TDLAS-Based Detection of Dissolved Methane in Power Transformer Oil and Field Application

Methane is one of the indicative gases in power transformer oil, and the detection of methane dissolved in oil with high accuracy is of great importance for dissolved gases analysis and fault diagnosis inside power transformers. Based on the Beer-Lambert spectral absorption law, dissolved methane de...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE sensors journal 2018-03, Vol.18 (6), p.2318-2325
Hauptverfasser: Jiang, Jun, Zhao, Mingxin, Ma, Guo-Ming, Song, Hong-Tu, Li, Cheng-Rong, Han, Xiao, Zhang, Chaohai
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Methane is one of the indicative gases in power transformer oil, and the detection of methane dissolved in oil with high accuracy is of great importance for dissolved gases analysis and fault diagnosis inside power transformers. Based on the Beer-Lambert spectral absorption law, dissolved methane detection with tunable diode laser absorption spectrum (TDLAS) is proposed in this paper for the advantages of high sensitivity and resolution. On the basis of wavelength modulation spectroscopy, a specialized TDLAS system was established. To meet the actual needs of field testing, the anti-vibration design of an integrated Herriott cell and a gas pressure (P)/temperature (T) setting are worked out. Photodetector, collimator, and Herriott cell are integrated into one component to reduce the effects of vibration. It is investigated that the temperature has little effect on the second harmonic amplitude in the range of 30 °C ~ 50 °C, and the vacuum pressure is reasonably set at about 1 kPa. Experimental results showed that the resolution of sensitivity could be reached as 6.8 mV/(μL/L), the maximum deviation was less than ±4 μL/L, and the response time is less than 5 min. In the end, field application was also carried out, proving it is a prospective online sensing technique to serve oil-immersed power transformers better.
ISSN:1530-437X
1558-1748
DOI:10.1109/JSEN.2017.2788871