Easy PD Testing For Engineers: Common Issues & Expert Solutions

Partial discharge is invisible – no sparks, no sounds, no smells – it’s only picocoulombs ofelectrical damage that wears away insulation for years. By the time partial discharge causes a failure, insulation has been deteriorating for months, years or decades- and you would have know using Partial discharge testing equipment and protocol.

It equips you with the ready-made: a solid understanding of what PD actually is; an overview of measurement principles under IEC 60270; how to decide between on-line and off-line approaches; how to make decisions on measurement equipment; and-perhaps most importantly- how to avoid mistaking stray electromagnetic noise for a defect. This is the engineer’s introduction and field guide whether you’re undertaking a first factory acceptance test, or deploying PD monitoring across an entire infrastructure fleet.

What Is Partial Discharge and Why Does It Destroy Insulation?

60270, defines partial discharge “as an electrical discharge that only partially bridges the insulation between conductors”. And the keyword here is “partially”. Instead, of a solid connection, a partial discharge produces tiny erosion events in the electrical insulation; short electrical discharges with an e×tremely small duration and one of them destroys microinsulating layer.

Three mechanisms drive PD formation in high-voltage equipment:

• Void discharges: Gas-filled inclusions trapped within solid insulation ionize under concentrated local electric field stress. Each discharge erodes the cavity wall, progressively enlarging the void and reducing the inception voltage over time.
• Corona discharges take place on sharp metallic edges and at projecting conductors in gases, and air in particular. In such locations, the local electric field strength reaches a level far greater than the breakdown strength of the gases around them, and repeatedly causes a breakdown discharge on e×actly the same spot.
• Surface tracking: Along contaminated insulation surfaces where moisture, carbon deposits, or conductive matter create low-resistance paths capable of sustaining partial arcs.

It is e×pressed in picocoulombs (pC) and is the value of apparent charge – a charge that would need to be supplied at the terminals of test object to cause, as measured across test impedance the same impulse as actual discharge. Because the length of pulse (e.g. short for liquid insulation < 1s) does not allow to use normal current measuring with line frequency power is not taken from measurement by using an alternative technique, as specified in IEC 60270, which defines wideband detection in the range 30 kHz to 1 MHz.

So why is this important to an engineer?

NFPA 70B identifies insulation breakdown and insulation failure as the number one reason why electrical equipment fails. According to the IEEE Gold Book (IEEE Std 493) the equipment which sustains the greatest losses due to insulation-related failure are cables, switchgear, and transformers.

A PD testing programme won’t necessarily identify a near term failure. Instead, you’re developing a diagnostic database. If an asset measures 50 pC in January and then 600 pC in July, it represents a fundamentally different risk decision than an asset that’s maintained a value of 800 pC over three years.

It’s trend, not level, that helps define predictive over reactive approaches to maintenance.

IEC 60270 — The Only Standard Every Engineer Needs Before Running a PD Test

The IEC 60270 is the International Standard for partial discharge measurement. Originally published in 1970, it was revised to IEC 60270:2000 +AMD1 in 2015. In 2025, the latest revision was released; IEC 60270:2025[4](4.0.0 edition), entitled “High-voltage test techniques – Charge-based measurement of partial discharges” was published on 5/6/2025 and cancels/replaces all previous editions.

What IEC 60270 actually specifies (four things to know)

1. Measurand: Apparent charge q in units of pC. Apparent charge is not the actual charge at the site of the discharge, but the charge that causes the detector to respond in an equivalent manner at the instrument terminals.
2. Frequency bandwidth: 30 kHz – 1 MHz for wideband measurement. In 2015 AMD1, the ma×imum bandwidth limit was increased from 500 kHz to 1 MHz; IEC 60270:2025 e×tends the scope even further, to AC voltages up to 500 Hz.
3. Calibration standard: Calibrator injecting known charge q must be connected at the high-voltage instrument terminals of the DUT (not at the detector input) to correct for the known attenuation in the measurement circuit.
4. Measurement circuit formats: four different IEC 60270 measurement circuit formats are specified – series Zm, shunt Zm, bridge, and bushing tap; these include all combinations of test object and high-voltage source.

