 5. LUF of all lines of IEEE 30 bus system.

It is observed from Fig. 9 that DE requires very less
Fig. 6, 7, and 8 show the objective function vs. crossover computation time, and computation time does not vary much
probability characteristics for step size = 0.1. It is observed with increase of CR. Fig. 10 compares the objective functions
that the objective minimization is better achieved with increase with respect to step size for CR = 0.1. It is observed that
in CR. However, increase in CR beyond 0.3 does not affect increase in step size increases the computation time without
the objective values much. much improvement in objective function values. Hence, the
MISHRA et al.: CONGESTION MANAGEMENT OF POWER SYSTEM WITH INTERLINE POWER FLOW CONTROLLER USING DISPARITY LINE UTILIZATION FACTOR AND MULTI-OBJECTIVE DIFFERENTIAL EVOLUTION 81

Crossover Probability is chosen to be 0.3 and step size is 0, 5.00E+01

as shown in Table III.

Objecve Funcon (p.u.)

4.00E+01
Installed IPFC capacity
Security margin Total active power loss Total acve power loss
3.00E+01 Security margin
25 21.576 21.4371 21.57 21.571
Security Margin (p.u.) and
Total Active Power Loss
Computaon me
20 15.4237 15.3378 15.4235 15.4232
2.00E+01
15
(MW)

1.00E+01
10
0.00E+00
5
0.1 0.3 0.6
0 Step Size
0.1 0.3 0.6 0.9
Crossover Probability Fig. 10. Objective functions vs. step size for CR = 0.1.

Fig. 6. Active power loss and security margin vs. CR for F = 0.1.
Thus, the voltage deviation of the overall system is reduced.
Table V shows a reduction in the values of all the objective
functions by optimal placement of IPFC. Thus, it is established
2.00E-06 1.64E-06 1.64E-061.64E-06 that by optimal tuning of IPFC using differential algorithm the
Capacity of Installed

system loss, voltage deviation, and security margin are reduced

IPFC (p.u.)

1.50E-06
with the use of minimum capacity of IPFC. Reduction in loss
1.00E-06 helps in congestion management of the system. Reduction in
2.91E-07 security margin protects the system against collapse. The lower
5.00E-07
the capacity of IPFC the lower the cost is. Hence, the overall
0.00E+00
0.1 0.3 0.6 0.9
system performance is improved at a minimum cost. Table VI
shows the effectiveness of IPFC in reduction of active and
Crossover Probability
reactive power loss.

Fig. 7. Capacity of installed IPFC vs cross probability for step size = 0.1. TABLE IV
IPFC PARAMETERS B EFORE AND A FTER T UNING FOR IEEE 30 B US
T EST S YSTEM U NDER N ORMAL L OAD C ONDITION
2.3796 2.3796 2.3795
IPFC Parameters Untuned IPFC Tuning of IPFC Using DE
2.38 Vse1 (p.u.) 0.0050 0.0011
Voltage Deviaon (p.u.)

Vse2 (p.u.) 0.0100 0.0081

2.37 Θse1 (degree) −159.8295 180
2.3566 Θse2 (degree) 180 −174.1833
2.36

2.35

2.34 1.1
0.1 0.3 0.6 0.9
Voltage Magnitude (p.u.)

1
Crossover Probability
0.9

Fig. 8. Voltage deviation vs. crossover probability for step size = 0.1. 0.8
Tuning of IPFC using DE
0.7 Without IPFC
0.6
1 11 21
45.676182
Computaon

50
Bus Number
Time (sec)

38.580945 39.732487 39.8068

40
Fig. 11. Voltage profile without and with DE tuned IPFC.
30
0.1 0.3 0.6 0.9
Crossover Probability
B. Results for 110% Loading
Simulation has been performed for 110% load on IEEE 30
Fig. 9. Computation time vs. crossover probability for step size = 0.1. bus test system. It is observed that with increase in load the
total real and reactive power loss increases. Optimal tuning
and placement of IPFC has been done to reduce loss, voltage
A. Results for Normal Loading deviation security margin and capacity of installed IPFC; the
The values of the IPFC parameters before and after tuning results are presented in Table VII, VIII, and Table IX. Fig. 12
are mentioned in Table IV. Fig. 11 shows a marked improve- shows the improvement in voltage profile with optimal tuning
ment in voltage profile of the buses with optimally tuned IPFC. and placement of IPFC.
82 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 1, NO. 3, SEPTEMBER 2015

TABLE V
R EAL P OWER L OSS , VOLTAGE D EVIATION S ECURITY M ARGIN AND 1.1

Voltage Magnitude (p.u.)

