Power Flow Control with Distributed Flexible

AC Transmission System (D-FACTS) Devices

Katherine M. Rogers, Student Member, IEEE, Thomas J. Overbye, Fellow, IEEE


Abstract—Distributed Flexible AC Transmission System (D-
FACTS) devices offer many potential benefits to power system
operations. This paper illustrates the flexibility of control that is
achievable with D-FACTS devices. The impact of installing D-
FACTS devices is examined by studying the sensitivities of power
system quantities such as voltage magnitude, voltage angle, bus
power injections, line power flows, and real power losses with
respect to line impedance. Sensitivities enable us to quantify the
amount of control D-FACTS devices offer to the system.
Independently controllable lines are selected for power flow
control, and appropriate locations to install D-FACTS devices for
line flow control are determined. Then, D-FACTS device settings
are selected to achieve desired line flow objectives.
Index Terms— power flow control, distributed flexible AC
transmission systems, controllability, linear sensitivity analysis
I. INTRODUCTION
weight, and made of easily purchased mass-produced parts.
The achievement of flexible line flow control through the use
of effectively placed and configured D-FACTS devices is
explored in this paper.
II. POWER FLOW CONTROL
Control of power flow requires the ability to maintain or
change line impedances, bus voltage magnitudes, and phase
angle differences. Power controller devices such as FACTS
devices [6], [7], [8] affect some or all of these parameters.
D-FACTS devices, the Distributed Static Series
Compensator (DSSC) in particular, are series power flow
control devices which change the effective impedance of
transmission lines through the use of a synchronous voltage
source (SVS) [9]. A D-FACTS device changes the effective
line impedance by producing a voltage drop across the line

A PPROXIMATELY two decades ago, Flexible AC

Transmission Systems (FACTS) were introduced. A
Flexible AC Transmission System incorporates power
electronics and controllers to enhance controllability and
increase transfer capability [1]. FACTS devices can improve
power system operation are by providing a means to control
power flow, to improve stability, and to better utilize the
existing transmission infrastructure. The benefits associated
with the use of FACTS devices have been demonstrated in
successful applications. One such application is in west Texas
where dynamic reactive compensation systems (DRCS)
correct abnormal voltages caused by the rapid changes in wind
production which is prevalent in the area [2].
Over the past two decades, technology in many areas of
electrical engineering has become faster, less expensive,
smaller, and more reliable. Advances in computing, wireless
communications, microprocessors, electronic devices, and
other technology advances have affected all aspects of life.
Improvements in electrical technology allow a revisit of
FACTS concepts from a fresh perspective.
Recently, Distributed Flexible AC Transmission System
(D-FACTS) devices [3], [4], [5] were introduced. D-FACTS
devices are power flow control devices which are small, light-
which is in quadrature with the line current. Thus, a D-
FACTS device provides either purely reactive or purely
capacitive compensation.
D-FACTS devices can be made to communicate wirelessly,
allowing them to receive commands for impedance injection
changes. In addition, D-FACTS devices can be configured to
operate autonomously during certain conditions [4]. D-FACTS
devices on different lines working together can be coordinated
in such a way that they can achieve some control objective.
Consider the simple 4-bus example below. D-FACTS
devices are first placed on line (1, 3) and then added to line
(2,4). The range of possible line flows due to D-FACTS
control is plotted in Figure 2 for each line in the system. It is
assumed that D-FACTS devices can change the line
impedance by +/-20% [4] of the uncompensated value.

The authors would like to acknowledge the support of the support of NSF
through its grant CNS-0524695, the Power System Engineering Research
Center (PSERC), City Water Light and Power (CWLP) in Springfield, IL, and
the Grainger Foundation.

The authors are with the University of Illinois Urbana-Champaign, Urbana, IL

61801 (e-mail: krogers6@illinois.edu; overbye@illinois.edu). Figure 1. 4-Bus System

Authorized licensed use limited to: Bharat University. Downloaded on June 17,2010 at 07:33:34 UTC from IEEE Xplore. Restrictions apply.

