IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 15, NO.
2, MAY 2000
Reactive Power Pricing: A Conceptual Framework
for Remuneration and Charging Procedures
Julián Barquín Gil, Member, IEEE, Tomás Gómez San Román, Member, IEEE, Juan José Alba Ríos, Member, IEEE,
and Pedro Sánchez Martín, Member, IEEE
Abstract—A new electrical sector regulation is being im-
plemented in different countries all over the world. The new
regulation stresses the role of competitive markets for the pro-
curement and remuneration of ancillary services. Among these
services stand the ones associated with reactive power supply and
transmission network voltage control in order to maintain the
required system security levels.
The object of this paper is twofold. On one hand, reactive power
supply and voltage control services, which today are bundled, are
decomposed in two types: (i) voltage profile management and re-
active dispatch and (ii) voltage regulation. A theoretical approach
based on marginal pricing is proposed in order to clarify the prin-
ciples to remunerate the suppliers and to charge the consumers of
these services. On the other hand, a practical organization of reac-
tive supply and voltage service markets is presented to be imple-
mented in a competitive environment.
Index Terms—Ancillary services, voltage control, reactive power
pricing, spot pricing, competitive markets.
NOMENCLATURE
Active and reactive power at bus
Voltage at bus
Reactive spot price, and losses and security compo-
nents, at bus
Cost function of device
System marginal active power price
Security component of the reactive spot price asso-
ciated to constraint
Constraint , and its level
I. INTRODUCTION
C URRENTLY, electric power systems all over the world are
moving from a regulated environment based upon tradi-
tional vertically integrated utilities toward a more decentralized
environment based on a much greater extent upon competitive
markets. Therefore, it is needed to address the issues related to
remuneration and charges of electrical services in this new and,
in many instances, still unknown world, in order to assure the
technical and economical power system feasibility [1]–[2].
Most of the effort has been aimed toward the proper treatment
of the remuneration of generation assets. Although generation
investments and fuel costs are the main system costs, there are
Manuscript received August 6, 1998; revised August 11, 1999.
The authors are with the Instituto de Investigación Tecnológica, Universidad
Pontificia Comillas, Alberto Aguilera 23, 28015 Madrid, Spain.
Publisher Item Identifier S 0885-8950(00)03776-7.
0885–8950/00$10.00 © 2000 IEEE
483
other less expensive generation services that should be provided
in order to maintain the system reliability and to meet the re-
quired security levels. Among these ancillary services are the
ones associated with reactive power support and voltage control
in the transmission network.
In vertically integrated utilities, the provision of these type of
services is closely linked to all the other aspects of electricity
generation, transmission and distribution: a central operator co-
ordinates the different suppliers that provide them, the remu-
neration that these suppliers receive for that is usually an indis-
tinguishable part of their total revenues. Only final customers
are usually charged with a penalty associated to the ratio reac-
tive-active energy consumption.
Whenever the electricity supply industry is based on com-
petitive markets, it seems reasonable to organize the ancillary
services provision around markets. Nevertheless, given the im-
portance of ancillary services for reliability and quality of ser-
vice, their complexity and their specific technical characteris-
tics, a significant degree of obligation and centralized control is
needed.
In this paper competitive pool-based generation markets are
considered, where there is a need for procedures to incentive
the participants in the market to provide reactive services and
ensure an adequate payment that guarantees the economic fea-
sibility of this business. That is the case of England and Wales,
Argentina, Australia and recently Spain. On the other hand, in
some other countries, like USA, for instance in California, it has
been proposed competitive schemes based on physical bilateral
contracts. In this later case, each transaction has consequences
that go beyond the involved parties in terms of additional voltage
control and reactive reserve requirements, etc. Thus, mecha-
nisms to ensure that participating agents assume the impact of
their transactions on the whole system are required [3].
