APPLIED GEOPHYSICS, Vol.13, No.2 (June 2016), P. 288-306, 17 Figures.

DOI:10.1007/s11770-016-0553-1

Geophysical monitoring technology for CO2

sequestration*
Ma Jin-Feng♦1, Li Lin1, Wang Hao-Fan1, Tan Ming-You2, Cui Shi-Ling2, Zhang Yun-Yin2, Qu Zhi-Peng2,
Jia Ling-Yun1, and Zhang Shu-Hai3

Abstract: Geophysical techniques play key roles in the measuring, monitoring, and verifying
the safety of CO 2 sequestration and in identifying the efficiency of CO2-enhanced oil
recovery. Although geophysical monitoring techniques for CO2 sequestration have grown
out of conventional oil and gas geophysical exploration techniques, it takes a long time to
conduct geophysical monitoring, and there are many barriers and challenges. In this paper,
with the initial objective of performing CO2 sequestration, we studied the geophysical
tasks associated with evaluating geological storage sites and monitoring CO2 sequestration.
Based on our review of the scope of geophysical monitoring techniques and our experience
in domestic and international carbon capture and sequestration projects, we analyzed the
inherent difficulties and our experiences in geophysical monitoring techniques, especially,
with respect to 4D seismic acquisition, processing, and interpretation.
Keywords: Carbon capture and storage, geophysical monitoring, 4D seismic monitoring,
CO2 saturation, reservoir pressure

Introduction increase in the atmospheric concentrations of greenhouse

gases (Chu, 2009; Sun, 2006). Higher temperatures and
extreme weather events have serious impacts on global
Responding to the devastating effects of climate natural ecosystems and a particularly significant impact
change is one of the greatest challenges facing the on food production, which threatens human survival and
world today because climate change poses the greatest sustainable development.
threat to the future sustainability of humanity (IPCC, Carbon capture and sequestration (CCS) or carbon
2014). Overwhelming scientific evidence has proved capture, utilization and sequestration (CCUS) is a
that human activities since the industrial revolution, technology that can capture CO 2 from coal-fired
in particular, those of the developed countries, have power plants or other large CO2 emission sources and
produced great amounts of cumulative CO2 emissions by store it safely underground. According to estimates
the use of fossil fuels, which has resulted in a significant from the International Energy Agency (IEA, 2013),

Manuscript received by the Editor November 5, 2015; revised manuscript received March 6, 2016.
*This work was supported by National 863 Program Grant 2012AA050103 and Grant 2011KTCQ03-09.
1. State Key Laboratory of Continental Dynamics, Department of Geology, Northwest University, Xi'an 710069, China.
2. SINOPEC Shengli Geophysical Research Institute, Dongying 257022, China.
3. SINOPEC Zhongyuan Geophysical Research Institute, Puyang 457001, China.
♦Corresponding author: Ma Jin-Feng (Email: [email protected])
© 2016 The Editorial Department of APPLIED GEOPHYSICS. All rights reserved.

288
 Ma et al.

CCS technology could reduce one-sixth of the global 2009; Lawton, 2010). In this paper, we focus on the
carbon emissions by 2050. It is currently recognized challenges and opportunities of geophysical technology
internationally as the most efficient way to quickly for CO2 sequestration.
reduce the greenhouse effect (Bikle, 2009). Although
CCS technology and other technologies such as
renewable energy, nuclear energy, or energy-efficient Feasibility of CO2 sequestration
technologies are key technologies for dealing with
climate change, CCS technology is the most direct and
critical technology for reducing carbon emissions from The CO2 sequestration process is equivalent to the
power generation, steel, metallurgy, glass, ceramics, inverse of the oil and gas development process, in which
cement, and chemical industries based on fossil fuels. captured CO2 is injected into underground geological
Since the success of CCS operations in the Weyburn traps for safe and permanent storage. At present, the
oil field in Canada, the Sleipner gas field in Norway, and CO2 sequestration method internationally recognized as
the In Salah saline aquifer storage in Algeria (Service, most cost-effective is to capture CO2 from large-scale
2009), more international CCS projects are entering into emission sources and inject it into depleted oil and gas
construction and operation phases. In October 2014, the fields or into those in the late development stage. These
world’s first one-million-scale post-combustion CO 2 oil fields have structural and stratigraphic traps that have
capture CCS project became operational at SaskPower’s stored oil and gas for millions of years without leakage,
Boundary Dam power station in Saskatchewan, Canada and which are ideal safe geological storage places for
(Reiner, 2016). This is regarded as a global milestone for CO2. In addition, depleted oil fields have accumulated a
“clean coal” technology. At the end of 2014, there were large number of drilling sites, core samples, geological
22 large-scale global integrated CCS or CCUS projects study results, geophysical measurements, formation
either in operation or under construction (GCCSI, 2014). and production test results, and other information
China currently has several small development projects regarding the oil exploration and development process.
for CO 2 capture, CO 2-enhanced oil recovery (EOR), Their use can save considerable expense with respect
and sequestration. Four of these projects are integrated to the selection and baseline safety monitoring of CO2
CCS projects, including CO 2 capture, transport, and geological storage sites. More importantly, the use of
storage. They include the CO2-EOR project in the Jilin CO2-EOR can obtain 10% more than recovery by water
oilfield, the Ordos Shenhua saline aquifer CO2 storage injection. This additional revenue from CO2-EOR can
projects, the Jingbian CCS Project of the Shaanxi compensate for the cost of the entire CCUS project. If
Yanchang Petroleum Group, and the Shengli oilfield we consider the environmental benefits, the use of CO2
post-combustion carbon capture CCUS Project. Of these flooding not only saves water but also reduces the waste-
four integrated CCS projects, the Jingbian and Shengli water discharge from water or chemical flooding.
CCS projects have been listed as low-carbon promotion Although an oil and gas reservoir is the best place
technologies by China’s National Development and to carry out CCUS, its capacity for CO2 sequestration
Reform Commission (NDRC, 2014, 2015) since 2014. is much smaller than that of a deep saline aquifer.
In June 2015, the Jingbian CCS Project became China’s The potential capacity of saline aquifers for CO 2
first CCS project recognized by the Carbon Sequestration sequestration is huge. How to reduce the cost of CO2
Leadership Forum (CSLF) (CSLF, 2015). On September storage in saline aquifers and then utilize the brine is
25, 2015, the Jingbian CCS Project of the Shaanxi also an important challenge for international research
Yanchang Petroleum Group was listed in the China−US and development.
Joint Presidential Statement on Climate Change. It is The conditions and requirements for CO2 sequestration
also China’s first large-scale commercial CCUS project include:
on which China and the U.S. will cooperate (Chinese (1) Storage site selection, determination of the
Government, 2015). lithology, structure, traps, capacity, permeability,
CCS projects have the potential to grow into a big saturation, temperature, pressure, and other reservoir
market and open up new areas for the application of parameters, and more accurate analyses of the rock
geophysical technologies. Geophysical monitoring is a fabric, porosity, and micro-cracks prior to CO2 injection;
key technology for identifying CO2 geological storage (2) An understanding of the reservoir geometry,
sites, monitoring CO 2 migration, and evaluating the the thickness and extension of the seal layers, and the
safety of geological storage underground (Wills et al, geometric features and characteristics of its faults and

