Ferroresonance in Voltage Transformers Explained

Introduction

A voltage transformer that was operating normally yesterday is found burned beyond recognition this morning — with no fault record in the protection relay, no overcurrent trip, and no external damage to the surrounding equipment. The substation operators are baffled. The protection engineer suspects insulation failure. But the real cause is something far more insidious, and it was present in the circuit design long before the transformer failed: ferroresonance.

Ferroresonance in voltage transformers is a nonlinear resonance phenomenon that occurs when the transformer’s saturable magnetic core interacts with the capacitance of the connected network — producing sustained, chaotic overvoltages and overcurrents that can reach 3–5 times normal operating levels, causing catastrophic insulation failure, thermal destruction, and protection system maloperation without triggering conventional overcurrent protection.

I’ve investigated ferroresonance incidents across MV industrial networks in Europe, the Middle East, and Southeast Asia, and the pattern is remarkably consistent: a network configuration change — a cable connection, a switching operation, a single-phase fault — triggers a resonance condition that the original design never anticipated. The result is a destroyed voltage transformer, a confused protection system, and an engineering team searching for answers in the wrong place. This article gives you the complete picture: what ferroresonance is, why it occurs, how to recognize it, and — most importantly — how to eliminate it from your network design. 🔍

Table of Contents

• What Is Ferroresonance and How Does It Differ from Linear Resonance?
• What Causes Ferroresonance in Voltage Transformers and Which Network Configurations Are Most Vulnerable?
• How Do You Identify Ferroresonance Conditions and Select the Right VT Specification?
• What Are the Proven Mitigation Strategies for Ferroresonance in MV Networks?
• FAQs About Ferroresonance in Voltage Transformers

What Is Ferroresonance and How Does It Differ from Linear Resonance?

To understand ferroresonance, you first need to understand why it is fundamentally different from the classical resonance that electrical engineers encounter in circuit theory. Linear resonance is predictable, calculable, and occurs at a single well-defined frequency. Ferroresonance is none of these things — and that unpredictability is precisely what makes it so dangerous. ⚙️

Classical Linear Resonance vs. Ferroresonance

In a standard LC circuit, resonance occurs at a single frequency:

$f_{\text{resonance}} = \frac{1}{2\pi \sqrt{LC}}$

At this frequency, inductive and capacitive reactances are equal and opposite, and the circuit impedance drops to its resistive minimum. The behavior is entirely predictable — given L and C, you can calculate exactly when and at what amplitude resonance will occur.

Ferroresonance replaces the linear inductance L with a nonlinear, saturable inductance — the magnetizing inductance of a voltage transformer core. This single substitution transforms the entire mathematical character of the problem:

Property	Linear Resonance	Ferroresonance
Inductance	Constant (linear)	Variable (nonlinear, core-dependent)
Resonant frequency	Single, fixed value	Multiple possible values
Amplitude	Predictable, calculable	Chaotic, unpredictable
Triggering	Requires exact frequency match	Can be triggered by transients
Stable states	One stable operating point	Multiple coexisting stable states
Damping effect	Reduces amplitude proportionally	May not prevent sustained oscillation
Self-sustaining	No — requires continuous excitation	Yes — can be self-sustaining

The Nonlinear Core: Why VTs Are Uniquely Vulnerable

Voltage transformers are designed to operate with their cores at relatively high flux densities — close to the knee point of the B-H magnetization curve¹ — to achieve accurate voltage measurement across a wide range. This design choice, which is essential for measurement accuracy, simultaneously makes VT cores highly susceptible to ferroresonance because:

• The core’s magnetizing inductance varies dramatically with flux level
• Small increases in applied voltage can drive the core into saturation
• Once saturated, the effective inductance drops sharply, shifting the resonant condition
• The circuit can lock into a new stable operating state at a much higher voltage level

The Multiple Stable States Problem

The most dangerous characteristic of ferroresonance is the existence of multiple stable operating states for the same circuit configuration. The nonlinear V-I characteristic of a saturating VT core produces a folded response curve with three intersection points against the capacitive load line:

• State 1: Normal operating point — low voltage, low current, linear core operation
• State 2: Unstable transition point — never observed in practice
• State 3: Ferroresonant operating point — high voltage, high current, saturated core

A circuit can jump from State 1 to State 3 in response to a transient disturbance — a switching operation, a fault, a lightning surge — and then remain locked in State 3 indefinitely, even after the triggering event has passed. This is why ferroresonance is self-sustaining: the circuit has found a new stable equilibrium that does not require the original trigger to maintain it.

