Electrical Power and Energy Systems 30 (2008) 563–568

Contents lists available at ScienceDirect

Electrical Power and Energy Systems

journal homepage: www.elsevier.com/locate/ijepes

Congestion management by determining optimal location of TCSC

in deregulated power systems
Hadi Besharat *, Seyed Abbas Taher
Department of Electrical Engineering, University of Kashan, Kashan, Iran

a r t i c l e i n f o a b s t r a c t

Article history: In a deregulated electricity market, it may always not be possible to dispatch all of the contracted power
Received 29 July 2007 transactions due to congestion of the transmission corridors. The ongoing power system restructuring
Received in revised form 18 June 2008 requires an opening of unused potentials of transmission system due to environmental, right-of-way
Accepted 15 August 2008
and cost problems which are major hurdles for power transmission network expansion. Flexible AC trans-
mission systems (FACTSs) devices can be an alternative to reduce the flows in heavily loaded lines, result-
ing in an increased loadability, low system loss, improved stability of the network, reduced cost of
Keywords:
production and fulfilled contractual requirement by controlling the power flows in the network. A
Congestion
Deregulated power system
method to determine the optimal location of thyristor controlled series compensators (TCSCs) has been
Flexible AC transmission systems (FACTS) suggested in this paper based on real power performance index and reduction of total system VAR power
Thyristor controlled series compensators losses.
(TCSC) Ó 2008 Elsevier Ltd. All rights reserved.

1. Introduction devices in electrical systems seems to be a promising strategy to

In a competitive electricity market, congestion occurs when the
transmission network is unable to accommodate all of the desired
transactions due to a violation of system operating limits. Conges-
tion does occur in both electrically bundled and unbundled sys-
tems but the management in the bundled system is relatively
simple as generation, transmission, and in some cases, distribution
systems are managed by one utility. The management of conges-
tion is somewhat more complex in competitive power markets
and leads to several disputes.
In the present day competitive power market, each utility man-
ages the congestion in the system using its own rules and guide-
lines utilizing a certain physical or financial mechanism [1].
The limitations of a power transmission network arising from
environmental, right-of-way and cost problems are fundamental
to both bundled and unbundled power systems. Patterns of gener-
ation that result in heavy flows tend to incur greater losses, and to
threaten stability and security, ultimately make certain generation
patterns economically undesirable [2,3]. Hence, there is an interest
in better utilization of available power system capacities by install-
ing new devices such as flexible AC transmission systems (FACTS).
FACTS devices by controlling the power flows in the network
without generation rescheduling or topological changes can im-
prove the performance considerably [4–6]. The insertion of such
* Corresponding author. Tel.: +98 3614460366; fax: +98 3615559930.
E-mail addresses: hadi181@yahoo.com (H. Besharat), sataher@kashanu.ac.ir
(S.A. Taher).
decrease the transmission congestion and to increase available
transfer capability. Using controllable components such as control-
lable series capacitors line flows can be changed in such a way that
thermal limits are not violated, losses minimized, stability margins
increased, and contractual requirement fulfilled, without violating
specific power dispatch. The increased interest in these devices is
essentially due to two reasons. Firstly, the recent development in
high power electronics has made these devices cost effective [7]
and secondly, increased loading of power systems, combined with
deregulation of power industry, motivates the use of power flow
control as a very cost-effective means of dispatching specified
power transactions. It is important to ascertain the location for
placement of these devices because of their considerable costs.
There are several methods for finding optimal locations of
FACTS devices in both vertically integrated and unbundled power
systems [8–12]. In [8], a sensitivity approach based on line loss
has been proposed for placement of series capacitors, phase shift-
ers and static VAR compensators. Other works in optimal power
flow with FACTS devices [9,10] have used optimization with differ-
ent objective functions. In [13,14], the optimal locations of FACTS
devices are obtained by solving the economic dispatch problem
plus the cost of these devices making the assumption that all lines,
initially, have these devices. In the presence of bilateral and multi-
lateral contracts it would be difficult to use this objective.
Congestion in a transmission system, whether vertically orga-
nized or unbundled, cannot be permitted except for very short
duration, for fear of cascade outages with uncontrolled loss of load.
Some corrective measures such as outage of congested branch,

