CHAPTER I

OVERVIEW

When there is a change in an electrical or an electromechanical system, such as a switching

action or a sudden change in load, the voltages, currents and speeds go through a transient
state before attaining steady-state conditions. The transients in electrical systems occur for a
short duration immediately after the change. The duration of the transients is very short and
depends on the system parameters such as resistances, inductances, capacitances, inertias of
moving parts etc. The systems keep changing until a steady-state is achieved. In any system,
transients may lead to instability, so controlling these transients increases system reliability
and their study permits the design of the appropriate protection. The steady state is
characterized by the voltages, currents and speeds being either constant or varying according
to a given law (e.g. sinusoidal).
There are several sources causing transients in electrical systems. Sources can be either
internal or external to the system. The table below outlines these sources.

Internal sources External Sources

 Capacitor switching  Lightning

 The operation of power semiconductor  External load removal or
switches connection
 Internal faults  Tap changing transformers
 Electrostatic discharge  Loose connections at the utility
 Relay operation end
 Circuit breaker or switchgear operation  External faults
 Load removal or addition  Human errors
 Arcing  Short circuits caused by animals
 Bad weather conditions
 Neighboring circuits

The effects of transients influence not only power lines but also signal and control lines. The
effects of transients can be divided into four types:

Intermittent interruptions - Results in a loss of data or corrupted data in data lines or

control lines.

Chronic degradation - The repetitive occurrence of transient events degrades the

components in an electrical circuit and diminishes performance, ultimately causing complete
circuit failure.
Latent failures - Similar to chronic degradation, except the component continues to operate
with derated performance.
Catastrophic failures - The component stops working immediately.

Systems are modeled by differential equations, and to solve these equations we can use two
methods. (i) the classical method and (ii) the method based on Laplace transform. It is
necessary to determine the initial conditions of the equations.
These conditions are determined from energetic considerations. Every state of a system
corresponds to an energy storage element (capacitor which stores electrostatic energy,
inductor which stores magnetic energy, rotating mass which stores kinetic energy). This
energy cannot change instantly as it requires an infinite power to do so (P = dE/dt -- > ∞ for
dE≠ 0 and dt=0). That is why the state of such a system will be the same just before (t = 0+)
and just after (t = 0+) the change. These conditions are specified from the following laws of
commutation.

First commutation law

The current i in an inductor L cannot change instantly as the emf induced e would be infinite,
that is
e = L di/dt  ∞ if di≠0 for dt=0 , thus p = e.i  ∞

Second commutation law

The voltage u of a capacitor C cannot change instantly as the current i would be infinite, that
is
i = C du/dt  ∞ if du≠0 for dt=0 , thus p = u.i  ∞

Third commutation law

The speed of rotation ω of a body of moment of inertia J cannot change instantly as the
accelerating torque would be infinite, that is
T = J dω/dt  ∞ if dω ≠0 for dt=0 , thus p = C. ω  ∞

a) Classical method:

The general solution of a linear differential equation is the sum of the homogeneous equation
solution and the solution due to the excitation signal (second member of the equation). The
integration constants are deduced from the initial conditions which are determined from the
commutation laws.

Example
For the following differential equation
find y(t) with y(0) and y’(0) given.

The solution of the homogeneous equation

is

where

To find r1 and r2 we solve the characteristic equation r2 + A r + B = 0

If r1 and r2 are real and different, yh = C1 exp (r1 t) + C2 exp (r2 t)

If r1 and r2 are complex (α ± j β), yh = eα t (C1 cos (β t) + C2 sin (β t)

If r1 = r2 = r, yh = er t (C1 + C2 t)

The specific solution depends on the forcing function f(t).

b) Méthod based on Laplace transform:

This method consists on transforming equations containing derivatives and integrals into
algebraic equations easier to solve. It is applicable to linear time invariant systems.

To this aim we define the Laplace transform F(s) of a function of time f(t) as
∫

where s=α+jω,
has a dimension of a frequency, and is called complex frequency.
We can show that
L ( f’(t) ) = p F(p) – f(0)

L ( f’’(t) ) = p2 F(p) - p f(0) – f’(0)

L ( ∫ f(t) dt ) = F(p) / p

As an example, the above differential equation

becomes,

p2 Y(p) - p y(0) – y’(0) + A (p Y(p) – y(0) ) + B Y(p) = F(p)

Knowing y(0), y’(0) and the function F(p), we can calculate algebreaicallyY(p). The
function y(t) can be deduced from inverse transforming Y(p) using tables.