Sources and References

Primary references — M. S. Naidu and V. Kamaraju, High Voltage Engineering, 5th ed., McGraw-Hill, 2013; E. Kuffel, W. S. Zaengl, and J. Kuffel, High Voltage Engineering: Fundamentals, 2nd ed., Newnes/Butterworth-Heinemann, 2000. Supplementary texts — J. L. Blackburn and T. J. Domin, Protective Relaying: Principles and Applications, 4th ed., CRC Press, 2014; A. R. Bergen and V. Vittal, Power Systems Analysis, 2nd ed., Prentice Hall, 2000. Online resources — IEC 60060-1 (High-voltage test techniques standard overview); IEEE Std C62.11 (metal-oxide surge arrester standard); CIGRE technical brochures on insulation coordination.

---

Chapter 1: Introduction to High Voltage Engineering

1.1 Importance of High Voltage in Power Systems

The transmission of electrical energy over long distances is fundamentally an exercise in minimizing resistive losses. For a given power \(P\) delivered at voltage \(V\) and current \(I\), the line losses scale as \(I^2 R\). Since \(P = VI\), raising the transmission voltage by a factor of \(k\) reduces the current by the same factor, and therefore reduces resistive losses by \(k^2\). This simple economic argument drives the continued push toward ever-higher transmission voltages.

Modern power systems operate across a hierarchy of voltage levels:

• Distribution: 4 kV to 35 kV (supplies residential and light commercial loads)
• Sub-transmission: 35 kV to 115 kV
• High voltage (HV) transmission: 115 kV to 230 kV
• Extra high voltage (EHV): 345 kV to 765 kV
• Ultra high voltage (UHV): above 1000 kV AC or 800 kV DC

Both alternating-current (HVAC) and direct-current (HVDC) transmission technologies are employed. HVAC dominates because it allows voltage transformation through conventional transformers, simplifies fault interruption (current crosses zero naturally), and supports reactive power management. HVDC is economical for bulk power over distances exceeding roughly 600–800 km (underground or submarine cable) or where asynchronous grid interconnection is required; it produces no capacitive charging current, enables precise power-flow control, and avoids the skin effect losses that limit AC cable current density.

1.2 Components of a High Voltage Power System

A high-voltage power system encompasses:

1. Generation: Synchronous generators at 11–25 kV, stepped up by generator-step-up (GSU) transformers.
2. Transmission: Overhead lines on lattice steel or concrete pole structures; underground or submarine cables for special applications.
3. Substations: Transformers, busbars, disconnect switches, circuit breakers, current transformers (CTs), voltage transformers (VTs), surge arresters, and grounding systems.
4. Insulation systems: Air gaps, ceramic/glass insulators on overhead lines, oil-paper insulation in transformers and cables, SF\(_6\) gas in gas-insulated switchgear (GIS), and solid polymeric insulation in modern cables.
5. Protection systems: Relays, circuit breakers, reclosers, fuses, and communication channels.

1.3 Overview of Insulation Systems

Every insulation system must withstand both the normal continuous operating voltage and the transient overvoltages arising from lightning strikes, switching operations, and power-frequency faults. The dielectric performance is characterized by:

• Basic Lightning Impulse Insulation Level (BIL): the crest value of a standard 1.2/50 µs lightning impulse voltage that the insulation withstands under specified test conditions.
• Basic Switching Impulse Insulation Level (BSL): the crest value of a standard 250/2500 µs switching impulse.

Insulation may be classified as self-restoring (air gaps, which recover after flashover) or non-self-restoring (solid or liquid dielectrics in transformers and cables, which suffer permanent damage).

---

Chapter 2: Generation of High Voltages for Testing

Part A: Generation and Measurement of High Voltages

2.1 Origin of High Voltage Stresses and Testing Requirements

Equipment insulation is stressed by three broad categories of voltage:

1. Power frequency (AC) — continuous operating voltage plus temporary overvoltages (TOV) at power frequency, lasting from cycles to minutes.
2. Switching impulses — generated by circuit breaker operations, line energization, and fault clearing; characterized by a slow front (tens to hundreds of microseconds) and long tail (milliseconds).
3. Lightning impulses — associated with direct or induced lightning strikes; extremely fast (microsecond) fronts with high peak amplitudes.

Testing laboratories reproduce these stresses to verify that equipment meets design and standard specifications before deployment. Tests fall into two broad categories:

2.1.1 HV Test Facilities

A typical HV laboratory contains:

• AC test transformers (single or cascade configuration)
• DC high-voltage sets with rectifier circuits or electrostatic generators
• Impulse voltage generators (Marx generators)
• Calibrated dividers and measuring systems
• Shielded rooms or outdoor test areas with sufficient clearances

2.2 Generation of High AC Voltages

2.2.1 Single Test Transformers

A single high-voltage test transformer is constructed similarly to a power transformer but with special attention to:

• Uniform insulation grading along the winding to minimize internal stress concentration
• Oil–SF\(_6\) or oil–paper insulation in the bushing and terminal region
• Low leakage inductance to maintain voltage waveform quality under capacitive test loads

The rated secondary voltage may range from tens of kilovolts to several megavolts for research facilities. The key limitation is the physical size of a single unit; above roughly 500 kV it becomes impractical to transport or erect a single transformer.

2.2.2 Cascade Transformers

To achieve test voltages above the practical single-unit limit, transformers are connected in cascade. In a three-stage cascade with each stage rated \(V_0\):

• Stage 1: primary energized from supply at \(V_{supply}\), secondary produces \(V_0\); the tank is at earth potential.
• Stage 2: mounted on top of stage 1 (insulated platform), its primary energized by an excitation winding on stage 1; secondary adds another \(V_0\).
• Stage 3: similarly excited by stage 2, producing a further \(V_0\).
• Total output: \(3V_0\).

The cascade principle avoids the insulation difficulties of a single transformer rated \(3V_0\) because each stage’s tank need only be insulated from the adjacent stage by \(V_0\).

2.2.3 Series Resonant Circuits

A series resonant test system consists of a variable-inductance reactor \(L\) in series with the capacitive test object \(C_{test}\). At resonance:

\[ f_0 = \frac{1}{2\pi\sqrt{LC_{test}}} \]

The quality factor \(Q = \frac{1}{R}\sqrt{\frac{L}{C_{test}}}\) determines the voltage magnification. The voltage across \(C_{test}\) is \(Q\) times the supply voltage. Resonant systems are advantageous for testing cables and capacitor banks (large \(C_{test}\)) because the reactive power is supplied by the resonant circuit rather than by the supply, drastically reducing the required supply kVA.

