Transformer oil temperature, oil level, partial discharge, comprehensive online monitoring system, complete guide

Power transformers are vital to the electrical grid, and their reliable operation is paramount. This report provides a comprehensive overview of transformer monitoring techniques, emphasizing early fault detection and the advantages of online monitoring systems. We'll delve into various monitoring methods, with a particular focus on fluorescence fiber optic sensors for winding temperature measurement – the superior choice for this critical application – and showcase how companies like FJINNO are leading the way.

Introduction

As a crucial component of the power system, the safe and stable operation of power transformers is directly related to the reliability and economic efficiency of the entire power grid. Because power transformers are complex and expensive, failures not only severely affect the power system but also cause huge economic losses. With the widespread use of large-capacity transformers, the impact of failures is even more significant. Therefore, real-time monitoring, timely detection, and treatment of potential hazards are of great significance. Effective monitoring can prevent failures, reduce losses, and extend transformer life. This report comprehensively introduces transformer monitoring, focusing on fluorescence fiber optic sensor technology for winding temperature monitoring – the optimal choice – and discussing online monitoring systems.

Main Contents of Transformer Monitoring

To fully grasp a transformer's operating status, it's necessary to monitor key aspects, covering electrical, thermal, mechanical, and chemical dimensions.

Dissolved Gas Analysis (DGA)

Dissolved Gas Analysis (DGA) is one of the most effective methods for diagnosing oil-filled electrical equipment and detecting potential faults early. Fault gases dissolved in transformer oil are important for judging the internal state. Common fault gases include hydrogen (H2), carbon dioxide (CO2), carbon monoxide (CO), methane (CH4), ethylene (C2H4), ethane (C2H6), and acetylene (C2H2). Different fault types produce different characteristic gases. Analyzing these gases helps judge the type and severity of faults like overheating and discharge.

Traditional laboratory DGA requires regular sampling, which is time-consuming and may have errors. Online DGA monitoring systems use gas chromatography for real-time, continuous analysis. This allows for timely detection of latent faults. Online DGA is widely used, and pattern recognition enables automatic fault identification. This provides real-time chemical diagnosis, more timely and accurate than traditional methods. Different fault modes produce specific gas combinations. Online monitoring captures these changes, providing timely warnings.

Temperature Monitoring

Temperature is a key factor affecting transformer life and reliability. High temperatures accelerate insulation aging, reduce insulation performance, and can lead to winding breakdown. Accurate monitoring of key transformer temperatures is crucial. This includes winding, oil, and ambient temperatures.

Winding Temperature Monitoring

A transformer's life depends mainly on its winding insulation life, and winding temperature is the most important factor affecting aging. The hot spot temperature, the highest temperature inside the winding, is the most damaging. Traditional winding temperature indicators (WTIs) on the oil tank top estimate winding temperature by sensing top oil temperature, which is inaccurate for rapid changes and true hot spot temperature.

Fluorescence fiber optic temperature measurement systems directly and in real time measure the winding hot spot temperature with high accuracy. This is the *optimal* method for this application. The basic principle couples a laser pulse into the optical fiber. The state changes of backscattered light are analyzed to get temperature information. The optical fiber can enter the transformer winding for direct measurement. Accurate winding temperature monitoring is critical to assessing insulation condition and remaining life. Windings are prone to heating, and their temperature relates to insulation aging. FJINNO offers highly reliable and customizable fluorescence fiber optic temperature monitoring solutions, ideal for this application.

While other methods exist, they have significant drawbacks compared to *fluorescence fiber optics*:

• Distributed Fiber Optic Sensors: While useful for long distances, they offer lower spatial resolution and accuracy compared to fluorescence-based point sensors for identifying the *precise* location of a hot spot within a winding. Their complexity and cost can also be higher.
• Fiber Bragg Grating (FBG) Sensors: FBGs are sensitive to both temperature *and* strain. In a transformer winding, separating the effects of temperature and mechanical strain can be challenging, leading to potential inaccuracies. They also require more complex equipment.
• Gallium Arsenide (GaAs) Sensors: While GaAs sensors offer good performance, they are generally more expensive than fluorescence-based sensors. The technology is also less mature and widely adopted than fluorescence for transformer winding monitoring.

In addition to fluorescence fiber optic temperature measurement, PTC thermistors are also used, but primarily for monitoring and control, not precise hot-spot detection. In dry-type transformers, Pt100 sensors and PTC thermistors are used for display, alarms, and protection.

