Professional Documents
Culture Documents
Distributed Acoustic Sensing in Geophysics Methods and Applications Yingping Li Full Chapter
Distributed Acoustic Sensing in Geophysics Methods and Applications Yingping Li Full Chapter
Distributed Acoustic Sensing in Geophysics Methods and Applications Yingping Li Full Chapter
Yingping Li
Martin Karrenbach
Jonathan B. Ajo-Franklin
Editors
This work is a co-publication of the American Geophysical Union and John Wiley & Sons, Inc.
This edition first published 2022
© 2022 American Geophysical Union
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise, except as permitted by law. Advice on how to obtain
permission to reuse material from this title is available at http://www.wiley.com/go/permissions.
The rights of Yingping Li, Martin Karrenbach, and Jonathan B. Ajo-Franklin to be identified as the editors of this work have been
asserted in accordance with law.
10 9 8 7 6 5 4 3 2 1
CONTENTS
Preface .................................................................................................................................................................... xv
Part II Distributed Acoustic Sensing (DAS) Applications in Oil and Gas, Geothermal,
and Mining Industries
5 Field Trial of Distributed Acoustic Sensing in an Active Room-and-Pillar Mine ........................................... 67
Xiangfang Zeng, Herbert F. Wang, Neal Lord, Dante Fratta, and Thomas Coleman
6 On the Surmountable Limitations of Distributed Acoustic Sensing (DAS) Vertical Seismic
Profiling (VSP) – Depth Calibration, Directionality, and Noise: Learnings From Field Trials ....................... 81
Albena Mateeva, Yuting Duan, Denis Kiyashchenko, and Jorge Lopez
7 Denoising Analysis and Processing Methods of Distributed Acoustic Sensing (DAS) Vertical
Seismic Profiling (VSP) Data........................................................................................................................... 93
Yuan-Zhong Chen, Guang-Min Hu, Jun-Jun Wu, Gang Yu, Yan-Peng Li, Jian-Hua Huang,
Shi-Ze Wang, and Fei Li
8 High-Resolution Shallow Structure at Brady Hot Springs Using Ambient Noise Tomography (ANT)
on a Trenched Distributed Acoustic Sensing (DAS) Array........................................................................... 101
Xiangfang Zeng, Clifford H. Thurber, Herbert F. Wang, Dante Fratta, and Kurt L. Feigl
v
vi CONTENTS
12 Coalescence Microseismic Mapping for Distributed Acoustic Sensing (DAS) and Geophone
Hybrid Array: A Model-Based Feasibility Study ........................................................................................... 161
Takashei Mizuno, Joel Le Calvez, and Daniel Raymer
15 Surface Wave Imaging Using Distributed Acoustic Sensing Deployed on Dark Fiber: Moving
Beyond High-Frequency Noise..................................................................................................................... 197
Verónica Rodríguez Tribaldos, Jonathan B. Ajo-Franklin, Shan Dou, Nathaniel J. Lindsey,
Craig Ulrich, Michelle Robertson, Barry M. Freifeld, Thomas Daley, Inder Monga, and Chris Tracy
16 Using Distributed Acoustic Sensing (DAS) for Multichannel Analysis of Surface Waves (MASW)............. 213
Chelsea E. Lancelle, Jonathan A. Baldwin, Neal Lord, Dante Fratta, Athena Chalari, and
Herbert F. Wang
17 A Literature Review: Distributed Acoustic Sensing (DAS) Geophysical Applications Over the
Past 20 Years ................................................................................................................................................ 229
Yingping Li, Martin Karrenbach, and Jonathan B. Ajo-Franklin
Index...................................................................................................................................................................... 293
LIST OF CONTRIBUTORS
vii
viii LIST OF CONTRIBUTORS
xiii
PREFACE
Distributed acoustic sensing (DAS) systems are opto- need for a book on DAS geophysical applications. We had
electronic instruments that measure acoustic interactions interesting discussions with many scientists and engineers
(distributed strain or strain rate) along the length of a working on the frontier of DAS geophysical applications
fiber-optic sensing cable. DAS observation systems can about the potential for a book. We specially recognize
record sound and vibration signals along several tens of Biondo L. Biondi, Thomas M. Daley, William Ellsworth,
kilometers of sensing optical fiber with fine spatial Mahmoud Farhadiroushan, Barry M. Freifeld, Albena
resolution (1–10 m) and over a wide frequency range Mateeva, Robert Mellors, Clifford H. Thurber, Herbert
(from millihertz to tens of kilohertz). DAS provides a Wang, and Mark E. Willis, as well as many others for their
large sensing aperture for acquiring high-resolution encouragement.
acoustic data in both time and space domains. The advan- During the 2017 AGU Fall Meeting in New Orleans,
tages of DAS technology have enabled its rapid adoption we fortunately got an opportunity to meet with the
across a range of applications, including geophysics AGU Books Editor, Dr. Bose, who was already aware
geohydrology, environmental monitoring, geotechnical of this rapidly growing scientific field. We discussed a
and civil engineering (railroad, tunnel, and bridge moni- potential book on DAS geophysical applications, and
toring), hazard mitigation and prevention, and safety she was very supportive and invited us to submit a book
and security fields. proposal for an AGU monograph. With no surprise, this
This monograph focuses on various DAS applications DAS book proposal received very positive comments and
in geophysics. The use of DAS in the oil, gas, geothermal, constructive suggestions from all reviewers. Several
and mining industries for high-resolution borehole and reviewers also asked for an opportunity to submit their
surface seismic imaging, and microseismic monitoring own contributions to this monograph. We are grateful
for hydraulic fractures has accelerated with improvements to those anonymous reviewers of the book proposal for
in the sensitivity of DAS instruments, advances in real- their positive comments and constructive suggestions that
time big data processing, and flexible and economic led this book to be initiated.
deployment of fiber-optic sensing cables. There is also This monograph is organized into four parts. Part I
growing interest in using DAS for critical geophysical starts with principles of DAS measurements and instru-
infrastructure applications, such as earthquake and ments. DAS interrogation units transmit a pulse of laser
near-surface passive seismic analysis, including the devel- light into the fiber. As this pulse of light travels down
opment of tailored or novel numerical techniques. This the fiber, interactions within the fiber result in light reflec-
book aims to engage both the scientific and industrial tions known as backscatter (Rayleigh scattering). Back-
communities to share their knowledge and experiences scatters are determined by tiny strain events within the
of using DAS for novel geophysical applications. fiber, which in turn are caused by localized acoustic
The origin of this book was the 2017 American energy. This backscattered light travels back up the fiber
Geophysical Union (AGU) Fall Meeting, when scientists toward the interrogation unit where it is sampled. Part II
and engineers from both industry and academia gathered introduces various DAS applications in the oil and gas,
in New Orleans to present their fantastic research outcomes geothermal, and mining industries. Part III looks at
on DAS instrumentations and applications in geophysics DAS applications in seismic monitoring. DAS microseis-
and seismology. As DAS technologies have continued to mic monitoring of hydraulic fracturing is an industry
advance, more and more successful geophysical DAS application but with passive seismic sources. The micro-
applications have been reported and published in different seismic DAS method has been shown to have sufficient
geophysical and seismological journals, abstracts, and pro- sensitivity to record very small magnitude microearth-
ceedings of technical conferences, such as the AGU, the quakes with DAS deployed in boreholes. Microseismic
Society of Exploration Geophysicists (SEG), the European DAS systems can be naturally extended to monitoring lar-
Association of Geoscientists and Engineers (EAGE), the ger earthquake activity, and slow deformation of Earth’s
Society of Petroleum Engineers (SPE), and the Seismolog- structure with large-scale fiber-optic networks. Part IV
ical Society of America (SSA). However, few DAS books discusses DAS environmental and shallow geophysical
are available on DAS principles, instrumentation, and geo- applications such as geological carbon dioxide sequestra-
physical applications. Many attendees at the DAS sessions tion. The final chapter presents a review of fiber optical
at the 2017 AGU Fall Meeting expressed that there was a sensing applications in geophysics including historical
xv
xvi PREFACE
developments and recent advances. The list of over 900 lit- This monograph will be the first comprehensive hand-
erature references of DAS and related technologies will book for anyone interested in learning DAS principles
benefit readers, especially newcomers who have just and applications. We hope that the book will have a wide
stepped into this fast-growing field. spectrum of readers – such as geophysicists, seismologists,
We would like to thank the AGU Books Editorial geologists, and geoscientists; environmental scientists; and
Board for supporting this monograph. Without the graduate and undergraduate students in geophysics and
efforts from contributing authors it would not have geoscience – with a common interest in DAS geophysical
been possible to accomplish this project. We would applications. This book also provides a common platform
also like to thank the many volunteer reviewers who to the scientific and industry communities to share state-of-
spent tremendous amounts of time and effort to ensure the-art DAS technology.
that each chapter is of the highest quality. We appre-
ciate Jonathan B. Ajo-Franklin, Biondo L. Biondi, Yingping Li
Mahmoud Farhadiroushan, Albena Mateeva, and BlueSkyDas (formerly Shell), USA
Siyuan Yuan for providing their pictures as candidates
for the book cover design. Thanks are also extended to Martin Karrenbach
the AGU Books editorial team at Wiley, especially Dr. OptaSense Inc. (A LUNA Company), USA
Rituparna Bose, Layla Harden, Noel McGlinchey, Jonathan B. Ajo-Franklin
Vaishali Rajasekar, Sangaprabha Mohan, Bobby Rice University and Lawrence Berkeley
Kilshaw, Nithya Sechin, and Emily Bae, for their National Laboratory, USA
organization, management, and cover design.
Part I
Distributed Acoustic Sensing (DAS)
Concept, Principle, and
Measurements
ABSTRACT
The distributed acoustic sensor (DAS) offers a new versatile tool for geophysical applications. The system allows
seismic signals to be recorded along tens of kilometers of optical fiber and over a wide frequency range. In this
chapter we introduce the concept of DAS and derive an expression for the system response by modeling the
superposition of the coherent backscatter fields along the fiber. Expressions are derived for converting the optical
phase to strain rate and equivalent particle motion. We discuss DAS signal processing and denoising methods to
deal with the random nature of the Rayleigh scatter signal and to further improve dynamic range and sensitivity.
Next we consider DAS parameters such as spatial resolution, gauge length and directionality in comparison with
geophones. We present some field trial results that demonstrate the benefits of the DAS for vertical seismic
profiling and microseismic detection. Finally we discuss the fundamental sensitivity limit of DAS. We consider
how the scattering properties of conventional fiber can be engineered to deliver a step-change DAS performance,
beyond that of conventional geophones and seismometers. Theoretical findings are illustrated by the field data
examples, including low-frequency strain monitoring and microseismic detection.
