Ma Guo-Ming, Zhao Shu-Jing, Jiang Jun, Song Hong-Tu, Li Cheng-Rong, Luo Ying-Ting, Wu Hao
State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources, North China Electric Power University, Beijing, 102206, P. R. China.
College of Automation Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing, 210016, China.
Sci Rep. 2017 Nov 2;7(1):14961. doi: 10.1038/s41598-017-13823-0.
Dissolved gas analysis (DGA) is widely used in monitoring and diagnosing of power transformer, since the insulation material in the power transformer decomposes gases under abnormal operation condition. Among the gases, acetylene, as a symbol of low energy spark discharge and high energy electrical faults (arc discharge) of power transformer, is an important monitoring parameter. The current gas detection method used by the online DGA equipment suffers from problems such as cross sensitivity, electromagnetic compatibility and reliability. In this paper, an optical gas detection system based on TDLAS technology is proposed to detect acetylene dissolved in transformer oil. We selected a 1530.370 nm laser in the near infrared wavelength range to correspond to the absorption peak of acetylene, while using the wavelength modulation strategy and Herriott cell to improve the detection precision. Results show that the limit of detection reaches 0.49 ppm. The detection system responds quickly to changes of gas concentration and is easily to maintenance while has no electromagnetic interference, cross-sensitivity, or carrier gas. In addition, a complete detection process of the system takes only 8 minutes, implying a practical prospect of online monitoring technology.
溶解气体分析（DGA）广泛应用于电力变压器的监测与诊断，因为电力变压器中的绝缘材料在异常运行条件下会分解出气体。在这些气体中，乙炔作为电力变压器低能火花放电和高能电气故障（电弧放电）的一个标志，是一项重要的监测参数。在线DGA设备目前所采用的气体检测方法存在交叉敏感性、电磁兼容性和可靠性等问题。本文提出一种基于可调谐半导体激光吸收光谱（TDLAS）技术的光学气体检测系统，用于检测变压器油中溶解的乙炔。我们在近红外波长范围内选择了波长为1530.370 nm的激光，使其对应乙炔的吸收峰，同时采用波长调制策略和赫里奥特池来提高检测精度。结果表明，该检测系统的检测限达到0.49 ppm。该检测系统对气体浓度变化响应迅速，易于维护，且不存在电磁干扰、交叉敏感性或载气问题。此外，该系统完整的检测过程仅需8分钟，这意味着在线监测技术具有实际应用前景。
