1

Abstract The IEEE Power System Stability Controls

Subcommittee developed a panel session for the Power
Engineering Society General Meeting in July 2008. One of the
contributions to this panel session is presented in this paper.

The objective of this paper is to describe black-start operations,
to discuss the studies that should be part of a black-start
planning process, and in particular to describe the system
dynamics and control aspects of the black start process.

Index Terms---Black start, power system restoration, motor
starting, overvoltage, protection, self-excitation.
I. INTRODUCTION
odern power systems are designed to have a high level
of reliability. System operation is closely monitored to
maintain safe operating margins. However, despite these
efforts, power system outages do occur. Most outages involve
only a portion of the power system, and this portion can be
restored with assistance from neighboring power grids. While
such restoration can be a complex process, the restoration
generally occurs in a relatively straightforward manner. The
tie lines from the outside power system to the blacked out
area are energized. Then tie-breakers are used to energize
the transmission high voltage grid. With significant circuits in
the transmission network energized, it is relatively
straightforward to pick up the subtransmission system, start to
restore some of system load, and supply auxiliary power to
selected power plants to bring generation back online. This is
an example of a top-down restoration, where the high voltage
grid is energized first and used to then energize the lower
voltage systems.
However, in the case of a wide spread blackout, there may
be no neighbor to help. In this case, system restoration must
begin from pre-selected generating units with the ability to
start themselves. These units, generally called black-start
units, are then used as the kernels to start the restoration
process. This can be considered a bottom-up type of
restoration, starting from an individual generating unit and
emanating outward towards the critical system load. Of
course to speed the restoration process, this black-start
sequence will likely be occurring simultaneously using several
generating units independently, where these independent
islands of generation and load will later be synchronized to


J. W. Feltes (email: james.feltes@siemens.com) and C. Grande-Moran
(email: carlos.grandemoran@siemens.com) are with Siemens Power
Technologies, Schenectady, NY 12305.
restore the original power system.

The objective of this paper is to describe black-start
operations, to discuss the studies that should be part of a
black-start planning process, and in particular to describe the
system dynamics and control aspects of the black start
process. This paper will not address other important
components of the restoration plan, such as communications
requirements, which are well documented in other references
[1] [8].
II. THE NEED FOR BLACK START ANALYSIS
The need for black start analysis is recognized in NERC
Standard EOP-005-1 System Restoration Plans [9]. The
stated purpose of this standard is to ensure plans, procedures,
and resources are available to restore the electric system to a
normal condition in the event of a partial or total shut down of
the system. Each transmission operator must have a
restoration plan to reestablish its electric system in a stable
and orderly manner in the event of a partial or total shutdown
of its system. Furthermore, each transmission operator and
balancing authority must verify the restoration procedure by
actual testing or by simulation. As testing is not always
practical or even possible, simulation is often the only feasible
approach. A further requirement is that the transmission
operator must document the cranking paths the number and
switching sequence of transmission elements involved,
including initial switching requirements, between each
blackstart generating unit and the unit(s) to be started.
NERC Standard EOP-009-0 Documentation of
Blackstart Generating Unit Test Results further addresses
black start plans, stating that a system Blackstart Capability
Plan (BCP) is necessary to ensure that the quantity and
location of system blackstart generators are sufficient and that
they can perform their expected functions as specified in
overall coordinated Regional System Restoration Plans [10].
Again, simulation is often required to demonstrate sufficient
performance.
III. BLACK START UNITS [11] [15]
Black start units are units that do not require off-site power
to start. Generally these fall into four categories:
Hydroelectric units. These units can be designed for
black start capability and have fast response
characteristics.
Diesel generator sets. Diesel sets usually require
only battery power and can be started very quickly.
They are small in size, and useful only for supplying
Black Start Studies for System Restoration
J.W. Feltes, Senior Member, IEEE, Carlos Grande-Moran, Senior Member, IEEE

M
2008 IEEE.
2
the power needed to start larger units. They
generally cannot be used to pick up any significant
transmission system elements.
Aero-derivative gas turbine generator sets. This type
of gas turbine typically requires only local battery
power to start. Often start-up is a remote operation.
These units can pick up load quickly.