2.3 Generation of High DC Voltages

2.3.1 Half-Wave and Full-Wave Rectifier Circuits

A simple half-wave rectifier with a smoothing capacitor \(C_s\) produces an average DC voltage approximately equal to the peak AC supply voltage but with significant ripple. The ripple voltage for a half-wave rectifier is:

\[ \Delta V \approx \frac{I_{load}}{f \cdot C_s} \]

where \(f\) is the supply frequency and \(I_{load}\) is the DC load current. Full-wave rectification halves the ripple for the same \(C_s\) and \(f\).

2.3.2 Voltage Doubler Circuit

In a voltage doubler (Greinacher circuit), two diodes and two capacitors are arranged so that during the negative half-cycle, capacitor \(C_1\) charges to peak supply voltage \(V_p\); during the positive half-cycle, the supply voltage adds to the voltage on \(C_1\), charging \(C_2\) to \(2V_p\). The output is therefore twice the peak supply voltage. This principle generalizes to the Cockcroft–Walton multiplier.

2.3.3 Cockcroft–Walton Voltage Multiplier

The Cockcroft–Walton (CW) multiplier consists of \(n\) stages, each comprising a diode–capacitor ladder. For an ideal multiplier with \(n\) stages, peak AC supply voltage \(V_p\), and negligible load current, the output voltage is:

\[ V_{out} = 2n V_p \]

Under load current \(I_{load}\), the output voltage drops and the ripple increases. For an \(n\)-stage CW multiplier with all capacitors equal to \(C\):

\[ V_{out} = 2nV_p - \frac{I_{load}}{fC}\left(\frac{2n^3}{3} + \frac{n^2}{2} - \frac{n}{6}\right) \approx 2nV_p - \frac{2n^3 I_{load}}{3fC} \]

The ripple voltage is:

\[ \Delta V = \frac{n(n+1)I_{load}}{2fC} \]

These expressions show that for fixed \(C\) and \(f\), there is an optimal number of stages \(n_{opt}\) that maximizes output voltage under load:

\[ n_{opt} = \sqrt{\frac{V_p f C}{I_{load}}} \]

2.3.4 Electrostatic Generators (Van de Graaff)

The Van de Graaff generator transports charge mechanically on a moving belt from a lower corona-discharge assembly (charging spray points) to an upper collector inside a spherical high-voltage terminal. The terminal voltage is limited by corona and flashover from the terminal to the structure. Tandem and Pelletron variants extend the range to tens of megavolts, used primarily in nuclear physics research rather than power engineering HV testing.

2.4 Generation of Impulse Voltages

2.4.1 Overvoltages and Transients in Power Systems

Lightning strikes on or near overhead lines cause traveling waves with amplitudes that may reach several megavolts at the point of strike but are attenuated, reflected, and refracted as they propagate through the network. Switching operations generate slower but still damaging transients. The standard characterizes these by:

• Lightning impulse: defined by IEC 60060-1 as a 1.2/50 µs waveform — the time to peak (front time) is 1.2 µs and the time to half-value on the tail is 50 µs (tolerances ±30% on front, ±20% on tail).
• Switching impulse: 250/2500 µs waveform.
• Chopped impulse: the lightning impulse is deliberately chopped (by a rod gap or triggered gap) at a specified time, producing a sharp voltage collapse. Chopped-wave tests assess the ability of transformer windings to withstand rapidly oscillating voltages during wave reflections.

2.4.2 Basic Impulse Insulation Levels

BIL values are selected from standard insulation levels (IEC 60071) and coordinated with the protective devices on the system.

2.4.3 Marx Generator Principle

The Marx generator, named after Erwin Marx (1923), is the standard method of producing high-impulse voltages. The principle is elegant: charge \(n\) capacitors in parallel to a modest voltage \(V_0\), then switch them in series to deliver \(nV_0\) across the test object.

Circuit construction: \(n\) stages, each comprising a capacitor \(C_s\) (stage capacitor), a series charging resistor \(R_s\) to the next stage, and a spark gap. During charging, all \(C_s\) charge to \(V_0\) in parallel. When the first (trigger) gap is fired, the potential at the top of stage 1 jumps to \(2V_0\), overvoltaging stage 2 gap, which fires, raising the next junction to \(3V_0\), and so on in rapid cascade. The erected voltage across the output is \(nV_0\).

Equivalent circuit during discharge: The erected Marx generator appears as a single charged capacitor:

\[ C_{eq} = \frac{C_s}{n} \]

in series with a total inductance \(L_{total}\) and total resistance \(R_{total}\), driving the load (wave-shaping circuit and test object). The wave-shaping network — comprising a front resistor \(R_f\) and a tail resistor \(R_t\) in conjunction with the load capacitance \(C_L\) — shapes the discharge into the desired 1.2/50 µs waveform.

The peak output voltage is:

\[ V_{peak} = \eta \cdot n V_0 \]

where the efficiency factor \(\eta < 1\) accounts for the voltage division between \(C_{eq}\) and the load capacitance \(C_L\):

\[ \eta = \frac{C_s / n}{C_s / n + C_L} = \frac{C_s}{C_s + n C_L} \]

The energy stored in the erected generator is:

\[ W = \frac{1}{2} C_{eq} (nV_0)^2 = \frac{1}{2} \cdot \frac{C_s}{n} \cdot n^2 V_0^2 = \frac{1}{2} n C_s V_0^2 \]

which equals the total energy stored in all \(n\) capacitors charged to \(V_0\) in parallel — confirming energy conservation.

2.5 Measurement of High Voltages

2.5.1 Sphere Gaps

A calibrated sphere gap measures peak voltages by exploiting the reproducible breakdown voltage between two identical metallic spheres separated by a known distance in clean air at standard conditions. Correction factors account for air density (temperature and pressure):

\[ k_d = \frac{\rho}{\rho_0} = \frac{p}{p_0} \cdot \frac{T_0}{T} = \frac{p \times 293}{101.3 \times T} \]

where \(p\) is absolute pressure in kPa, \(T\) is temperature in kelvin, and \(p_0 = 101.3\) kPa, \(T_0 = 293\) K are standard conditions. The corrected breakdown voltage is \(V_{breakdown} = k_d \times V_{table}\).

2.5.2 Resistive Voltage Dividers

A resistive divider consists of a high-voltage arm \(R_1\) and a low-voltage arm \(R_2\). The division ratio is:

\[ k = \frac{R_1 + R_2}{R_2} \approx \frac{R_1}{R_2} \]

For impulse voltage measurement, parasitic capacitances of the high-voltage resistor distort the impulse waveshape. The response time of the divider must be much smaller than the impulse front time.

2.5.3 Capacitive and Mixed Dividers

Capacitive dividers are preferred for impulse and AC voltage measurement. The ratio:

\[ k = \frac{C_2}{C_1 + C_2} \approx \frac{C_2}{C_1} \quad (C_1 \ll C_2) \]

Mixed (damped capacitive) dividers combine the low response time of capacitive dividers with resistive damping to suppress oscillations. They provide excellent fidelity for both AC and impulse waveforms.