0142-0615/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijepes.2008.08.007
564 H. Besharat, S.A. Taher / Electrical Power and Energy Systems 30 (2008) 563–568

using FACTS devices, operation of transformer taps, re-dispatch of Bus-i Z ij = rij + jxij Bus-j
generation and curtailment of pool loads and/or bilateral contracts
can relieve congestion.
A method to determine the optimal location of TCSC has been
suggested in this paper. The approach is based on the sensitivity jxc
of the reduction of total system VAR power loss and real power
performance index. In Section 2 static modeling of TCSC is ob- jBsh jBsh
tained. In Section 3 the objective function for using in OPF is pre-
sented. The optimal location is based on the minimizing the
production and device cost. The proposed method has been dem-
Fig. 2. Model of transmission line with TCSC.
onstrated on two 5-bus systems. The results show that above algo-
rithm is suitable for relieving congestion and getting economical
results. Also at the end, line outage as a contingency analysis has Bus-i Zij = rij + jxij Bus-j
been discussed.

2. Static modeling of TCSC

Fig. 1 shows a simple transmission line represented by its Sic S jc

lumped p equivalent parameters connected between bus-i and
bus-j. Let complex voltages at bus-i and bus-j are Vi \ di and Vj \ dj, Fig. 3. Injection model of TCSC.
respectively. The real and reactive power flow from bus-i to bus-j
can be written as The real and reactive power injections at bus-i and bus-j can be ex-
pressed as
Pij ¼ V 2i Gij  V i V j ½Gij cosðdij Þ þ Bij sinðdij Þ ð1Þ
Pic ¼ V 2i DGij  V i V j ½DGij cos dij þ DBij sin dij  ð11Þ
Q ij ¼ V 2i ðBij þ Bsh Þ  V i V j ½Gij sinðdij Þ  Bij cosðdij Þ ð2Þ
Pjc ¼ V 2j DGij  V i V j ½DGij cos dij  DBij sin dij  ð12Þ
where dij = di  dj. Similarly, the real and reactive power flow from
Q ic ¼ V 2i DBij  V i V j ½DGij sin dij  DBij cos dij  ð13Þ
bus-j to bus-i is
Q jc ¼ V 2j DBij þ V i V j ½DGij sin dij þ DBij cos dij  ð14Þ
Pji ¼ V 2j Gij  V i V j ½Gij cosðdij Þ  Bij sinðdij Þ ð3Þ xc r ij ðxc  2xij Þ xc ðr 2ij  x2ij þ xc xij Þ
where DGij ¼ and DBij ¼ .
Q ji ¼ V 2j ðBij þ Bsh Þ þ V i V j ½Gij sinðdij Þ þ Bij cosðdij Þ ð4Þ ðr2ij þ x2ij Þðr2ij þ ðxij  xc Þ2 Þ ðr 2ij þ x2ij Þðr2ij þ ðxij  xc Þ2 Þ

The model of transmission line with a TCSC connected between 3. Objective function
bus-i and bus-j is shown in Fig. 2. During the steady state the TCSC
can be considered as a static reactance-jxc. The real and reactive Due to high cost of FACTS devices, it is necessary to use cost-
power flow from bus-i to bus-j, and from bus-j to bus-i of a line hav- benefit analysis to analyze whether new FACTS device is cost
ing series impedance and a series reactance are [15] effective among several candidate locations where they actually
installed. The TCSC cost in line-k is given by [16]
Pcij ¼ V 2i G0ij  V i V j ðG0ij cos dij þ B0ij sin dij Þ ð5Þ
C TCSC ðkÞ ¼ c  xc ðkÞ  P 2L  Base power ð15Þ
Q cij ¼ V 2i ðB0ij þ Bsh Þ  V i V j ðG0ij sin dij  B0ij cos dij Þ ð6Þ
where c is the unit investment cost of FACTS, xc(k) is the series
Pcji ¼ V 2j G0ij  V i V j ðG0ij cos dij  B0ij sin dij Þ ð7Þ
capacitive reactance and PL is the power flow in line-k.
Q cji ¼ V 2j ðB0ij þ Bsh Þ þ V i V j ðG0ij sin dij þ B0ij cos dij Þ ð8Þ The objective function for placement of TCSC will be
X
The active and reactive power loss in the line having TCSC can be min C i ðPi Þ þ C TCSC : ð16Þ
Pi
written as i

