International Journal of Greenhouse Gas Control 51 (2016) 357–368

Contents lists available at ScienceDirect

International Journal of Greenhouse Gas Control

journal homepage: www.elsevier.com/locate/ijggc

Effect of pressure on corrosion behavior of X60, X65, X70, and X80

carbon steels in water-unsaturated supercritical CO2 environments
Minghe Xu a , Weihong Li b , Yi Zhou c , XiaoXian Yang a , Zhe Wang a,∗ , Zheng Li a
a
State Key Laboratory of Power Systems, Department of Thermal Engineering, Tsinghua University, Beijing, PR China
b
Institute of Gas Turbine, Department of Thermal Engineering, Tsinghua University, Beijing, PR China
c
School of Materials Science and Engineering, Tsinghua University, Beijing, PR China

a r t i c l e i n f o a b s t r a c t

Article history: A systematic study was performed to investigate the effect of pressure and water content on the general
Received 8 February 2016 corrosion and localized attack of carbon steels during the transport of CO2 for the CCS industry. Corrosion
Received in revised form 2 June 2016 experiments were conducted on X60, X65, X70, and X80 carbon steels in water-unsaturated supercritical
Accepted 6 June 2016
CO2 with SO2 and O2 in conditions of 50 ◦ C and 8–12 MPa. The general corrosion rates increased with water
Available online 20 June 2016
content and ranged from 0.02 mm/year to 0.31 mm/year at 8 MPa and from 0.01 mm/year to 0.94 mm/year
at 10 MPa. Localized corrosion was determined by measuring the penetration rate and pitting factor.
Keywords:
The penetration rates ranged from 0.2 mm/year to 3.25 mm/year at 8 MPa and from 0.04 mm/year to
Supercritical CO2 transport
Pressure
6.02 mm/year at 10 MPa. Meanwhile, the pitting factors ranged from 4 to 70 at 8 MPa and from 3 to 43 at
Water content 10 MPa, with water contents varying from 1600 ppm to 3000 ppm. The degrees of general corrosion and
Localized corrosion localized attack were higher at 8 MPa than those at 10 MPa, with water contents ranging from 1600 ppm
Carbon steel to 2600 ppm, but lower at a water content of 3000 ppm.
© 2016 Elsevier Ltd. All rights reserved.

1. Introduction
Carbon capture and storage (CCS) has been recently identified
as a possible means to reduce global carbon emissions. In the CCS
process, CO2 is captured from power plants or other industrial
factories, transmitted to the storage site, and deposited within
a geological reservoir (Zhang et al., 2006). Considering the large
quantities of CO2 transmitted over long distances, carbon steel
pipeline transport is recognized as the most reasonable and cost-
effective option. CO2 transported through pipelines is typically
compressed into a supercritical state (greater than 7.38 MPa and
31.1 ◦ C) to prevent pressure drops and two-phase flow regime (Gale
and Davison, 2004).
Since CO2 transportation is the critical connecting link between
the capture and storage processes, this process should be eco-
nomical, effective, and environmentally safe. However, pipeline
networks implemented to transport supercritical CO2 are likely to
suffer significant levels of internal pipeline corrosion when free
water exists within the system (Choi et al., 2010). The aqueous
phase forming on the surface of the internal pipeline will be sat-
∗ Corresponding author.
E-mail addresses: zhewang@tsinghua.edu.cn, wangzhe988@gmail.com
(Z. Wang).
urated with CO2 , creating carbonic acid (H2 CO3 ), hence posing a
corrosive threat to pipeline steel. Other contaminants present in
the CO2 stream, such as O2 , H2 O, SOx , NOx , Ar, and N2 , also influ-
ence and potentially accelerate the corrosion process. For example,
acidic impurities, such as SOx , may dissolve in water to become sul-
furous acid and sulfuric acid, which cause a considerable threat to
the durability of pipelines (Xiang et al., 2011, 2013b). The species
and contents of impurities in transported CO2 vary depending on
the application of different combustion and flue gas cleaning tech-
nologies (de Visser et al., 2008).
CO2 transportation is a vital aspect of the CCS cycle, but the study
of transportation process safety has been somewhat overlooked
before (Hua et al., 2015a; Xiang et al., 2013c). Fortunately, more
researchers have recently focused on the internal pipeline corro-
sion issue during CO2 transportation process (Hua et al., 2014a,
2014b, 2015b, 2015c, 2015d; Sim et al., 2013a, 2013b; Xiang et al.,
2012a, 2012b, 2013a; Zhang et al., 2012). A summary of recent stud-
ies is provided in Table 1. Several researchers investigated the effect
of impurities on the corrosion behavior of carbon steel pipelines
(Hua et al., 2015a; Sim et al., 2013a; Xiang et al., 2011, 2012a),
whereas other researchers concentrated on the effects of opera-
tion conditions, such as temperature and pressure (Choi et al., 2013;
Hua et al., 2014a; Xiang et al., 2012a, 2012b, 2013a). However, no
general consensus on the actual water contents and operation con-

http://dx.doi.org/10.1016/j.ijggc.2016.06.002
1750-5836/© 2016 Elsevier Ltd. All rights reserved.
358 M. Xu et al. / International Journal of Greenhouse Gas Control 51 (2016) 357–368

Table 1
Summary of literature relating to corrosion behavior in pressurized CO2 in recent years.a

Steel p/bar T/◦ C t/h I CR/(mm/y) Ref.