---

Chapter 3: Gaseous Insulation and Breakdown Mechanisms

Part B: Conduction and Breakdown in Dielectrics

3.1 Source of Charge Carriers in Gases

Under normal conditions, dry air at atmospheric pressure is an excellent insulator. However, ionizing agents continuously create electron–ion pairs:

• Cosmic rays and natural radioactivity produce roughly 5–10 ion pairs per cm³·s at sea level.
• UV radiation and X-rays eject photoelectrons from gas molecules.
• Thermal ionization is negligible at room temperature but important at arc temperatures (>3000 K).
• Field emission (Fowler–Nordheim) from cathode surface irregularities under intense electric fields.

These background charge carriers are the seed electrons that initiate avalanche processes.

3.2 Collision Mechanisms and Ionization

3.2.1 Townsend First Ionization Coefficient

When an electron gains sufficient energy from the applied electric field between collisions, it can ionize a neutral gas molecule, creating a new electron–ion pair. The Townsend first ionization coefficient \(\alpha\) is defined as the number of ionizing collisions made by one electron per unit path length in the direction of the field:

\[ \alpha = A p \exp\left(-\frac{Bp}{E}\right) \]

where \(E\) is the electric field, \(p\) is the gas pressure, and \(A\) and \(B\) are gas-specific constants (for air: \(A \approx 15\) cm\(^{-1}\)torr\(^{-1}\), \(B \approx 365\) V cm\(^{-1}\) torr\(^{-1}\)).

The number of electrons produced by an initial electron \(n_0\) after traveling a distance \(d\) is:

\[ n = n_0 e^{\alpha d} \]

This exponential growth is the electron avalanche.

3.2.2 Secondary Ionization Processes (Townsend Second Coefficient)

The secondary ionization coefficient \(\gamma\) (sometimes called \(\gamma_s\) or simply the second Townsend coefficient) quantifies secondary emission processes, primarily:

• Positive ions striking the cathode and releasing electrons
• Photoemission from the cathode due to photons generated in the avalanche

The number of secondary electrons released at the cathode per primary avalanche electron (on average) is \(\gamma\).

3.3 Townsend Breakdown Theory

In a uniform-field gap of spacing \(d\), with an initial photo-current \(I_0\) at the cathode, the sustained current is:

\[ I = \frac{I_0 e^{\alpha d}}{1 - \gamma(e^{\alpha d} - 1)} \]

The denominator approaches zero when \(\gamma(e^{\alpha d} - 1) = 1\), i.e., when:

At the breakdown condition, the current becomes theoretically infinite (limited only by the circuit impedance), and a self-sustaining glow or arc forms.

3.4 Paschen’s Law and the Paschen Curve

Paschen’s law states that the breakdown voltage \(V_b\) of a uniform-field gap is a function only of the product of pressure and gap spacing \((pd)\):

\[ V_b = f(pd) = \frac{Bpd}{\ln(Apd) - \ln\left[\ln\left(1 + \frac{1}{\gamma}\right)\right]} \]

This follows from the Townsend criterion. The Paschen curve (plot of \(V_b\) vs. \(pd\)) exhibits a characteristic minimum at \((pd)_{min}\):

\[ (pd)_{min} = \frac{e}{A} \ln\left(1 + \frac{1}{\gamma}\right) \]\[ V_{b,min} = \frac{eB}{A} \ln\left(1 + \frac{1}{\gamma}\right) \]

For air, this minimum occurs at approximately \(pd \approx 0.567\) torr·cm with \(V_{b,min} \approx 327\) V.

3.5 Streamer Breakdown Theory

The Townsend theory applies well to gaps where \(pd < 200\) torr·cm (roughly, gaps shorter than a few cm at atmospheric pressure). For longer gaps at atmospheric pressure, the observed breakdown mechanism differs fundamentally: breakdown occurs in a time much shorter than the ion transit time, and the formative time lag does not correlate with \(\gamma\) processes. The streamer theory, developed independently by Raether, Meek, and Loeb, explains this behavior.

3.5.1 Streamer Mechanism

As an electron avalanche propagates from cathode to anode, it builds up a space charge at its head. The avalanche head consists of electrons (fast, moving toward anode) while the positive ion tail is left behind. When the avalanche charge is sufficiently large, the space-charge field at the avalanche head becomes comparable to the applied field, and a highly conducting streamer filament can propagate.

Meek’s condition for streamer inception is approximately:

\[ \alpha d_{cr} \approx 18 \text{ to } 20 \]

This means the avalanche must multiply to contain approximately \(e^{18} \approx 6.6 \times 10^7\) to \(e^{20} \approx 5 \times 10^8\) electrons before the streamer mechanism dominates.

More precisely, Meek’s condition requires that the radial space-charge field of the avalanche equals the applied field:

\[ \frac{e \cdot N}{4\pi\varepsilon_0 r^2} = E_0 \]

where \(N = e^{\alpha d_{cr}}\) is the avalanche charge and \(r\) is the avalanche radius.

3.5.2 Statistical and Formative Time Lags

The statistical time lag \(t_s\) is the time from voltage application until a suitable initiating electron appears in the gap. At moderate overvoltages, initiating electrons are scarce and \(t_s\) can be substantial; at high overvoltages or under UV irradiation, \(t_s\) is negligible.

The formative time lag \(t_f\) is the time from the appearance of an initiating electron to the completion of the discharge channel. In the Townsend regime, \(t_f\) is large because it requires multiple cathode-crossing cycles. In the streamer regime, \(t_f\) is very short (nanoseconds to tens of nanoseconds).

The total time lag: \(t = t_s + t_f\). Impulse time–voltage (t–V) characteristics plot the breakdown voltage against the total time lag, defining the volt–time characteristic used in coordination of surge arresters and insulation.

3.6 Corona Discharge

3.6.1 Corona on Coaxial Cylindrical Geometry

The electric field between an inner conductor of radius \(r\) and outer conductor of radius \(R\) in a coaxial arrangement is:

\[ E(r) = \frac{V}{r \ln(R/r)} \]

where \(V\) is the applied voltage. The field is maximum at the inner conductor surface (\(\rho = r\)) and decreases as \(1/\rho\). This non-uniform field is why corona occurs first at the smaller conductor.

Corona onset field: For a cylindrical conductor of radius \(r_0\) in air at standard conditions, Peek’s empirical formula gives the critical surface field for corona onset:

\[ E_{onset} = E_0 \left(1 + \frac{m_r \delta^{1/2}}{r_0^{1/2}}\right) \]

where \(E_0 \approx 3\) MV/m, \(m_r\) is a surface roughness factor, and \(\delta\) is the relative air density.