PL ¼ Pij þ Pji ¼ G0ij ðV 2i þ V 2j Þ  2V i V j G0ij cos dij ð9Þ

4. Optimal location of TCSC
Q L ¼ Q ij þ Q ji ¼ ðV 2i þ V 2j ÞðB0ij þ Bsh Þ þ 2V i V j B0ij cos dij ð10Þ
r ðx x Þ
where G0ij ¼ r2 þðx ijx 2 and B0ij ¼ r2 þðxij xc Þ2 . 4.1. Reduction of total system VAR power loss
ij ij cÞ ij ij c

The change in the line flow due to series capacitance can be rep-
resented as a line without series capacitance with power injected
at the receiving and sending ends of the line as shown in Fig. 3.
Bus-i Yij = Gij + jB ij
Here, we look at a method based on the sensitivity of the total
system reactive power loss with respect to the control variable of
the TCSC. For TCSC placed between buses i and j we consider net
line series reactance as a control parameter. Loss sensitivity with
Bus-j respect to control parameter of TCSC placed between buses i and
j can be written as

oQ L r 2ij  x2ij
aij ¼ ¼ ½V 2i þ V 2j  2V i V j cos dij   ð17Þ
oxij ðr2ij þ x2ij Þ2

4.2. Real power flow performance index sensitivity indices

jBsh jB sh
The severity of the system loading under normal and contin-
gency cases can be described by a real power line flow perfor-
Fig. 1. Model of transmission line. mance index [17], as given below
 H. Besharat, S.A. Taher / Electrical Power and Energy Systems 30 (2008) 563–568 565

XNL  2n Table 1

wm PLm
PI ¼ max ð18Þ Bid prices of generators
m¼1
2n PLm
Generator Bid prices ($/h) Pimin Pimax
where PLm is the real power flow and Pmax Lm is the rated capacity of 1 0:11P 21þ 5P 1 þ 150 15 180
line-m, n is the exponent and wm a real non-negative weighting 2 0:085P 22 þ 1:2P 2 þ 60 20 200
coefficient which may be used to reflect the importance of lines. 3 0:1225P 23 þ P 3 þ 335 15 300
PI will be small when all the lines are within their limits and
reach a high value when there are overloads. Thus, it provides a
good measure of severity of the line overloads for given state of about the optimal location for those devices in a deregulated envi-
the power system. Most of the works on contingency selection ronment. These factors as mentioned in the previous section, can
algorithms utilize the second order performance indices which, be computed for a base-case power flow solution. The FACTS de-
in general, suffer from masking effects. The lack of discrimination, vice should be placed on the most sensitive lines. With the sensi-
in which the performance index for a case with many small viola- tivity indices computed for TCSC, following criteria can be used
tions may be comparable in value to the index for a case with one for its optimal placement:
huge violation, is known as masking effect. By most of the opera-
tional standards, the system with one huge violation is much more (a) In reactive power loss reduction method, TCSC should be
severe than that with many small violations. Masking effect to placed in a line having the most positive loss sensitivity
some extent can be avoided using higher order performance indi- index.
ces, that is n > 1. However, in this study, the value of exponent (b) In PI method, TCSC should be placed in a line having most
has been taken as 2 and wi = 1. negative sensitivity index.
The real power flow PI sensitivity factors with respect to the
parameters of TCSC can be defined as