Figure 2. Line Flow Control of 4-bus System
A certain range of flows is achievable by D-FACTS control
of line (1,3). When D-FACTS are added to a second line, line
(2, 4), more control is possible. However, lines do not have
equal potential to provide control or to be controlled.
D-FACTS devices control of one line affects the flows on
all lines. The impact that the control of one line flow has on
other line flows is specific to the system. If a system has only
one loop, the flows are completely coupled and cannot be
controlled independently. The extent to which line flows can
be controlled independently is one measure of D-FACTS
potential. A method for identification of independently
controllable line flows is presented in Section IV. D-FACTS
devices should be placed at effective locations throughout the
system and set to achieve the desired control; this control is
described in Section V.
III. LINE IMPEDANCE SENSITIVITIES
Sensitivities are linearized relationships often used in power
systems analysis [10]. Sensitivities reveal the impact of a
small change in a variable on the rest of the system. Linear
approximations in nonlinear systems provide insight into
relationships which may otherwise be difficult to characterize.
Line impedance sensitivities are fundamental to the analysis of
placing and setting D-FACTS devices for line flow control.
The equations from which the sensitivities are derived in are
given in Appendix A.
A. Power Injection and State Variable Sensitivities
The negative inverse of the power flow Jacobian, J,
describes the way the state variables change in a solution of
the power flow due to bus power injection mismatch.
−1
∆𝒔(𝜃,𝑉) = −𝑱 ∙ 𝒇(𝑝,𝑞)
The power injection to impedance sensitivity matrix 𝚼 is the
derivative of the entries of f(p,q) with respect to impedance:
∆𝒇(𝑝,𝑞) = 𝚼 ∙ ∆𝒙
Authorized licensed use limited to: Bharat University. Downloaded on June 17,2010 at 07:33:34 UTC from IEEE Xplore. Restrictions apply.
The state to impedance sensitivity matrix Φ describes how
the state variables change in a solution of the power flow due
to a small impedance change:
𝚽 = −𝑱−1 ∙ 𝚼 (3)
∆𝒔(𝜃,𝑉) = 𝚽 ∙ ∆𝒙 (4)
The matrix Φ is the only full matrix needed, and its
computation involves 𝚼 and the inverse of J. The dimension
of the columns of 𝚼 is the number of lines equipped with D-
FACTS devices, k. The rows of 𝚼 are sparse since not every
bus is connected to each of the k lines. Thus, each column of
𝚼 is a sparse vector, and sparse vector methods may be used to
compute Φ using the fast-forward and full back schemes as
described in [11].
B. Power Flow Sensitivities
The relationships between state variables and real power
flows are represented by the power flow to state sensitivity
matrix Σ:
∆𝑷𝒇𝒍𝒐𝒘 = 𝚺 ∙ ∆𝒔(𝜃 ,𝑉) (5)
Each row in the power flow to impedance sensitivity matrix
Γ has one non-zero element corresponding to the line
impedance on the same line. Since a line has a sending end
and a receiving end power flow, each column in Γ has two
non-zero elements.
∆𝑷𝒇𝒍𝒐𝒘 = 𝚪 ∙ ∆𝒙 (6)
C. Loss Sensitivities
The sensitivity of losses (24) to real power line flows Τ is a
row vector of all ones with dimension of twice the number of
lines in the system, also the dimension of Pflow.
∆𝑃𝑙𝑜𝑠𝑠 = 𝚻 ∙ ∆𝑷𝒇𝒍𝒐𝒘 (7)
The aforementioned sensitivities define the complete
relationship between system losses and the reactive part of line
impedance. The total sensitivity of losses to line impedance Κ
has dimension equal to the number of lines with D-FACTS
devices:
𝚱=𝚻 𝚺∙𝚽+𝚪 (8)
∆𝑃𝑙𝑜𝑠𝑠 = 𝐊 ∙ ∆𝒙 (9)
The elements in Κ give the change in system losses due to a
small change in x and a solution of the power flow. Solution
of the power flow equations is important to consider.
Otherwise, our ability to analyze the impact of impedance-
changing devices would be limited to the direct sensitivities of
real power losses to line impedance found from Τ·Γ.
Including indirect sensitivities Τ·Σ·Φ in the analysis allows
representation of the impact of lines on other lines. The total
sensitivity representation (8) allows us to consider the use of a
D-FACTS device to provide control not only for the line on
which it is placed but also for other lines in the system.
(1)
IV. CONTROL POTENTIAL OF D-FACTS DEVICES
For any power system, it is useful to be able to determine
(2) the control potential available from D-FACTS devices.
Analysis of the control of power systems with FACTS devices independently.
[12], [13], [14] has been examined, but primarily with respect Consider the 4-bus example. The sensitivity matrix
to transient stability, where FACTS devices can be used for 𝚺 ∙ 𝚽 + 𝚪 for the system is given as follows,
control of certain modes of the system. 𝑥(1,2) 𝑥(1,3) 𝑥(2,3) 𝑥(2,4) 𝑥(3,4)
In this work, we are interested in the ability of D-FACTS 𝑃𝑓𝑙 (1,2) 1.23 1.43 −1.29 −1.84 −0.13
devices to provide control over line flows throughout the 𝑃𝑓𝑙 (1,3) −1.22 −1.42 1.24 1.88 0.13
system. When effective line impedances change, power flows (12)
𝑃𝑓𝑙 (2,3) 0.63 0.73 −3.77 3.67 0.27
redistribute in the system. Our perspective is to show through
steady-state analysis the ability of D-FACTS devices to 𝑃𝑓𝑙 (2,4) 0.63 0.74 2.46 −5.55 −0.41
control the way power flows distribute throughout the system. 𝑃𝑓𝑙 (3,4) −0.60 −0.71 −2.37 5.35 0.39

