Sei sulla pagina 1di 6

1

Abstract-- This paper presents a procedure for minimizing


the costs involved with the reactive power voltage control
using the synchronous generators/compensators as ancillary
service. The procedure combines an optimal power flow
algorithm and the secondary voltage control algorithm, the
first aiming to minimize the reactive power costs
injected/absorbed by the synchronous generators/
compensators, while the second one assuming proportional
loading of the generators with respect to their maximum
capability limit. The reactive power of a generator can be
controlled by the system operator by specifying a certain
voltage at the generator terminals.

Index Terms -- ancillary services, deregulation, reactive
power market, secondary voltage control.
I. INTRODUCTION
HE voltage is a local electrical quantity which differs
from one nod to another because of the voltage drops
occurring on the power grid branches. For this reason, the
requirement for reactive power necessary to control the
voltage is different along the system.
Handling the reactive power in a power market, governed
by specific market mechanisms, transparent and
competitive, depends to a great extent by the possibility of
obtaining the primary information, that is the operating
point on the P-Q diagram of the synchronous generator and
the measurement of the reactive energy.
Currently, the implementation of a reactive power market
mechanism used for voltage control is restricted, in
several power systems, by the limited measurement
capabilities of the reactive energy and the real-time control
possibilities of the nodal voltages by the system operator. If
this inconvenient can be eliminated by additional
investments, also with the help of the metering and
computation advancements, exact identification of the
operation point on the P-Q characteristic, pushing the
operation to the maximum limit, both on the overexcitation
and underexcitation, would be possible.
However, the more loaded in reactive power is a
generator, the higher the operating costs. Remuneration of
generators for the reactive power provided as ancillary
service needs an appropriate mechanism. Since some
generators are strategically located and they can offer high

The work has been co-funded by the Sectoral Operational Programme
Human Resources Development 2007-2013 of the Romanian Ministry of
Labour, Family and Social Protection through the Financial Agreement
POSDRU/89/1.5/S/62557 and by CNCSIS Grant IDEI 1402 TAMPERE.
The authors are with the Electrical Power Systems Department of Power
Engineering Faculty from University Politehnica of Bucharest, 313, Spl.
Independentei, RO-060042, Bucharest, Romania (e-mail: lucian@ieee.org).
prices, an uniform cost based remuneration mechanism
might lead to high costs of the reactive power for which the
system operator have to pay in order to provide the voltage
control service. The proposed algorithm assumes application
of a pay-as-bid remuneration mechanism.
Although a large number of researches have been
dedicated to the consideration and remuneration of the
reactive power as ancillary service, so far only UK has
implemented a reactive market, while other countries uses
an embedded cost method. However, many countries
remunerate the reactive power only when produced closed
to the capability limit.
Attempts are being made by the engineers to design the
best reactive power market structure. However, few of them
are focusing on approaching the voltage control based on
area control. In [8] the authors present the design of a
localized competitive market for the reactive power
ancillary services at the level of individual voltage-control
area, by considering uniform prices in each control area, but
without considering proportional loading of the generators.
II. THE LOADING DIAGRAM OF THE SYNCHRONOUS
GENERATOR
The feasible operating domain of the synchronous
generator is defined by active power limits,
min max
P P P s s , and reactive power limits given by reactive
power functions, i.e.
min max
(.) (.) Q Q Q s s (Fig. 1,a).
Therefore, an active generated (notified) power
A
not
P has a
corresponding reactive power
min min max max
( ) , ( )
A A A A
g not not
Q Q Q P Q Q P
(
e = =

. The reactive
power produced/absorbed by a synchronous generator is
limited by the armature current (L1), field current (L2) or by
the underexcitation limiter (L4).
Let us consider that the generator operates to the
overexcitation limit, in the point (
A
not
P ,
max
A
Q ). If the system
operator would call the generator to increase the generated
reactive power by Q A value, then the operating point must
move in the way of active power decrease by P A . Mainly,
the positive value Q A is determined using the sensitivity
| |
1
max
Q P

c c in one or more points (Fig. 1,b). Similarly, if
the generator operates to the underexcitation limit in the
point (
A
not
P ,
min
A
Q ), and the power system shows a reactive
power surplus of Q A that must be absorbed by the
synchronous generator, the generator should reduce the
active power by P A , which can be calculated using the
sensitivity | |
1
min
( ) Q P P

