Insulation Coordination

Station Lightning Insulation Coordination

Part One of the series. Selecting BILs for substation equipment against lightning surges entering from the connected lines: the overall procedure, single-phase station modelling in EMTP®, why arrester location and separation distance govern the protection, how surge and power-frequency voltage combine, and the crest-voltage calculation at the transformer — the open-circuit \(2E_A\) result, the K₂ arrester correction, arrester lead length, transformer capacitance, and clearance and altitude correction.

Reading time ≈ 45 min · Part One of the series

Section 1

What Station Lightning Insulation Coordination Does

Station lightning insulation coordination is the process of selecting suitable insulation levels for substation equipment when lightning surges enter the station from the connected overhead lines.

The main insulation level here is the BIL (Basic Lightning Impulse Level) — the standard lightning-impulse withstand level, associated with the standard \(1.2/50\ \mu\text{s}\) impulse waveshape. Different equipment may have different requirements: transformers and their bushings, circuit breakers, disconnectors, bus supports, surge arresters, GIS/AIS equipment, open bus sections and line entrances.

The purpose is to ensure the surge voltages appearing at these equipment terminals stay below the insulation strength, with an acceptable margin or reliability target.

The whole study follows one sequence: define the incoming lightning surge and select the arrester rating; represent the station layout by its bus lengths, travel times and equipment locations; calculate the surge voltages at the transformer, breaker, bus junctions and open points; compare those voltages with the insulation strength of each item; and finally select the next available standard BIL and the required clearances. The sections below follow that order — understanding it before the equations begin makes the rest of the page easier to follow.

What this page teaches

the overall procedure for station lightning insulation coordination;
how the station is modelled in a digital transient program (EMTP® / ATP);
why arrester location and separation distance govern the protection;
how surge voltage and power-frequency voltage are combined;
the crest-voltage calculation at the transformer, including capacitance;
clearance and altitude correction, and when to shorten the procedure.

Section 2

The Starting Point of the Study

Before station BILs can be selected, two things are normally already known: the incoming surge to be applied to the station, and the arrester rating / protective level. Once these are chosen, the selection of equipment BILs can begin.

The station type and layout are usually known too — single-bus, double-bus, ring bus, breaker-and-a-half, GIS, distribution or transmission. And the candidate BILs are often limited: for transformers perhaps 1–3 possible values, for other equipment often only 2, and for circuit breakers frequently a single value fixed by rating and standard practice.

So in practice, BIL selection is usually a checking process rather than a free optimisation. The available equipment BILs are limited to a small number of standard values, and the purpose of the study is to check whether the calculated surge stress can be coordinated with those available insulation levels.

Section 3

Why "Station BIL" Tables Should Not Be Used Blindly

Some standards give typical station insulation levels. These can be conservative and are widely used, especially at lower voltages — but they should not be used without checking the actual layout and surge behaviour. A station BIL table is useful for preliminary guidance, but it does not represent the actual station: it does not account for the arrester-to-transformer distance, the bus lengths, open points, the transformer capacitance, the arrester lead length or the incoming-surge steepness. So it should not be used as a replacement for a station-specific insulation-coordination check.

Why a table is not enough

The surge voltage at each item of equipment depends strongly on arrester location, separation distance, reflections, station layout, number of connected lines, transformer capacitance, arrester lead length and incoming-surge steepness. A general table is useful for initial guidance, but it should not replace a station-specific study where the consequence of failure matters — especially for transformers, whose insulation is non-self-restoring and expensive to replace.

Section 4

The Overall Procedure

The procedure is a sequence of engineering checks. The five core steps are summarised here and then taken one at a time:

The five steps

Evaluate opened-circuit-breaker protection — decide if arresters are needed across an open breaker.
Select the incoming surge — from a reliability criterion (MTBF).
Select candidate BILs — from standard equipment insulation levels.
Evaluate contingency conditions — lines out of service, bus open, etc.
Select arrester rating and preliminary location.

Section 5

Step 1 — Opened-Breaker Protection

The first step is to evaluate whether protection is needed across an opened circuit breaker, which can see high voltage across its contacts during surge conditions. If arresters are required for opened-breaker protection, they should be included in the initial station study — because adding arresters later changes the surge-voltage distribution throughout the station. Opened-breaker protection should not be treated as an afterthought.

