Partial discharge (PD) testing has become a cornerstone of modern electrical condition monitoring, providing early warning of insulation degradation before it leads to catastrophic failure. This technique, grounded in decades of high-voltage engineering research, allows maintenance teams to identify microscopic electrical sparks occurring within the insulation of cables, transformers, switchgear, and rotating machines. By detecting these discharges during controlled voltage stress, engineers can pinpoint defects such as voids, cracks, or contamination that would otherwise remain hidden until a flashover occurs. The effectiveness of partial discharge testing lies not just in its ability to identify faults, but in its capacity to quantify the severity and location of damage, enabling truly condition-based maintenance. This article explores the principles, methods, advantages, limitations, and real-world applications of PD testing, providing a comprehensive guide for electrical professionals.

What Is Partial Discharge?

Partial discharge is defined by the International Electrotechnical Commission (IEC) as a localized electrical discharge that partially bridges the insulation between conductors. Unlike a full breakdown, PD events are small—often measured in picocoulombs—and occur repeatedly at weak points within the dielectric material. Common sources include:

  • Voids – gas-filled cavities in solid or liquid insulation where the electric field strength exceeds the breakdown strength of the gas.
  • Treeing channels – microscopic conductive pathways that grow over time due to electrical stress, often branching from contaminants.
  • Surface discharges – tracking along the surface of insulators, often caused by moisture or pollution.
  • Corona – discharges in air around sharp edges of energized conductors.

Each of these types emits high-frequency electrical pulses, acoustic noise, light, and chemical byproducts. Partial discharge testing capitalizes on these emissions to detect and locate insulation weaknesses.

How Partial Discharge Testing Works

The fundamental principle behind PD testing is to apply a voltage stress slightly above the normal operating level and measure the resulting discharge activity. The procedure typically follows these stages:

  1. Voltage application – A test voltage, often 1.2 to 1.5 times the rated voltage, is applied to the equipment using a resonant test set or variable-frequency transformer.
  2. Sensing – Discharges are detected using capacitive couplers, high-frequency current transformers (HFCTs), or transient earth voltage (TEV) sensors placed on the equipment or its grounding system.
  3. Signal acquisition – The high-frequency signals are captured by a PD detector and digitized for analysis.
  4. Measurement and classification – Software algorithms distinguish between internal PD, external noise, and interference. Parameters such as magnitude (in pC), phase-resolved pattern, and repetition rate are recorded.
  5. Localization – For long assets like power cables, time-domain reflectometry (TDR) or acoustic triangulation is used to pinpoint the exact defect location.

Modern PD testing systems can perform measurements both online (while equipment is energized) and offline (during planned outages). The choice depends on the asset type, accessibility, and risk tolerance.

Online vs. Offline PD Testing

Online PD testing is performed under normal operating conditions, making it ideal for continuous monitoring of critical assets such as transmission cables and transformers. It captures discharge activity under actual service stress, including load variations and temperature cycles. However, online measurements are more challenging due to background electrical noise from the grid and other equipment.

Offline PD testing isolates the equipment from the system and applies a controlled test voltage. This eliminates external noise and allows for more accurate measurement and calibration. Offline tests are standard for factory acceptance testing (FAT) and periodic maintenance of medium-voltage switchgear and rotating machines. The disadvantage is that the equipment must be taken out of service, and the test conditions may not perfectly replicate real-world stresses.

Both methods have proven effective, and many asset owners combine them: periodic offline baseline measurements supplemented by online continuous or periodic monitoring.

Types of Partial Discharge and Their Signatures

Recognizing the different forms of PD is essential for interpreting test results. Each type produces a characteristic phase-resolved pattern:

  • Internal discharge (cavity discharge) – Occurs in voids within solid insulation. Pattern shows symmetrical pulses in both positive and negative half-cycles, often with a "rabbit-ear" shape. Magnitude increases with voltage but saturates when the void becomes fully ionized.
  • Surface discharge – Appears as pulses clustered near the zero crossings of the voltage waveform, with asymmetry between polarities. Often associated with contamination or moisture on insulator surfaces.
  • Corona discharge – In air, corona produces sharp, narrow pulses in the negative half-cycle (for positive DC) or symmetric patterns at AC crests. It is usually harmless on overhead lines but indicates problems on enclosed equipment.
  • Treeing – Electrical trees grow over time and produce erratic, high-magnitude discharges with multiple pulses per cycle. The pattern can vary as the tree branches.

Advanced PD monitoring systems use machine learning algorithms to automatically classify these patterns, reducing the need for expert interpretation and enabling early warning systems.

Advantages of Partial Discharge Testing in Fault Identification

The primary strength of PD testing is its ability to detect incipient faults that other diagnostic methods—such as insulation resistance (IR) or tan delta measurements—may miss. Specific advantages include:

  • Early fault detection – PD activity often begins months or years before a catastrophic breakdown. Catching it early allows planned repairs rather than emergency shutdowns. For example, in medium-voltage cables, PD testing can identify water tree degradation at a stage where replacement is still avoidable.
  • Cost savings – Unplanned outages in industrial plants can cost hundreds of thousands of dollars per hour. PD testing prevents these losses by enabling predictive maintenance. The cost of a test is typically 1–5% of the asset replacement value.
  • Non-destructive methodology – When performed within the recommended voltage levels, PD testing does not damage the insulation. This contrasts with high-potential (hipot) testing, which can stress good insulation and cause premature aging.
  • Quantitative severity assessment – Unlike simple go/no-go tests, PD testing provides a trendable metric (pC magnitude, repetition rate, number of sources). This allows engineers to track deterioration over time and set alarm thresholds.
  • Localization capability – Using modern sensors and signal processing, PD testing can pinpoint fault locations within a few centimeters on cables and a few meters on transformers. This drastically reduces repair time.