289
 Geophysical monitoring technology for CO2 sequestration

fractures; al., 2009; Spetzler et al., 2008; Zhang et al., 2015), and
(3) An understanding of the geological structure of passive seismic monitoring (Verdon et al., 2010; Ugalde
saline aquifers near major CO2 injection reservoirs and et al., 2013);
lithology information about their caprocks; 2) Borehole geophysical monitoring techniques (Xue
(4) Determination of the distribution of residual oil et al, 2006) including time lapse logging techniques,
and CO2-EOR efficiency; multi-level borehole temperature and pressure
(5) Verification of the integrity of the caprock and well monitoring techniques;
bore, and detection of any CO2 leakage and migration 3) Rock physics experiments (Brown, 2002; Xue and
underground; Lei, 2006; Martínez and Schmitt, 2013), i.e., experiments
(6) Verification of the stability of the CO2 plume and that simulate reservoir temperature, injection pressure,
its prevention from spreading by secondary seals if CO2 and other parameters under CO2 flooding conditions;
breaks through the first caprock; confirmation of the 4) Resistivity methods (Kiessling et al., 2010;
reservoir pressure and the actual CO2 storage capacity of Bergmann et al., 2012), including surface and borehole
the reservoir; monitoring techniques;
(7) Forecast of surface deformation and the rock 5) Gravity monitoring (Alnes et al., 2011; Gasperikova
mechanics of overlying layers; and Hoversten, 2008);
(8) An understanding of the mechanisms of long- 6) Remote sensing (Verkerke et al., 2014) and surface
term CO2 geological storage and geochemical reactions deformation monitoring (He et al., 2014; Samsonov et
during storage (Matter and Kelemen, 2009); al., 2015).
(9) Leakage risk assessment; Geophysical monitoring is performed throughout
(10) The development and application of near- every stage of CO 2 sequestration, i.e., prior to CO 2
surface and atmospheric rapid monitoring techniques injection (baseline), during injection (monitoring) and
and an understanding of the environmental effects of after closure (post-closure monitoring). After the closure
CO2 leakage with respect to the underground geological of a CO2 sequestration project, the security and safety of
structure. the CO2 storage must be monitored over the long term
The most important scientific questions are how (Ma and Zhang, 2010; Hao and Yang, 2012). During this
to improve the efficiency of CO 2-EOR during CO 2 stage, seismic monitoring is the most effective technique
sequestration and how to ensure that CO2 will be safely for subsurface monitoring.
retained for at least 200 to 1000 years. Geophysical Seismic monitoring for CO2 sequestration is a long-
techniques will play a vital role, and the biggest term process that mainly employs 4D (also known as
challenges in the development of these techniques time-lapse) seismic monitoring techniques. The 4D
include: how to measure and monitor CO 2 -EOR seismic differences of amplitude and small travel time
efficiency; how to verify the safety of CO2 sequestration within the reservoir are obtained by comparing seismic
within reservoirs thousands of meters deep; how to monitoring and baseline data or different monitoring
confirm that the amount of CO2 stored in a deep reservoir datasets, and are used to estimate the vertical and
is equal to that of the injected CO2; and how to detect horizontal distribution of the CO2 plume. 4D seismic
potential fast and slow leakage points. monitoring for CO2 sequestration has been successfully
applied in Canada (White, 2009), Norway (Chadwick et
al., 2010), Australia (Urosevic et al., 2010; Pevzner et
Geophysical monitoring techniques in al., 2011), the European Union (Kiessling et al., 2010;
CO2 sequestration Ivanova et al., 2012), and the United States (Finley,
2014). Figure 1 shows amplitude differences from 4D
seismic monitoring data in Canada’s Weyburn field.
Current geophysical monitoring techniques for CO2 Nearly all global CO 2 sequestration projects have
sequestration include: implemented either baseline seismic monitoring or
1) Four-dimensional (4D) seismic technology, seismic surveys to evaluate storage sites.
including 4D and three-component seismic techniques Compared with surface 4D seismic monitoring
(4D3C), 4D and three-component vertical seismic techniques, it is more difficult to implement vertical
profile technology (4D3CVSP), 4D and nine–component seismic profiling (VSP) and cross-borehole seismic
seismic techniques (4D9C) (White, 2012, 2013; Davis tomography during CO2 injection and especially after
et al., 2003), crosswell seismic tomography (Onishi et site closure. For example, existing VSP, micro-seismic

290
 Ma et al.

monitoring, and downhole monitoring techniques are operation. There could be a risk of a large amount of
carried out at monitoring wells at small test areas. It is CO2 being leaked from the CO2 injection, production, or
difficult to put seismic sources and receivers in both CO2 monitoring wells. Also, if CO2 injection is stopped, water
injection wells and producing wells. If CO2 injection is could gush up from the borehole, possibly preventing
stopped or the production of oil is stopped to monitor future injection of CO2 such that the operation would
seismic data, the bottomhole pressures of the injection have to be permanently shut down.
and production wells will differ from those in active Of the current CO2 sequestration projects in the world,

2002 2004 2007

-5000 0

0 km 2.5 km 5 km

Fig.1 Differences of seismic amplitude at the top of the Marly unit of the Weyburn field, Canada. The labels 2002, 2004,
and 2007 indicate the year in which the 3D seismic monitoring data was acquired: 2.80 million tons of CO2 were injected
in 2002, 3.70 million tons of CO2 were injected in 2004, and 7.40 million tons of CO2 were injected in 2007, minus the
baseline 3D seismic data acquired in 1999. The black lines represent horizontal production wells and green lines
represent horizontal CO2 injection wells (White, 2009). The yellow areas represent CO2 distribution areas.

the most comprehensive seismic monitoring techniques

have been applied in Canada’s Weyburn field (White,
2013), including five 3D3C seismic surveys, three 3D
VSP surveys, three 3D9C surveys, and five passive
seismic surveys before and during CO2 injection. Before
injecting CO2, a new borehole, Well 4-23 (Figure 2), was
drilled and logged at the edge of the 3D seismic survey
area (Ma and Morozov, 2010), and integrated well logs
were measured in this well, including fast and slow shear
wave velocities. Later, Well 4-23 was converted to a Well 4-23

water-injection well. Unfortunately, this new well is not

a core drill well. Core samples have been acquired from
two other core boreholes (Figure 2), and a rock physics
experiment was conducted at the Colorado School of
Mines (Brown, 2002).
WAG injector
In CO 2 sequestration projects being conducted in CO 2 injector
Water injector
depleted oil fields, there is often a shortage of drill Horizontal injector
Sample well
cores, as well as of dipole sonic logs. Alternatively, there Receiver location
Source location
are drill core samples, but no dipole sonic logs. These
shortcomings in the data affect the ability to correspond Fig.2 4D seismic surveys (Phase 1A) with well locations in
core samples to well logs and influences the accuracy of the Weyburn field. WAG indicates water-alternating gas (CO2).

291
 Geophysical monitoring technology for CO2 sequestration

subsequent fluid substitution calculations. sources and the replanting of receivers, it is very difficult
In order to avoid seasonal surface velocity changes, to achieve identical repeatable geometries. That is, it
4D seismic monitoring in the Weyburn field is on a is difficult to maintain consistency in the source and
fixed annual schedule from the end of November to receiver locations in repeated 3D seismic surveys (Ma
early December. Surface velocity changes could cause et al., 2009). Figures 3 and 4 show the differences in
differences between repeated seismic surveys. In actual the shot and receiver locations between the baseline and
vintage seismic acquisitions, due to the use of dynamite monitoring seismic surveys in the Weyburn field.

Distance (m
Distance (m)
20
20 19
19 18
18 17
17 16
16 15
15 14
14 13
13 12
12 11
11 10
10 9
9 8
8 7
7 6
6 5
5 4
4 3
3 2
Shot 2 Shot 1
Repeat Shot 1 0
0

Fig.3 Differences in the source coordinates between baseline and monitoring seismic surveys (Ma et al., 2009). Left: 2001 monitoring
data minus the 1999 baseline source coordinates; right: 2002 monitoring data minus the 1999 baseline source coordinates.

Distance (m) Distance (m)

20 20
19 19
18 18
17 17
16 16
15 15
14 14
13 13
12 12
11 11
10 10
9 9
8 8
7 7
6 6
5 5
4 4
3 3
2 2
Receiver
1 Receiver 1
0 0

Fig.4 Differences in receiver coordinates between the baseline and monitoring seismic surveys (Ma et al., 2009). Left: 2001
monitoring data minus the 1999 baseline receiver coordinates; right: 2002 monitoring data minus the 1999 baseline receiver
coordinates.

292
 Ma et al.

Nevertheless, attempts have been made to maintain of the 1999 baseline data, the amplitude and travel-
consistency in the source and receiver locations of the time differences between the monitoring and baseline
monitoring seismic surveys in the Weyburn field with data were obtained (Li, 2003), which gave researchers
those of the baseline seismic surveys. Table 1 lists the some confidence and laid the foundation for conducting
seismic acquisition parameters. The first 3D seismic subsequent 3D seismic monitoring. Based on the lessons
monitoring data were acquired in the year 2001, after learned from the changing geometry in the year 2001,
one million tons of CO2 had been injected and stored the monitoring seismic geometry for the year 2002 were
underground (Table 1). Although the 3D seismic changed to be consistent with the 1999 baseline seismic
geometry of the 2001 monitoring data differs from that geometry (Table 1).

Table1 4D seismic acquisition parameters in the Weyburn field (Ma et al., 2009)
Parameters\Year Baseline (1999) Monitor (2001) Monitor (2002)
Shot number 630 882 630
Receiver station 986 986 986
Sample rate (ms) 2 2 1
Maximum offset 2152.87 3445.84 2105.627
Maximum fold 77 132 78
Source type Dynamite, 1 kg, 12 m Dynamite, 1 kg, 12 m Dynamite, 1 kg, 12 m
Mitcham, 3C OYO, 3C OYO, 3C
Receiver type Frequency 10 Hz Frequency 10 Hz Frequency 10 Hz
Damping 70% Damping 1% Damping 0.7%
Source interval (m) 160 160 160
Receiver interval (m) 160 160 160
Patch 19 lines × 39 stations 19 lines × 39 stations 19 lines × 39 stations

The 4D seismic monitoring used for CO2 sequestration saturation by seismic inversion when CO2 saturation has
(Calvert, 2005; Johnston, 2013) differs from that surpassed 20%.
used to monitor flooding of water, heavy oil, steam In the Nagaoka Pilot Project, Kim et al. (2011)
assisted gravity drainage (SAGD), and steam flood. The confirmed that resistivity changes significantly with
elastic properties of supercritical CO2 under reservoir CO2 saturation, as shown in Figure 5. As a consequence,
temperature and pressure are similar to those of gas. monitoring resistivity could be an effective way to
When CO 2 is injected into a reservoir, the P-wave identify CO2 saturation. However, insufficient resolution
velocity of the reservoir decreases rapidly so that it is of resistivity imaging could pose a problem. Also, when
easy to measure variations of the seismic amplitude CO2 is injected into a reservoir, it dissolves in water and
and travel-time delay during CO2 flooding. Taking the
Weyburn field carbonate reservoir as an example, we 4.5 8.0
calculated that the P-wave velocity would change by Sonic Vp
4.0 Reduction resistivity 7.5
10% from that of the fluid substitution model during CO2
P-wave velocity (km/s)

flooding. In carbonate rock physics experiments, Wang

Resistivity (Ohm-m)