1.1. DISTRIBUTED ACOUSTIC SENSOR (DAS) random scatter centers along the fiber. We use the term
PRINCIPLES AND MEASUREMENTS acoustics in a broad physical sense here, like any
propagation of mechanical disturbances (Lewis,
In this chapter, we consider the principles and perfor- 1985). We review different DAS systems, including
mance of distributed and precision engineered fiber optic direct-intensity-detection and phase-detection schemes,
acoustic sensors for geophysical applications (Hartog where we derive a mathematical relationship for
et al., 2013; Parker et al., 2014). In particular, system optical phase recovery. Our aim is to explain the nature
parameters such as spatial resolution, dynamic range, sen- of the distributed acoustic signal and describe the
sitivity, and directionality are examined for seismic and natural limitations for DAS measurements. Such infor-
microseismic measurements. mation is needed to optimize DAS recording para-
In this first section, we consider the measurement meters for geophysical applications. Examples of
principle of DAS, which uses naturally occurring DAS parameter optimization for seismic applications
can be found in Section 1.2. We also present some
examples of active and passive seismic field data in
Sections 1.2 and 1.3.
Silixa Ltd. Elstree, UK
Distributed Acoustic Sensing in Geophysics: Methods and Applications, Geophysical Monograph 268, First Edition.
Edited by Yingping Li, Martin Karrenbach, and Jonathan B. Ajo-Franklin.
© 2022 American Geophysical Union. Published 2022 by John Wiley & Sons, Inc.
DOI:10.1002/9781119521808.ch01
3
4 DISTRIBUTED ACOUSTIC SENSING IN GEOPHYSICS
Optical fiber
Acoustic field
Optical phase shift between beginning
and end of pulse
1.1.1. DAS Concept et al., 2000), along multi-kilometer fiber cables (Juarez
et al., 2005; Shatalin et al., 1998).
The principle of distributed sensing is based on optical The principle of the COTDR system can be understood
time domain reflectometry (OTDR), as indicated in by analyzing the radiation generated by localized scatter
Figure 1.1. When a laser pulse travels down an optical centers (Taylor & Lee, 1993). Here, the coherent scattered
fiber, a tiny portion of the light is naturally scattered light can be represented as the result of two reflections
through Rayleigh, Raman (Dakin & Culshaw, 1989), with random amplitude and phase. When the fiber is
and Brillouin (Parker et al., 1998) interactions and returns strained, the backscatter intensity varies in accordance
to the optoelectronic sensor unit. The measurement loca- with the strain rate (Figure 1.2), but with an unpredictable
tion can be determined from the time taken for the laser amplitude and phase, which changes along the fiber
pulse to travel down the sensing fiber, and the backscatter (Shatalin et al., 1998). As a result, the signal cannot be
light to return to the optoelectronic sensor unit. effectively accumulated for multiple seismic pulses: the
Figure 1.1 shows the basic principle of DAS, where the fiber response to strain is highly nonlinear, and therefore
sensing fiber is excited with a coherent laser pulse and the the changes in amplitude and phase cannot be directly
Rayleigh backscattered interference along the fiber is matched to the original strain affecting the fiber. The next
detected and digitized. An acoustic wave elongates the section discusses ways of addressing this. Therefore,
fiber and so changes the optical phase shift between back- COTDR systems are not that useful for seismic
scatter components from the leading and trailing parts of applications.
the optical pulse. As a result of interference, the intensity With the phase DAS technique, the method for optical
of the returning light changes from pulse to pulse. It is also phase analysis is a key feature of system design. All tech-
possible to determine the optical phase to recover acoustic niques rely on phase modulation between the beginning
phase so there are two classes of DAS, based on the detec- and end of a pulse, which can be considered as a double
tion of: (i) optical intensity and (ii) optical phase. With the pulse. Such modulation can be performed before or after
intensity DAS technique, also referred to as coherent opti- light propagation over optical fiber, as indicated in
cal time domain reflectometry (COTDR), a perturbation Figure 1.3. We have limited our discussion to schemas
along the fiber is detected by measuring the changes in that have been patented and implemented in practice.
the backscatter intensity from pulse to pulse, as indicated In one scheme, which is similar to that used for multi-
in Figure 1.2. COTDR has been used for the detection of plexed interferometer sensors (Dakin, 1990), two laser
temperature changes (Rathod et al., 1994; Shatalin et al., pulses with different frequencies may be sent down the
1991) and acoustic vibration (Juškaitis et al., 1992; Posey fiber (Figure 1.3a). In this case, the acoustic phase shift
SEISMIC DATA FROM DISTRIBUTED AND ENGINEERED FIBER OPTIC ACOUSTIC SENSORS 5
Laser Modulator
Fiber
under Test
Signal a.u.
Digitiser Receiver
Distance
Delay
Fiber
Digitiser Receiver under Test
b Interferometer Farhadiroushan et
with 3x3 coupler al., 2010
and embedded Circulator
delay
Laser MOD
R1
Digitiser
R2 Fiber
R3 under Test
Receivers Delay
c Heterodyne Hartog & Kader
2012
Circulator
Laser AOM
f1 f2
Digitiser
Fiber
Balanced under Test
Receivers
d Different Farhadiroushan et
frequency al., 2010
comparator
Circulator
Handerek, 2016
Laser MOD
f±Δf
Crickmore & Ku,
Fiber 2017
Digitiser Receiver under Test
Figure 1.3 DAS schemas: MOD—intensity and frequency modulator; AOM—acousto-optic modulator.
6 DISTRIBUTED ACOUSTIC SENSING IN GEOPHYSICS
e(z)
Scattering r(z) Doppler scattering r(z)exp[iΩ(z)t]
ν(z) z = c tʹ / 2
𝜓0 L0
I(z,t)
will be transferred to a frequency difference and can be Let us calculate how the intensity of backscattered light
measured in the photocurrent radio frequency domain. changes when a section of fiber is moving with speed v(z)
Other solutions, such as that shown in Figure 1.3b, con- under a seismic wave (Figure 1.4). The Rayleigh centers
tain an embedded delay line that defines the spatial reso- will move with the fiber, so the frequency of the backscat-
lution. We will focus our analysis on this class of systems. tered light will experience a Doppler shift Ω(z) propor-
Another configuration uses optical heterodyne, as shown tional to its speed, like for Brillouin scattering (Hartog,
in Figure 1.3c, where the backscatter signal is continu- 2017). The aim of DAS can be considered as the measure-
ously mixed with a slightly frequency shifted local oscilla- ment of Doppler shift for Rayleigh scattering derived
tor laser. In this case, the elongation along the fiber is from the detected photocurrent. The phase shift can be
measured by computing the difference of the accumulated measured between two separate points in space, and then
optical phase between two sections of fiber, and the meas- the resultant Doppler shift can be recovered with spatial
urement is carried out at differential frequency f1 − f2. integration, as will be shown later in the text. The first step
Although this technique offers a flexible spatial resolu- is to analyze changes in intensity between different optical
tion, it requires a laser source with extremely high coher- pulses to derive the fiber speed information, which will be
ence to achieve reasonable signal-to-noise ratio (SNR) equal to the ground speed in a seismic wave.
performance over several tens of kilometers of fiber. Consider a coherent optical pulse e(t ) that is launched
The details of the heterodyne concept are thoroughly cov- into a single-mode optical fiber. The backscattered optical
ered elsewhere (Hartog, 2017). Another method involves field E(t ) at time t for light reemerging from the launch
sending multiple pulses of different frequencies, either in end can be expressed as a superimposition of delayed par-
series or from pulse to pulse, and then computing the tial fields backscattered with a reflection coefficient r0(z)
phase of the backscatter signal, as indicated in along the fiber axis z (Shatalin et al., 1998). This ampli-
Figure 1.3d. The phase calculation in this case is similar tude coefficient represents coupling between the forward
to first case (Figure 1.3a). and backward modes. For a speed of light in the fiber
c ≈ 2 108m/c, and wave propagation constant β, we
z
can use group and phase delays 2z/c and 2 β x dx ,
1.1.2. DAS Interferometric Optical Response 0
respectively. So, the emerging field will depend on inter-
The theoretical concept of DAS is based on the assump- ferometer optical delay, or gauge length, L0 as:
tion that the Rayleigh centers have no microscopic
motion, but they are “frozen” inside glass during manu- L
2z 2z 2L0
facture. In this case, the positions of the centers depend E t = e t − +e t − − r0 z
on the macroscopic motion of fiber and can coincide with c c c
0
the ground speed around a buried fiber (v). There are two z
time scales of relevance to DAS: (1) as optical pulse travels exp 2i β x dx dz
with speed c, significantly faster than ground motion, this
0
dictates the spatial resolution; (2) seismic motion is
responsible for interference changes pulse to pulse, which (1.1)
can be used to recover the seismic signal. All parameters For a regular fiber, the phase shift term in Equation 1.1
for both fast and slow motions are summarized in the can be separated into a constant part and a part changing
table of variables at the end of the chapter. with “slow” time t, representing pulse-to-pulse parameter
SEISMIC DATA FROM DISTRIBUTED AND ENGINEERED FIBER OPTIC ACOUSTIC SENSORS 7
variation with Doppler shift frequency Ω(z), which is pro- shift ΔΩ(z) = Ω(z) − Ω(z − L0) can be represented via var-
portional to scattering particle velocity v(z) and wave- iation of intensity I(z, t) = E(z, t)E(z, t)∗ . The expression in
length frequency ω. braces in Equation 1.6 represents a two-beam interfer-
z ence, so the intensity will vary harmonically depending
on the phase. As we are interested in the intensity change,
β x, t dx = β0 z + Ω z t (1.2)
only the interference term needs be taken into considera-
0 tion, which can be reshaped using the intensity derivative:
4π neff K ε
Ω z =ω v z c= vz (1.3) ∂I ∂E z, t ∂E z, t ∗
λ = E z, t ∗
+ E z, t (1.7)
Here the strain coefficient Kε relates the physical and ∂t ∂t ∂t
optical length of fiber, neff is fiber effective refractive Then using convolution properties ∂[a b(t)]/∂t =
index, and λ is the laser wavelength. Equations 1.2–1.3 a ∂b(t)/∂t, we can find intensity variation via phase shift
represent a well-known dualism, when a change in inter- Φ of backscattered light where there is argument of back-
ference can be considered not only as a result of a change scattering complex function:
in phase, but also as a beating of a frequency due to a
Doppler shift. The concept finds application in Doppler ∂I ∗
lidars, where Rayleigh scattering light contains wind = 2ΔΩ z r z r z − L0 sin ψ 0 + Φ (1.8)
∂t
speed information, so the height distribution of the speed
∗
can be detected using OTDR (Garnier & Chanin, 1992). Φ = ΔΩ z t + Arg r z r z − L0 (1.9)
The DAS conception is somewhat different: we do not The COTDR signal can be deduced from Equation 1.8
measure the absolute velocity of Rayleigh scatterers, if we set L0 = 0 and ψ 0 = 0. Even such a simple setup
but the difference in such velocity along the gauge length. can deliver information on the Doppler shift and
Another difference is that Rayleigh centers are frozen in a hence the ground speed v(z) through the intensity varia-
glass of fiber at a melting point of about 800 . Their move- tion ∂I/∂t Δv in accordance with Equations 1.3, 1.8.