oPI  5. Simulation results
bk ¼ ð19Þ
oxck xck ¼0
5.1. Determining optimal location of TCSC
The sensitivity of PI with respect to TCSC parameter connected be-
tween bus-i and bus-j can be written as The approach has been examined on two 5-bus systems. MAT-
X
NL  4 POWER, a toolbox of MATLAB, has been used for simulations [18].
oPI 1 oPLm
¼ wm P3Lm max ð20Þ One of the approach solvers for OPF in MATPOWER is based on lin-
oxck m¼1 PLm oxck ear programming. The application of linear programming tech-
niques is desirable, due to their robustness, high speed and high
The real power flow in a line-m can be represented in terms of real
dimension of the matrix problems that they can handle. The bid
power injections using DC power flow equations [17] where s is
prices by generators for each 5-bus system are given in Table 1
slack bus, as
8 where P is in MW and $ is a monetary unit which may be scaled
>
> P
N
by any arbitrary constant without affecting the results, and Pimin
>
> Smn Pn for m–k
>
> and Pimax, are the generation power limits of each generator.
>
> n¼1
>
> The first 5-bus system is shown in Fig. 4. Bus-1 has been taken
< n–s
PLm ¼ ð21Þ as a reference bus.
> P
> N
The load flow of first 5-bus system is shown in Table 2. From the
>
> Smn Pn þ Pj for m ¼ k
>
> load flow, it was found that real power flow in line 2–5 was
>
> n ¼ 1
>
> 1.034 pu which is more than its line loading limit.
:
n–s
The sensitivities of reactive power loss reduction and real
Using Eq. (21), the following relationship can be derived, power flow performance index with respect to TCSC control
8  parameter has been computed and are shown in Table 3. The sen-
> oP
< Smi oxoPi þ Smj ox j for m–k sitive line in each case is presented in bold type. It can be observed
oPLm ck ck
¼   ð22Þ from Table 3 (column 3) that placement of TCSC in line-3 is suit-
oxck >
: Smi oxoPi þ Smj oxoPj þ oxoPj for m ¼ k
ck ck ck able for reducing the total reactive power loss. System power flow
  result after placing TCSC in line-3 is shown in Table 4. The value of
oP i  oP j 
The terms oxck 
and oxck 
can be derived as below control parameter of TCSC for computing power flow is taken as
 xck ¼0 xck ¼0
oPi  oP ic  0.2885 pu. It can be observed from Table 4 that congestion has
oxck 
¼ oxck 
xck ¼0 xck ¼0 been relieved. Placement of TCSC in line-1 also will reduce the total
ðx2 r 2 Þ
ð23Þ
r x
¼ 2ðV 2i  V i V j cos dij Þ ðr2 þx
ij ij ij ij
2 Þ2  V i V j sin dij ðr 2 þx2 Þ2
ij ij ij ij
 
oP j  oP 
ox 
¼ ox jc 
ck xck ¼0 ck xck ¼0
ð24Þ
r x ðx2 r 2 Þ 1 2 3
¼ 2ðV 2j ij ij
 V i V j cos dij Þ ðr2 þx ij ij
2 Þ2 þ V i V j sin dij ðr 2 þx2 Þ2
ij ij ij ij

4.3. Criteria for optimal location of TCSC

The location of FACTS devices can be based on static or dynamic

performance of the system. The sensitivity factor methods are gen-
5
erally used to find the best location to enhance the static perfor-
4
mance of the system. The sensitivity for determining optimal
location of Facts devices can at best give an approximate idea Fig. 4. First 5-bus system.
566 H. Besharat, S.A. Taher / Electrical Power and Energy Systems 30 (2008) 563–568

Table 2 Table 6
Power flow result of first 5-bus system Total cost for optimal location of TCSC in first system

Line ij Power flow (pu) Method Total cost

1 2–1 0.07798 VAR reduction 2250.11
2 2–5 1.034 PI 2276.78
3 3–5 0.08441
4 5–4 0.40379
5 1–4 0.4145
6 3–2 0.51559

1 3 4

Table 3
Calculated sensitivity indices

Line ij aij bij

1 2–1 0.008057 0.0789
2 2–5 0.970852 1.95327
3 3–5 0.00784 0.10536 5
4 5–4 0.261704 0.34953 2
5 1–4 0.967394 0.41433
6 3–2 0.240349 0.45582 Fig. 5. Second 5-bus system.