T OTAL C APACITY OF I NSTALLED IPFC WITH U NTUNED AND
WITH DE T UNED IPFC 1

0.9
Untuned Tuning of IPFC
Parameters
IPFC Using DE 0.8
Real power losses (MW) 21.909 21.4371 Tuning of IPFC using DE
Voltage deviation of all buses (p.u.) 2.3889 2.3566 0.7
Security margin of all lines (p.u.) 18.2714 15.3378 Without IPFC
Total capacity of installed IPFC (MVA) 0.000406 2.91e−7 0.6
1 11 21
Bus Number
TABLE VI
C OMPARISON OF T OTAL R EAL AND R EACTIVE P OWER L OSS IN THE Fig. 12. Voltage profile without and with DE tuned IPFC under 110% load
SYSTEM W ITHOUT IPFC, WITH U NTUNED IPFC condition for IEEE 30 bus test system.
AND WITH DE T UNED IPFC

Tuning of IPFC TABLE X

Without
Untuned IPFC PARAMETERS B EFORE AND A FTER T UNING FOR 125% L OAD
Parameters Using Differential
IPFC IPFC
Algorithm
Real power losses (MW) 22.941 21.909 21.4371 IPFC Parameters Untuned IPFC Tuning of IPFC Using DE
Reactive power losses (MVar) 107.370 101.334 100.154 Vse1 (p.u.) 0.0050 0.0010
Vse2 (p.u.) 0.0100 0.0079
Θse1 (degree) −167.9689 180
Θse2 (degree) 180 −175.0117
TABLE VII
IPFC PARAMETERS B EFORE AND A FTER T UNING FOR 110% L OAD
C ONDITION TABLE XI
C OMPARISON OF R EAL P OWER L OSS , VOLTAGE D EVIATION S ECURITY
IPFC Parameters Untuned IPFC Tuning of IPFC Using DE M ARGIN AND T OTAL C APACITY OF I NSTALLED IPFC WITH U NTUNED
Vse1 (p.u.) 0.0050 0.0012 AND DE T UNED IPFC FOR 125% L OADING C ONDITION
Vse2 (p.u.) 0.0100 0.0033
Θse1 (degree) −140.1182 180 Untuned Tuning of IPFC
Θse2 (degree) 180 −153.393 Parameters
IPFC Using DE
Real power losses (MW) 34.115 33.934
Voltage deviation of all buses (p.u.) 2.5861 2.5613
Security margin of all lines (p.u.) 21.2293 18.3804
TABLE VIII
Total capacity of installed IPFC (p.u.) 0.000245 0.00009827
C OMPARISON OF T OTAL R EAL P OWER L OSS AND R EACTIVE P OWER
L OSS W ITHOUT IPFC, WITH U NTUNED AND WITH DE T UNED
TABLE XII
Without Untuned Tuning of IPFC
Parameters C OMPARISON OF T OTAL R EAL P OWER L OSS AND R EACTIVE P OWER
IPFC IPFC Using DE
L OSS W ITHOUT IPFC, WITH U NTUNED AND WITH DE T UNED IPFC
Real power losses (MW) 27.806 26.294 26.079
Reactive power losses (MVar) 127.295 118.994 118.507
Without Untuned Tuning of IPFC
Parameters
IPFC IPFC Using DE
Real power losses (MW) 36.074 34.115 33.934
TABLE IX Reactive power losses (MVar) 160.733 149.791 149.180
C OMPARISON OF R EAL P OWER L OSS , VOLTAGE D EVIATION S ECURITY
M ARGIN AND T OTAL C APACITY OF I NSTALLED IPFC WITH U NTUNED
AND WITH DE T UNED IPFC 1.1
Voltage Magnitude (p.u.)