Larger gas turbines. These units are not in
themselves black start capable, but are coupled with
on-site diesel generator sets to make the plant a black
start source. The diesels are started and used to
energize plant auxiliary buses and start either the gas
turbine or steam turbine. A gas turbine is generally
quicker to bring on-line. Time to restart and
available ramping capability will be a function of
how long the unit was off-line.
IV. SYSTEM ENERGIZATION [16]- [18]
In the case of a total system outage, system restoration
must begin from the black start unit(s), restoring the power
system outward towards critical system loads. As the black
start units themselves can only supply a small fraction of the
system load, these units must be used to assist in the starting
of larger units, which need their station service loads to be
supplied by outside power sources. Full restoration of system
load can only occur when these larger units can come on-line.
Thus, the restoration plan following a system blackout should
include self-starting units that can be used to black start large
steam turbine driven plants located electrically close to these
units. Another objective for many systems is the supply of
auxiliary power to nuclear power stations in need of off-site
power to supply critical station service loads. Other priority
loads include military facilities, law enforcement, hospitals
and other public health facilities, and communication
facilities.
In this paper we are focusing on the most common
definition of black start analysis, the use of a self-starting unit
to start a larger unit through the energization of a small
portion of the transmission system and the starting of motors
and other auxiliary loads at the plant to be started.
The typical black starting scenario includes the self-starting
unit(s), the transmission lines that will transport the power
supplied to the large motor loads in the power plant to be
black started, and at least three transformer units. These
would include the generator step-up transformers of the black
start generating unit, the generator step-up transformers of the
steam turbine unit involved, and one or more auxiliary
transformers serving motor control centers (MCC) at the
steam plant. The transmission lines used for the black start
may be either an overhead line(s) or high voltage underground
cable(s). The load to be black started includes very large
induction motors, ranging from a few hundred HP to several
thousands HP and also plant lighting and small motor load.
The black start plan describes the steps that the
transmission operators need to take to restore the isolated
power system from the black start unit. This includes
sequentially energizing transformers, transmission lines, and
potentially shunt compensation and load pickup, to supply
power to the steam unit auxiliary loads allowing that unit to
begin operation.
The key concerns that are stressed here is the control of
voltage and frequency. Both voltage and frequency must be
kept within a tight band around nominal values to guard
against damage to equipment and to ensure the progress in the
restoration process. Any equipment failure will severely
hinder the restoration process and may require starting over
with a revised plan. System protection operation can also
occur if voltage or frequency goes outside acceptable ranges,
again with the potential to set back or stop the process [19].
Black start plans must be validated either by tests or by
simulation. Testing of the starting of black start units is
straightforward. Testing of line energizations is more
complex, as it involves de-energization of parts of the
transmission system such that they can be connected to the
black start unit. This must be accomplished without impact to
loads. This may not always be possible. Restoration plans
that require the energization of load at a particular step in the
plan cannot be tested beyond that step, as it is never
acceptable to submit loads to outage and pickup as part of a
test. Thus simulation is usually required to validate the plan,
rather than field testing, which is covered in other references.
A black start study is performed to verify the feasibility of a
black start plan in terms of both steady state and transient
operating conditions.
The steady state analysis of this isolated power system
includes:
Voltage control and steady state overvoltage
(Ferranti effect) analysis
Capability of the black starting units to absorb
reactive power (vars) produced by charging
currents of the transmission system.
Step-by-step simulation of the black start plan
being tested to ensure its feasibility and
compliance with required operational limits.
Verification of the robustness of the tested black
start plan to ensure its ability to compensate for
the unavailability of key components to be used in
the plan
Demonstration of generation and load matching
capability
Voltage control analysis determines the black starting
generating unit voltage reference set-point and off-nominal
tap setting for all transformers that are part of the plan. This
insures proper control of voltage and provides the needed
terminal voltage to start up the large induction motor loads at
the black started plant. Note that transformer tap settings that
are appropriate for normal conditions, generally having
significant current flows, may result in high system voltages
under the lightly loaded black start condition. Since most taps
on generator step-up transformers and station auxiliary
transformers cannot be changed under load, the selection of
3
transformer taps must be a balance between the needs of the
black start period and later operation when the power system
is supplying a significant amount of load.