Table 4 Table 7
Power flow result of first 5-bus system after placing TCSC in line-3 Power flow result of second 5-bus system

Line ij Power flow (pu) Line ij Power flow (pu)
1 2–1 0.07614 1 1–2 1.0181
2 2–5 0.99956 2 1–3 0.48796
3 3–5 0.08441 3 2–3 0.43934
4 5–4 0.40379 4 2–4 0.33076
5 1–4 0.41123 5 2–5 0.76539
6 3–2 0.47879 6 3–4 0.1133
7 4–5 0.14286

system reactive power loss but it will be less effective than placing
a TCSC in line-3 as can be seen from its sensitivity factors. Table 8
It can be observed from Table 3 (column 4) that placing a TCSC Calculated sensitivity indices
in line-5 is optimal for reducing the PI and congestion relief.
Line ij aij bij
System power flow result after placing TCSC in line-5 is shown in
Table 5. The value of control parameter of TCSC for computing 1 1–2 1.20822 3.45
2 1–3 0.19303 1.11
power flow is taken as 0.0423 pu. It can be observed from Table
3 2–3 0.18757 0.609
5 that congestion has been relieved. 4 2–4 0.10456 0.124
Placement of TCSC on line-3 will reduce the PI value but it will 5 2–5 0.60931 1.39
be less effective than placing a TCSC in line-5 as can be seen from 6 3–4 0.05629 0.15
7 4–5 0.0368 0.18
its sensitivity factors. Total costs of two methods are shown in
Table 6. It can be observed from Table 6 that reduction of total sys-
tem VAR power loss method is more economical than PI method
for placing the TCSC and congestion management. is presented in bold type. It can be observed from Table 8 (column
The second 5-bus system is shown in Fig. 5. Bus-1 has been ta- 3) that placement of TCSC in line-6 will reduce the total system
ken as a reference bus. VAR loss but it will be less effective than placing a TCSC in line-7
The load flow of second 5-bus system is shown in Table 7. From as can be seen from its sensitivity factors. System power flow result
the load flow, it was found that real power flow in line 1–2 was after placing TCSC in line-7 is shown in Table 9. The value of con-
1.0181 pu which is more than its line loading limit. The sensitivi- trol parameter of TCSC for computing power flow is taken as
ties of reactive power loss reduction and real power flow perfor- 0.17815 pu. It can be observed from Table 9 that congestion has
mance index with respect to TCSC control parameter has been been relieved. From the calculated sensitivity factors bij of Table
computed and are shown in Table 8. The sensitive line in each case

Table 9
Table 5 Power flow result of second 5-bus system after placing TCSC in line-7
Power flow result of first 5-bus system after placing TCSC in line-5
Line ij Power flow (pu)
Line ij Power flow (pu)
1 1–2 0.99956
1 2–1 0.10893 2 1–3 0.50718
2 2–5 0.99956 3 2–3 0.48785
3 3–5 0.08798 4 2–4 0.37522
4 5–4 0.37453 5 2–5 0.65385
5 1–4 0.46051 6 3–4 0.17992
6 3–2 0.51202 7 4–5 0.25336
 H. Besharat, S.A. Taher / Electrical Power and Energy Systems 30 (2008) 563–568 567

