Insulation Coordination

Station Lightning Coordination: Voltage Behind the Arrester

Part Two of the series. The voltage at equipment behind the arrester — the breaker, a disconnector, a bus support — is a travelling-wave timing problem, not a simple clamp. This self-study derives the single-line breaker voltage E_B = E + E_A/2, shows how transformer capacitance delays arrester operation through the early- and late-operation conditions and the critical distance, examines the modest effect of the number of connected lines, and treats the arrester voltage–current characteristic, the Z = R_A reflected-crest condition and the estimation of arrester current.

Reading time ≈ 40 min · Part Two of the series

Section 1

Voltage Behind the Arrester

Part One established that a surge arrester limits voltage at its own terminals, while equipment behind the arrester can see a higher voltage through travelling-wave effects. This part calculates that voltage behind the arrester — the "breaker" voltage — and shows how transformer capacitance, arrester operating time and the arrester voltage–current characteristic change it.

Recap of Part One

Part One explained that the arrester clamps the voltage at its own terminals, but the transformer or breaker may still experience a different voltage because lightning surges travel and reflect inside the station. Part Two focuses on the voltage behind the arrester — especially at the breaker or other equipment connected beyond the arrester location.

The location behind the arrester is described generically as the breaker location, but it may equally represent a circuit breaker, a disconnecting switch, a bus support, an open bus section, or any point on the station bus away from the arrester. The central question is:

What “behind the arrester” means

In this page, “behind the arrester” means the part of the station on the protected side of the arrester — a breaker, bus section, disconnecting switch or other equipment electrically downstream of the arrester location. It refers to the electrical position, not the physical direction.

The central question

What voltage appears at equipment located behind the arrester when a lightning surge enters the station?

What Part Two covers

the breaker voltage \(E_B = E + E_A/2\) for a single-line station;
how transformer capacitance delays arrester operation and changes the result;
the early- and late-operation conditions, and the critical distance;
the modest effect of the number of connected lines;
the arrester voltage–current characteristic and the \(Z = R_A\) condition;
estimating the arrester discharge current and why it is needed.

Section 2

Why the Voltage Behind the Arrester Matters

The arrester clamps voltage at the arrester location, but the voltage at another item of equipment is not necessarily equal to it, because surge waves travel through the station and reflect from discontinuities. The voltage behind the arrester depends on the incoming-surge crest \(E\) and steepness \(S\), the arrester discharge voltage \(E_d\) and surge voltage \(E_A\), the arrester-to-equipment travel time \(T_B\), the transformer capacitance \(C_T\), the surge impedance \(Z\), the number of connected lines \(n\), the arrester lead travel time \(T_A\), and the arrester voltage–current characteristic.

Why the breaker can see more than the arrester

The arrester limits the voltage at its own terminals. The breaker is separated from the arrester by a bus section, so the surge and its reflected wave need time to travel between the two points. During that travel time the incoming surge keeps rising, so when the reflected wave returns to the breaker the breaker voltage can be higher than the arrester voltage.

Arrester protection is local

The further the equipment is from the arrester, the more important travelling-wave effects become. Protection is strongest at the arrester terminals and weakens with distance.

Section 3

The Single-Line Station

The first case is a single-line station. The incoming surge has a linearly rising front of steepness \(S\) (unlimited magnitude in the idealised case), and the arrester is first assumed to be a constant-voltage device, \(E_A = \text{constant}\). Real arresters are not perfect constant-voltage devices, but the assumption builds the basic travelling-wave concept. The equipment behind the arrester is a generic breaker point, at travel time \(T_B\) from the arrester:

\[ T_B = \frac{d_B}{300} \qquad (d_B\text{ in m},\; T_B\text{ in }\mu\text{s, for AIS}) \]

\(T_B\): travel time between breaker and arrester
\(d_B\): physical separation distance

Section 4

Breaker Voltage Without Transformer Capacitance

For the constant-voltage arrester with no transformer capacitance, the maximum surge voltage behind the arrester depends on the incoming surge, the arrester voltage and the number of connected lines. For a single line (\(n=1\)):

\[ E_B = E + \frac{E_A}{2} \qquad (n = 1) \]

\(E_B\): crest surge voltage at the breaker location
\(E\): incoming-surge crest voltage
\(E_A\): surge voltage at the arrester

For more than one line (\(n > 1\)) the maximum tends toward \(E_B = E\). So additional connected lines can reduce the voltage behind the arrester — though the effect is not large in all cases.