Ferroresonance Modes

Ferroresonance manifests in four distinct modes, each with characteristic waveform signatures:

Mode	Frequency Content	Waveform Character	Typical Trigger
Fundamental mode	Power frequency (50/60Hz)	Distorted sinusoid, sustained	Single-phase switching
Subharmonic mode	fn/n (e.g., 16.7Hz, 25Hz)	Periodic, low-frequency oscillation	Cable energization
Quasi-periodic mode	Multiple frequencies	Complex, irregular	Network reconfiguration
Chaotic mode	Broadband spectrum	Completely irregular, unpredictable	Multiple simultaneous triggers

What Causes Ferroresonance in Voltage Transformers and Which Network Configurations Are Most Vulnerable?

Ferroresonance does not occur randomly — it requires a specific combination of circuit conditions to be present simultaneously. Understanding these conditions is the foundation of both risk assessment and prevention. 🔬

The Three Essential Ingredients

Every ferroresonance incident requires all three of the following conditions to coexist:

1. A Saturable Nonlinear Inductance:
The voltage transformer’s magnetic core. Electromagnetic VTs (inductive VTs) are inherently susceptible. Capacitive Voltage Transformers (CVTs) have a fundamentally different circuit topology that provides natural immunity to most ferroresonance modes.

2. A Capacitance in Series or Parallel:
The capacitance can originate from multiple sources:

• Underground cable charging capacitance (most common in MV networks)
• Busbar and switchgear stray capacitance
• Grading capacitors in circuit breakers and disconnectors
• Power factor correction capacitor banks
• Shunt capacitance of overhead lines

3. A Low-Loss Circuit Path:
Ferroresonance is sustained by the energy exchange between the nonlinear inductance and the capacitance. Sufficient damping resistance in the circuit will prevent sustained oscillation — but many MV network configurations, particularly isolated neutral systems and lightly loaded cable networks, provide very little natural damping.

Network Configurations with Highest Ferroresonance Risk

Isolated Neutral (IT) Systems — Highest Risk:
In an isolated neutral MV network, the phase-to-earth capacitance of the cable network forms a direct resonant circuit with the VT magnetizing inductance. Single-phase switching operations — opening one phase of a disconnector while the other two remain closed — apply the full line voltage across the VT through the cable capacitance, creating ideal ferroresonance conditions.

Resonant Earthed (Petersen Coil) Systems — High Risk:
The Petersen Coil² is tuned to compensate the network capacitance, which means the residual capacitance after compensation is very small. This small residual capacitance can resonate with VT magnetizing inductance at or near power frequency — a particularly dangerous condition because the resonance is close to the fundamental mode.

Solidly Earthed Systems — Lower Risk (but not immune):
Solid earthing provides a low-impedance path that damps ferroresonance significantly. However, ferroresonance can still occur during switching operations that temporarily isolate a VT from the earth reference, or in cable-fed systems with high charging capacitance.

Triggering Events

Triggering Event	Ferroresonance Risk	Explanation
Single-phase disconnector operation	Very High	Temporarily applies voltage through capacitance only
Single-phase fuse operation	Very High	Creates unbalanced capacitive coupling
Cable energization with VT connected	High	Cable capacitance charges through VT magnetizing branch
Single-phase-to-earth fault clearing	High	Sudden voltage redistribution across healthy phases
Transformer energization	Medium	Inrush current drives VT core into saturation
Lightning or switching surge	Medium	Transient pushes circuit from normal to ferroresonant state

Why Underground Cable Networks Are Particularly Dangerous

The proliferation of underground cable networks in modern MV distribution systems has dramatically increased ferroresonance risk compared to traditional overhead line systems. The reason is straightforward: underground cables have 10–50 times higher capacitance per unit length than equivalent overhead lines.