3.5 7.0
et al. (1998) also found that P-wave velocity could drop 3.0 6.5
an average of 9% for high-porosity core samples and
2.5 6.0
drop an average of 4% for low-porosity core samples.
In their sandstone rock physics experiment in the 2.0 5.5
Nagaoka Pilot Project, Xue and Li (2006) found that 1.5 5.0
by increasing CO2 saturation, P-wave velocity may be
1.0 4.5
reduced by more than 10% (Figure 5). However, when 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
CO 2 saturation surpasses 20%, the P-wave velocity CO 2 saturation
of sandstone sample no longer declines. As a result, it Fig.5 P-wave velocity and resistivity varying with CO2
then becomes a technical problem of how to invert CO2 sequestration (Kim et al., 2011)
293
 Geophysical monitoring technology for CO2 sequestration

forms carbonic acid. Carbonic acid dissolves minerals, out in order to accurately calibrate the baseline and
which causes salinity to increase and a decrease in the monitoring seismic data, including all the well logs
resistivity of the formation water. As such, if the actual measured prior to CO2 injection or in open hole wells.
monitored reservoir resistivity is affected by the CO2 and After CO 2 injection, the temperature, oil saturation,
formation water resistivity, it becomes difficult to obtain pressure, resistivity, gas-to-oil ratio, and brine salinity
an accurate measurement of CO2 saturation. will be changed. Even the porosity could be changed by
The change in pore pressure or confining pressure fracturing, acidification, and carbonic acid dissolution.
(overburden pressure minus pore pressure) during CO2 Therefore, time-lapse well logging is necessary to
injection is a main factor affecting the elastic parameters obtain accurate reservoir parameters. However, due to
of a reservoir as well (Figure 6). The impact of pressure the cementing and casing of the wells, time-lapse well
on the elastic parameters, such as P-wave velocity, logging will be influenced by the steel casing pipe,
could be greater than that of CO2 saturation, especially making it difficult and expensive to perform accurately.
in the late development stages of an oil field and in the At present, of all the CO2 sequestration projects in the
early stages of CO2 injection. Since shear wave velocity world, time-lapse well logging has been conducted only
is not sensitive to CO2 saturation, pressure change is at the Nagaoka Pilot Project (Xue et al., 2006), and
the main factor influencing shear wave velocity. For neutron, induction, and sonic logs (Nakajima and Xue,
example, in the Jingbian CCS Project of the Shaanxi 2013) were measured only several to 40 times. For this
Yanchang Petroleum Group, when CO2 injection began reason, the use of Gassmann’s theory to calculate fluid-
in September, 2012, the reservoir pressure had already substitution well logs remains the primary method of
dropped from the original 12 MPa to 2−3 MPa. In the estimating the elastic parameter curves of a reservoir at
Gao 89 CO2 injection area of the SINOPEC Shengli oil different stages of CO2 sequestration.
field, the reservoir pressures dropped from 42.6 MPa to When time-lapse well logging cannot be carried
28.1−32.2 MPa when CO2 injection began. In Canada’ out, rock physics experiments must be performed to
s Weyburn field, the original reservoir pressure was 15 establish a relationship between reservoir pressure and
MPa, and when CO2 injection began, the bottomhole the elastic parameters (Brown, 2002). Then, the well
pressure of the injection well was 23 MPa, and the logs can be corrected or fitted to reservoir conditions
pressure of the producing well was 8 MPa (Ma and that are consistent with the seismic data acquisition time,
Morozov, 2010). In addition, reservoir temperature, i.e., reservoir parameters such as temperature, injection
brine salinity, gas-to-oil ratio, and the crude oil API also pressure, and oil saturation, so that the well logs match
affects the elastic parameters of the reservoir. the 4D seismic data. Further, reservoir models could
be developed based on the corrected well logs to
3600
interpret 4D seismic data and perform seismic inversion
2100 (Mezghani et al., 2004; Roggero et al., 2007).
P-Wave Velocity (m/s)

3400
S-wave Velocity, m/s

2000 Passive seismic monitoring can be used to monitor

3200 possible CO2 breakthrough locations caused by high
1900
3000 injection pressure and the occurrence of micro-cracks,
Vp
Vs1
1800 faults, and breakthroughs of the caprock. It can also
3800
Vs2 1700 be used to monitor and evaluate borehole integration
2600
0 5 10 15 20 25 30 35 caused by earthquake by monitoring any damage to the
Confining = Differential Pressure (MPa) cementing and casing, which could lead to CO2 leakage
Fig.6 Changes in the P-wave and shear wave velocities from the wellbore.
with the confining pressure of a dolomite dry sample in
the Marly unit of the Weyburn field (Brown, 2002).
Technical problems of geophysical
Other geophysical methods are generally used in methods for monitoring CO2
conjunction with 4D seismic monitoring to obtain
reservoir parameters at different stages of CO2 injection.
sequestration
These geophysical methods should be conducted at
the same time as the seismic acquisitions. These data The technical problems associated with the use of
can then be used to calibrate the 4D seismic data. For geophysical methods for CO 2 sequestration can be
example, time-lapse well logging must be carried divided into the four aspects of acquisition, equipment,

294
 Ma et al.

processing, and interpretation: after the injection of one million tons of CO2. While the
(1) Acquisition source and receiver locations of the 2001 monitoring
Of fundamental importance is maintaining the survey were carried out as near as possible to the
same geometries and source and receiver locations. location of the 1999 baseline survey, in order to increase
For example, in 4D seismic surveys, current onshore the seismic stacking fold, a swath of the baseline survey
acquisition methods cannot realize repeatable shot and was split into two or three sections in the monitoring
receiver locations for monitoring and baseline surveys. survey, as shown in Figure 7 (Ma et al., 2009). Even so,
Nevertheless, as early as in the beginning of the second the seismic differences between the 2001 monitoring
phase of the Weyburn project, geophysicists had thought data and the 1999 baseline data were well within the
to plant permanent geophones and to use a Vibroseis. post-stack seismic section (Li, 2003). The distribution
However, the use of a Vibroseis in a monitoring survey of CO2, as determined by the differences in the seismic
produces seismic signal differences between the baseline amplitude map, was affected by the change in geometry.
dynamite sources and the Vibroseis monitoring. This is why, for 2002, the monitoring acquisition
In the Weyburn field, the baseline 3D seismic survey geometry was changed back to be consistent with that of
was acquired in December 1999, and the first 3D seismic the 1999 baseline (Table 1).
monitoring survey was carried out in December 2001

154 154
156 153 156 153
1999 158 155154 2002 158 155154
160159157156 153 160159157156 153
162 158 155154 162 158 155154
164 161160 157156 153 164 161160 157156 153
166165163162 159158 155154 166165163162 159158 155154
168 164 161160 157156 153 168 164 161160 157156 153
170 167166 163162 159158 155154 170 167166 163162 159158 155154
172171169168 165164 161160 157156 153 172171169168 165164 161160 157156 153
174 170 167166 163162 159158 155 174 170 167166 163162 159158 155
176 173172 169168 165164 161160 157 176 173172 169168 165164 161160 157
178177175174 171170 167166 163162 159 178177175174 171170 167166 163162 159
180 176 173172 169168 165164 161 180 176 173172 169168 165164 161
182181179178 175174 171170 167166 1632139163 182181179178 175174 171170 167166 1632139163
180179177176 173172 169168 165 180179177176 173172 169168 165
182181 178 175174 171170 167 182181 178 175174 171170 167
180 177176 173172 169 180 177176 173172 169
182181179178 175174 171 182181179178 175174 171
180 177176 173 180 177176 173
182181179178 175 182181179178 175
180 177 180 177
182181179 182181179

Receiver Receiver
Shot Shot
Bad trace Bad trace

154
156 153
2001 158157155154 2001
160 156 153
162 159158 155154
164163161160159 157156 153
166 162 158 155154
168167165164 161160 157156 153
170169 166 163 162 159158 155
154
172 168 165164 161160 157156 153
174173171170 167166 163162 159158 155
169 165
172 168 164 161160 157 176
174173171170 167166 163162 159 178 175
172171169168 165164 161 180179177176
174173 170 167166 1632139163 182 178 175 2139163
172 169168 165 184183181180 177176
174 171170 167 186185 182181179178177175
173172 169 188 184 180 176
174173171 190 187186 183182 179178 175
192191189188 185184 181180 177176
194193 190 187186 183182 179178 175
192 189188 185184 181180 177
194193191190 187186 183182 179
192 189188 185184 181
194193191190 187186 183
192 189188 185
Receiver 194193191190189187 Receiver
Shot 192 Shot
194193191
Bad trace Bad trace

Fig.7 One swath shot of 2139163 in the 1999 baseline, 2001 monitoring, and 2002 monitoring surveys in the
Weyburn field. Blue numbers indicate receiver locations and red numbers source locations. Note that the
same source position was shot twice in 2001. As such, two swath shots in the 2001 monitoring survey can be
combined into one swath shot, as that in the 1999 baseline and 2002 monitoring surveys.