These benefits make PD testing a cornerstone of condition-based maintenance programs in utilities, data centers, oil & gas facilities, and manufacturing plants worldwide.

Limitations and Challenges

Despite its effectiveness, PD testing is not a universal panacea. Several inherent limitations must be managed:

  • Need for specialized equipment and expertise – PD signals are weak (microvolt to millivolt scale) and easily masked by noise. Proper sensor selection, grounding, and shielding are critical. Technicians require training in high-voltage safety and signal interpretation. Many organizations outsource PD testing to specialized firms like Power Diagnostics.
  • Complex and inaccessible equipment – In switchgear with many compartments or in long cable runs, it can be difficult to place sensors close to potential defect sites. Acoustic sensors may not penetrate metal enclosures. For transformers, internal PD detection requires built-in sensors or temporary access through oil valves.
  • Interpretation challenges – Distinguishing between dangerous PD and harmless background noise (e.g., from power electronics or arcing in motors) remains a significant challenge. Pattern recognition algorithms help, but false positives still occur, especially during online testing in electrically noisy environments.
  • Limited effectiveness on certain defect types – Slow-deterioration mechanisms like thermal aging of paper insulation may not produce persistent PD until very late. Also, DC systems require different PD detection techniques because of the absence of a 50/60 Hz reference.
  • Environmental and operational factors – Humidity, temperature, and load variations affect PD activity. A single measurement may not reflect worst-case conditions. Continuous monitoring or periodic testing under consistent conditions is necessary for reliable trending.

Recognizing these limitations allows asset managers to design testing programs that combine PD testing with complementary diagnostics such as dissolved gas analysis (DGA) for transformers or tan delta for cables.

Applications Across Electrical Assets

Partial discharge testing is applied to a broad range of high-voltage equipment. Below are the most common applications with specific considerations for each:

Power Cables

MV and HV cables are among the most PD-prone assets, especially XLPE-insulated cables that suffer from water treeing. Offline PD testing with Very Low Frequency (VLF) voltage sources is standard for cable circuits. Online PD testing using HFCT clamps on ground leads can detect defects during operation. PD localization in cables is achieved through time-domain reflectometry with high accuracy. IEEE Standard 400.2 provides guidelines for field testing of shielded power cable systems.

Power Transformers

In transformers, PD can occur in the winding insulation, tap changers, bushings, or core gaps. Detection is typically performed using capacitive coupling through the bushing tap or by installing ultra-high-frequency (UHF) sensors inside the tank. Online PD monitoring is common for large power transformers to provide continuous surveillance. The results are combined with DGA to differentiate between PD and other fault types.

Switchgear and GIS

Gas-insulated switchgear (GIS) and air-insulated switchgear (AIS) are routinely tested for PD using TEV sensors that measure the transient earth voltage induced by discharges. These sensors are non-invasive and can be placed on the external surface of metal-clad equipment. Pattern recognition helps identify whether the PD source is inside the switchgear (e.g., loose particle, sharp edge) or external (e.g., corona from an incoming cable).

Rotating Machines

Stator winding insulation in generators and large motors is susceptible to PD from slot discharges, end-winding contamination, and insulation delamination. Testing is performed offline with a 50/60 Hz source connected phase-to-phase, using coupling capacitors at the machine terminals. Online partial discharge testing for rotating machines is also established, with sensors installed in the neutral or on surge capacitors. IEC 60034-27 covers the measurement of PD on rotating machines.

Standards and Best Practices

Following recognized standards ensures that PD test results are repeatable, comparable, and reliable. The key standards include:

  • IEC 60270 – The global standard for high-voltage partial discharge measurements. It defines measurement circuits, calibration procedures, and bandwidth limits (typically 100–500 kHz for narrowband or 100 kHz–1 MHz for wideband).
  • IEEE 400 Series – Specifically for power cable systems. IEEE 400.2 covers VLF PD testing; IEEE 400.3 deals with online PD testing.
  • IEC 60034-27 – Offline and online PD measurement on rotating electrical machines.
  • CIGRE Technical Brochures – Provide detailed guidance on PD testing in GIS, transformers, and cables.

Best practices include: always calibrate the PD detection system at the test site using a known charge injector; perform baseline tests on new or refurbished equipment; compare results over time rather than relying on single measurements; and use at least two different sensor types (e.g., HFCT + acoustic) for confirmation of ambiguous signals.

Conclusion

Partial discharge testing has proven to be one of the most effective tools for identifying insulation faults in high-voltage electrical equipment. Its ability to detect, locate, and quantify deterioration at an early stage makes it indispensable for asset reliability and safety programs. While challenges such as noise interference and the need for skilled interpretation remain, advancements in sensor technology, machine learning, and digital signal processing continue to improve the accuracy and accessibility of PD testing. When integrated into a comprehensive condition-based maintenance strategy—alongside other diagnostic techniques and adherence to international standards—PD testing delivers significant cost savings, reduces unplanned downtime, and extends the operational life of critical power system assets. Electrical engineers and maintenance managers who invest in PD testing capabilities will be well-equipped to prevent failures and ensure the long-term health of their infrastructure.