A high-voltage (HV) or medium-voltage (MV) transformer almost never fails without warning signs: overheating, moisture in the oil, dissolved gases, phase imbalance, harmonics, tap changer drift, ventilation failure, or a Buchholz alarm. The Eziwan monitoring system centralizes these measurements in real time to help operators detect faults before a shutdown, document trends, and ensure the safe maintenance of industrial substations.
Why Monitor a High-Voltage/Extra-High-Voltage Transformer Continuously?
Power transformers are critical assets: they serve as the interface between the upstream grid and industrial, commercial, or energy facilities. Their unavailability can result in lost production, a site shutdown, a protection trip, a fire hazard, or a major repair on equipment that is difficult to replace quickly.
In many facilities, however, monitoring is still limited to three levels:
- electrical protection devices that trip only when a fault has already become severe;
- on-site inspections involving the local reading of thermometers, relays, and meters;
- periodic oil analyses performed several times a year.
These approaches remain useful, but they are not sufficient to detect all deviations between inspections. Slow overheating, an increase in humidity, harmonic drift, or wear on the load regulator can develop over weeks before becoming apparent during an inspection.
Continuous monitoring enables a shift from a reactive approach to condition-based maintenance: maintenance is triggered by the actual condition of the asset, not just by a schedule or a failure.
Transformer Faults to Be Detected Early
A transformer can experience several types of faults. Some are sudden and must be handled by protective relays. Others develop gradually and are well suited for monitoring.
| Defect Category | Weak Signals | Main Risk |
|---|---|---|
| Oil or winding overheating | high temperature, heavy load, insufficient ventilation | accelerated aging of insulation |
| Oil degradation | moisture, acidity, dissolved gases | loss of dielectric strength |
| Internal fault | gases, Buchholz trip, abnormal pressure | internal arc, dielectric breakdown |
| Load imbalance | unbalanced currents, loaded neutral | overheating and additional losses |
| Harmonics | High THD, dominant harmonics | losses, overheating, aging |
| OLTC fault | Abnormal operations, switching times, contact current | incorrect voltage setting, mechanical fault |
| Auxiliary fault | ventilation, oil pump, control power supply | loss of cooling or monitoring |
| Adverse environmental conditions | ambient temperature, humidity, intrusion | aging, safety, unavailability |
The goal is not to replace electrical protection systems. They remain essential. The objective is to supplement these protection systems with a monitoring layer that identifies trends before they reach a critical threshold.
Monitoring Architecture for HTA/HTB Stations
Eziwan acts as a data collection and transmission gateway. The gateway retrieves measurements from sensors, relays, network analyzers, substation controllers, and third-party equipment, then transmits them to a centralized dashboard via a secure tunnel.
This architecture allows you to monitor a workstation locally or remotely without exposing the equipment to the Internet. Encrypted outbound traffic is easier to secure than direct inbound access to the workstation’s relays or controllers.
Priority measurements to collect
The value of transformer monitoring depends on the choice of data points. It is better to collect less data—but high-quality data—than to report hundreds of values without context.
| Measure | Possible source | Use |
|---|---|---|
| High oil temperature | built-in sensor, PT100, converter | thermal monitoring of the transformer |
| Estimated winding temperature | thermal indicator, relay, calculation | assessment of insulation aging |
| Load per phase | power analyzer, current transformers | overload and unbalance detection |
| Voltage per phase | power analyzer | power quality monitoring |
| Voltage and Current THD | Power quality analyzer | Harmonic detection |
| Buchholz gas | Buchholz relay | Alert for possible internal fault |
| Oil level | Contact or analog sensor | Leak, expansion, tank anomaly |
| Oil moisture | In-line sensor or periodic analysis | Dielectric risk |
| DGA | Dissolved gas sensor | Detection of thermal or electrical faults |
| OLTC position | On-load tap changer | Voltage regulation monitoring |
| Number of OLTC operations | internal counter | usage-based maintenance |
| Forced ventilation | contact, auxiliary controller | cooling verification |
| Substation room temperature | ambient sensor | environmental conditions |
tip: Good practice Associate each point with a unit, a frequency, a criticality level, a threshold, a deviation rule, and an expected action. A useful alarm tells you what to do, not just that a value is in the red.
Oil and Winding Temperature
Temperature is one of the most important indicators for a transformer. A sustained rise in temperature accelerates the aging of the oil-impregnated paper insulation and reduces the safety margin. Monitoring must distinguish between oil temperature, ambient temperature, load, and estimated winding temperature.
