Introduction
An electrical arc is a luminous, high-temperature discharge of electricity that occurs when current flows through an ionized gas path between two conductors. While arcs are essential for some industrial processes like welding, in electrical power systems they represent a serious hazard. Arc faults—unintended electrical discharges—are a leading cause of electrical fires worldwide. In Vietnam, for instance, electrical faults accounted for 74.8% of fire incidents in 2024, with the majority attributed to dangerous arc faults.
Arc detection is the technology designed to identify these hazardous electrical arcs before they can cause damage, injury, or loss of life. This article explores the principles, methods, applications, and future of arc detection technology.
What Is Arc Detection?
Arc detection refers to the process of identifying the presence of an electrical arc in a circuit or system. Once detected, the system can trigger protective measures—typically disconnecting the circuit to stop the arc. Arc detection is implemented in various devices, including:
- Arc Fault Detection Devices (AFDDs): Specialized protective devices that detect and interrupt arc faults
- Arc Fault Circuit Interrupters (AFCIs): Circuit breakers with integrated arc detection capability
- Dedicated arc detectors: Standalone systems for specialized applications like RF infrastructure protection
Why Arc Detection Matters
The importance of arc detection stems from the destructive nature of electrical arcs:
Fire Prevention: Arcs generate intense heat—temperatures can reach thousands of degrees Celsius—which can easily ignite surrounding materials. Damaged cable insulation, which can occur through aging, improper installation, or even rodent damage, can create arc faults. If power is not immediately interrupted, this discharge generates sufficient heat to start a fire.
Equipment Protection: In high-power environments, arcing can cause catastrophic damage to expensive equipment. The ARC Sentry system, originally developed for the National Synchrotron Light Source II, detects optical evidence of arc propagation and issues an interlock signal to stop propagation before hardware damage occurs.
Personnel Safety: Beyond fire hazards, arcs pose direct risks of electric shock and burns.
Challenging Detection: Not all arcs are easy to detect. Series arcs—where the arc occurs in series with the load—are particularly problematic because there is little change in load current when an arc occurs, making them difficult to detect with conventional methods.
How Arc Detection Works
Arc detection technologies can be broadly categorized by the physical phenomenon they monitor and the signal processing methods they employ.
Detection Modalities
Current-Based Detection: Most common in circuit protection devices. Arc faults generate distinct high-frequency current signatures that differ from normal load currents. These high-frequency components can be detected and analyzed.
Voltage-Based Detection: Monitoring voltage patterns across the arc gap can provide additional discrimination capability.
Electromagnetic Field Sensing: Arcs generate electromagnetic fields. Some detectors use field sensors to detect the characteristic random chaotic patterns associated with arcing.
Optical Detection: Arcs emit light, including ultraviolet and visible radiation. Optical sensors can detect this light emission. Some systems use fiber optic cables to capture ultraviolet signals for arc detection. Infrared sensors can also detect the heat concentration caused by arcs.
Electromagnetic Radiation Detection: Arcs produce electromagnetic radiation (EMR) across a wide spectrum. Advanced systems can even use antenna arrays to estimate the direction of arrival of arc-induced EMR for fault localization.
Signal Processing Algorithms
Once a signal is acquired, sophisticated algorithms are required to distinguish genuine arcs from normal operations and transient events.
Fast Fourier Transform (FFT): FFT is widely used because it can efficiently analyze high-frequency signals generated outside of normal AC signals. A study implementing 2048-point FFT using radix-2 and radix-4 architectures on FPGA detected arcs up to 122,880 Hz. FFT is particularly suitable for serial arc identification.
Filter-Based Methods: Mixed filter approaches combine average filters (for noise removal) and band-pass filters (for isolating specific frequency components). After normalization and FFT, total spectral energy is calculated—arcs and steady states can be easily discriminated by calculating spectral energy.
High-Pass Filtering: Some algorithms use high-pass filters (HPF) to detect series arc currents by analyzing differences in high-frequency amplitudes. When arc waveforms exceed a certain threshold compared to the average of normal waveforms, the algorithm triggers detection.
Adaptive Thresholding: Advanced algorithms use adaptive thresholds derived from scaled moving averages of signal power. This allows dynamic adaptation to changes in baseline noise, reducing false alarms in noisy environments. Time thresholds are also incorporated to discriminate between arcs and short-duration transient events like switching operations.
Discrete Wavelet Transform (DWT): DWT-based algorithms can detect arc faults within milliseconds while preventing false detection for short noise signals.
Artificial Intelligence: Modern designs integrate AI-powered detection. Microcontrollers running AI algorithms in real time can correlate current and voltage patterns to improve accuracy and reduce false positives.
Applications Across Industries
Arc detection technology has found applications across diverse sectors:
Residential and Commercial Buildings
AFDDs are increasingly mandated in building electrical installations. Siemens’ SENTRON AFDD, using patented SIARC technology, can detect all arc fault types and disconnect circuits within one second. These devices detect both parallel and series arc faults and can replace existing MCBs or connect to them.
Photovoltaic Systems
Solar power systems present unique challenges for arc detection. Texas Instruments’ TIDA-010231 reference design implements a four-channel analog front end for DC arc detection in photovoltaic systems, supporting DC voltages up to 1000V and currents up to 10A. Arcing is detected by analyzing AC noise present on the DC current between solar panels and inverters.
Aerospace
The aerospace industry’s transition toward electrical propulsion has led to high-voltage DC electrical systems in aircraft, raising critical safety concerns. Novel arc detection algorithms analyze high-frequency current signals across multiple frequency bands, capitalizing on the distinct spectral signatures of arcs.
RF and Microwave Infrastructure
High-power RF environments are vulnerable to arcing damage. The ARC Sentry system protects critical RF infrastructure with features including multi-channel support for up to 24 channels, rapid detection and shutdown in less than 3 microseconds, and high-speed sampling to reduce false alarms.
Railway and Low-Voltage Systems
Mixed filter and frequency analysis techniques have been proposed for arc diagnosis of low-voltage electric wires in railway applications.
Standards and Regulations
Arc detection devices are governed by international standards. AFDDs must comply with IEC 62606 (“General requirements for Arc Fault Detection Devices”), while installation requirements follow IEC 60364-4-42 for electrical safety.
Challenges and Future Trends
Despite significant advances, several challenges remain:
False Alarms: Differentiating arcs from normal transient events like motor startups or switching operations remains a key challenge. Modern algorithms address this through multi-band analysis, adaptive thresholding, and time-domain discrimination.
Series Arc Detection: Series arcs remain difficult to detect due to minimal current change. Advanced frequency-domain analysis and AI are increasingly employed to address this limitation.
Real-Time Processing: Detection must occur fast enough to prevent damage. Some systems achieve detection and shutdown in microseconds.
Cost and Integration: Reducing the cost and footprint of arc detection while maintaining reliability drives innovation. Compact designs like Siemens’ single-module-width AFDD save 50% of distribution board space.
Emerging Trends:
- AI and machine learning for improved discrimination
- Multi-sensor fusion combining current, voltage, optical, and electromagnetic sensing
- Localization capabilities to identify the exact location of arc faults
- Integration with smart grid and IoT systems for remote monitoring
Arc detection technology has evolved from simple analog filters to sophisticated systems combining multi-band frequency analysis, adaptive algorithms, and artificial intelligence. As electrical systems become more complex and pervasive—from residential buildings to aircraft and solar farms—the importance of reliable arc detection will only grow. By identifying and interrupting hazardous arcs before they can cause harm, these technologies play a vital role in protecting lives, property, and critical infrastructure.
