Transformer oil samples for Dissolved Gas Analysis (DGA) should be collected using clean, gas-tight sampling equipment, following the applicable sampling standard and the laboratory’s instructions. The sample must be representative of the transformer oil, protected from contamination and contact with air, correctly labelled, and transported and stored under appropriate conditions before testing.
Even small errors during sampling, handling or transport can affect the measured gas concentrations and lead to incorrect conclusions about transformer condition. Following recognised standards and consistent sampling procedures helps utilities and maintenance teams obtain more reliable results, identify developing faults and make better-informed maintenance decisions.
DGA measures gases dissolved in transformer insulating oil. These gases can provide evidence of electrical and thermal activity inside the transformer, including overheating, partial discharge and arcing.
Commonly measured gases include:
These gases must be interpreted as a pattern, alongside concentration trends, transformer history and operating conditions. No single gas automatically confirms a specific fault.
If air enters the sample, gas escapes during handling, or the sample becomes contaminated, the laboratory result may not accurately represent the oil inside the transformer. That can complicate diagnosis and lead to unnecessary investigations, or delay the response to a developing problem.
Several standards are relevant, but they serve different purposes.
|
Standard |
Relevance |
|---|---|
|
IEC 60475:2022 |
Guidance on sampling insulating liquids, including liquids taken from electrical equipment. |
|
IEC 60567:2023 |
Guidance on sampling free gases and analysing free and dissolved gases in mineral oils and other insulating liquids. The 2023 edition refers oil sampling from equipment to IEC 60475:2022. |
|
ASTM D923-15(2023) |
Covers sampling electrical insulating liquids and includes techniques for DGA, water and particle analysis. |
|
IEEE C57.104-2019 |
Provides guidance on interpreting gases generated in mineral-oil-immersed transformers, including DGA limitations and diagnostic methods. |
Sources: IEC 60475 and IEC 60567 information, ASTM D923 and IEEE C57.104-2019.
Important distinction: IEEE C57.104 focuses on interpreting DGA results; it should not be treated as a standalone, step-by-step oil sampling procedure. For sampling, consult the applicable sampling standard, your laboratory’s instructions and your organisation’s approved procedures. Always check the current editions and any applicable local requirements.
Sampling should be performed by trained, authorised personnel using the equipment manufacturer's instructions and the site's safety procedures. Transformer sampling points may be pressurised, hot or located near hazardous electrical equipment, so safety requirements take priority over the sampling process.
Use a clean, dry sampling device and a container approved for the required test by the laboratory. DGA samples commonly use gas-tight syringes or suitable gas-tight cylinders, depending on the method and laboratory requirements.
Before sampling, confirm that:
Do not assume that a container suitable for routine oil quality tests is also suitable for DGA. Different tests may require different containers and handling methods.
Use the designated sampling point specified by the transformer manufacturer, asset owner or approved sampling procedure. The aim is to obtain a sample that represents the insulating liquid being assessed.
Record the sampling point and use a consistent location for routine trending wherever practicable. Samples taken from different locations or under different conditions may not be directly comparable.
Prevent air from entering the sampling system and minimise exposure of the oil to the atmosphere. Air contact can affect the sample's gas composition, while poor sealing or unsuitable handling can allow dissolved gases to escape.
Follow the approved procedure for preparing the sampling point and collecting the sample. Avoid turbulence, leaks, unnecessary transfers between containers and any handling that could introduce air or cause gas loss.
Follow the laboratory's instructions for preparing the sampling point, collecting the oil and filling the selected container. The required method depends on the sampling equipment and transformer arrangement.
The sample should represent the transformer oil at the time of collection, rather than stagnant oil or material affected by contamination at the sampling connection. Do not improvise the process if the sampling valve, container or equipment behaves unexpectedly.
After collection, secure the container according to the approved method. Check for leakage, damaged seals, incorrect filling, visible contamination or other signs that the sample may have been compromised.
For containers designed to exclude air, check that they have been prepared and sealed in accordance with the laboratory's requirements. If the container is damaged or the sampling process was questionable, document the issue and ask the laboratory whether a new sample is required.
