Reactive Compensation at Transmission Level

Dr K Rajamani, Power System Consultants, Mumbai

1.0 Introduction
With the opening of the Indian economy, the industrial sector is experiencing
a healthy growth. To meet the power requirements of the growing economy,
the power sector has to keep pace with this development. The private sector
is entering in the generation area as IPPs (Independent Power Producers).
However, the power has to be taken from the generating stations to the load
centres through a reliable transmission network. With most of the state
networks operating region-wise, secure operation of the network with many
parties involved (State Electricity Boards, Private licensees, National Thermal
Power Corporation, National Hydro Power Corporation, IPPs, Captive power
plants within big industries, wind farms, etc.) is a complex issue. In this
regard establishment of NPTC (National Power Transmission Corporation) is
a welcome development. With the proposed upgradation of all state and
regional LDCs (Load Despatch centres), on line control of the network to
avoid potential or impending disasters will be feasible. The present situation
in which grid collapses are assumed / inevitable, waiting to happen, may not
be acceptable in the future. However, transmission planning has never been
given the priority which generation planning always enjoyed. There have
been instances when generating stations with huge expenditure have been
completed without adequate transmission lines for transfer of power. Even
within transmission planning, reactive compensation for existing networks
receives less fiscal priority than say, constructing a new line.
2.0 VAR Flow Problems
The reactive power flows from generating stations and other devices to the
load centres through 400 kV, 220 kV, 132 kV, subtransmission and
distribution networks. This free flow of the reactive power results in
undesirable voltage levels in all parts of the network. Unlike active power,
which can be sent hundreds of kilometers, it is not practical to transmit
reactive power any appreciable distance because of resulting voltage
gradient. System voltage is like our body temperature. When the
temperature dips to 95° F or increases to 103° F, our body is sending a signal
for the impending problem requiring drastic action. However, even when the
400 kV system voltage dips to 350 kV, it does not create such a panic, which
it should. The very low voltages at EHV levels make it very difficult to set

1 of 7
October 1999, IEEMA Journal, Page 26 to 30
distance protection especially under heavy load condition. In an ideal
situation, the compensation shall be such that irrespective of the magnitude of
power flow on the lines, the voltage profile shall be near normal i.e. flat
voltage profile. Holding the voltage near normal values increases the
transient stability limit of the system allowing the system to handle more
power without the threat of instability. A sound and time tested policy in this
regard is that every voltage level shall be self sufficient in VAR support even
under heavily loaded conditions and reactive flows in the interconnecting
transformers should be minimum. This basic flaw in the Indian EHV network
planning may be one of the prime reasons for frequent grid collapses due to
initiating events, which are not uncommon in other countries. As an example,
reactive compensation available at EHV level in USA (NEMA survey, figures
upto 1994) is given in Table I. By year 2003, the shunt capacitive
compensation for NYPP (New York Power Pool) alone is expected to reach
the figures as shown in Table II.
Table I
kV MVAR
System Voltage Shunt Capacitor Shunt Reactor Series Capacitor
115-161 38,433 4,774 1,690
230 17,170 4,163 392
345 9,429 18,022 6,370
500 7,544 20,967 34,079
765 0 8,700 0

Table II
System Voltage Shunt Capacitor
(kV) (MVAR)
115-161 37,700
230 24,000
345 8,900

To us it may appear surprising to note the extensive shunt capacitive

compensation even at EHV level. But considering the vastness of their
network and all pervading interconnections in USA, the above level of
compensation is required to assure reliable operation of the network.
Unless the syndrome of one voltage level expecting reactive support from
other voltage levels is broken, one of the main causes for grid collapses in
India will not be addressed. By conducting load flow studies, the estimate of
compensation required can be arrived at. The logs of actual reactive flows on

2 of 7
October 1999, IEEMA Journal, Page 26 to 30
 the interconnecting transformers provide a clue for making initial estimates.
The main thrust is to make each voltage levels self-sufficient in reactive.
3.0 VAR Compensation Devices
There are many ways of providing compensation and each has its own
advantages and disadvantages. A brief description of some is given below.
3.1 Shunt Reactors
They are generally provided on 400 kV lines as part of a line without separate
switching arrangement. They are basically designed to absorb charging
MVAR from the line during light load period. When the line carries heavy
load, they are a drag on the voltage, as they cannot be independently
switched off. Neutral shunt reactor is provided only in case single-phase
auto-reclosure is planned for that line; otherwise it does not play any part
during system operation.
3.2 Shunt Capacitors
From individual blocks, with series-parallel combination, capacitors even up
to 400 kV can be built up and connected to the EHV network. Its advantage
is that it is static equipment with no moving parts. Its disadvantage is that
when reactive support is most needed, its reactive support is less, as the
reactive output from the unit is proportional to the square of applied voltage.
If split into different banks, each bank can be switched on or off as per
requirements. For EHV capacitor closing, recent trend is to use point on
wave closing of breaker (in addition to preinsertion resistors) so that inrush
currents during switching is minimized.
3.3 Synchronous Condensers
This is conventional synchronous machine but for producing only reactive
power. It is a rotating equipment. With the advent of brushless synchronous
condensers, the maintenance requirement is low. Its advantage is that when
the voltage is low, it naturally pushes more reactive into the system and when
the voltage is high, it absorbs reactive from the system. Incidents of voltage
collapse have been reported in USA, Japan, France, Sweden and Canada in
the recent past. Some system planners believe that more use of synchronous
condensers than capacitors could have mitigated these problems. One other
advantage, which has not been emphasized, especially in the Indian context,
is its stored energy (H). When the system islands or about to collapse, the
first second is very important and the rate of change of frequency (df/dt)
decay in this period determines where the system is heading. The only

