POWER SYSTEM STABILITY

Prof. M. K. Verma
Department of Electrical Engineering
Institute of Technology
Banaras Hindu University
Varanasi (INDIA)
 Stability of LTI systems

• With zero initial conditions, the system is said to be stable if

for bounded inputs, outputs are bounded - BIBO stability.
• With finite initial conditions and zero inputs, the system is
said to be stable if outputs →0 as t→∞, it is the case of
asymptotic stability.
• Stability depends on the location of the roots of the
characteristic equation.
1. Left-half s-plane roots – stability
2. Simple roots on imaginary axis – marginal stability
3. Right-half s-plane roots – instability
4. Multiple roots on imaginary axis - instability
 Stability assessment of LTI systems

Transfer function approach

• Gives BIBO stability ignoring initial conditions.
• Asymptotic stability for the system with finite initial
conditions, can’t be assessed.
• Post-disturbance input/output behaviour is studied.
Selection of state variables is arbitrary.
• Transfer function approach may not be useful for stability
assessment of nonlinear systems.
 Stability assessment of LTI systems (contd.)
State space approach
• Gives BIBO stability as well as asymptotic stability.
• State vector is uniquely defined.
• Transfer function is also uniquely defined.
• In addition to post-disturbance input/output behaviour,
dynamics of state variables representing internal variables
of the system responsible for instability, may also be
studied.
• Thus state space approach more suitable for determining
dynamic behaviour of multi-input/multi-output systems.
• Stability of nonlinear systems may also be assessed by
state space approach.
 Stability assessment LTI systems (contd.)
• Routh-Hurwitz criterion.
• Root locus techniques.
• Nyquist stability criterion
• Lyapunov’s first method of stability assessment
 If eigen values of system state matrix have negative real
part, the system is stable.
 If eigen values of system state matrix have positive real
part, the system is unstable.
 If eigen values lie on the imaginary axis, it is not possible
on the basis of first approximation to say anything in the
general.
 Stability of nonlinear systems

• Lyapunov’s definition – If states remain within

Domain or Region of Attraction (ROA) of equilibrium
point, system is said to be stable else unstable.
• If states converge to equilibrium point as t→∞ it is the
case of asymptotic stability.
• If states remain within Domain without converging to
equilibrium point, a self-sustained oscillations of fixed
frequency and amplitude may occur, known as emergence
of periodic orbits or limit cycle.
 Lyapunov’s classifications of stability

• Local Stability – States lie within small neighborhood of

post-disturbance equilibrium point. Domain is small.
• Finite Stability - States remain within finite neighborhood
of post-disturbance equilibrium point. Domain is finite.
• Global Stability – States remain within large
neighborhood of equilibrium point. Domain is very large.
 Power System Stability?
• It is ability of power system to regain
operating equilibrium following
disturbance.
• Post-disturbance operating equilibrium may
be same as pre-disturbance operating
equilibrium or different.
 Classifications of Power System Stability

• Angular stability – Ability of the synchronous generators to

maintain synchronism between them following disturbance
– Can be examined by plotting rotor angle (voltage angle) vs. time.
- Short term phenomenon (period of interest is few seconds)
• Voltage stability – Ability of the power system to maintain
bus voltages within acceptable limits.
- Short term (period of interest is few seconds) or long term
(period of interest – several minutes or hours)
• Frequency stability – Ability of the power system to maintain
frequency within acceptable limits
- Short term or long term phenomenon
 Reliability, Security and Stability
• Reliability of a power system refers to probability of its
satisfactory operation over sufficiently long period. It represents
ability of power system to supply adequate power to consumers
on a nearly continuous basis with few interruptions.
• Security of a power system refers to its ability to survive
imminent disturbances (contingencies) without interruption of
customer service. It relates to robustness of the system to
imminent disturbances and, hence, depends on the system
operating conditions as well as the probability of occurrence of
imminent disturbances.
• Stability of power system refers to the continuance of intact
operation following a disturbance. It depends on the operating
condition and the nature of disturbance.
Modeling of Power System Components
for Stability Studies (continued)
Angular Stability
 Types of Angular stability
• Small Signal (Small Disturbance) Stability
- Mainly due to insufficient damping torque – results in local (plant
mode) oscillations and inter-area oscillations.
-Power system may be represented by linear model.
- Stability may be assessed by eigen theory.
• Transient (Large Disturbance) Stability
- Mainly due to insufficient synchronizing torque
- Power system to be represented by non-linear model. Non-linear
equations are to be solved by numerical methods such as Euler’s
method, predictor-corrector method, Runge-Kutta method,
numerical integration method.
- Can also be assessed by a direct approach using Transient Energy
Function
Rotor dynamics and swing equation

where,
 Angular stability assessment using swing equation

• For pre-disturbance condition,

- θ changes but δ remains constant.