In Section II of this paper, the two main voltage services, re-
lated on one hand to the voltage profile management and the
reactive power dispatch, and in the other hand to the voltage
control and dynamic security aspects, are described. Section III
proposes a possible organization of two different reactive mar-
kets: i) a reactive energy market based on losses spot prices,
and ii) a reactive capacity market based on a reactive regulating
capacity payment. Then, in Sections IV and V, the theoretical
support of spot pricing understanding and computation is pro-
vided. Finally, in Section VI, based on the markets organiza-
tion proposals previously presented, practical remuneration and
charging procedures for reactive supply and voltage control ser-
vices are proposed.
484
II. REACTIVE SUPPORT AND VOLTAGE CONTROL SERVICES
Reactive power management and voltage control services in
the transmission network can be decomposed into two main
types:
a) Voltage profile management and reactive dispatch:
This service is oriented toward the optimization of
the system active power losses cost while keeping
steady-state system security in the face of possible
contingencies (for instance, by keeping the voltage
profile and the reactive power reserves within margins).
It changes on an hourly basis or slower. Generators,
capacitors, reactances and distribution load management
provide this service. The role of the system operator is
to perform the reactive power dispatch by taking into
account the reactive control elements that belong to dif-
ferent agents: generation, transmission and distribution
utilities, as well as the cost of using them. The procedure
to solve the optimal dispatch must be transparent, nondis-
criminatory among the agents and technically sound.
Optimal Power Flow (OPF) methods have been used
for this purpose. This type of services has a similarity
with the active power economic dispatch related to the
implementation of the hourly pool-based energy market.
b) Voltage control: also known as voltage regulation, it is
the service provided to keep the network voltages in a
dynamic time frame (seconds to minutes). Generators,
SVCS, or other equipment capable of fast regulation can
provide it. It is oriented toward system dynamic security
and voltage quality. Although the economic effect is un-
deniable, it is also difficult to evaluate. This type of ser-
vice can be considered analogous to active power reserve
and frequency-control services (primary and secondary
AGC frequency regulation).
III. ORGANIZATION OF THE REACTIVE SUPPLY AND VOLTAGE
SERVICE MARKETS
Any reactive power market differs from active power markets
in certain characteristics derived from the different economic
and physical properties of active and reactive power flows:
a) The local geographic character of the reactive power
market versus the system wide character of the active
power market.
b) The relatively smaller investments in new equipment
needed to supply reactive power as compared to those
associated with the active power generation.
Reactive smaller investments ease competition, as more
agents can participate in the market (for instance, investing in
SVCS). On the other hand, the local character of the reactive
flows can cause that, in some moments, just one (or a few)
generator can provide the required reactive energy, leading to
monopolistic behavior. One way to avoid that possibility is by
requiring longer term bids than in active power markets. In that
way, generators can not bid, in any case, their reactive energy
higher than the cost of alternative reactive power generation
means. Actually, the market size grows because it is “enlarged”
in the temporal direction, trying to compensate for the loss of
competition among agents trough, so to speak, the space.
IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 15, NO. 2, MAY 2000
Economic theory shows that, under fairly general conditions,
the optimal level of remuneration and charging is provided when
pricing the reactive power injection or absorption (a measure of
the level of service) by its spot price [4]–[7]. The spot price at
bus is the cost of supplying a unitary increment (the last one)
of reactive power demand at bus . It can be decomposed in a
part related to the system losses cost (including generator losses)
and a part related to the security margins enhancement. A more
thoroughly discussion is presented in Section IV.
The component of the reactive spot price associated with
losses has a relatively smooth time variation. Its value is in the
order of something less than one US dollar per MVAr-h. On the
other hand, the component of the reactive spot price associated
with the system security is usually a very small quantity, but
in periods when the system security is threatened when it can
reach very high values. This different behavior suggests the
creation of two different reactive power markets: a reactive
energy market related to the losses cost minimization and a
reactive capacity market related to voltage security aspects.