Corona onset voltage for a coaxial geometry:

\[ V_{onset} = E_{onset} \cdot r_0 \cdot \ln\frac{R}{r_0} \]

3.6.2 Positive and Negative Corona

The polarity of the high-field electrode (inner conductor) determines the corona morphology:

• Positive corona (conductor at high positive potential): Electrons drift rapidly toward the conductor, leaving positive ions in the outer region. The discharge takes the form of streamers (Hermstein glows at low fields, streamers at higher fields). Positive corona is more uniform and produces less radio-frequency (RF) interference than negative corona.
• Negative corona (conductor at high negative potential): Electrons are emitted from the high-field electrode into the gas. Several modes exist: Trichel pulses (regular, repetitive pulses at lower fields), negative glow, and negative streamers at higher fields. Negative corona generates stronger RF interference (Trichel pulse current spikes) but is more stable.

3.6.3 Applications: Overhead Lines and GIS

Overhead lines: Corona on transmission line conductors causes energy loss (corona loss), audible noise (AN), and radio/TV interference (RI). Bundled conductors (2, 3, or 4 sub-conductors per phase) are used on EHV lines to increase the equivalent conductor radius and reduce the surface field below the corona onset threshold.

Gas-insulated switchgear (GIS): SF\(_6\) gas is used at pressures of 3–5 bar in coaxial electrode configurations. The geometry allows very high voltage operation in compact equipment. Corona activity in SF\(_6\) is strongly suppressed by electron attachment; residual partial discharges from metallic particles or surface defects are detected by ultra-high frequency (UHF) sensors for condition monitoring.

---

Chapter 4: Breakdown and Failure in Solid and Liquid Dielectrics

4.1 Breakdown in Solid Dielectrics

4.1.1 Dielectric Strength

The intrinsic dielectric strength of a solid material is the maximum electric field (V/m) the material can sustain without irreversible breakdown. For common insulating materials:

Material	Dielectric Strength (kV/mm)
Polyethylene (XLPE)	18–25
Epoxy resin	15–20
Polypropylene film	20–30
Porcelain	6–12
Glass	10–15
Oil-impregnated paper	10–18

Practical design values are 2–5 times lower due to field enhancement at voids, interfaces, and electrode irregularities.

4.1.2 Intrinsic Breakdown

Intrinsic breakdown occurs when the applied field is large enough to directly accelerate conduction-band electrons (either thermally generated or field-injected) to energies sufficient to cause impact ionization. This is a purely electronic process independent of time, temperature, or electrode geometry (for ideal, defect-free samples). The intrinsic strength is the upper bound for a given material.

4.1.3 Electromechanical Breakdown

Electromechanical breakdown affects soft or rubbery solid dielectrics. The electrostatic compression pressure on a slab of thickness \(d\), permittivity \(\varepsilon\), and Young’s modulus \(Y\):

\[ P_{electrostatic} = \frac{\varepsilon E^2}{2} = Y \cdot \frac{\delta d}{d} \]

As voltage increases, the slab thins, increasing the field for the same voltage, creating positive feedback. The critical (breakdown) field is:

\[ E_{EM} = \left(\frac{Y}{\varepsilon}\right)^{1/2} \cdot (1/e)^{1/2} \approx 0.6 \sqrt{\frac{Y}{\varepsilon}} \]

4.1.4 Thermal Breakdown

Dielectric losses (from conduction current and polarization hysteresis) heat the material. At sufficiently high temperatures, the conductivity increases exponentially, creating a positive thermal feedback. Thermal equilibrium requires that the rate of heat generation equals the rate of heat dissipation:

\[ \sigma(T) E^2 = \nabla \cdot (\kappa \nabla T) \]

where \(\sigma(T) = \sigma_0 e^{b(T-T_0)}\) is the temperature-dependent conductivity (with coefficient \(b \approx 0.1\) K\(^{-1}\) for many polymers). When the applied field exceeds the thermal breakdown field \(E_{th}\), no stable solution exists and runaway heating leads to carbonization and permanent failure.

4.1.5 Partial Discharge Detection

Partial discharges (PD) are localized dielectric discharges within a portion of the insulation volume that do not bridge the entire gap. Over time, PD erodes solid dielectrics through:

• Bombardment of the void walls by ions and electrons
• Byproduct chemicals (ozone, nitric acid) attacking polymer chains
• Thermal damage

PD is measured by:

• IEC 60270 electrical method: a coupling capacitor and measuring impedance detect the charge pulse associated with each PD event. Charge is expressed in picocoulombs (pC).
• UHF method: broad-band electromagnetic sensors detect the radiated RF pulse from PD events, particularly in GIS.
• Acoustic emission: piezoelectric sensors detect ultrasonic waves generated by PD in oil-filled equipment.

4.1.6 Discharge in Cavities

A gas-filled spherical cavity embedded in a solid dielectric experiences an enhanced electric field relative to the background field \(E_0\):

\[ E_{cavity} = \frac{3\varepsilon_r}{2\varepsilon_r + 1} E_0 \]

For a high-permittivity solid (\(\varepsilon_r \gg 1\)), \(E_{cavity} \approx 1.5 E_0\). The cavity has a much lower breakdown voltage than the surrounding solid. Once the cavity gas breaks down, the PD current flows, charges the cavity walls, and partially reduces the cavity field. This leads to a repetitive PD mechanism synchronized with the AC cycle.

4.1.7 Treeing and Tracking

Electrical treeing: Under high AC or DC fields, PD events in a microdefect gradually erode the surrounding polymer, extending the damaged region in a branching, tree-like pattern. Once the tree bridges the insulation, catastrophic breakdown occurs. Trees grow slowly (hours to years) under service conditions.

Surface tracking: Along contaminated or wet surfaces, a conductive layer (from pollution, condensation, or salt deposits) creates leakage current paths. Dry band arcing creates carbonized tracks that eventually bridge across the insulator, causing flashover. Pollution flashover is a major concern for outdoor insulators in coastal, industrial, and desert environments.

4.2 Breakdown in Liquid Dielectrics

4.2.1 Role and Properties of Insulating Liquids

Mineral oil remains the dominant insulating liquid in power transformers and oil-filled cables. Key properties:

• Dielectric constant: \(\varepsilon_r \approx 2.2\) (low, minimizing capacitive loading)
• Breakdown strength: 10–30 kV/mm (for carefully purified, degassed oil under standard gap conditions)
• Tan delta (loss factor): < 0.001 for clean oil at power frequency
• Water content: even trace water (5–10 ppm by mass) dramatically reduces breakdown strength
• Dissolved gas: gas bubbles nucleate breakdown; dissolved gas analysis (DGA) monitors fault indicators (hydrogen, acetylene, ethylene)

Alternative insulating liquids include synthetic esters (more biodegradable), silicone oil, and vegetable oil (FR3).

4.2.2 Charge Injection and Conduction

In a clean, highly purified liquid, the dominant charge carriers are generated by:

• Field ionization of the liquid molecules at very high fields
• Dissociation of trace impurities and ionic contaminants
• Electrode injection: Schottky emission or field emission from electrode surfaces

The conduction is generally ohmic at low fields and space-charge limited at higher fields.