Load flow simulations can be used to calculate the
receiving end bus voltage of the transmission line(s) when the
black starting unit energizes the unloaded generator step-up
transformer and transmission line(s). The charging current
generated by an unloaded transmission lines will result in a
rise in voltage along the line. This is particularly true when
underground cables are used to transport the power supplied
by the black starting unit as they have significantly more
charging capacitance. The charging requirements can be large
enough to have the black starting unit absorbing reactive
power. There could be, for extreme conditions, the potential
for self-excitation, which is discussed later.
The steady state analysis of a black start plan should
include a step-by-step simulation of the plan to verify its
compliance to required operational limits on voltage and
power flows. The robustness of the plan to a loss of a system
component is also valuable, as it is reasonable to assume that
the events leading to the blackout could result in some
equipment unavailability during the restoration period.
Generally thermal overloads are not an issue, as the system is
lightly loaded, although this may become a concern as
restoration progresses and load is picked up.
Once the steady state analysis has been completed, a
dynamic analysis of the black start plan follows. The dynamic
analysis starts from an initial steady state operating point
representing a step in the plan. This initial system operating
condition is usually obtained from the system steady state
analysis. One key simulation starts from the state of the
isolated power system prior to the start up of the largest
induction motor load and simulates the starting of that motor.
The importance of accuracy in equipment modeling cannot
be over-emphasized. Whereas in normal operation, a large
number of units are controlling voltage and frequency and the
effect of an individual unit is generally not as significant,
under a black start condition, the black start unit is solely
responsible for these two tasks. Thus the modeling of the
generator, excitation system and speed governor is very
important. The modeling of equipment that does not
generally operate under normal conditions such as over- and
under-excitation limiters can be quite important. Governor
modeling must take into account whether the machine is
operated in an isochronous or droop control mode, as will be
discussed later in this paper. The accuracy of dynamic
modeling parameters of any large motors to be started are also
important to motor starting simulations.
The dynamic analysis of a black start plan includes some or
all of the following functions which will be explained below:
Load-frequency control
Voltage control
Large induction motor starting
Motor starting sequence assessment
Self-excitation assessment
System stability
Transient overvoltages
Because frequency may deviate significantly from its
nominal value, the effect of frequency variation on system
impedances must be modeled.
A. Load-frequency control
During the restoration process, the black start generating
unit will typically be used to pick up large induction motors,
e.g., boiler feed pump, forced draft and induced draft fan
motors. The frequency of the black start system will be
controlled by the speed governor of the turbine driving the
black start synchronous generator.
It is standard practice in multi-machine power systems with
units operating in parallel that all prime movers supplying the
mechanical power to the generators coupled to these units be
operating on a droop governing mode. This provides a stable
sharing of the electric system load among all units. However,
because of the proportional characteristic of the droop speed
governor control, a steady state frequency error will remain in
the system. Supplementary frequency control, in the form of a
pure integral controller mode, will follow the primary
frequency control action of the speed governors to remove
this undesirable steady state frequency error. Typical steady
state regulating droop (R) for speed governors is 5% on a
system frequency base of 50 Hz or 60 Hz and a turbine MW
rating base power. In black start plans, however, it is
imperative that system frequency regains its scheduled value
following the start up of motors or pick up of other load.
This frequency control should be automatic, since the crew
in charge of the black start unit will be operating under
extreme emergency conditions, which can lead to undesirable
operating errors. Thus, the automation of the frequency
control process can be carried out by the prime mover speed
governor operated in a constant frequency or isochronous
control mode. In this pure integral control mode, the steady
state frequency error is zero because of the resetting
characteristic of the pure integral control. Most, if not all,
modern diesel engines, gas turbines and hydraulic turbines are
furnished with digital speed governors where a droop or
isochronous operating mode is carried out by a simple change
in command.
Once the system has more than one generating unit on-line,
all speed governors should be operated in a droop control
mode, unless it is decided in the restoration plan that one of
the largest units will maintain the control of the system
frequency. The automatic generation control (AGC) will be
disabled under this extreme contingent scenario.
In summary, the preferred control mode for speed
governors associated with black start unit is isochronous or
constant frequency control. When additional units are added,
the preferred control mode for speed governors is droop
control mode. In some cases, it may be preferable to keep one
large unit in isochronous control mode. Units should not be
operated in parallel with more than one unit in isochronous
control mode.