Section 5

What \(E_B = E + E_A/2\) Means

This result can seem surprising: the voltage behind the arrester may be greater than the incoming-surge crest \(E\), depending on reflections and arrester operation. It does not mean the arrester is failing — it means the breaker point is not located exactly at the arrester terminal, so the incoming and reflected waves can combine there:

\[ E_B > E_A, \qquad\text{and sometimes}\qquad E_B > E \]

In plain terms

For a single-line station the breaker voltage can exceed the arrester voltage because the breaker sees the combination of the original incoming surge and the voltage reflected from the arrester–transformer point. This is why the breaker voltage must be calculated separately and cannot be assumed equal to the arrester protective level.

This is precisely why equipment behind the arrester must be checked — the arrester rating alone does not guarantee its protection.

Section 6

Bringing in the Transformer Capacitance

How the capacitance changes the timing

The transformer capacitance affects how quickly the arrester voltage rises. A larger capacitance absorbs part of the surge current in the first instant, delaying the moment the arrester reaches its operating voltage. This changes the reflected wave and can reduce the breaker voltage in some cases.

The simple result above ignores transformer capacitance. But in a real station the transformer may be electrically close to the arrester or on the same bus, and its surge capacitance \(C_T\) affects the voltage waveform at the arrester — and therefore the reflected wave travelling back to the breaker. When \(C_T\) is included, the breaker voltage may decrease compared with the no-capacitance case. The reduction depends on the time constant \(ZC_T\), the time to crest of the incoming surge, the breaker-arrester distance and the arrester operating time.

To explain the mechanism, a simplified circuit places the transformer capacitance at the arrester location, neglects the arrester lead, and combines arrester and capacitance at one point (an explanatory model, not a full station). A worked example uses \(E = 1600\ \text{kV}\), \(S = 1000\ \text{kV}/\mu\text{s}\), \(Z = 400\ \Omega\), \(E_A = 300\ \text{kV}\), \(T_B = 0.5\ \mu\text{s}\). The capacitor changes how quickly the arrester voltage rises — if it is significant, the arrester may operate later because the capacitance initially absorbs part of the surge current.

Capacitance does not have one universal effect

At the transformer terminal, capacitance can increase the crest voltage through charging and reflection effects. Behind the arrester, it can reduce the breaker voltage in some cases, because it delays arrester operation and changes the reflected wave. The effect must always be interpreted by location.

Transformer voltage versus breaker voltage

The two should not be read the same way. The transformer-side voltage is strongly influenced by the open-circuit effect and transformer capacitance (Part One). The breaker-side voltage is governed more by reflection timing, the arrester operating time and the round-trip travel time between breaker and arrester.

Section 7

Arrester Operating Time \(t_A\)

The arrester operating time \(t_A\) is the time when the voltage at the arrester reaches \(E_A\). Before operation, the arrester voltage is set by the incoming surge and the capacitor charging; after operation, the arrester holds the voltage at approximately \(E_A\) (under the constant-voltage assumption). The relation between \(t_A\), \(ZC_T\), \(S\) and \(E_A\) is not always explicit and may require iteration — which is why the original treatment provides a graph for the ratio \(t_A/ZC_T\).

Section 8

Reflected-Voltage Behaviour

When the incoming surge reaches the arrester–capacitance combination, a reflected voltage is produced. Initially the capacitance behaves almost like a short circuit (its voltage cannot change instantly), giving an initial negative reflection; as it charges, the reflected voltage may cross zero and later become positive. This behaviour is governed by the \(ZC_T\) time constant. The reflection arrives back at the breaker after a round-trip time \(2T_B\), so the breaker voltage is the sum of the original incoming surge and the reflected voltage from the arrester–transformer point.

Section 9

Two Breaker-Voltage Conditions

Timing quantities — read these first

\(t_f\) — the time to crest of the incoming surge.
\(t_A\) — the arrester operating time, when the arrester starts to clamp the voltage.
\(T_B\) — the travel time between the breaker and the arrester.
\(T_A\) — the arrester lead travel time (added later, in the lead-length section).
\(2T_B\) — and \(2(T_B+T_A)\) once the lead is included — the round-trip time for a reflected wave to travel out to the arrester path and back to the breaker.

