Chirped-pulse phase-sensitive optical time domain reflectometry

Abstract

The world today works thanks to big infrastructures, which provide safe energy and transport to its citizens. Such infrastructures (dams, pipelines, bridges, railways, roads...) typically present huge dimensions. Thus, the monitoring of its proper functioning and structural health and protecting them from possible threats is particularly difficult. Distributed optical fiber sensors are a reliable and efficient solution for this problem, since they allow the measurement of vibrations, strain and temperature along all the points of a conventional telecom fiber. Distributed optical fiber sensors based on Rayleigh scattering are particularly useful when real time measurements are required (i. e. vibration detection). In this work, a study about different solutions and alternatives to φOTDR technology limitations has been realized. A new technique, derived from φOTDR, that offers performance features that are significantly superior to the φOTDR features, has been proposed. To do this, first, a detailed study of the fundamentals and the state-of-the-art of the distributed monitoring techniques based on Optical Time Domain Reflectometry (OTDR), and particularly about the phase-sensitive implementation of OTDR (φOTDR), has been realized. The limitation in range and resolution of φOTDR systems associated to the onset of nonlinear effects such as Modulation Instability (MI) has been studied. A traditional φOTDR presents a maximum spatial resolution of tens of meters for a sensing range about a few tens of kilometers (if no distributed amplification technique is implemented). Two techniques for mitigating the non-desired MI effect were proposed. First, the impact of the probe pulse shape in the backscatter trace of φOTDR-based sensing systems is studied. The results show that Gaussian and triangular-shaped pulses present higher robustness against MI than the conventional square-shaped pulses. Secondly, a new technique based on the concept of Chirped Pulse Amplification (CPA) which achieves millimetric resolutions has been proposed. This new φOTDR performance opens to a wide range of applications where high spatial resolutions are required. Another important limitation of these sensors has also been studied: its non-linear behavior when a perturbation is applied. Unless phase recovery techniques or frequency sweeps are implemented (increasing the complexity, cost and measurement time), quantifiable temperature and strain sensing is not possible using φOTDR technology. In the same way, real acoustic sensing is not possible either. To solve this, two techniques are proposed. First, the possibility of using Phase Reconstruction Using Optical Ultrafast Differentiation (PROUD) for recovering the complex field of the backscattered φOTDR signals is analyzed. Implementing PROUD, linear measurements would be possible without the intrinsic complexity of traditional coherence detection. Secondly, the use of chirped-pulses in φOTDR sensors is proposed. The new technique has been named Chirped-Pulse φOTDR. This new technique allows for the measurement of distributed strain and temperature changes, in a single shot and without the requirement of a frequency scan or coherent detection. Temperature/strain resolutions of 0.5mK/4nε and real acoustic sensing have been demonstrated along this work. The limitations of this technique have also been studied and some solutions proposed. A method for mitigating the induced uncertainty introduced by the laser phase noise is proposed. Furthermore, the sensing range of this new sensor is increased using distributed amplification based on stimulated Raman scattering, achieving a sensing length of 75 km with 10 m spatial resolution