A. Identification of Independently Controllable Line Flows where the sending and receiving end line flows of a line
In some scenarios, it may be clear which lines need to be cannot be independently controlled, so they are not both
targeted for control. The need to operate the system securely shown. The coupling indices are given by the following
is costly but crucial. D-FACTS devices can be used to relieve symmetric matrix:
a known overloaded element such as a line or transformer. 𝑃𝑓𝑙 (1,2) 𝑃𝑓𝑙 (1,3) 𝑃𝑓𝑙 (2,3) 𝑃𝑓𝑙 (2,4) 𝑃𝑓𝑙 (3,4)
The ability to relieve an overloaded element through the use of 𝑃𝑓𝑙 (1,2) 1.00 −1.00 −0.01 0.50 −0.50
D-FACTS control is by itself a strong advantage. Since an 𝑃𝑓𝑙 (1,3) −1.00 1.00 0.03 −0.51 0.51
overloaded line or transformer can prevent many power 𝑃𝑓𝑙 (2,3)
(13)
−0.01 0.03 1.00 −0.87 0.87
transfers from being able to take place, reducing the flow 𝑃𝑓𝑙 (2,4) 0.50 −0.51 −0.87 1.00 −1.00
through the overloaded element by even a few percent
𝑃𝑓𝑙 (3,4) −0.50 0.51 0.87 −1.00 1.00
improves the operation of the power grid.
From a broader perspective, D-FACTS devices can be used The coupling indices indicate that the flow on lines (1,2) and
throughout the system to provide the most comprehensive (1,3) cannot be independently controlled. Similarly, the flows
control. In order to provide the most complete and effective on (2,4) and (3,4) cannot be independently controlled. On the
control for the entire system, it is necessary to identify how other extreme, the coupling index between lines (1,2) and
the control of line flows are related to each other. (2,3) is very small which indicates a high ability to be
The coupling of the control of line flows is important to independently controlled.
understand so that money and control effort are not wasted in The coupling indices provide insight into which line flows
attempts to independently control line flows which are highly should be targeted for control. For the 4-bus system, we will
coupled. The following matrices show trivial cases where choose to target lines (1,3) and (2,3) for control since they
control of line flows are completely decoupled (a) and have a small coupling index. The results of this control are
decoupled (b): given in Section V.
𝐚. 𝑥1 𝑥2 𝑥3 𝐛. 𝑥1 𝑥2 𝑥3 B. Identification of Effective D-FACTS Locations
𝑃𝑓𝑙 ,1 1 0 0 𝑃𝑓𝑙 ,1 1 1 1 D-FACTS devices are unique because they are well-suited
𝑃𝑓𝑙 ,2 𝑃𝑓𝑙 ,2 (10)
0 2 0 2 2 2 to be placed at multiple locations in the system where their use
𝑃𝑓𝑙 ,3 0 0 1 𝑃𝑓𝑙,3 1 1 1 could be the most beneficial. Comparatively, if only one
FACTS device is used, all support goes to the same place.
In the completely decoupled case, the vectors are orthogonal
However, reactive power support is most effective locally.
and the angle between them is exactly 90 degrees. In the
Sensitivities can be used to identify lines with a high impact
completely coupled case, the row vectors are perfectly aligned
for particular applications. Lines with higher sensitivities are
and the angle between them is exactly zero degrees. When the
able to provide more control, whereas lines with sensitivities
row vectors are perfectly aligned but point in opposite
of zero have no impact. The locations for D-FACTS devices
directions, the angle between them is 180 degrees, but they are
are found by determining the lines with the highest
still completely coupled. Thus, coupling can be determined by
sensitivities for the objective.
comparing the cosine of angles of vectors [15].
Consider again the 4-bus system and its line flow
The cosine of the angle between two row vectors v1 and v2,
sensitivities given by (12). If the objective is to control the
𝒗𝟏 ∙ 𝒗𝟐 flow of line (1,3), the relevant sensitivities are the second row
𝑐𝑜𝑠𝜃𝒗𝟏 𝒗𝟐 = (11)
𝒗𝟏 𝒗𝟐 of (12):
𝑥 1,2 𝑥 1,3 𝑥 2,3 𝑥 2,4 𝑥 3,4
of the total power flow to impedance sensitivity matrix
𝑃𝑓𝑙 1,3 −1.22 −1.42 1.24 1.88 0.13 (14)
𝚺 ∙ 𝚽 + 𝚪 will be called the coupling index. The coupling
index has values between -1 and 1. When the coupling index In (14), the greatest impact on flow (1,3) comes from line
has an absolute value of 1, there is complete correlation, either impedance (2,4); the next most effective line impedance is line
positive or negative, between the ways the two line flows (1,3), etc. For controlling multiple line flows, D-FACTS
respond to D-FACTS control. When the coupling index is locations are similarly determined by finding the highest
zero, the line flows have the ability to be controlled sensitivities of the control objective to line impedance.