c c [3].
Optimizing the costs of reactive power
for the coordinated voltage control service
Lucian Toma, Member, IEEE, Mircea Eremia, Senior Member, IEEE,
Constantin Bulac, Member, IEEE, and Ion Tritiu, Member, IEEE,
T
2
O
Q P
min min
( ) Q P
max min
( )
P
min
A
P
L2
L1
P
max
(L3)
Q
L4
(L5)
O
Q
O
Q
I
AP
AQ
AP
AQ
P
not
B
Q
as
I
II
P
not
A
B
max
Q
max
A B
A
Q
min
( ) P
not
Q P
min min
( )
Q P
max min
( )
c
II
c
I
c
u
Q
I
Q
a.s.
m.u./MVAr
Q
Q P
min min
( ) Q P
max min
( ) Q
I
Q
II
c
0
a.s.
a)
b)
c)

Fig. 1. Definition of the reactive power cost regions on the loading diagram
of the synchronous generators.

Figure 1,c shows the cost regions in terms of the reactive
power generated/absorbed by the synchronous generator:
- region 0, where 0
as
Q Q s s : the reactive power
generated in this region is used by the producer to
supply the internal auxiliary services
. . a s
Q ; in this
region the producer is not financially remunerated
because the reactive power is not injected into the
electrical network;
- region I, where
as I
Q Q Q < s : in this region, called in
Romania primary regulation band, the generator injects
reactive power into the network without affecting the
active power generation, but the active power losses in
the windings begin to increase; although these losses are
much smaller comparative to the active power losses in
the power grid, they have to be taken into account;
therefore, if this issue is seen from a market point of
view, two approaches can be implemented: financially
remuneration to a regulated fixed cost and remuneration
to the market clear price, respectively;
- region II, where
max I
Q Q Q < s : in this region, called
also in Romania secondary regulation band, the
generator is subjected to higher stress operation; first,
the closer the operating point is to the limit, the greater
the power losses in rotor and stator, and the generator is
subjected to accelerated aging; on the other hand, in
order to make possible the increase in reactive power,
the decrease in the active power might be required;
however, it is difficult to determine these
inconveniencies and this is why a real cost reflecting the
real effort cannot be determined;
- region u, where
min
0 Q Q s s : in this domain the
generator absorbs reactive power, and its operation
assumes some power losses; the operation in the
underexcitation region is restricted by the
underexcitation limiter; despite the fact that in order to
absorb more reactive power it might be requited to
reduce the generated active power, this issue is not
applicable in this region; it should be mentioned that the
hydro power plants can operate in this domain, while
operation of thermal power plants in this region might
cause critical problem into the machine.
III. COST BASED OPTIMIZATION OF REACTIVE POWER
A. Objective function
In a power market in which the system operator can have
a greater control on the generation sources, the reactive
power optimization problem becomes simple for this one
comparative to the active power markets. However, the
system operator does not have enough information to
estimate the generation costs of each generator, so that, for
the financial remuneration of the producers, the system
operator can make use of a market in which the producers
submit offers. For this reason, the offer model for reactive
power can be formulated according to a pre-established
design. If the power system configuration/characteristics
would allow a competitive environment, the producers
could be free to offer any price they wish; otherwise, the
offered price need to be limited to a certain value or
penalized by certain mechanisms.
According to the earlier description of the operating
regions on the synchronous generator loading diagram, the
approximated costs of the reactive power generated/
absorbed by a generator is defines as [7, 8]:

( )
1, 2,
, . ., ,
0
0, , , ,
d d d
i i
u i a s i I i
Q Q
i i u i i I i i II i i i
Q Q Q
CQ c c Q c Q c Q Q = + + +
} } }
(1)
where:
0,i
c is the availability cost offered by generator i, in m.u.;
, u i
c the offered costs for the reactive power absorbed
by the generator i in underexcitation, in
m.u./MVArh;
, I i
c the offered costs for the reactive power produced
by the generator i in the primary band, in
m.u./MVArh;
, II i i
c Q the opportunity cost of the reactive power
generated by the generator i in the secondary
band, in (u.u./MVArh)/MVArh; because this cost
includes also the reactive power, the cost function
will have a quadratic form;
3
, u i
Q the reactive power in the underexcitation state of
the generator i, in MVAr;
. ., a s i
Q the reactive power necessary to cover the
ancillary consumption of the generator i, in
MVAr;
1,i
Q the reactive power generated in the primary band
of the generator i, in MVAr;
2,i
Q the reactive power generated in the secondary
band of the generator i, in MVAr.
while m.u. stands for a monetary unit.