Table 10 power flow of line 2–5 will be 0.99956 pu and congestion

Power flow result of second 5-bus system after placing TCSC in line-2 will be relieved. Therefore, by setting the installed TCSC in
Line ij Power flow (pu) line 3–5 congestion has been relieved.
1 1–2 0.99956 (2) Outage of line 2–5
2 1–3 0.50751 System power flow by opening line 2–5 is shown in Table 12
3 2–3 0.42892 (column 3) in which no congestion can be observed.
4 2–4 0.32527 (3) Outage of line 5–4
5 2–5 0.76276
6 3–4 0.12135
System power flow by opening line 5–4 is shown in Table 12
7 4–5 0.14546 (column 4). From this table, it can be observed that conges-
tion has not been occurred.
(4) Outage of line 1–4
System power flow by opening line 1–4 is shown in Table 12
Table 11
Total cost for optimal location of TCSC in second system
(column 5). From this table, it can be observed that by open-
ing line 1–4, line 2–5 has been congested. Now if we set the
Method Total Cost control parameter of TCSC, xc, in line 3–5 to 0.774045 pu
VAR reduction 5040.51 then power flow of line 2–5 will be 0.99956 pu and conges-
PI 4929.44 tion will be relieved.
(5) Outage of line 3–2
System power flow by opening line 3–2 is shown in Table 12
8 (column 4) it can be observed that placement of TCSC in line-7 (column 6). From this table, it can be observed that conges-
will reduce the PI but it will be less effective than placing a TCSC tion has not been occurred.
in line-2. System power flow result after placing TCSC in line-2 is
shown in Table 10. The value of control parameter of TCSC for com-
puting power flow is taken as 0.014315 pu. It can be observed from 6. Conclusion
Table 10 that congestion has been relieved. Total costs of two
methods are shown in Table 11. It can be observed from Table 11 Congestion management is an important issue in deregulated
that PI method in this case is more economical than reduction of power systems. FACTS devices such as TCSC by controlling the
total system VAR power loss method for installing the TCSC and power flows in the network can help to reduce the flows in heavily
congestion relief. loaded lines. Because of the considerable costs of FACTS devices, it
is important to obtain optimal location for placement of these
5.2. Single line outage as a contingency analysis devices.
In this paper, two sensitivity-based methods have been devel-
In a power system, if a line is corrupted, its power flow will be oped for determining the optimal location of TCSC in an electricity
shared among other lines of the system. This will lead to possible market. In a system, first two optimal locations of TCSC can be
overloading of some of the lines. decided based on the sensitivity factors aij and bij and then optimal
Here, we consider first 5-bus system and study the effect of line location is selected based on minimizing production cost plus de-
outage on this network. First we consider the case that by using vice cost. Test results obtained on two 5-bus power systems show
reactive power loss reduction method (the economical method that sensitivity factors along with TCSC cost could be effectively
for first 5-bus system) where installing TCSC in line 3–5, conges- used for determining optimal location of TCSC.
tion has been relieved. System power flow of this case was shown The effect of TCSC on line outage in order to relive congestion
in Table 4. Then by opening each of the lines of the system, we con- has also been studied. It can be observed from the results of line
sider the effect of opened line on remaining of the system. If there outage that we can relieve congestion by setting the installed TCSC.
is still congestion in the network, then we try to set the installed
TCSC in such a way that congestion is relived. If congestion still Appendix A. First 5-bus system
persists in the system, we shall install a new TCSC by using reactive
power loss reduction method or open the congested line/s. In this The two lines 1–2 and 3–5 are of impedance 0.0258 + j0.866 pu
section we consider these five cases: each while other four lines have an impedance of 0.0129 +
j0.0483 pu each, all to a 100 MW base. Bus-1 has been taken as a
(1) Outage of line 2–1 reference bus. The line flow limit is set to 100 MW.
System power flow by opening line 2–1 is shown in Table 12.
From Table 12 (column2), it is found that by opening line 2–
Appendix B. Second 5-bus system
1, line 2–5 has been congested. Now if we set the control
parameter of TCSC, xc, in line 3–5 pu to 0.5283 pu then
The second 5-bus system is shown in Fig. 5. System line data is
shown in Table B. Bus-1 has been taken as a reference bus. The line
flow limit is set to 100 MW.
Table 12
Table B
Power flow result of first 5-bus system after single line outage
Second 5-bus system line data
Line i  j Power flow (pu)
ij rij xij
Case 1 Case 2 Case 3 Case 4 Case 5
1–2 0.0023 0.06
2–5 1.0701 – 0.62334 1.431 0.54894 1–3 0.063 0.24
3–5 0.1269 0.6974 0.08611 0.13453 0.6 2–3 0.0011 0.12
5–4 0.4813 0.0164 – 0.81648 0.43405 2–4 0.0017 0.18
1–4 0.3378 0.8547 0.81263 – 0.38409 2–5 0.012 0.12
3–2 0.4731 0.0974 0.51389 0.46457 – 3–4 0.012 0.03
1–2 – 0.5025 0.48699 0.36839 0.05106 4–5 0.021 0.24
568 H. Besharat, S.A. Taher / Electrical Power and Energy Systems 30 (2008) 563–568

