Fault diagnosis of main equipment in power plant
The typical faults of the main equipment of thermal power plant and the fault diagnosis method used, the problems existing in the existing fault diagnosis system and the fault diagnosis of thermal power plant equipment are described, and the development trend of the fault diagnosis system is pointed out.
The main equipment of large thermal power plants includes boilers, steam turbines, and generators to complete the conversion process from heat to mechanical energy to electrical energy. The coupling between the equipment and the equipment, the complexity of the system, and the special working environment of the equipment at high temperature, high pressure, and high speed rotation determine that the thermal power plant is a production site with high failure rate and great fault hazards. These failures will have significant economic losses and social consequences. Therefore, through advanced technical means, the state parameters of the equipment are tested, monitored, and analyzed to determine whether the equipment is abnormal or faulty, the site and cause of the failure, and the deterioration trend of the failure, in order to determine the reasonable overhaul time.
Typical faults of the main equipment of thermal power plants and their diagnostic methods:
1. Major Faults and Diagnosis Methods of Boilers
(1) Major malfunctions
1 superheater leak. The superheater leakage explosion area is concentrated on the side of the hot superheater under the high temperature superheater. The main reason is that the height of the furnace is low, which causes overheating in the area. There are also problems with insufficient material selection and welding quality.
2 Save coal leak. The main reason is the thinning of the pipe wall caused by fly ash wear, especially in the wall pipe, furnace wall leakage and bending is common.
The water leaked from the cold wall. The main reason is local overheating and corrosion. Local overheating is caused by water cycle failure and internal scaling, while flame deflection or burning area smoke temperature over Gaoze causes high temperature corrosion of water cold walls.
4 Duster failure. The main reasons are that the flue gas flow rate is too fast, the particle size is large, the dust concentration is large, and the exhaust temperature is lower than the dew point temperature.
(2) Diagnostic methods
In boiler fault diagnosis, the physical diagnosis method is: infrared temperature measurement technology, The specific application scope includes the identification and control of boiler flame and combustion state, fatigue damage of thermal equipment, thermal mechanical characteristics of thermal equipment, heat leakage and heat insulation diagnosis and evaluation of thermal system, and thermal pollution control of boilers; Ultrasonic diagnosis method can detect the leakage problem of all parts of the boiler. The principle of ultrasonic detection is that ultrasonic detection instruments can detect the leakage of almost all pressure vessels, vacuum systems, and pipeline valves, and ordinary detection instruments are difficult to do. This point. You can detect the leakage of a target with a diameter of a few microns or less using a three-point scanning module. The leakage diagnosis and location of the equipment can be carried out quickly. High sensitivity, depending on its model, can distinguish smaller, farther, higher distance leaks. The host adjusting knob can block all environmental noise. This is an advantage over other testing products. The detection range is large and the distance is far. According to the different models, different production equipment is detected. Use safe. Since the leak measurement is non-contact, the ultrasonic detection instrument is very safe to use.
Due to its unique performance, it plays a huge role in many aspects such as petrochemicals, metallurgy, electricity, food, automobiles, railways, and military research. Moreover, the structure of the modern ultrasonic detector is gradually becoming miniaturized and intelligent, its performance is constantly improving, and its use is also more flexible and convenient. Therefore, the application range of ultrasonic detection technology is constantly expanding and its application level is constantly improving.

2. Main Faults and Diagnosis Methods of Steam Turbine Units
(1) Main faults of steam turbines
1 is unbalanced. It is mainly caused by machining errors and permanent bending of the rotor, blade shedding, coupling drift, poor manufacturing of long shaft, uneven heating, or thermal bending caused by material defects.
2 is not right. The main reason is that the sliding pin system is jammed, the temperature difference between cylinders is excessive, the unit's lift speed is too fast, the vacuum drop is too large, the foundation is not evenly sinking, and so on, so that the unit expansion is not smooth and the bearing seat is not evenly expanded, and the coupling is eccentric.
3 Rotors touch. The main reasons are the improper design and adjustment of the closure gap, the poor expansion of the cylinder, the excessive thermal deformation of the cylinder, the excessive vibration of the unit and the increase of the axial displacement.