After integration of expression (1) and summing up the
expressions
i
CQ for all generators participating to the
voltage control through the offers process, the total cost of
the reactive power provided by all producers is given by the
expression:
0, , , , , . .,
2
1
, , . ., , , ,
( )
1
( ) ( )
2
p
n i u i u i i I i I i i a s i
i
II i I i i a s i II i II i i I i
c c Q c Q Q
CQ
c Q Q c Q Q =
| | + o + o +
|
=
|
+o + o
|
\ .

(2)
where:
o is a binary variable which defines the state of a
generator at a certain instant and the operating
region on the loading diagram
min,
,
1, if 0
0, otherwise
i i
u i
Q Q s <
o =


, ,
,
1, if
0, otherwise
as i i I i
I i
Q Q Q s <
o =


, max,
,
1, if
0, otherwise
I i i i
II i
Q Q Q s <
o =


n
p
the number of participants to the reactive control
process;
and
, I i
Q is the maximum value of the reactive power
corresponding to the primary band, in MVAr;
max,i
Q the maximum reactive power limit that a
generator can produce, in MVAr.
The optimization problem assumes the minimization of
the objective function, i.e:
min CQ
subject to equality and inequality constraints:
B. Relational constraints
The operating point corresponding to the generated
reactive power can be placed in one of the regions defined
in Figure 1; therefore, the sum of coefficients that defines
the state of a generator is at most 1:
1
u I II
o + o + o s (3)
If a generator is shut down or it generates reactive power
only for its own power plant consumption, the sum of
coefficients is 0.
The reactive power generated/absorbed by a generator is
defined as the sum of the values falling in the operating
regions defined in Figure 1,a:

, 1, 2, i u i i i
Q Q Q Q = + + (4)
Because the reactive power generated/absorbed can fall
only in one of the defined regions, the following constraints
are defined:

min, ,
0
u i u i
Q Q o s <

, . . 1, , , I i a s i I i I i
Q Q Q o s < o (5)

, 2, , max, II i I i II i i
Q Q Q o s < o
C. Nodal active and reactive powers balance
Irrespective of the operating conditions, physically, the
sum of powers entering in a node is equal to the sum of
powers going out from that node. For this reason, for
mathematical solution of the load flow problem,
optimization or other type of problems, the power balance
must be satisfied in every node. This constraint can by
formulated by:
| |
| |
, ,
1
, ,
1
cos( ) sin( )
sin( ) cos( )
n
g i c i i k ik i k ik i k
k
n
g i c i i k ik i k ik i k
k
P P UU G B
Q Q UU G B
=
=
= u u + u u
= u u u u

(6)
where:
, g i
P is active power entering in the node i;
, g i
Q reactive power entering in the node i;
, c i
P active power going out of the node i;
, c i
Q reactive power going out of the node i;
ik
G real part of the term
ik
Y from the nodal
admittances matrix;
ik
B imaginary part of the term
ik
Y from the nodal
admittances matrix;
i
U amplitude of nodal voltage i;
k
U amplitude of nodal voltage k;
i
u argument of nodal voltage i;
k
u argument of nodal voltage k;
D. Reactive powers limits
As mentioned before, the synchronous generators have
limitation as regards the lagging or leading operation. For
this reason, in the optimization problem the reactive power
of each generator i must be restricted to the capability limits:

min, max, i i i
Q Q Q s s (7)
E. Voltage limits
For security reasons, under normal operating conditions,
the voltage should not exceed the admissible limits:

min, max, i i i
U U U s s (8)
IV. THE THEORY OF SECONDARY VOLTAGE CONTROL
The hierarchical voltage control, based on electrical
network partitioning and automatic coordination of the
reactive power, called coordinated (secondary) voltage
control, was implemented so far in France, Italy and
Belgium, while other countries have performed studies to
analyze its effectiveness and benefits. The coordinated
voltage control assumes grouping the generators into several
areas, while the generators are proportionally loaded in
reactive power so that the voltage of a reference bus, called
pilot bus, is kept to a fixed value. Also, the secondary
voltage control assumes minimization of reactive power
flows between control areas.
4
A. The partitioning algorithm
For the partitioning of the power system into control
areas the following procedure is applied:
1. Calculate the base load flow;
2. Determine the Jacobian matrix corresponding to the
current operating point;
3. The network buses are grouped into control areas;
4. The voltage sensitivities with respect to the reactive
power produced by the generators are determined for
each area with problems;
5. The control capability of a certain generator from a
specific area is given by the product between the
generators reactive power band (
max min
Q Q ) and its
capability to participate to the voltage control;
6. Step 4 is repeated for all the necessary areas.
B. The theory of electrical distances between the power
system buses
The electrical distance concept is based on the
sensitivities of the bus voltages with respect to the reactive
power variations at the generator buses expressed in matrix
form as | | U Q c c . These sensitivities are obtained by
inverting the component | | Q U c c of the Jacobian matrix
used in the Newton-Raphson method based power flow
computation.
From the U-Q control point of view, the electrical
coupling between two buses can be quantified as the
measure in which the voltages of the two buses influence
each other. These influences (sensibilities) can be obtained
from the | | U Q c c matrix by dividing each column to the
corresponding diagonal term. Thus, the voltage sensitivity
matrix (also known as the mitigation matrix) containing the
sensitivities between all buses,
ij
o , is calculated.
The change of the voltage at bus i to the change of
voltage at bus j can be defined as [10]:

i ij j
U U A = o A (9)
where

j
i
ij
j j
U
U
Q Q
| | | | c
c
o = | |
| |
c c
\ . \ .
(10)
Generally
ij ji
o = o . To obtain symmetric proprieties for
two buses influence, the electrical distance between nodes i
and j is defined as:
log( )
ij ji ij ji
D D = = o o (11)
The
ij
D function is positive and symmetric and it defines
the influence of the buses from the voltage point of view.
The algorithm for the computation of the electrical
distance is the following:
1. The Jacobian matrix is computed and the
| |
QU
J Q U = c c component is retained;
2. The
QU
J matrix is inverted in order to obtain
| |
1
UQ QU
J J U Q

= = c c , where its elements are of the


form
i j
U Q i j
J U Q = c c ;
3. The mitigation matrix
ij
o between all buses is
determined as:
i j j j
ij U Q U Q
J J o = ;
4. The electrical distance
ij
D is computed by:
log( )
ij ij ji
D = o o (12)
The maximum efficiency of the voltage control is
obtained by using the control groups in the order given by
the sensitivities of the bus voltages. In this regard, the
electrical distance is normalized so that the most influential
bus to be quantized by 1, obtaining thus a bus hierarchy.
The normalized electrical distances are computed using:

1
/ { ,..., }
ij ji i in
D D Max D D = (13)
C. Adapting the Newton-Raphson method for the
hierarchical voltage - reactive power control system
The Newton-Rhapson method applied to compute the
load flow in a power grid consists in solving a nonliniar
algebraic system of equations containing the real and
reactive power balance equations.
( , ) 0
imp
Pi i i
f P U P = u = , i C G e (14)
( , ) 0
imp
Qi i i
f Q U Q = u = , i C e (15)
where ( , )
i
P U u and ( , )
i
Q U u are nodal powers and are
calculated as the sum of real and reactive powers changed
between bus i and the rest of the system, as shown in
equation (6), while

imp
i gi ci
imp
i gi ci
P P P
Q Q Q
=
=
(16)
where C represents the set of load buses, and G is the set of
generator buses.
In order to implement the secondary voltage control,
which involves the proportional loading of all the control
generators from a control area, it is necessary to complete
the mathematical model formed by equations (14) and (15)
with (
c p
n n ) supplementary equations that represent the
typical secondary voltage constraint bounds:

1
1, max
, max
... ; 1, ...,
j
c
j
c
j
j
n
a
j j
n
Q
Q
j n
Q Q
= = = (17)
where: n
p
is the number of pilot buses;
n
c
the number of control generators;
n
a
the number of control areas;

j
c
n the number of control generators from area
j.
while the subscript max stands for the maximum reactive
power limit.
Under these circumstances, the new mathematical
model can be defined as:

,max 1,max 1
( , ) 0, 1, 2,..., ( )
( , ) 0, 1, 2,..., ( )
, 1,..., ; 1, 2,...,
sp
Pi i i u g c p
sp
Qi i i u g p
j j j j j
c c c a c c c
f P P U i n n n n
f Q Q U i n n n
f Q Q Q Q j n c n
+ +

= u = = + + +

= u = = + +

= = =

(16)
where: n
u
is the number of load buses;
n
g
the number of generator buses;
and subscript c stands for the control generators.
In order to solve the nonlinear equations (16), the
Newton-Raphson method is applied. Hence, the following
matrix equation is obtained:
5
( , ) ( , ) ( , )
( , ) ( , ) ( , )
i i i
i i
i i i
i i
j
c c j j j
c c c
P U P U P U
U Q
P
Q U Q U Q U
Q U
U Q
Q f
f f f
U Q
c u c u c u (
(
cu c c
(
A ( Au (
( c u c u c u
( (
A A =
(
( (
cu c c
(
( (
A

(
c c c
(
cu c c (

(4.24)
equation that allows, through an iterative process, the
computation of the unknown variables U and and the
adjustment of the reactive powers produced by the control
generators, Q
c
.
V. CASE STUDY
A. Description of the test system
For the case study, the CIGRE Nordic32 test system was
chosen [9]. This test system is characterized by a generation
surplus in the Northern area and higher load in the Southern
part.
CS
31 1 14 3
4 2 15
32
5 6
17 16
13 12 18 19
21
20
28
29
30
27 26
25 22
23
11 7
9
24
10
8
SOUTH
CENTRAL
NORTH
EXTERNAL

Fig. 2. NORDIC32 power system.
B. System partitioning
In order to simulate the secondary voltage control system
implemented on the Nordic32 test system, the network is
partitioned into 2 control areas (Table 1). Area 1 consists of
the North area and the External area, while Area 2 consists
of the South and Central areas. In order to have a larger
number of generators, the External area was considered as
part of the whole system. Note that partitioning the network
into 3 control areas proved to be unfeasible.

TABLE 1. NORDIC32 TEST SYSTEM PARTITIONING
Area 1 Area 2
Controlled
generators
Non-controlled
generators
Controlled
generators
Non-controlled
generators
1, 2, 3, 4, 5, 15, 32 6, 13, 16, 31 18, 20, 29, 30 8, 9, 21, 26, 27
C. Producers offers for the reactive power market
The price offers corresponding to the domains from the
capability diagram for all the producers are shown in Table
2 [8]. In the absence of any credible information it was
assumed that the average reactive power consumption of a
power plant is roughly 10% of the maximum value. Hence,
it was assumed that Q
a.s.
=0.1Qmax and it was not
considered for financial remuneration. Moreover, the limit
between the primary and the secondary band was assumed
to be at 75% from the maximum value, while for the
simulations, Q
I
=0.75Qmax was considered.

