Only condition-based maintenance, grounded in intelligent online monitoring of transformers, has true practical significance. It not only avoids the blind and mandatory nature of preventative testing and periodic maintenance, saving significant manpower, material resources, and financial resources, but also enhances the power supply reliability of transmission and transformation equipment, extends the service life of transformers, and has broad development prospects.
Transformers accelerate the construction of smart grids and have broad future development prospects.
As the State Grid Corporation of China accelerates the construction of a robust smart grid, the technological advancement of smart transformers, one of the core components of the smart grid, plays a significant role in promoting its development. Online monitoring and real-time feedback/interaction functions of transformers are giving traditional transformers a boost in intelligent development.
Transformer online monitoring involves monitoring the transformer's operating status based on its various mechanical and electrical characteristics, including analysis of dissolved gases in the oil, partial discharge, core grounding current, winding deformation, and vibration. The development of modern sensing technology, microelectronics, and computer technology has laid a solid foundation for the condition monitoring of high-voltage equipment such as transformers. The core difference between intelligent transformers, as crucial equipment in substations, and traditional transformers lies in their intelligent online monitoring capabilities.
1. Advantages of online transformer monitoring
Transformers, operating in the power grid for extended periods, inevitably accumulate various defects. The process by which these defects develop into faults can be broadly categorized into two types: gradual faults and sudden faults. Gradual faults primarily manifest as a gradual decline in insulation performance during normal operation due to factors such as insulation aging and moisture absorption, ultimately leading to a fault. Although the probability of such faults is high, they exhibit certain patterns and can be detected and avoided in a timely manner through online monitoring. Timely and accurate prediction of potential transformer faults is not only a key focus for power grid maintenance personnel but also crucial for ensuring the safe and stable operation of the power grid.
As common methods of transformer maintenance, preventative testing and periodic maintenance tend to be somewhat arbitrary and mandatory. Establishing a predictive maintenance approach—"maintaining when necessary" rather than "maintaining only when scheduled"—has become an inevitable trend in condition-based maintenance. This approach is based on the current condition of the equipment, using intelligent monitoring methods to conduct longitudinal and lateral comparative analyses, identify early signs of faults, and make trend judgments on the location, severity, and development of faults, serving as the fundamental basis for transformer maintenance.
Compared with traditional periodic maintenance, condition-based maintenance of transformers has the following advantages: it enhances the power supply reliability of transmission and transformation equipment; it enables early detection of faults, helping to develop more reliable maintenance plans; it reduces maintenance costs and difficulties, and reduces equipment maintenance risks; it saves a lot of manpower and material resources; and it extends the entire life cycle of transformers.
2. Composition of the online transformer monitoring system
Transformer condition-based maintenance in the power grid mainly consists of two parts: condition monitoring and condition diagnosis. Monitoring methods generally include routine monitoring, online monitoring, and special monitoring. Online transformer monitoring is based on the condition information acquired by various sensors. A computer monitoring system then analyzes and calculates the monitored data to make a scientific assessment and prediction of the transformer's operating status. Finally, a comprehensive intelligent diagnosis is performed by the monitoring and diagnostic system to provide a decision-making plan and determine the optimal maintenance time. The technological advantage of online monitoring lies in its comprehensive judgment and decision-making capabilities resulting from the organic combination of monitoring and diagnostic functions.
Transformer online monitoring systems can be divided into two types: centralized and decentralized. Centralized systems involve periodic or roving monitoring of all tested equipment; decentralized systems use specialized testing instruments to measure signals locally.
Online transformer monitoring, based on microprocessor technology, integrates sensors, data collection, communication systems, and data analysis functions. By continuously monitoring state parameters over a period of time, it promptly captures early warning signs of transformer faults and determines the transformer's operating status based on parameter trends. This not only effectively prevents the occurrence and development of various transformer faults but also reduces economic losses caused by unexpected power outages. Real-time monitoring of operating transformers compensates for the shortcomings of conventional testing and detection methods. Although there are still certain limitations in predicting internal sudden faults based on captured transformer dynamic information, it remains the most effective technical basis for formulating transformer maintenance plans and has significant guiding value.
The online monitoring system utilizes highly reliable, maintenance-free sensors and monitoring instruments with self-diagnostic capabilities. If a problem arises with the monitoring instrument itself, it can automatically issue an audible and visual alarm signal. Therefore, it eliminates the misjudgment problem caused by malfunctions in conventional detection methods.
The investment cost of a transformer online monitoring system mainly includes two parts: sensors and intelligent monitoring software. According to statistics from relevant departments at home and abroad, the cost of a complete transformer online monitoring system is generally about 1% of the transformer price. However, because it can detect early transformer faults in a timely and accurate manner, it can reduce the operating and maintenance costs by 75%, saving an annual cost equivalent to 2% of the transformer price, and can extend the service life by about 10 years, making its economic benefits quite considerable.