X70 100 50 120 H2 O, SO2 , O2 0.005–1.46 Xiang et al. (2012a)

X70 100 50 24–192 H2 O, SO2 , O2 0.7–1.94 Xiang et al. (2012b)
X70 100 25–93 120 H2 O, SO2 0.7–2.95 Xiang et al. (2013a)
X70 120 50 454 H2 O, SO2 , O2 1.16 Xiang et al. (2013b)
X70 100 50 288 H2 O, SO2 , O2 0.2–0.876 Xiang et al. (2011)
X65 80 35–50 14–48 H2 O 0–0.112 Hua et al. (2015a)
X65 80 50 6.5–48 H2 O 4–10.8 Hua et al. (2014a)
X65 80 35 48 H2 O, SO2 , O2 0–0.7 Hua et al. (2015b)
X65 80 35 96 H2 O, SO2 , O2 0.25–0.7 Hua et al. (2014b)
X65, 5Cr 80 35 48 H2 O, O2 0.3–1.7 Hua et al. (2015c)
Carbon steel 80 40 168 H2 O 0.045–2.48 Sim et al. (2013b)
X65 40–80 50 24 H2 O 0.2–20 Choi et al. (2013)
X65 80 50 24 H2 O, O2 0.3–1 Choi et al. (2010)
a
p: pressure; T: temperature; t: test period; I: impurities; CR: corrosion rate.