In the Weyburn project, the biggest problem in greater volumes of CO2 injection. Consequently, in CCS
subsequent seismic monitoring was that the CO2 had projects, the baseline 3D seismic survey area must be
migrated outside the baseline 3D seismic survey with designed to be big enough so that the CO2 plume does
295
 Geophysical monitoring technology for CO2 sequestration

not move outside the 3D seismic monitoring area during CO2. When we injected CO2 again, the backwater was so
long-term CO2 injection. However, if we increase the strong from the borehole that CO2 could not be injected.
area of a 3D seismic survey, the cost of 4D seismic Two years later, this well could again be injected with
acquisition and the investment necessary in the entire CO2, but the daily CO2 injection volume is small (Figure
CCS project will also increase. 9). Therefore, stopping CO2 injection during seismic
For CCS projects undertaken in China, only the acquisition can result in having to abandon the injection
Sinopec Shengli Oil Company has acquired a baseline well. In our experience, when acquiring seismic data,
and an overlaid 3D seismic survey in the Gao 89 CO2- particularly when acquiring monitoring seismic data,
EOR area. The baseline 3D seismic survey was acquired CO2 and water injection wells should not be shut down.
in the winter of 1992 and the overlaid or monitoring 3D The bottom borehole pressure can be maintained at that
seismic survey was acquired in the winter of 2011. CO2 of the injection or production pressures. For example,
injection began in 2007. By 2011, about 60,000 tons in the Weyburn field, CO2 and water injection continues
of CO2 had been injected into the reservoir. However, during the acquisition of monitoring seismic data.
the geometries of the baseline and overlaid 3D seismic
surveys are quite different. The acquisition pattern in 45586-03
45587-01 45586-01 92
1992 was 4 lines × 6 shots, the fold-coverage was 20, 45586-04
the common-depth-point (CDP) bin size was 25 × 100 m,
45587-04 45543-06 45543-05
and the sample rate was 4 ms. In 2011, the acquisition 45544-05
45543-02
51-01 51-03

pattern was 18 lines × 12 shots, the fold-coverage Q44-03 45543-03

45543-08 45543-07 51-02 38081-02
was increased to 225, the CDP bin size was 25 × 25 45544-04
Q44-01 45543-04
Q44-02
m, and the sample rate was 2 ms. Besides, the source 45544-08
45543 45543-10 38079-01 38079-02
and receiver locations of the 1999 baseline 3D seismic 45543-01
survey did not coincide with those of the 2011 overlaid 45544-07 45545-07
38079
3D survey. 45543-09
45545-05
45545-06
45542-04 45545-03
Most depleted oil fields have 3D seismic data, so if we Q46-02 45545 38134-0538134-06
Q46-01
take the 3D seismic data acquired earlier as the baseline Q46-3 45545-01
45547-04 45547-08 45545-02
data when carrying out CO2-EOR or sequestration, then 45545-04 38134-03 38134-04

the monitoring seismic acquisition parameters are often

higher than those of the baseline, such as in high-density Fig.8 Well distributions of the Jingbian CCS Project of the
and wider azimuth acquisitions. It would not be an easy Shaanxi Yanchang Petroleum Group (Ma et al., 2014). The
task to make the monitoring geometry match the baseline yellow symbols indicate CO2 injection wells; blue symbols
geometry. In addition, with the ongoing construction of water injection wells, and red solid circles producing wells.
a well site, including the installation of water, oil, CO2,
and sewage pipelines, monitoring seismic acquisition 12
45543
must be done very carefully when using a dynamite 45543-03
Injection Pressure (MPa)

10
source. To avoid damaging the underground facilities 45543-05
45543-08
of oil fields, dynamite source locations must be moved 8 45543-09
from their designed locations when acquiring monitoring 6
seismic data, causing the source locations to be different
4
from those of the baseline.
When acquiring monitoring 3D seismic data at the 2
Gao 89 area in 2011, the geophysical crew asked the oil 0
company to shut down the oil and CO2 injection pumps
2012-09

2012-12

2013-03

2013-06

2013-09

2013-21

2014-03

2014-06

2014-09

2014-12

2015-13

at the well site in order to reduce background noise. This

YY-MM
resulted in one of the four CO2 injection wells never
being able to be injected with CO2 again. Similar things Fig.9 Injection pressure of five CO2 injection wells in the
happened in the Jingbian field of the Shaanxi Yanchang Jingbian field of the Shaanxi Yanchang Petroleum Group (Ma
Petroleum Group at its first CO2 injection well 45543-03 et al., 2014). Note that the red line represents the first CO2
(Figure 8). When the CO2 source was changed from injection well 45543-03. In September 2012, CO2 injection
began at this well, and then, following the replacement of the
food-grade CO2 to captured CO2 from the Yulin Coal
CO2 source, it could not be injected further until March 2012.
Chemical Factory, we temporarily stopped injecting

296
 Ma et al.

Of course, if operations continue, well site pumping Oil and water injection pumping were not shut down
equipment produces loud noise during seismic data during this acquisition, so the seismic data would reflect
acquisition. For example, in the second CO2-EOR and the pressure during production. However, the pumping
storage at a site of the Shaanxi Yanchang Petroleum noise interfered with the seismic signal and affected the
Group, i.e., a 3D seismic baseline area of 10.68 km2 was quality of the seismic data (Figure 10).
acquired in the Changguanmiao area of the Wuqi field.

1.4 1.4

1.6 1.6

1.8 1.8
Time (seconds)

Time (seconds)
2

2.2 2.2

2.4 2.4

2.6 2.6

2.8 2.8

Fig.10 Typical pumping noise in a shot gather of a baseline 3D seismic survey in the Wuqi field.

In the Saskpower’s Boundary Dam CCS saline aquifer depends on costs. Seismic acquisition costs must be
storage site, 630 permanent receivers are embedded at a effectively reduced in order to decrease the cost of
depth of 20 meters in a 6.25-km2 area, in order to ensure the entire CCS project. However, unlike 4D seismic
consistent receiver locations for the 4D seismic survey. monitoring for water flooding and steam injection for
A Vibroseis truck was used in the monitoring seismic heavy recovery, the major reservoir for the CO2 injection
survey after CO2 injection in April 2015. Prior to CO2 must be monitored as well as other reservoirs near the
injection, vintage baseline 3D seismic data was acquired major reservoir and saline aquifers. This is because
in March 2012, April 2013, and November 2013 with the injected CO2 might leak into nearby reservoirs, so
a dynamite source. These three baseline 3D seismic seismic monitoring is required to track and determine
datasets are used to evaluate the repeatability of the 4D CO2 leakage pathways. Furthermore, the optimal way to
seismic data acquisition and processing (Rostrona et al., store CO2 is in multiple reservoir units and aquifers, also
2014; White et al., 2014), and especially to evaluate the known as the stack storage method. Using this storage
impact of seasonal variation on 4D seismic acquisition. method, we can effectively reduce the investment
Before CO2 injection, except for three baseline 3D associated with geological, geophysical, drilling, and
seismic acquisition sessions, passive seismic, resistivity/ other investigations. As such, these nearby reservoirs
magnetotelluric, gravity and other surface monitoring and aquifer parameters must be studied as an integral
techniques were carried out in the saline aquifer storage part of major oil and gas reservoirs. So, when acquiring
site of the Boundary Dam power station. In the borehole, 4D seismic monitoring data, the acquisition parameters
crosshole seismic, VSP, joint surface and borehole must be taken into account, particularly the stacking fold
resistivity, online temperature and pressure monitoring, of the major reservoir and its nearby reservoirs, which
and borehole passive seismic monitoring techniques may increase the cost of seismic acquisition.
were performed. An advanced time-lapse logging (2) Geophysical equipment
technique was designed as well (White et al., 2014). The majority of geophysical monitoring equipment
With respect to 4D seismic data acquisition for CO2 for conventional oil and gas exploration can be used
sequestration, the optimal time interval of repeated 3D in CO 2 sequestration. However, the research and
seismic monitoring and observation geometry surveys development of a borehole monitoring device for CO2