ment follows the movement of the fiber, and hence very Unfortunately, the proportionality factor contains an
low Doppler frequencies (down to mHz) can be measured. oscillation term, so we cannot distinguish positive speed
For simplicity of further calculations, the reflective from negative.
coefficient r0(z) can be redefined as the effective reflective The result of computer modeling of a COTDR response
coefficient r(z): on a differential Ricker wavelet for ground speed (Hartog,
r z = r0 z exp β0 z (1.4) 2017) is presented in Figure 1.5. The right side shows 1D
seismic wave moving in the z direction (in m) with a reflec-
Then, to extract the Doppler shift from the intensity tion from an interface with a positive reflection coeffi-
equation, we need to control the phase shift ψ 0 between cient. Below the image is a time series of apparent
delayed optical fields in the interferometer. So Equa- velocity, when units are normalized to the expected opti-
tion 1.1 can be rearranged using Equations 1.2–1.4 to: cal phase shift in radians between points separated by
gauge length 10 m. The left side of the figure corresponds
E z, t = e z + e z − L0 exp iψ 0 r z exp i Ω z t
to the relative pulse-to-pulse variation of the COTDR sig-
(1.5) nal calculated in accordance with Equations 1.8–1.9. The
sign of response changes randomly in accordance with an
Here the convolution symbol is used to simplify the
optical pulsewidth of 50 ns or 5 m. As a result, the signal
expression, and the OTDR scale z = 2ct for the “fast”
cannot be effectively accumulated for multiple seismic
time t is used. The convolution commutes with
pulselosityes because of the temperature drift between
translations (Goodman, 2005), meaning that Equation 1.5
seismic shots. Temperature drift changes the phase con-
can be converted using a(z1 − z2) b(z1) = a(z1) b(z1 − z2)
stant of the fiber β0 and, in accordance with Equation 1.4,
to:
the effective reflection coefficient r(z) also changes. As a
E z, t = e z r z exp i Ω z t + r z − L0 result of such drift, every seismic shot will have a unique,
random, alternating, speckle-like signature that cancels
exp i Ω z − L0 t + ψ 0 (1.6) the averaging sum. Fortunately, this problem can be over-
come by optical phase recovery, when, after similar aver-
Let us consider first the simple case of short pulse aging, average values appear. Thus, the actual DAS
e(z) = δ(z) when δ is the Dirac delta function. Then con- output will be a combination of fiber speed information
volution can be removed from Equation 1.5 because and the unaveraged portion of the random COTDR
δ(z) a(z) = a(z), and the distance variation of Doppler signal.
8 DISTRIBUTED ACOUSTIC SENSING IN GEOPHYSICS
50 50
100 100
150 150
200 200
Depth, m
Depth, m
250 250
300 300
350 350
400 400
450 450
0 25 50 75 100 125 150 0 25 50 75 100 125 150
Time Time
0.03 0.03
Acoustic
Acoustic
0 0
–0.02 –0.02
0 25 50 75 100 125 150 0 25 50 75 100 125 150
Time Time
Figure 1.5 COTDR response (Equation 1.6) shown in the left panel of the simulated signal of a ground velocity
wavelet shown in the right panel. The signals’ cross-section along the white line is shown in the bottom panels
in radians. Source: Based on Correa et al. (2017).
1.1.3. DAS Optical Phase Recovery The same data can be extracted directly from phase infor-
mation, as is clear from Equation 1.11.
The randomness of the COTDR signal can be reduced So far, we have analyzed the short pulse case, where the
through proper control of the external interferometer pulsewidth is significantly smaller than the external inter-
phase shift ψ 0, which can be achieved in many ways. ferometer delay. In reality, such pulses cannot deliver sig-
All these methods are based on the fact that COTDR nificant optical power, which is necessary for precise
intensity is random in distance but will vary harmonically measurements. Fortunately, Equations 1.10–1.11 can be
depending on the phase, as follows from Equation 1.1 (see generalized for a nonzero length optical pulse e(z) directly
Figure 1.6). So, phase control can reveal phase informa- from Equation 1.5 in the same way that an optical inco-
tion regardless of the random nature of the signal. herent image was obtained in Goodman (2005) using cor-
We will start our phase analysis with a simple, although relation averaging (a r1)(a r2) = a2 r1r2 . This
not very practical, approach, where the phase shift ψ 0 is expression is valid for an uncorrelated field, generated by
locked onto a fringe sin(ψ 0 + Φ) ≡ 1. Such an approach random reflection points r1(z1)r2(z2) = δ(z1 − z2). This
was used earlier to analyze the spatial resolution in phase calculation confirms that Equation 1.11 remains the
microscopy (Rea et al., 1996). Then Equations 1.8 and 1.9 same, as it represents averaging over different harmonic
can be averaged over an ensemble of delta correlated signals, but Equation 1.10 will be reshaped to:
backscattering coefficients r(u)r(w) = ρ2δ(u − w) as:
∂I z, t 2
∂I z, t = 2ρ2 e z ΔΩ z (1.12)
= 2ρ ΔΩ z
2
(1.10) ∂t
∂t
Intensity, I(z,t)
Distance, z/L0
8
6
4
Phase shift,
2
𝜓0 + ΔΩ(z)t + Arg[r(z)r(z – L0)*]
0 π 2π
Figure 1.6 Intensity changes are irregular along distance but harmonic along phase shift axis.
2 2 2
1 1 1
DAS signal
DAS signal
DAS signal
0 0 0
–1 –1 –1
–2 –2 –2
Simulation Simulation Simulation
Averaging 100 Averaging 100 Averaging 100
–3 Theory –3 Theory –3 Theory
20 40 60 80 20 40 60 80 20 40 60 80
Distance, m Distance, m Distance, m
Figure 1.7 Comparison of DAS theoretical response (Equation 1.13) with simulation for a 3 × 3 coupler.
COTDR signal, which will be overlaid on the DAS signal Equation 1.18 gives an idea of this. If the phase is a
with nonzero pulsewidth. This is a natural limit for smooth function, we can differentiate in time Φ(t) before
increasing SNR by extending pulsewidth; we have a com- unwrapping. Then, the first differential linear term is
promise between SNR and signal quality at around removed, and condition becomes more relaxed:
L0 = 2τ. Finally, we can expect that the theoretical expres-
sion (Equation 1.13) can be used for spatial resolution − π ≤ Φ t + 2Δt − 2Φ t + Δt + Φ t < π (1.19)
analysis for different phase recovery algorithms after a
proper averaging. So, the second order tracking algorithm can be
obtained by differentiating the signal before unwrapping:
Re(Z), Im(Z)
0
–2
0 100 200 300 400 500 600 700
Time, samples
20
Acoustic, rad
Figure 1.8 Comparison of first and second order tracking algorithms for DAS.
1.1.5. DAS Signal Processing and Denoising using weighted spatial averaging (Farhadiroushan et al.,
2010). The maximum SNR is realized when the weighting
In all phase-detection schemes, the change in optical factor of each channel is chosen to be inversely proportional
phase between the light scattered in two fiber segments to the mean square noise in that channel (Brennan, 1959),
is determined, meaning we are measuring the determinis- meaning the squared interference visibility, V2, can be used
tic phase change between two random signals. The ran- for the weighting factor as:
domness of the amplitude of the scattered radiation
imposes certain limitations on the accuracy of the sensor, A z V2 z p z
through the introduction of phase flicker noise. The Az ≈ (1.21)
V2 z p z
source of flicker noise is an ambiguity: when the fiber is
stretched, the scattering coefficient varies, and can The averaging function p(z) = 5m should optimally be
become zero. In this case, the differential phase detector chosen to be compatible with the pulsewidth τ(z) = 50ns,
generates a noise burst regardless of which optical setup which should be around half the interferometer length
is used. The amplitude of such noise increases with L0 = 10m. With this width of the averaging function, it
decreasing frequency (as is expected for flicker noise) has no significant effect on the spatial resolution of the
when the phase difference is integrated into the displace- DAS. Modeling with and without weighted averaging is
ment signal. presented in Figure 1.9, which demonstrates that signifi-
From a quantum point of view, we need, for successive cant noise reduction can be achieved. It should be noted
phase measurements, a number of interfering photon that this noise reduction is particularly marked in compar-
pairs scattered from points separated by the gauge length ison with the coherent OTDR response, by contrasting
distance. In some “bad” points, there are no such pairs, as with Figure 1.5. Nevertheless, weighted averaging sup-
one point of scattering is faded. A natural way to handle presses rather than completely removes the effect of
this problem is to reject “bad” unpaired photons by con- flicker noise, and some channels still demonstrate exces-
trolling the visibility of the interference pattern. As a sive noise (in addition to shot noise). Hence, the response
result, the shot noise can increase slightly as the price over all depths at a given time for Figure 1.9 will contain
for the dramatic reduction of flicker noise. The rejection spikes for faded channels.
of fading points can be practically implemented by assign- As is explored in Section 1.3, the problem of flicker
ing a weighting factor to each measurement result and noise can be overcome by introducing engineered bright
performing a weighted averaging. scatter zones along the fiber with constant spatial separa-
This averaging can be done over wavelength if a multi- tion and uniform amplitude. Such scatter zones also
wavelength source is used. Alternatively, we can slightly sac- reflect more photons, and so improve the shot noise detec-
rifice spatial resolution and solve the problem by denoising tion limitation. In addition, the use of such engineered
12 DISTRIBUTED ACOUSTIC SENSING IN GEOPHYSICS
50 50
100 100
150 150
200 200
Depth, m
Depth, m
250 250
300 300
350 350
400 400
450 450
0 25 50 75 100 125 150 0 25 50 75 100 125 150
Time Time
0.03 0.03
Acoustic
Acoustic
0 0
–0.02 –0.02
0 25 50 75 100 125 150 0 25 50 75 100 125 150
Time Time
Figure 1.9 The left-hand panel shows modeling of raw DAS acoustic data (Equation 1.12); the right-hand panel
shows the same shot with weighted averaging denoising (Equation 1.13) applied. The signals’ cross-section
along the white line is shown in the bottom panels in radians. The modeled source is shown in the right panel
of Figure 1.5.
meaning a time integrated DAS signal can be considered 1.2. DAS SYSTEM PARAMETERS AND
as an output of a huge caliper that is measuring fiber elon- COMPARISON WITH GEOPHONES
gation between two points with sub-nanometer precision.
This measuring principle is different from that of a geo- In this section, we consider how DAS parameters (such
phone but is similar to an electromagnetic linear strain as spatial resolution), gauge length, frequency response,
SEISMIC DATA FROM DISTRIBUTED AND ENGINEERED FIBER OPTIC ACOUSTIC SENSORS 13
and SNR enable DAS to become an effective tool for seis- is represented (Equation 1.5) as a convolution of a point
mic measurements. Field data are also presented, with spread function with v(z).