Untuned Tuning of IPFC 1

Parameters
IPFC Using DE
Real power losses (MW) 26.294 26.079 0.9
Voltage deviation of all buses (p.u.) 2.4609 2.4557
Security Margin of all lines (p.u.) 19.4269 16.5855 0.8
Total capacity of installed IPFC (p.u.) 0.0002437 0.000157 Tuning of IPFC using DE
0.7
Without IPFC
0.6
1 11 21
C. Results for125% Loading
Bus Number
Simulation results for 125 % load on IEEE 30 bus system
Fig. 13. Voltage profile without and with DE tuned IPFC.
listed in Table X, XI, and XII show a marked reduction in
the objective function values after optimal tuning of IPFC. It
is observed that voltage deviation, security margin, capacity
VI. C ONCLUSION
of installed IPFC, and real and reactive power loss further
decreases. Comparisons of voltages are shown in Fig. 13, A disparity line utilization factor for the optimal placement
which shows that after incorporating the IPFC in the system, of IPFC for congestion management has been proposed. The
voltage profile is improved. IPFC is accordingly placed in the lines with highest DLUF
MISHRA et al.: CONGESTION MANAGEMENT OF POWER SYSTEM WITH INTERLINE POWER FLOW CONTROLLER USING DISPARITY LINE UTILIZATION FACTOR AND MULTI-OBJECTIVE DIFFERENTIAL EVOLUTION 83

value. It is established that placement of IPFC using DLUF

effectively reduces line congestion and power loss. A multi-
objective function comprising reduction of active power loss,
minimization of total voltage deviations, and minimization
of security margin with the usage of minimum value of
installed IPFC is considered for the optimal tuning of IPFC
using differential evolution algorithm. The proposed method
is implemented for IEEE 30 bus test system. The results are
presented and analyzed under normal loading, 110% loading,
and 125% loading conditions to ascertain the effectiveness of
the proposed method on the power system performance. It is
observed that placement of IPFC by the proposed methodology
causes an effective reduction in congestion in the lines.
The results of LUF calculation before and after the compen-
sation process show reduction of loading in the congested line.
Thus, it is found that placement of IPFC at the location where
DLUF is maximum is the best location for the placement of
IPFC in terms of reduction of congestion. Simulation results
demonstrate the effectiveness and accuracy of the differential
evolution algorithm technique to achieve the multiple objec-
tives and to determine the optimal parameters of the IPFC
under different loading conditions. A reduction in real power
loss, voltage deviation, and security margin is achieved with
much smaller capacity of installed IPFC. Reduction in loss
helps in congestion management of the system. Reduction in Fig. A1. IEEE 57 bus test system with IPFC installed at line connected
between buses 14–46 and 13–14.
security margin protects the system against collapse. The lower
the capacity of IPFC, the lower is the cost. Hence, the overall
system performance is improved at a minimum cost. TABLE AI
LUF VALUES OF A LL L INES OF 57 B US T EST S YSTEM W ITHOUT AND
WITH O PTIMALLY P LACED IPFC