TABLE 2. PRODUCERS OFFERS
Gen c
0
c
u
c
I
c
II

1 0.77 0.75 0.75 0.25
2 0.42 0.42 0.42 0.19
3 0.5 0.54 0.54 0.28
4 0.69 0.68 0.68 0.39
5 0.65 0.77 0.77 0.27
6 0.88 1.03 1.03 0.50
8 0.5 0.65 0.65 0.34
9 0.77 0.9 0.9 0.48
13 0.73 1.12 1.12 0.4
15 0.43 0.41 0.41 0.19
16 0.91 1.29 1.29 0.7
18 0.85 1.17 1.17 0.59
20 0.73 1.03 1.03 0
21 0.9 1.26 1.26 0.55
26 0.92 1.11 1.11 0.71
27 0.5 0.76 0.76 0.33
29 0.76 1.05 1.05 0.48
30 0.93 1.16 1.16 0.52
31 0.4 0.41 0.41 0.2
32 0.96 0.86 0.86 0.46
D. Results of the reactive power market
The mathematical models were implemented in Matlab,
while for the optimization problem, the fmincon function
was used.
Table 3 shows the voltage and the reactive power results
obtained for the 3 cases: the base case (init), the optimized
case having as objective the minimization of the reactive
power total cost (opt) and the case with secondary voltage
control operation without cost optimization (SVC).
The operation of a power system with secondary voltage
control implies the proportional loading of the control
generators with respect to the maximum capability limit.
Analyzing the results, it can be observed that there are
important individual modifications of the reactive power
generated by both the controlled generators and the non-
controlled generators. Nevertheless, the total generated
reactive power did not change significantly.
Evaluation of the reactive power total costs gives the
value 1344.8m.u. CQ = for the base case,
1143.02m.u. CQ = for the optimized case and
1193.21 m.u. CQ = for the case with secondary voltage
control system. The optimization mathematical model was
also combined with the mathematical model of the
secondary voltage control resulting a total cost of
1152.94 m.u. CQ = Mention that the availability cost was
not considered in these simulations.
It is easy to see that the total cost of reactive power can
be minimized by modifying the injected/absorbed reactive
power. This is possible by changing the voltage imposed at
the generators terminals. In the absence of an optimization
procedure, the application of the secondary voltage control
system leads to minimization of the reactive power cost.
6
TABLE 3. POWER FLOW DATA FOR THE INITIAL, OPTIMIZED AND
SECONDARY VOLTAGE CONTROL OPERATING REGIMES
Bus
U
init

