Understanding Partial Discharge Detection in Medium Voltage Switchgear: Technologies, Standards, and Best Practices

Home » News » Understanding Partial Discharge Detection in Medium Voltage Switchgear: Technologies, Standards, and Best Practices

Partial discharge (PD) is one of the most insidious threats to the reliability and safety of medium voltage (MV) switchgear. Left undetected, PD activity progressively degrades insulation materials, eventually leading to complete dielectric breakdown, catastrophic equipment failure, arc flash incidents, and unplanned outages that can cost industrial facilities and utilities millions of dollars in downtime and repair expenses. Understanding how to detect, measure, and interpret partial discharge signals is therefore a cornerstone of modern power grid condition monitoring strategy.

What Is Partial Discharge?

Partial discharge is a localized electrical discharge that only partially bridges the insulation between conductors. Unlike a full breakdown or short circuit, PD does not immediately cause failure—but it is a reliable precursor. PD occurs when the local electric field exceeds the dielectric strength of a small void, crack, or contamination site within the insulation system, while the bulk of the insulation remains intact.

In medium voltage switchgear (typically rated 1 kV to 38 kV), PD can originate from multiple sources: voids in epoxy resin insulation, delamination in busbar sleeves, surface tracking across contaminated insulators, floating metal particles, loose connections, and corona discharge from sharp conductor edges. Each type produces distinct electrical, electromagnetic, and acoustic signatures that can be detected and analyzed using specialized grid monitoring equipment.

Why PD Detection Matters for Asset Owners

The economic case for continuous PD monitoring is compelling. According to industry data from CIGRE and IEEE, approximately 40% of all MV switchgear failures originate from insulation defects that could have been detected through PD monitoring months or even years before failure. A single unplanned outage at a medium-sized industrial plant can cost between $50,000 and $500,000 per hour in lost production. When you factor in equipment replacement costs, safety investigations, and reputational damage, the return on investment for a comprehensive online monitoring system becomes self-evident.

Beyond pure economics, PD monitoring addresses critical safety concerns. Arc flash incidents in MV switchgear can release energy equivalent to several kilograms of TNT, causing severe burns, hearing damage, and fatalities. Insurance underwriters increasingly require evidence of condition-based maintenance programs that include PD surveillance as a prerequisite for coverage.

Partial Discharge Detection Technologies

1. Ultra-High Frequency (UHF) Detection

UHF detection measures electromagnetic waves in the 300 MHz to 3 GHz range that are emitted by PD events. Because PD pulses have extremely fast rise times (typically sub-nanosecond), they radiate electromagnetic energy across a broad spectrum. UHF sensors—typically installed inside the switchgear compartment through dielectric windows or built into spacer insulators—capture these signals with high sensitivity and excellent immunity to external corona and switching noise.

UHF technology is particularly effective in gas-insulated switchgear (GIS) and air-insulated switchgear (AIS) with metal-clad construction, where the metallic enclosure acts as a Faraday cage that attenuates external interference. The technique can detect PD magnitudes as low as 0.5 pC and provides localization capabilities through time-of-flight analysis using multiple sensors. Integrated PD monitoring devices that combine UHF sensing with other modalities offer the most comprehensive diagnostic picture.

2. Transient Earth Voltage (TEV) Measurement

TEV measurement is a non-intrusive technique that detects the transient voltage pulses appearing on the earthed metal surfaces of switchgear when internal PD occurs. These pulses, typically in the range of 1 mV to several volts, propagate along the enclosure surface and can be detected by capacitive coupling probes temporarily or permanently attached to the switchgear exterior.

The key advantage of TEV is that it requires no internal sensor installation, making it ideal for retrospective deployment on energized equipment and for periodic walk-around surveys using portable instruments. TEV amplitude correlates with PD severity, though the relationship is not linear and requires calibration. Modern online monitoring systems combine continuous TEV monitoring with trend analysis algorithms that detect the characteristic “rising edge” pattern of developing PD faults long before they become critical.

3. Acoustic Emission (AE) Detection

Acoustic PD detection relies on the fact that every partial discharge event generates a pressure wave—essentially a miniature sonic boom—as the localized heating causes rapid expansion of the surrounding medium. AE sensors, typically piezoelectric transducers operating in the ultrasonic range (20 kHz to 300 kHz), are mounted on the switchgear enclosure or directly on high-voltage components.

Acoustic monitoring excels at locating PD sources through triangulation using multiple sensors with known geometry. It is completely immune to electromagnetic interference, making it the preferred technique in electrically noisy environments such as arc furnace installations, rolling mills, and traction power substations. However, acoustic signals attenuate significantly at material interfaces (metal-to-air, metal-to-epoxy), so sensor placement and coupling are critical design considerations.

4. High-Frequency Current Transformer (HFCT) Measurement

HFCT sensors clamped around the earth connection of switchgear or cable terminations detect the high-frequency current pulses that flow to ground during PD events. This technique is particularly useful for monitoring cable terminations, where PD often initiates at stress cone interfaces. HFCT-based monitoring can be integrated with cable sheath and partial discharge monitoring systems to provide end-to-end insulation health assessment for the entire cable-switchgear assembly.

