Optical fiber sensors: overview and recent advances Claudio Oton - PowerPoint PPT Presentation

Optical fiber sensors: overview and recent advances Claudio Oton Scuola Superiore Sant’Anna, Pisa, Italy 18 th Annual Workshop of the IEEE Photonics Benelux Chapter Mons, Belgium 22 May 2015

Outline  Introduction to optical fiber sensors • Rayleigh • Raman • Brillouin • FBG  Applications and market  Recent advances  Conclusions 2 Claudio Oton - Optical fiber sensors

Optical fiber sensors  Distributed  Discrete Input light 50 km fiber Reading unit Continuous profile information Distance (km) 3 Claudio Oton - Optical fiber sensors

Optical fiber sensors  Distributed  Discrete Input light Reading unit Discrete parameter information (if many, quasi-distributed ) Distance (km) 4 Claudio Oton - Optical fiber sensors

Backscattering in an optical fiber In absence of elements along the fiber Laser Detector 3 spectral bands: • Rayleigh (elastic scattering) • Brilluoin (acoustic phonons) • Raman (optical phonons) 5 Claudio Oton - Optical fiber sensors

Rayleigh scattering Elastic scattering produced by the atoms 𝛽 𝑆 = 𝐷 C constant (0.7-0.9 dB mm 4 /km) 𝜇 4 Some of the scattered ligth is guided backwards 𝛿 = 𝑇𝛽 𝑆 S : capture factor g ~ 10 -4 km -1 Reflected photon is in phase with the incident 6 Claudio Oton - Optical fiber sensors

Distributed Rayleigh backscattering L P Laser Detector 𝑂 𝐹 𝑗 𝑓 𝑘𝜚 𝑗 𝐹 = 𝑗=0 If L coh << L P  incoherent 𝑂 𝐽 = 𝐽 𝑗 𝑗=0 Typically  P  10ns, resolution  1m Typical OTDR (Optical time domain reflectometry) 7 Claudio Oton - Optical fiber sensors

Coherent Rayleigh scattering If L coh >> L P  Coherent L P 𝑂 𝑂 𝐹 𝑗 𝑓 𝑘𝜚 𝑗 𝐹 = 𝐽 = 𝐽 𝑗 𝑗=0 𝑗=0 Speckles Speckles harm the OTDR trace for loss estimation Coherent (Phase) OTDR Trace 0.25 But speckles are phase sensitive 0.2 0.15 Amplitude (V) They change with variations of strain, temperature, etc 0.1 0.05 Phase OTDR or  -OTDR 0 Distributed acoustic sensor (DAS) 100 200 300 400 500 600 700 800 900 1000 Distance(m) 8 Claudio Oton - Optical fiber sensors

Distributed acoustic sensor Thousands of microphones along the fiber! Geological surveys A. Masoudi, M. Belal, T. Newson, Meas. Sci. Tech. 24 (2013) 9 Claudio Oton - Optical fiber sensors

Raman scattering Interaction with optical phonons Raman Stokes Raman Anti-Stokes 𝑄 = 𝑓 −ℎΔ𝜑 𝑆 𝐵𝑇 𝑙𝑈 Sensitivity: 0.035 dB/K 𝑄 𝑇 Δ𝜑 𝑆 ~13THz for silica glass 10 Claudio Oton - Optical fiber sensors

Raman distributed temperature sensors Experimental result Spontaneous Raman is incoherent TCC at 50°C 60 Raman traces are smooth TCC at 26°C TCC at 10°C Temperature [°C] TCC at -10°C 40 20 0 -20 0 5 10 15 Distance [km] Single-ended RDTS RDTS P AS is very weak configuration Sensing fiber Sensing fiber Pump powers typically high (>1W) Double-ended RDTS RDTS configuration Sensing fiber Sensing fiber (immune to wavelength  Multimode fiber dep. loss variations)  Broadband laser 11 Claudio Oton - Optical fiber sensors

Brillouin scattering Interaction with acoustic phonons (long-range vibrations) Speed of sound  5200 km/s Doppler effect: 2𝑊 𝑏 𝑜 𝜇 0  10GHz (80pm) Δ𝑤 = ν B dependent on temperature and strain   Can be a strain/temperature distributed sensor ~ 0.05 MHz / m B   12 Claudio Oton - Optical fiber sensors

Stimulated Brillouin scattering A counter-propagating probe beam in the Stokes band can be amplified 13 Claudio Oton - Optical fiber sensors

Brillouin Optical Time Domain Analyisis (BOTDA) A pump pulse and a cw probe can extract the gain profile Pump Probe Typical BOTDA setup Fiber 20 km EDFA  MZM CW Laser DC RF EDFA MZM t FBG Waveform Oscilloscope generator 14 Claudio Oton - Optical fiber sensors

Fiber Bragg grating sensors Typical bandwidth  100-200 pm Monitoring the peak position, we can sense vibration and temperature Typical strain response:  1 pm/ m Typical temperature response:  10 pm/K Advantage: fast measurements 15 Claudio Oton - Optical fiber sensors

Multiplexed FBG sensors WDM (limited total grating number) WDM & TDM (many more gratings, using pulsed source) WDM & SDM (FBG sets read in sequence) 16 Claudio Oton - Optical fiber sensors