Authorized licensed use limited to: Bharat University. Downloaded on June 17,2010 at 07:33:34 UTC from IEEE Xplore. Restrictions apply.
 V. TRANSMISSION LINE POWER FLOW CONTROL A. 4-bus Test System
Once appropriate lines are targeted for control and effective The results of control in the 4-bus case are presented here.
locations for D-FACTS devices are selected, the problem of The approach described above is used to determine the D-
power flow control needs to be solved. The goal of the FACTS device settings. Lines (1,3) and (2,3) are first targeted
problem can be stated as a desire to attain specified line flows for control; their flow coupling index is 0.03. D-FACTS
on any number of independently controllable lines through the devices are placed on lines (1,3), (2,3), and (2,4). The results
control of line impedance settings of D-FACTS devices on a for the four possible control scenarios are shown in Table I. In
specified number of lines. all cases, the control objective is achieved.
It is not always possible to achieve a specified power flow Table I. Decoupled Line Flow Control
on a line, so the line flow control equation, Pflow,calc(x) = Control Case 1
Objective: Lower Flows on (1,3) and (2,3)
Pflow,spec(x) does not always have a solution. This is acceptable
because line flow control is merely an additional benefit. The Original Flows New Flows
Line (1,3): 40.69 MW Line (1,3): 37.82 MW
level of importance of a solution of the power balance Line(2,3): 80.17 MW Line (2,3): 66.88 MW
equations is much higher than the line flow control equations.
Control Case 2
For any power system application, the power balance Objective: Raise Flows on (1,3) and (2,3)
equations f(p,q) must always be satisfied, but if some control
Original Flows New Flows
over the power flow on a line can be achieved, that can be Line (1,3): 40.69 MW Line (1,3): 42.53 MW
done as well. Line(2,3): 80.17 MW Line (2,3): 93.97 MW
Optimization methods are useful for problems that do not Control Case 3
have a solution [16]. The line flow control problem can be Objective: Raise Flow on (1,3) and Lower Flow on (2,3)
examined in an optimization framework which reflects the Original Flows New Flows
intuition behind what is being accomplished with D-FACTS Line (1,3): 40.69 MW Line (1,3): 43.60 MW
Line(2,3): 80.17 MW Line (2,3): 67.32 MW
devices. The objective is to choose D-FACTS line impedance
settings to minimize the differences between the actual power Control Case 4
Objective: Lower Flow on (1,3) and Raise Flow on (2,3)
flows and the desired power flows. The objective function is
Original Flows New Flows