4 Leaf blades fall off. It is mainly due to the lack of strength of the design, the fluctuation of steam parameters, the long-term operation of the unit under large changes in the power grid and low load, and the unit's stay time at a certain speed below the rated speed is too long.
The steam vibrates. The main reason is that due to the low critical speed of the high school pressure rotor, and the large change in deflection under the hot state of the high school pressure rotor, the bearing elevation changes greatly, the quality of processing and installation of the long shaft is poor, and the radial or axial error in the center of the coupling is too large., Shaft balance and the deterioration of the middle state make the high pressure rotor unstable.
(2) Main failure of generators
Oil membrane oscillations. The main reason is that the first-order critical speed of the rotor of the generator is too low, and the stability of the three-axis wedge is poor. When the load is reduced, the instability speed is too low.
The stator coil insulation failure. It is mainly due to wear, aging, pollution, and corrosion that invalidates the insulation, causing partial discharge and temperature increase, and damage to the insulation layer due to tightness, cooling water leakage, fatigue wear and other reasons, resulting in line rod displacement and short-circuit.
The stator coil is overheating. It is mainly due to certain defects in the manufacturing or installation process that short-circuit between turns and cause local overheating.
4 Rotor winding failure. Including ground, interramp short circuit and fault. Ground and interramp short circuit barriers are mainly caused by the reduction of insulation and damage, and joint welding and thermal deformation can cause wire breakage.
The cooling water system is malfunctioning. It is mainly due to cooling water leakage, pipe foreign body blockage, wrong movement, etc., cooling water flow reduction and interruption, and due to material and installation defects caused by stator leakage.
(3) Diagnostic methods
Vibration method is one of the most popular and mature methods used in fault diagnosis of steam turbine units. The application of thermodynamic analysis to diagnose turbine unit performance faults is also an important means, as well as oil liquid analysis, acoustic emission method, non-destructive testing technology. The acoustic emission method is mainly used for the fault detection and leakage detection. Hitachi set up a test piece on the high pressure rotor of the 350MW steam turbine to perform acoustic emission and recording at the bearing tiles at both ends to diagnose the grinding of the rotor. The application of nondestructive testing is very important in the life diagnosis of steam turbine unit. The current non-destructive evaluation methods include hardness measurement, electrical resistance, ultrasound, tissue contrast, crystal grain deformation, microscopic observation and measurement, and X-ray analysis.
Nowadays, dozens of fault diagnosis systems for steam turbine units have been developed at home and abroad. There are mainly expert systems for vibration diagnosis of turbogenerator sets developed by Radial Corporation of the United States, fault diagnosis systems for turbogenerator sets of Westinghouse, rotating mechanical fault diagnosis systems of Bendy Corporation, and falling performance of the stator coil insulation of IRIS Corporation of Canada. Online monitoring system. All kinds of generator sets and large high-voltage motor stator winding stations are on-line testing issues that are of great concern to the power generation industry and large production enterprises. These devices often appear in the tank coil loosening, long-term thermal deterioration, winding pollution caused by tidal gas and oil pollution, load circulation, poor impregnation, improper coating of the end of the groove winding, and insufficient spacing of the end winding coil. From the above problems, it is not difficult to see that the insulation deterioration of the stator winding accounts for 40 % of the motor faults. These failures will cause huge losses to the enterprise.
From the point of view of motor development, the maintenance methods are also constantly changing, from bad repairs(expensive maintenance costs and replacement costs of components
Use), preventive maintenance(regular downtime, regular testing, regular maintenance, regular replacement, some downtime, testing, and maintenance are unnecessary, expensive) to the current predictive maintenance(stop as needed, bad then repair that. According to the failure warning plan when the downtime and downtime, the need for reliable online detection equipment, diagnosis of the possible failure of the precursor. ) The failure of the generator and motor 50 comes from the problem of bearings and vibrations, 40 from the problem of stator winding insulation, and 10 from the problem of rotor winding. According to the above three questions, we use three different methods to detect and monitor. Vibration problem is a mechanical fault We can use on-line vibration monitoring, winding insulation problem We can use on-line release monitoring, rotor winding problem We can use on-line flux monitoring. The online monitoring method can identify the main causes of faults for enterprises, do not affect the normal income of power plants and enterprises, avoid sudden failures, and extend the time for two downtime, which is conducive to the implementation of the "predictive maintenance" plan. Maintenance work can be arranged to be appropriate, and maintenance work is based on the actual condition of the motor itself rather than on a calendar, and online monitoring of the sea can provide potential failure warnings in advance. 24 kV 4-6 years early warning, 13.8 kV 3-4 years early warning,
6.9 KV 1-2 years of early warning, 4 kV-several weeks of early warning. Electrical interference(noise) is often present because the generator is connected to the power system. Therefore, the generator will encounter special difficulties in online annual monitoring under normal operating conditions. The noise comes from the corona, sliding ring or commutator electric spark of the power system, electric spark generated by poor electrical contact, and other electric tool operations. This kind of noise will cause the pulse of the traditional monitoring instrument to be blurred. In fact, the noise is measured, but it will mislead the technicians to set up a very interesting position in the winding. The consequence is that the normal winding is judged to be defective. Thus gives the abnormal false alarm, reduces the credibility of the online monitoring. The IRIS continuous online monitoring system can effectively separate the noise from the local playback sound, and the amount of the coil can be clearly seen through the local discharge table.