Typical cases to look for are:
- High oil temperature despite a normal load;
- Overheating due to inadequate ventilation;
- Estimated winding temperature deviation;
- Abnormal phase-to-phase voltage difference;
- Local temperature at the substation too high;
- Thermal cycles occurring more frequently than usual;
- Load restart after a trip, followed by a rapid temperature rise.
Temperature alone is not enough. It must be considered in relation to the load, the cooling method, the outdoor temperature, and substation events. A high temperature during a short load peak does not have the same significance as a persistently high temperature under average load conditions.
Buchholz, Pressure, and Internal Defects
The Buchholz relay is a key component for oil-immersed transformers equipped with a conservator. In particular, it detects gas buildup or abnormal oil movement that may indicate an internal fault.
Supervision must be reported at least:
- gas alarm;
- activation;
- contact status;
- event timestamp;
- correlation with load and temperature;
- communication status;
- acknowledgment and operator comment.
A Buchholz event should never be treated as a mere computer alarm. It requires a clear operating procedure tailored to the type of transformer, the load conditions, and the manufacturer’s instructions.
DGA: Dissolved Gas Analysis
Dissolved gas analysis, often referred to as DGA for Dissolved Gas Analysis, helps diagnose certain internal faults in oil-filled transformers. The gases detected can provide clues about partial discharge, overheating, arcing, or insulation degradation.
Online DGA sensors do not always replace a comprehensive laboratory analysis, but they do provide a continuous trend. This trend is valuable when a gas concentration rises rapidly or when the profile of a gas mixture changes.
| Gas or indicator | Possible interpretation | Analysis considerations |
|---|---|---|
| Hydrogen | Possible partial discharge or electrical fault | Correlate with other gases |
| Methane | Possible low-temperature heating | Depends on oil context |
| Ethane | Possible thermal fault | Trend is more significant than isolated value |
| Ethylene | Possible more pronounced overheating | Monitor rate of change |
| Acetylene | Possible electric arc | Critical signal—confirm promptly |
| Carbon monoxide | Possible cellulose degradation | Related to paper aging |
| Moisture | Possible decrease in dielectric strength | Depends on temperature and history |
The interpretation of DGA results must be guided by recognized methods, such as IEC 60599 or the diagnostic practices used by transformer specialists. Eziwan can centralize and log the data, but technical decisions must take into account the transformer, its oil, its age, previous test results, and the manufacturer’s recommendations.
OLTC Support Switcher
The on-load tap changer, or OLTC, is one of the most heavily used mechanical components in a power transformer. It adjusts the turns ratio to keep the voltage within an acceptable range, but each operation causes mechanical and electrical wear.
The key points to monitor are:
- current position;
- total number of operations;
- number of operations per period;
- operation time;
- motor current;
- switching fault;
- deviation between setpoint and actual position;
- voltage regulator alarms;
- abnormal frequency of tap changes.
An unusual increase in the number of tap changes may indicate a control problem, unstable upstream voltage, or overly aggressive settings. OLTC maintenance should therefore be based on actual usage, not just on a scheduled timeline.
Power Quality, Harmonics, and Unbalance
Transformers are affected by power quality. Nonlinear loads, variable-speed drives, rectifiers, inverters, charging stations, and electronic equipment can generate harmonics. These disturbances increase losses, cause the windings to overheat, and can accelerate aging.
The key measures are:
- phase-to-phase and phase-to-neutral voltage;
- current per phase;
- active, reactive, and apparent power;
- power factor;
- voltage and current imbalance;
- voltage THD;
- current THD;
- harmonics by order;
- sags and surges;
- frequency;
- outages or micro-outages.
The EN 50160 standard describes the characteristics of the voltage supplied by public power grids under normal conditions. For an internal industrial power system, it can serve as a reference, but the operating thresholds must be adapted to the connection agreement, the requirements of the utility or transmission company, the site’s equipment, and power quality studies.
High-Voltage and Medium-Voltage Substations: Compliance, Traceability, and Audits
Monitoring alone does not ensure regulatory compliance. However, it does provide the historical data and logs needed to document operations.
For high-voltage (HV) delivery substations, the NF C 13-100 covers substations supplied by a public HV distribution network up to 33 kV. Internal high-voltage installations may also be subject to other standards, depending on their voltage, intended use, and design.