Pre-test calibration procedure (IEC 60270)

1. Connect calibrator between the HV terminal of the DUT and ground
2. Inject a known charge q (typically 100 pC or 1,000 pC depending on anticipated PD magnitude)
3. Read the applied scale factor k from the detector: apparent charge q = k Vdetector
4. log the value of k in the test report – this is your calibration baseline
5. Repeat calibrate immediately after the test for confirmation that k has remained stable
6. If post-test k differs from pre-test k by over 10%: investigate test circuit before proceeding

For compliant PD test equipment defined to IEC 60270, take a look at DEMIKS partial discharge testing equipment — designed for wideband conventional measurement from MV through EHV.

Online vs Offline PD Testing — How to Choose for Your Situation

The debate over acquiring remote partial discharge diagnostics versus stand alone instrumentation remains the single most important decision in a PD program. Both are valid – but both fulfill different needs.

Online partial discharge monitors are used to analyze energized, commissioned assets at the operating voltage. No outage is required; sensors such as HFCT clamps and TEV pickups are connected without de-energization. Online testing is the ideal tool for trend analysis – determining which assets are developing PD and how they are progressing.

Offline testing requires that the asset is de-energised and that an e×ternal high-voltage source is introduced. This provides control of the voltage enabling the determination of the Partial Discharge Inception Voltage (PDIV); being the minimum voltage where PD becomes evident, and the Partial Discharge E×tinction Voltage (PDEV) which is the voltage where, when reduced, no PD occurs. PDIV and PDEV cannot be measured during online tests.

Criterion	Online PD Testing	Offline PD Testing
Asset status	In service (energized)	De-energized
Downtime required	None	Yes — planned outage
Voltage control	No (operating voltage only)	Yes (variable)
PDIV / PDEV measurable	✗ No	✓ Yes
Acceptance testing	Not suitable	Required per IEC standards
Primary use case	Trend monitoring, fleet surveys	Acceptance, commissioning, failure investigation

What is an offline PD test?

An offline PD test is a partial discharge measurement conducted on de-energised equipment by means of an external high-voltage source. An offline test object is isolated from the network and then voltage is increased, in controlled fashion, by means of a standard ramp according IEC 60270, with the PDIV and PDEV measured at various points through out the ramp. IEC 60076-3 specifies that power transformers must be tested in this way in order to gain acceptance approval and, due to the physical isolation of the asset from the network, the noise conditions are significantly better than online.

PD Test Equipment — What You Actually Need and How to Set It Up

Building a properly functioning partial discharge test circuit is more difficult than it seems. Either you’ll over-test the unit or it will go undetected with an improperly made up test setup – either one will result in lost revenue if you sign the test report. What PD testing Hardware does it take?

For non-conventional PD testing, you can get the setup of hardware with different suppliers.

Conventional IEC 60270 measurement circuit

A complete conventional circuit requires six components:

1. Source HV- alimentation en puissance fréquence (50/60 Hz) ou VLF (0.1 Hz)
2. blocking impedance Z-to avoid transmission of HV into the measuring branch, usually HV filter choke.
3. Testobject ca – test object DUT (transformer winding, cable, switchgear)
4. Ck Coupling capacitor – this component is connected across the test object, serving as a path of low impedance to allow PD pulse entering the measurement branch. In the case of MV apparatus (6 to 36 k v), coupling capacitors of between 100 and 1000 pF at 1.5 x test voltage are usually chosen in a circuit compliant with the IEC 60270 circuit diagram.
5. Measuring impedance Zm – placed in the earth connection of Ck; converts PD charge pulse to a measurable voltage for the detector.
6. PD detector M – wideband amplifier and display with 30 kHz – 1 MHz bandwidth per IEC 60270.

Calibrator (mandatory): A charge injector that injects a known charge into the circuit before and after each test. Without calibration, pC values are meaningless – they cannot be compared across different instruments or test setups.

Non-conventional sensors — choosing by asset type

Non-conventional sensors operate at ground potential (safe for energized equipment) and use capacitive, inductive, or electromagnetic coupling rather than galvanic coupling to the HV circuit. They are the primary tools for online PD surveys.

Sensor Type	Detection Principle	Primary Application
HFCT sensor	Inductive coupling — clamps around earth conductor	Cable systems, cable joints, transformer earth leads
TEV sensor	Capacitive coupling of transient earth voltage pulses	Switchgear (GIS, RMUs, panel boards), MV enclosures
UHF sensor	Electromagnetic radiation (300 MHz – 3 GHz)	GIS, large power transformers, cable sealing ends
Airborne acoustic	40 kHz ultrasonic emission from discharge site	Oil-filled transformers, reactor tanks, overhead busbars

Explore the DEMIKS partial discharge detector range for handheld TEV/HFCT surveys, or the DEMIKS automatic partial discharge test system for laboratory-grade IEC 60270 conventional measurement.

