Detection Of Fault Characteristic Gas In Transformer Oil Based On Photoacoustic Spectroscopy

The safe operation of power system is the key to directly affect the power transportation.As an indispensable equipment of power system,power transformer is the necessary condition for the reliable operation of power system when it runs safely.During the long-term operation of transformer,it often fails because of the influence of external factors and aging of key components.Therefore,the normal operation of transformer protection and reasonable maintenance of insulation system can largely ensure that transformer has a relatively long service life,and preventive and predictive maintenance is the key to improve the service life of transformer and improve the reliability of power supply.Therefore,it is particularly important to study the detection of dissolved fault characteristic gases in oil and transformer fault diagnosis.When common detection methods can not simultaneously meet the high sensitivity and loss-free detection of dissolved gases in oil,it is further impossible to use existing technology to achieve long-term reliable online monitoring of dissolved aging characteristic gases in oil.Using the advantages of photoacoustic spectroscopy to remedy these shortcomings,the photoacoustic spectroscopy technology is studied to dissolve fault characteristic gas in transformer oil,which lays a foundation for accurate detection and diagnosis of dissolved fault gas in transformer oil.The main contents of this paper are as follows:(1)The interaction between molecule and laser is analyzed theoretically,and the specific principle of producing absorption spectrum is given.From the macroscopic point of view,the basic principle of photoacoustic spectrum detection is analyzed concretely,and the mechanism of photoacoustic signal generation is analyzed,which lays a theoretical foundation for the understanding of later experimental phenomena.(2)The importance of component selection in the whole process of photoacoustic signal detection is expounded.Among them,the selection of lasers is particularly important.Quantum cascade lasers have the largest wavelength range and large adjustable range,and their power meets the practical application.Therefore,it is the best choice for excitation of photoacoustic spectral signals.Secondly,in the design of photoacoustic cell,considering the basic requirements and the particularity of transformer fault gas detection,differential photoacoustic cell is the best choice.Then,in the selection of acoustic sensors,electret microphone is the simplest,but it can not resist interference;so the best choice is the silicon cantilever beam acoustic sensor.In addition,the methods of oil and gas targeting for common faults are summarized.In this chapter,the key points in the experimental process are analyzed from the practical components,and an experimental platform for photoacoustic spectroscopy detection is established.(3)From the experimental point of view,photoacoustic spectroscopy technology is studied in the detection of transformer fault gas.The infrared absorption spectrum of the object to be measured is determined,and the best characteristic spectrum is determined to be 2103.3 cm-1.The influence of laser modulation depth,temperature and pressure on the photoacoustic signal of measured material is studied experimentally.It is found that there is a linear function between the modulation depth and the photoacoustic signal.It is found that the photoacoustic signal decreases with the increase of temperature,so the temperature should be kept constant as far as possible during the experiment.In this case,the low temperature environment inside the photoacoustic cell is maintained;the influence of the pressure inside the photoacoustic cell on the photoacoustic signal is studied experimentally.The results show that the absorption coefficient of the gas increases with the increase of the pressure,but the photoacoustic signal decreases.The main reason is that the response sensitivity of the silicon cantilever beam decreases with the increase of the pressure.Decrease of starting signal.Carbon monoxide standard gas samples with different concentrations were used for detection experiments.The results show that there is a linear function relationship between the photoacoustic signal and the concentration in a certain concentration range.The fitting coefficient can reach 0.9687 by using one-dimensional linear regression method.In addition,the long-term monitoring data of oil and gas in two operating transformers are analyzed,and the probable fault types in transformers are predicted by the ratio of several gases,and the relevant treatment suggestions are put forward.In this paper,photoacoustic spectroscopy technology and its application in transformer fault gas detection are discussed.The research results provide a useful reference for further study of other fault characteristic gas detection in the future,and for more field detection and diagnosis technology research in the future.