The calculation splits into two conditions, controlled by comparing the arrester operating time \(t_A\) with \(t_f - 2T_B\), where \(t_f\) is the time to crest of the incoming surge. The quantity \(t_f - 2T_B\) is the time remaining before the surge reaches crest after the reflected wave has returned to the breaker.

Table 1 — The early- and late-operation conditions.

Condition	When	Breaker Crest Voltage
1 — arrester operates early	\(t_A < t_f - 2T_B\)	\(E_B = E + \tfrac{E_A}{2}\) (single line) — controlled by the arrester-reflection effect
2 — arrester operates late	\(t_A \ge t_f - 2T_B\)	crest occurs at \(t_f\); voltage is lower, but never below \(E_A\)

Early operation — in words

The arrester operates early enough that its reflected wave returns to the breaker before the incoming surge has reached its crest. The breaker voltage is then governed by the arrester reflection while the incoming wave is still rising — this is the \(E_B = E + \tfrac{E_A}{2}\) case for a single line.

Late operation — in words

The incoming surge reaches its crest before the arrester reflection can fully control the breaker voltage. The breaker voltage is then governed more by the incoming-surge crest and less by early arrester clamping — the crest occurs at \(t_f\), and the voltage is lower but never below \(E_A\).

So larger capacitance ⇒ later arrester operation ⇒ the breaker voltage may be reduced. But the breaker voltage cannot fall below the arrester voltage: if the calculation gives less than \(E_A\), the actual value is taken as \(E_B = E_A\).

Section 10

A Useful Limiting Case

When \(2T_B > t_f\), the reflected wave returns after the incoming surge has already reached crest, so:

\[ 2T_B > t_f \;\;\Rightarrow\;\; E_B = E \]

The breaker sees essentially the incoming-surge crest before the reflected wave can modify it. This shows why separation distance matters: a longer separation delays the reflected wave.

Section 11

Worked Example — Capacitance Reduces Breaker Voltage

Take \(E = 2300\ \text{kV}\), \(E_A = 900\ \text{kV}\), \(ZC_T = 2.4\ \mu\text{s}\), \(T_B = 0.2\ \mu\text{s}\), \(V_{pv} = 0\), \(S = 2000\ \text{kV}/\mu\text{s}\). The time to crest is:

\[ t_f = \frac{E}{S} = \frac{2300}{2000} = 1.15\ \mu\text{s} \]

Solving for the arrester operating time by iteration gives \(t_A = 1.12\ \mu\text{s}\). Then \(t_f - 2T_B = 1.15 - 0.4 = 0.75\ \mu\text{s}\). Since \(t_A > t_f - 2T_B\), the late-operation condition applies:

\[ E_B = 1223\ \text{kV} \quad\text{(late operation)} \qquad\text{vs}\qquad E_B = 1700\ \text{kV} \quad\text{(simple expression)} \]

So including transformer capacitance significantly reduces the calculated breaker voltage in this case.

Section 12

The Critical Distance

What the critical distance is

The critical distance is the separation at which the calculation changes from the early-operation case to the late-operation case. Below it, the reflected wave returns early enough to influence the breaker voltage before the incoming-surge crest. Beyond it, the breaker may see a higher voltage because the incoming surge reaches crest before the reflected wave can control it.

The two conditions meet when \(t_A = t_f - 2T_B\). Rearranging gives the critical travel time and distance:

\[ T_B = \frac{t_f - t_A}{2} \qquad\Rightarrow\qquad d_B = 300\,T_B \]

In the previous example the critical distance is \(4.5\ \text{m}\). At this distance both equations give the same breaker voltage; it separates the early-operation and late-operation regions.

Section 13

More Than One Connected Line

If more than one line is connected at the arrester location, the additional lines provide surge paths for waves to leave the station, which affects the arrester voltage and the reflected voltage returning to the breaker. The theory is the same, using the same two conditions (\(t_A < t_f - 2T_B\) and \(t_A \ge t_f - 2T_B\)), with the arrester operating time found from the corresponding \(n\)-line relation (again via a quick-estimation graph).

Why the reduction is only modest

Additional connected lines give the surge energy extra paths to divide between, which can reduce the voltage behind the arrester. The reduction is usually modest, though, because the breaker voltage is still mainly governed by the incoming surge, the arrester operation and the local travelling-wave timing.