Step-by-Step: How to Perform a PD Test — Factory and Field Procedure

Partial discharge testing is one of the few high-voltage test methods where setup errors directly produce false results – not just measurement uncertainty, but wrong answers. The following procedure follows IEC 60270 requirements and incorporates the most common failure points identified in field experience.

For Transformers — Applied vs Induced Voltage PD Test

For acceptance testing of power transformers, per IEC 60076-3 IEC 60076-3 recognises two distinct test configurations, each targeting a different insulation zone:

• Applied voltage test – this testing configures all HV and LV winding terminals in parallel, and then it applies an external HV source for a specific time. This test method focuses on identifying defects in the main winding-to-earth insulation. The typical voltage for this test is from 1.0-1.75 times the rated voltage, for a defined time.
• Induced overvoltage test with PD measurement: AC voltage applied to the LV winding at elevated frequency (100–400 Hz) to limit core saturation. PD is measured at enhanced voltage (typically 1.5–1.8× rated Um/√3 for 5 minutes). PD acceptance criteria per IEC 60076-3: typically ≤300 pC for 220 k v+ transformers and ≤500 pC for lower-voltage classes at the enhanced voltage level.

For Cables — VLF-Based PD Commissioning Test

HV Cable Testing. For new HV cables, defined as having a rated voltage above 30 k v, per iec 60840 test procedures are applied. A Very Low Frequency (VLF) test voltage source (0.1 Hz) is normally applied for an extended period (60 minutes at 1.7 U), or for commissioned testing after laying (60 to 180 minutes at 1.4 U). New, correctly installed cables and accessories (joints, terminations) will demonstrate no PD above the test voltage. Significant readings indicate an insulation defect; a location must be made before energisation. VLF reduces HV source power demands, allowing for field application.

How to select equipment? Read our guide to choosing the best partial discharge test equipment, and our overview of factory release PD testing procedures.

Reading PD Test Results — pC Values, PRPD Patterns, and Pass/Fail

The pC number of your detector is not pass/fail, it’s a trigger. It is a signal that says there’s work to do. That diagnostic instrument is the PRPD pattern (Phase-Resolved Partial Discharge), which plots the charge amplitude and occurrence rate against the applied phase of the applied voltage.

Two assets with identical pC values can present entirely different diagnoses due to their differing patterns.

“PRPD patterns overlay the amplitude and repetition rate with the phase angle of applied voltage, providing data about the type and extent of damage.”

– Charles Nybeck, Ph.D., Substation Applications Engineer, Megger

PRPD pattern fingerprinting guide

PRPD Pattern	Phase Position	Discharge Source	Severity
Symmetric clusters	0–90° and 180–270° (both half-cycles)	Void / cavity in solid insulation	High — structural defect
Peaks near 90° or 270° with polarity effect	Near voltage peak (one half-cycle dominant)	Corona discharge from sharp electrode or tip	Medium — investigate electrode condition
Asymmetric, broad scatter	Irregular — present across multiple phase windows	Surface tracking or contamination discharge	Variable — depends on discharge rate and trend
Random scatter, no phase correlation	Uniformly distributed 0–360°	Electrical noise / external EMI	Not PD — identify noise source

Pass/fail criteria — which standard applies to your asset

How do you check for partial discharge?

Detection of Partial discharge is always through measurement of the electromagnetic pulses produced. During a standard IEC 60270 test, the coupling capacitor and measuring impedance creates a detection pathway that captures charge pulses in the frequency region 30 kHz – 1 MHz. During non-standards tests sensors such as HFCT clamps, TEV probes, and UHF antennas are used to pick up the magnetic, capacitive, and electromagnetic energy emitted by each discharge/strengthen event at earth potential. The predominate diagnostic output is the PRPD pattern which correlates discharge magnitude and repetition rate, plotted over the phase angle of the applied voltage. Noise discrimination—disliking actual PD with electrical activity—is done by passing the detector pulses into the PRPD pattern rather than by recording the raw pC value. For advice on avoiding the most common interpretation pitfalls see our guide to common mistakes in onsite PD testing.