f0, where L is the number of line flows to be controlled: Line (1,3): 40.69 MW Line (1,3): 36.58 MW
𝑳
2
Line(2,3): 80.17 MW Line (2,3): 94.51 MW
𝑓0 = 𝑷𝒇𝒍𝒐𝒘,𝑐𝑎𝑙𝑐 𝒙 -𝑷𝒇𝒍𝒐𝒘,𝑠𝑝𝑒𝑐 𝒙 (15)
𝑖 Lines (1,2) and (2,3), with a flow coupling index of -1.0, are
𝑖=1
also targeted for control. The results are given in Table II.
The line flow control problem may be stated as follows:
Table II. Coupled Line Flow Control
min 𝑓0 Control Case 1
Objective: Lower Flows on (1,2) and (1,3)
𝑠𝑡 𝒇 𝑝,𝑞 (𝒔 𝜃 ,𝑉 ) = 0
(16) Original Flows New Flows
𝒙 ≤ 𝒙𝑚𝑎𝑥 Line (1,2): -40.57 MW Line (1,2): -32.44 MW
𝒙 ≥ 𝒙𝑚𝑖𝑛 Line(1,3): 40.69 MW Line (1,3): 32.61 MW
The first constraint of (16) represents the AC power balance Control Case 2
equations. The next two constraints are constraints on how Objective: Raise Flows on (1,2) and (1,3)
much D-FACTS devices are able to change the line Original Flows New Flows
impedances. The gradient of f0 is given by the following, Line (1,2): -40.57 MW Line (1,2): -48.71 MW
Line(1,3): 40.69 MW Line (1,3): 48.80 MW
∇𝑓0 = 2𝝆(𝒙)𝑨′′ (17)
Control Case 3
where the matrix A’’ is formed from elements of the power Objective: Raise Flow on (1,2) and Lower Flow on (1,3)
flow to impedance total sensitivity matrix, 𝚺 ∙ 𝚽 + 𝚪. Thus, Original Flows New Flows
D-FACTS devices are able to control line flows on any lines Line (1,2): -40.57 MW Line (1,2): -42.37 MW
Line(1,3): 40.69 MW Line (1,3): 42.73 MW
with high enough sensitivities, not just their own line.
Important connections exist between sensitivities and Control Case 4
optimization theory [17], [18]. The sensitivities which Objective: Lower Flow on (1,2) and Raise Flow on (1,3)
determine independently controllable line flows and effective Original Flows New Flows
D-FACTS locations also exactly provide the gradient needed Line (1,2): -40.57 MW Line (1,2): -42.33 MW
Line(1,3): 40.69 MW Line (1,3): 42.36 MW
to solve (15) using steepest descent. Steepest descent steps are
given by the following, where α is a positive, scalar step size: The negative sign on the flows of line (1,2) indicates that the
flow is in the other direction. The control objectives in Table
𝒙𝑣+1 = 𝒙𝑣 − 𝛼∇𝑓0 (18) II can not all be accomplished because lowering (raising) one
Knowledge of the total sensitivity of an equation to the control flow always lowers (raises) the other flow. In Cases 3 and 4,
variables is enough to know how to minimize that function. however, the control objective can still be accomplished.
Minimizing the objective function is equivalent to controlling Thus, the decoupled line flows (1,3) and (2,3) in Table I are
real power line flows with D-FACTS devices.