Correct collection is only half the process. Storage, transport and documentation can also affect sample quality.
Each sample should be traceable to the correct transformer and sampling event. Include the information required by the laboratory and your asset management process, such as:
These details help engineers interpret the result in context and compare it with earlier samples.
Keep the container sealed and protected from damage, contamination and unsuitable environmental conditions. Avoid unnecessary delays and follow the laboratory's requirements for transport, storage temperature and delivery time.
Do not transfer the sample into another container unless the laboratory's approved procedure specifically allows it. Every additional transfer creates another opportunity for contamination, air ingress or gas loss.
Use a laboratory with an appropriate method for the insulating liquid and test being requested. Ensure that the laboratory receives the sample identification, requested tests and relevant equipment history.
If there has been a leak, the sample may contain air, or the container has been damaged, tell the laboratory before relying on the result. The laboratory can advise whether the sample remains suitable for analysis or should be recollected.
DGA results are most useful when assessed alongside historical measurements, gas generation rates, operating conditions and other diagnostic information.
A single result can provide valuable evidence, but it may not tell the whole story. Consistent sampling and documentation make it easier to distinguish a genuine change in transformer condition from a difference caused by sampling or handling.
Consider a transformer that has been sampled routinely for several years. A new sample is collected using a poorly sealed connection, and air enters the sampling system.
The laboratory reports a change in gas concentrations. Without checking the sampling history, the maintenance team might interpret the result as evidence of a new transformer issue and initiate additional investigations.
The opposite problem is also possible: if dissolved gases escape before analysis, the measured concentrations may underrepresent the condition of the oil. This could complicate the assessment of a developing fault.
In either case, the appropriate response is to review the sample's integrity, compare the result with historical data and consult the laboratory or a qualified transformer specialist before deciding on corrective action.
The practical lesson: reliable DGA depends on both reliable analysis and reliable sampling. A laboratory cannot fully correct for a sample that no longer represents the oil inside the transformer.
Routine laboratory DGA provides a snapshot of transformer condition at the time the sample was collected. Online DGA monitoring complements this approach by tracking gas concentrations over time, helping asset managers identify changes between scheduled sampling events.
Insulect supplies online DGA monitoring solutions, including Qualitrol Serveron systems for hydrogen-only and multi-gas monitoring. These systems can support earlier fault detection and condition-based maintenance, particularly for critical transformers where changes in gas levels need closer attention.
Online monitoring does not eliminate the need for laboratory testing, correct sampling or expert interpretation. Instead, it provides another source of condition data that can help teams investigate abnormal trends and plan maintenance more effectively.
Use this checklist as a high-level quality check alongside your approved site procedure and laboratory instructions.
The required volume depends on the laboratory's analytical method and sampling container. Confirm the required volume and container type with the laboratory before collecting the sample.
Yes. Air ingress and exposure can alter the sample's gas composition, while dissolved gases may escape during unsuitable handling. This can make the results less representative of the transformer oil.
Sampling arrangements and site procedures vary. Follow the transformer manufacturer's instructions, the asset owner's safety requirements and the approved sampling procedure. Only trained, authorised personnel should perform the work.
Send it as promptly as practicable, following the laboratory's specified transport and storage requirements. The appropriate timeframe depends on the container, test method and laboratory procedures.
There is no single interval suitable for every transformer. Sampling frequency should reflect factors such as asset criticality, age, operating conditions, historical DGA results and the organisation's maintenance strategy. Abnormal results may justify additional testing or more frequent monitoring following an engineering assessment.
The two approaches serve complementary purposes. Laboratory testing provides periodic measurements, while online monitoring tracks gas concentrations over time. The appropriate approach depends on transformer criticality, risk, budget and the level of condition visibility required.
Reliable transformer maintenance starts with trustworthy condition data. Correct oil sampling and handling help ensure that DGA results provide a meaningful picture of transformer health, supporting better fault diagnosis and maintenance planning.
Insulect supplies DGA monitoring systems and transformer monitoring solutions to help utilities, generators and industrial operators improve asset visibility and detect developing faults.
Contact Insulect to discuss your transformer monitoring requirements and identify an approach suited to your assets.