[kV]
U
opt

[kV]
U
SVC

[kV]
Q
g,init

[MVAr]
Q
g,opt

[MVAr]
Q
g,SVC

[MVAr]
,
max
g SVC
Q
Q
1 404 403.90 403.77 94.05 50 74.52 0.149
2 146.9 146.57 146.43 84.99 69.01 59.62 0.149
3 148.85 148.85 148.03 43.96 79.21 44.71 0.149
4 150.8 149.835 149.57 82.08 35 52.16 0.149
5 143 143 142.86 44.64 53.3 44.71 0.149
6 138.23 137.21 138.10 125 95.68 125 1
7 124.87 123.76 123.77 0 0 0 0
8 130 130 129.99 79.01 81.40 84.61 0.423
9 128.64 127.35 127.75 100 78.33 100 1
10 127.85 127.55 127.31 0 0 0 0
11 128.8 128.49 127.74 0 0 0 0
12 231.71 229.76 231.16 0 0 0 0
13 242 238.61 242 145.77 130.18 151.88 0.357
14 145.93 145.90 145.67 0 0 0 0
15 404 404 404 -2.58 40 59.62 0.149
16 403.31 416.49 413.76 -30 150 109.31 0.729
17 397.56 395.72 397.19 0 0 0 0
18 404 401.52 402.78 113.25 17.5 29.82 0.17
19 405.46 405.96 406.19 0 0 0 0
20 400 400 400 -8.69 30 51.11 0.17
21 400 400 400 265.28 216.09 253.80 0.725
22 396.10 395.87 394.96 0 0 0 0
23 395.27 394.74 393.76 0 0 0 0
24 398.63 397.91 395.42 0 0 0 0
25 396.29 396.15 395.01 0 0 0 0
26 408 408 406.61 304.58 309.79 299.01 0.498
27 408 408 403.31 217.33 226.58 202.56 0.579
28 392.82 391.88 386.69 0 0 0 0
29 400.17 398.88 391.75 0 95.80 51.11 0.17
30 400 392.72 386.46 177.01 83.35 102.22 0.17
31 404 404 401.40 54.22 159.71 82.65 0.331
32 404 397.77 398.17 193.89 100 149.04 0.149
Q 2083.77 2100.93 2127.44
VI. CONCLUSIONS
Implementation of a market for reactive power needs
significant investments in the communication infrastructure,
similar to the secondary frequency infrastructure. Optimal
use of reactive power reserve through application of an
optimization algorithm can be done only in real time, which
needs real-time computation equipment, located at the
power system dispatching center, as well as the existence of
a communication system between the system operator and
the reactive power sources.
Besides the technical advantages, the issues earlier
presented are additional reason to justify the investments in
a reactive power - voltage control system, being almost
common with the reactive power market infrastructure. In
order to allow the improvement of the voltage control
efficiency and also the market competitiveness, a greater
number of control reactive power sources is needed; this is
possible only if the synchronous generators would be
equipped with advanced regulation equipment.
VII. REFERENCES
[1] Hao, S. A reactive power management proposal for transmission
operators, IEEE Trans. Power Syst., Vol. 18, No. 4, November 2003.
[2] da Silva, E.L., Hedgecock, J., Mello, J.C., da Luz, J.C.F. Practical
cost-based approach for the voltage ancillary service, IEEE
Transactions on Power Systems, Vol. 16, No. 4, November 2001.
[3] Gross, G., Tao, S., Bompard, E., Chicco, G. Unbundled reactive
support service: key characteristics and dominant cost component,
IEEE Transactions on Power Systems, Vol. 17, No. 2, May 2002.
[4] Hao, S., Papalexopoulos, A. Reactive power pricing and
management, IEEE Transactions on Power Systems, Vol. 12, No. 1,
February 1997.
[5] El-Samahy, I., Bhattacherya, K., Canizares, C. A unified framework
for reactive power management in deregulated electricity markets,
Proceedings of Power System Conference and Exposition, Atlanta,
SUA, 29 October 1 November, 2006.
[6] Chicco, C., Gross, G. Current issues in reactive power
management: a critical overview, Proceedings of IEEE PES General
Meeting, Pittsburgh, SUA, 20-24 July, 2008.
[7] Bhattacherya, K., Zhong, J. Reactive power as ancillary service,
IEEE Transactions on Power Systems, Vol. 16, No. 2, May 2001.
[8] Zhong, J., Nobile, E., Bose, A., Bhattacherya, K. Localized reactive
power markets using the concept of voltage control areas, IEEE
Transactions on Power Systems, Vol. 19, Nr., 3, August 2004.
[9] CIGRE Long Term Dynamics. Phase II. Final Report, Task Force
38.02.08, March 1995.
[10] Lagonotte, P., Sabonnadiere, J. C., Leost, J. Y., Paul, J. P. Structural
analysis of the electrical system: Application to secondary voltage
control in France, IEEE Trans. Power Syst., Vol. 4, pp. 479486,
May 1989.
[11] Toma, L., Eremia, M., Bulac, C. Aspects of ancillary services,
Power Systems Conference PSC 2003, Timioara, Romania, 6-7
November 2003.
VIII. BIOGRAPHIES

Lucian Toma received the B.Sc. and Ph.D. degree
in electrical power engineering from the University
Politehnica of Bucharest in 2002 and 2010
respectively. Currently he is lecturer at University
Politehnica of Bucharest, Power Engineering
Faculty, Electrical Power Systems Department. His
fields of interest include power market, ancillary
services, power systems dynamics, computer
modeling of power systems, steady-state and
transient stability assessment.


Mircea Eremia received the B.Sc. and Ph.D.
degree in electrical engineering from the
Polytechnic Institute of Bucharest in 1968 and 1977
respectively. He is currently Professor at the
Electrical Power Systems Department from
University Politehnica of Bucharest. His area of
research includes transmission and distribution of
electrical energy, power system stability and
FACTS applications in power systems.

Constantin BULAC was born in Romania in 1957.
He graduated at the Polytechnic Institute of
Bucharest in 1982 and has received the Ph.D.
degree in electrical engineering from the University
Politehnica of Bucharest in 1998. He is Professor
within the Electrical Power Engineering
Department from University Politehnica of
Bucharest. His research interests include stability of
power systems, FACTS devices and artificial
intelligence applications in power systems.

Ion Tritiu received the B.Sc. and Ph.D. degree in
electrical engineering from the Polytechnic Institute
of Bucharest in 1990 and 1998 respectively. He is
currently Associate Professor at the Electric Power
Engineering Department from University
Politehnica of Bucharest. His fields of interest
includes transmission and distribution of electrical
energy, distribution networks reconfiguration and
distributed generation.

Potrebbero piacerti anche