A PPENDIX LUF LUF with

From Bus To Bus
Line No. Without Opt. Placed
(SB No) (RB No.)
IPFC IPFC
C ASE S TUDY FOR IEEE 57 B US T EST S YSTEM 1 1 2 2.2141 1.360
2 2 3 1.0286 1.047
3 3 4 0.6947 0.693
An IEEE 57 bus test system is shown in Fig. A1 in which 4 4 5 0.3229 0.322
bus no. 1 is considered as a slack bus and bus nos. 2, 3, 6, 8, 5 4 6 0.2115 0.214
6 6 7 0.2522 0.164
9, 12 are considered as PV buses while all other buses are load 7 6 8 0.5275 0.427
buses. This system has 80 interconnected lines. LUF values of 8 8 9 1.8483 1.830
all the lines, without and with optimal placement of IPFC, are 9 9 10 0.1792 0.186
10 9 11 0.1904 0.164
presented in Table AI. It is found that line connected between 11 9 12 0.1205 0.050
buses 14–46 (line 59) is the most congested line. All possible 12 9 13 0.0897 0.046
DLUF index calculation for line 14–46 are shown in Table 13 13 14 0.3954 0.415
14 13 15 0.5226 0.624
AII as test cases. 15 1 15 1.7232 1.585
Table AI shows that line 59 is the most congested line 16 1 16 0.8385 0.834
17 1 17 0.9834 0.950
connected to load bus. In the 57 bus system, three lines 18 3 15 0.5463 0.382
have been connected to line 59. So, three test cases for IPFC 19 4 18 0.3160 0.344
placement are considered, as shown in Table AII. DLUF is 20 4 18 0.3160 0.344
21 5 6 0.8674 0.852
calculated for each test case and it is observed that congestion 22 7 8 0.9475 0.787
in line 59 is reduced most when the second line used for 23 10 12 0.2954 0.168
IPFC placement is Line 13, and where the DLUF value is 24 11 13 0.1092 0.091
25 12 13 0.8694 0.169
maximum. Hence, lines 59 and 13 have been selected for 26 12 16 0.3787 0.358
optimal placement of IPFC. It is observed from Table AII that 27 12 17 0.5314 0.477
placement of IPFC at the location where DLUF is maximum 28 14 15 0.6985 0.629
29 18 19 0.0391 0.040
reduces the congestion in line 59 from 1.23 to 0.053. Table 30 19 20 0.0079 0.006
AIII shows that after the placement of IPFC using DLUF, 31 21 20 0.0655 0.200
line congestion, line losses, and voltage deviation and security 32 21 22 0.0249 0.200
33 22 23 0.1188 0.138
margin are considerably reduced.
84 CSEE JOURNAL OF POWER AND ENERGY SYSTEMS, VOL. 1, NO. 3, SEPTEMBER 2015