Section 6

Step 2 — Select the Incoming Surge (MTBF)

The incoming surge is the lightning surge that enters the station from the connected line. It is selected from a reliability criterion such as the MTBF (Mean Time Between Failures). It should represent a severe but credible surge — not simply the maximum possible stroke — linked to the desired reliability level:

\[ \text{higher reliability target} \;\Rightarrow\; \text{more severe design surge} \]

and a lower acceptable failure rate implies a higher required insulation strength or better protection.

Section 7

Step 3 — Candidate BILs and Chopped-Wave Tests

Candidate BILs are taken from standard equipment insulation levels. For circuit breakers at high system voltage, two switching-impulse levels may be listed — one for the closed breaker and one for the opened breaker; at \(362\ \text{kV}\) and above this distinction becomes important. Circuit breakers are also subjected to chopped-wave tests:

\[ 1.15 \times \text{BIL} \;\text{chopped at}\; 3\ \mu\text{s} \qquad\qquad 1.29 \times \text{BIL} \;\text{chopped at}\; 2\ \mu\text{s} \]

These matter because transformer and equipment insulation may experience fast-front and chopped waves from arrester operation or flashover. For disconnecting switches, the switching-impulse withstand across the opened switch is about 10% higher than the value to ground — logical, because open contacts can be stressed by voltage appearing on both sides.

Section 8

Step 4 — Contingency Conditions

During thunderstorms the normal assumption is all lines in service. Contingencies — one or two lines out, a bus section open, a transformer disconnected, a temporary switching or maintenance configuration — often have low probability during a storm and are not always studied in detail. But if a contingency is likely or operationally important, it should be evaluated.

The reason: the number of connected lines affects surge behaviour. Connected lines provide additional surge paths and change reflections; with fewer lines connected, some locations may see higher surge voltages. Station configuration is therefore part of lightning insulation coordination.

Section 9

Step 5 — Arrester Rating and Preliminary Location

The arrester rating is selected from application requirements: maximum continuous operating voltage, temporary-overvoltage duty, energy duty, protective level, system earthing, and switching- and lightning-surge duty. The preliminary location is then chosen — line-entrance arresters at the line entrances; otherwise priority normally goes to transformer protection. Simple low-voltage stations may put arresters on the bus; large breaker-and-a-half stations often place the arrester near the transformer (e.g. on the transformer bus).

The guiding principle

The arrester protects best where it is electrically close to the equipment being protected. Electrically close does not only mean physically close — the effectiveness depends on the complete surge path between the arrester and the equipment. A long arrester lead, an indirect connection or a long bus section all increase the effective travel time, which lets travelling-wave reflections raise the equipment voltage above the arrester protective level.

Section 10

Why Arrester Location Is Critical

The arrester limits the voltage at its own terminals. It does not instantly force the same voltage at every point in the station. A lightning surge travels along the bus as a wave: if the transformer, breaker or open bus end is some distance from the arrester, the wave needs time to reach it. During that time, reflections occur at open ends, capacitances and changes in surge impedance, and the incident and reflected waves can add together at the equipment terminal. So the transformer or breaker voltage can be higher than the arrester voltage:

\[ V_{\text{equipment}} \neq V_{\text{arrester}} \]

The difference grows when the separation distance is larger, the surge steepness is higher, the arrester lead is longer, the equipment capacitance produces oscillation, or the terminal behaves like an open circuit. This is exactly why station insulation coordination must consider the station layout — not just the arrester rating.

Section 11

Digital Transient Modelling (EMTP® / ATP)

The recommended method is to model the station in a digital transient program such as EMTP® or ATP, so surge voltages can be calculated throughout the station. A single-phase model is generally sufficient for lightning surge studies, because the behaviour is dominated by travelling-wave phenomena on one phase rather than balanced three-phase power-frequency behaviour.

EMTP® is preferred because the station behaves as a travelling-wave network during a lightning surge: the bus sections, transformer capacitance, arrester characteristics, open points and reflections all shape the voltage waveshape, and these effects are difficult to represent accurately with a simple static calculation.