4
B. Voltage Control
Control of voltage is obtained through the generators
excitation system. The excitation system must be operated in
automatic control, that is, with the automatic voltage regulator
(AVR) in service. The system voltage will be a function of
the generator terminal voltage. Thus the generator scheduled
voltage may need to be adjusted throughout the restoration
process, as load is picked up, and also coordinated with any
changes in transformer tap positions. Such adjustments
should be an integral part to the restoration plan. The changes
in voltage that will be seen upon the starting of large motors
or the pickup of large blocks of load require that the
excitation system respond in a quick, well-tuned manner.
C. Motor Starting
A black start units primary function is generally to start up
the auxiliary load of a larger power plant. This auxiliary load
is made up of lighting and motor load used, for the most part,
in the start up of steam generators and fuel systems. The
motor load is made up of a large number of small and medium
size HP motors and a few large HP motors ranging anywhere
from several hundred HP and up to several thousand HP. Fuel
and feed-water pump motors, and forced and induced draft
fan motors belong to this large HP group. It is this latter group
which presents the greatest challenge to the reactive power
resources available in any well-designed black start plan [20].
The method used for starting up these large motors is often
across-the-lines, that is, a hard start, but occasionally motors
may also be soft started, that is, they are started at reduced
voltage. Thus, it is extremely important to properly identify
the motor starting method since this will greatly impact the
depth of the dip in voltage seen during the black start process.
Also, it is very important to gather motor data that will
assist the analyst conducting dynamic studies to verify the
viability of a given black start process. The information
needed to establish the dynamic model for the large induction
motors participating in the black start process includes: motor
plus mechanical load inertia, starting or locked rotor torque,
starting or locked rotor current and associated power factor,
pull-out torque, full-load torque, full-load current and
associated power factor. All of these data should be at rated
voltage and frequency. From this motor performance data,
parameters for the stator and rotor circuits are estimated. The
dynamic model for these large induction motors should
include both inertial and rotor circuit flux dynamics.
Verification of this dynamic model must result in a close
matching of the speed-torque characteristic, particularly at
starting, pull-out and full-load operating points. In addition, it
is important to include the mechanical load damping effect in
the inertial model of the mechanical load, which for most
centrifugal pumps and fans follows a quadratic speed-torque
characteristic.
The motor starting sequence is another variable that must
be verified in any black starting process. Is it feasible to start
up first the lower HP motor in the large motor group or is it
better to begin the motor starting process with the largest
motor in the group down to the smallest? This is the question
that must be answered by a dynamic simulation of the various
motor starting sequences that may be selected for a black
starting plan.
The voltage dip caused by the starting of these large
induction motors must also be accurately quantified because
magnetic contactors used by the already on-line motors open
up generally around 80% of their terminal voltage. The IEEE
Standard 399-1997 recommends a minimum terminal voltage
of 80% of rated voltage. Occasionally there may be magnetic
contactors that can hold their contacts for voltages as low as
70%; however, the number of cycles that this operating
condition can be sustained is low. In addition, the life
expectancy of the insulation of the stator and rotor windings is
reduced as a result of the large currents circulating through
these windings. In situations where this is found, the motor
manufacturer must be consulted to avoid a catastrophic failure
at worst or shorted turns at best. Undervoltage protection
settings should also be verified to avoid opening of circuit
protection caused by undervoltage relay action.
The accelerating time period required by an induction
motor depends in great measure on the combined inertia of
the motor and its mechanical load. The longer the
accelerating period, the higher the heating experienced by the
stator and rotor windings. When accelerating periods last a
few tens of seconds, motor manufacturer data on allowed
motor heating should be consulted to avoid a significant loss
of useful operating life of the winding insulating material.
