Reliability analysis and comparison of demodulation methods for Raman distributed temperature sensor

Abstract

We presented theoretical and experimental analysis of the reliability of the three frequently-used demodulation methods in Raman distributed temperature sensor (RDTS), including single-ended RDTS using Anti-Stokes signal only and using ratio between Anti-Stokes signal to Stokes and double-ended RDTS using Anti-Stokes signal only. In particular, the effects of temperature fluctuations, bending losses and variations of the attenuation at Anti-Stokes wavelength on the sensing performance are investigated at different temperatures. Experimental results demonstrate that the temperature influence at $60\,^{\circ }\hbox {C}$ for single-ended RDTS using Anti-Stokes only are 1.1, 7.6 and $7.5\,^{\circ }\hbox {C}$, respectively, when temperature fluctuates about $1\,^{\circ }\hbox {C}$, bending losses are 0.22 dB and variations of the attenuation at AS wavelength are 0.22 dB, while those for single-ended RDTS using ratio between Anti-Stokes to Stokes are 1.2, 1.1 and $9.0\,^{\circ }\hbox {C}$, respectively. The influence of temperature fluctuations at $60\,^{\circ }\hbox {C}$ for double-ended RDTS using Anti-Stokes signal only are $1.2\,^{\circ }\hbox {C}$, and the effects of the bending losses and variations of the attenuation at AS wavelength can be almost canceled out. Experimental results show good agreement with the theoretical analysis, allowing for guidance in applied conditions.