3. Major faults and diagnostic methods of Transformers
(1) Major malfunctions
The coil is short-circuited between turns. The main reasons are aging insulation or poor heat dissipation or long-term overloading. Due to short-circuit Electrodynamic damage, inter-ramp insulation, defects in winding materials or processes, dampness, atmospheric or operating overvoltage attacks.
2 The winding is broken. The main reasons are short-circuit electric power to break the coil, poor welding, short-circuit between the turns.
The wound broke through to the ground. The main reasons are the aging of the main insulation, the dampness of the insulating oil, the entry of impurities in the winding, the deformation of the coil when the overvoltage is short-circuited, and the failure of the cooling system, the blockage of the cooling pipeline, and the failure of the protection to produce a whole or partial overheating. Insulation damage.
4 Short circuit between windings. The main reason is similar to ground breakdown. It may also be a short circuit between the leads or the casing, and the oil surface is too low.
5 Local short circuit or local melting of iron chip. The main reason is that the screw insulation of the iron core or choke iron is damaged. Metal chips short-circuit the iron chip at the fault, the insulation between the pieces is seriously damaged, and the grounding method does not correctly constitute the current loop.
The contact of the splitter switch is poor and the surface of the contact melts. The main reason is that there are defects in the structure assembly. After switching the joints, the contact is not reliable, the spring pressure of the moving contact is not enough, and the installation and adjustment of the loading voltage regulator is improper.
The casing flashes or explodes against the ground. The main reason is the surface area gray, dirty, crack, seal is not strict, respirator configuration is improper.
The normal oil temperature of 8 ~ load rises. The main reason is short circuit between winding turns, increased loss, poor contact at large current connection, low oil position, and poor cooling effect.
(2) Diagnostic methods
In Transformer fault diagnosis, the commonly used methods are vibration analysis method, oil gas analysis method, partial discharge method, restoration voltage method, ultrasonic heterodyne method, frequency response analysis method, and infrared diagnosis technology. At present, more applications are mainly infrared diagnostic techniques and ultrasonic aberrations. Here is a general introduction to the ultrasonic heterodyne monitoring method. The ultrasonic heterodyne method has tested electrical equipment such as Transformers, high-voltage transmission insulators, Transformers, and switchgear, and found that the sensitivity of the detector is very high. The earphones provided by the instrument can clearly hear the ultrasonic waves that we can not hear when the electrical device discharges. At the same time, the instrument's screen shows specific measurements that can be stored in the instrument's memory. Back to the office, it can be downloaded to the computer through supporting software for trend analysis, comparison of the status of similar products, or these data can be exported and edited into device diagnostic reports.
At the same time, the instrument can be used for the internal fault detection of electrical equipment. It can detect the early faults in the equipment and avoid accidents by touching the surface of the equipment or near the gap between the equipment and the surface. It can also be used to accurately locate SF6 leaks of electrical equipment filled with SF6 gases, but its measurement sensitivity is 1 × 10-2 ～ 1 × 10-3 CC / SEC. Very small leaks in the test can not be accurately positioned remotely. Through the actual test in the field, we think that the ultrasonic detector is simple and easy to use, and is very suitable for the daily inspection and evaluation of the operating health of electrical equipment. And has the powerful data processing ability, provides the reliable basis for the equipment state overhaul. At present, the world's most advanced UP10000 model can quickly identify and identify potential problems, store data, quickly change sensitivity, download through RS232 output or customize your instrument settings through compact flash memory cards, set on/off control, recording and other functions.