D. Self Excitation
As noted above, energization of a transmission line or cable
will result in a rise in voltage along the line or cable due to
charging currents. The charging requirements can be large
enough to result in the black starting unit absorbing reactive
power. There is the potential for self-excitation if the
charging current is high relative to the size of the generating
unit. The result can be an uncontrolled rise in voltage and
could result in equipment failure. Such an undesirable
operating condition may occur when the effective charging
capacitive reactance of the transmission system used in the
black start operation, as seen by the black starting unit, is less
than the q-axis generator reactance Xq. In generating units
with no negative field current capability, d-axis self-excitation
cannot be controlled by the excitation system, and thus the
machine terminal voltage rises almost instantly for cases
where the capacitive reactance is less than the d-axis
reactance Xd. Generator excitation systems with negative
field current capability delay but do not prevent the onset of
q-axis self-excitation. It is worth noting that most generating
units installed in the last 40 years do not have negative field
current capability. Thus, it is extremely important to verify the
reactive power capability of the black starting unit when
operated at a leading power factor [21].
Self-excitation can also occur from the load end through
inadvertent loss of supply, with the opening of a transmission
line or cable at the sending end, leaving the line connected to
5
a large motor or a group of motors.
E. System Stability
Finally, system stability is checked to make sure that
voltage and rotor angle stability are maintained during the
implementation of the black start plan. Angular stability is
assessed only when more than one generating unit is used in
the black start plan. Otherwise, frequency stability is what
matters in the stability assessment of the plan.
F. Cold Load Pickup [17], [22]
While black start analysis often concentrates on the starting
of large motors, the pick up of other loads is often also part of
the plan. If the load has been de-energized for several hours
or more, the inrush current upon re-energizing the load can be
as high as eight to ten times normal. The magnitude and
duration of the inrush current that flows when a feeder is re-
energized after a prolonged outage is a function of the type of
load served by the feeder. This could include lighting,
motors, and also thermostatically controlled loads such as air-
conditioners, refrigerators, freezers, furnaces, and electric hot
water heaters.
There are different components of the load which
contribute to the total inrush current. An example is the
component due to the filaments of incandescent lights. The
resistance of the filament is very low until it warms to
operating temperature. This low resistance results in a very
high inrush current up to ten times normal. This high
current flows for a short period, approximately one-tenth of a
second.
Another component of the inrush is due to the starting of
motors when the load is picked up. When a motor starts, the
current drawn will be typically five to six times normal, until
the motor accelerates up to its operating speed. This may take
as long as several seconds for large industrial type motors.
A third component of inrush current is that due to
thermostatically controlled loads, which turn on and off
automatically to hold temperature to a desired, preset value.
Under normal operating conditions, probably only about one-
third of these loads will be connected at any instant in time.
But after a lengthy interruption of service, they will all have
their thermostat contacts closed, waiting to run as soon as
power is restored. As a result, these thermostatically
controlled loads will be perhaps three times greater than they
normally would be for the first half hour or so after being
energized. Most thermostatically controlled load also contain
small single phase motors, which will draw five or so times
running current until they are accelerated up to running speed
in perhaps a half second. This results in the initial current
drawn by some thermostatically controlled loads to be as high
as fifteen times normal current for the first half second
following energization.
A summary of the magnitude and duration of inrush current
for some of the various types of load s is shown in Figure 1.

Figure 1. Load Variation Following Cold Load Pick Up
G. Transient Overvoltages
Energizing equipment during black start conditions can
result in higher overvoltages than during times of normal
operation. Overvoltages can lead to equipment failure or
damage that may hinder the successful implementation of the
utility restoration plan. These transient overvoltages originate
from energizing operations and equipment non-linearity
(transformer saturation). Transformer inrush currents can
have a high magnitude and contain a significant harmonic
content. These harmonics can interact with system inductive-
capacitive resonances, potentially resulting in resonant
overvoltages that will be sustained for seconds because of the
lack of sufficient damping. The equivalent system
inductances are relatively high during a restoration process
because of the relatively few generators on-line and/or
because of the relative sparseness of the grid. Therefore, the
first system resonant frequency can be much lower than
during normal system operation. Large capacitances also
contribute to the low resonant frequencies.
The following material is a brief discussion of voltage
stresses in electric power systems. Voltage stresses can be
classified as:
Continuous steady state power frequency
Temporary overvoltages
Switching surges
Lightning surges
All (except lightning) are a concern for a black start
restoration of a power system.
