Transformer Condition Monitoring: How DGA and Online Monitoring Help Prevent Failures

Power transformers are among the most critical and expensive assets in electrical networks. A transformer failure can result in unplanned outages, equipment damage, production losses and lengthy replacement periods.

The challenge is that many transformer problems develop gradually before a major failure occurs.

Insulation deterioration, overheating, moisture contamination, partial discharge and on-load tap changer wear can all produce warning signs long before the transformer is forced out of service. The purpose of transformer condition monitoring is to detect these changes early enough for operators to investigate and respond.

Modern monitoring combines traditional inspection with technologies such as dissolved gas analysis, temperature monitoring, moisture measurement, partial discharge detection and OLTC monitoring.

Instead of relying only on fixed maintenance intervals, asset owners can use condition data to move toward a predictive maintenance transformer strategy based on the actual health of the equipment.

Why Transformer Condition Monitoring Matters

Traditional transformer maintenance often follows a calendar-based approach. Inspections, oil sampling and electrical tests are completed at predetermined intervals regardless of whether the transformer condition has changed.

This approach remains important, but it has limitations.

A transformer may develop a rapidly progressing fault between scheduled inspections. Conversely, equipment in good condition may undergo unnecessary maintenance simply because a predetermined interval has been reached.

A transformer monitoring system provides another layer of information by tracking operating conditions and signs of deterioration over time.

Depending on the transformer criticality and monitoring configuration, operators may observe:

• Dissolved gases in transformer oil

• Top-oil and winding temperatures

• Moisture levels

• Partial discharge activity

• Load current

• Cooling-system operation

• OLTC operation

• Bushing condition

• Oil level and pressure

• Ambient temperature

The objective is not simply to collect more data. The objective is to identify abnormal changes that may indicate developing problems.

For utilities, renewable energy projects, mining operations and industrial facilities, this can help maintenance teams prioritise inspections around asset condition and operational risk.

DGA Turns Transformer Oil Into an Early Warning System

Dissolved Gas Analysis, commonly known as DGA, is one of the most established techniques for assessing the internal condition of an oil-filled transformer.

Electrical and thermal faults can break down transformer oil and insulation materials, producing gases that dissolve in the oil.

A dissolved gas analysis transformer assessment examines the type, concentration and development of these gases.

What Can DGA Reveal?

Different fault mechanisms can produce different gas patterns.

Gases commonly evaluated include hydrogen, methane, ethane, ethylene, acetylene, carbon monoxide and carbon dioxide.

The results may provide evidence associated with conditions such as:

• Local overheating

• High-temperature thermal faults

• Low-energy electrical discharge

• Arcing

• Insulation-paper degradation

• Abnormal oil decomposition

For example, an increase in gases associated with thermal degradation may indicate overheating, while certain combinations of gases can suggest electrical discharge activity.

However, a DGA transformer diagnosis should not normally be based on one gas value in isolation.

Gas ratios, total combustible gases, transformer design, loading history, previous results and the rate of gas generation all contribute to interpretation.

Why the Trend Is Often More Important Than One Reading

A single laboratory result provides a snapshot.

Trend data provides context.

Suppose a particular gas concentration has been relatively stable for several years and suddenly begins increasing. That change may be more significant than a higher concentration that has remained stable for a long period.

For this reason, trending is a central part of transformer condition assessment.

Periodic laboratory DGA can establish long-term trends, while online DGA monitors can provide more frequent measurements for critical transformers where faster detection is valuable.

Temperature and Moisture Reveal Transformer Ageing Conditions

Transformer insulation life is strongly influenced by operating temperature and moisture.

Monitoring these parameters provides valuable information about both immediate operating conditions and long-term asset health.

Temperature Monitoring Shows the Transformer’s Thermal Stress

Transformer temperature depends on loading, losses, ambient conditions and cooling-system performance.

Typical measurements can include:

• Top-oil temperature

• Winding temperature

• Ambient temperature

• Cooling-system status

Higher-than-expected temperatures may indicate overloading, restricted cooling, fan or pump problems, abnormal losses or internal deterioration.