4.2.3 Breakdown Theories in Liquids

Cavitation/Bubble theory: A gas bubble (from dissolved gas or local vaporization) has much lower dielectric strength than the surrounding liquid. The bubble elongates in the field direction, and the field in the bubble tip is enhanced. An electron avalanche develops in the bubble, extending it further. When the conducting channel bridges the gap, breakdown occurs.

Suspended particle theory: Conducting or semiconducting particles in the oil experience dielectrophoretic forces and align in chains along the field. A chain bridging the gap provides a conductive path and triggers breakdown. Particle contamination dramatically reduces the oil’s dielectric strength.

---

Chapter 5: Overvoltages and Insulation Coordination

Part C: Insulation Coordination and Power System Protection

5.1 Causes of Overvoltages in Power Systems

Overvoltages are classified by their origin, duration, and waveshape:

Category	Cause	Typical Duration	Typical Amplitude
Temporary (TOV)	Single-phase faults, load rejection, ferroresonance	Seconds–minutes	1.0–1.4 p.u.
Slow-front (switching)	Circuit breaker operations, line energization	ms	2–4 p.u.
Fast-front (lightning)	Direct strikes, back-flashovers, induced surges	µs	3–6 p.u. or more
Very fast transients	GIS disconnector operations	ns	1.5–2.5 p.u.

5.1.1 Lightning Overvoltages

A direct lightning strike to a phase conductor injects a current pulse (typically 10–50 kA crest) that creates a voltage wave traveling in both directions from the strike point. The voltage at the strike point is:

\[ V = Z_{surge} \cdot \frac{I_{stroke}}{2} \]

where \(Z_{surge} \approx 250\text{–}400\) Ω is the characteristic impedance of the overhead line. For a 30 kA stroke, \(V \approx 30\text{–}60\) MV — far exceeding any practical line insulation. The ground wire (overhead shielding wire) intercepts most direct strikes, but back-flashover from the tower top back to the phase conductor can still occur when the tower ground impedance is high.

5.1.2 Switching Overvoltages

Line energization (closing a breaker onto an uncharged line) creates a traveling wave transient. The worst case is re-energization of a trapped-charge line, where the receiving-end voltage can reach 3 p.u. Pre-insertion resistors (hundreds of ohms, inserted briefly by controlled closing) attenuate switching surges on EHV systems.

5.1.3 Temporary Overvoltages (TOV)

Ground faults on ungrounded or resonantly grounded systems raise the unfaulted phase voltages by \(\sqrt{3}\) (factor of 1.73). Load rejection on long lines causes the receiving-end voltage to rise due to the Ferranti effect (line charging current through line inductance). TOV determines the maximum continuous voltage rating of surge arresters.

5.2 Electric Stress and Strength Concepts

Insulation coordination balances the electric stress (applied voltage) against the dielectric strength (withstand capability), taking into account the statistical distributions of both:

• Stress: The probability distribution of overvoltage amplitudes on the system (derived from electromagnetic transient simulations or statistical analysis of historical data).
• Strength: The probability distribution of withstand voltages for the insulation (established by repeated impulse testing; characterized by the V50 = 50% flashover voltage and the standard deviation).

The risk of failure is the integral of the product of the stress and strength distributions over all voltage levels. Insulation coordination chooses the withstand level such that the risk of failure meets the utility’s reliability target.

5.3 External and Internal Insulation

External insulation (air gaps, overhead line insulators) is self-restoring: after flashover the insulating medium recovers. It is influenced by pollution, humidity, and altitude.

Internal insulation (solid and liquid dielectrics inside equipment) is non-self-restoring. Flashover or breakdown causes permanent damage. Internal insulation design uses larger safety margins than external insulation.

Clearances: The minimum air gap distances between live parts and earthed structures for various voltage classes are specified in IEC 60071-2 and national standards. They ensure that the probability of flashover from the design overvoltage is acceptably low.

---

Chapter 6: Overvoltage Protection Devices

6.1 Surge Impedance and Surge Propagation

Traveling voltage waves on transmission lines follow the wave equation. For a lossless line with inductance \(L'\) per unit length and capacitance \(C'\) per unit length, the characteristic (surge) impedance is:

\[ Z_0 = \sqrt{\frac{L'}{C'}} \]

For overhead lines, \(Z_0 \approx 250\text{–}400\) Ω; for underground cables, \(Z_0 \approx 20\text{–}60\) Ω (lower due to the much higher capacitance per unit length).

At a discontinuity where two lines of impedances \(Z_1\) and \(Z_2\) are joined, the refraction (transmission) coefficient is:

\[ \tau = \frac{2Z_2}{Z_1 + Z_2} \]

and the reflection coefficient is:

\[ \rho = \frac{Z_2 - Z_1}{Z_2 + Z_1} \]

At an open circuit (\(Z_2 \to \infty\)): \(\tau = 2\), meaning the voltage doubles. At a short circuit (\(Z_2 = 0\)): \(\tau = 0\), the voltage is absorbed. These reflection phenomena explain voltage magnification at open-circuit terminations and at transformer terminals.

6.2 Surge Protective Devices

6.2.1 Rod and Horn Gaps

Simple rod gaps (set at a fixed clearance) protect line insulators and transformers by providing a deliberate flashover path. The gap flashes over at a voltage below the equipment BIL, but the power-frequency follow current must be interrupted by the circuit breaker. Horn gaps exploit the rising arc column: as the arc heats and rises, it lengthens, increasing the arc voltage until natural extinction occurs.

6.2.2 Silicon Carbide (SiC) Gapped Arresters

Early surge arresters used a series combination of spark gaps and SiC nonlinear resistors. The gaps reduce the power-frequency voltage across the SiC, while the SiC limits the surge current and therefore limits the protective voltage. The gaps introduce a voltage across the arrester from gap sparkover to current interruption, creating a coordination challenge. The series gap must extinguish the power-frequency current (arc quenching requires careful design).

6.2.3 Metal Oxide (ZnO) Gapless Arresters

Modern surge arresters use zinc oxide (ZnO) varistor disks, which exhibit an extremely nonlinear voltage–current relationship:

\[ I = k V^{\alpha} \]

where the nonlinearity exponent \(\alpha\) is 25–50 for ZnO (compared to \(\alpha \approx 5\) for SiC). This means that over a modest voltage range, the current changes by many orders of magnitude. The consequence:

• At normal operating voltage, the leakage current through the arrester is only a few milliamperes — negligible.
• During a surge, the arrester clamps the voltage to its protection level while passing kiloamperes of surge current.
• No series gap is needed; the arrester is directly connected between line and earth.

The continuous operating voltage (COV) is the maximum voltage the arrester can sustain continuously. The protection level (residual voltage \(V_{res}\)) is the voltage across the arrester at the specified discharge current (e.g., 10 kA).