Switching surges are the transient overvoltages that
immediately follow the opening or closing of a circuit breaker
or other switching device. Switching surges have high
frequency (100 Hz to 10 kHz) components that decay quickly,
typically within two to three cycles of the power frequency,
and are followed by a normal steady state voltage. Switching
surges typically contain only one, or just a few, peaks that are
of interest, as shown in Figure 2. This sample waveform is
from a simulation of the energizing of a typical overhead line.
6
(f il E FIG l ) LINOGB
0 10 20 30 40 50 60 70 80
[ms]
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
[V]

Figure 2. Switching surge from an EMTP simulation of the energizing of a
typical overhead line where the voltage is in per unit.
The magnitude and waveshape of the switching surge
depends upon the angle of the power frequency source voltage
wave at the instant of circuit breaker closure. Thus, many
simulations must be made with various closing times to obtain
the overvoltage results as a statistical distribution. Surge
arresters are effective in limiting the peak of the switching
surges. Cables and transformers often do not have a
switching surge withstand voltage rating.
The classification of temporary overvoltages (TOV)
includes many types of events where the voltage exceeds the
rated value for three cycles or more. TOV encompasses
power frequency phenomenon such as a Ferranti rise on an
open-ended line or cable, the overvoltage on an unfaulted
phase during a single line to ground fault, and other events.
Temporary overvoltages can also follow switching surges.
TOV results from switching circuits that saturate the core of a
power transformer, such as when cables and transformers are
energized together. The harmonic rich transformer inrush
currents interact with the harmonic resonances of the power
system. The resonant frequencies are a function of the series
inductance associated with the system's short circuit strength
and by the shunt capacitances of the cables and lines. Higher
inductances (relatively weak systems) and higher capacitances
(long cables) yield lower resonant frequencies and a higher
chance of TOV.
Figure 3 shows an example of a temporary overvoltage
taken from a simulation of the energizing of a large
transformer.
Like switching surges, this type of temporary overvoltage
can be dependent upon the circuit breaker closing times. In
contrast to switching surges with one predominant peak,
temporary overvoltages can have hundreds of peaks of about
the same magnitude if the TOV duration is several seconds. It
is rare to find TOV ratings for cables and transformers.
0.0 0.1 0.2 0.3 0.4 0.5
[s]
-3
-2
-1
0
1
2
3
[V]

Figure 3. Example of a temporary overvoltage in per unit of nominal peak phase
voltage.
The expected TOV magnitude and duration is often a major
concern for surge arresters. MOV type surge arresters have
little effect upon temporary overvoltages that are below about
1.6 per unit. Silicon carbide type surge arresters are not
affected by TOV levels below the 60 Hz sparkover level.
However, if the TOV repeatedly exceeds the sparkover level,
then the multiple discharges may result in excessive energy
absorption and arrester failure.
V. BLACK START EXAMPLE
An example of a simulation of a black start is presented in
this section. The scenario is one where a fast starting gas
turbine driven generating unit is used as a black start unit to
start up a combined cycle power plant. Generator step-up and
auxiliary transformer at the black start plant and the plant to
be started are used for stepping up and stepping down voltage
in this example. These plants are connected by underground
HV cables. Tap setting in all transformers with tap adjustment
capability and the voltage reference set-point of the black start
unit were set to ensure terminal voltages at the large induction
motor units used in the black start were close to their nominal
value.
The black start plan begins with the across-the-line starting
of a 2500 HP motor. The motor performance during the
starting period is shown in Figure 4. Terminal voltage, motor
reactive power, electrical torque and motor slip are shown.
Note the dip in motor terminal voltage and that the demand
for reactive rises during the period following the lowest
voltage at the motor terminals. The air gap torque increases
significantly during the acceleration period, as expected, to
overcome the mechanical load torque that opposes developed
electromagnetic torque.
Finally, the dynamic response of the black start unit during
the starting of the large induction motor is shown in Figure 5.
The performance of the excitation system is shown as it works
to control the black start unit terminal voltage. Note the fast
response and large field forcing capability to pull up the
machine terminal voltage from the dip caused by the large
reactive power demand imposed by the starting motor. The
unit also sees a significant voltage rise caused by the rapid
reduction in reactive power as the motor locks in to its
operating speed. Electric power demand also increases during
this period as the motor is accelerating and moving towards its
steady state operating point.