A practical observation

The breaker voltage is somewhat insensitive to the number of connected lines. More than one line gives a small reduction in breaker voltage — not a dramatic one in the examples.

Section 14

Worked Example — One-Line and Two-Line Station

Data: \(E = 1600\ \text{kV}\), \(S = 1000\ \text{kV}/\mu\text{s}\), \(E_d = 300\ \text{kV}\), \(V_{pv} = 100\ \text{kV}\), \(C_T = 3\ \text{nF}\), \(T_B = 0.2\ \mu\text{s}\), \(Z = 400\ \Omega\). So \(ZC_T = 1.2\ \mu\text{s}\), \(E_A = E_d + V_{pv} = 400\ \text{kV}\), and \(t_f = E/S = 1.6\ \mu\text{s}\).

One-line station (\(n=1\))

The graph gives \(t_A/ZC_T = 0.64\), so \(t_A = 0.768\ \mu\text{s}\). Since \(t_f - 2T_B = 1.2\ \mu\text{s}\) and \(t_A < 1.2\), the early-operation equation applies:

\[ E_B = 800\ \text{kV} \;\;\Rightarrow\;\; E_b = E_B - V_{pv} = 800 - 100 = 700\ \text{kV} \]

Critical distance, one line

At the critical condition \(2T_B = t_f - t_A = 1.6 - 0.768 = 0.832\ \mu\text{s}\), so \(T_B = 0.416\ \mu\text{s}\) and \(d_B = 124.8\ \text{m}\), where both equations give \(E_b = 1133\ \text{kV}\). For \(T_B = 0.8\ \mu\text{s}\), \(t_f - 2T_B = 0\) and the breaker voltage to ground reaches \(E_b = 1600 - 100 = 1500\ \text{kV}\) — essentially the incoming-surge crest minus the power-frequency voltage.

Two-line station (\(n=2\))

The graph now gives \(t_A = 0.855\ \mu\text{s}\). For the original \(T_B = 0.2\ \mu\text{s}\) the early condition still applies and \(E_b = 700\ \text{kV}\). The critical \(T_B\) becomes \(0.3725\ \mu\text{s}\), at which \(E_b = 1045\ \text{kV}\); for larger distances \(E_b\) rises toward the incoming-surge value — e.g. \(1208\ \text{kV}\) at 180 m and \(1500\ \text{kV}\) at 240 m. The two-line voltage is below the one-line voltage, but the reduction is modest.

Section 15

Large Transformer Capacitance

With a large capacitance, \(C_T = 20\ \text{nF}\) (so \(ZC_T = 8.0\ \mu\text{s}\)) and \(n = 1\), the arrester operating time becomes \(t_A = 1.88\ \mu\text{s}\), and the critical \(T_B\) turns negative — meaning the early-operation equation does not apply. The late-operation equation must be used, and the breaker voltage stays limited to the arrester voltage until the separation distance exceeds about \(24\ \text{m}\).

A useful suppression effect

Large transformer capacitance can delay arrester operation and suppress the voltage behind the arrester for short separation distances — the breaker voltage is held at \(E_A\) until the separation grows large enough.

Section 16

Including Arrester Lead Length

With the arrester lead length included, the effective travel time becomes \(T_B + T_A\) (where \(T_A\) is the arrester lead travel time), and the two conditions become \(t_A \le t_f - 2(T_B + T_A)\) and \(t_A > t_f - 2(T_B + T_A)\). The implication: longer leads raise the effective separation, and so potentially the equipment voltage.

Keep arrester leads short

Arrester leads should be short, straight and directly connected — lead length adds directly to the effective separation distance.

Section 17

A Limitation: the Transition Is Not Perfectly Smooth

The two formulations give good estimates but do not produce a perfectly smooth transition between equations. Around the boundary \(t_A = t_f - 2(T_B + T_A)\), small discontinuities or abrupt changes can appear in the calculated result — a limitation of the simplified method. For accurate studies the station should be modelled in EMTP® or ATP.

Section 18

Effect of the Actual Arrester Characteristic

In simple terms

A real arrester does not clamp at one fixed voltage — its voltage depends on the current flowing through it. The slope of the arrester voltage–current curve affects the reflected voltage, and the crest of that reflection occurs when the line surge impedance equals the local slope resistance of the characteristic.