297
 Geophysical monitoring technology for CO2 sequestration

injection, production, and monitoring wells has taken (3) Data Processing
the technical high ground in equipment development. In 4D seismic data processing for CO2 sequestration,
Online monitoring data in the borehole may include the the objective is to obtain the differences between
VSP, temperature, pressure, and time lapse logs. These repeated seismic surveys, and then use these seismic
instruments must be placed at different well depths and differences to determine the effects of CO 2 -EOR,
require miniaturization, high-temperature resistance, CO2 distribution underground, and abnormal pressure
and corrosion protection from carbonic acid, as they distributions, and to confirm the safety of the CO 2
must be placed permanently in the borehole. This is storage. Since P-wave velocity obviously declines in
because the monitoring of reservoir parameters from the reservoir after CO2 injection, it should be easy to
the borehole may continue to be required after the CO2 measure the seismic differences before and after CO2
sequestration project is completed. These parameters are injection. This does not mean that we can process
very important in calibrating seismic monitoring data reasonable seismic differences. When processing 4D
and making well log corrections post-closure of the CO2 seismic data, we should attribute different processing
sequestration. flows to the characteristics of the seismic data. These
The current temperature and pressure monitoring processing methods are not unique.
method is to sink monitoring instruments into the bottom In the Weyburn field (Ma et al., 2009), although 3D
of the borehole and pull them out to read data after a seismic acquisition was carried out in December, we see
period of CO2 injection. While VSP and time-lapse well that the first arrivals are different at the same source and
logging are currently only carried out in monitoring same receiver locations of the vintage 3D seismic shot
wells, if they could be carried out in CO2 injection wells, gathers (Figures 11 and 12). For this reason, we must
we could obtain more accurate and valuable reservoir calculate the average first-break differences between the
parameters at these wells under high-pressure conditions. baseline and monitoring 3D seismic data, and correct
From the Weyburn Project, we learned the importance the first arrival data by applying the average first-break
of geophysicists using fiber optic geophones and differences to the monitoring seismic data so that the
permanently burying them behind the casing to measure baseline and monitoring data have the same first-break.
VSP, as they did in the Boundary Dam aquifer storage site Thus, we can process vintage 3D seismic data sets
(White et al., 2014). The fiber optic geophone is small, separately using the same processing flow.
has high sensitivity, has strong anti-electromagnetic To ensure that the vintage poststack seismic profiles
interference, can withstand high temperature and are comparable, we applied the static correction of the
pressure, has no electrical leakage, and is easy to reuse. baseline (1999) data sets to that of the monitoring data
It may be used to achieve permanent, real-time online sets after making the first-arrival correction, and vice
measurement. Crosswell resistivity monitoring is also versa. Only the data quality of the static correction of the
facing equipment problems with respect to online baseline was better than those of the monitoring data sets
monitoring in the well during CO2 injection. in the Weyburn field.

Source Source
Chan 1 3 6 8 11 13 Chan 1 3 6 8 11 13
740 740

780 780

820 820

860 860
Time (ms)

Time (ms)

900 900

940 940

980 980

1020 1020

1060 1060

Fig.11 Comparison of the same shot gathers by Fig.12 Comparison of the same shot gathers by
inserting them into the display. The first trace was inserting them into the display. The first trace was
acquired in 1999, the second trace in 2001, and so on. acquired in 1999, the second trace in 2002, and so on.

298
 Ma et al.

157
156 157
159 156
158 159
161 157 158
160 156 161 157
163 159 160 156
162 158 163 159
165 161 157 162 158
164 160 156 165 161 157
167 163 159 164 160 156
166 162 158 167 163 159
169 165 161 157 166 162 158
168 164 160 156 169 165 161 157
171 167 163 159 168 164 160 156
170 166 162 158 171 167 163 159
173 169 165 161 157 153 170 166 162 158
172 173
175
174
171
170
168
167
166
164
163
162
160
159
158
156
155
154
152 Time (ms) 175
172
171
169
168
167
165
164
163
161
160
159
157
156
155
153
152
Time (ms)
177 173 169 165 161 157 153 174 170 166 162 158 154
179
178
176
175
174
172
171
170
168
167
166
164
163
160
159
156
155
152 0 179
177
176
175
173
172
171
169
168
167
165
164
161
160
157
156
153
152
181
180
179
177
175
173
172
171
169
168
167
165
164
162
163
161
160
158
159
157
156
154
154
152 -2 181
180
178
177
174
173
172
170
169
168
166
165
163
164
162
161
159
160
158
157
155
154
154 16
182 178 174 170 166 156 179 175 171 167 156 152
181
180
179
177
176
175
173
172
171
169
168
167
165
164
162
163
161
160
158
160
158
156
154
153
152 -4
182
181
180
178
177
176
174
173
172
170
169
168
166
165
163
164
162
161
159
160
158
158
156
154
153
14
182 178 174 170 166 155 179 175 171 167 156 152
184
183
181
180
179
177
176
175
173
171
169
172 2116171168
167
165
164
162
164
162
160
158
157
156
154
153
-6 184
182
181
180
178
177
176
174
173
170
169
172 2116171168
166
165
163
164
162
162
160
158
157
155
154
153
12
186 178 174 170 166 159 155 183 179 175 171 167 160 156

190
188
187
185
184
183
181
180
179
177
176
175
173
172
171
169
168
168
166
164
162
163
161
160
158
157
-8 188
186
185
184
181
180
178
177
176
174
173
172
170
169
168
166
166
164
164
162
161
159
158
157
155
10
186 182 178 174 170 159 187 183 179 175 171 160
189
188
187
185
184
183
181
180
179
177
176
175
173
172
172
170
168
166
167
165
164
162
163
161
-10
190
189
188
186
185
184
182
181
180
178
177
176
174
173
172
170
170
168
168
166
165
163
162
161
159
8
190 186 182 178 174 170 187 183 179 175 172 164
189
188
187
185
184
183
181
180
179
177
176
176
174
172
171
169
168
166
167
165
-12
190
189
188
186
185
184
182
181
180
178
177
176
174
174
172
170
169
167
168
166
165
163
6
190 186 182 178 174 170 187 183 179 176 171
189
188
187
185
184
183
181
180
180
178
176
175
173
172
171
169
-14
190
189
188
186
185
184
182
181
180
178
178
176
174
173
172
170
169
167
4
190 186 182 178 174 187 183 180 175
189
188
187
185
184
184
182
180
179
177
176
175
173
-16
190
189
188
186
185
184
182
182
180
178
177
176
174
173
171
2
190 186 182 178 187 184 179 175
189
188
188
186
184
183
181
180
179
177
-18
190
189
188
186
186
184
182
181
180
178
177 0
190
186
185
182
181 Receiver 188
186
183
182
179
Receiver -2
188
187
184
183 -20 190
188
185
184
181
190
189
186
185 Shot 190
187
186
183
Shot -4
188
187
Bad trace -22 189
188
185
190
189 190
187 Bad trace -6
-24 189

Fig.13 Differences in the first arrival for the same shot (2116171) between the baseline and monitoring 3D seismic data
sets in the Weyburn field. Left: first arrival of 2001 (Monitor) minus that of 1999 (baseline); right: first arrival of 2002
(Monitor) minus that of 1999 (baseline).

When separately calculating the static correction of and monitoring 3D seismic data are relatively large.
the baseline and monitoring 3D seismic data sets, we Given these large differences in the static corrections,
found their differences to be quite large. Such large if we applied the same static corrections from either
differences in the static correction also affects the the baseline or monitoring data to both the baseline
comparison of the vintage common midpoint (CMP) and monitoring data, this would cause large errors in
gathers and poststack seismic data. In Figures 14 and the data even if we had applied first-arrival correction.
15, we compared the differences in the static correction Such errors would lead to the need for a larger residual
in the vintage seismic data sets acquired in 1999, 2001, static correction, and this might lead to new errors in the
and 2002 without any first-arrival correction. The seismic amplitude and other attributes.
differences in the static correction between the baseline

25

10 20

15

5 10
Static difference (ms)

Static difference (ms)

0 0

-5

-5 -10

-15

-10 -20
Receiver Receiver
-25

Fig.14 Differences in receiver static corrections between the baseline and monitoring data sets in the Weyburn field (Ma et al.,
2009). Left: 1999 (baseline) minus 2001 (monitoring); right: 1999 (baseline) minus 2002 (monitoring).

299
 Geophysical monitoring technology for CO2 sequestration

8 4

4 2

Static difference (ms)

Static difference (ms)

0 0

-4 -2

-8 -4
Shot Shot

Fig.15 Differences of source static correction between the baseline and monitoring data sets in the Weyburn field
(Ma et al., 2009). Left: 1999 (baseline) minus 2001 (monitoring); right: 1999 (baseline) minus 2002 (monitoring).