Authorized licensed use limited to: Bharat University. Downloaded on June 17,2010 at 07:33:34 UTC from IEEE Xplore. Restrictions apply.
shown to be independently controllable whereas the coupled line impedances are often set to their limits, settings can
line flows (1,2) and (1,3) are not independently controllable. potentially be chosen without completely solving the problem
for the optimal line impedance settings.
B. Utility Test System
Table III. Utility System D-FACTS Line Impedance Settings
The ability of D-FACTS devices to improve power system
Low
operation is evaluated for a system based upon a small North From To Bus Original High Limit D-FACTS
Limit
Bus # # x (+20%) Setting
American utility system provided by a US Midwest utility (-20%)
who is closely involved in this project. The system of Figure 2 23 0.1498 0.17976 0.11984 Low Limit
3 has 48 buses and 65 lines, although not all are in the utility’s 6 15 0.00034 0.000408 0.000272 High Limit
area. Buses 9 through 48 are in the utility’s area. The total 8 13 0.0559 0.06708 0.04472 High Limit
10 46 0.02316 0.027792 0.018528 High Limit
load is 571.44 MW. 13 15 0.01308 0.015696 0.010464 High Limit

VI. D-FACTS CONTROL FOR A GENERAL PROBLEM

The same control approach is extended to other power
system problems as follows
min 𝑓2 (𝒔 𝜃,𝑉 , 𝒙)
𝑠𝑡 𝒇 𝑝,𝑞 (𝒔 𝜃 ,𝑉 ) = 0
(19)
𝒙 ≤ 𝒙𝑚𝑎𝑥
𝒙 ≥ 𝒙𝑚𝑖𝑛
where f2 is the objective function for the problem of interest
and D-FACTS devices are placed at locations in the system
determined by the sensitivities of the objective function f2 to
line impedance which are furthest from zero.
The direction of steepest descent is given by – 𝛻f2, where
𝛻f2 is the total derivative of the objective function with respect
to x. Line impedance settings to minimize f2 are
Figure 3. Utility 48-bus Test System 𝒙𝑣+1 = 𝒙𝑣 − 𝛼 ∙ 𝛻𝑓2 (20)
The overloaded transformer between buses 7 and 8 is where α is a positive, scalar step size. D-FACTS devices may
targeted for control. It is assumed that the only lines which then implement the final line impedance settings. This
may be equipped with D-FACTS devices are lines in the approach can be used to implement D-FACTS applications
utility’s area. Results for different numbers of lines with D- such as loss minimization and voltage control, briefly
FACTS devices are shown in Figure 4. described below.
288 A. Loss Minimization and Voltage Control
282
For loss minimization, f2 is the losses equation (24). The
MW Flow

total sensitivity of (24) to line impedances is given by 𝛻f2 = Κ,

276
270
264
0 5 10 15 20 25
Number of Lines with D-FACTS Devices
Figure 4. Utility Case, Line Flow Control
Significant flow reduction is achieved through the use of D-
FACTS devices on the five most sensitive lines. The flow
through the overloaded transformer is reduced from 284.91
MW to 270.90 MW (4.92%). The addition of D-FACTS
devices to increasingly less sensitive lines is not beneficial, as
shown by the flat portions of Figure 4.
Table III summarizes the results of the scenario where D-
FACTS devices are placed on the five most sensitive lines. For
the utility test system for this scenario, the D-FACTS settings
are all at their limits. When settings are at their limits, the
benefit of having D-FACTS devices will increase if the
amount of possible line impedance change increases. Also, if
where Κ comes from (8) and (9).
For voltage control including both raising and lowering
system voltages, f2 is the sum of the differences of the bus
voltages from specified values. The gradient 𝛻f2, is given by
30 𝛻f2, = 2𝜼(𝒙)𝚽V where ΦV, the sensitivities of voltages with
respect to line impedance, are the lower section of the state to
impedance sensitivity matrix, Φ= [𝚽θ , 𝚽V ]T .
B. Comments on Other Solution Methods
The general problem can potentially be solved using other
methods. The steepest descent optimization approach used in
this paper is a logical choice because it requires only
knowledge of the sensitivities and the ability to solve the
power flow, and it guarantees movement toward the optimum.
The ability to guarantee descent is important since the goal is
to determine the extent of D-FACTS abilities.
One approach, often using Newton’s method, treats the
effective reactances of D-FACTS devices as state variables
and solves the modified power flow equations for the line
impedances in addition to the other state variables. Problems