3 of 7
October 1999, IEEMA Journal, Page 26 to 30
 countermeasure available during this period is the kinetic energy (H) stored in
the rotating masses of generating units. If H for the system is high, df/dt will
be low and vice versa. Though synchronous condensers are meant for
producing reactive power, by providing a large fly-wheel, its H can be
increased and will increase the net system inertia. By introducing a gear
between the synchronous condenser (similar to between high-speed gas
turbine and low speed generator), the speed at which the flywheel rotates can
be made higher which increases H substantially as it is proportional to the
square of the speed. Conversely the weight of the flywheel can be lower for
the same H with the provision of the gear. An extra support for the rotor
between the machine and the flywheel may be needed but this is not a major
issue. If a very large number of High Inertia Brushless Synchronous
Condensers (HIBSC) are used for reactive compensation at all voltage levels,
it will enhance the system stability during the crucial first one second after a
major system disturbance. The system planners should seriously consider
this alternative in view of the frequent grid collapses in India. The utilities can
jointly standardize the unit ratings (i.e. 25 MVA, 50 MVA, etc with H = 8
sec,…) and starting methods so that manufacturers can offer off-the–shelf
tested designs expeditiously. The voltage ratings of synchronous condensers
typically range from 6.6 kV to 15kV. Hence connection to EHV network needs
step-up transformers. But with the recent introduction of Extra High voltage
Synchronous machines, the condensers can be directly connected to the EHV
network without step-up transformers. This may be one of the interesting
applications for Extra High Voltage Synchronous machines especially for
problem prone grids like in India.
3.4 Statcons
Static Condenser is a static version of synchronous condenser and is part of
the family of FACTS (Flexible AC Transmission System) devices. Since they
are just now introduced in the market with very limited operating experience,
they may not find favour with Indian utility planners at this moment.
3.5 Static VAR Device (SVD)
Static VAR Device is also a part of FACTS family. The most popular
configuration is FC + TCR (Fixed Capacitor and Thyristor Controlled Reactor).
This static device imitates a synchronous condenser (without its H) in that it
can supply both leading and lagging VARS as per requirements. By
controlling the firing angle of TCR, reactive output from the unit can be varied

4 of 7
October 1999, IEEMA Journal, Page 26 to 30
 over a wide range. Since the voltage rating of thyristor is limited, SVD is
generally connected to the network at 33 kV or below.
4.0 Note on OLTC
There is a widespread belief that OLTC (On Load Tap Changer) is a magic
pill for voltage control. But use of OLTC by itself does not produce reactive
power and the voltage control is obtained by rescheduling reactive flows. For
example, in an industrial plant, operation of tap changer of a plant transformer
to improve the down stream voltage may increase the reactive draw from the
grid. OLTC by itself can not be a substitute for sound reactive compensation.
In this context the following comments may be pertinent:
1:N

|————
V1 ————|
V2
TRANSFORMER
Let the allowable operating voltage range be 90% to 110%.
(i) If V1 < 90% or V2 < 90%: Add capacitor.
(ii) If V1 > 110% or V2 > 110%: Add reactor.
(iii) If 110% >V1 > 90% and 110% >V2 > 90%: ,
(a) If V2 > V1 Decrease N
(b) If V2 < V1 Increase N
The function of OLTC is to obtain near normal voltage (say 100%) when the
voltages in the beginning are within the range of say, 90% to 110%. If the
voltages are outside the range from the beginning, shunt compensation shall
be switched on before OLTC operates. Indiscriminate operation of OLTC
without adequate reactive support is claimed to be one of the principle causes
for voltage collapse.
5.0 Conclusions
The main objective of this article is to refocus out attention on reactive
compensation at EHV level. For too long reactive planning has remained in
the background. Even under the limited reactive planning, the main action
has been to provide a few shunt reactors at 400 kV level and capacitors at
distribution levels. With the evolving complexity of regional power networks, it
is essential to provide compensation at every transmission level so that one
voltage level does not overburden the other levels to ensure security under

5 of 7
October 1999, IEEMA Journal, Page 26 to 30
severe system disturbances. Though static capacitors are in wide use, a
strong plea to consider high inertia brushless synchronous condenser as
another important alternative in the Indian context is made.

6 of 7
October 1999, IEEMA Journal, Page 26 to 30
 Comments from Scrutineers’ and Author’s Replies
1.0 Scrutineers’ Comment
A few references be given for further reading.

Author’s Reply
The references may be useful for understanding VAR compensation
alternatives:
(i) Reactive Power Control in Electric Power System, T J E Miller, John
Wiley, 1982;
(ii) Static shunt Devices for Reactive Power Control, CIGRE 31-08, 1974;
(iii) Power Electronics as a Work Horse for Power Systems, IEEE
Spectrum, July 1985.

2.0 Scrutineers’ Comment

Some data on prevailing Indian practice be given.

Author’s Reply
Except for providing shunt reactors at EHV level and shunt capacitors at
distribution level, as mentioned in the article, there is no concerted effort for
reactive flow control. Of course some utilities may have initiated some action
in this regard but this is more an exception than rule. Finally the proof of the
pudding is the voltage profile, which presently can be anywhere between 80%
to 120% even at the network level, but the saddest part is that this band is
accepted as normal.

7 of 7
October 1999, IEEMA Journal, Page 26 to 30