When the acquisition interval between repeated Vibroseis should become the predominant source in
3D seismic surveys is not long, the differences in the CO2 sequestration projects. However, it is still difficult
static corrections should not be large. However, if the to completely remove harmonic noise in the current
acquisition interval is longer, so will be the differences seismic processing method when using the Vibroseis.
in the static corrections. For example, in the Gao 89 area Hammer et al. (2004) compared Vibroseis and explosive
of the Sinopec Shengli oil field, baseline 3D seismic shot gathers generated at the same source location, from
data was acquired in 1992, and the overlaid (monitoring) which we can see the differences caused by different
3D seismic data was acquired in 2011. The larger static sources (Figure 16). Therefore, the differences associated
correction differences in the baseline and monitoring with the use of Vibroseis, explosive, and other hammer
data that occurred in such a long interval are likely sources may result in errors in 4D seismic interpretation.
mainly caused by the declining water table. Increasing Processing and comparisons of repeated 3D seismic
temperature has led to climate change and ground water data sets could be carried out for shot gathers, CMP
evaporation, and the overexploitation of ground water gathers, and poststack and migrated data sets. Ideally,
has also resulted in the continuous drawdown of the repeated seismic data sets should be compared with
water table. This situation is particularly serious in the shot gathers. That is, if amplitude changes and travel-
north and northwest of China. The water table in many time delays caused by CO2 injection could be observed
areas of the Ordos Basin in northwest China has dropped or processed by the comparison of shot or CMP gathers,
50−100 meters or more. This decline in the water table then subsequent pre-stack seismic inversions could
will result in a frequency change in the seismic data in produce better results. However, in actual 4D seismic data
vintage seismic data acquisitions. acquisitions, even if the source and receiver locations can
A change in the seismic sources will directly affect the be repeated before and after CO2 injection, the impact of
results of the 4D seismic data processing. For example, bad traces, pumping interference, and the interference of
in Australia’s CO2CRC Otway Project, baseline 3D alternating currents on each 3D seismic survey will not
seismic data was acquired in January 2008 using a be the same. Thus, with interference noise occurring at
hammer source, while monitoring 3D seismic data was different traces, bad traces in the baseline data would not
acquired in 2009 after injecting 35,000 tons of CO 2 correspond to those in the monitoring data. Baseline bad
using a Vibroseis (Urosevic et al., 2010; Pevzner et traces may correspond to monitoring normal traces, and
al., 2011). For the sake of maintaining the same source vice versa, and cause difficulties in making comparisons
locations and to minimize environmental damage, the of repeated shot or CMP gathers.

300
 Ma et al.

(a)
NW SE
Peg 968 938 908 878 848 Land Ice 798
0.0
Dyke reflection

0.05
Time (seconds)

0.1

Line 2 shot gather: vibroseis shot 100 m

0.15
(b) NW SE
Peg 968 938 908 878 848 798
0.0
Dyke reflection

0.05
Time (seconds)

0.1

100 m
Line 2 shot gather: explosive shot
0.15

Fig.16 Comparison of Vibroseis and explosive shot gathers (Hammer et al., 2004).

CDP traces of repeated poststack and migrated seismic monitoring and processing is dependent on whether the
data can correspond, which will make it easy to perform 4D seismic processing results are consistent with those
cross-equalization and other comparisons. However, of geological analysis, CO2 flooding, tracing of CO2,
the ability to evaluate the effect of 4D seismic data geochemical analysis, and reservoir numerical simulation.

Iline_no
Xline_no 4
14 17 21 24 27 31 34 37 41 44 47 51 54

600
700
800
900
1000
1100
Time (ms)

1200
1300
1400
1500
1600
1700
1800

Fig. 17 Comparison of poststack 3D seismic baseline and monitoring data in the Weyburn
field, where the first trace is 1999 baseline data, the second trace is 2001 monitoring data,
and the third trace is 2002 monitoring data. The top of Marly unit is around 1140 ms.

301
 Geophysical monitoring technology for CO2 sequestration

This is the most difficult challenge in 4D seismic it is difficult to estimate CO2 saturation near injection
monitoring and processing. Sometimes, in our experience, wells. For reservoirs with natural or hydraulic fractures,
even after processing 4D seismic data multiple times, we the difficulty of well log prediction is how the open
still could not achieve satisfactory results. or closed micro-fractures caused by the injection or
Often, the evaluation results of 4D seismic processing production pressures will impact the compressional,
is affected by the available geological knowledge during shear wave, and density logs (Shen et al., 2009; Wei et
CO2 flooding. For example, it is generally believed that al., 2013).
CO2 saturation is high near an injection well. While 4D The greatest difficulty in 4D seismic interpretation
seismic differences must show anomalies near injection is determining how to obtain CO2 saturation and pore
wells, the injected CO 2 might not be gathered near pressure from the amplitude differences and time-
the injection well, and could have migrated elsewhere delay information (Ivanova et al., 2012; Grude et al.,
along a highly porous, fractured channel or a quick 2013). Amplitude variation with offset (AVO) inversion,
breakthrough. If the CO2 injection region has a certain acoustic impedance inversion, elastic impedance
level of geological structure, the CO2 will migrate to inversion and other reservoir prediction technologies
the high point of the structure. In this case, 4D seismic are the main methods for obtaining these parameters
differences in the vicinity of the injection well might (Lumley, 2010; Meadows and Cole, 2013; Gong et
not be observed. It is also likely that when 4D seismic al., 2013; Huang et al., 2015). However, it is difficult
differences cannot be observed, the injected CO2 could to discriminate between the impacts of CO2 saturation
have broken through the caprock and moved into the and pore pressure in seismic difference data when
aquifer near the major oil reservoir. using only compressional wave information. Taking
It is tricky to process baseline and monitoring 3D into account that shear velocity is not sensitive to fluid
seismic data with different geometries. In this case, we saturation, the use of joint compressional and converted
may use geometry degradation or other methods to make wave interpretation, and the use of 4D-converted wave
the baseline and monitoring 3D geometries consistent. information to predict pore pressure may be one solution
Or, we may make baseline CMP traces correspond for predicting pore pressure (Yang et al., 2015).
to those of the monitoring data. Then, 4D seismic Binding and mashing the results of numerical reservoir
processing can be carried out. simulation is an important method in 4D seismic
(4) Interpretation interpretation (Huang et al., 1997; Johnston, 2013; Riazi
The goal of seismic interpretation for CO2 sequestration et al., 2013). Numerical reservoir simulation can be
is to eliminate interference from the seismic differences used to predict the characterization of CO2 migration,
in order to identify the CO2 and pressure distributions. distribution, and storage. It can also be used to predict
The ultimate goal is to use seismic information to verify the actual CO2 sequestration capacity as well. We may
whether the amount of injected CO2 equals the amount obtain the CO2 saturation and pressure distribution from
of the carbon stored. If the amount of CO2 injected does numerical reservoir simulation. When the CO2 saturation
not equal the amount of stored CO2, we must determine and pressure effects cannot be distinguished from the
from the 4D seismic information whether CO2 leakage 4D seismic data, and the seismic data resolution is not
occurred, where it originated, whether the CO2 is leaking enough to distinguish thin interbed and reservoir vertical
into nearby saline aquifers, and whether the CO 2 is heterogeneity, the CO2 sweep range and CO2 saturation
trapped by the secondary seals or is leaking to the surface. and pore pressure distribution models provided by
4D seismic interpretation includes baseline and numerical reservoir simulation can be verified and
monitoring data and differential data interpretation. improved from 4D seismic monitoring data. Then, the
In this case, synthetic seismograms must be made to CO 2 saturation volume can be obtained to calculate
calibrate these data, respectively. In the absence of the CO2 sequestration capacity and reservoir pressure
time-lapse logging data, an important step in the 4D volume to predict the risk of CO2 leakage.
seismic data calibration and interpretation becomes the For example, from the amplitude differences of
prediction of well logs with the CO2 injection pressure the Marly unit in the Weyburn field (Figure 1), it
and saturation at the seismic monitoring stages. To do so, is difficult to explain the CO 2 sweep efficiency and
well curves are logged at the baseline stage to predict the CO2 sequestration capacity. In fact, there are 16 CO2-
compressional and shear wave velocities and the density EOR patterns in the Phase 1A area (White et al.,
at the monitoring stages using fluid substitution methods. 2004). Different patterns have different CO2 flooding
However, when there is no time-lapse well logging data, efficiencies. Although the CO 2 flooding patterns