Authorized licensed use limited to: Bharat University. Downloaded on June 17,2010 at 07:33:34 UTC from IEEE Xplore. Restrictions apply.
include that Newton’s method does not guarantee descent,
may not converge, and may not exhibit expected behavior if
started far from the solution. If second order sensitivities can
be calculated or approximated, the class of Newton-like
methods [18] may be worthwhile to investigate. Newton-like
methods also alleviate some of the problems with pure
Newton’s method.
VII. CONCLUSION
D-FACTS devices have the unique ability to be
incrementally installed on multiple lines throughout a system
to provide power flow control wherever needed. Effective D-
FACTS device locations and independently controllable flows
can be identified from sensitivities. After D-FACTS devices
are installed in certain fixed locations, their control objective
can easily be changed to target other lines flows. Thus, D-
FACTS devices can provide widespread, versatile control for
power systems.
In this paper, the successful control of line flows with D-
FACTS devices is presented for two test systems. A general
approach for line flow control with D-FACTS devices is
developed. The use of sensitivities in solving nonlinear
problems can be extrapolated to any application of interest and
to any system.
VIII. APPENDIX A
The AC power injection equations for real power P and
reactive power Q at a bus i are stated in (21) and (22),
𝑛
𝑃𝑖,𝑐𝑎𝑙𝑐 =𝑉𝑖 𝑉𝑗 𝐺𝑖𝑗 𝑐𝑜𝑠 𝜃𝑖 -𝜃𝑗 +𝐵𝑖𝑗 𝑠𝑖𝑛 𝜃𝑖 -𝜃𝑗 (21)
𝑗 =1
𝑛 [14]
𝑄𝑖,𝑐𝑎𝑙𝑐 =𝑉𝑖 𝑉𝑗 𝐺𝑖𝑗 𝑠𝑖𝑛 𝜃𝑖 -𝜃𝑗 -𝐵𝑖𝑗 𝑐𝑜𝑠 𝜃𝑖 -𝜃𝑗 (22)
𝑗 =1
where n is the number of buses. Power balance is expressed by
the vector f(p,q)(s(θ,V)) = ∆𝒑, ∆𝒒 𝑇 which must equal zero,
where s(θ,V) = 𝜽, 𝑽 𝑇 is a vector of bus voltage magnitudes
and angles, G+jB is the system admittance matrix,
∆𝑝𝑖 =𝑃𝑖,𝑐𝑎𝑙𝑐 -(𝑃𝑖,𝑔𝑒𝑛 -𝑃𝑖,𝑙𝑜𝑎𝑑 ), and ∆𝑞𝑖 =𝑃𝑖,𝑐𝑎𝑙𝑐 -(𝑄𝑖,𝑔𝑒𝑛 -𝑄𝑖,𝑙𝑜𝑎𝑑 ).
All real power line flows for the system comprise Pflow, and
system losses are the summation of all real power flows.
𝑃𝑓𝑙𝑜𝑤 ,𝑖𝑗 =-𝑉𝑖2 𝐺𝑖𝑗 + 𝑉𝑖 𝑉𝑗 𝐺𝑖𝑗 𝑐𝑜𝑠 𝜃𝑖 -𝜃𝑗 +𝐵𝑖𝑗 𝑠𝑖𝑛 𝜃𝑖 -𝜃𝑗 (23)
𝑛 𝑛
𝑃𝑙𝑜𝑠𝑠 = 𝑃𝑓𝑙𝑜𝑤 ,𝑖𝑗 𝑖≠𝑗 (24)
𝑖 𝑗
IX. REFERENCES
[1] FACTS Working Group, “Proposed Terms and Definitions for Flexible
AC Transmission System (FACTS)”, IEEE Transactions on Power
Delivery, Vol. 12, Issue 4, October 1997, p 1848-1853.
[2] P. Hassink, D. Matthews, R. O'Keefe, F. Howell, S. Arabi, C. Edwards,
E. Camm, “Dynamic Reactive Compensation System for Wind
Generation Hub,” IEEE PES Power Systems Conference and
Exposition, 2006, p 440-445.
Authorized licensed use limited to: Bharat University. Downloaded on June 17,2010 at 07:33:34 UTC from IEEE Xplore. Restrictions apply.
[3] D. Divan, “Improving Power Line Utilization and Performance With D-
FACTS Devices,” IEEE PES General Meeting, June 2005, p 2419 -
2424.
[4] D. M. Divan, W. E. Brumsickle, R. S. Schneider, B. Kranz, R. W.
Gascoigne, D. T. Bradshaw, M. R. Ingram, I. S. Grant, “A Distributed
Static Series Compensator System for Realizing Active Power Flow
Control on Existing Power Lines,” IEEE Transactions on Power
Delivery, Vol. 22, No. 1, Jan 2007, p 642 - 649.
[5] H. Johal, D. Divan, “Design Considerations for Series-Connected
Distributed FACTS Converters,” IEEE Transactions on Industry
Applications, Vol. 43, No. 6, Nov/Dec 2007, p 654 - 661.
[6] D. G. Ramey, R. J. Nelson, J. Bian, T. A. Lemak, “Use of FACTS