A real arrester does not clamp at a perfectly constant voltage — its voltage depends on its current (higher current ⇒ higher discharge voltage), described by the voltage–current characteristic. Consider a surge of impedance \(Z\), crest \(E\) and steepness \(S\) arriving at an arrester at the end of a line. Computing the reflected voltage from the total arrester voltage and the incoming surge, the crest of the reflected voltage occurs when:

\[ Z = R_A, \qquad R_A = \frac{de_A}{di_A} \]

\(R_A\): slope (dynamic) resistance of the arrester V–I characteristic
\(Z\): line surge impedance

The maximum reflected voltage is therefore not necessarily controlled by the maximum arrester current — it can be set by the point on the characteristic where the slope resistance equals the line surge impedance.

Section 19

What \(Z = R_A\) Means

The condition \(Z = R_A\) means the reflected voltage peaks when the line surge impedance equals the local dynamic resistance of the arrester characteristic — which occurs before the arrester current reaches its maximum. So the maximum voltage along the line may be set by a relatively low arrester-current point, not by the peak current. In practice, however, arrester characteristics are fairly flat, so using the maximum discharge voltage (at maximum current) is usually conservative and accurate enough.

Why the conservative choice is acceptable

In most practical cases, detailed arrester voltage–current modelling changes the result only slightly. Using the maximum arrester discharge voltage (the value at maximum arrester current) normally gives a conservative estimate of the protected-equipment voltage.

Section 20

Worked Example — Arrester Characteristic

With \(E = 2300\ \text{kV}\) and \(Z = 400\ \Omega\), the slope resistance drops below \(400\ \Omega\) at \(i_A = 0.05\ \text{kA}\), where \(e_A = 715.5\ \text{kV}\). The reflected-voltage crest is \(e_1 = 348\ \text{kV}\), so the maximum line voltage is:

\[ E_{\max} = 2300 + 348 = 2648\ \text{kV} \quad\text{(dynamic characteristic)} \]

Ignoring the dynamic characteristic, the maximum arrester current is \(i_A = 9.26\ \text{kA}\) with \(e_A = 895\ \text{kV}\), giving:

\[ E_{\max} = 2750\ \text{kV} \quad\text{(maximum current)} \;\approx\; +4\% \]

So ignoring the detailed characteristic gives a result about 4% higher — conservative in this example.

Section 21

Practical Conclusion on Arrester Characteristics

For this case, detailed dynamic arrester characteristics are not necessary — using the maximum arrester voltage associated with maximum current gives a conservative result. In the 138 kV case from Part One, including actual characteristics caused only a modest reduction in breaker voltage. So the simplified equations need no major modification for arrester characteristics: detailed V–I modelling improves accuracy, but the maximum discharge voltage is often conservative enough for simplified calculations.

Section 22

Estimating the Arrester Discharge Current

Assuming a transformer in the studied circuit, the arrester current can be estimated from a relation that does not consider reflections from the struck point but still gives a good estimate (and is recommended). Conceptually, the current is the driving surge voltage minus the arrester voltage, divided by the effective surge-impedance path:

\[ I_A \approx \frac{\text{driving surge voltage} - \text{arrester voltage}}{\text{effective surge impedance}} \]

The arrester current must be estimated because the discharge voltage of a metal-oxide arrester depends on current, \(E_d = f(I_A)\). A detailed study therefore uses an iterative loop:

The iterative arrester logic

estimate the arrester current \(I_A\);
read the arrester discharge voltage \(E_d\) from the V–I curve;
calculate the surge voltage at the equipment;
check against the BIL or withstand requirement, and iterate.

Why the current must be estimated

The arrester current is needed because the discharge voltage is current-dependent. The estimated current is used to read or interpolate the arrester protective voltage from the manufacturer’s voltage–current characteristic, and that voltage is then used in the insulation-coordination calculation.

This is why the arrester is not simply represented by a fixed protective level in a detailed study.