• During disturbance,

where, = Instance of occurrence of disturbance

and, = Absolute angular position of rotor at instant of occurrence of
disturbance

-Both θ and δ change during disturbance.

 Angular stability assessment using swing equation
(contd.)
• If rotor speed comes back to synchronous speed at under post-
disturbance condition,

and, = Absolute angular position of rotor when synchronism is

recovered
=constant
- θ still changes but δ becomes constant.
• Thus, the ability of machine to regain synchronism following
disturbance may be determined by plotting δ vs. t. – a famous curve
that is popularly known as SWING CURVE.
• If rotor angle δ becomes constant following disturbance, machine is
stable else it is unstable.
 Angular stability assessment using swing equation
(contd.)
• As ability of the machine to regain synchronism depends on rotor
angle δ, this type of power system stability has been named as
ANGULAR STABILITY.
• Post-disturbance rotor angle δ may be same as its pre-disturbance
value or different in case of oscillation of δ under disturbance.
• Post-disturbance rotor angle δ will be different from its pre-
disturbance value in case of monotonous increase of rotor angle
during disturbance.
• Thus, post-disturbance equilibrium may not be same as pre-
disturbance equilibrium, for stability.
 Relationship between and

• It is important to find relative swing of one machine w.r.t. other

machines.
• If keeps on changing following disturbance and never
attains a constant value, machine-i will be out of synchronism from
machine-j.
• In order to examine rotor angular deviations of one machine w.r.t.
other machines, H-constant of all the machines must be expressed on
a common system MVA base (a typical value is 100 MVA) instead of
expressing them on individual machine MVA ratings.
• As H is inversely proportional to MVA base taken;
 Coherency of synchronous machines
• Synchronous machines that swing together (i.e. the machines which
have same change in rotor angle δ following disturbance) are called
coherent machines.
• A group of n coherent machines can be replaced by a single
equivalent machine whose swing equation is given by:
(in p.u.)
where, ,
, ,

• This reduces computational requirement as only one swing equation is

to be solved instead of solving n swing equations.
 Swing equations for noncoherent machines
• Let machines 1 and 2 be noncoherent. The swing equations for these
machines neglecting damping may be written as;

(in p.u.) (1)

(in p.u.) (2)

Multiplying (1) by and (2) by and subtracting resulting

equations;
(in p.u.)

where, ,

,
 Swing equations for noncoherent machines (contd.)
• This further reduces the number of swing equations to be solved.
• For stability, a machine must remain in synchronism with other
machines.
• Thus, study of variations of (i.e. relative swing between
machines) following disturbance is more important compared to
study of variations of rotor angles of individual machines.
 Angular stability studies - Classical approach

Assumptions:
• Mechanical power input to generator is assumed to be controlled
either manually or considered constant – Speed governor
dynamics is ignored.
• Each generator is represented by a constant voltage source
behind transient reactance – Excitation system dynamics is
ignored.
• Damping torque may be considered or neglected.
• All loads are considered as constant impedance loads. Thus, load
impedances may be included in matrix.
• Rotor angle δ is assumed to coincide with machine internal
voltage angle.
Small Signal Stability Studies for Single Machine
Infinite Bus (SMIB) System (classical
representation) – State Space Approach
Eδ V0

jX

G P G+jQG

PD+jQD
𝐸𝑉
𝑃 =𝑃 = sin 𝛿
𝑋
Δ𝑃 = Δ𝛿 = cos 𝛿 = 𝐾 Δ𝛿, where, 𝐾 = Synchronizing power coefficient
𝑑Δ𝛿 𝑑Δ𝜃
= = Δ𝜔 = Δ𝜔
𝑑𝑡 𝑑𝑡
2𝐻 𝑑 Δ𝛿
= Δ𝑃 − 𝐾 Δ𝛿 − 𝐾 Δ𝜔 𝑖𝑛 𝑝. 𝑢.
𝜔 𝑑𝑡
⋅ 0 Δ𝜔
Δ𝛿 0
Δ𝛿 𝐾 𝜔 𝐾 𝜔 𝜔
⋅ = + Δ𝑃
Δ𝜔 − − Δ𝜔
2𝐻 2𝐻 2𝐻
Small Signal Stability Studies for Single Machine
Infinite Bus (SMIB) System (classical
representation) – Transfer Function Approach