6.2.4 Energy Absorption in Surge Arresters

During a surge event, the arrester must absorb the energy in the surge without overheating. The energy absorbed is:

\[ W = \int_{0}^{\infty} V(t) \cdot I(t) \, dt \]

where \(V(t)\) is the voltage across the arrester (approximately equal to the protection level \(V_p\) during the surge) and \(I(t)\) is the discharge current. For a simplified rectangular pulse of amplitude \(I_{peak}\) and duration \(\tau\):

\[ W \approx V_p \cdot I_{peak} \cdot \tau \]

The total energy must remain within the arrester’s rated energy absorption capability (joules or joules per joule/kV of MCOV). Modern ZnO arresters for station class applications are rated at tens to hundreds of kilojoules.

6.2.5 Time–Voltage Coordination

Protection coordination requires that the arrester protection level \(V_p\) be lower than the equipment BIL by a sufficient margin — the protective ratio or coordination factor \(K_c\):

\[ K_c = \frac{BIL}{V_p} \geq 1.20 \text{ (recommended minimum for station equipment)} \]

The volt–time characteristics of both the arrester (relatively flat — ZnO responds nearly instantaneously) and the insulation gap (decreasing withstand at shorter times) must be coordinated to ensure the arrester clamps the voltage before the insulation breaks down.

---

Chapter 7: Introduction to Overcurrent Protection

7.1 Basic Definitions in Power System Protection

Key terms:

• Primary protection: the main protection system for a given zone, designed to operate first.
• Backup protection: operates if the primary fails; may be local (on the same relay) or remote (on an adjacent zone’s relay).
• Selectivity (discrimination): the ability to isolate only the faulted element without disturbing healthy parts of the network.
• Speed: faster operation limits equipment damage and improves transient stability.
• Sensitivity: the ability to detect minimum fault currents, including those through high fault impedances.
• Reliability: the certainty that the relay will operate correctly for faults within its zone (dependability) and will not operate incorrectly for external faults (security).

7.2 Faults in Power Systems

Power system faults are abnormal conditions involving breakdown of insulation between phases or between phase and earth. Classifications:

Fault Type	Approximate Frequency
Single line-to-ground (SLG)	70–80%
Line-to-line (LL)	10–15%
Double line-to-ground (DLG)	10%
Three-phase (3Φ)	3–5%

Three-phase faults are the most severe but least frequent; single line-to-ground faults are the most common. Fault impedance, fault location, and system configuration (grounding method) all affect fault current magnitude.

The symmetrical three-phase fault current is bounded by the Thevenin equivalent at the fault point:

\[ I_f^{3\phi} = \frac{V_{prefault}}{Z_{Thevenin}} \]

For ground faults, sequence network theory (using positive, negative, and zero sequence networks) is required.

7.3 Types of Protective Relays

7.3.1 Electromechanical Relays

The induction-disk relay (idisk) operates on the same principle as the induction motor: two time-displaced fluxes (from the relay CT secondary current and its time-shifted replica) create a torque on an aluminum disk proportional to the product of the fluxes. The disk accelerates and closes the trip contact after a time inversely related to the current. These relays are robust and have served well for decades but have limited adaptability.

7.3.2 Static (Solid-State) Relays

Static relays use analog electronic circuits (op-amps, comparators, timers) to replicate the overcurrent characteristic. They offer faster operation, lower burden on CTs and VTs, and more precise settings than electromechanical types.

7.3.3 Digital and Numerical Relays

Modern numerical relays digitize the CT and VT secondary waveforms, apply digital filtering (Fourier-based extraction of fundamental component), and execute protection algorithms in software. They offer:

• Multiple protection functions in one device
• Extensive event recording and disturbance recording
• Self-monitoring and diagnostics
• Communication via IEC 61850 (GOOSE messaging, sampled values)

---

Chapter 8: Instrumentation Transformers

8.1 Current Transformers (CTs)

8.1.1 Principles and Equivalent Circuit

A current transformer (CT) reproduces a scaled replica of the primary current in its secondary circuit for measurement and protection. The ideal CT has a turns ratio \(N_1 : N_2\) such that:

\[ I_2 = \frac{N_1}{N_2} I_1 \]

The equivalent circuit of a real CT is analogous to a standard transformer with the following key elements (referred to the secondary):

• \(R_1, X_1\): primary resistance and leakage reactance (usually negligible)
• \(R_m, X_m\): core loss and magnetizing reactance (the excitation branch, connected in shunt)
• \(R_2, X_2\): secondary resistance and leakage reactance
• \(Z_b\): burden (load impedance — relay coils, meters, leads)

Because the primary current \(I_1\) is dictated by the power system (not by the CT secondary burden), the CT operates as a current source rather than a voltage source. The secondary current:

\[ \vec{I}_2 = \frac{N_1}{N_2}\vec{I}_1 - \vec{I}_e \]

where \(\vec{I}_e\) is the excitation current (the magnetizing current that the core requires). This excitation current is the source of errors.

8.1.2 Ratio Error and Phase Angle Error

The ratio error (also called current error) is:

\[ \varepsilon = \frac{(N_1/N_2) I_2 - I_1}{I_1} \times 100\% \]

For a well-designed metering CT, this is less than 0.1–0.5% at rated current.

The phase angle error \(\delta\) is the phase angle between the reversed secondary current and the primary current. For a purely resistive burden and a core with conductance \(G_m\) and susceptance \(B_m\):

\[ \delta \approx \frac{I_m \cos\delta - I_c \sin\delta}{I_1} \approx \frac{I_m}{I_1} \]

where \(I_m\) is the magnetizing component and \(I_c\) is the core loss component of \(I_e\).

8.1.3 CT Saturation

Under fault conditions, the primary current may contain a DC offset component (asymmetrical fault current). The DC component drives the CT core into saturation because the core flux must integrate the applied voltage:

\[ \Phi(t) = \frac{1}{N_1} \int V_1 \, dt \]

During DC offset decay (time constant \(\tau = L/R\) of the primary circuit, typically 20–100 ms), the core flux can build up to several times the rated flux, driving the core into saturation. During saturation periods, the CT secondary current collapses, causing protection relays to see distorted or reduced current. The CT must be sized with an adequate accuracy limit factor (ALF) and, for protection applications, sufficient remnant flux margin.

The knee point voltage \(V_k\) of a protection CT (IEC 61869 Class P or PX) is the secondary voltage at which a 10% increase in voltage causes a 50% increase in excitation current — the onset of saturation.

8.2 Voltage Transformers (VTs) and CVTs

8.2.1 Electromagnetic VTs

Voltage transformers (VTs, also called potential transformers, PTs) are essentially precision power transformers with the secondary burden limited to meter and relay coil impedances (typically a few to tens of VA). The turns ratio gives:

\[ V_2 = \frac{N_2}{N_1} V_1 \]

Standard secondary voltages: 110 V or 115 V (line-to-line), 63.5 V or 57.7 V (phase-to-earth for a 110 V system).