Section 12

Modelling the Buses, Transformers, Equipment and Surge

Buses

Station buses are modelled as distributed-parameter lines, each described by a surge impedance, a length and a propagation velocity. For an air-insulated station the velocity is about \(300\ \text{m}/\mu\text{s}\) (\(\approx 1000\ \text{ft}/\mu\text{s}\)), so the travel time for a bus of length \(l\) is:

\[ T = \frac{l}{v} = \frac{l}{300} \qquad (l\text{ in m},\; T\text{ in }\mu\text{s}) \]

Transformers

Transformers are modelled by their surge capacitance to ground, typically between \(1\ \text{nF}\) and \(10\ \text{nF}\) (use \(2\)–\(4\ \text{nF}\) if unknown). This capacitance affects the surge waveshape and crest at the transformer terminal — the maximum crest may occur near \(4\ \text{nF}\), or near \(1\ \text{nF}\) or less when separation distances are long. So transformer capacitance should not be ignored.

Other equipment

Breakers, disconnectors, instrument transformers, bus supports and bushings can also be modelled by surge capacitances, but in many conservative studies these are neglected except in special cases — one important exception being GIS disconnecting-switch operation, where high frequencies and equipment capacitance matter more.

The incoming surge

The incoming surge is applied to the model, assumed to arrive from a distant struck point (so reflections from the struck point are ignored). It is described by a steepness \(S\), a crest voltage \(E\) and a long / effectively infinite tail, and is assumed to ride on a power-frequency voltage \(V_{pv}\) of opposite polarity — a conservative assumption, because it increases the stress to ground.

Section 13

Surge-Voltage Evaluation and BIL Selection

With the surge applied, voltages are measured throughout the station — at transformer, breaker and disconnector terminals, bus supports, open bus points, arrester terminals and line entrances. The calculated voltages usually do not have the standard \(1.2/50\ \mu\text{s}\) waveshape, so they must be related to the standard insulation-withstand basis. This relating of a non-standard wave to a standard withstand is one of the key difficulties of the whole subject.

Section 14

Self-Restoring and Non-Self-Restoring Insulation

Self-restoring insulation

Self-restoring insulation recovers after flashover — air gaps, external clearances, some external support surfaces. It can be evaluated by methods ranging from subjective judgement to the leader progression model (the development of a leader discharge in air) and the destructive effect method (derived from it, for non-standard waveshapes).

Non-self-restoring insulation

Non-self-restoring insulation does not recover after breakdown — the key example is transformer internal insulation, whose failure is permanent and serious. Here only a subjective evaluation is possible, and the crest voltage is usually compared with the \(3\ \mu\text{s}\) chopped-wave test, because the transformer may see fast-front or chopped surges in service.

Margins — and why they differ

For non-self-restoring insulation a margin of typically 15% to 20% is recommended: \(\text{BIL} \ge (1.15\text{ to }1.20)\times\text{equivalent surge stress}\), because transformer failure is permanent and costly. For self-restoring insulation, fixed margins are more questionable — it is often better to use a higher reliability target (a higher MTBF, hence a more severe design surge) than to apply an arbitrary voltage margin.

Section 15

Clearance Estimation

Clearances are estimated from the highest equivalent crest voltage of the standard \(1.2/50\ \mu\text{s}\) wave, divided by a negative-polarity CFO gradient. Typical gradients are \(540\) to \(750\ \text{kV/m}\) (the higher value for a rod-plane gap), with a suggested value of \(605\ \text{kV/m}\):

\[ d = \frac{V_{\text{eq}}}{G_{\text{CFO}}} = \frac{V_{\text{eq}}}{605} \]

\(d\): required clearance (m)
\(V_{\text{eq}}\): equivalent crest voltage (kV)
\(G_{\text{CFO}}\): CFO gradient (kV/m), depends on gap configuration

This is an approximate estimate, to be applied with engineering judgement.

Section 16

Altitude Correction

BILs are defined for standard sea-level conditions. At altitude the air density is lower, which reduces the strength of external insulation, so BILs and clearances must be corrected. Strength decreases approximately linearly with the relative air density \(\delta\), so the required values are divided by \(\delta\):

\[ \text{BIL}_{\text{alt}} = \frac{\text{BIL}_{\text{sea level}}}{\delta} \qquad\qquad d_{\text{alt}} = \frac{d_{\text{sea level}}}{\delta} \]

Because \(\delta < 1\) at altitude, the required BIL or clearance increases.

Section 17

If the Required BILs or Clearances Are Excessive

Add arresters inside the station

Additional arresters can be installed closer to the protected equipment, reducing separation distance and limiting local surge voltages — shorter arrester-to-equipment distance gives lower surge voltage at the equipment. The cost: extra arresters need space, money, and coordination with operating voltage and energy duty.