A. The Reactive Energy Market
The reactive energy market would be based on long-term
bids provided by generators and other control elements to the
System Operator. The bid format includes the margin of the
reactive power variation (generation and absorption) and the
losses curve, which relates the internal equipment losses to the
reactive power, produced or absorbed by the control element.
These loss curves will be priced at the marginal price of the
hourly active energy market. The System Operator should dis-
patch the system including the losses generator’s reactive power
curves as an additional cost to be minimized. Both generator in-
jections and reactive power demands shall be remunerated or
charged by multiplying the reactive power amount by the cor-
responding losses minimization spot price. Distribution utilities
or large customers would adjust their reactive power demands
taking into account the current reactive power spot prices.
Fig. 1 represents the information exchange between the
agents in order to implement the proposed reactive energy
market. Generators and other control elements send to the
System Operator the reactive power generation and absorption
limits and the internal loss curves, and distributors and large
customers send the estimated reactive consumption. Then, the
System Operator performs an optimal reactive dispatch taking
into account all the available reactive sources and regulating
equipment including those owned by the transmission network
utility. From the solution of this optimization problem losses
reactive spot prices at each system bus are obtained. The
System Operator sends to generators and control elements the
voltage reference values that they must control. It also sends
the reactive spot prices that would be used to remunerate the
generators and control elements and to charge the demands.
B. The Reactive Capacity Market
The reactive capacity market would be also based on
long-term capacity bids provided by generators and SVC’s
to the System Operator to ensure system voltage security.
For selected bids there is a long-term obligation for voltage
regulation in their connection buses. The control element will
BARQUÍN GIL et al.: REACTIVE POWER PRICING
Fig. 1. Organization of the reactive energy market.
receive a capacity payment for this service. In order to avoid
the undesirable effects of the high volatility of security reactive
spot prices, this regulating service would be remunerated by a
capacity payment. The total equipment remuneration should be
proportional to the impact of the equipment on the expected
nonsupplied energy, and it should take into account the amount
of the reactive power capacity provided and its type of control
(time constant of the AVR response, integration in a secondary
voltage control loop, etc.). In Section V a theoretical discussion
is presented in order to compute the value of the capacity
payment.
IV. OPTIMALITY CONDITIONS AND SPOT PRICING
COMPUTATION
A power system can be described as a set of generators and
loads connected through a transmission network. Assume that
the power system is working in its steady-state optimal oper-
ation point. This operation point is found solving the optimal
reactive dispatch problem. If a load increases its reactive power
demand in a small amount, the rest of the system shall change in
order to supply the additional demand keeping optimality con-
ditions. It can be shown that the increase of the profit of the load
must be equal to the increase of the cost of the rest of the system.
This incremental cost is known as the reactive power spot price
[6]–[7]. The reactive spot price at node is denoted by .
Two kind of incremental system costs can be separated: those
related to system losses and those related to voltage security.
As a consequence, the reactive spot price can be also decom-
posed into two components: a losses component and a secu-
rity component . term
OPF algorithms can solve the optimal reactive dispatch
problem and they provide the optimal set points for the control
variables and also the reactive spot prices at each system bus.
In the sequel, it shall be assumed that the OPF optimizes the
operating cost of the system minimizing system losses cost
(including generator losses bids), whilst the level of security is
set by a suitable set of operational constraints [8]. Under these
assumptions, approximate formulae for the reactive marginal
prices are provided in the sequel.
485
A. Computation of Reactive Spot Prices at Generator Buses
Consider firstly a bus in which is connected a reactive power
source, generator or SVC, with enough reactive margin. Usu-
ally, any reactive load increment at that bus shall be almost to-
tally provided by the reactive source equipment connected to
the same bus, and therefore the reactive spot price is the deriva-
tive of the equipment operating cost curve. In general, the oper-
ating reactive costs are due to the internal losses associated with
the generation or absorption of reactive power. See appendix
B for illustrative purposes. For a given injected active power,
these costs can be written as . (In the case of gen-
erators, the loss function can be very approximately written as
[12]. Only the second term is rel-
evant. Besides, the generator is likely to be working at its max-
imum or minimum output). They depend on both the injected re-
active power and the terminal voltage . The reactive mar-
ginal price is:
(1)
The sensitivity is the variation in the generation
plant voltage when the injected reactive power changes. Usu-
ally, the second term is smaller than the first one, although there
might be exceptions.