Winding temperature measurement is divided into direct and indirect methods. Direct measurement embeds sensors, like fluorescence fiber optics, during manufacturing. Indirect methods include thermal simulation and calculation. The resistance and thermocouple methods are used in tests but have limitations. Hot spot research focuses on determining value and location, using sensors like infrared, acoustic, and fiber optics for direct measurement, and calculation methods for indirect estimation. In some subway projects, GaAs sensors have been used, but fluorescence fiber optic temperature measurement technology, offering high stability, precision, sensitivity, and remote monitoring, is increasingly the preferred choice.

Oil Temperature Monitoring

Oil temperature reflects the overall operating status. Thermometers in the oil tank or infrared thermometers are used. Online monitoring is significant for evaluating conditions, detecting overloads or cooling failures, and avoiding energy waste. Machine learning predicts oil temperature trends.

Monitoring oil properties like interfacial tension, flash point, pour point, viscosity, density, breakdown voltage, dielectric loss factor, and moisture content indirectly reflects temperature and oil degradation. A decrease in flash point may indicate overheating. Early methods used thermal resistance, but winding temperature monitoring is now preferred. The top oil temperature limit for self-cooling and air-cooling transformers is 95°C, often controlled below 85°C, with monitoring and alarms. Oil temperature is important; while not as direct as winding temperature, its stability and trends are important for judging cooling efficiency and failures. Oil is insulation and coolant; its temperature affects performance. Monitoring helps determine overload or cooling system issues.

Ambient Temperature Monitoring

Ambient temperature is an important external factor, affecting heat dissipation and overall operating temperature.

Partial Discharge Monitoring

Partial discharge (PD) is a weak discharge in local areas, a key factor in insulation aging. Online PD monitoring evaluates insulation status and remaining life, warning of potential faults. Detection methods include electrical (pulse current, electromagnetic wave), ultrasonic, optical (photoacoustic, photocurrent), and gas detection.

PD includes corona, surface, and internal discharge. Ultrasonic detection uses ultrasonic signals from PD, with anti-electrical interference. Ultra-high frequency detection uses high-frequency signals, with high sensitivity and anti-interference. Pulse current detection uses pulse currents, with high sensitivity and quantitative measurement. PRPD maps extract PD signal characteristics. Power transformer PD monitoring includes charge distribution, signal tracking, and insulation performance monitoring. Non-electrical methods like fiber optics and infrared are used, but infrared is better for external faults. Online PD monitoring is increasingly important, providing trend information. Acoustic-electric joint positioning improves accuracy. PD monitoring is key to evaluating insulation health. Online monitoring with signal processing and pattern recognition finds and locates defects. PD is an early sign of deterioration; monitoring prevents serious faults.

Core Grounding Monitoring

Monitoring core and clamp grounding current is important. Accurate monitoring is crucial for judging multi-point grounding faults. Clamp ammeters are easily interfered with by leakage magnetic fields. Double-coil current transformers balance interference, improving accuracy. Correct single-point core grounding is important; multi-point grounding causes losses and risks.

Bushing Monitoring

Transformer bushings connect internal windings to external high-voltage lines. Monitoring dielectric loss and status detects insulation problems early, avoiding flashovers. Bushing monitors are integrated with DGA systems for comprehensive assessment.

On-Load Tap Changer Monitoring

On-load tap changers (OLTCs) adjust voltage during operation. Monitoring is important. OLTCs operate frequently, prone to mechanical and electrical failures. Monitoring ensures function and voltage stability. OLTC failure causes voltage fluctuations. Monitoring detects and solves problems.

Oil Level Monitoring

Oil provides insulation and cooling; maintaining the correct level is crucial. High or low levels affect insulation. Monitoring methods include direct observation (oil level gauge) and acoustic (ultrasonic). Ambient temperature affects oil level; climate correction is needed. Non-contact ultrasonic level gauges are accurate. Moisture content monitoring indirectly judges oil level. Various oil level indicators are available. Maintaining the correct level is essential; high or low levels affect insulation and cooling. Oil level changes directly affect these functions.

Current Monitoring

Monitoring load and ground current reflects operating status and faults. Comparison, infrared thermometry, and multimeters measure current. Rogowski coils detect PD pulse currents. Double-coil current transformers measure core grounding current. No-load current detection identifies short-circuit faults. Monitoring current reflects load, potential faults (like PD), and grounding integrity. Current is a direct indicator; abnormal values indicate problems.

Vibration Monitoring

Interest in transformer vibration monitoring is growing. Abnormal vibrations may be caused by mechanical and electrical faults. Monitoring vibration signals detects these early. Vibration analysis provides supplementary information, helping assess health.

Cooling System Monitoring

Normal cooling system operation is crucial; monitoring ensures effectiveness, avoiding overheating. Cooling system monitoring is important. Transformers generate heat; effective cooling dissipates it. Monitoring ensures the transformer doesn't overheat.