Improve the lightning performance of adjacent lines

Alternatively, reduce the severity of the incoming surge (while keeping the target MTBF) by improving the line or towers near the station: reduce tower footing resistance, improve tower earthing, add overhead ground wires near the station, install line-entrance arresters, or use gaps at the first and second structures for wood-pole lines without ground wires. Lower footing resistance reduces backflashovers near the station; ground wires reduce direct exposure of the phase conductors — so the surge entering the station becomes less severe.

Section 18

When to Shorten the Procedure, and the Role of Hand Methods

For new voltage levels or unfamiliar designs, the full procedure should be followed. With enough experience it can be shortened for similar stations, studying only the significantly different features — for standardised distribution layouts a detailed study may be needed only once. Even so, the recommended approach remains to perform the study in EMTP® or ATP, because even small stations model easily and the voltages compute quickly and accurately.

Simplified hand-calculation methods are still useful — for quick checks, early-stage design, protection and BIL estimates, sanity checks on computer results, and simple low-voltage layouts. But they are not always truly simple: travelling-wave behaviour, reflections, arrester operation and equipment capacitance cannot be reduced to a trivial rule without losing accuracy. Their purpose is to support engineering judgement and give independent checks — not to replace EMTP®.

To position it clearly: the simplified method suits early estimates, lower-voltage standardised stations and sanity checks — but for important high-voltage substations, large AIS layouts, GIS layouts or stations with unusual arrangements, a detailed transient study should be preferred.

Section 19

Stress at Equipment — the Crest-Voltage Circuit

The stress calculation begins with a general station circuit for an \(n\)-line station. The other connected lines are represented by an equivalent resistor at the arrester tap point:

\[ R = \frac{Z}{n-1} \]

\(R\): equivalent resistance of the other lines at the arrester tap
\(Z\): surge impedance of a connected line or bus
\(n\): number of connected lines

This represents the fact that other lines provide paths for surge waves to travel away from the station — more connected lines generally reduce the surge build-up at some points. Two simplifying assumptions are made: the phase-conductor and bus surge impedances are assumed equal (their difference is usually not significant here), and travel times come from physical distance and velocity, \(T = l/300\) for an AIS. The important travel times are arrester-to-transformer, the arrester lead, and arrester-to-breaker/bus junction.

Section 20

Transformer Representation

In the simplified circuit the transformer is represented by its surge capacitance to ground \(C_T\), ranging from \(1\ \text{nF}\) to more than \(10\ \text{nF}\). The maximum crest at the transformer may occur for \(C_T \approx 4\ \text{nF}\), or for \(C_T \le 1\ \text{nF}\) when separation distances are long. The "transformer" in the circuit is generic — it may represent an actual transformer, an open line end, an opened bus end or any open termination. With no transformer present, \(C_T = 0\) and the model becomes the open-circuit case.

Section 21

Incoming-Surge Definition

The incoming surge has a crest voltage \(E\), a steepness \(S\) and an effectively infinite tail, arriving from a distant struck point and superimposed on a power-frequency voltage \(V_{pv}\) of opposite polarity. The calculation is deliberately separated: first compute the surge component in the station, then combine the power-frequency voltage to obtain the voltage to ground.

Section 22

Voltage Definitions at Each Point

Each physical point has two related voltages: the surge component and the voltage to ground (including the power-frequency effect). The arrester is the special case, because its surge calculation must include \(V_{pv}\) so the correct arrester current flows.

Table 1 — Surge voltage and voltage-to-ground at each point.

Point	Surge Voltage	Voltage to Ground	Relation
Arrester	\(E_A\)	\(E_d\) (discharge voltage)	\(E_A = E_d + V_{pv}\)
Breaker	\(E_v\)	\(E_b\)	\(E_b = E_v - V_{pv}\)
Transformer	\(E_T\)	\(E_t\)	\(E_t = E_T - V_{pv}\)
Arrester-bus junction	\(E_I\)	\(E_j\)	\(E_j = E_I - V_{pv}\)

Section 23

Why Surge and Power-Frequency Voltage Are Separated

The surge calculation is performed using the travelling-wave surge component — but the insulation is stressed by the voltage to ground. Because the lightning surge is assumed to be superimposed on an opposite-polarity power-frequency voltage, the surge component and the power-frequency component must be combined correctly — which is exactly why the notation distinguishes between the surge voltage and the voltage to ground. Here the surge rides on a power-frequency voltage of opposite polarity, so the voltage to ground is found by subtracting \(V_{pv}\) from the surge voltage, e.g. \(E_t = E_T - V_{pv}\).