TABLE AI R EFERENCES
(C ONTINUED )
[1] X. P. Zhang and L. Yao, “A vision of electricity network congestion
LUF LUF with management with FACTS and HVDC,” in The Third International
From Bus To Bus Conference on Electric Utility Deregulation and Restructuring and
Line No. Without Opt. Placed
(SB No) (RB No.) Power Technologies, IEEE PES, IET, CSEE, Nanjing, China, 2008, pp.
IPFC IPFC
116–121.
34 23 24 0.0565 0.078
[2] M. Gitizadeh and M. Kalantar, “A new approach for congestion man-
35 24 25 0.1653 0.165
agement via optimal location of FACTS devices in deregulated power
36 24 25 0.1653 0.165
systems,” in The Third International Conference on Electric Utility
37 24 26 0.6851 0.089
Deregulation and Restructuring and Power Technologies, IEEE PES,
38 26 27 0.1095 0.089
IET, CSEE, Nanjing China, 2008, pp. 1592–1597.
39 27 28 0.2021 0.183
40 28 29 0.2578 0.237 [3] F. Qian, G. Tang, and Z. He, “Optimal location and capability of
41 7 29 0.6102 0.639 FACTS devices in a power system by means of sensitivity analysis
42 25 30 0.0840 0.083 and EEAC,” in The Third International Conference on Electric Utility
43 30 31 0.0415 0.041 Deregulation and Restructuring and Power Technologies, IEEE PES IET
44 31 32 0.0257 0.026 CSEE, Nanjing China, 2008, pp. 2100–2104.
45 32 33 0.0430 0.043 [4] P. Ye, Y. Yang, T. Wang, F. Sun, and H. Zhao, “Decomposition control
46 34 32 0.0834 0.094 of UPFC for optimal congestion dispatch,” in The Third International
47 34 35 0.0913 0.093 Conference on Electric Utility Deregulation and Restructuring and
48 35 36 0.1591 0.159 Power Technologies, IEEE PES, IET, CSEE, Nanjing China, 6–9 Apr.
49 36 37 0.2597 0.193 2008.
50 37 38 0.3742 0.235 [5] M. Mandala and C. P. Gupta, “Congestion management by optimal
51 37 39 0.0395 0.041 placement of FACTS device,” in PEDES and 2010 Power India, New
52 36 40 0.0389 0.035 Delhi, India, 2010, pp. 1–7.
53 22 38 0.1433 0.157 [6] S. S. Reddy, M. S. Kumari, and M. Sydulu, “Congestion management in
54 11 41 0.0930 0.109 deregulated power system by optimal choice and allocation of FACTS
55 41 42 0.1034 0.104 controllers using multi-objective genetic algorithm,” in Transmission and
56 41 43 0.1302 0.133 Distribution Conference and Exposition, IEEE PES, New Orleans, LA,
57 38 44 0.2488 0.241 Apr. 2010, pp. 1–7.
58 15 45 0.5156 0.381 [7] N. Acharya and N. Mithulananthan, “Locating series FACTS devices
59 14 46 1.2301 0.053 for congestion management in deregulated electricity markets,” Electric
60 46 47 0.6038 0.636 Power Systems Research, vol. 77, no. 3–4, pp. 352–360, Mar. 2007.
61 47 48 0.2786 0.316 [8] J. Zhang, “Optimal power flow control for congestion management by
62 48 49 0.1220 0.119 interline power flow controller (IPFC),” in International Conference on
63 49 50 0.0755 0.123 Power System Technology 2006, Chongqing, China, 2006, pp. 1–6.
64 50 51 0.1785 0.120 [9] K. H. Mohamed, K. S. Rama Rao, and K. N. Md. Hasan, “Application
65 10 51 0.9562 0.312 of particle swarm optimization and its variants to interline power flow
66 13 49 0.3884 0.293 controllers and optimal power flow,” in 2010 International Conference
67 29 52 0.2874 0.202 on Intelligent and Advanced Systems (ICIAS), IEEE, Kuala Lumpur,
68 52 53 0.1184 0.140 Malaysia, 2010, pp. 1–6.
69 53 54 0.1310 0.097 [10] L. Kirschner, D. Retzmann, and G. Thumm, “Benefits of FACTS for
70 54 55 0.2471 0.141 power system enhancement,” in Transmission and Distribution Confer-
71 11 43 0.2405 0.162 ence and Exhibition: Asia and Pacific, IEEE/PES, Dalian, China, 2005,
72 44 45 0.3737 0.362 pp. 1–7.
73 40 56 0.0415 0.035 [11] M. A. Abdel-Moamen, and N. P. Padhy, “Optimal power flow incorporat-
74 56 41 0.0595 0.061 ing FACTS devices-bibliography and survey,” in Proceedings of 2003
75 56 42 0.0148 0.017 IEEE PES Transmission and Distribution Conference and Exposition,
76 39 57 0.0376 0.041 2003, pp. 669–676.
77 57 56 0.0320 0.030 [12] N. G. Hingorani and L. Gyugyi, Understanding FACTS: Concepts and
78 38 49 0.1835 0.054 Technology of Flexible AC Transmission System. New York, NY, USA:
79 38 48 0.3734 0.298 IEEE, 2000.
80 9 55 0.3367 0.227 [13] S. Teerthana, A. Yokoyama, “An optimal power flow control method of
power system using interline power flow controller,” in Proceedings of
IEEE Region 10 Conference TENCON 2004, Nov. 2004, pp. 343–346.
TABLE AII [14] R. Storn and K. Price, “Differential evolution—a simple and efficient
DLUF VALUE C ALCULATION FOR L INE 59 OF 57 B US T EST S YSTEM heuristic for global optimization over continuous spaces,” Journal of
Global Optimization, vol. 11, no. 4, pp. 341–359, 1997.
Line 1 Line 2 [15] R. Storn, “On the usage of differential evolution for function optimiza-
LUF LUF LUF Line 1 tion, ” in Biennial Conference of the North American Fuzzy Information
Sl. No. SB No– SB No– DLUF
Line 1 Line 2 with IPFC Processing Society (NAFIPS), Berkeley, CA, 1996, pp. 519–523.
RB No. RB No.
Case 1 14–46 46–47 1.230 0.603 0.627 0.166 [16] K. P. Wong and Z. Dong, “Differential evolution, an alternative approach
Case 2 14–46 14–15 1.230 0.698 0.532 0.160 to evolutionary algorithm,” in Proceedings of the 13th International Con-
Case 3 14–46 13–14 1.230 0.395 0.834 0.053 ference on Intelligent Systems Application to Power Systems, Arlington,
VA, 2005, pp. 73–83.
[17] E. Acha, C. Fuerte-Esquivel, H. Ambriz-Perez, and C. Angeles, FACTS:
Modeling and Simulation in Power Networks. New York: Wiley, 2004.
TABLE AIII
[18] X. P. Zhang, “Modeling of the interline power flow controller and
R EAL AND R EACTIVE P OWER L OSS , VOLTAGE D EVIATION S ECURITY
the generalized unified power flow controller in Newton power flow,”
M ARGIN W ITHOUT AND WITH IPFC FOR N ORMAL L OADING
Transmission and Distribution, vol. 150, no. 3, pp. 268–274, May 2003.
[19] M. A. Perez, A. R. Messina, and C. R. Fuerte-Esquivel, “Application
Without Optimal Placement
Parameters of FACTS devices to improve steady state voltage stability,” in Power
IPFC of IPFC
Engineering Society Summer Meeting, IEEE, vol. 2, Seattle, WA, Jul.
Real power losses (MW) 42.258 38.11
16–20, 2000, pp. 1115–1120.
Voltage deviation of all buses (p.u.) 6.4029 5.06
[20] M. Gitizadeh, “Allocation of multi-type FACTS devices using multi-
Security margin of all lines (p.u.) 25.0588 24.83
objective genetic algorithm approach for power system reinforcement,”
Reactive power loss (MVar) 166.112 146.724
Electrical Engineering, vol. 92, no. 6, pp. 227–237, 2010.
MISHRA et al.: CONGESTION MANAGEMENT OF POWER SYSTEM WITH INTERLINE POWER FLOW CONTROLLER USING DISPARITY LINE UTILIZATION FACTOR AND MULTI-OBJECTIVE DIFFERENTIAL EVOLUTION 85