References [10] Momoh JA, Zhu JZ. A new approach to optimal power flow with phase shifter.
In: IEEE international conference on systems, vol. 5; 1998. p. 4794–9.
[11] Wu G, Yokoyama A, He J, Yu Y. Allocation and control of FACTS devices for
[1] Kumar Ashwani, Srivastava SC, Singh SN. A zonal congestion management
steady state stability enhancement of large scale power system. In: IEEE
approach using real and reactive power rescheduling. IEEE Trans Power Syst
international conference on power system technology, vol. 1; 1998. p. 357–
2004;19(1):554–62.
61.
[2] Vries LJ. Capacity allocation in a restructured electricity market: technical and
[12] Liu JY, Song YH. Comparison studies of unified power flow controller with
economic evaluation of congestion management methods on interconnectors.
static var compensators and phase shifters. Electric Machines Power Syst
In: Proceedings of the 2001 IEEE Porto power tech conference.
1999;27:237–51.
[3] Lommerdal M, Soder L. Simulation of congestion management methods. In:
[13] Lie TT, Deng W. Optimal flexible AC transmission systems (FACTS) devices
Proceedings of the 2003 Bologna power tech.
allocation. Int J Electrical Power Energy Syst 1999;19(2):125–34.
[4] Baldick R, Kahn E. Contract paths, phase shifters and efficient electricity trade.
[14] de Oliveira EJ, Lima WM. Allocation of FACTS devices in a competitive
IEEE Trans Power Syst 1997;12(2):749–55.
environment. In: 13th PSCC; 1999. p. 1184–90.
[5] Bladow J, Montoya A. Experiences with parallel EHV phase shifting
[15] Verma KS, Singh SN, Gupta HO. FACTS devices location for enhancement of
transformers. IEEE Trans Power Deliv 1991:1096–100.
total transfer capability. In: Power engineering society winter meeting, IEEE,
[6] Galiana GD. Assessment and control of the impact of FACTS devices on power
vol. 2; 2001. p. 522–7.
system performance. IEEE Trans Power Syst 1996;11(4):1931–6.
[16] Singh SN, David AK. Placement of FACTS devices in open power market. Adv
[7] Larsen E, Millers N, Nilsson S, Lindgren S. Benefits of GTO-based compensation
Power Syst Control Operation Manage 2000;1:173–7.
systems for electric utility applications. IEEE Trans Power Deliv 1992;7(4):
[17] Wood AJ, Wollenberg BF. Power generation. Operation and control. New
2056–64.
York: John Wiley; 1996.
[8] Preedavinchit P, Srivastava SC. Optimal reactive power dispatch considering
[18] MATPOWER, a MATLAB power system simulation package. Version 3.0.0.
FACTS devices. Electric Power Syst Res 1998;46(3):251–7.
http://www.pserc.cornell.edu/matpower.
[9] Huang G, Hsieh SC. Fast textured algorithms for optimal delivery problems in
deregulated environments. IEEE Trans Power Syst 1998;13(2):493–500.