A common mistake

Do not compare the surge component directly with the equipment BIL without accounting for the power-frequency reference used in the arrester calculation. The insulation sees the voltage to ground, not the bare surge component.

Section 24

How the Equations Are Developed

The equations are built up in stages, introducing assumptions progressively. The first assumption is zero arrester lead travel time (\(T_A = 0\)) — no lead length initially. The second is a constant-voltage arrester, which holds a constant voltage independent of current; not fully realistic (real arrester voltage depends on discharge current), but a useful starting point. The development then adds the actual arrester voltage–current characteristic, the arrester lead length, and the transformer capacitance.

Section 25

Transformer as an Open Circuit — \(E_T = 2E_A\)

The first case takes \(T_A = 0\) and \(C_T = 0\) — no lead length, the transformer as an open circuit. An open end behaves like a reflecting point for voltage waves: when a surge reaches it, the reflected voltage has the same polarity as the incoming voltage, so the two add together. This is why, in the ideal open-circuit case, the voltage at the open end can approach twice the arrester voltage. The maximum voltage at the transformer can then reach:

\[ E_T = 2E_A \qquad\text{at}\qquad \frac{S\,T_T}{E_A} = \frac{n+1}{4} \]

\(E_T\): surge voltage at the transformer
\(E_A\): surge voltage at the arrester
\(S\): surge steepness
\(T_T\): travel time from arrester to transformer
\(n\): number of connected lines

What \(E_T = 2E_A\) really means

This does not mean the arrester failed — it clamps voltage at its own location. But the transformer is separated by a travelling-wave path, and at the open end the wave reflects with the same polarity, so the incident and reflected waves add and the voltage can double. \(E_T > E_A\) is possible, reaching \(E_T = 2E_A\) in the ideal open-circuit case — the fundamental reason separation distance matters.

Section 26

From Constant-Voltage Arrester to the \(K_2\) Correction

The initial equations assume a constant-voltage arrester, \(E_A = \text{constant}\) regardless of current — a conservative assumption. Real metal-oxide arresters have a voltage–current characteristic (discharge voltage rises with current), and the interaction with travelling waves can reduce the calculated transformer voltage compared with the constant-voltage model, with a considerable reduction especially for \(n = 1\). The constant-voltage equations are kept but multiplied by a correction factor \(K_2\):

\[ E_{T,\max} = 2K_2 E_A \]

Table 2 — Correction factor \(K_2\) versus number of connected lines.

Number of lines \(n\)	\(K_2\)
1	≈ 0.91
4	≈ 0.97

So \(E_{T,\max} = 2K_2 E_A\) is a little below the ideal \(2E_A\).

Section 27

Including Arrester Lead Length

When the arrester lead length is included, the effective travel time becomes \(T_T + T_A\), where \(T_A\) is the arrester lead travel time. Longer leads increase the voltage stress at the protected equipment.

The arrester lead is part of the surge path, not a minor detail. During a fast lightning surge even a short lead has inductance and travel time, so the effective protection point is not the arrester block itself but the complete arrester connection. Long or indirect leads therefore raise the voltage seen by the protected equipment.

Keep arrester leads short

Arrester leads must be as short and direct as possible. Even a physically short lead can be significant when the surge front is very steep.

Section 28

The Transformer Capacitance Effect

In plain terms: at the first instant of a fast surge, the transformer capacitance does not allow its voltage to change immediately, so it initially behaves differently from a simple open end. As the capacitance charges, the reflected wave changes. This interaction between the incoming wave, the arrester and the transformer capacitance can increase the transformer crest voltage and change its time to crest.

Looking more closely: when a fast surge reaches the capacitance the capacitor at first behaves almost like a short circuit (its voltage cannot change instantly), giving an initial negative reflection; as it charges it behaves more like an open circuit, giving a positive reflection. The interaction of the incoming surge, arrester operation, transformer capacitance and reflections between transformer and arrester can produce higher crest voltages. In the idealised example the transformer voltage could theoretically approach:

\[ E_T \approx 2E_A + 0.348\,S Z C \]

\(S\): surge steepness
\(Z\): surge impedance
\(C\): transformer capacitance

A conceptual expression, not a design formula

This expression is included only to show the physical reason why transformer capacitance can increase the crest voltage. Treat it as a conceptual explanation, not a universal design equation — for design checks, use the regression equations below or an EMTP® simulation.