[21] R. Benadid, M. Boudour, and M. A. Abido, “Optimal placement of

FACTS devices for multiobjective boltage dtability problem,” in Pro-
ceedings of IEEE Power Systems Conference and Exposition, Seattle,
WA, 2009, pp. 1–11.
[22] O. O. Obadina and G. J. Berg, “Determination of voltage stability limit
in multimachine power systems,” IEEE Transactions on Power Systems,
vol. 3, no. 4, pp. 1545–1554, Nov. 1988.

Akanksha Mishra was born in Cuttack, India, in

1982. She received her bachelor degree in electrical
engineering from Kalinga Institute of Industrial Te-
chnology, Bhubaneswar, India in 2004 and master’s
degree in power electronics and drives in 2006 from
the same institute. She is presently pursuing her Ph.D.
from Gandhi Institute of Technology and Manage-
ment, Visakhapatnam, India. Her research interests
are FACTS devices, power electronics and power
system stability. She has published several research
papers in national and international conferences.

G.V. Nagesh Kumar (M’06) was born in Visakha-

patnam, India in 1977. He received the B.E. degree
from College of Engineering, Gandhi Institute of
Technology and Management, Visakhapatnam, India
and M.E. degree from the College of Engineering,
Andhra University, Visakhapatnam. He received his
doctoral degree from Jawaharlal Nehru Technologi-
cal University, Hyderabad. He is also working as an
associate professor in the Department of Electrical
and Electronics Engineering, GITAM University,
Visakhapatnam. His research interests include gas
insulated substations, fuzzy logic, high voltage testing, and wavelets and
FACTS devices. He has published research papers in national and international
conferences and journals.