Δ𝛿(𝑠) 𝜔
𝑇𝐹 = =
Δ𝑃 (𝑠) 𝑠 + 2𝜔 𝜁𝑠 + 𝜔
𝐾𝜔
𝜔 =
2𝐻
1 𝐾
𝜁=
2 𝐾 2𝐻𝜔
 Methods for transient stability assessment

• Equal Area Criterion (SMIB system)

• Step-by-step (point-by-point) method
• Euler’s method
• Predictor-corrector method
• Runge-Kutta method
• Numerical Integration method
• Using Transient Energy Function (Lyapunov’s
method)
 Equal Area Criterion of Stability

• Equal Area Criterion is a graphical approach for

determining transient stability of SMIB system / two
machine system.
• It assesses stability of SMIB system / two machine
system without solving swing equation.
• The method of stability assessment is based on
‘Principle of Conservation of Energy’.
Equal Area Criterion of Stability (contd.)
 Equal Area Criterion of Stability (contd.)
• Integrating swing equation between two points 1 and 2 and
neglecting damping,

• Multiplying both sides by ,

• Above equation tells that kinetic energy gained between any

two points of power angle curve is equal to potential energy lost
between those points or vice-versa.
• It means the sum of K.E and P.E. remains constant during
acceleration/deceleration of rotor.
 Equal Area Criterion of Stability (contd.)

• If rotor speed equal to synchronous speed at two points 1 and 2,

Under this condition,

• If points a and e are considered points 1 and 2 where

,

• Similarly, as at points e and f, .

Equal Area Criterion of Stability (contd.)
 Ecrqual Area Criterion of Stability (contd.)
• As during fault,

• As per ‘Equal Area Criterion of Stability’,

 Ecrqual Area Criterion of Stability (contd.)

• Using ‘Equal Area Criterion of Stability,

• Unlike the case of three phase dead short circuit at generator

terminals, a literal form of solution is not possible in this case.
 Step-by-step method for obtaining swing curve

Assumptions:
• Accelerating power computed at the beginning of a sub-
interval is considered constant from the middle of
previous sub-interval to middle of sub-interval
considered.
• Rotor angular speed computed at the middle of a sub-
interval is considered constant throughout the sub-
interval.
Step-by-step method (contd.)
 Step-by-step method (contd.)
•
• Change in δ in interval,

• Change in δ in interval,

• Therefore,
(1)
where,
• Swing curve is plotted using (1).
• If any sub-interval begins at transition time (i.e. time of occurrence of
disturbance or time of disturbance clearance, average of two values of
is taken.
 Angular stability studies for multi-machine system
• For multi-machine system,

where is an element of modified matrix that

includes load impedance and is total number of generators present
in the system.
• Generated complex voltages are calculated using,

where, is an element of original matrix and n is

the total number of buses present in the system.
 Angular stability studies for multi-machine system
(contd.)
• Generator terminal voltages as well as other bus voltages,
are obtained from base case load flow
solution.
• In Classical Approach of angular stability studies, magnitude of
generated EMF, (i.e. ) are kept constant
corresponding to their values calculated using base case load flows.
• The initial value of rotor angle δ (at the instant of occurrence of
disturbance) is taken as angle of generated EMF (i.e.
) calculated using base case load flows.
• Load impedance at bus-i is calculated using,
 Small signal stability studies for multi-machine
system (Classical Approach)
• State equations are given by

where,

• Location of eigen values of state matrix that is size

determines small signal stability of the system.
 Transient stability studies for multi-machine system
(Classical Approach)
• Swing equations are given by,

where,
• Swing equation for each generator for each generator may be
converted to two first order differential equations,