302
 Ma et al.

differ, their 4D seismic responses are similar. This is a or injection and geophysical methods are used to predict
characteristic non-uniqueness of 4D seismic difference the known CO2 reserves. These existing geophysical
data. Through numerical reservoir simulation, we can methods must be amended or improved with respect to
establish the correct reservoir model and reduce the non- known targets.
uniqueness of 4D seismic interpretation. Geophysical monitoring is required prior to CO 2
Considering that injected CO2 could leak into a nearby injection, during injection, and after site closure. It takes
saline aquifer, it is necessary to make a saline aquifer a long time to monitor the safety of CO2 sequestration.
prediction, which is similar to reservoir prediction, in The quality and effectiveness of 4D seismic monitoring
order to estimate both the amount of CO2 that might leak is directly affected by the acquisition methods used, the
into the saline aquifers and the pore pressure. Then, we monitoring equipment used, and the surface conditions.
can determine whether the caprock for the saline aquifer Processing 4D seismic data requires different processing
is safe and whether there is the risk that the pore pressure flows depending on the observed geometries and seismic
is high enough to break the caprock and cause the CO2 source types. The interpretation of 4D seismic data
to continue to move upward. Consequently, rock physics requires consistency with the fields and techniques of
experiments, fluid substitution, and other methods geology, well logging, determinations of CO2 flooding
applied in major reservoirs are also required in the study effect, and numerical reservoir simulation. Acquisition,
of upper saline aquifers. processing, and interpretation involve many problems
Technically speaking, conventional 3D seismic that must be resolved. The primary challenge is how
interpretation techniques are suitable for 4D seismic to measure, monitor, and verify the safety of CO 2
interpretation. Currently, some interpretation methods sequestration using advanced equipment, while realizing
for CO2 sequestration are being tested and explored, the lowest cost and designing the best monitoring
such as seismic attributes, absorption, and attenuation. program. This brings new challenges in the development
Of course, the comprehensive utilization of other of geophysical equipment, data processing, and
geophysical information, such as the inverted density interpretation techniques.
from repeated gravity monitoring, the inversion of CO2 With the global demand to cut carbon emissions
saturation by the resistivity method, locating fractures and the urgent need to address climate change, CCS
from microseismic monitoring, and making geophysical technology is currently the most direct and effective
results consistent with those from other disciplines and method for rapidly reducing carbon emissions and
technologies will all help to improve the accuracy of 4D is currently facing unprecedented opportunities for
seismic interpretation. development. The Chinese government has promised
to reach a peak in its carbon emissions by 2030 and
will launch a national cap and trade plan in 2017. This
Conclusions will bring great opportunities for the development of
CCS technology in China. With more CCS projects
being put into operation, geophysical monitoring for
Geophysical techniques, especially 4D seismic CO 2 sequestration will bring more opportunities for
monitoring, are the most effective and reliable techniques geophysical applications. The upcoming release of
for the measurement, monitoring, and verification of CO2 the International Organization for Standardization’s
sequestration. They represent the most important aspect (ISO’s) standards regarding carbon dioxide capture,
of the safety monitoring system of entire CCS projects. transportation, and geological storage (ISO/TC265) will
The greatest value in carrying out CO2-sequestration reflect the essential nature of geophysical monitoring
geophysical monitoring using oil and gas geophysical in CCS projects. With CCS, traditional geophysical
methods is to improve the accuracy of existing techniques for oil and gas exploration and resource
geophysical methods in reservoir prediction. This exploration have been given a new mission to expand
means that conventional reservoir prediction is used to their range and application.
predict unknown oil and gas reserves. The predicted
oil and gas reserves might vary greatly from the actual
reserves. A very long period of oil and gas exploitation Acknowledgements
is required to confirm the presence of predicted reserves.
However, reservoir prediction for CO2 sequestration is
used to determine the amount of potential CO2 storage With thanks for the support from the Shaanxi

303
 Geophysical monitoring technology for CO2 sequestration

Yanchang Petroleum Group, Geophysical Research Gasperikova, E., and Hoversten, G. M., 2008, Gravity
Institute of Sinopec Shengli Oilfield, Geophysical monitoring of CO 2 movement during sequestration:
Research Institute and Geophysical Company of Sinopec Model studies: Geophysics, 73(6),WA105–WA112.
Zhongyuan Oilfield. We also express our sincere thanks Gong, X., Zhang, F., Li, X., and Chen, S., 2013, Study of
to Professor Huang Xuri for his critical and constructive S-wave ray elastic impedance for identifying lithology
comments on the paper. and fluid: Applied Geophysics, 10(2), 145–156.
GCCSI-The Global CCS Institute, 2014, The Global Status
of CCS: http://www.globalccsinstitute.com/publications/
References global-status-ccs-2014
Grude, S., Landrø, M., and Osdal, B., 2013, Time-lapse
pressure–saturation discrimination for CO2 storage at the
Alnes, H, Eiken, O., Nooner, S., and Sasagawa, G., 2011, Snøhvit field: International Journal of Greenhouse Gas
Results from Sleipner gravity monitoring: updated Control, 19, 369–378.
density andtemperature distribution of the CO2 plume. Hammer, P. T. C, Clowes, R. M., and Ramachandran, K.,
Energy Procedia, 4, 5505–5511. 2004, High-resolution seismic reflection imaging of a
Bergmann, P., Schmidt-Hattenberger, C., Kiessling, D., thin,diamondiferous kimberlite dyke: Geophysics, 69(5),
Rücker, C., Labitzke, T., Henninges, J., Baumann, 1143–1154.
G., and Schütt, H., 2012, Surface-downhole electrical Hao, Y., and Yang, D., 2012, Research progress of carbon
resistivity tomography applied to monitoring of CO2 dioxide capture and geological sequestration problem
storage at Ketzin, Germany: Geophysics, 77(6), B253– and seismic monitoring research: Progress in Geophysics
B267. (in Chinese), 27(6), 2369–2383.
Bikle, M. J., 2009, Geological carbon storage: Nature He, Y., Zhang, B., Duan, Y., Xue, C., Yan, Q., He, C.,
Geoscience, 2, 815–818. and Hu, T., 2014, Numerical Simulation of Surface
Brown, L. T., 2002, Integration of rock physics and and Downhole Deformation Induced by Hydraulic
reservoir simulation for the interpretation of time-lapse Fracturing: Applied Geophysics, 11(1), 63–72.
seismic data at Weyburn Field, Saskatchewan: Master’s Huang, H., Wang, Y., Guo, F., Zhang, S., Ji, Y., and Liu, C.,
thesis, Colorado School of Mines. 2015, Zoeppritz equation-based prestack inversion and its
Calvert, R., 2005, Insights and methods for 4D reservoir application in fluid identification: Applied Geophysics,
monitoring and characterization: SEG Distinguished 12(2), 199–211.
Instructor Short Course, DISC series, 8, 2–3. Huang, X., Meister, L., and Workman, R., 1997, Reservoir
Chadwick, A., Williams, G., Delepine, N., Clochard, V., Characterization by Integration of Time-lapse Seismic
Labat, K., Sturton, S., Buddensiek, M., Dillen, M., and Production Data: SPE-38695, SPE Annual Technical
Nickel, M., Lima, A. L., Arts, R., Neele, F., and Rossi, Conference and Exhibition, San Antonio, Texas, USA, 5
G., 2010, Quantitative analysis of time-lapse seismic –8 October.
monitoring data at the Sleipner CO2 storage operation: IEA-International Energy Agency, 2013, Technology
The Leading Edge, 29(2), 170–177. Roadmap: Carbon Capture and Storage 2013 edition.
Chinese Government, 2015, China-U.S. Joint Presidential http://www.iea.org/publications/freepublications/
Statement on Climate Change: http://www.gov.cn/ publication/technology-roadmap-carbon-capture-and-
xinwen/2015-09/26/content_2939222.htm storage-2013.html.
Chu, S., 2009, Carbon capture and sequestration, Science, IPCC-Intergovernmental Panel on Climate Change, 2014,
325, 1599. IPCC Climate Change 2014 Synthesis Report, in Core
CSLF-Carbon Sequestration Leadership Forum, Active Writing Team, Pachauri, R. K., and Meyer, L., Eds:
and Completed CSLF Recognized Projects, 2015. http:// Cambridge Univ. Press. http://www.ipcc.ch/report/ar5/
www.cslforum.org/projects/jingbian_ccs.html syr/
Davis, T. L., Terrell, M. J., Benson, R. D., Cardona, R., Ivanova, A., Kashubin, A., Juhojuntti, N., Kummerow,
Kendall, R. R., and Winarsky, R., 2003, Multicomponent J., Henninges, J., Juhlin, Ch., Lüth, S., and Ivandic,
seismic characterization and monitoring of the CO2 flood M., 2012, Monitoring and volumetric estimation of
at Weyburn Filed, Saskatchewan: The Leading Edge, injected CO2 using 4D seismic, petrophysical data, core
22(7), 696–697. measurements and well logging: a case study at Ketzin:
Finley, R. J., 2014, An overview of the Illinois Basin Germany: Geophysical Prospecting, 60(5), 957–973.
–Decatur project: Greenhouse Gases: Science and Johnston, D. H., 2013, Practical applications of time-lapse
Technology, 4(5), 571–579. seismic data: 2013 SEG Distinguished Instructor Short