This theoretical maximum does not normally occur exactly — the surge has a finite tail, the arrester is not a perfect constant-voltage device, the voltage builds up over time, and reflections modify the waveshape. In practice the maximum transformer voltage is often in the range \(2.2E_A\) to \(2.6E_A\) — higher than the simple open-circuit \(2E_A\), which is why transformer capacitance matters.

Section 29

A Regression Equation for the Transformer Voltage

Because exact analytical equations with capacitance are difficult, a curve-fitting (regression) approach is used. The transformer voltage ratio \(E_T/E_A\) is expressed as a function of a single dimensionless parameter:

\[ \frac{E_T}{E_A} = f\!\left(\frac{S\,(T_T + T_A)}{E_A}\right) \]

\(E_T,\ E_A\): surge voltage at transformer and at arrester
\(S\): incoming-surge steepness
\(T_T\): arrester-to-transformer travel time
\(T_A\): arrester lead travel time

A larger \(S(T_T+T_A)/E_A\) gives a larger transformer voltage. This one parameter combines surge steepness, separation distance, arrester lead length and arrester protective voltage — the four things that drive the stress.

Section 30

Worked Example — 138 kV Transformer Voltage

Take \(E_d = 267\ \text{kV}\), \(V_{pv} = 80\ \text{kV}\), \(S = 1400\ \text{kV}/\mu\text{s}\), a 12 m arrester-to-transformer distance, \(T_A = 0.01\ \mu\text{s}\) and \(T_T + T_A = 0.05\ \mu\text{s}\). First the arrester surge voltage:

\[ E_A = E_d + V_{pv} = 267 + 80 = 347\ \text{kV} \]

\[ \frac{S(T_T+T_A)}{E_A} = \frac{1400 \times 0.05}{347} = 0.20173 \;\;\Rightarrow\;\; \frac{E_T}{E_A} = 1.5903 \quad(n=1) \]

\[ E_T = 1.5903 \times 347 = 552\ \text{kV} \;\;\Rightarrow\;\; E_t = E_T - V_{pv} = 552 - 80 = 472\ \text{kV} \]

So the transformer experiences about \(472\ \text{kV}\) to ground in this example.

Section 31

Worked Example — Maximum Permissible Separation Distance

The same method runs in reverse. If the maximum permissible transformer voltage to ground is \(E_t = 504\ \text{kV}\), then:

\[ E_T = E_t + V_{pv} = 504 + 80 = 584\ \text{kV} \;\;\Rightarrow\;\; \frac{E_T}{E_A} = \frac{584}{347} = 1.683 \]

\[ T_T + T_A = 0.07476\ \mu\text{s} \;\;\Rightarrow\;\; d = v(T_T + T_A) = 300 \times 0.07476 = 22.4\ \text{m} \]

So the maximum arrester-to-transformer separation is about \(22.4\ \text{m}\) — showing how the method defines an allowable separation distance.

Section 32

Waveshape and Time to Crest at the Transformer

Including transformer capacitance makes the crest voltage larger, the oscillating frequency lower, and the waveshape different. This matters because transformer insulation strength depends not only on crest voltage but also on time to crest and waveshape — for non-self-restoring insulation the evaluation is not a simple crest comparison, and the waveshape should be considered, especially against chopped-wave tests.

To estimate the time to crest, the section between arrester and transformer is reduced to an inductance \(Z(T_T+T_A)\) and a capacitance \(\tfrac{T_T+T_A}{Z}\), with the transformer capacitance \(C_T\). The capacitor voltage in the oscillatory circuit is:

\[ E_c = V\,(1 - \cos\omega t) \]

The time to crest is about half the oscillation period. For \(C_T = 2\)–\(4\ \text{nF}\) and \(T_T = 0.05\)–\(0.10\ \mu\text{s}\), the estimated time to crest exceeds the actual value by about \(2\%\) to \(14\%\) — i.e. the estimate is conservative. The time to crest matters because the standard BIL is based on a standard waveshape, while the actual terminal voltage may be faster or chopped and stress the insulation differently — so both crest voltage and time to crest should be considered, especially for non-self-restoring insulation.