VI. SUMMARY
It must be noted that restoration actions involve very
unusual conditions, especially for local generation. Such
factors as the ability to operate in islanded conditions with
stable frequency and voltage control, availability of
synchronizing equipment at key substations to permit
paralleling of separate sections and the validity of
assumptions on feasibility of generation to operate at unusual
points of their capability are important considerations in
assuring that restoration plans are realizable.
Restoration actions determined from a restoration study are
not necessarily what will actually be executed should a major
7
breakdown occur. They are based on a given set of
assumptions on available transmission, amount of cold load to
be picked up, etc. Actual conditions could differ from these
assumptions. The value of restoration studies is in
demonstrating the logic behind particular steps being taken,
i.e. the cause and effect reasoning behind the choice and
sequence of operator actions and the results of those actions
on the power system. With this understanding the operating
staff will be able to adapt to differences in the actual versus
assumed conditions.
This paper has described black-start operations and
discussed the studies that should be part of a black-start
planning process. In particular, it attempted to describe the
system dynamics and control aspects of the black start
process.
This overview should be very helpful to utility staff
involved in the development of restoration plans.
Development of thorough restoration plans and the testing of
those plans through simulation and drills will help to minimize
disruption of service to loads and the risk of damage to
equipment following power system outages.

Figure 4. Motor terminal voltage, reactive power, electric torque and slip

Figure 5. Black start unit terminal voltage, field voltage and speed
VII. ACKNOWLEDGMENTS
The authors would like to acknowledge F. P. deMello for his
guidance on the issues of power system restoration and
control. They would also like to acknowledge Daniel Durbak
for contributing to the section on transient overvoltages.
VIII. REFERENCES
[1] M. M. Adibi, Power system restoration, methodologies and
implementation strategies, in IEEE Series in Power Engineering, P.M.
Anderson, Ed., 2000
[2] M. M. Adibi and L. H. Fink, Power system restoration planning,, IEEE
Trans. Power Systems., vol. 9, no. 1, pp. 2228, Feb. 1994.
[3] R. J. Kafka, D. R. Penders, S. H. Bouchey, and M. M. Adibi, System
restoration plan development for a metropolitan electric system, IEEE
Trans. Power Apparatus and. Systems, vol. PAS-100, no. 8, pp. 3703
3713, Aug. 1981.
[4] W. C. Bryson and A. P. Hayword, Restoration of service on a
metropolitan power system, Transactions of the AIEE, vol. 59, 1940.
[5] IEEE Committee Report, System Restoration-Deploying the Plan:
Current Operational Problems Working Group, IEEE Trans. Nov. 1982,
pp. 4623-4671.
[6] M. M. Adibi and R. J. Kafka, Power system restoration issues, IEEE
Computer Application in Power, vol. 4, no. 2, pp. 1924, Apr. 1991.
[7] IEEE Committee Report, New approaches in power system restoration,
IEEE Trans. Power Systems., vol. 7, no. 4, pp. 14281434, Nov. 1992.
[8] J.W. Feltes, C. Grande-Moran, P. Duggan, S Kalinowsky, M. Zamzam,
V.C. Kotecha, F.P. de Mello, Some considerations in the development of
restoration plans for Electric utilities serving large metropolitan areas,
IEEE Trans. Power Systems, May 2006, Volume: 21, Issue: 2, pp. 909
915.
[9] NERC Standard EOP-005-1, System Restoration Plans,
ftp://www.nerc.com/pub/sys/all_updl/standards/rs/EOP-005-1.pdf.
[10] NERC Standard EOP-009-0, Documentation of Blackstart Generating
Unit Test Results, ftp://www.nerc.com/pub/sys/all_updl/standards/rs/EOP-
009-0.pdf.
[11] F. P. de Mello and J. C. Westcott, Steam plant startup and control in
system restoration, IEEE Trans. Power Systems, vol. 9, no. 1, pp. 93
101, Feb. 1994.
[12] R. R. Lindstrom, Simulation and field tests of the black start of a large
coal-fired generating station utilizing small remote hydro generation,
IEEE Trans. Power Systems, Volume 5, Issue 1, Feb. 1990, pp 162 168.
[13] M. M. Adibi, G. Adsunski,, R. Jenkins, P. Gill, Nuclear plant
requirements during power system restoration, IEEE Trans. Power
Systems, Volume 10, Issue 3, Aug. 1995, pp. 1486 1491.