Power Flow Controllers to Enhance Transmission Transfer Limits,”
Proceedings of the American Power Conference, 1994, p 712-718.
[7] L. Gyugyi, C. D. Schauder, K. K. Sen, “Static Synchronous Series
Compensator: A Solid-State Approach to the Series Compensation of
Transmission Lines,” IEEE Transactions on Power Delivery, Vol. 12,
No. 1, Jan. 1997, p 406 - 417.
[8] L. Gyugyi, C. D. Schauder, S. L. Williams, T. R. Rietman, D. R.
Torgerson, A. Edris, “The Unified Power Flow Controller: A New
Approach to Power Transmission Control,” IEEE Transactions on
Power Delivery, Vol. 10, No. 2, Apr. 1995, p 1085 - 1097.
[9] L. Gyugyi, “Dynamic Compensation of AC Transmission Lines by
Solid-State Synchronous Voltage Sources,” IEEE Transactions on
Power Delivery, Vol. 9, No. 2, Apr. 1994, p 904 - 911.
[10] O. Alsac, J. Bright, M. Prais, B. Stott, “Further Developments in LP-
Based Optimal Power Flow,” IEEE Transactions on Power Systems,
Vol. 5, No. 3, Aug 1990, p 697 - 711.
[11 ] W. F. Tinney, V. Brandwajn, S.M. Chan, “Sparse Vector Methods,”
IEEE Transactions on Power Apparatus and Systems, Vol. PAS-104,
No. 2, Feb 1985, 295 - 301.
[12] X. R. Chen, N. C. Pahalawaththa, U. D. Annakkage, C. S. Kumble,
“Controlled series compensation for improving the stability of multi-
machine power systems,” IEE Proceedings: Generation, Transmission
and Distribution, v 142, n 4, Jul, 1995, p 361-366.
[13] H. Okamoto, A. Kurita, Y. Sekine, “A method for identification of
effective locations of variable impedance apparatus on enhancement of
steady-state stability in large scale power systems,” IEEE Transactions
on Power Systems, v 10, n 3, Aug. 1995, p 1401-7.
X. R. Chen, N. C. Pahalawaththa, U. D. Annakkage, C. S. Kumble,
“Output feedback TCSC controllers to improve damping of meshed
multi-machine power systems,” IEE Proceedings: Generation,
Transmission and Distribution, v 144, n 3, May, 1997, p 243-248.
[15] A. M. A. Hamdan, A. M. Elabdalla, “Geometric Measures of Modal
Controllability and Observability of Power System Models,” Electric
Power Systems Research, v 15, n 2, Oct 1988, p 147-155.
[16] T. J. Overbye, “A Power Flow Measure for Unsolvable Cases,” IEEE
Transactions on Power Systems, Vol. 9, No. 3, Aug 1994, p 1359-1365.
[17] D. G. Luenberger, Linear and Nonlinear Programming, 2nd ed, Kluwer
Academic Publishers Group, 2003.
[18] D. P. Bertsekas, Nonlinear Programming, 2nd ed, Massachusetts:
Athena Scientific, 1999.
Katherine M. Rogers (S’05) received the B.S. degree in electrical
engineering from the University of Texas at Austin in 2007 and the M.S.
degree from the University of Illinois Urbana-Champaign in 2009 and is
currently working toward the Ph.D. degree at the University of Illinois
Urbana-Champaign. Her interests include sensitivity analysis, power system
analysis, and power system protection.
Thomas J. Overbye (S’87-M’92-SM’96-F’05) received the B.S., M.S. and
Ph.D. degrees in electrical engineering from the University of Wisconsin-
Madison. He is currently the Fox Family Professor of Electrical and
Computer Engineering at the University of Illinois Urbana-Champaign. He
was with Madison Gas and Electric Company, Madison, WI, from 1983-1991.
His current research interests include power system visualization, power
system analysis, and computer applications in power systems.