A typical 11kV XLPE cable has a charging capacitance of 0.2–0.4 μF/km. A 5km cable feeder therefore presents 1–2 μF of capacitance to the network — more than sufficient to form a resonant circuit with the magnetizing inductance of a standard electromagnetic VT at power frequency.

Customer Story: A protection engineer named David, managing a 33kV industrial substation at a petrochemical complex in Rotterdam, Netherlands, experienced three VT failures in eighteen months — all on the same busbar section fed by a 4.2km underground cable. Each failure occurred during a switching operation, with no fault record and no overcurrent trip. Post-incident analysis identified ferroresonance as the cause: the cable capacitance (1.68 μF total) was resonating with the VT magnetizing inductance at 47Hz — close enough to fundamental frequency to sustain the oscillation indefinitely. The VT insulation was being destroyed by sustained 2.8 per-unit overvoltage. Bepto supplied replacement VTs with factory-fitted damping resistors in the open-delta secondary winding, which eliminated all subsequent ferroresonance incidents. ✅

How Do You Identify Ferroresonance Conditions and Select the Right VT Specification?

Ferroresonance risk assessment is a quantitative engineering process — not a qualitative judgment. The following framework gives you the tools to evaluate risk before equipment is specified and installed, rather than after the first VT failure. 📐

Step 1: Characterize the Network Capacitance

Calculate the total phase-to-earth capacitance at the VT installation point:

$C_{\text{total}} = C_{\text{cable}} + C_{\text{busbar}} + C_{\text{switchgear}} + C_{\text{other}}$

For cable networks:
$C_{\text{cable}} = c_{\text{specific}} \times L_{\text{cable}}$

Where c_specific is the cable’s capacitance per unit length (from cable datasheet, typically 0.15–0.45 μF/km for MV XLPE cables) and L_cable is the total connected cable length in km.

Step 2: Determine the Critical Capacitance Range

The ferroresonance risk zone is defined by the capacitance range within which the network capacitive reactance can resonate with the VT magnetizing reactance at or near power frequency:

$C_{\text{critical}} = \frac{1}{\omega^{2} \times L_{m}}$

Where Lm is the VT magnetizing inductance (obtainable from the no-load loss test data or magnetizing current specification). If C_total falls within $0.1 \times C_{\text{critical}} ;\text{to}; 10 \times C_{\text{critical}}$, ferroresonance risk is significant and mitigation measures are required.

Step 3: Assess Neutral Earthing Configuration

Neutral Earthing	Ferroresonance Risk	Recommended VT Type
Isolated (IT)	Very High	CVT or VT with damping resistor
Resonant earthed (Petersen coil)	High	VT with damping resistor, anti-ferroresonance design
High-impedance earthed	Medium–High	VT with damping resistor
Low-impedance earthed	Medium	Standard VT with open-delta secondary
Solidly earthed	Low	Standard VT — verify for cable-fed applications

Step 4: Select VT Type Based on Risk Assessment

Electromagnetic VT (Inductive VT) — Standard Design:

• Susceptible to ferroresonance in isolated and resonant earthed networks
• Requires additional mitigation measures (damping resistors, anti-ferroresonance devices)
• Lower cost, suitable for solidly earthed systems with low cable capacitance

Electromagnetic VT with Anti-Ferroresonance Design:

• Core designed to operate at lower flux density — further from saturation knee point
• Increased magnetizing inductance reduces resonance risk
• Suitable for medium-risk applications in isolated neutral systems

Capacitive Voltage Transformer (CVT):

