Distributed detection of temperature gradients with single-wavelength phase-sensitive OTDR and speckle analysis methods

Abstract

A simple and easy to implement method for temperature gradients detection in real time with single-wavelength ΦOTDR derived from the speckle analysis theory was presented and demonstrated.The method relies solely on a low-cost post-processing of the standard ΦOTDR traces (already acquired for vibration detection).Could be implemented without affecting the distributed vibration detection and with a close to zero cost.A successful test of it has been performed by measuring the temperature decrease of water into an oven as hot-spot. Testing its reliability in a temperature-controlled oven hot-spot.    MOTIVATION• Traditional ΦOTDR allows for high-bandwidth vibration detection.  But provides no information on temperature changes along the fiber.•

Enhanced Distributed Fiber Optic Vibration Sensing and Simultaneous Temperature Gradient Sensing Using Traditional C-OTDR and Structured Fiber with Scattering Dots

“…Due to the pulse coherence, the Rayleigh backscatter waves originating from the many scatterers in a fiber segment which is currently being irradiated by the pulse and thus forming the respective current resolution cell, have a defined phase relationship. The interference of the many individual backscatter waves result in a speckle-like pattern from this composite interferometer depending on the phase differences of all pairs of backscattered waves [9,11,14,59]. These phase differences are each highly sensitive to mechanical and thermal influences, resulting in significant flucuations of the effective backscatter intensity in the time and the spatial domain, respectively.…”

“…A limiting factor is the sampling rate, so that phase changes can be monitored sufficiently. However, since realistic temperature changes for most monitoring purposes lead to phase changes below a few Hz, this should not limit the applicability of the proposed method for even long range monitoring applications [59,60] since the sampling frequency is in the kHz region. One important exception would be fiber optic fire detection, which can involve high frequency signals due to the large temperature gradients.…”

“…The current temperature gradients can then be determined from the slowly varying component of the measurement data and for vibration sensing the higher frequency bands can be used since vibrations are modeled as elastic processes. Previously published work demonstrated that local temperature gradients could be detected from analyzing the intensity output of traditional C-OTDR [59,60]. However, averaging in the time and spatial domain was necessary due to the randomly distributed scatterers in the used standard optical fibers and the ensuing intensity fluctuations; the temperature gradients could also not be precisely quantified.…”

“…This together with the regularized transfer function also enables C-OTDR based quantitative distributed temperature gradient sensing (DTGS) simultaneous to distributed vibration detection under certain conditions. The distributed detection of local temperature gradients by evaluating C-OTDR intensity changes was previously proposed by Garcia-Ruiz et al [59,60], though their method required averaging in the time and space domains due to fading and temperature changes were not accurately quantified. In this paper, we show that absolute temperature changes can be quantitatively determined in real time by observing the corresponding, comparatively slow sinusoidal intensity modulation at discrete sensor positions in the middle of the aforementioned sensing zones, given that monotonous temperature change can be assumed during the time interval of interest and that there are no mechanical perturbations in the corresponding frequency range.…”