Est´ andares

tection System

3.5.1 Introduction

Distributed optical fibre sensors based on Brillouin scattering, with sub-metre spatial resolution capabilities, have been previously reported, but were confined to relatively short sensing lengths, and focused on either measuring the Brillouin frequency shift [5] or power [6]. Several applications, such as the continuous monitoring of temperature/strain in underground power cables, live optical links and large scale structures, require sensors with relatively high spatial resolution combined with long range [7][8]. In this section, the optimised CBS, along with the microwave detection system, was experimentally demonstrated as a long range high spatial resolution distributed temperature sensor.

3.5.2 Experimental Details and Measurements

The experimental arrangement for the coherent detection of anti-Stokes spontaneous Brillouin backscatter, using the microwave detection system, is illustrated in figures 3.1 and 3.5. The source was a tuneable laser at 1533.2nm, with ∼1MHz linewidth, and 100µW CW output. Two EDFAs and an acousto-optic modulator were used to generate a probe pulse of 25mW and 20ns with a repetition rate of ∼80Hz, which was launched into the 32km sensing fibre. A single pass EDFA preamplifier was used to amplify the weak backscattered signal generated in the sensing fibre prior to mixing with a 1.8mW optical LO. A 20GHz lightwave detector and the microwave detection system allowed the collection of time-domain traces centred at the desired RF frequencies. The sensing fibre was standard SMF in 3 sections. The first 30km remained on the original spools at room temperature. The next 400m were placed in an oven at 60◦_{C. The subsequent 1.6km}

was maintained at room temperature and zero strain as a reference. To determine the spatial resolution and accuracy of the system, Brillouin frequency measurements were taken between 30 and 30.8km, where the central 400m length of fibre was heated to 60◦C, and the remaining fibre was at room temperature (20◦C). The temperature

change along the sensing fibre was determined by analyzing the frequency shift of the Brillouin backscatter. Brillouin spectra were built from 15 separate backscatter traces, each averaged 215 _{times, taken every 10MHz, starting at 10.99GHz. A Lorentzian} curve was fitted to each spectrum and the peak frequency was evaluated at each point along the sensing fibre. First, the spatial resolution was evaluated by measuring the rise time at a heated section; then the temperature change at the heated section 30km down the sensing fibre was measured. The frequency resolution along the sensing fibre was evaluated every 5km, averaged over a length of 500m, and then converted to the corresponding temperature resolution.

3.5.3 Experimental Results

The spatial resolution was limited by the rise time of the available modulator AOM used to generate the probe pulse, which limits the pulse width to ∼20ns. For clarity, figure 3.8 shows a 10-90% rise-time measured at the front end of the sensing fibre. It agrees with the expected performance, which is governed by the duration of the pulse and not by the electronics of the detection system.

Figure 3.8: A (10-90%) rise time of temperature change at 600m down the sensing fibre indicating spatial resolution of ∼2m.

Chapter 3 Optimising the Coherent Brillouin Sensor 50 Figure 3.9 illustrates the temperature change at the heated section 30km down the sensing fibre.

Figure 3.9: Temperature change at the heated section 30km down the sensing fibre.

The sensor was able to record temperature changes of less than 1.1◦_{C at up to 20km}

distance. The error increased with distance, but was less than 1.7◦_{C at the end of the}

sensing length. Figure 3.10 illustrates the temperature resolution along the sensing fibre.

Figure 3.10: The RMS temperature errors for the 20ns pulse width measurements along the sensing fibre at 5km intervals, averaged over a length of 500m.

The system’s accuracy may be improved by performing more time-domain trace averages, which will reduce noise but increase the measurement time. The rise time of the currently used RF diode rectifier (∼10ns) places an upper limit on the system’s spatial resolution, but the detection system is now potentially capable of∼60cm spatial resolution, provided its BPF bandwidth is tuned to its maximum (∼80MHz) and a faster microwave diode rectifier is used.

3.5.4 Discussion

A microwave detection system was demonstrated for coherent detection of backscat- tered spontaneous anti-Stokes Brillouin signals for distributed temperature/strain sen- sors. The aim was to improve the spatial resolution of the previously reported results by an order of magnitude, and this was achieved. The design represents an important and necessary advance for constructing a commercial sensor, as it dispenses with the previous need for an expensive ESA. It also promises practical high spatial resolution and accurate long range Brillouin based distributed optical sensing systems, with poten- tial for further improvements in spatial resolution. As a future application of this new system, the set-up was used to characterise fibres designed to have a high SBS threshold for use in high power fibre lasers. This is reported in the next section.

Chapter 3 Optimising the Coherent Brillouin Sensor 52