Temperature should therefore be interpreted together with load.

A transformer operating at high temperature under heavy load may be behaving differently from one showing the same temperature at relatively low load.

Long-term thermal performance is also relevant to transformer losses and efficiency, because excessive temperature increases conductor resistance and contributes to insulation ageing.

Online monitoring makes it possible to compare temperature against actual operating conditions rather than relying only on occasional site observations.

Moisture Can Reduce Insulation Reliability

Moisture is another important transformer condition indicator.

Water may exist in both the insulating oil and the solid cellulose insulation. Excessive moisture can reduce dielectric strength and accelerate insulation deterioration.

Moisture levels can change with transformer temperature because water moves between the oil and paper insulation as thermal conditions vary.

For this reason, moisture readings should be interpreted in context rather than treated as an isolated number.

Online moisture monitoring can be particularly useful where transformers are exposed to changing loads, ageing sealing systems or operating environments where moisture ingress is a concern.

Partial Discharge Can Identify Developing Insulation Problems

Partial discharge is a localised electrical discharge that does not completely bridge the insulation between conductors.

Although an individual discharge may contain relatively low energy, persistent activity can damage insulation over time.

Sources can include voids within insulation, defects around high-field regions, contamination or deterioration of insulation components.

Why Partial Discharge Monitoring Is Valuable

Partial discharge monitoring can help identify insulation problems before they develop into more serious faults.

Depending on transformer design and monitoring strategy, detection may involve electrical, acoustic or other sensing technologies.

The presence of partial discharge does not automatically mean that transformer failure is imminent.

Engineers need to consider:

• Discharge magnitude

• Repetition rate

• Pattern

• Location

• Operating voltage

• Loading condition

• Changes over time

As with DGA, the trend is important.

Stable low-level activity may require observation, while rapidly increasing or changing activity may justify further investigation.

DGA and partial discharge information can also complement one another. If both monitoring methods indicate abnormal electrical activity, maintenance teams have stronger evidence for targeted inspection or diagnostic testing.

OLTC Monitoring Targets One of the Transformer’s Most Active Components

The on-load tap changer, or OLTC, allows transformer voltage ratio to be adjusted while the transformer remains energised.

Unlike the transformer main tank, which contains relatively few moving components, an OLTC performs repeated mechanical and electrical switching operations during service.

This makes it an important focus for power transformer monitoring.

What Should Be Monitored on an OLTC?

Depending on the equipment design, monitoring may evaluate:

• Number of tap operations

• Tap position

• Motor operating current

• Operating time

• Oil temperature

• Contact wear indicators

• Vibration or acoustic behaviour

• OLTC oil condition

• Abnormal operation alarms

Changes in motor current or operating time, for example, can indicate increasing mechanical resistance or problems within the drive mechanism.

Operation counters can also help maintenance teams understand actual OLTC duty instead of assuming that all tap changers experience similar wear.

For transformers that regulate frequently because of changing renewable generation, industrial loads or network voltage conditions, monitoring actual OLTC activity can be especially useful.

From Monitoring Data to Predictive Maintenance

Installing sensors does not automatically create a predictive maintenance program.

The real value comes from turning measurements into maintenance decisions.

A successful predictive maintenance transformer strategy combines multiple information sources rather than relying on one alarm.

Combine Multiple Indicators

Consider a transformer showing increasing top-oil temperature.

Temperature alone may not identify the cause.

If the monitoring system also shows normal loading but declining cooling-system performance, the investigation can focus on fans, pumps or radiators.

If elevated temperature appears alongside increasing DGA gases associated with thermal faults, the issue may require a different level of investigation.

Similarly, rising moisture combined with changes in dielectric condition may justify more attention than either measurement considered separately.

Avoid Treating Every Alarm as a Failure

Monitoring systems require suitable alarm thresholds and escalation logic.

If limits are too sensitive, operators may receive frequent nuisance alarms and eventually ignore important warnings.

If limits are too broad, developing problems may remain unnoticed.