Fiber Bragg grating sensors Typical sensing parameters:  Strain/Vibration  Temperature Special FBGs:  Pressure  Acceleration  Chemical substances  Electrical current  Magnetic field 17 Claudio Oton - Optical fiber sensors

Application sectors Fire detection Gasoducts Geothermal Oleoducts Solar power plants Structural health Industrial plants Oil rigs Railtrack monitoring Fracking Wind farms Aeronautic FBG-based (strain, vibration, pressure...) 18 Claudio Oton - Optical fiber sensors

Distributed fiber sensor market  Over 1.5 Billion$ distributed fiber optic sensors market forecast in 2013-2017 in strategic industrial sectors Photonic Sensor Consortium Market Survey Report, http://www.igigroup.com/st/pages/photonic_sensor_report.html 19 Claudio Oton - Optical fiber sensors

Distributed fiber sensor market Photonic Sensor Consortium Market Survey Report, http://www.igigroup.com/st/pages/photonic_sensor_report.html 20 Claudio Oton - Optical fiber sensors

The hype cycle 21 Claudio Oton - Optical fiber sensors

Challenges  Cost  Sensing distance  Speed (strain/vibration)  Cost  Spatial resolution (cracks are small)  Cross sensitivity (temperature & strain)  Cost ...did I mention cost? 22 Claudio Oton - Optical fiber sensors

How to improve SNR?  Increase peak power (nonlinear effects!)  Increase measurement time (I lose speed!)  Spatial averaging (I lose spatial resolution!)  Any other idea? 23 Claudio Oton - Optical fiber sensors

How to improve SNR? 3 weighing tests C B A Weighing scale 10 9 8 7 Simple but inaccurate 6 5 4 3 unknown weights 3 2 1 x y z A + s x + y = W y + z = W B + s  SNR improves! x , y , z x + z = W C + s 24 Claudio Oton - Optical fiber sensors

SNR improvement: Coding T R Single pulse response samples  M 1     Acquired Samples y [ i jH ] p x [ i kH ]  j k M  k 0 M-bit moving window P = { p 0 , p 1 , p 2 , p 3 , … … , p M-1 , p 0 , p 1 , p 2 , … p M-1 } Example of backscattered trace with the 7-bit binary Reshaping Pattern P = { 0,1,1,1,0,1,0 }   Decoding: p p ... p p   Cyclic Coefficients Matrix 0 1 M 2 M 1   MxM linear system p p ... p p    1 2 M 1 0 1 *       Y S * X X S Y Y p p ... p p * X  2 3 0 1   : : : : :      p p : p p    M 1 0 M 3 M 2 s  M 1  s  y C Theoretical Coding Gain gain 2 M x 25 Claudio Oton - Optical fiber sensors

Raman DTS with cyclic coding 63-bit code 0.7 (a) Stokes 0.6 SNR can be improved without 0.5 Voltage [V] increasing peak power 0.4 0.3 Anti-Stokes 0.2 0.1 0 5 10 15 20 25 Simple decoding: one matrix multiplication Distance [km] 10 Conventional RDTS Normalize intensity [dB] Simplex-coded RDTS 0  Less noise -10  Longer distances Experimental  -20 Faster measurements Coding Gain: ~6dB  Lower peak powers -30 0 5 10 15 20 25 Distance [km] M. Soto, T. Nannipieri, A. Signorini, et al. Opt. Lett. 36 (13) 2557 (2011) 26 Claudio Oton - Optical fiber sensors

Fast BOTDA with coding Subsecond measurements achieved M. Taki, Y. Muanenda, C. J. Oton, et al, Opt. Lett. 38 (15) 2877 (2013) 27 Claudio Oton - Optical fiber sensors

Dynamic BOTDA sensing Probe fixed at max. slope 200 Hz sampling rate, 12Hz vibration detected Natural vibration modes can be detected R. Bernini, A. Minardo, and L. Zeni. Opt. Lett. 34 , (17) 2613 (2009) 28 Claudio Oton - Optical fiber sensors

Dynamic BOTDA through phase modulation f RF = 850 MHz  1.6 kHz sampling rate  1m resolution  160 m length Immune to gain variations J. Urricelqui, A. Zornoza, M. Sagues, A. Loayssa, Opt. Express 20 , (24) 26942 (2012) 29 Claudio Oton - Optical fiber sensors

BOTDA with better SNR  45km sensing length  No pol. scrambler needed  SNR improvement A. Lopez-Gil, A. Dominguez-Lopez, S. Martin-Lopez, M. Gonzalez-Herraez, J. Lightwave Tech. (in print, 2015) 30 Claudio Oton - Optical fiber sensors

High-spatial resolution BOTDA Can we make resolution < 1m? Use shorter pulses? Phonon lifetime:  10ns Intrinsic limitation of BOTDA spatial resolution 31 Claudio Oton - Optical fiber sensors

Sub-meter BOTDA Differential pulse pair (DPP) technique Substracting slightly different pulses  15 cm resolution achieved!  L = 1km  SNR penalty W. Li, X. Bao, Yun Li, L. Chen Opt. Express 16 , (26) 21616 (2008) 32 Claudio Oton - Optical fiber sensors