Section 33

Practical Engineering Lessons

The key lessons

Arrester rating alone is not enough — the equipment voltage can exceed the arrester voltage through travelling-wave effects.
Arrester location is critical — place it electrically close to the protected equipment, normally near the transformer terminal or bus.
Transformer capacitance must be considered — a transformer is not simply an open end; \(C_T\) raises the crest and changes the waveshape.
Separation distance must be limited — the design may need a maximum allowable arrester-to-transformer distance.
Arrester leads must be short — lead length adds to the effective travel time.
EMTP® / ATP is preferred — hand methods are for quick estimates and sanity checks, not a substitute for a proper transient study.

Section 34

Summary of Key Equations

Equation Summary

Travel time (AIS)

\( T = \dfrac{l}{300} \)

Other-lines resistance

\( R = \dfrac{Z}{n-1} \)

Arrester surge voltage

\( E_A = E_d + V_{pv} \)

Transformer voltage to ground

\( E_t = E_T - V_{pv} \)

Open-circuit maximum

\( E_{T,\max} = 2E_A \)

With arrester correction

\( E_{T,\max} = 2K_2 E_A \)

Dimensionless parameter

\( \dfrac{S\,(T_T + T_A)}{E_A} \)

Capacitance contribution

\( E_T \approx 2E_A + 0.348\,SZC \)

Clearance estimate

\( d = \dfrac{V_{\text{eq}}}{605} \)

Altitude correction

\( \text{BIL}_{\text{alt}} = \dfrac{\text{BIL}}{\delta} \)

Section 35

Reader Should Remember

Station lightning insulation coordination is not just selecting a BIL from a table. The actual surge voltage at station equipment depends on the incoming-surge severity, the arrester protective level and location, the arrester lead length, the separation distance, the number of connected lines, reflections, the transformer capacitance, the station layout, altitude and the insulation type.

The single most important message

A surge arrester limits voltage at its own terminals, not automatically at every item of equipment in the station. So the arrester must be located close enough to the protected equipment — especially the transformer, whose non-self-restoring insulation warrants a margin of typically 15% to 20%. For detailed studies, EMTP® or ATP is preferred; simplified equations are for early design and sanity checks, not a replacement for a proper transient study.

Reader should remember

A surge arrester protects best at its own terminals. The voltage at the transformer, breaker or open bus end can be higher, because lightning surges travel as waves and reflect inside the station. So the arrester location, lead length, separation distance, transformer capacitance and station layout must all be considered before selecting the BIL.

This is Part One of a multi-part self-study series on station lightning insulation coordination, covering the overall procedure, station modelling, arrester location and the first part of the crest-voltage calculation. Later parts continue with the remaining station voltages — arrester-bus junction and breaker voltage — insulation-strength evaluation and final BIL / clearance selection.

Section 36

Key Symbols

Table 3 — Key symbols used on this page.

Symbol	Meaning
BIL / BSL	Basic Lightning Impulse Level / Basic Switching Impulse Level
MTBF	Mean Time Between Failures (reliability criterion)
\(E,\ S\)	Incoming-surge crest voltage; incoming-surge steepness
\(E_d,\ E_A\)	Arrester discharge voltage to ground; surge voltage at arrester
\(V_{pv}\)	Power-frequency voltage component (opposite polarity)
\(E_T,\ E_t\)	Surge voltage at transformer; voltage to ground at transformer
\(E_v,\ E_b\)	Surge voltage at breaker; voltage to ground at breaker
\(E_I,\ E_j\)	Surge voltage at arrester-bus junction; voltage to ground there
\(T_T,\ T_A\)	Arrester-to-transformer travel time; arrester lead travel time
\(C_T,\ Z\)	Transformer surge capacitance; surge impedance
\(n,\ K_2\)	Number of connected lines; arrester voltage–current correction factor
\(\delta\)	Relative air density (altitude correction)

Part One Reading now

Procedure, Modelling & First Crest

The overall procedure, single-phase EMTP® station modelling, arrester location, and the first part of the transformer crest-voltage calculation.

Series progress 1 of 8

Part Two

Voltage Behind the Arrester

The breaker voltage E_B = E + E_A/2, transformer-capacitance effects, the early/late operation conditions and the arrester voltage–current characteristic.