B. Computation of Reactive Spot Prices at Load Buses
Consider now a load bus where no reactive generation
equipment is connected. Then, when increasing the reactive
power load, assuming that the rest of the loads remain constant,
the system cost is going to increase because three reasons:
a) The increment of reactive power generation.
b) The increment of the system active power losses produced
for the increment of reactive power flows.
c) The possible re-dispatch caused by some system con-
straints.
The reactive power spot price can be decomposed as:
(2)
The weight factors indicate in which amount each reac-
tive power generation equipment responds to the assumed re-
active demand increment in the absence of system constraints.
The term is the network active power losses incre-
ment caused by the assumed reactive load increment, being
the system marginal active power price. (This price is assumed
to be almost equal to the active power spot price.) Finally, the
represents the marginal contribution of the system
constraint to the system operation costs. This term is dif-
ferent from zero only when the constraint is active. A nu-
merical example is presented in appendix A in order to clarify
this spot price decomposition.
It is interesting to compare formula (2) with formula (3) for
the active power spot price at bus :
(3)
486
The active power spot price can be understood as the sum of
the system marginal active power price plus the incremental net-
work losses cost plus the incremental cost associated to network
active constraints [9], [11]. The main difference with respect to
the reactive spot price formula is the substitution of the system
marginal cost for the weighted sum of the reactive costs of the
generators, which responds to the load reactive increment. That
substitution embodies the local character of the reactive power
flows and it is the main reason because reactive prices are in
general much smaller than active prices.
C. Losses and Security Reactive Prices
The previously presented decomposition of the reactive spot
price into a losses component and a security component has im-
portant advantages from the point of view of organization of
reactive markets. System operational constraints are related to
maintain the system under specific security levels, such as to
impose a certain level of reactive power reserves in each system
area or to keep system voltages above security lower limits. As-
suming that all the system constraints are security-related, it is
straightforward the following term identification for the reactive
price losses component and the security component
(4)
The security component can rise to high values, specially if
the reactive power constraints impose the dispatching of higher
cost generating units that those which would be needed other-
wise.
V. REACTIVE SECURITY PRICING AND CAPACITY
REMUNERATION
The security component of the reactive spot price can be
obtained from the solution of the optimal reactive dispatch
problem. Specifically, in this problem, each operational security
constraint can be written as:
(5)
where is a function of the network variables, and the con-
stant is related to the strength of the constraint. For in-
stance, could be the required amount of reactive power
reserve in a system area.
The dual variable of the constraint obtained through the solu-
tion of the dispatch optimization problem will be used to com-
pute the value of as it was presented in Section IV. In each
operational condition, for instance in each hour, according to
the marginalist theory, the remuneration that should be paid to
an equipment connected to bus and providing reactive power
is:
(6)
IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 15, NO. 2, MAY 2000
Therefore, the total amount during a considered period is cal-
culated as:
(7)
However, the security component shows a high degree
of volatility, reaching high values in some critical periods (for
instance, when reactive reserves are needed), and going down
zero in most of time.
To avoid this undesirable effect, it is proposed that the re-
muneration associated with reactive reserves and reactive regu-
lating capability would be made through a “capacity payment”.
In order to compute the capacity payment, let us assume
that under ideal conditions for perfect competition, the optimal
amount of investment is given by the profit obtained by the
equipment when it is paid at the spot price [10]. On the other
hand, to invest in a new equipment connected at bus with
an available reactive capacity ( ) and a specific type of
regulating control (AVR, SVC, secondary voltage loop, etc.) is
tantamount to increase the critical operational constraint
in a quantity (for instance, an equivalent increase in
the reactive reserve in an area).