Accuracy classes (IEC 61869-3): 0.1, 0.2, 0.5, 1, 3 (metering) and 3P, 6P (protection).

8.2.2 Capacitive Voltage Transformers (CVTs)

Above 100–145 kV, electromagnetic VTs become expensive and physically large. Capacitive voltage transformers (CVTs) use a capacitive voltage divider (\(C_1\) and \(C_2\)) to step down the voltage to an intermediate level (typically 5–15 kV), followed by a small electromagnetic voltage transformer to reach the 110 V secondary:

The intermediate voltage:

\[ V_{int} = \frac{C_1}{C_1 + C_2} V_1 \]

A series reactor \(L_{comp}\) is tuned to resonate with \(C_2\) at power frequency, which compensates for the capacitive loading and improves accuracy. CVTs are standard at 115 kV and above.

---

Chapter 9: System Protection Components

9.1 Circuit Breakers

A circuit breaker (CB) is the primary fault-interrupting device. It must:

1. Carry rated current continuously.
2. Close onto a fault without damage.
3. Interrupt the fault current at a current zero.

Arc interruption: When CB contacts separate, an arc forms. The arc is extinguished by cooling the arc column (reducing ion concentration) and increasing the dielectric strength of the gap faster than the transient recovery voltage (TRV) rises after current zero. Interrupting media:

• Air-blast: forced compressed air extinguishes the arc; used historically at EHV.
• SF\(_6\) (puffer and self-blast): SF\(_6\) gas is blown through the arc. Dominant technology for HV CBs above 36 kV.
• Vacuum: arc products condense on the contacts rapidly; used for medium voltage (up to 36 kV).
• Oil (bulk oil, minimum oil): arc energy vaporizes oil, forming hydrogen (excellent arc-quenching gas); older technology.

CB ratings: Breaking current (kA symmetrical), making current (kA peak), TRV envelope, operating duty cycle (e.g., O–0.3s–CO–3min–CO).

9.2 Reclosers

A recloser is an automatic circuit breaker with a built-in control that enables it to open on a fault and reclose automatically after a preset time. The logic typically allows 2–4 reclosing attempts before locking out:

• First trip + first reclose (instantaneous — typically < 0.5 s dead time)
• Second trip + second reclose (delayed — e.g., 5 s dead time)
• Third trip + lockout

This strategy is highly effective for distribution systems because 70–80% of distribution faults are transient (lightning-caused flashover, momentary tree contact) and clear when the line is de-energized briefly. A recloser minimizes the number of customers who experience sustained outages.

9.3 Fuses

Fuses provide simple, low-cost overcurrent protection for distribution feeders, laterals, and equipment. Types:

• Expulsion fuses: the arc in the fiber fuse tube generates high-pressure gas that blasts out the arc products, extinguishing the current at a natural zero. Used on distribution feeder laterals and as transformer fuses.
• Current-limiting fuses: a fine silver element in a quartz sand medium melts rapidly, creating many series arcs and a very high arc voltage. This voltage limits the peak fault current (current limitation) and clears the fault within the first half-cycle. Used to protect capacitor banks, cables, and transformers.

Fuse coordination with reclosers: The recloser’s fast curve must operate before the fuse blows for transient faults (fuse saving), while the fuse must clear permanent faults on downstream laterals before the recloser locks out.

9.4 Disconnect Switches (Disconnectors)

Disconnectors (isolators) open and close circuits under no-load or very small load conditions. They are not rated for fault interruption. Their functions:

• Isolation: provide visible open gaps for safety maintenance
• Bus transfer: switch between bus sections
• Line charging switching: some types are rated to interrupt small capacitive currents

In GIS, three-position disconnect/earth switches combine the disconnector and earthing switch functions.

---

Chapter 10: Overcurrent Protection Coordination

10.1 Zones of Protection and Coordination Principles

Power system protection is organized into overlapping zones of protection, each covering one element (generator, transformer, bus, line). Adjacent zones overlap at the circuit breakers:

• Within a zone, faults are cleared by the primary protection of that zone.
• Zone overlap ensures there are no unprotected points (all faults are within at least one zone).
• The overlap also means some faults are within two zones; only the zone whose primary fault falls more precisely in its coverage clears first.

Primary vs. backup protection: Primary protection clears faults quickly (e.g., 3–5 cycles for high-speed pilot protection). If the primary relay or the associated circuit breaker fails, backup protection — on adjacent relays or on a breaker-failure relay — clears the fault after a longer delay.

10.2 Overcurrent Relay Characteristics

10.2.1 Definite-Time and Inverse-Time Relays

A definite-time overcurrent relay trips after a fixed time delay \(T\) if the current exceeds the pickup setting \(I_{pickup}\), regardless of the actual current magnitude (above pickup). It is simple to coordinate but allows longer fault duration at high fault currents.

An inverse-time relay operates faster at higher currents, with an operating time that decreases as current increases. This characteristic naturally aligns with the equipment damage curves (high currents cause faster thermal damage, so faster clearing is appropriate).

10.2.2 IEC Standard Time–Current Characteristics

The IEC 60255-151 standard defines several inverse-time characteristic families. For all, the operating time is:

\[ t = \frac{TDS \cdot K}{\left(\dfrac{I}{I_{pickup}}\right)^{\alpha} - 1} \]

where \(TDS\) is the Time Dial Setting (also called TMS, time multiplier setting), \(I\) is the measured current, \(I_{pickup}\) is the relay pickup current, and \(K\) and \(\alpha\) are characteristic constants:

Characteristic	\(K\)	\(\alpha\)
Standard inverse (SI)	0.14	0.02
Very inverse (VI)	13.5	1.0
Extremely inverse (EI)	80.0	2.0
Long-time inverse	120.0	1.0

Very inverse (VI) characteristic: operating time falls as \(1/I\), which aligns with fuse-melting characteristics; used when the relay needs to coordinate with downstream fuses.

Extremely inverse (EI) characteristic: operating time falls as \(1/I^2\), closely matching the thermal damage characteristics of cables, and is used when the source impedance is relatively low compared to the fault impedance.

10.2.3 Time–Current Characteristic (TCC) Curves

TCC curves plot operating time (y-axis, log scale) vs. current (x-axis, log scale) and are the primary tool for coordination:

• Device curves (relay, fuse, breaker) are plotted on the same TCC graph.
• Coordination time interval (CTI): The minimum time separation between the operating time of the backup relay and the operating time of the primary device, including circuit breaker interrupting time and relay overshoot. Typical CTI: 0.2–0.4 s for electromechanical relays; 0.1–0.2 s for numerical relays.
• Coordination criterion: For every fault current in the coordination range, the primary device must operate before the backup device by at least the CTI.