In environments connected to Enedis or RTE, the applicable requirements depend on the connection point, the contract, the voltage level, the protection scheme, and the operating rules. The Eziwan monitoring system helps provide actionable traceability, but the exact reference framework must be validated against the contractual documents and the requirements of the relevant grid operator.
The following are useful elements for an audit:
- temperature history;
- alarm logs;
- Buchholz events;
- DGA trends;
- protection and auxiliary status;
- power quality;
- maintenance reports;
- interventions and acknowledgments;
- PDF or CSV exports;
- monitoring availability;
- threshold and configuration changes.
Traceability Info A monitoring report is most useful when it is time-stamped, consistent, legible, and linked to maintenance actions. The evidence is not limited to a graph; it must document the condition of the asset and the decisions made.
Predictive Alerts and Drift Rules
A transformer alert should not be based solely on an absolute threshold. Slow deviations are often more useful than sudden overshoots, as they allow time to plan.
Examples of useful rules:
- Oil temperature exceeding a threshold for a defined period of time;
- Abnormal temperature rise under steady load;
- Significant current imbalance between phases;
- Current THD rising over several days;
- Number of OLTC operations exceeding the usual pattern;
- Rapid increase in a DGA gas;
- Rising oil moisture content;
- Loss of communication with a relay;
- Ventilation failure under high load;
- Substation room temperature exceeding the authorized threshold.
| Rule | Type | Recommended Action |
|---|---|---|
| Persistently high oil temperature | time-delayed threshold | check load and cooling |
| Rapid rise in DGA gas | drift | initiate transformer diagnostics |
| Buchholz alarm | critical event | apply operating procedure |
| OLTC operating too frequently | behavioral anomaly | check voltage regulation |
| Persistent current imbalance | grid quality | analyze loads by phase |
| Ventilation unavailable | auxiliary | perform maintenance |
| High THD | power quality | harmonic analysis or filtering |
The threshold must be tailored to the transformer, its cooling method, its age, its load, and the manufacturer’s recommendations. Copying a generic threshold from one site to another can cause false alarms or mask a real risk.
Predictive Maintenance: How Data Is Really Making a Difference
Predictive maintenance is not a magic solution. It becomes useful when the measurements are reliable, comparable, and linked to concrete actions.
In particular, it allows you to:
- prioritize which transformers to inspect;
- schedule an oil analysis before a failure occurs;
- anticipate the need for an OLTC adjustment;
- detect recurring overloads;
- justify increased ventilation;
- verify the effect of harmonic correction;
- document accelerated aging;
- reduce unnecessary rounds;
- plan outages with greater precision.
The most robust approach combines online measurements, field inspections, laboratory oil analyses, manufacturer recommendations, and the experience of electrical teams.
Example of a pinout diagram for a high-voltage transformer
| Item | Source | Frequency | Threshold or rule | Usage |
|---|---|---|---|---|
| High oil temperature | PT100 | 1 min | threshold and drift | thermal monitoring |
| Local ambient temperature | sensor | 5 min | high threshold | cooling context |
| Phase current L1 L2 L3 | network analyzer | 1 min | imbalance | transformer load |
| Current THD | network analyzer | 5 min | quality threshold | harmonics |
| Buchholz alarm | dry contact | event | immediate | internal fault |
| Oil level | sensor or contact | 5 min | low threshold | leak or anomaly |
| OLTC position | regulator | event | inconsistency | voltage regulation |
| OLTC operation counter | regulator | 1 h | maintenance threshold | actual wear |
| Ventilation fault | auxiliary contact | event | immediate | cooling |
| Oil moisture | DGA or third-party sensor | 1 h | drift | dielectric strength |
This plan can then be expanded to include the criticality of the unit, the transformer’s power rating, the type of oil, the presence of a preservative, the cooling method, and the maintenance strategy.
CMMS, SCADA, and Central Monitoring Integration
Eziwan can function as a dedicated dashboard for transformers, but it can also feed data into a SCADA system, a CMMS, or an energy monitoring tool.
Common integrations include:
- Automatic creation of CMMS tickets;
- CSV export for oil analysis and reporting;
- Webhook to on-call system;
- API to data platform;
- Alarm forwarding to SCADA;
- Monthly PDF export for operational review;
- Access log for OT cybersecurity.
| Destination | Data Sent | Purpose |
|---|---|---|
| CMMS | alarms, faults, OLTC counters | maintenance planning |
| SCADA | critical statuses and measurements | real-time operations |
| Data platform | time series | advanced analysis |
| On-call | qualified alerts | rapid response |
| PDF report | monthly summary | audit and assurance |
Integration should minimize noise. Not all measurements warrant a SCADA alarm. Some should remain in the history log and only be flagged in the event of an abnormal trend.
