INSULATION COORDINATION

1. Breakdown Probability under Impulse Voltages

Suppose that the breakdown voltage of a gap under DC voltage is Vs. When an impulse voltage of a peak
value higher than VS is applied to a gap, as shown in Fig. 1, there is a certain probability but not a
certainty that breakdown will follow. For breakdown it is essential that the spark develops during the
interval of overvoltage [V(t)— VS] duration, i.e. the overvoltage duration must exceed the time lag [t < (t2
—t1)]. For a given impulse voltage waveshape the overvoltage duration will increase with the voltage
amplitude (Vp).

Because of the statistical nature of the time lags, when a given number of impulses of an amplitude Vp,
exceeding the static value Vs , is applied to a gap only a certain percentage will lead to breakdown. We
therefore obtain a breakdown probability P for each given applied maximum impulse voltage Vp

Figure 1. Breakdown under impulse voltage.

The breakdown probability for a given impulse voltage is obtained by applying a large number of
identical impulses (e.g. 100) and taking the ratio of the number of impulses that lead to breakdown to
the total number applied. Figure 2 illustrates an example of the breakdown probability distribution
function for impulse voltages of amplitude Vp . Vb-100 represents the 100% breakdown voltage, i.e. each
voltage application of this magnitude leads to breakdown. Vb-50 is the 50% breakdown voltage, i.e. one-
half of the applied voltages at this level lead to breakdown.
 Figure 2. Breakdown probability under impulse voltage. Vb0 impulse withstand level; Vb-50 50%
impulse breakdown; Vb-100 100% breakdown.

Figure 3. Gaussian probability distribution curve.

Vb0 represents the highest impulse voltage that does not lead to breakdown. It is known as the “impulse
withstand level” and is important in the design of insulations. The breakdown probability will usually
have a normal (Gaussian) distribution as shown in Fig. 3. It is necessary to make large number of
experiments in order to determine the exact values of Vb-0 and Vb-100 . The asymptotic approach of the
curve in Fig. 3 to the values Vb-0 and Vb-100 also shows this fact. However, once Vb-50 and standard
deviation (S) values are known, Vb-0 and Vb-100 can be approximately obtained from

Vb-0 = Vb-50 — 3S , Vb-100 = Vb-50 + 3S

2. Insulation coordination
Insulation coordination is the correlation of insulation of electrical equipment with the characteristics of
protective devices such that the insulation is protected from excessive overvoltages. Insulation
coordination comprises the adequacy of the dielectric strength of an insulating arrangement to the
voltages that can appear in the system.

The aim is to reduce (to an economically and operationally acceptable level) the probability that voltage
stresses would cause damage or hinder the continuity of service. In a substation, for example,
transformers, circuit breakers, bus supports, etc., should have insulation strengths sufficiently larger
than the overvoltages that they may experience.

Insulation coordination also takes into account the characteristics of the protective devices. Electric
systems insulation designers have two options: (i) choose insulation levels for components that would
withstand all kinds of overvoltages, (ii) consider and devise protective devices that could be installed at
the sensitive points in the system that would limit overvoltages there. The first alternative is
unacceptable especially for e.h.v. and u.h.v. operating levels because of the excessive insulation
required and the associated cost. Hence, there is a need to use protective devices. The actual
relationship between the insulation levels and protective levels is a question of economics.

3. Insulation level
‘Insulation level’ is defined by the values of test voltages which the insulation of equipment under test
must be able to withstand.

In the earlier days of electric power, insulation levels commonly used were established on the basis of
experience gained by utilities. As laboratory techniques improved, so that different laboratories were in
closer agreement on test results, an international joint committee was formed and was charged with the
task of establishing insulation strength of all classes of equipment and to establish levels for various
voltage classification. In 1941 a detailed document was published giving basic insulation levels for all
equipment in operation at that time. The presented tests included standard impulse voltages and one-
minute power frequency tests.

In today’s systems for voltages up to 245 kV the tests are still limited to lightning impulses and one-
minute power frequency tests. Above 300 kV, in addition to lightning impulse and the one-minute
power frequency tests, tests include the use of switching impulse voltages. Tables given below list the
standardized test voltages for ≤245 kV and above ≥300 kV respectively, suggested by IEC for testing
equipment.

Basic Lightning Impulse Insulation Level is the crest value of a standard lightning impulse for which the
insulation does not exhibit disruptive discharge. (also named Basic Insulation Level (BIL) or Lightning
Impulse withstand voltage). BIL is the impulse voltage which the insulation of any electrical equipment
for a given rated voltage must be able to withstand.

The values given in column 3 of Tables 7.2 and 7.3 correspond to impulse test voltage (1.2/50 usec) and
are ‘Basic Lightning Insulation Levels’ (BILs).