DAS output compared to geophone data. Spatially integrated signal (Equation 1.27) was mod-
eled for 10 m gauge length and 50 ns pulsewidth, as shown
in Figure 1.5 (right panel). The results of modeling
1.2.1. DAS Optimization for Seismic Applications (Equation 1.25) are presented in Figure 1.11 (left panel),
and the result is converted to geophone-style data
Distributed fiber sensors measure physical parameters (Equation 1.26) in the right panel. From a practical point
of an external environment continuously through the inte- of view, low temporal frequencies, out of the range of
gration properties of light traveling along a lengthy opti-
interest, can be filtered out, and also spatial antialiasing
cal path. This is quite different from point sensors, such as filtering can be used. It is worth mentioning that the right
geophones, which make an inertial measurement of panel of Figure 1.11 demonstrates the real change in
ground speed at fixed positions (SEAFOM, 2018). The
polarity of the reflected seismic pulse. Also, spatial inte-
DAS records a local strain rate, which can be converted gration (Equation 1.26) acts as statistical averaging,
into particle velocity to allow direct comparison with geo- which eliminates the randomness of the “staircasing” in
phone data. Following Jousset et al. (2018), we can
Figure 1.5 left panel.
approximately represent DAS signal A(z, t) via ground The most valuable geophysical information is delivered
displacement u(z, t), where FS is the DAS sampling fre- by sound waves with frequencies below FMAX = 150Hz, as
quency and L0 is the gauge length.
higher frequencies are attenuated by the ground. For a
A z, t u z, t + 1 F S − u z, t speed of soundC = 3000m/s, this corresponds to an acous-
(1.23) tic wavelength C/FMAX = 20m, so Nyquist’s limit dictates
− u z − L0 , t + 1 F S − u z − L0 , t
that LG ≤ C/2FMAX = 10m is the maximum spacing of
If FS ∞, L0 0, then the DAS signal can be pre- conventional sensors. Formally, the linear spline approx-
sented in a double differential form: imation G(z) of conventional antenna velocity v(z) output
can be represented using expressions from (Unser,
∂ ∂ ∂
A z, t u z, t = v z, t (1.24) 1999), as:
∂z ∂t ∂z
These simplified expressions (Equations 1.23–1.24) give G z = θ z + LG − θ z θ z + LG − θ z
us a qualitative sense of the DAS algorithm output. For a comb z LG vz (1.28)
subsequent quantitative analysis, we shall need the
detailed expression that was obtained in the previous sec- The spatial spectral response of DAS in acoustic angu-
tion. Namely, for a nonzero interferometer gauge length lar wavenumber Kz can be represented by Fourier trans-
L0 and optical pulsewidth τ, averaged over random scat- form ℑ(Kz) following Goodman (2005):
tering DAS output, A(z) can be represented by Equa- ℑG K z = sinc K z LG 2 comb K z LG 2π ℑ Kz
tion 1.15 in expanded form: (1.29)
1
Az = τ z δ z − δ z − L0 vz (1.25) Such spectral responses can be normalized for a con-
A0 F S stant signal ℑ(K) = 1 (see black line in Figure 1.12).
where FS is sampling frequency and A0 = 115nm is a scale The comb function in (Equation 1.29) is responsible for
constant (Equation 1.14). So, the velocity field can be the repeating of the spatial spectrum with a shift of 2π/
recovered by spatial integration starting from a motion- Λ, as is shown by the dotted line. To prevent aliasing,
less point as: the signal spectrum should be inside Nyquist’s limit,
z
which is shown by the gray vertical line.
Let us compare the conventional velocity sensor with
Az = A u du = A z θ z (1.26) the DAS spectrum, calculated from the spatial resolution
0 expression (Equation 1.25), by Fourier transform as:
Then DAS signal (Equation 1.25) can be transformed ℑ A Kz = sinc K z τ 2 sin K z L0 2 ℑ Kz (1.30)
using shift invariant a(z1) b(z1 + z2) = a(z1 + z2)
b(z1) to: Two cases are presented in Figure 1.12: when the opti-
cal pulse length is almost equal to the interferometer
1
Az = τ z θ z + L0 − θ z vz (1.27) gauge length τ = L0, and when it is half the interferometer
A0 F S gauge length τ = L0/2 (see dashed and solid blue lines,
where θ(z) is the Heaviside step function, whose value is respectively). The absolute value is presented in the figure
zero for a negative argument. As expected, the DAS signal to aid comparison between curves. In the second case, we
14 DISTRIBUTED ACOUSTIC SENSING IN GEOPHYSICS
50 50
100 100
150 150
200 200
Depth, m
Depth, m
250 250
300 300
350 350
400 400
450 450
0 25 50 75 100 125 150 0 25 50 75 100 125 150
Time Time
0.03 0.03
Acoustic
Acoustic
0 0
–0.02 –0.02
0 25 50 75 100 125 150 0 25 50 75 100 125 150
Time Time
Figure 1.11 Acoustic measurements using DAS: The left panel represents strain rate measurement and the right
panel displays ground speed measurement, the transform to which comprises filtering and integration. The
signals’ cross-section along the white line is shown in the bottom panels in radians. The modeled source is
shown in the right panel of Figure 1.5.
10 m aliasing
1.0 DAS τ=5m Lo=10m
DAS τ=10m Lo=10m
Sensors array 10m
0.8 Sensors array aliasing
Spectal responce
0.6
0.4
0.2
0
0.05 0.10 0.15 0.20
Acoustic wavenumber, K/2π, 1/m
Figure 1.12 Comparison of DAS spectral response with that from a 10 m sensor antenna array.
have a gain, which is highlighted by the blue filling. This Practically, however, this is not the case, as the geophone
gain can be explained by signal smearing over a noise rises at low frequencies, and this can be character-
long pulse. ized by some high-pass (HP) filters that limit the range
It seems from Figure 1.12 that DAS low frequency sen- to frequencies around 10 Hz (see dotted line in
sitivity is significantly lower than that of a geophone. Figure 1.13). However, DAS has the potential to increase
SEISMIC DATA FROM DISTRIBUTED AND ENGINEERED FIBER OPTIC ACOUSTIC SENSORS 15
1.4
10 Hz HP DAS Ns =5m Lo=30m
DAS Ns=5m Lo=10m
1.2
Geophones array 10m
Geophones array HP
1.0
Spectal responce
0.8
0.6
0.4
0.2
0
0.01 0.02 0.03 0.04 0.05
Acoustic wavenumber, K/2π, 1/m
Figure 1.13 Low spatial frequency gain in DAS by using long interferometer.
the spectral response at low frequencies by increasing respect to incident angle Γ can be found by transformation
interferometer length—for example, from L0 = 10m to of the strain tensor components with rotation using geo-
L0 = 30m (see Figure 1.13). So, potentially, the DAS metrical consideration. For a longitudinal (P) apparent
response can be synthesized from two measurements: with wave, it will be cos2Γ, and for transversal (S) wave sin
a short interferometer gauge length to deliver high spatial Γ cos Γ, similar to Benioff (1935) (see Figure 1.14).
frequency bandwidth, and a long one to deliver low fre- Detailed analysis and diagrams for Rayleigh and Love
quency. As the result, full-frequency coverage can be as waves can be found in Martin et al. (2018).
good as from a geophone antenna, or possibly even better, In vertical seismic profiling (VSP), in the vertical part of
as will be shown in a later SNR comparison. An addi- the well, both cable and seismic waves are in the same
tional advantage over geophones is the large dynamic direction for near-offsets, so the DAS is more sensitive
range of DAS at low frequencies, which will be discussed to P-waves, in which the acoustic displacement vector
later. coincides with the fiber direction. In other applications,
such as fracking, the microseismic source is usually on a
side of the cable, so shear waves can be effectively
1.2.2. DAS Directionality in Seismic Measurements
detected.
In the previous section, we analyzed the correspondence Cable orientation is responsible not only for acoustic
between DAS and geophones in the one-dimensional case amplitude but also for acoustic spatial resolution, even
and found that “geophone-style” velocity data can be for the same acoustic wavelength. The cable acts as an
extracted from DAS signals by spatial integration. How- acoustic antenna where the signal varies rapidly in space
ever, in 3D analysis, we should consider that DAS is not a if the P-wave and cable direction coincide, but the signal
velocity sensor but a differential strain sensor. This is a remains the same over distance if the acoustic wave front
fundamental difference: DAS can measure a component is parallel to the cable. To take this effect into considera-
of 3D tensor (strain) but not 3D vector (velocity). tion, we need to expand the expression for acoustic wave-
Directionality of the DAS response depends on the fiber length Kz along the cable for Equations 1.29–1.30 as:
optic cable configuration and the cable design, as the 2πF
device itself is sensitive only to fiber elongation. We will K z = K cos Γ = cos Γ (1.31)
C
start our consideration where the fiber is placed linearly
inside a cable, with no slippage between fiber and cable, For a harmonic wave, directionality will directly affect
nor between the cable and the ground. In this case, fiber not only the spatial resolution but also the temporal fre-
displacement will follow ground displacement, and sensi- quency. After Fourier transfer in the time domain, Equa-
tivity will depend on the relative position of fiber and seis- tion 1.30, in the absence of aliasing, can be presented as:
mic source. A similar mechanical principle was used for
the electromagnetic linear strain seismograph to measure ℑA K z , F = sinc K z L0 4 sin K z L0 2 ℑ K z, F
variations in the distance between two points of the (1.32)
ground (Benioff, 1935). DAS directional response with
16 DISTRIBUTED ACOUSTIC SENSING IN GEOPHYSICS
VSP
source
┌
180 0
315
225
270
Figure 1.14 DAS with linear optical cable is more sensitive to P-wave in VSP configuration and to S-wave in
microseismic events.
700 700
F = C K/2π, C = 3000m/c
600 600
500 500
Frequency, F, Hz
Frequency, F, Hz
400 400
300 300
F=C/Lo
200 200
100 100
K/2π=1/Lo
0 0
–0.2 –0.1 0 0.1 0.2 –0.2 –0.1 0 0.1 0.2
Acoustic wavenumber, K/2π, 1/m Acoustic wavenumber, K/2π, 1/m
Figure 1.15 2D spectral representation on upgoing and downgoing acoustic waves: The left panel represents
original signal and the right panel represents filtered signal, where the spatial filter shape is shown by a blue
wavy line.