Section 23

Engineering Interpretation of the Breaker Voltage

The voltage behind the arrester is not controlled by any single factor — it comes from the interaction of the incoming-surge crest and front steepness, the arrester voltage, the separation distance, the transformer capacitance, the number of connected lines, the arrester lead length and the reflection timing. The critical comparison is:

\[ t_A \quad\text{versus}\quad t_f - 2(T_B + T_A) \]

If the arrester operates early, one equation applies; if it operates late, another. This makes the problem a travelling-wave timing problem, not just a voltage-clamping problem.

Section 24

Practical Design Lessons

The key lessons

Equipment behind the arrester must still be checked — breaker and bus voltages can be high if the separation distance is long.
Transformer capacitance can reduce breaker voltage (by delaying arrester operation) — yet may increase transformer terminal voltage; interpret capacitance effects by location.
The number of connected lines has a modest effect — consider it, but do not rely on it as the main protection.
Arrester lead length matters — the effective travel time is \(T_B + T_A\); keep connections short and direct.
Simplified methods have limits — the two-equation transition is not perfectly smooth; prefer EMTP® for important substations.
Maximum arrester discharge voltage is often conservative — detailed V–I characteristics refine but are not always essential for preliminary checks.

Section 25

Summary of Key Equations

Equation Summary

Time to crest

\( t_f = \dfrac{E}{S} \)

Arrester surge voltage

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

Breaker voltage to ground

\( E_b = E_B - V_{pv} \)

Travel time (AIS)

\( T_B = \dfrac{d_B}{300} \)

Breaker voltage (early, n=1)

\( E_B = E + \dfrac{E_A}{2} \)

Operation condition

\( t_A \;\lessgtr\; t_f - 2(T_B + T_A) \)

Critical travel time

\( T_B = \dfrac{t_f - t_A}{2} \)

Lower limit / long distance

\( E_B = E_A \;;\;\; E_B \to E \)

Reflected-crest condition

\( Z = R_A = \dfrac{de_A}{di_A} \)

Arrester current

\( I_A \approx \dfrac{\Delta V}{Z_{\text{eff}}} \)

Section 26

Reader Should Remember

The voltage behind the arrester is a travelling-wave timing problem. The breaker voltage can exceed both the arrester voltage and even the incoming-surge crest, depending on reflection timing. Transformer capacitance generally reduces the breaker voltage by delaying arrester operation, while the number of connected lines and the detailed arrester characteristic have only modest effects.

The single most important message

Equipment behind the arrester is not automatically protected. Whether the early- or late-operation equation applies depends on the comparison \(t_A\) versus \(t_f - 2(T_B + T_A)\) — so separation distance and arrester lead length must be controlled, and important substations modelled in EMTP®.

Carry these points forward

The voltage behind the arrester is controlled by travelling-wave timing. The key comparison is between the arrester operating time and the time left before the incoming surge reaches crest after the reflected wave returns. Transformer capacitance, arrester lead length and breaker separation distance can all change the breaker voltage — so the breaker voltage must be calculated and should never be assumed equal to the arrester protective level.

This is Part Two of the station lightning insulation coordination series, covering the voltage behind the arrester, breaker voltage, transformer-capacitance effects and arrester current. Later parts continue with the remaining station voltages, insulation-strength evaluation and final BIL / clearance selection.

Section 27

Key Symbols

Table 2 — Key symbols used on this page.

Symbol	Meaning
\(E,\ S\)	Incoming-surge crest voltage; incoming-surge steepness
\(t_f\)	Time to crest of the incoming surge, \(E/S\)
\(E_A,\ E_d\)	Surge voltage at the arrester; arrester discharge voltage to ground
\(E_B,\ E_b\)	Surge voltage / voltage to ground at the breaker (behind the arrester)
\(V_{pv}\)	Power-frequency voltage component (opposite polarity)
\(T_B,\ T_A\)	Breaker-to-arrester travel time; arrester lead travel time
\(C_T,\ Z\)	Transformer surge capacitance; surge impedance
\(ZC_T\)	Surge-impedance–capacitance time constant
\(t_A\)	Arrester operating time (when the arrester voltage reaches \(E_A\))
\(n\)	Number of connected lines
\(R_A,\ i_A\)	Slope (dynamic) resistance of the arrester characteristic; arrester current

Part Two Reading now

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.

Series progress 2 of 8

Part Three

Insulation Strength & BIL Selection

Turning surge stress into strength — safety factors, the 1.15 BIL rule, the transformer chopped-wave / full-wave / long-tail criteria, bushings and clearances.