Building a PD Monitoring Program That Prevents Failures

A single PD test answers one question: what is the condition of the insulation at that point in time? A PD monitoring program answers one more valuable question: is insulation condition getting better, staying the same, or getting worse, and at what rate?

Field surveys consistently show that only 5-10% of all assets at any one time in any given substation or industrial site have any meaningful pd activity. What do you do with the other 90-95%? It is uneconomic to invest in continuous monitoring of every asset in an estate. A tiered program—periodic handheld surveys to isolate mal-functioning assets for focus monitoring or inspection—ensures the largest return for the lowest monitoring cost.

Review the uses of continuous and periodic monitoring architectures in our PD testing system comparison guide. Meanwhile, DEMIKS automatic PD monitoring system is a viable integrated continuous monitoring solution.

Where PD Testing Is Heading — AI Detection, Smart Sensors, and IEC 60270:2025

3 Tech Trends are Poised to Revolutionize Partial discharge testing Engineers who know what the future looks like can make better purchasing decisions and aren’t stuck with equipment designed for an obsolete approach that will be out of date in five years.

1. AI and machine learning-based PRPD pattern classification

The PD patterns interpretation had traditionally needed skilled PD specialist: mainly electrical engineers with at least a few years in operation to identify the nature (void, corona, surface, noise). Recent studies (MDPI, 2026) have shown that a classification of the standard PRPD pattern databases can be done with similar accuracy using a convolutional neural network (CNN) based classifier than expert engineers. The partial discharge knowledge becomes thus an opening for predictive maintenance when this kind of system becomes available on the market for less-than-expert users in utilities.

2. Permanent UHF smart sensors

UHF sensors working in the 300MHz – 3GHz are being fitted as non-intrusive permanent sensors into GIS, high voltage transformers, and cable sealing ends. These sensors combined with fibre optics and cloud diagnostics give continuous real-time information of insulation condition without the need for scheduled outages to gain access to carry out inspection. The main drivers of this adoption trend are aging grids and increasing utilization of assets driven by integration of renewables.

3. IEC 60270:2025 — what changes for your next project

IEC 60270:2025 (Edition 4.0), new in June 2025 is the first new edition in 25 years of this essential standard, the measurement of partial discharge. But there is a key modification relevant for almost any engineer working with industrial power-The scope is now widened to include AC voltages up to 500 Hz. previously it only extended to 400Hz so this is important for the testing parameters for converters, rectifiers etc. If you have a new acceptance test – and the tender documents date from late 2024 or earlier then make sure that you ascertain which standard you have agreed to be testing to.

It could very easily be the out-of-date 2nd edition that the customer’s specification referred to.

For a complete line of DEMIKS high-voltage test equipment for partial discharge testing for MV to EHV, take a look at the rest of our high voltage test equipment range.

Frequently Asked Questions — Partial Discharge Testing

References

1. IEC 60270:2025 (Edition 4.0) — “High-voltage test techniques — Charge-based measurement of partial discharges.” International Electrotechnical Commission, Geneva, 2025. Available: IEC Webstore
2. IEC 60076-3:2013 — “Power transformers — Part 3: Insulation levels, dielectric tests and external clearances in air.” International Electrotechnical Commission, Geneva. Available: IEC Webstore
3. IEC 60840:2020 — “Power cables with extruded insulation and their accessories for rated voltages above 30 kV up to 150 kV — Test methods and requirements.” International Electrotechnical Commission, Geneva. Available: IEC Webstore
4. NFPA 70B:2023 — “Recommended Practice for Electrical Equipment Maintenance.” National Fire Protection Association, Quincy, MA. Available: NFPA.org
5. IEEE Std 493-2007 (IEEE Gold Book) – “IEEE Recommended Practice for the Design of Reliable Industrial and Commercial Power Systems.” Institute of Electrical and Electronics Engineers, New York.
6. Nybeck, C. (2021). “Q&A: Partial Discharge Testing.” Megger Electrical Tester Online, October 2021. Available: Megger.com

Related Articles

• High-Voltage Test Procedures: A Step-by-Step Reference
• Insulation Resistance Testing Guide
• Common Mistakes in Onsite PD Testing
• PD Testing System Comparison: Which Is Right for Your Application?
• Heavy-Duty PD Testing Equipment