This expression (Equation 1.32) represents spatial filter- It is interesting to mention that, for uniform strain,
ing in the 2D Fourier domain as shown in Figure 1.15 for where C ∞, we have |ℑA(F)| cos2Γ as expected from
L0 = 10m for upgoing and downgoing waves, together Benioff (1935). The result of modeling of Equation 1.33 is
with white phase noise. As far as harmonic waves can presented in Figure 1.16 for different incident angles. An
be represented as single lines, the result of spatial filtering increase in angle expands the measurement frequency
is an intensity modulation of these lines. Such modulation range but reduces low-frequency SNR at the same time.
is equivalent to temporal frequency modulation, so we DAS directivity can be significantly modified through
can combine Equation 1.31 and Equation 1.32 to get: appropriate cable design, which is currently a developing
area. For example, field tests have shown that the helical
cos Γ L0 cos Γ L0
ℑA F = sinc πF sin πF ℑF placement of fiber within a squeezable material can
2C C deliver omnidirectional sensitivity (Hornman et al.,
(1.33) 2013) for P-waves. The angular dependence will be
SEISMIC DATA FROM DISTRIBUTED AND ENGINEERED FIBER OPTIC ACOUSTIC SENSORS 17
1.0
0° Incidence
0.8 45° Incidence
0.6
SNR
0.4
0.2
0
0 50 100 150 200 250 300 350 400 450 500 550 600
Frequency, F, Hz
Figure 1.16 Normalized SNR curve (SNR vs. frequency) for a 3000 m/s wave speed of P-wave at 0 and 45
incident angles.
different for S-waves (Abbott et al., 2019), and a different the clamps, then lines on the VSP has a staircase-like
helix pitch can be useful to optimize performance for dif- shape. The period of the steps equals the distance between
ferent wave types (Baird, 2020). Additional complexity the clamps, which is about 10 m. Fortunately, this value
comes from cable construction, and the Poisson s ratio does not significantly exceed the DAS resolution and
of the cable itself can affect the angular signature practically does not degrade the quality of the VSP
(Wuestefeld & Wilks, 2019). An even more sophisticated pattern.
approach can be used to measure inertial acceleration— A typical VSP seismic shot response for permanently
by using a dedicated non-isotropic cable, where a dense installed fiber optic cable behind the casing is shown in
mass compresses the fiber along the cable Figure 1.18 for both the raw acoustic data and with the
(Farhadiroushan et al., 2017). Such a solution can be used denoising algorithm applied.
for multi-component seismic acquisition, including for An important practical question is the ability of the
analysis of microseismic events. DAS to perform measurements on both single-mode
This analysis demonstrates that the DAS broad spectral (SM) and multi-mode (MM) optical fiber, since MM fiber
response can potentially correspond to conventional geo- has been deployed in many legacy installations. It was
phones and seismometers. In the next section, we will pro- found experimentally that seismic data can be recorded
vide some examples of how such promises can be fulfilled equally well on both SM and MM fiber. This is achieved
in field measurements. as the fundamental mode LP01 size diameter in MM fiber
(14 μm) is nearly matched to SM fiber (10 μm), and, there-
1.2.3. DAS Field Data Examples fore, the DAS performance using MM fiber is similar to
that from SM fiber. Strictly speaking, the SNR in MM
DAS seismic services were introduced to deliver better fiber can be slightly worse at the near end of the fiber
characterization of geophysical properties by dramati- because of the optical coupling loss, and slightly better
cally increasing the spatial density of acquired data. at the far end of the fiber because its larger core diameter
DAS technology enables the collection of seismic data allows higher optical power transmission along the fiber.
with a wide range of source types. For in-well measure- Similar performance for SM and MM fiber was observed
ments, the optical fibers are embedded within ruggedized in field experiments (see Figure 1.19) when two fibers were
downhole cables that can be conveyed loosely in the well placed side by side in an optical cable. These results show
(wireline), or clamped to tubing and/or cemented with the the feasibility of retrofitting DAS to existing MM fiber
completion, thereby offering a permanent sensor array installations and so utilizing distributed temperature sen-
(Figure 1.17). The usual assumption is that the stretching sing infrastructure to perform the full scope of DAS ser-
of the optical cable coincides with the deformation of the vices, which include not only seismic measurements but
ground in the acoustic wave. In turn, the length of the also well diagnostics and flow monitoring (Finfer
optical fiber tracks the length of the optical cable due to et al., 2014).
internal friction. When the cable is attached to a pipe, For VSP applications in vertical wells, the direction of
the pipe deformation coincides with the ground deforma- the well and fiber optic cable coincides for near-offsets
tion. If the cable is poorly connected to the pipe between with the seismic wave propagation, and so DAS is mostly
18 DISTRIBUTED ACOUSTIC SENSING IN GEOPHYSICS
300 300
Channel index
Channel index
600 600
900 900
1200 1200
Figure 1.18 The left-hand panel shows a single shot of raw acoustic data; the right-hand panel shows the same shot
with denoising applied from Miller et al. (2016).
sensitive to P-waves. This effect was tested by comparison seismic signal at every point along the optical fiber with
with transverse and vertical geophones (see Figure 1.20). each source activation, leading to much greater receiver
The geophone with transverse orientation (left panel) has coverage than is achievable with conventional borehole
not detected the P-wave, whereas the geophone with ver- seismic methods. A typical result of DAS 3D VSP is pre-
tical orientation (central panel) has. The DAS (right sented in Figure 1.21.
panel) also has detected the P-wave as expected. It is Fine spatial resolution, in combination with good sen-
worth noting that the case for far offsets is more compli- sitivity and dynamic range, gives DAS a significant
cated (Mateeva et al., 2014). advantage for hydraulic fracture monitoring and the
A comparison of DAS data (converted from strain rate detection of microseismic events, particularly where a
to particle velocity) to co-located geophones indicates that geophone chain cannot be readily positioned, such as in
the DAS data is consistent with geophone response. As a a treatment well. Figure 1.22 shows a waterfall plot (depth
result, a 3D VSP image can be collected from multiple vs time), recording strong acoustic signals, corresponding
dynamite shots in a similar manner as for conventional to fluid placement across individual clusters, while, at the
geophones (Miller et al., 2016). The DAS records the same time, detecting small microseismic events.
SEISMIC DATA FROM DISTRIBUTED AND ENGINEERED FIBER OPTIC ACOUSTIC SENSORS 19
Mulitimode Singlemode
500 200
400
1000
600
1500
800
2000
1000
2500
1200
3000 1400
0 500 1000 1500
500 1000 1500
(msec) (msec)
VERTICAL
Depth (m)
DAS
P-wave
P-wave
900
0.0 1.0 2.0 0.0 1.0 2.0 0 500 1000 1500 2000
TIME (S) TIME (S) (msec)
Figure 1.20 Directionality of DAS response: The left and central panels represent geophones with transverse and
vertical orientation, and the right panel represents the DAS signal for VSP.
Optical fiber can also be used for offset well monitoring, geophones and can offer the benefits of wide aperture
as indicated in Figure 1.23. Here, the optical fiber cable, monitoring along the entire borehole with broad fre-
cemented behind the casing in the originally treated well, quency response. Improvements in optical fibers and
is used to monitor microseismic events and strain while an cable designs offer new possibilities for the DAS monitor-
offset well is being treated. The shape and arrival time of ing of geophysical properties.
P- and S-waves can be used for microseismic event picking
and localization. The data can be used for optimizing the
well spacing, cluster spacing, and stimulation parameters. 1.3. DAS WITH PRECISION ENGINEERED FIBER
In summary, DAS is a new, versatile technology that
can be deployed in many different configurations along In this chapter, we consider how the scattering proper-
boreholes where geophones cannot readily be deployed. ties of conventional fiber can be engineered to deliver bet-
The frequency response of DAS is comparable with ter DAS performance (Figure 1.24). We will show how an
20 DISTRIBUTED ACOUSTIC SENSING IN GEOPHYSICS
2934 m
500
1000
Depth (m)
1500
2000
2500
3000
1000
500
0 1000
–500 500
0
(m) –1000 –1000 –500 (m)
Figure 1.21 3D VSP: Two intersecting images processed from DAS seismic data acquired with the dynamite shot
positions indicated from Miller et al. (2016).
Pump
Acoustic,
17,000 nm
50
To DAS 16,000
Flow 40
noise 15,000
14,000 30
Depth, ft
13,000 20
12,000
10
11,000
0
Fracking
event 10,000
–10
0 0.05 0.10 0.15 0.20
Time, s
Figure 1.22 DAS hydraulic fracture monitoring in the treatment well with a fine spatial resolution and wide
dynamic range for simultaneous cluster fluid allocation and microseismic monitoring.
SNR improvement can be achieved, along with a wider but higher scattering fiber to generate more light. These
dynamic range, using engineered fiber with precisely uni- two apparently contradictory requirements can be bal-
form scattering centers. This approach differs from a sim- anced by engineering bright scatter centers in the fiber,
ple increase in irregular backscattering intensity without introducing significant excess loss for the forward
(Westbrook et al., 2017). We also consider the trade-offs propagating light. This can be achieved, for example, by
between spatial resolution, signal-to-noise performance using fiber Bragg grating technology.
and frequency response, and present data acquired from For long fiber lengths, 100 times more light than Ray-
several different seismic and microseismic surveys. leigh level can be safely used (Farhadiroushan et al.,
DAS performance is largely governed by how much 2021). That gives 20 dB reduction of acoustic noise caused
light can be usefully collected from the optical fiber. In by quantum shot noise at frequencies of around 1 kHz.
general, we require low-loss fiber for long range sensing, This improvement can be even more at low frequencies
SEISMIC DATA FROM DISTRIBUTED AND ENGINEERED FIBER OPTIC ACOUSTIC SENSORS 21
5200 5200
Fracking
5300 5300
event
Depth, m
5400 5400
5500 5500
5600 5600
5700 5700
5800 5800
5900 5900
Figure 1.23 DAS hydraulic fracture monitoring in the offset (a) with a fine spatial resolution for microseismic
monitoring (c), and localizing of microseismic events in time and space (b).
Standard fiber
Engineered fiber
Figure 1.24 DAS with standard fiber and engineered fiber with precision bright scatter center zones.
as pink noise is suppressed by the regular structure of scat- where e(z) is a coherent optical pulse and Ω(z) v(z) is the
tering. So, noise reduction can be more than 30 dB at Doppler shifted angular frequency, which is proportional
around 1 Hz. This prediction was successfully confirmed to the local acoustic speed—see Figure 1.25.
in field surveys and are presented at the end of the chapter. The scattering coefficient for engineered fiber can be repre-
sented by a spatially periodic function (Farhadiroushan
et al., 2021), meaning a reflection coefficient r(z) can be repre-
1.3.1. Precision engineered fiber concept sented by a set of defined scatter center zones separated by
sampling distance LS.
We will start our consideration from Equation 1.6 in
M
Section 1.1 (titled ‘Distributed Acoustic Sensor (DAS)
r z =R δ z − jLS = R comb z LS (1.35)
Principles and Measurements’), which represents the scat- j=0
tered E(z, t) field as a convolution of input optical field
with scattering coefficient r(z), for a gauge length L0. where comb(z ) is the Dirac comb function, or sampling
operator. If the gauge length is s times larger than
E z, t = e z r z exp i Ω z t + r z − L0 sampling distance, L0 = s LS, s = 1, 2…, then r(z) =
exp i Ω z − L0 t + ψ 0 (1.34) r(z − L0), and the reflectivity function r(z)can be taken
out of the brackets:
22 DISTRIBUTED ACOUSTIC SENSING IN GEOPHYSICS
v(z + 2Ls)
v(z)
v(z + Ls)
τ (z)
z z + Ls z + 2Ls
Figure 1.25 Optical fiber with defined scatter center zones and the corresponding Doppler shifted angular
frequency sampled between the zones. The length occupied by optical pulse is less than the distance between
the zones. The gray line corresponds to spatially integrated DAS output, following a linear spline approximation.