10.3 Time-Dial Setting Selection

To coordinate two inverse-time overcurrent relays in series (relay R1 upstream, relay R2 downstream), the procedure is:

1. Select R2 settings (pickup and TDS) based on the minimum fault current at the downstream bus and the maximum load current.
2. For the maximum fault current at the location of R2, read R2’s operating time from its TCC.
3. Add the CTI to get the required operating time of R1 at that fault current.
4. Select R1’s TDS such that its TCC passes through (or above) the required operating time at that fault current.
5. Verify that R1’s pickup is set above the maximum load current and below the minimum fault current seen by R1.

10.4 Pickup Setting and Sensitivity

The pickup current \(I_{pickup}\) of an overcurrent relay must satisfy two constraints simultaneously:

\[ I_{max,load} < I_{pickup} < \frac{I_{fault,min}}{CTI_{factor}} \]

where \(I_{max,load}\) is the maximum load current through the CT, and \(I_{fault,min}\) is the minimum fault current for faults within the protection zone. A minimum sensitivity requirement (the relay must operate for the minimum fault) dictates that:

\[ \frac{I_{fault,min}}{I_{pickup}} \geq 2 \text{ (minimum recommended)} \]

This ratio is often called the relay sensitivity factor or earth-fault factor (for ground relays).

10.5 Directional Overcurrent Relays

In a looped or meshed network, fault current can flow in either direction through a given line. A non-directional overcurrent relay cannot distinguish between forward (into the line) and reverse (back-feeding from the other end) faults. A directional overcurrent relay uses a polarizing quantity (voltage, or zero-sequence current for ground protection) to determine the current direction and only trips for forward faults.

For ground fault protection, the zero-sequence current \(I_0 = (I_a + I_b + I_c)/3\) or the residual current \(I_r = I_a + I_b + I_c = 3I_0\) is the operating quantity, and the zero-sequence voltage \(V_0\) or the zero-sequence current from a directional source provides polarization.

---

Chapter 11: Pollution Effects and Insulation in Service

11.1 Pollution Flashover of Outdoor Insulators

Outdoor insulators (on overhead lines and substations) accumulate pollution from industrial emissions, salt spray, bird droppings, dust, and cement alkali. When wet (rain, fog, condensation), the pollution layer dissolves and forms a conductive surface film. The surface leakage current heats dry bands, which arc over, extending and contracting as the AC cycle progresses. Eventually, the arc bridges the full insulator, causing pollution flashover.

Specific creepage distance (SCD): The total leakage path length (creepage distance) divided by the rated voltage is the specific creepage distance, in mm/kV. IEC 60815 defines pollution levels (a through d, or I through IV), each requiring a minimum SCD:

Pollution Level	Typical Environment	Minimum SCD (mm/kV)
Very light	Desert, low humidity	16
Light	Agricultural, low traffic	20
Medium	Industrial, moderate coast	25
Heavy	Coastal, heavy industry	31
Very heavy	Severe coastal	35

Mitigation: RTV (room-temperature vulcanizing silicone) coatings on porcelain and glass insulators impart hydrophobicity — the surface repels water droplets rather than forming a continuous film. Silicone rubber composite insulators (polymer insulators) are inherently hydrophobic and are widely used in polluted areas.

11.2 Altitude and Air Density Correction

Air density decreases with altitude. The dielectric strength of air is approximately proportional to density. At altitude \(h\) (meters above sea level), the correction factor for air clearances and dry flashover voltages is:

\[ K_a = e^{-h/8150} \]

Equipment installed at high altitude must either have larger air clearances or be tested at the appropriate corrected voltage. For example, at 1000 m altitude, \(K_a \approx 0.88\), meaning clearances must be approximately 14% larger than at sea level.

---

Chapter 12: Integrated Protection System Design

12.1 Protection System Architecture

A complete protection system comprises:

1. Instrument transformers (CTs and VTs/CVTs): signal conditioning
2. Protective relays: fault detection and decision logic
3. Trip circuits: relay output contacts energize the circuit breaker trip coil
4. Circuit breakers: physical interruption of fault current
5. DC auxiliary supply: battery-backed 110/125 V DC system powers all protection and control equipment, independent of the AC supply

The protection system must operate correctly even when the AC system is dead (during faults), so reliable DC power is non-negotiable.

12.2 Coordination Between Protection and Insulation

The goal of the integrated system is to ensure that:

• Surge arresters clamp transient overvoltages below the equipment BIL.
• Overcurrent and distance relays clear sustained faults before insulation thermal or mechanical damage thresholds are exceeded.
• Reclosers and fuses cooperate to minimize the extent and duration of service interruptions.

The protection coordination study and the insulation coordination study are conducted together when designing a new substation or upgrading an existing one, since both determine the voltage and current stresses on equipment and the required withstand capabilities.

12.3 Breaker Failure Protection

If a circuit breaker fails to interrupt a fault (due to mechanism failure, loss of SF\(_6\) pressure, or trip coil failure), the fault persists and must be cleared by tripping all breakers surrounding the failed breaker’s bus zone. A breaker failure relay starts a timer when a trip command is issued. If the current through the breaker has not fallen below a threshold (indicating successful interruption) within the timer period (typically 100–200 ms for HV systems), the breaker failure relay sends trip commands to the adjacent breakers.

---

Summary: Key Equations and Relationships

The following equations represent the core quantitative results of ECE 464:

Townsend breakdown criterion:

\[ \gamma\left(e^{\alpha d} - 1\right) = 1 \]

Townsend first ionization coefficient:

\[ \alpha = Ap\exp\!\left(-\frac{Bp}{E}\right) \]

Paschen’s law (breakdown voltage vs. \(pd\)):

\[ V_b = \frac{Bpd}{\ln(Apd) - \ln\!\left[\ln\!\left(1 + \frac{1}{\gamma}\right)\right]} \]

Streamer inception (Meek’s condition):

\[ \alpha d_{cr} \approx 18\text{–}20 \]

Coaxial field distribution:

\[ E(r) = \frac{V}{r\ln(R/r)} \]

Cockcroft–Walton output voltage:

\[ V_{out} = 2nV_p - \frac{2n^3 I_{load}}{3fC} \]

Marx generator peak output:

\[ V_{peak} = \eta \cdot nV_0, \quad \eta = \frac{C_s}{C_s + nC_L} \]

CT ratio error:

\[ \varepsilon = \frac{(N_1/N_2)I_2 - I_1}{I_1} \times 100\% \]

IEC inverse-time overcurrent relay:

\[ t = \frac{TDS \cdot K}{\left(I/I_{pickup}\right)^{\alpha} - 1} \]

Surge arrester energy absorption:

\[ W = \int V(t)\cdot I(t)\,dt \]

Refraction coefficient at line junction:

\[ \tau = \frac{2Z_2}{Z_1 + Z_2} \]