OT Security for the Substation
A high-voltage/extra-high-voltage substation is a sensitive environment. Monitoring connectivity must be secured from the design phase onward.
Best Practices:
- No public ports exposed to relays, PLCs, or gateways;
- Encrypted outbound VPN tunnel;
- Traffic filtering by IP, port, and protocol;
- Separation between the monitoring network and the protection network;
- Remote access limited to authorized personnel;
- Strong authentication for maintenance;
- Connection logging;
- Rapid revocation of vendor access;
- Configuration backups;
- Documentation of routes and equipment;
- Monitoring of tunnel status.
To learn more about secure remote access, see the guide on OT network security without public ports and industrial connectivity solutions.
Typical Deployment with Eziwan
A transformative monitoring project must be structured to avoid blind spots.
1. Job Description
Identify the transformers, protective devices, Buchholz relays, existing sensors, network analyzers, auxiliary equipment, OLTC regulators, PLCs, and available communication systems.
2. Criticality Analysis
Classify assets based on their impact: production loss, safety, redundancy, replacement time, power, age, failure history, and operational constraints.
3. Mapping Points
Define the essential metrics, thresholds, deviation rules, frequencies, alert recipients, and expected actions.
4. Installing the gateway
The Eziwan gateway is connected to the station’s sensors, relays, analyzers, or PLCs. Depending on the architecture, it can use Ethernet, Modbus TCP, Modbus RTU, digital inputs, analog inputs, or third-party gateways.
5. Securing Communication
The outbound VPN tunnel is configured, access is filtered, user permissions are restricted, and logging is enabled.
6. Configuring Dashboards
The views are organized by substation, transformer, criticality, and measurement category: thermal, oil, DGA, OLTC, grid quality, auxiliaries, and alarms.
7. Field Tests
Every critical measurement is verified: unit, scale, contact direction, timestamp, alarm, acknowledgment, CMMS ticket, and behavior in the event of a loss of communication.
8. Periodical Review
Thresholds and deviation rules are adjusted based on observations of actual behavior. A transformer is more than just a spec sheet: its operational history matters.
Common Pitfalls
Several mistakes can significantly reduce the value of a transformative supervision project.
- confusing electrical protection with predictive monitoring;
- monitoring temperature without tracking the load;
- creating too many unprioritized alarms;
- ignoring the tap changer;
- overlooking auxiliary cooling systems;
- interpreting a DGA without historical data or expertise;
- mixing critical measurements with secondary information;
- failing to test Buchholz contacts;
- granting overly broad remote access to the substation;
- failing to document thresholds;
- failing to link alerts to the CMMS;
- comparing very different transformers without context.
Caution: Important Note Monitoring must never inhibit or bypass electrical protective measures. It complements protection, operation, and maintenance, but does not replace the substation’s safety functions.
How Eziwan Fits Into Your High-Voltage/Extra-High-Voltage Systems
Eziwan provides the connectivity, data collection, and monitoring layer between substation equipment and operations teams. The Eziwan gateway transmits field data, the cloud platform centralizes dashboards and historical data, and connectivity solutions secure remote access across multiple sites.
This approach is suitable for manufacturers, infrastructure operators, energy facilities, data centers, pumping stations, private high-voltage networks, and multi-station facilities. It allows for the modernization of monitoring systems without immediately replacing existing relays, sensors, or analyzers.
Conclusion
Monitoring high-voltage (HV) and extra-high-voltage (EHV) transformers and substations makes it possible to detect weak signals before they develop into major faults: overheating, moisture, dissolved gases, Buchholz alarms, unbalance, harmonics, auxiliary faults, or tap changer wear.
With Eziwan, critical metrics are centralized in a secure dashboard, logged, correlated, and transformed into actionable alerts. Electrical teams gain greater visibility, can better prioritize maintenance, and have robust traceability for their audits, insurers, and operational requirements.
Further Reading
- Modbus RTU to Cloud — Upload Modbus measurements from your relays and analyzers to a cloud database
- Energy Monitoring — Manage and optimize your electricity consumption in real time
- Cloud-based industrial data logger — Record and archive field data from your transformers
- Industrial protocols — Modbus, IEC 61850, DNP3, and other protocols used in high-voltage substations
- Cloud SCADA — cloud-hosted SCADA architecture for multi-site monitoring