Development Of 35kV Oil-immersed Power Transformer With Built-in Distributed Optical Fiber And Analysis Of 3D Internal Temperature Distribution

As the key equipment of power grid,transformer plays an important role in the security and stability of power supply in one region.But at present,the on-line monitoring methods for the internal temperature of transformer are basically the point-type measurements.And also,the detection results will change with the different sensor installation positions,leaving a large number of monitoring blind areas.In recent years,the rapidly developed distributed optical fiber temperature sensing technology can realize the continuous temperature detection along the optical fiber laying path.This technology has great application potential for the distributed online monitoring of the whole space area inside the transformer.The revelation of temperature distribution inside the transformer also serves as an important reference for both the manufacturers and the on-site operation and maintenance personnel.In this paper,a 35 kV ONAN transformer prototype has been upgraded with internally integrated distributed optical fiber sensor(DOFS).Through temperature rise test,the temperature field distribution inside an operating power transformer has been revealed.The main research contents and achievements of this paper are as follows:Firstly,to illustrate the principle of distributed optical fiber sensing,the sensing mechanism and system based on Rayleigh scattering,Brillouin scattering and Raman scattering are introduced respectively.And the demodulation algorithms of Raman scattering are analyzed emphatically.Then,the security and stability of the DOFS in the oil-immersed transformer are studied.The feasibility of the optical fiber is verified from three aspects: the electric field stability after the fiber installation,the compatibility with transformer oil and the quality monitoring scheme during the optical fiber laying process.Based on the former work,the fiber laying and leading out scheme of three-phase winding,three-phase iron core,oil tank and clamp parts are studied,and the development and ex-factory type test of the 35 kV transformer prototype are completed.Before the temperature rise test,Gaussian convolution is used to reduce the background noise of the entire optical fiber,which furtherly improves the temperature detection accuracy.Finally,during the whole process of the short-circuit temperature rise test,by establishing the spatial position function of optical fiber and matching with the measured data points,the real-time three-dimensional temperature field of transformer winding and iron core is realized.The local overheating regions development process of different transformer components during the whole test are furtherly studied.At the same time,the hot spots of each phase winding and iron core during the transformer operation are successfully located and tracked,which furtherly reveals the evolution of hot spots in each region and breaks the traditional cognition that the hot spots of winding are always located at the top of winding(hot spots of high(low)voltage winding are respectively located at90%(80%)of winding height).Based on the analysis of the feld measured data,the IEC traditional hot spot model and winding temperature distribution model are modified.The "Three-Section" winding temperature distribution model is proposed,and the general calculation formula of winding hot spot is given,which provides an important reference for both the transformer manufacturers and field operation and maintenance personnel.