• Above equations may be solved by numerical methods such as

Euler’s method, Predictor-corrector method, Runge-Kutta method,
Numerical integration method to obtain SWING CURVE.
 Preventive and Corrective measures against
Angular Instability
• Paralleling of transmission lines
• Use of transformers of lower leakage reactance
• Placement of series capacitors
• Placement of Flexible AC Transmission System
(FACTS) controllers
• Use of Power System Stabilizer (PSS) for damping
rotor oscillations
• Fast exciter with AVR feature
• Turbine valve control
• Use of fast circuit-breakers
Voltage Stability
 Voltage Stability and Voltage Security

• VOLTAGE STABILITY is the ability of the

system to maintain voltages within acceptable
limits.
• VOLTAGE SECURITY is the ability of a system
not only to operate stably, but also to remain
stable (as far as the maintenance of system voltage
is concerned) following any reasonable credible
contingency or adverse system change.
 Classification of Voltage Stability
• Large-disturbance voltage stability
-system’s ability to maintain steady voltages following
large disturbances
-examination of non-linear response of power system
-short term phenomenon (generally)

• Small-disturbance voltage stability

-system ability to maintain steady voltages under small
disturbances
-system equations can be linearized for analysis
-long term phenomenon (generally)
 Classification of Voltage Stability (contd.)

• Short-term voltage stability

-Involves dynamics of fast acting load components such as induction
motors, electronically controlled loads, and HVDC converters.
-The study period of interest is in the order of several seconds
-Analysis requires solution of appropriate differential equations.

• Long-term voltage stability

-Involves dynamics of slower acting equipments such as tap-
changing transformers, thermostatically controlled loads and
generator current limiters.
-The study period of interest may extend to several or many minutes.
-Static analysis can be used to estimate stability margins.
 Bifurcation and Voltage Stability

• The change in qualitative behavior of the power system is

known as bifurcation.
• Out of different types of bifurcations, Saddle-node-
bifurcation and Hopf bifurcation have been found to cause
instability in the power system.
• The system equilibrium disappears at the Saddle-node-
bifurcation point. One of the eigen values of the system
Jacobian matrix reaches origin. This bifurcation under all
limits is called Limit-induced bifurcation.
• At Hopf bifurcation one pair of complex eigen values of
the system Jacobian matrix reaches imaginary axis. This
may result into periodic orbits.
 Static and Oscillatory Voltage Stability

• STATIC VOLTAGE STABILITY is attributed to the

occurrence of saddle-node-bifurcation, when singularity
of Jacobian matrix results in disappearance of steady-
state solutions. At this point one of the eigen value of
Jacobian matrix reaches origin.
• OSCILLATORY VOLTAGE STABILITY is attributed
to occurrence of Hopf bifurcation, when one pair of
complex eigen values of system state matrix reaches
imaginary axis. This may result into emergence of
periodic orbits.
Static and Oscillatory Voltage Stability Margins
AB = Static voltage stability margin (Static loading margin)
AC = Oscillatory voltage stability margin (Dynamic loading
margin)

HB
Voltage SNB
(V)

A C B
Loading
factor (λ)
 Causes of Voltage Instability

• Difficulty in transmission of reactive power

under heavy loads
E δ V 0

jX

G Ps+jQs

Pr+jQr
 Causes of Voltage Instability (contd.)

• High reactive power consumption at

heavy loads

-The reactive power consumption in lines increases

quadratically with increase in line current, thus
restricting reactive power delivery to loads.
 Causes of Voltage Instability (contd.)

• Hitting of Q-limit of generators

-The reactive power generation capability of

synchronous generators is more limited
compared to real power generation capability
due to armature current, field current, end-
region heating and stability limits.
 Causes of Voltage Instability (contd.)

• Occurrence of contingencies
-Occurrence of contingencies may reduce voltage stability
margin.

• Presence of constant power loads

Constant
Contingency impedance

Intact case

V
Constant
power

λ
 Causes of Voltage Instability (contd.)