Basic Switching Impulse Insulation Level (BSL) (also named Switching Impulse withstand voltage) is the
rest value of a standard swithcing impulse for which the insulation does not exhibit disruptive discharge.
There are two approaches for studying the insulation coordination. The first one, which has been used
since the early days of insulation coordination, is designated as deterministic or conventional method.
The conventional approach is very conservative. Here the insulation is designed to provide a sufficiently
high withstand level so as to withstand any overvoltage that might be experienced. A considerably large
margin of protection is left between the electrical stress and the electrical strength.

The second one is the statistical approach. Here a certain number of flashovers are tolarated depending
on the desired performance. Present-day practices of insulation coordination rely on a statistical
approach. This approach requires a knowledge of the probability distribution of overvoltages and also a
knowledge of the breakdown probability of the insulation as a function of voltage.

In order to determine with some degree of confidence the probability distribution of switching
overvoltages on a system, it is necessary to perform a large number of computations. With today’s
availability of digital computers the distribution of overvoltages can be calculated. Amplitudes of
switching impulse voltages are influenced by the impedances and switching conditions of the system.
The values of the overvoltages depend on the relative timing of the switching event with respect to the
power frequency voltage wave. Generally, the closing time of the breaker is not synchronized with the
power frequency voltage wave shape. It is random, and each switching operation will result in a
different overvoltage.

One of the computer programs which can simulate the power system behavior under switching
conditions is EMTP. EMTP provides a statistical switch for overvoltage switching studies. The closing
time of a statistical switch is selected randomly by the program and the overvoltages due the closing of
the switch are calculated. The probability distribution of the closing time of the switch may be assumed
to be uniform or normal (Gaussian). Different switching cases can be evaluated randomly, with the
computer program, to determine the probability function of occurrence of switching overvoltages.

Statistical Method 1: Risk of Failure Calculation

For the purpose of coordinating the electrical stresses with electrical strengths, it is convenient to
represent the overvoltage distribution in the form of a probability density function (Gaussian
distribution curve as shown in Fig. 4) and the insulation breakdown probability by the cumulative
distribution function (shown in Fig. 4). The knowledge of these distributions enables us to determine the
‘risk of failure’.

Figure 4 Method of describing the risk of failure. 1. Overvoltage distribution–Gaussian

function. 2. Insulation breakdown probability–cumulative distribution)

If Vk is the kth value of overvoltage, the probability of occurrence of overvoltage is p0(Vk) du, whereas the
probability of breakdown is Pb(Vk). The probability that the gap will break down at an overvoltage Vk is
Pb(Vk) p0(Vk)du. For the total voltage range, the total probability of failure (risk of failure) is obtained as
∞
𝑅 = ∫ 𝑃𝑏 (𝑉𝑘 )𝑃𝑜 (𝑉𝑘 )𝑑𝑢
0

The risk of failure will thus be given by the shaded area under the curve R.
-- o --
Statistical Method 2: Statistical Safety Factor Calculation

In this simplified method, a ‘statistical overvoltage’ Vs is defined as shown in Fig.5(a). Vs is the voltage
value which is likely to be exceeded by 2% of all overvoltages. In Fig.5.a. only 2 percent (shaded area) of
all the overvoltages has an amplitude larger than Vs.

Fig. 5(b) illustrates another definition used in the simplified method. ‘Statistical withstand voltage’ Vw
(also named ‘statistical BIL’) is the crest value of a standard lightning impulse for which the insulation
exhibits a 90% probability of withstand (a 10% probability of failure).

The ‘statistical safety factor’ is defined to be the ratio of the ‘statistical withstand voltage’ Vw to
‘statistical overvoltage’ Vs.
-- o --

Similar to the concept of ‘risk of failure’, the ‘statistical safety factor’ concept permits a quantitative
evaluation of the risk of breakdown. The ‘statistical safety factor’ calculation is relatively simple since
the complete distribution functions for overvoltage and withstand of insulation is not required.

Figure 5 ‘Statistical overvoltage’ and ‘statistical withstand voltage’.

In Fig.6 functions PO(V) and Pb(V) are plotted for three different cases of insulation strength but keeping
the distribution of overvoltage occurrence the same. The density function PO(V) is the same in (a) to (c)
and the cumulative function Pb(V) is gradually shifted to the right (towards high values of V). This
corresponds to increasing the insulation strength by either using thicker insulation or material of higher
insulation strength. As a result of the relative shift of the two curves ( PO(V) and Pb(V) ) the ‘risk of
failure’ and the ‘statistical safety factor’ will vary.
 Figure 6 The statistical safety factor and its relation to the risk of failure

The relation between the ‘statistical safety factor’ and the ‘risk of failure’ is plotted in Figure 6(d). It is
clear that an increase in the statistical safety factor corresponds to a reduction of the risk of failure. At
the same time, this will cause an increase in the cost of insulation.

Non-Self-Restoring Insulations

The above treatment applies to self-restoring insulations. In the case of non-self-restoring insulations
the electrical withstand is expressed in terms of actual (deterministic) breakdown values.