E z, t = R e z exp i Ω z t + exp i Ω z − L0 t + ψ 0 The main parameter for spatial resolution is still the gauge
length L0, and the sampling distance can be chosen to have
comb z LS (1.36) two points per gauge length LS = L0/2. We are considering
To prevent cross-interference and fading, the spatial here the physical spatial sampling, which is defined by the
length of the optical pulse should be smaller or equal to optical configuration, keeping in mind that the photocur-
the distance between scatter center zones, so the spatial rent sampling can have a higher rate. The difference from
sampling of the optical field (Equation 1.35) can be repre- conventional fiber is an absence of averaging, as the detected
sented by a train of pulses: signal is deterministic for engineered fiber, and excessive
noise from non-averaged components will hence disappear.
M
Also, the generated optical field can be significantly larger
E z, t = R e z − jLS exp i Ω jLS t
j=0
than with conventional Rayleigh backscattering, so the shot
+ exp i Ω jLS − L0 t + ψ 0 (1.37) noise limitation can be reduced significantly.
The velocity field can be recovered by spatial integra-
The optical pulses from each zone are separated tion starting from a motionless point as:
(see Figure 1.25), so the maximum signal intensity and z
maximum SNR can be delivered if the pulsewidth is equal
to the sampling distance, or τ(z) = θ(z + LS) − θ(z), Az = A u du = A z θ z (1.40)
where θ(z) is the Heaviside step function whose value is 0
0 for negative argument and 1 for positive argument. In So Equation 1.39 can be transformed to:
this case, intensity can be calculated from the interference
between pulses with the same index j, and, for each pulse, 1
Az = θ z + LS − θ z θ z + L0 − θ z
an acoustic signal A(z) = F ∂Φ/∂t, where Φ = ΔΩ(z)t can A0 F S
be recovered from Equation 1.34 using A0 from Equa- comb z LS v z
tion 1.14 as: (1.41)
50 50
100 100
150 150
200 200
Depth, m
Depth, m
250 250
300 300
350 350
400 400
450 450
0 25 50 75 100 125 150 0 25 50 75 100 125 150
Time Time
0.03 0.03
Acoustic
0 Acoustic 0
–0.02 –0.02
0 25 50 75 100 125 150 0 25 50 75 100 125 150
Time Time
Figure 1.26 Acoustic measurements using DAS with precision engineered fiber: The left panel represents strain rate
measurement (Equation 1.39) and the right panel displays ground speed measurement (Equation 1.41) after filtering
and integration. The signals’ cross-section along the white line is shown in the bottom panels in radians. The
modeled source is shown in the right panel of Figure 1.5.
out, and also spatial antialiasing filtering can be used. It is spectral density is zero for maximum frequency, seen by
worth mentioning that the right panel of Figure 1.26 is comparing the position of the black and gray vertical lines
very similar to the original pulse (Figure 1.5), which in Figure 1.27. This advantage can explain the absence of
demonstrates the real change of polarity of the reflected “staircasing” and the smooth output in Figure 1.26 right
seismic pulse. Compared with Figure 1.10 (conventional panel. An additional advantage of high sampling is that,
fiber), Figure 1.26 shows better SNR and signal amplitude for a typical L0 = LG = 5m, the sampling is twice or even
stability than with conventional fiber, and a more uniform three times smaller than the sensor separation in a geo-
size of the step in the “staircase” in the left panel, which phone array. This spatial frequency margin is useful
can be easily filtered out. because DAS timing is different from analog geophones.
The spatial spectral response in the wavenumber For a geophone antenna, we can filter out high-frequency
domain Kz can be represented by Fourier transform ℑ: space-time components in the time domain by electrically
filtering individual channels before sampling to prevent
ℑG K z = sinc K z LS 2 sin K z L0 2 spatial aliasing. This approach is ineffective for DAS
when the time sampling acts directly on the rapidly chan-
comb K z LS 2π ℑ Kz (1.42)
ging photocurrent. The problem can be solved for DAS by
where ℑ(Kz) is the spatial spectral response of the seismic mechanical filtering in the acoustic area using a special
wave. Comparisons of DAS with engineered fiber spectral design of the sensing cable, as in Carroll & Huber
response for spatial sampling equal to the gauge length (1986). An alternative approach involves some oversam-
and half of gauge length are presented in Figure 1.27 pling in the spatial domain, and the result is not com-
based on Equation 1.41. For the high spatial sampling, pletely independent. Subsequent filtering then removes
we have a gain in the frequency range, which is high- high spatial frequencies and prevents aliasing.
lighted by the gray filling. Moreover, it is easy to filter Finally, we can neglect the comb function in Equa-
out the aliased component for high sampling as the tion 1.42, following which Equation 1.42 is exactly
24 DISTRIBUTED ACOUSTIC SENSING IN GEOPHYSICS
0.5
0
0 0.05 0.10 0.15 0.20
Acoustic wavenumber, K/2π, 1/m
Figure 1.27 Comparison of DAS with engineered fiber spectral response for special sampling equal to gauge length
(black) and half of gauge length (gray).
equivalent to the expression for a conventional fiber expressed in Equation 1.43. A simple practical implemen-
(Equation 1.30) with pulsewidth equal to the scattering tation for optimizing both low and high spatial frequency
period LS = τ = 5m. can be realized by sliding a leaky distance integration of
DAS with engineered fiber combines the benefits of a DAS signal similar to how it was done for velocity recov-
distributed sensor, giving full coverage, with the high sen- ery (Equation 1.41).
sitivity of point sensors such as geophones. The scatter The ultimate spectral response of DAS with standard
centers are precisely engineered along the length of the (Equation 1.30) and engineered (Equation 1.43) fiber
fiber and not distributed randomly as for standard fiber compared to that from a geophone array is shown in
(see Figure 1.28). This allows the backscattered signal Figure 1.28. The pulsewidth of the DAS is the same as dis-
to be downsampled precisely and optimum spectral tance between scatter centers in engineered fiber
response to be obtained. τ = LS = 5m, and the gauge length is the same as the dis-
The DAS signal with engineered fiber, as expressed in tance between geophones LG = L0 = 10m. In summary,
Equation 1.39, can be considered as a staircase function downsampling of the DAS signal with engineered fiber
with differential velocity sampling LS: when sampled over can improve the spectral response as compared to stand-
each staircase distance LS, the expression in the square ard fiber with the same gauge length. However, DAS with
brackets will be eliminated from Equation 1.39, and, standard fiber can provide a wide spectral response with-
therefore, the corresponding sinc function in Equa- out aliasing, as is shown in Figure 1.28.
tion 1.43 will also be eliminated. As a result, the DAS
signal with engineered fiber will be defined by (v(z) − 1.3.2. Sensitivity and Dynamic Range
v(z − L0)), or comb filters in the spectral domain:
sin j K z L0 2 DAS sensitivity can be calculated for a fundamental
Specral Responce = MAX (1.43) limit—the shot noise generated by the number of photons
j j
detected. Let us estimate the photon number N per second
Equation 1.43 also includes a gain that can be obtained based on input peak power P0 = 1 W, which is near to the
from synthetic gauge length optimization. With this maximum optical connector power damage threshold (De
approach, low spectral frequencies can be measured by Rosa, 2002). The backscattered intensity can be found
adding a few consecutive downsampled signals. From a from the typical scattering coefficient for SM fiber
physical point of view, it means that the combination of RBS = 82dB for a 1 ns pulse (Ellis, 2007). For an optical
multiple gauge lengths L0 can be used to form a single pulsewidth τ = 50ns, the energy quant for λ = 1550nm is
long gauge length. The SNR for the resultant gauge length hυ = 1.28 10−19 J. We consider a relatively short fiber
j L0 will decrease proportionally to j in a shot noise lim- length, L = 2000m, to neglect nonlinear effects (Martins
ited DAS—see denominator in Equation 1.43. High spa- et al., 2013) and suppose that light is collected over an
tial frequencies can still be measured with original gauge integration length LP = 5m:
length L0 without any loss of spatial resolution. Poten-
P0 τRBS LP
tially, we can maximize the spectral response by choosing N= = 6 109 s − 1 (1.44)
a proper averaging factor j for any spectral band, as is hν L
SEISMIC DATA FROM DISTRIBUTED AND ENGINEERED FIBER OPTIC ACOUSTIC SENSORS 25
Geophones antenna
LG=10 m aliasing LS=5 m aliasing
1.0
Spectral responce
DAS with standard fiber 0.6
0.4
τ
0.2
DAS with engineered fiber
0
0 0.05 0.10 0.15 0.20
z z + LS z + L0 z + 2L0
Acoustic wavenumber, K/2π, 1/m
Figure 1.28 Ultimate SNR spectral response of DAS with standard and engineered fiber and geophone antenna.
Pulse width of DAS is the same as distance between scatter centers along engineered fiber—5 m, and gauge
length of DAS is the same as distance between geophones—10 m.
The shot or Poisson noise limit for phase measurement conventional Rayleigh backscattering, so the noise will
Φmin is proportional to 1 N , where the coefficient be 10 times smaller.
depends on the phase-detection approach. For a classical Another advantage of DAS with engineered fiber is a
phase-locked homodyne, only half of the photons reach wider dynamic range that is defined as the ratio of the
the photodetector when the interferometer operates in maximum detectable signal to the noise level. The typical
quadrature, and so the noise is 2 N. For both hetero- geophone bandwidth is ΔF = 100Hz, so the minimum
dyne and/or homodyne phase detection, the photons strain level εmin detectable for DAS for gauge length
number halves again (Kazovsky, 1989), as sine and cosine L0 = 10m within the same detection bandwidth is:
signal components should be measured independently, Φmin A0 ΔF
and so the noise rises to 4 N . Direct photodetection εmin = 0 01nanostrain (1.46)
L0
at λ = 1550nm is not sufficiently sensitive, so DAS usually
uses an erbium doped fiber amplifier (EDFA) to boost the where A0 = 115nm is the elongation corresponding to one
signal, which introduces additional noise. This noise can radian phase shift (Equation 1.14).
be simply represented by a noise figure NF ≈ 3, which Experimental measurements with conventional fiber
can be reached with appropriate optical filtering as DAS found a value three times higher, at 0.03nanostrain
explained in Kirkendall & Dandridge (2004). In this case, (Miller et al., 2016). In this case, there was some extra
the shot noise limit is then given by: flicker noise, as discussed earlier (see Figure 1.11). Here,
a spiky noise structure corresponds to algorithm disconti-
nuities that amplify photodetector noise, with a spectrum
Φmin =
1 4N F
10 − 4 rad Hz (1.45) after DAS signal time integration, which is F−1. The typ-
V N ical low frequency limit when excessive noise starts to
dominate over shot noise is between 10 and 100 Hz,
where visibility, V = 0.5, includes all other system imper- depending on the fiber conditions.
fections such as polarization mismatch. Equation 1.45 For engineered fiber (Farhadiroushan et al., 2021),
represents the white noise level for 1 second time integra- reflectivity can be engineered to be hundreds of times higher
tion of the DAS signal. For engineered fiber, the number than the normal Rayleigh level, without any significant pro-
of photons can be up to 100 times larger than for blems with crosstalk, such that R = 100 RBS τ = − 45dB.