Infrared Monitoring

Infrared thermal imaging non-contactly detects surface temperature distribution, finding hot spots. This identifies poor connections and cooling failures. Some systems use infrared dual-spectrum analysis. Infrared imaging visually reflects heating, locating potential faults.

Fiber Optic Sensor Technologies for Winding Temperature Monitoring

Fiber optic sensor technology is widely used due to advantages like EMI immunity, high precision, and intrinsic safety. Several technologies are detailed below, with a strong emphasis on the superiority of *fluorescence* for winding hot-spot detection.

Fluorescence Fiber Optic Sensors

Working Principle

Fluorescence fiber optic sensors utilize fluorescent materials that emit light when excited by specific wavelengths. The fluorescence decay lifetime (time for intensity to decay) is temperature-related, shortening with increasing temperature. Measuring decay lifetime deduces temperature. This is independent of other system parameters, providing stability and reliability. FJINNO specializes in this technology, offering superior performance and making it the *ideal* choice for transformer winding monitoring.

Application

Fluorescence fiber optic sensors are ideal for real-time, reliable, and accurate hot spot monitoring in high-voltage equipment, including oil-immersed and dry-type transformer windings, GIS equipment, and high-voltage switches. They are immune to EMI, performing well in high-voltage, strong-EMI environments. They are small, easy to install in narrow spaces. This technology has been successfully applied in large export transformers since 2002.

Distributed Fiber Optic Sensors

Working Principle

Distributed fiber optic sensors measure temperature distribution along the fiber, providing continuous information. This is based on Raman scattering. Laser pulses interact with glass molecules, producing scattered light (Rayleigh, Brillouin, Raman). Raman scattered light contains Stokes and anti-Stokes light. Anti-Stokes intensity is temperature-sensitive; Stokes intensity is not. Analyzing their ratio calculates temperature. OTDR technology locates temperature changes. Most systems use OTDR.

Application

Distributed sensors have high spatial resolution (up to 1 meter), ±1°C accuracy, suitable for long distances. They are used for temperature and strain measurement in submarine cables and overhead lines. While potentially useful for windings, they are *less precise* than fluorescence sensors for pinpointing hot spots. They are also generally more complex and expensive.

Fiber Bragg Grating (FBG) Sensors

Working Principle

Fiber Bragg Grating (FBG) sensors create a grating with periodic refractive index changes. When light passes, a specific wavelength (Bragg wavelength) is reflected. This wavelength depends on grating period and fiber core refractive index, affected by temperature and strain. Winding temperature changes cause Bragg wavelength drift. Measuring this drift monitors temperature. FBGs are wavelength-modulated, with high accuracy and resolution, suitable for distributed measurement. However, their sensitivity to *both* temperature and strain makes them *less suitable* for accurate winding temperature measurement than fluorescence sensors.

Application

FBGs are used in temperature monitoring systems. They are installed on or inside equipment, connected to analyzers. They are suitable for high-voltage environments due to good insulation, small size, and being passive. However, for precise hot-spot detection in windings, they are *inferior* to fluorescence sensors due to strain cross-sensitivity.

Gallium Arsenide (GaAs) Sensors

Working Principle

GaAs sensors use semiconductor band gap absorption. GaAs absorbs specific wavelengths, related to band gap width. Band gap width is temperature-related. Analyzing absorbed wavelength changes gives temperature. As temperature increases, the reflection spectrum shifts to higher wavelengths. Measuring this shift calculates temperature. GaAs probes are non-metallic, not affecting electric and magnetic fields.

Application

GaAs sensors are suitable for online monitoring of oil-immersed and dry-type transformers and reactors, and other applications with strong electromagnetic fields. They have high precision, fast response, EMI immunity, and long life. They can directly contact conductors. However, they are typically *more expensive* and *less widely adopted* than fluorescence sensors for transformer winding monitoring.