Therefore, the marginal benefit associated with the new in-
vestment, expressed for instance as a decrement of the cost of
the expected nonsupplied energy (NSE) would equal the mar-
ginal cost of investment. Thus
Total investment
(8)
Remember that the marginal saving of having an additional
MVAr in a given system bus is , then the total savings pro-
vided by the equipment are (formula 7) and the marginal
saving:
(9)
Note that this formula can be used to establish a capacity pay-
ment in each bus of the system. For instance, during a given
period, six months or one year, the expected security compo-
nent associated with reactive reserves in each system area of the
reactive power prices are computed, and capacity payments are
finally calculated as:
Capacity payment at bus (10)
VI. REMUNERATION AND CHARGING OF REACTIVE SUPPLY
AND VOLTAGE CONTROL SERVICES
Based on the markets organization proposal presented in Sec-
tion III and the theoretical concepts on reactive spot pricing de-
tailed in Sections IV and V, the following practical remuneration
and charging procedures for reactive supply and voltage control
services are proposed:
a) Remuneration for the reactive energy provision, which is
paid to generators and other voltage control equipment.
BARQUÍN GIL et al.: REACTIVE POWER PRICING 487

The amount of the remuneration is set by the losses re-

active spot prices ($/MVAr-h) times the injected reactive
power. The spot prices are computed by using the in-
ternal losses curves declared by the reactive sources in
their long-term bids and the marginal price of the active
energy hourly market.
b) Remuneration for the regulating reactive power capacity,
which is paid to generators and other voltage control
equipment. The amount of the remuneration is associated
with the available reactive capacity (in MVAR), and the
control (time constant, integration in a secondary voltage
control, etc.). Local or regional capacity payments
should be calculated. To receive this capacity payment
a long-term obligation to provide the regulating service Fig. 2. Base case.
should be agreed upon the supplier agents and the System
Operator.
and the quality of the control (dynamic response) is proposed.
c) The payments made by large customers and distribution
The theoretical basis for the computation of this capacity pay-
utilities shall be associated with their reactive energy
ment based on the improvement of the expected nonsupplied en-
consumption times the corresponding losses reactive
ergy associated with the installation of new control equipment
spot prices. Additionally, bounds on the coso could be
is presented. This approach allows the spatial differentiation of
established in order to charge an extra-payment associ-
capacity payments for each system electrical area depending on
ated to the security spot price if the power factor bounds
its particular voltage security characteristics.
are not met. The customer agent could sign interruption
Future work will deal with practical implementation issues
contracts with the System Operator in order to avoid
of this reactive market approach to real power systems where
those extra-payments. Usually distribution utilities can
numerical values for losses spot prices and capacity payments
influence the required level of service by making con-
will be obtained.
tracts with embedded generators, by capacitor banks
switching, etc.
d) The difference between the total remuneration (a and b) APPENDIX A
and the reactive energy payments made by large cus- The purpose of this appendix is to illustrate the spot price
tomers and distribution utilities (c) is the bundled part of decomposition shown in Section IV. Consider the small system
the service. This part could be dealt as a charge to all pool shown in Fig. 2.
participants proportional to the sold or bought active en- The figures between round brackets are the active and reac-
ergy amount, or included as an uplift in the pool market tive power (MW and MVAR) generated by the generators or
price. consumed by the load. The following figure is the bus voltage
(p.u.). The last two figures between square brackets are the ac-
tive and reactive spot prices ($/MW-h and $/MVAr-h).
