Detection of Dissolved Acetylene in Power Transformer Oil Based on Photonic Crystal Fiber

Dissolved gas analysis (DGA) is an important online monitoring technique for power transformers. When there are some faults in oil-immersed transformers, insulation materials will decompose into various small-molecule gases and continually dissolve in transformer oil. In order to figure out the composition and the concentration of dissolved gases, an optical gas sensing system based on photonic crystal fiber was developed. Hollow core photonic crystal fiber (HC-PCF) used in the system played the role of an absorption gas cell. The gas diffusion time was studied theoretically and experimentally, and the result showed that by using a 1.0 m length long HC-PCF, the theoretical diffusion time for acetylene was about 3.47 hours. Various acetylene gas samples got detected. Through Allan variance analysis, the sensitivity of the proposed sensing system could reach the best with the highest signal-to-noise ratio (SNR) when the average time was 29 s, and a minimum detection limit of 4.5 ppm could be achieved. The proposed sensing system showed tremendous advantages, such as low detection limit, no cross interference, high sensitivity. This gas sensing technique has a good application prospect in DGA for power transformers.