A better strategy considers absolute values, rate of change, transformer operating conditions and historical baseline data.

The system should help engineers identify meaningful deviations rather than simply generate large volumes of data.

Online Monitoring or Periodic Testing?

Not every transformer requires the same monitoring architecture.

Periodic oil sampling, thermography, inspections and electrical testing may be appropriate for lower-criticality assets.

Critical transformers may justify continuous online monitoring because their failure consequences are greater.

Factors influencing the decision include:

• Transformer MVA and voltage

• Asset age

• Loading

• Network criticality

• Availability of redundancy

• Replacement lead time

• History of abnormal results

• Environmental conditions

• Consequence of failure

A utility grid transformer supplying a critical load may require a more comprehensive monitoring package than a smaller transformer with full redundancy.

The correct approach is risk-based.

Online monitoring should be selected because it improves asset-management decisions not simply because additional sensors are available.

What to Include in a Transformer Monitoring System Specification

Monitoring requirements are best considered during transformer specification rather than added after manufacturing.

A power transformer specification checklist should identify which sensors, communication interfaces, alarms and monitoring functions the project requires.

For critical transformers, consider defining requirements for DGA, winding and oil temperature, moisture, cooling equipment, OLTC status, alarms and communication with the plant SCADA or asset-management platform.

Other important questions include data storage, communication protocol, cybersecurity, remote access, alarm management and integration with existing monitoring systems.

The supplier should also define sensor maintenance and calibration requirements.

Monitoring equipment itself must remain reliable throughout the transformer lifecycle.

Condition Monitoring Supports Better Lifecycle Decisions

Transformer condition monitoring is not only about avoiding catastrophic failure.

It can also improve lifecycle decisions.

Condition data can help asset owners determine whether to continue operation, increase inspection frequency, schedule maintenance during a planned outage, perform additional diagnostic testing or begin planning transformer replacement.

This becomes particularly valuable for ageing assets.

A transformer approaching its expected design life does not necessarily need immediate replacement if condition data remains favourable. Conversely, a younger transformer experiencing rapidly deteriorating indicators may deserve more attention.

Condition monitoring therefore helps shift maintenance planning from assumptions about age toward evidence about actual equipment condition.

It can also support better electrical equipment procurement because operational experience from existing transformers can inform monitoring requirements for future assets.

Building a More Reliable Transformer Maintenance Strategy

The strongest transformer maintenance programs combine several layers of information.

Routine inspection remains important. Laboratory oil testing remains important. Electrical diagnostic testing remains important.

Online monitoring adds continuous or high-frequency information that these traditional methods cannot always provide.

DGA can identify developing thermal and electrical faults. Temperature monitoring reveals thermal stress. Moisture indicates insulation risk. Partial discharge monitoring can identify emerging dielectric problems, while OLTC monitoring provides insight into one of the transformer’s most mechanically active systems.

When these data sources are combined with transformer operating history and engineering assessment, maintenance teams can make better-informed decisions.

The result is not maintenance-free transformers.

The result is maintenance targeted at the right equipment, at the right time, for the right reason.

Conclusion

Transformer condition monitoring provides asset owners with visibility into conditions that may otherwise remain hidden until a transformer develops a serious problem.

Dissolved gas analysis, temperature, moisture, partial discharge and OLTC monitoring each provide a different view of transformer health.

Their greatest value comes when the information is trended and analysed together.

For utilities, renewable energy projects, industrial facilities and other critical power systems, condition-based monitoring can support earlier fault detection, better maintenance planning and stronger lifecycle asset management.

Rather than waiting for a transformer to fail before identifying a problem, monitoring allows operators to recognise deterioration while there is still time to investigate and respond.

Improve Visibility Into Your Power Transformer Assets

Leistung Energie provides power transformers and electrical equipment solutions for utility, renewable energy, industrial, mining and infrastructure applications.

If your project requires transformer monitoring, protection or condition-monitoring provisions, these requirements should be defined early during equipment specification.

Contact Leistung Energie Australia to discuss your transformer requirements, monitoring strategy and upcoming power infrastructure project.