VII. CONCLUSIONS
The generator 1 has an active power limit of 500 MW, a reac-
Two practical proposals have been made to organize the pro- tive power limit of 300 MVAR, and a cost of 43.333 $/MW-h.
curement, remuneration and charging of reactive and voltage The reactive power cost is assumed to be constant and equal
control services. to 1.333 $/MVAr-h. The generator 2 has also and active power
The first proposal deals with the organization of a reactive limit of 500 MW, an reactive power limit of 300 MVAR, an ac-
energy market based on long-term bids associated with internal tive power cost of 46.667 $/MW-h and a reactive power cost
losses of reactive control equipment for producing or absorbing of 0.667 $/MVAr-h. The operating point has been computed by
reactive energy. The system operator will perform periodically using an OPF.
optimal reactive dispatches minimizing operational costs related When the reactive power demand in bus 3 is increased by
to reactive generation bids plus network losses costs. Control 1 MVAr, the OPF yields the operating point shown in Fig. 3.
elements that produce or absorb reactive energy will be remu- Note that the generator 1 has increased its reactive power output
nerated at the losses component of the reactive spot price in in 0.68 MVAr and the generator 2 in 0.68 MVAr. This are the
their corresponding connection buses and reactive loads will be quantities denoted by , and in Section IV. The marginal
charged at the losses component of the reactive spot price at their generator 2 supplies the additional losses (0.03 MW). is the
connection buses. This settlement procedure will be performed active power cost of the marginal generator 2. Therefore, by
every hour according to the value of the energy market price as applying formula (2), it is found that
a reference cost for losses.
The second proposal deals with the remuneration of voltage
regulation services concerning system security. A remuneration
based on a capacity payment which takes into account the avail- This quantity is equal, when taking into account the lineariza-
able total regulating capacity in MVAr, its system allocation, tion round-off errors, to the one shown in Fig. 3. Note that the
488
Fig. 3. Increasing the reactive demand at bus #3.
Fig. 4. Reactive power limit activated in gen. #1.
Fig. 5. Losses curve.
total payments to the generators ($213.16) are less than the total
payment by the load ($285.3). This is related to the problem
of revenue adequacy (general issue in spot price remuneration
mechanisms), which is addressed in point of the proposed re-
muneration scheme.
To study the effects of the constraints, it is considered the case
when the reactive limit of generator 1 is precisely 104.99 MVAr,
that is the reactive power generation in the base case. The OPF
output is shown in Fig. 4. Note that, although the operating point
does not change, the bus 1 reactive spot price is increased to
2.033 $/MVAr-h. This is because generator 1 is not longer able
to supply any reactive power demand in bus 1. The increment of
the spot prices from those of Fig. 2 is the spot price component
related to the system constraints. So:
IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 15, NO. 2, MAY 2000
TABLE I
APPENDIX B
The purpose of this appendix is to discuss the cost of gener-
ating reactive power. This cost is mainly due to the active losses
in the generator and in the step up transformer caused by the
reactive power. These losses can be divided in Joule eddy, hys-
teresis and stray losses, and the losses of the excitation system
[12].
From the transmission network point of view the injected ac-
tive and reactive power and the voltage in the high voltage bus
characterize the generation plant. The generator operator tries
to minimize the generation plant total losses while keeping the
above constraints. The net result is that, for a given injected ac-
tive power, high side voltage and transformer tap, the generation
unit losses depends almost quadratically of the injected reac-
tive power, with minimum losses for a slightly capacitive power
factor.
The Fig. 5 shows the losses curves for a typical generator
whose data are provided the Table I, for different values of the
injected reactive power, when the high voltage bus is assumed
to be constant at 400 kV.
ACKNOWLEDGMENT
The authors gratefully acknowledge the help from Miss Laura
Gonzalez Ruiz de Valbuena and Mr. David Soler.
REFERENCES
[1] C. M. Bastos and M. A. Abdalá, “Reform of the Electric Power Sector
in Argentina,” Editorial Antartica S.A., Buenos Aires, 1996.