26 DISTRIBUTED ACOUSTIC SENSING IN GEOPHYSICS
As a result, sensitivity is ten times higher, at around 1pico- detectors have limited environmental isolation
strain, which corresponds to a 100x (20 dB) improvement (Matichard et al., 2015); DAS can be potentially used
in acoustic signal sensitivity. for such applications.
It is important to compare the shot noise level of DAS We now turn our attention to the increase in dynamic
with the noise level of high-sensitivity geophones and range achievable using DAS with an engineered fiber.
seismometers. The DAS white noise value should be The acoustic algorithm transforms DAS intensity signals
added to flicker noise with coefficient μ and corrected into a phase shift proportional to the fiber elongation
for spatial filtering (Equation 1.46) as: value. The algorithm is based on an ambiguous function
such as ATAN(x), which give a valid result only inside a
Φmin A0 + μF − 1 limited region (Itoh, 1982). As was analyzed in Section 1.1
zmin = (1.47)
sinc πFL
2C
0
sin πFL
C
0
(titled ‘Distributed Acoustic Sensor (DAS) Principles and
The comparison in Figure 1.28 demonstrates that DAS Measurements’), a set of different algorithms can be used,
sensitivity is compatible with geophones. The noise spec- depending on the order of phase tracking. For the first and
trum data for Sercel SG5-SG10 was adapted from Fou- second order, we have:
gerat et al. (2018), and the seismometer Streckeisen − π ≤ A1 t < π (1.48)
STS-2 data from Ringler & Hutt (2010) and Wielandt &
1
Widmer-Schnidrig (2002). − π ≤ A2 t + − A2 t < π (1.49)
The sensitivity of DAS can even be improved at low fre- Fs
quencies by extending the gauge length from L0 = 10m to For limits (Equations 1.48–1.49), it is clear that the
L0 = 30m, but at the cost of increased noise at frequencies maximum recoverable strain ε1,2 will depend on the algo-
of more than 70 Hz. Also, 30 m data for DAS with engi- rithm order 1 or 2, and can also be increased using a
neered fiber is presented with synthetic gauge length opti- higher sampling frequency Fs. For a harmonic signal
mization (Equation 1.44). It is worth mentioning that this cos(2πFt), we can normalize strain results as:
optimization can be effectively applied to DAS with engi-
neered fiber only, as it has no significant pink noise and A0 F S
ε1 ≤ (1.50)
can be effectively spatially averaged. As is clear from 2L0 F
Figure 1.29, the performance of DAS with engineered A0 F S 2
fiber can reach seismometers, and it is deep below Peter- ε2 ≤ (1.51)
4πL0 F
son’s low noise model level (Peterson, 1993). So, the engi-
neered fiber antenna is an equivalent of a set of multiple The maximum strain comparison for the first and sec-
seismic stations and can be used for passive seismic appli- ond order tracking algorithm ε1 and ε2 (Equations 1.50–
cations. Moreover, DAS with engineered fiber has unique 1.51) is presented in Figure 1.30 for FS = 50kHz and
sensing capability below 1 Hz, where gravitational wave L0 = 10m. The second order algorithm can deliver
Ground displacement noise, Zmin, m/Hz0.5
Geophone SG5-SG10
Low Noise Model
10–7 DAS Lo=10m
Eng. DAS Lo=10m
Seismometer STS-2
10–8
Eng. DAS Lo=30m
10–9
10–10
10–11
10–12
0.1 1 10 100
Frequency, F, Hz
Figure 1.29 Displacement noise comparison of DAS (with and without engineered fiber) with seismometer and
geophone. 30 m DAS data are for synthetic gauge length.
SEISMIC DATA FROM DISTRIBUTED AND ENGINEERED FIBER OPTIC ACOUSTIC SENSORS 27
10–1
10–3
Strain, ε
2.2 10–4
10–4
10–5
2.9 10–6
10–6
0.1 1 10 100 1000
Frequency, F Hz
Figure 1.30 Maximum strain comparison of first and second order algorithms for DAS.
measurements of fiber strain up to fiber breakage point overall sensitivity of a DAS system is a function of the
(~10%) at frequencies of around 10 Hz. coupling, cable, fiber, electronics, and digital signal pro-
We can now estimate the maximum DAS dynamic cessing, field data is most convincing, and, in the next sec-
range D as: tion, we will discuss some examples of high definition
ε1,2 seismic and microseismic data that demonstrate the ben-
D = 20 log 10 (1.52) efits of the engineered fiber DAS solution as compared to
εmin
conventional DAS and geophones.
Using the real noise level εmin = 0.03nanostrain from
Miller et al. (2016), we can estimate D = 99dB for a max- 1.3.3. Field Trial Results
imum value ε1 = 2.9μstrain. This estimation gives the
practical upper limit for seismic DAS at 100 Hz using A comparison of DAS with standard and engineered
Rayleigh scattering. Generally speaking, the second order fiber for a seismic sweep signal is presented in
tracking algorithm has limited applicability for a conven- Figure 1.31. This measurement was provided using two
tional DAS because flicker noise pulses can reach π and different fibers placed side by side in the same optical
destroy measurements in accordance with Equation 1.49. cable, so the elongation of both fibers was identical.
Nevertheless, 120dB was achieved in Parker et al. (2014) The top graphs (a) and (c) demonstrate the difference
when the fiber elongation zone was significantly smaller between the time-distance representation; the right panel
than the gauge length and pulsewidth, such that the flicker (c), which represents engineered fiber, is visibly cleaner
noise was suppressed. However, when a continuous seis- than the left panel (a). The detected seismic signal has
mic signal expands the reflectivity zone, then the reflection the same shape (around 10 nm peak to peak for a channel
can disappear, and the signal has ambiguity. Fortunately, 898) for engineered (d) and standard (b) fiber, except
in engineered fibers, the scatter center zones are well noise. Some change in amplitude (20%) can be explained
defined, and so the reflectivity change is negligible. As a by incomplete averaging of the DAS signal over distance,
result, we can optimistically estimate a maximum as is shown in Figure 1.7. There was less variation in the
D = 167dB for engineered fiber using εmin = 1picostrain amplitude level for engineered fiber, and this stability can
and maximum ε2 = 220μstrain—see Figure 1.30. be important for 3D VSP, as was shown in Figure 1.21.
The dynamic range of DAS with engineering fiber was A comparison of DAS acoustic noise with standard and
tested during a dry alluvium geology series of chemical engineered fiber is presented in Figure 1.32. Noise spectral
explosions, including 50,000 kg TNT-equivalent at 300- density versus distance is practically constant for engi-
m depth-of-burial (Abbott et al., 2019). “Two orders of neered fiber (b) but varies significantly for standard fiber
magnitude more data relative to traditional geophones/ from channel to channel along distance (a). In other
accelerometers” was successfully recorded. words, we can conclude that standard DAS noise depends
Summarizing, we can conclude that theoretical estima- on fiber randomness and can be far from the average
tions demonstrate that the performance of DAS with engi- value, but engineered fiber DAS noise is predictable.
neered fiber can potentially exceed that of conventional The SNR difference is emphasized by the signal Fourier
geophones and seismometers. In general, given that the transform in the bottom chart (c): the noise reduction
28 DISTRIBUTED ACOUSTIC SENSING IN GEOPHYSICS
1000 1000
Rad
Distance, m
Distance, m
800 0.2
800
0.1
600 0
600
–0.1
Acoustic, rad
0.2 0.2
Acoustic, rad
0 0
–0.2 –0.2
0 0.5 1.0 1.5 2.0 0 0.5 1.0 1.5 2.0
Time, s Time, s
Figure 1.31 Comparison of DAS with Rayleigh scattering [(a) and (b)] and engineered fiber [(c) and (d)] for a seismic
sweep signal. Acoustic signals are measured in optical phase radians.
(a) (b)
200 200
150 150
Spectrum
Frequency, Hz
Frequency, Hz
100 600
100
400
50 50
200
0 0 0
0 10 20 30 40 50 0 10 20 30 40 50
(c) –40 Distance, m Distance, m
Standard fiber
Spectrum, dB
–100
Figure 1.32 Comparison of DAS noise spectrums with Rayleigh scattering (a) and engineered fiber (b). Panel (c)
represents acoustic noise spectrum density with respect to 1 rad/Hz0.5. Source: Based on Richter et al. (2019).
SEISMIC DATA FROM DISTRIBUTED AND ENGINEERED FIBER OPTIC ACOUSTIC SENSORS 29
for engineered fiber is nearly 20 dB as was expected from A typical microseismic event is presented in
shot noise estimation (Equation 1.46). Figure 1.33, where both S- and P-waves are clearly visible,
Fine spatial resolution in combination with good sensi- such that the distance from observation well to fracking
tivity gives DAS a significant advantage for detection of event can be easily detected. Figure 1.34 shows how the
microseismic events, particularly where a geophone chain same installation can be used to detect a “frac hit,” where
cannot be readily positioned. Such measurements are used a fracking zone and strain extends slowly from the well
in fracking jobs, where a wireline fiber optic cable is undergoing treatment to the observation well. This new
pumped down into an already completed observation well data allows completion engineers to map the depth, azi-
(Richter et al., 2019). This gives the possibility to deter- muth, and speed of the fractures and feed that information
mine the frack height and well interference with unprece- back into the fracture models to validate and optimize the
dented clarity. designs for the next operation.
17,000
Pump To DAS Acoustic,
nm
16,000 5
Fracking 4
15,000
Depth, ft
event
3
14,000 2
13,000 1
0
12,000
–1
0 0.1 0.2 0.3 0.4 0.5 0.6
Time, s
Figure 1.33 Microseismic event in observation well detected by DAS with engineered fiber.
Strain 14,200
10
field
14,100
Depth, ft
5
14,000
0
13,900
–5
13,800
–10
13,700
13,600 –15
13,500 –20
–2
Strain rate, nstrain/s
–4
–6
–8
–10
–12
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4
Time, hours
Figure 1.34 Example of low frequency (down to millihertz level) “slow strain” data, showing a fracking hit on an
observation well from a well undergoing treatment. The time-averaged signal cross-section along the white line is
shown in the bottom panel.
30 DISTRIBUTED ACOUSTIC SENSING IN GEOPHYSICS
15 400 700 1000 1300 300 400 500 600 700 800 900 1000 1100 1200 1300
0
500
1000
1500
2000
Figure 1.35 Comparison of geophones (left panel) and DAS with engineered fiber (right panel).