Complementary Monitoring for Complete Switchgear Health

While PD detection is essential, a truly comprehensive switchgear condition monitoring strategy must incorporate additional parameters that either correlate with or contribute to PD development:

Temperature Monitoring: Hot spots at connection points accelerate insulation aging and create thermal stress that promotes void formation—a direct precursor to PD. Solutions like the SCYC-PWTM2304 Passive Wireless Online Temperature Monitoring Device provide continuous thermal surveillance of critical connection points (busbar joints, cable terminations, circuit breaker contacts) without requiring auxiliary power wiring. For ring main units, the SCYC-CW30 Passive Wireless Temperature Monitoring System offers equivalent capability optimized for compact RMU configurations. Both systems use surface acoustic wave (SAW) sensor technology, which is inherently passive—no batteries, no CT energy harvesting required—ensuring maintenance-free operation over decades.

Grounding Current Monitoring: Elevated grounding currents in transformer cores can indicate developing inter-laminar insulation faults that create circulating currents, localized heating, and potential PD initiation. The DTE2100 Online Monitoring Device for Grounding Current of Transformer Core provides real-time data on core grounding current magnitude, harmonic content, and trend analysis, enabling early detection of core faults that might otherwise go unnoticed until catastrophic failure occurs.

Arrester Monitoring: Deteriorating surge arresters lose their ability to clamp overvoltage transients, exposing switchgear insulation to voltage stresses far beyond design limits—a direct trigger for PD initiation. The DT801 Online Monitoring Device for Zinc Oxide Surge Arresters continuously tracks resistive leakage current, which is the most reliable indicator of MOV (metal oxide varistor) degradation, allowing replacement before arrester failure compromises switchgear protection.

Industry Standards Governing PD Detection

Several international standards define PD measurement methodologies and acceptance criteria relevant to MV switchgear:

  • IEEE C37.20.2-2015 — Standard for Metal-Clad Switchgear. Specifies design and production testing requirements, including dielectric and PD tests, for MV metal-clad switchgear used in North American markets.

Best Practices for PD Monitoring Program Implementation

Baseline Assessment

Before deploying continuous online monitoring, conduct a comprehensive baseline PD survey of all critical switchgear assets. This should combine off-line PD testing (where outages can be scheduled) with on-line techniques such as TEV and acoustic scanning. The baseline establishes the “normal” PD signature for each asset, against which future measurements are compared. Equipment with elevated baseline PD levels should be flagged for increased monitoring frequency or prioritized for internal inspection.

Sensor Placement Strategy

PD sensor placement should follow a risk-based approach. For UHF sensors, position them near known high-stress regions: cable termination compartments, bushing interfaces, busbar support insulators, and circuit breaker spouts. For TEV sensors on metal-clad switchgear, place probes on the front panel near each circuit breaker compartment and on side panels adjacent to busbar chambers. Acoustic sensors should be positioned with direct line-of-sight (through metal) to the suspected PD source, avoiding air gaps and gasketed joints where possible.

Alarm Thresholds and Trending

Data Management and Analysis

Modern PD monitoring generates substantial volumes of data—phase-resolved PD patterns (PRPD), pulse waveforms, trend logs, and spectra. Effective data management requires centralized storage, automated pattern recognition algorithms, and integration with computerized maintenance management systems (CMMS). Machine learning techniques, particularly convolutional neural networks trained on labeled PRPD pattern databases, are increasingly being deployed to automatically classify PD types (internal void, surface discharge, corona, floating particle) and assess severity without human interpretation.

Case Example: PD-Induced Failure in 11 kV Switchgear

Consider a typical scenario at a water treatment plant where three 11 kV incoming feeder switchgear panels had been in service for 18 years without incident. A routine thermographic survey detected a 35°C temperature rise at the busbar spout connection of Feeder 2 compared to adjacent panels. Follow-up TEV PD measurement revealed intermittent PD activity with peak amplitudes of 28 dBmV—well above the 15 dBmV threshold considered normal for this equipment class. An internal inspection during a scheduled outage confirmed severe tracking and erosion of the epoxy bushing at the spout interface. Had the PD activity continued unchecked, full dielectric breakdown was estimated to occur within 3–6 months based on the observed deterioration rate. The bushing was replaced, and a continuous PD and temperature monitoring system was installed on all three feeders to prevent recurrence.

The Future of PD Detection

The trajectory of PD detection technology is toward multi-parameter sensor fusion, edge computing, and cloud-based analytics. Next-generation systems will combine UHF, TEV, acoustic, and HFCT data streams with temperature, humidity, and load current inputs to build a holistic insulation health index—a single number that quantifies the probability of failure within the next maintenance interval. Edge computing nodes will perform real-time PD classification and severity assessment locally, transmitting only actionable alerts and summary data to the cloud, reducing bandwidth requirements and enabling deployment at remote substations with limited connectivity.

Conclusion

Partial discharge detection has evolved from a specialized laboratory technique into a mainstream condition monitoring tool that is essential for any organization that depends on medium voltage switchgear reliability. The availability of multiple complementary detection technologies—UHF, TEV, acoustic, and HFCT—means that a suitable monitoring approach exists for virtually every switchgear installation, regardless of age, configuration, or operating environment. When combined with complementary monitoring parameters such as temperature, grounding current, and arrester condition, PD monitoring provides a comprehensive window into switchgear health that enables truly predictive maintenance.

Related Products from Qingdao Britop

Qingdao Britop offers a comprehensive range of power grid monitoring solutions designed to complement partial discharge detection in medium voltage switchgear applications:

Contact Qingdao Britop today to discuss your switchgear condition monitoring requirements and learn how our power grid monitoring products can help you achieve the reliability, safety, and operational efficiency your facility demands.


This article is part of the Qingdao Britop Industrial Knowledge Base series.

Related Products:

Related Products

Scroll to Top