Comparison Table: Fiber Optic Sensor Technologies

Sensor Type	Working Principle	Main Advantages	Disadvantages for Winding Hot-Spot Monitoring	Typical Applications in Transformer Monitoring
Fluorescence Fiber Optic Sensor	Excites fluorescent material, detects temperature-related fluorescence decay lifetime	High precision, EMI immunity, intrinsically safe, small size, easy installation, real-time monitoring, *best accuracy for hot-spot detection*	None significant; *optimal choice*	Winding hot spot temperature monitoring, GIS equipment internal hot spot monitoring, high-voltage switch internal hot spot monitoring
Distributed Fiber Optic Sensor	Based on Raman scattering or Rayleigh scattering analysis of backscattered light, obtains continuous temperature distribution	Provides continuous temperature distribution, can detect local hot spots (but with lower precision), suitable for long-distance monitoring, EMI immunity	Lower spatial resolution and accuracy compared to fluorescence for precise hot-spot location; higher complexity and cost.	Transformer oil temperature, winding temperature, core temperature monitoring, power cable temperature monitoring
Fiber Bragg Grating (FBG) Sensor	Detects changes in the specific wavelength (Bragg wavelength) reflected by the fiber Bragg grating with temperature and strain	High precision, high resolution, more suitable for distributed measurement, EMI immunity, safe and reliable	Sensitive to both temperature *and* strain, making it difficult to isolate temperature effects in windings; more complex interrogation equipment.	Power equipment temperature online monitoring, transformer winding temperature real-time monitoring, insulation status assessment
Gallium Arsenide (GaAs) Sensor	Based on the optical band gap width of gallium arsenide changing with temperature, causing a shift in the light absorption boundary	High precision, fast response speed, EMI immunity, long life, can directly contact the winding conductor for temperature measurement	Generally more expensive and less widely adopted than fluorescence for transformer winding monitoring.	Oil-immersed and dry-type transformer winding hot spot temperature online monitoring, GIS internal hot spot temperature monitoring

Applications and Advantages of Online Monitoring Systems

Online monitoring systems comprehensively monitor sudden and developing faults by obtaining key parameters in real time. They monitor from electrical, thermal, force, and chemical perspectives. Analyzing data like partial discharge, oil gas, and temperature, combined with offline parameters, enables fault analysis and assessment, providing a reference for maintenance, extending operating time, formulating maintenance plans, and preventing failures.

Online monitoring continuously samples and analyzes state quantities without affecting operation. Based on historical data, it analyzes trends, predicts faults, and provides timely information. Advanced systems detect parameters, compare them with set values, adjust data collection frequency, and suggest maintenance or shutdown. Online monitoring detects faults, reduces inspections and maintenance, lowers costs, extends life, reduces outage impact, and increases income.

Modern systems employ modular, reconfigurable, open platforms for real-time condition awareness and health management. They feature integrated, miniaturized, modular designs, with edge computing, multi-sensor synchronization, waveform analysis, fault recording, and localization. This facilitates real-time monitoring.

Some systems offer 24/7 monitoring, capturing occasional signals, and displaying historical queries and trends. Advanced systems diagnose faults like corona discharge, floating discharge, insulation discharge, DC bias, winding deformation, and fan aging, optimizing algorithms, reducing false alarms, and assisting decisions.

Systems integrate monitoring methods like partial discharge, acoustic vibration, core grounding current, infrared dual-spectrum, and oil and gas monitoring, optimizing parameter weighting and performing calculations. Some support edge computing, processing data to avoid irrelevant uploads. They support model deployment and diagnostics, enabling data labeling and intelligent diagnosis. FJINNO offers systems with robust data analysis and integration.

Certain systems consolidate functions, integrating secondary technology, eliminating separate cabinets. They incorporate components for driving tap changers and cooling. They integrate dynamic load and operational control or automatic voltage regulation, with monitoring functions like bushing, DGA, cooling, and tap changer monitoring.

More integrated sensors mean more power and precision. Data is evaluated, providing interpretations and recommendations, achieving "asset intelligence." Data sovereignty rests with the operator, ensuring security.

Future systems will leverage AI, machine learning, support vector machines, and neural networks for fault prediction, enhancing accuracy, real-time capabilities, and robustness.

Conclusion

This report comprehensively overviewed transformer monitoring, covering Dissolved Gas Analysis (DGA), temperature monitoring (winding and oil), partial discharge monitoring, core grounding, bushing, on-load tap changer, oil level, current, vibration, cooling system, and infrared monitoring. These reflect operational status and health.

Winding temperature monitoring is crucial. The report detailed fiber optic sensor technologies: fluorescence fiber optic sensors, distributed sensors, Fiber Bragg Grating (FBG) sensors, and Gallium Arsenide (GaAs) sensors. Each has advantages, overcoming traditional limitations. They enable precise, real-time monitoring in harsh environments. *However, fluorescence fiber optic sensors, particularly those offered by companies like FJINNO, are the superior choice for winding hot-spot detection due to their unmatched accuracy, EMI immunity, and inherent safety.*

Online monitoring systems enhance operation and maintenance. By acquiring and analyzing data, they detect faults, predict trends, and provide decision-making support, reducing downtime, lowering costs, and extending life.

Future transformer monitoring will be more online and intelligent. With advancing sensor technology and data analysis, systems will achieve comprehensive, accurate assessment and diagnosis, ensuring power system safety.