• Fundamentally different circuit topology — capacitive divider with intermediate transformer
• Immune to most ferroresonance modes due to the series capacitor in the primary circuit
• Preferred for HV and EHV applications (≥66kV) and high-risk MV configurations
• Higher cost but eliminates ferroresonance risk entirely

Customer Story: Sarah, procurement manager at an EPC contractor in Singapore handling a 22kV industrial distribution system for a semiconductor manufacturing facility, initially specified standard electromagnetic VTs throughout the switchgear. The network comprised 8.5km of underground cable in an isolated neutral configuration — a textbook ferroresonance risk scenario. Bepto’s engineering team flagged the risk during the technical review and recommended anti-ferroresonance VTs with factory-fitted open-delta damping resistors. The additional cost was less than 8% of the total VT procurement budget. The facility has operated for three years without a single VT failure or ferroresonance event. 💡

Step 5: Verify Environmental and Installation Requirements

• Outdoor installations in humid or coastal environments: IP65 minimum, stainless steel terminal boxes, hydrophobic silicone insulator housing
• High-pollution environments (industrial, chemical): Creepage distance ≥ 25mm/kV, pollution class IV
• High-altitude installations (>1000m): Apply IEC altitude correction factors for dielectric strength
• Seismic zones: Verify mechanical withstand rating per IEC 60068-3-3³

What Are the Proven Mitigation Strategies for Ferroresonance in MV Networks?

Ferroresonance mitigation is not a single solution — it is a layered engineering strategy that addresses the phenomenon at the circuit level, the equipment level, and the operational level simultaneously. The most effective protection schemes combine multiple mitigation layers. 🛡️

Mitigation Strategy 1: Open-Delta Secondary Damping Resistor

The most widely applied and cost-effective mitigation for electromagnetic VTs in MV networks. The principle is straightforward: connect a resistor across the open corner of the open-delta (broken-delta) secondary winding to provide a continuous energy dissipation path that prevents sustained ferroresonance oscillation.

Resistor sizing:
The damping resistor must be sized to provide sufficient damping without overloading the VT secondary under earth fault conditions (when the open-delta voltage rises to 3× normal):

$R_{\text{damping}} = \frac{\left(3 \times V_{\text{secondary,rated}}\right)^{2}}{P_{\text{VT,thermal limit}}}$

Typical values range from 25Ω to 100Ω for standard MV VTs, with power ratings of 50W to 200W continuous.

Important constraints:

• The resistor must be permanently connected — switching it out during normal operation defeats its purpose
• The resistor value must be verified against the specific VT’s magnetizing characteristic — too high a resistance provides insufficient damping; too low overloads the VT winding

Mitigation Strategy 2: Anti-Ferroresonance VT Core Design

Modern anti-ferroresonance VTs use core designs that operate at significantly lower flux density than standard VTs — typically 60–70% of the flux density used in conventional designs. This moves the operating point further from the saturation knee point, increasing the voltage margin before ferroresonance can be triggered.

Key design features:

• Larger core cross-section — reduces flux density at rated voltage
• Higher quality grain-oriented silicon steel⁴ — sharper knee point, more predictable saturation behavior
• Optimized winding geometry — reduces leakage inductance⁵ that can contribute to resonance

Mitigation Strategy 3: Neutral Earthing Modification

Changing the network neutral earthing arrangement is the most fundamental mitigation — it addresses the root cause rather than the symptom:

• Converting from isolated to low-impedance earthed: Dramatically reduces ferroresonance risk by providing a low-impedance path that damps oscillations
• Neutral earthing resistor (NER): Adding a resistance between the neutral point and earth provides damping without the fault current implications of solid earthing
• Detuning the Petersen coil: In resonant earthed systems, adjusting the coil inductance away from exact resonance reduces the risk of fundamental-mode ferroresonance

Mitigation Strategy 4: Switching Sequence Optimization

Many ferroresonance incidents are triggered by specific switching sequences that can be avoided through operational procedures:

• Always switch three-phase simultaneously — avoid single-phase switching operations on circuits containing VTs in isolated neutral systems
• De-energize VTs before cable switching — disconnect VTs from the busbar before energizing or de-energizing long cable feeders
• Use circuit breakers instead of disconnectors — circuit breakers interrupt all three phases simultaneously, eliminating the unbalanced switching condition that triggers ferroresonance

Mitigation Strategy 5: Surge Arresters and Overvoltage Protection

While surge arresters do not prevent ferroresonance, they provide a critical last line of defense against the overvoltages it produces:

• Install metal oxide surge arresters (MOV) directly at the VT primary terminals
• Select arrester energy rating based on ferroresonance overvoltage duration — standard lightning arresters may be inadequate for sustained ferroresonance overvoltages
• Verify arrester continuous operating voltage (COV) is appropriate for the network earthing configuration

Mitigation Effectiveness Summary

Mitigation Strategy	Effectiveness	Cost	Implementation Complexity
Open-delta damping resistor	High	Low	Simple — retrofit possible
Anti-ferroresonance VT design	High	Medium	Requires VT replacement
Capacitive VT (CVT)	Very High	High	Requires VT replacement
Neutral earthing modification	Very High	Medium–High	Network-level change
Switching sequence procedures	Medium	Very Low	Operational — no hardware
Surge arresters at VT terminals	Low (protective only)	Low	Simple — retrofit possible

Installation and Commissioning Checklist

1. Verify open-delta wiring — confirm the secondary open-delta connection is correctly made before energization; an incorrectly wired open-delta provides no ferroresonance protection
2. Measure damping resistor value — verify installed resistance matches the specified value within ±5%
3. Check resistor thermal rating — confirm the resistor’s continuous power rating is adequate for earth fault conditions
4. Test surge arrester condition — perform leakage current test before energization
5. Document cable capacitance — record total connected cable length and calculated capacitance for future network change assessments
6. Establish switching procedures — document approved switching sequences that avoid single-phase operations on VT-connected circuits

Common Mistakes That Allow Ferroresonance to Persist

• Treating VT failures as insulation defects — repeatedly replacing failed VTs without investigating ferroresonance as the root cause is the most expensive mistake in MV network maintenance
• Removing damping resistors to reduce VT loading — some operators disconnect damping resistors to extend VT life under earth fault conditions, unknowingly eliminating the only ferroresonance protection in the circuit
• Extending cable networks without reassessing VT compatibility — adding cable feeders increases network capacitance; a VT that was safe with 2km of cable may be at risk with 6km
• Specifying standard VTs for isolated neutral cable networks — this combination is a known high-risk configuration that requires explicit ferroresonance mitigation from the design stage
• Ignoring subharmonic and chaotic ferroresonance modes — protection relays tuned to detect fundamental-frequency overvoltages will not detect subharmonic ferroresonance, which can destroy a VT at voltages that appear normal to standard monitoring equipment

Conclusion

Ferroresonance is a predictable, preventable phenomenon — but only if it is recognized and addressed at the design stage, before the first VT failure provides the evidence that the risk was real. The combination of saturable VT cores, network capacitance, and low-damping circuit configurations creates the conditions for self-sustaining overvoltages that conventional protection cannot detect or interrupt. Assess your network capacitance, specify the correct VT type for your neutral earthing configuration, install open-delta damping resistors as standard practice in isolated neutral systems, and establish switching procedures that eliminate single-phase operations on VT-connected circuits. Eliminate the conditions for ferroresonance, and your voltage transformers will deliver accurate measurements and reliable protection performance throughout their operational life. 🔒

FAQs About Ferroresonance in Voltage Transformers

1. Understanding the relationship between magnetic flux density and field intensity in transformer cores. ↩

2. A method for earthing the neutral point of a distribution network using a variable reactor. ↩

3. International standards for seismic testing methods for equipment and systems. ↩

4. Specialized electrical steel processed to align magnetic properties in the rolling direction. ↩

5. The unintended magnetic flux that does not link both primary and secondary windings. ↩