304
 Ma et al.

Course, DISC series, 13, 181–183. seismic data for an improved reservoir characterization,
Kiessling, D., Schmidt-Hattenberger, C., Schuett, H., SPE 90420, SPE Annual Technical Conference and
Schiling, F., Krueger, K., Schoebel, B., Danckwardt, Exhibition, Houston, Texas, USA, 26–29 September.
E., and Kummerow, J., 2010, The CO2SINK GROUP, Nakajima, T., and Xue, Z., 2013, Evaluation of a resistivity
Geoelectrical methods for monitoring geological model derived from time-lapse well logging of a pilot-
CO2 storage, First results from crosshole and surface- scale CO2 injection site, Nagaoka, Japan: International
downhole measurements from the CO2SINK test site at Journal of Greenhouse Gas Control, 12, 288–299.
Ketzin (Germany): International Journal of Greenhouse NDRC-The National Development and Reform
Gas Control, 4, 816–826. Commission of the People’s Republic of China, A
Kim, J., Matsuoka, T., and Xue, Z., 2011, Monitoring and catalogue for national promoting key low-carbon
detecting CO2 injected into water-saturated sandstone technologies, 2014. http://www.sdpc.gov.cn/zcfb/
with joint seismic and resistivity measurements: zcfbgg/201409/t20140905_625023.html
Exploration Geophysics, 42(1), 58–68. NDRC-The National Development and Reform
Lawton, D., 2010, Carbon capture and storage: Commission of the People’s Republic of China, 2015,
opportunities and challenges for geophysics: CSEG Implementation guide for national promoting key low-
Recorder, 35(6), 7–10. carbon technologies: China Financial and Economic
Li, G., 2003, 4D seismic monitoring of CO2 flood in a thin Publishing House.
fractured carbonate reservoir: The Leading Edge, 22(7), Onishi, K., Ueyama, T., Matsuoka T., Nobuoka, D., Saito,
690–695. H., Hiroyuki, A., and Xue, Z., 2009, Application of
Lumley, D., 2010, 4D seismic monitoring of CO 2 crosswell seismic tomography using difference analysis
sequestration: The Leading Edge, 29(2), 150–155 with data normalization to monitor CO 2 flooding in
Ma, J., Gao, L., and Morozov, I. B., 2009, Time- an aquifer: International Journal of Greenhouse Gas
lapse repeatability in 3C-3D dataset from Weyburn Control, 3(3), 311–321.
CO 2 sequestration project: 2009 Canadian Society Pevzner, R., Shulakova, V., Kepic, A., and Urosevic,
of Exploration Geophysicists CSEG/CSPG/CWLS M., 2011, Repeatability analysis of land time-lapse
Convention, AVO. Calgary, Canada. seismic data: CO2CRC Otway pilot project case study:
Ma, J., and Morozov, I. B., 2010, AVO modeling in Geophysical Prospecting, 59(1), 66–77.
of Pressure-Saturation effects in Weyburn CO 2 Reiner, D. M., 2016, Learning through a portfolio of carbon
Sequestration: The Leading Edge, 29(2), 178–183. capture and storage demonstration projects: Nature
Ma, J., Wang, X., Gao, R., Zhang, X., Wei, Y., Wang, Z., Energy, 1, 1–7, doi:10.1038/nenergy.2015.11
Ma, J., Huang, C., Zhao, X., Jiang, S., Liu, L., Li, L., Yu, Riazi, N., Lines, L., and Russell, R., 2013, Integration of
H., and Wang, H., 2014, Jingbian CCS Project, China: time-lapse seismic analysis with reservoir simulation,
Second Year of Injection, Measurement, Monitoring and GeoConvention 2013, Calgary, Canada.
Verification: Energy Procedia, 63, 2921–2938. Roggero, F., Ding, D. Y., Berthet, P., Lerat, O., Cap, J., and
Ma, J., and Zhang, X., 2010, Geophysical methods for Schreiber, P. E., 2007, Matching of production history
monitoring CO2 sequestration: Status, challenges and and 4D seismic data -application to the Girassol Field,
countermeasures: China Population Resources and Offshore Angola, SPE109929, SPE Annual Technical
Environment, 2010 China Sustainable Development Conference and Exhibition, Anaheim, California, USA,
Forum Special Issue, 20, 223–228. 11–14 November.
Martínez, J. M., and Schmitt, D. R., 2013, Anisotropic Rostrona, B., White, D., Hawkesc, C., and Chalaturnyka,
elastic moduli of carbonates and evaporates from the R., 2014, Characterization of the Aquistore CO2 project
Weyburn-Midale reservoir and seal rocks: Geophysical storage site, Saskatchewan, Canada: Energy Procedia,
Prospecting, 61(2), 363–379. 63, 2977–2984.
Matter, J. M., and Kelemen, P. B., 2009, Permanent storage Samsonov, S., Magdalena, Czarnogorska, M., and White,
of carbon dioxide in geological reservoirs by mineral D. J., 2015, Satellite interferometry for high-precision
carbonation: Nature Geoscience, 8, 837–841. detection of ground deformation at a carbon dioxide
Meadows, M. A., and Cole, S. P., 2013, 4D seismic storage site: International Journal of Greenhouse Gas
modeling and CO 2 pressure-saturation inversion at Control, 42, 188–199.
Weyburn Field, Saskatchewan: International Journal of Service, R., 2009, Carbon sequestration: Science, 325, 1644
Greenhouse Gas Control, 16S, S103–S117. –1645.
Mezghani, M., Fornel, A., Langlais, V., and Lucet, N., Shen, X., Ma, J., and Pan, B., 2009, Pressure-dependent
2004, History matching and quantitative use of 4D AVO response of fractured-aperture rock: CPS/SEG
305
 Geophysical monitoring technology for CO2 sequestration

2009 Conference and Exhibition, Geophysics’ Challenge, (Editor), Best Practices for Validating CO2 Geological
Opportunity, and Innovation, April 25–27, Beijing, Storage: Geoscience Publishing, Canada, 155–210.
China. White, D. J., 2013, Seismic characterization and time-lapse
Spetzler, J., Xue, Z., Saito, H., Nobuoka, D., Azuma, imaging during seven years of CO2 flood in the Weyburn
H., and Nishizawa, O., 2008, Case Story: Time-Lapse field, Saskatchewan, Canada: International Journal of
Seismic Crosswell Monitoring of CO 2 injected in an Greenhouse Gas Control, 16S, S78–S94.
Onshore Sandstone Aquifer: Geophysical Journal White, D. J., Roach, L. A. N., Roberts, B., and Daley, T.
International, 172, 214–225. M., 2014, Initial Results from Seismic Monitoring at
Sun, S., 2006, Geological problems of CO2 underground the Aquistore CO2 Storage Site, Saskatchewan, Canada:
storage and its significance on mitigating climate change: Energy Procedia, 63, 4418–4423.
China Basic Science, 8(3), 17–22. Wills, P. B., Hatchell, P. J., and Hansteen, F., 2009,
Ugalde, A., Villaseñor, A., Gaite, B., Casquero, S., Martí, D., Practical seismic monitoring of CO2 sequestration: 2009
Calahorrano, A., Marzán, I., Carbonell, R., and Estaún, A. SEG Summer Research Workshop, CO2 Sequestration
P., 2013, Passive Seismic Monitoring of an Experimental Geophysics, 23–27 August, Banff, Canada.
CO 2 Geological Storage Site in Hontomín (Northern Xue, Z., and Lei, X., 2006, Laboratory study of CO 2
Spain): Seismological Research Letters, 84(1), 75–84. migration in water-saturated anisotropic sandstone: based
Urosevic, M., Pevzner, R., Kepic, A., Shulakova, V., on P-wave velocity imaging: Exploration Geophysics,
Wisman, P., and Sharma, S., 2010, Time-lapse seismic 37(1), 10–18.
monitoring of CO 2 injection into a depleted gas Xue, Z., Tanase, D., and Watanebe, J., 2006, Estimation
reservoir—Naylor Field, Australia: The Leading Edge, of CO 2 saturation from time-lapse CO 2 well logging
29(2), 164–169. in an onshore aquifer, Nagaoka, Japan: Exploration
Verdon, J. P., Kendall, J., White, D., Angus, D. A., Fisher, Q. Geophysics, 37(1), 19–29.
J., and Urbancic, T., 2010, Passive seismic monitoring of Yang, Y., Ma, J., and Li, L., 2015, Research progress
carbon dioxide storage at Weyburn: The Leading Edge, of carbon dioxide capture and storage technique and
29(2), 200–206. 4D seismic monitoring technique: Advances in Earth
Verkerke, J. L., Williams, D. J., and Thoma, E., 2014, Science, 30(10), 1119–1126.
Remote sensing of CO 2 leakage from geologic Zhang, L., Ren, B., Huang, H., Li, Y., Ren, S., Chen, G.,
sequestration projects: International Journal of Applied and Zhang, H., 2015, CO 2 EOR and storage in Jilin
Earth Observation and Geoinformation, 31, 67–77. oilfield China: Monitoring program and preliminary
Wang, Z., Cates, M. E., and Langan, R. T., 1998, Seismic results: Journal of Petroleum Science and Engineering,
monitoring of a CO2 flood in a carbonate reservoir: A 125, 1–12.
rock physics study: Geophysics, 63(5), 1604–1617.
Wei, J., Di, B., and Ding, P., 2013, Effect of crack aperture
on P-wave velocity and dispersion: Applied Geophysics,
10(2), 125–133. Ma Jin-Feng graduated from Department of Applied
White, D. J., 2009, Monitoring CO2 storage during EOR Geophysics, Changchun College of
at the Weyburn-Midale Field: The Leading Edge, 28(7), Geology in 1985. In 1991 and 1998,
838–842. he earned his MS and Ph.D degree in
White, D. J., Hirsche, K., Davis, T., et al., 2004, Theme Department of Geology, Northwest
2: Prediction,Monitoring and Verification of CO 2 University. He is a professor of geophysics
movements, in Wilson M., and Monea M, eds. IEA GHG at Department of Geology, Northwest
CO2 Monitoring and Storage Project Summary Report University. His main research interests
2000–2004: PTRC, Regina. are geophysical monitoring for CO2 sequestration, seismic
White, D. J., 2012, Geophysical Monitoring in Hitchon, B. reservoir prediction.

306