The results from a VSP survey in a carbon sequestration Baird, A. (2020). Modelling the response of helically wound DAS
well (Correa et al., 2017) demonstrate that DAS with engi- cables to microseismic arrivals. Paper presented in First EAGE
neered fiber has the potential to provide similar, or even Workshop on Fibre Optic Sensing (Vol. 2020, No. 1, pp. 1–5).
superior, quality data sets as compared to conventional geo- European Association of Geoscientists & Engineers.
Benioff, H. (1935). A linear strain Seismograph. Bulletin of the
phones. An important aspect is that, due to the higher spatial
Seismological Society of America, 25(4), 283–309.
sampling, DAS data has the capability to provide more
Brennan, D. G. (1959). Linear diversity combining techniques.
detailed velocity information as compared to geophones. Proceedings of the IRE, 47(6), 1075–1102. doi: 10.1109/
This conclusion was expected from the preceding theory JRPROC.1959.287136
and is illustrated in Figure 1.35, which demonstrates even Carroll, J., & Huber, D. (1986). A fiber-optic hydrophone with a
a finer reflection structure from DAS than from geophones. mechanical anti-aliasing filter. Journal of Lightwave Technol-
In summary, we have estimated the main DAS perfor- ogy, 4(1), 83–86.
mance parameters for standard and engineered fiber and Correa, J., Egorov, A., Tertyshnikov, K., Bona, A., Pevzner, R.,
provided field data that correspond to the theoretical pre- Dean, T., et al. (2017). Analysis of signal to noise and direc-
dictions of improved sensitivity and dynamic range. tivity characteristics of DAS VSP at near and far offsets—A
CO2CRC Otway Project data example. The Leading Edge,
36(12), 994a1–994a7. doi: 10.1190/tle36120994a1.1
ACKNOWLEDGMENTS Crickmore, R. I., & Hill, D. J. (2010). U.S. Patent No. 7,652,245.
Washington, DC: U.S. Patent and Trademark Office.
The authors would like to thank Roman Pevzner and Crickmore, R., & Ku, E. (2017). U.S. Patent Application No. 15/
Valeriya Shulakova (of CO2CRC, Curtin University) 309,076.
and Thomas M. Daley, Barry M. Freifeld, Jonathan Dakin, J. P. (1990). Distributed fibre optic sensor system. UK
Ajo-Franklin, and Shan Dou (of the Lawrence Berkeley Patent, GB2222247A.
National Laboratory) for the use of raw and processed Dakin, J., & Culshaw, B. (Eds.). (1989). Optical fiber sensors:
surface seismic data and the presented field geometry. Systems and applications (Vol. 2, Chap. 15). Artech House
The authors would also like to acknowledge the signifi- Optoelectronics. Norwood, Massachusetts.
De Rosa, M., Carberry, J., Bhagavatula, V., Wagner, K., & Sar-
cant help and support from their colleagues at Silixa. In
avanos, C. (2002). High-power performance of single-mode
addition, the authors thank the reviewers for their helpful
fiber-optic connectors. Journal of Lightwave Technology, 20
comments in preparing this manuscript. (5), 851.
Ellis, R. (2007). Explanation of reflection features in optical fiber
REFERENCES as sometimes observed in OTDR measurement traces. Corn-
ing White Paper.
Abbott, R. E., Mellors, R. E., & Pitarka, A. E. (2019). Distrib- Farhadiroushan, M., Finfer, D., Strusevich, D., Shatalin, S., &
uted acoustic sensing observations and modeling of the DAG Parker, T. (2021). Non-isotropic acoustic cable. U.S. Patent
series of chemical explosions. Paper presented in CTBT Sci- Application No. 15/804,657.
ence & Technology 2019 Conference, T2.3-P12. https:// Farhadiroushan, M., Parker, T. R., & Shatalin, S. (2010). Method
ctnw.ctbto.org/ctnw/abstract/32643 and apparatus for optical sensing. WO2010136810A2.
Another random document with
no related content on Scribd:
of infamous crime, person unable to read Constitution
in English, and to write his name.
COLORADO.
Citizen or alien, male or female, who has declared intention 4
months prior to election.
CONNECTICUT.
Citizen of United States.
DELAWARE.
Citizen who has paid registration fee of $1.
FLORIDA.
Citizen of United States.
GEORGIA.
Citizen of the United States who has paid all
his taxes since 1877.
IDAHO.
Citizen of the United States, male or female.
Under guardianship, idiots, insane, convicted of felony,
treason, or embezzlement of public funds, polygamist or
bigamist.
ILLINOIS.
Citizen of the United States.
Convicted of felony.
INDIANA.
Citizen of United States, or alien who has declared intention
and resided 1 year in United States and 6 months in State.
IOWA.
Citizen of the United States.
KANSAS.
Citizen of United States, alien who has declared intention, or
[under] treaties with Mexico.
KENTUCKY.
Citizen of the United States.
LOUISIANA.
Citizen of United States or alien who has declared intention.
Idiots, insane, convicted of treason, embezzlement of public
funds, all crime punishable by imprisonment in penitentiary,
persons unable to read and write, and not owning property in
the State assessed at $300, or not the son or grandson of a
citizen of the United States prior to January 1, 1867, person
who has not paid pool tax.
MAINE.
Citizen of the United States.
MARYLAND.
Citizen of the United States.
MASSACHUSETTS.
Citizen of the United States.
MICHIGAN.
Citizen or inhabitant who has declared intention under United
States laws 6 months before election and lived in State
2½ years.
MINNESOTA.
Citizen of United States or alien who has declared intention,
and civilized Indians.
Convicted of treason or felony, unless pardoned, persons
under guardianship or insane.
MISSISSIPPI.
Citizen of the United States.
Missouri.
Citizen of United States or alien who has declared intention
not less than 1 year or more than 5 before offering to vote.
MONTANA.
Citizen of the United States.
NEBRASKA.
Citizen of United States or alien who has declared intention.
Convicts.
NEVADA.
Citizen of the United States.
NEW HAMPSHIRE.
Citizen of United States.
Paupers (except honorably discharged United States soldiers
and sailors), persons excused from paying taxes at their own
request,
NEW JERSEY.
Citizen of the United States or alien who has declared
intention 30 days prior to election.
NEW YORK.
Citizen who shall have been a citizen for 90 days.
NORTH CAROLINA.
Citizen of the United States.
NORTH DAKOTA.
Citizen of the United States, alien who has declared
intention 1 year, and civilized Indian.
OHIO.
Citizen of the United States.
OREGON.
Citizen of Unite States or alien who
has declared intention 1 year preceding election.
PENNSYLVANIA.
Citizen of the United States at least 1 month, and if 22 years
old or more, must have paid tax within 2 years.
RHODE ISLAND.
Citizen of the United States.
SOUTH CAROLINA.
Citizen of the United States.
SOUTH DAKOTA.
Citizen of the United States or alien
who has declared intention.
Under guardianship, idiots, insane, convicted of treason
or felony, unless pardoned.
TENNESSEE.
Citizen of the United States who has paid poll tax of
preceding year.
TEXAS.
Citizen of the United States or alien who has declared
intention.
UTAH.
Citizen, male and female.
VERMONT.
Citizen of the United States.
VIRGINIA.
Citizen of the United States.
WASHINGTON.
Citizen of the United States.
Indians not taxed, idiots, insane, persons convicted of
infamous crimes.
WEST VIRGINIA.
Citizen of the State.
WISCONSIN.
Citizen of the United States or alien who has declared
intention.
WYOMING.
Citizen of the United States, male and female.
Alabama, nine;
Arkansas, seven;
California, eight;
Colorado, three;
Connecticut, five;
Delaware, one;
Florida, three;
Georgia, eleven;
Idaho, one;
Illinois, twenty-five;
Indiana, thirteen;
Iowa, eleven;
Kansas, eight;
Kentucky, eleven;
Louisiana, seven;
Maine, four;
Maryland, six;
Massachusetts, fourteen;
Michigan, twelve;
Minnesota, nine;
Mississippi, eight;
Missouri, sixteen:
Montana, one;
Nebraska, six;
Nevada, one;
New Hampshire, two;
New Jersey, ten;
New York, thirty-seven;
North Carolina, ten;
North Dakota, two;
Ohio, twenty-one:
Oregon, two;
Pennsylvania, thirty-two;
Rhode Island, two;
South Carolina, seven;
South Dakota, two;
Tennessee, ten:
Texas, sixteen:
Utah, one;
Vermont, two;
Virginia, ten;
Washington, three;
West Virginia, five;
Wisconsin, eleven; and
Wyoming, one.
"SECTION 2.
That whenever a new State is admitted to the Union the
Representative or Representatives assigned to it shall be in
addition to the number three hundred and eighty-six.
"SECTION 3.
That in each State entitled under this apportionment, the
number to which such State may be entitled in the Fifty-eighth
and each subsequent Congress shall be elected by districts
composed of contiguous and compact territory and containing as
nearly as practicable an equal number of inhabitants. The said
districts shall be equal to the number of the Representatives
to which such State may be entitled in Congress, no one
district electing more than one Representative.
"SECTION 4.
That in case of an increase in the number of Representatives
which may be given to any State under this apportionment such
additional Representative or Representatives shall be elected
by the State at large, and the other Representatives by the
districts now prescribed by law until the legislature of such
State in the manner herein prescribed, shall redistrict such
State; and if there be no increase in the number of
Representatives from a State the Representatives thereof shall
be elected from the districts now prescribed by law until such
State be redistricted as herein prescribed by the legislature
of said State; and if the number hereby provided for shall in
any State be less than it was before the change hereby made,
then the whole number to such State hereby provided for shall
be elected at large, unless the legislatures of said States
have provided or shall otherwise provide before the time fixed
by law for the next election of Representatives therein.
"SECTION 5.
That all Acts and parts of Acts inconsistent with this Act are
hereby repealed."
{678}
{679}
Section 11 provides that "the enlisted force of the Corps of
Engineers shall consist of one band and three battalions of
engineers. … Each battalion of engineers shall consist of one
sergeant-major, one quartermaster-sergeant, and four
companies. Each company of engineers shall consist of one
first sergeant, one quartermaster-sergeant, with the rank,
pay, and allowances of sergeant, eight sergeants, ten
corporals, two musicians, two cooks, thirty-eight first-class
and thirty-eight second-class privates."
{680}
"The national verdict of 1896 has for the most part been
executed. Whatever remains unfulfilled is a continuing
obligation resting with undiminished force upon the Executive
and the Congress. But fortunate as our condition is, its
permanence can only be assured by sound business methods and
strict economy in national administration and legislation. We
should not permit our great prosperity to lead us to reckless
ventures in business or profligacy in public expenditures.
While the Congress determines the objects and the sum of
appropriations, the officials of the executive departments are
responsible for honest and faithful disbursement, and it
should be their constant care to avoid waste and extravagance.
Honesty, capacity and industry are nowhere more indispensable
than in public employment. These should be fundamental
requisites to original appointment and the surest guarantees
against removal.