1
.
1
5
0
0
0
.
9
0
0
0
0
TIME (SECONDS)
0.0
3.0000
6.0000
9.0000
12.000
15.000
18.000
21.000
24.000
27.000
30.000
SLIP
ELECTRIC TORQUE
TERMINAL VOLTAGE
ABSORBED REACTIVE POWER
1
.
0
4
0
0
0
.
9
9
0
0
0
TIME (SECONDS)
0.0
3.0000
6.0000
9.0000
12.000
15.000
18.000
21.000
24.000
27.000
30.000
FIELD VOLTAGE
TERMINAL VOLTAGE
SPEED DEVIATION
8
[14] M.M. Adibi, D. P. Milanicz, T.L.Volkmann, Remote cranking of steam
electric stations, IEEE Trans. Power Systems, Volume 11, Issue 3, Aug.
1996 pp. 1613 1618.
[15] D. Lindenmeyer, A. Moshref, M.C Schaeffer, A. Benge, Simulation of
the start-up of a Hydro Power plant for the emergency power supply of a
nuclear power station, IEEE Trans. Power Systems, Volume 16, Issue 1,
Feb 2001, pp.163 169.
[16] IEEE Committee Report, Special considerations in power system
restoration, IEEE Trans. Power Systems, Volume 7, Issue 4, Nov. 1992,
pp. 1419 1427.
[17] IEEE Committee Report, Special considerations in power system
restoration the second working group report, IEEE Trans. Power
Systems, Volume 9, Issue 1, Feb. 1994, pp. 15 21.
[18] B. Delfino, G.B. Denegri, M. Invernizzi, A. Morini, E. Cima Bonini, R.
Marconato, P. Scarpellini, Black-start and restoration of a part of the
Italian HV network: modeling and simulation of a field test, IEEE Trans.
Power Systems, Volume 11, Issue 3, Aug. 1996, pp. 1371 1379.
[19] M.M. Adibi, D.P. Milanicz, Protective system issues during restoration,
IEEE Trans. Power Systems, Volume 10, Issue 3, Aug. 1995, pp. 1492
1497.
[20] R.D Shultz, G.A. Mason, Blackstart utilization of remote combustion
turbines, analytical analysis and field test, IEEE Trans. Power Apparatus
and Systems, Volume PAS-103, Issue 8, Aug. 1984, pp. 2186 2191.
[21] F. P. de Mello, L. M. Leuzinger, and R. J. Mills, Load rejection
overvoltages as affected by excitation system control, IEEE Trans. Power
Apparatus and Systems, vol. PAS-94, no. 2, pp. 280287, Mar./Apr. 1975.
[22] J. E. McDonald, A. M. Bruning, and W. R. Mahieu, Cold load pickup,
in Proc. IEEE PES Winter Power Meeting, New York, 1979, Paper F79-
180-1.
IX. BIOGRAPHIES
James W. Feltes (M79, SM94) received his BSEE degree with honors from
Iowa State University in 1979 and his MSEE degree from Union College in
1990. He joined PTI, now Siemens PTI, in 1979 and is currently a Senior
Manager in the Consulting Department. At PTI, he has participated in many
studies involving planning, analysis and design of transmission and distribution
systems. He is an instructor in several of the courses taught by PTI. He is a
member of several IEEE committees, working groups, and task forces dealing
with power system stability and control. He is a Senior Member of the IEEE
and is a registered professional engineer in the State of New York.

Carlos Grande-Moran ( M74, SM89) received his Diploma Engineer degree
in Electrical and Mechanical Engineering from the University of El Salvador in
1974, his MEEE and MESE from Iowa State University and the University of
Virginia in 1976 and 1977, respectively and a Ph.D. in Electrical Engineering
from Iowa State University in 1982.
From 1990 to 2000, he worked in GE Power Systems Energy Consultants
and GE Power Systems Generator Engineering.
Dr. Grande joined PTI, now Siemens PTI, in 2000 and is currently an Senior
Staff Consultant working in the areas of power system dynamics, power system
planning, analysis and control of subsynchronous resonance in electric power
systems, and power system operations and control applications.