3. Transformer Online Monitoring Methods
Transformer online monitoring systems typically diagnose transformer faults not based on the absolute values of measured parameters, but rather on the trends in these parameters. Based on the data collected by sensors, the monitoring system performs complex data analysis and processing to ultimately derive trend-based fault predictions for the transformer. Its basic monitoring procedure is: sensor—data collection—data storage—condition analysis—fault classification—fault determination—solution proposal.
Real-time monitoring data is taken from the transformer itself and its auxiliary components. The database stores comprehensive information about the transformer, including raw parameters, historical data, operating status, monitoring data, knowledge and experience, and diagnostic results.
State analysis is based on artificial neural network analysis. Data processing and fault classification generally employ Fast Fourier Transform or advanced wavelet transform methods. For numerous and complex data such as core, winding, oil temperature, and load, the pre-analysis results are transformed into a form suitable for artificial neural network processing through feature analysis by the artificial neural network processing unit, thereby improving its computational and inference speed.
Fault classification mainly distinguishes the nature of transformer faults, such as winding overheating, core overheating, gas in oil, insulation defects, core grounding current, partial discharge faults, etc.
The monitoring and diagnostic system ultimately uses the analysis results and the knowledge and experience stored in the database as criteria to determine the fault and propose maintenance solutions for the transformer. All information from the monitoring and diagnostic system can be uploaded and retrieved via the network, greatly facilitating power grid operation and maintenance and meeting the requirements of a smart grid.
The scope of online transformer monitoring is extremely broad, encompassing almost all potential fault issues. These include hot spot monitoring, total gas content in the oil, partial discharge, dielectric loss and capacitance of bushings, cooling system function, oil humidity and acidity, load current, humidity and migration of paper insulation, oil temperature at the top and bottom of the windings, defects in the dielectric and power system, clamping force of components, performance and defects of the on-load tap changer system, core grounding faults and winding defects, and oil level in the conservator.
4. Specific applications of online transformer monitoring
A power equipment integrator used FourFaith Communication's industrial wireless router to build a stable and reliable wireless communication network based on a wireless communication operator's network for real-time monitoring of transformers. The design concept is as follows: a transformer characteristic gas detection module detects the concentration of characteristic gases in the transformer's insulating oil and transmits the data to the main control module via a CAN bus. The main control module determines whether the transformer has reached a fault warning level based on its settings, and then transmits the data to the monitoring center via the FourFaith industrial wireless router, thus achieving remote monitoring functionality.
By combining real-time operating systems and wireless network technology and applying them to the remote monitoring of transformers, the concentration of dissolved gases in the transformer's mineral insulating oil can be monitored online. The data can then be transmitted wirelessly to the monitoring center for processing, which can greatly improve the stability and safety of power supply, prevent major accidents, and efficiently solve the problem of unified management and monitoring of power facilities caused by geographical dispersion. At the same time, it can alleviate the management pressure on users and improve management efficiency.
5. Development Trends of Online Transformer Monitoring
Judging from the application of transformer online monitoring technology, although there are still shortcomings such as weak anti-electromagnetic interference capability, short service life, and high price, all these problems will be gradually solved in the course of development due to its unparalleled advantages over conventional detection methods.
On the technological front, hardware will evolve towards greater intelligence and higher maintenance-free operation. The adoption of intelligent sensors will effectively suppress electromagnetic interference; with the widespread use of new materials, the lifespan of equipment such as gas-permeable membranes for chromatographic analysis will also be extended. Fault analysis software systems will be organically integrated with offline testing, equipment status, and operational data to conduct comprehensive online diagnostics, and online monitoring data from different systems will be shared via network.
At the policy level, driven by the demand for smart grid construction and the increasing R&D efforts of various research institutes and online monitoring companies, the application of online monitoring technology is steadily advancing. This will not only prompt changes in transformer maintenance procedures and management methods, but will also inevitably lead to the establishment of a standardized condition-based maintenance management system and the eventual unification of technical standards.
From an economic perspective, with the increased R&D efforts of domestic online monitoring equipment manufacturers and software developers, competition from similar foreign products, and the strong demand from the power grid, not only will the prices of their hardware and software products drop significantly, but their cost-effectiveness will also improve substantially.
6. Conclusion
In conclusion, only condition-based maintenance built upon intelligent online monitoring of transformers has true practical significance. It not only avoids the blind and mandatory nature of preventative testing and periodic maintenance, saving significant manpower, material resources, and financial resources, but also enhances the power supply reliability of transmission and transformation equipment, extends the service life of transformers, and has broad development prospects.