[2] P. Fox-Penner, “Electric Utility Restructuring. A Guide to the Competi-
tive Era,” Public Utilities Report Inc., Vienna, Virginia, 1997.
BARQUÍN GIL et al.: REACTIVE POWER PRICING
[3] S. H. Hao and A. Papalexopoulos, “Reactive Power Pricing and Manage-
ment,” IEEE Transactions on Power Systems, vol. 12, no. 1, Feb 1997.
[4] M. C. Caramanis, et al., “Optimal pricing: practice and theory,” IEEE
Transactions on Power Apparatus and Systems, vol. PAS-101, no. 9,
1982.
[5] M. Rivier, J. I. Pérez-Arriaga, and G. Luengo. JUANAC: a model for
computation of spot prices in interconnected power systems. presented
at 10th Power Systems Computation Conference
[6] M. L. Baughman and R. Siddiqi, “Real time pricing of reactive power:
theory and case study results,” IEEE Transactions on Power Systems,
vol. 6, no. 1, Feb. 1991.
[7] W. Hogan. Electric power spot pricing in a network: DC vs AC price
models. presented at 15th Annual International Conference of the Inter-
national Association for Energy Economics
[8] N. H. Dandachi, M. J. Rawlins, O. Alsac, M. Prais, and B. Stott, “OPF
for Reactive Pricing Studies on the NGC System,” IEEE Transactions
on Power Systems, vol. 11, no. 1, Feb. 1996.
[9] M. Rivier and I. J. Pérez-Arriaga. Computation and decomposition of
spot prices for transmission pricing. presented at Power Systems Com-
putation Conference
[10] J. I. Pérez Arriaga, “Curso de Doctorado Mercados de Energía Eléc-
trica,” (in Spanish), Universidad Pontificia Comillas, Instituto de Inves-
tigación Tecnológica, Madrid, 1994.
[11] M. Rivier, J. I. Pérez-Arriaga, P. Sánchez, A. Ramos, and T. Gómez.
An improved version of the model JUANAC: Applications to network
adequacy and economic studies in large interconnected power systems.
presented at 4th Probabilistic Methods Applied to Power Systems Con-
ference
[12] J. Barquín, D. Soler, O. Largo, G. Relaño, and I. de la Fuente. On the cost
of the reactive power generation and voltage support service. presented
at Bulk Power Services and Control IV
489
Julián Barquín Gil received the Ingeniero Industrial Degree and the Doctor
Ingeniero Industrial Degree from the Universidad Pontificia Comillas, Madrid,
in 1988 and 1993 respectively; and the Licenciado en Ciencias Físicas Degree
in 1994. He belongs to the research staff at the Instituto de Investigación Tec-
nológica. His present interests include control, operation and planning of power
systems.
Tomás Gómez San Román obtained the Degree of Doctor Ingeniero Industrial
from the Universidad Politécnica, Madrid, in 1989, and the Degree of Ingeniero
Industrial in Electrical Engineering from the Universidad Pontificia Comillas,
Madrid, in 1982. He is a research fellow and since 1994 Director of the Instituto
de Investigación Tecnológica. He has been involved in more than thirty research
projects with Spanish and European utilities. His areas of interest are planning
and operation of transmission and distribution systems, power quality and reg-
ulatory issues.
Juan José Alba Ríos obtained the Degree of Doctor Ingeniero Industrial and
the Degree of Ingeniero Industrial from the Universidad Pontificia Comillas,
Madrid, in 1996 and 1986, respectively. He has been Researcher of the Instituto
de Investigación Tecnológica until September 1997. His areas of interest are
artificial intelligence, knowledge engineering, machine learning, planning and
search problems, applications to power system operation and other industrial
problems.
Pedro Sánchez Martín obtained the Degree of Ingeniero Industrial from the
Universidad Pontificia Comillas in 1993. He is Research Assistant at the Insti-
tuto de Investigación Tecnológica. His areas of interest include the operation
and planning of power systems.