• Reverse operation of ON Load Tap-Changer

(OLTC)
Preventive and Corrective Measures for Voltage
Stability and Voltage Security Enhancement

• Placement of series and shunt capacitors

• Placement of Synchronous Condensers
• Placement of FACTS controllers
• Under-voltage load shedding
• Blocking of tap-changers under reverse operation
• Generation rescheduling
 Conclusions
• An LTI system is said to be stable if for bounded input, output is
bounded, and for zero input output becomes zero under steady-
state conditions.
• The Bounded Input Bounded output (BIBO) stability may be
determined theoretically using transfer function, or
experimentally by giving standard test signals (bounded inputs)
and observing response.
• A system may respond due to initial conditions present on its
elements even in the absence of inputs. However, such responses
must decay finally to zero in case of LTI systems if it is stable.
Such type of stability is called ‘Asymptotic Stability’.
• Asymptotic stability of LTI systems can’t be obtained by transfer
function approach.
• State space approach determines BIBO stability as well as
Asymptotic stability for LTI systems.
 Conclusions (contd.)
• An nonlinear stable may not converge to equilibrium point under
zero inputs. However, its states must remain within the Domain or
Region of Attraction (ROA) of post-disturbance equilibrium point.
• Lyapunov defined stability of nonlinear system (applicable for linear
systems, too) as its ability to remain within the domain of post-
disturbance equilibrium point. If states converge to post-disturbance
equilibrium point, it is the case of ‘Asymptotic Stability’.
• Based on size of the domain, Lyapunov defined three types of
stability (viz. Local stability, Finite stability and Global stability).
• Stability of a power system is its ability to regain certain operating
equilibrium (not necessarily same as pre-disturbance equilibrium)
following disturbance. If states converge to post-disturbance
equilibrium point, it is the case of Asymptotic Stability.
• Power system stability may be classified of three types (viz. Angular,
Voltage and Frequency stabilities).
 Conclusions (contd.)
- ‘Angular Stability’ is ability of synchronous machines present in the
system to regain synchronism following disturbance.
- ‘Voltage Stability’ is ability of the system to maintain bus voltages
within acceptable limits.
- ‘Frequency Stability’ is ability of the system to maintain frequency
within acceptable limits.
• Angular Stability is a short term phenomenon – period of interest is
few seconds.
• Voltage Stability as wells as Frequency Stability may be either short
term (period of interest –few seconds or long term phenomenon –
period of interest – several minutes or hours).
• Depending upon amount of disturbance, Angular Stability may be
classified in two types (viz. Small Signal Stability or Small
Disturbance Stability and Transient Stability or Large Disturbance
Stability).
 Conclusions (contd.)
• Depending upon amount of disturbance Voltage Stability may be
classified in two types (viz. Small Disturbance Voltage Stability and
Large Disturbance Voltage Stability).
• Based on time frame Voltage Stability may be classified in two types
(viz. Short term Voltage Stability and Long term Voltage Stability).
• Classical approach of angular stability studies considers only rotor
dynamics and ignores all other associated dynamics.
• Classical approach of angular stability studies assumes constant or
manually controlled mechanical power input to generator, represents
generators by constant voltage source behind transient reactance and
loads as of constant impedance types. Damping may be considered
or neglected.
• Swing equations of coherent generators may be clubbed together to
replace these machines by single equivalent machines.
 Conclusions (contd.)
• Small signal stability may be assessed by transfer function approach
or state space approach.
• Transient stability of SMIB system or two machine system may be
obtained without solving swing equations by ‘Equal Area Criterion
of Stability’.
• Transient stability of multi-machine system may be obtained by
numerical techniques such as Euler’s method, Predictor-Corrector
method, Runge-Kutta method, numerical integration method.
• Approximate swing curve may be obtained (thus, transient stability
may be assessed) by step-by-step (point-by-point) method.
• Remedial measures against angular stability includes placement of
series capacitors in lines, use of transformers with low leakage
reactance, paralleling of transmission lines, use of fast circuit
breakers.
 Conclusions (contd.)
• Voltage stability may be assessed by plotting real or reactive power
demand vs. bus voltage magnitude – a curve popularly known as
nose curve.
• Static Voltage Stability may be attributed to Saddle-node-
bifurcation, whereas, Oscillatory Voltage Stability may be attributed
to Hopf bifurcation.
• Causes of voltage instability include high reactive power
consumption in the network under heavy loads, difficulty in
transmission of reactive power particularly in stressed conditions,
occurrence of contingencies, dominance of constant power loads in
the system, use of OLTC under stressed conditions.