Problems in Fault Diagnosis of Thermal Power Plant Equipment
Although there are many diagnostic methods and diagnostic systems used in the fault diagnosis of thermal power plant equipment, and good application results have been achieved, but there are also many problems in the practical application, mainly in the following aspects.
1. Detection Means
The inference mechanism of fault diagnosis has reached a very high level, but the acquisition of symptoms has become a bottleneck, that is, the biggest problem is that the detection method can not meet the needs of diagnosis and can not truly reflect the characteristics of the fault.
2. Complex failure mechanism
Understanding fault mechanism is the premise of fault diagnosis. At present, it is difficult to explain the complex failure of some power plant equipment theoretically, and the understanding of its mechanism is not profound.
3. Artificial intelligence applications
Expert system has been successful in the application of artificial intelligence in the fault diagnosis of major equipment in power plants, but there are still some key artificial intelligence application problems that need to be solved, mainly the expression and acquisition of knowledge, self-learning, intelligent identification, and information Fusion.
4. Singularity of diagnostic methods
At present, the diagnosis methods used in the fault diagnosis system of thermal power plant equipment include fuzzy logic method, fault tree analysis method, expert system, artificial neural network, etc.. However, a single diagnostic method is often difficult to achieve the desired diagnostic effect.
5. Fault location
The current fault diagnosis system is often only to the fault type recognition part, can not determine the specific location of the fault, and the state of the equipment to predict the functional research is insufficient.

III. Development of Fault Diagnosis of Thermal Power Plant Equipment
1. Development Trends of Fault Diagnosis Systems
(1) Fault diagnosis system of layered distributed structure
The structure and function of each subsystem of the thermal power unit are distributed and multilayered. This structural hierarchical relationship requires that the diagnostic system be distributed and multilayered and consists of a global diagnostic system and a sub-diagnostic system. The global diagnostic system is responsible for the management of diagnostic tasks, including the decomposition of overall tasks into sub-tasks and the assignment of tasks to sub-diagnostic systems. These tasks are often coupled with each other. After the completion of the diagnosis subsystem, the final conclusion is given through the synthesis of the conclusions of each sub-diagnosis system. The distributed fault diagnosis expert system has the characteristics of high reasoning efficiency, fast diagnosis speed, reliable system, and good timeliness.
(2) Integrated Fault Diagnosis System
Due to the uniqueness of the current diagnostic system in reasoning methods, it is greatly limited to solve the diagnostic problems of complex systems. In the future, the fault diagnosis system of thermal power units will adopt different inference models according to the characteristics of different subsystems, and even use several different inference models for mixed reasoning. The advantages of various propulsion models will be fully utilized, thus improving the reasoning speed and accuracy.
(3) Construction of large monitoring diagnostic centres
In the same power grid, there are many thermal power units of the same type operating at the same time. The benefits of constructing a large monitoring and diagnosis center are very obvious: 1. It is easy to centralize the data of the unit's operation and the information of the unit's health status; It is convenient to share existing knowledge among multiple units and multiple power plants, and facilitate the perfection of knowledge base; 3 It is helpful for the load scheduling of the unit.
(4) Autonomous closed-loop diagnostic system
A fully autonomous, closed-loop fault diagnosis system can complete continuous fault diagnosis without the participation of personnel, form a decision, and then the diagnosis system issues corresponding control orders to apply appropriate control to the unit. To realize autonomous closed-loop diagnosis, it is necessary to have mature and advanced diagnostic techniques. The knowledge base of the diagnostic system must be complete, and the diagnostic system should have a learning mechanism that can diagnose unpredictable failures.
(5) Remote fault diagnosis system based on Internet
With the development of computer network technology, Internet-based remote application system becomes possible. Combining the fault diagnosis system with the Internet can not only obtain a large number of fault cases and diagnosis experiences, but also share diagnostic resources, achieve expert remote consultation, and improve the level of fault diagnosis.
Provide all the above inspection, monitoring instrument level pre-sale after-sales service.