ditions to prevent or reduce corrosion to appreciable levels has Table 2

Main elemental composition (wt.%) of four carbon steels.
been achieved.
Among all potential impurities, water content is the most critical X60 X65 X70 X80
factor because of the ability water has in allowing the occurrence C 0.068 0.07 0.088 0.039
of corrosion and potentially enhancing the degree of corrosion Si 0.15 0.34 0.32 0.174
attack. Determining the influence of water content on corrosion S 0.001 0.002 0.002 0.001
and the upper limit of water in the presence of different impurities Mn 0.989 1.039 1.538 1.830
Mo 0.002 0.105 0.004 0.290
and in various operation conditions has been the focus of certain
Cr 0.034 0.224 0.176 0.281
studies (Xiang et al., 2012a; Hua et al., 2014a; Sim et al., 2013b). Cu 0.010 0.014 0.192 0.222
Xiang et al. (2012a) observed the increasing trend of the corro- Fe Balance
sion rate (CR) with the increase in the water content at 10 MPa
and 50 ◦ C. They indicated the existence of a critical relative humid-
ity, above which the CR of the X70 steel increased significantly et al. (2012) reported that the corrosion mechanism of X65 carbon
with the water content. The critical relative humidity was sug- steel corrosion in CO2 -saturated water in low CO2 partial pressure
gested to be between 50% and 60%. However, they did not employ (1 MPa) was basically the same in comparison with the corrosion in
the upper limit of water to avoid corrosion occurrence. Hua et al. supercritical CO2 (9.5 MPa), while the corrosion rate in supercritical
(2014a) investigated the corrosion behavior of the X65 carbon steel CO2 was much higher. However, Choi and Nešić (2011) pointed out
in supercritical CO2 -saturated water and water-containing super- that the corrosion rates of carbon steel exposed to CO2 -saturated
critical CO2 at 8 MPa, with the temperature of 35 ◦ C and 50 ◦ C. They water with different CO2 partial pressures (4, 6, and 8 MPa) at 50 ◦ C
revealed that a water content of 1600 ppm can prevent corrosion did not significantly increase with CO2 partial pressure. Regard-
at 50 ◦ C, but a water content of 300 ppm cannot even prevent cor- less, assessing the dependence of corrosion behavior of various
rosion at 35 ◦ C. The experiments were conducted in CO2 with the carbon steels on pressure is essential for the purpose of provid-
water as the sole impurity. However, as other impurities like SO2 ing a reference for the selection of operation conditions in the CCS
and O2 usually exist in the actual CO2 stream, the upper limit of industry.
water content is very likely to be less than 1600 ppm (Hua et al., Consequently, the present study aims to understand the effect of
2015b). Another research series of tests were conducted with the pressure on the general and localized corrosion for X60, X65, X70,
addition of 1000 ppm O2 into the CO2 stream (Hua et al., 2015c). and X80 carbon steels exposed to water-unsaturated supercritical
Results showed that no corrosion was observed when the water CO2 with impurities of 3000 ppm SO2 and 1000 ppm O2 for 72 h. The
content was less than 1200 ppm for X65 and 5Cr steels. Neverthe- effect of pressure on the upper limit of water content, below which
less, the existing literature proposed that the extent of corrosion no appreciable levels of corrosion are observed, is also discussed in
attack cannot be neglected not only in water-saturated supercriti- this study.
cal CO2 , but also in water-unsaturated supercritical CO2 (Hua et al.,
2014a, 2014b, 2015b, 2015c, 2015d; Sim et al., 2013a, 2013b; Xiang 2. Material and methods
et al., 2012a, 2012b, 2013a; Zhang et al., 2012).
In terms of the operation conditions, only a few studies inves- 2.1. Materials and preparation
tigated the influence of temperature and pressure on internal
pipeline corrosion (Choi et al., 2013; Hua et al., 2014a; Xiang et al., The X60, X65, X70, and X80 pipeline steels were
2013a). Xiang et al. (2013a) investigated the effect of temperature machined into tablet specimens with a size of approximately
on X70 steel corrosion in water-saturated supercritical CO2 . The 20 mm × 20 mm × 2.6 mm. The chemical composition of the
average corrosion rates initially increased and then decreased with pipeline steels is listed in Table 2. The microstructures of these
temperature, reaching the peak value at 75 ◦ C. However, localized steels are shown in Fig. 1. The chemical composition of the four
attack was not evaluated in the research to obtain a full view of materials is close, while the microstructure is different. Generally,
the extent of corrosion. However, Hua et al. (2014a) illustrated that the four materials all contain the ferrite phase, but with differ-
the temperature of 35 ◦ C tended to induce more severe degrada- ent ferrite grain forms. X60 steel is composed of large volume
tion of X65 steel than 50 ◦ C in a water-unsaturated supercritical quasi-polygonal ferrite in small grain size and carbon-enriched
CO2 stream. bainite (indicated by the circles in Fig. 1a), while X65 steel is
When consolidating the literature on the effect of pressure composed of polygonal ferrite and pearlite. The grain size of X65
on the corrosion of carbon steel in water-containing supercriti- steel polygonal ferrite is larger than that of X60 steel ferrite, and
cal CO2 typical in CCS, sparse research has been conducted. Zhang the volume percentage of X65 steel pearlite is the highest of the
 M. Xu et al. / International Journal of Greenhouse Gas Control 51 (2016) 357–368 359

Fig. 1. Microstructure of X60, X65, X70 and X80 steel. (a) X60 steel; (b) X65 steel; (c) X70 steel; (d) X80 steel.

four materials. X70 steel is mainly composed of acicular ferrite, Table 3

Test conditions.
granular ferrite, and a small percent of polygonal ferrite, while X80
steel is composed of acicular ferrite, polygonal ferrite, granular No. T/◦ C p/MPa t/h H2 O/ppm SO2 a /ppm O2 a /ppm Speed of
ferrite, a small percent of granular pearlite, and carbon-enriched rotation/rpm
bainite (indicated by the circles in Fig. 1d). The volume percent 1 50 8.0 72 1600 3000 1000 120
of X80 steel polygonal ferrite seems larger than that of X70 steel 2 50 8.0 72 2000 3000 1000 120
polygonal ferrite. 3 50 8.0 72 2600 3000 1000 120
4 50 8.0 72 3000 3000 1000 120
Prior to the corrosion tests, all specimens were polished succes-
5 50 10.0 72 1600 3000 1000 120
sively on finer grades of silicon carbide papers up to 1200 grit and 6 50 10.0 72 2000 3000 1000 120
then washed with acetone and anhydrous ethanol. Samples were 7 50 10.0 72 2600 3000 1000 120
subsequently stored in a desiccator for at least 24 h and weighed 8 50 10.0 72 3000 3000 1000 120
9 50 12.0 72 2000 3000 1000 120
using an electronic balance to an accuracy of 0.01 mg before being
a
hung inside the autoclave. Two samples for each type of pipeline SO2 and O2 concentrations herein are the theoritical gas concentration and may
steel, i.e., a total of eight samples, were placed within the auto- subject to undergo minimal change due to the SO2 oxidation.

clave for each individual test, generating a total surface area of

approximately 20 cm2 for each type of pipeline steel exposed to
the supercritical CO2 mixture.
2.2. Autoclave testing procedure
Fig. 2 provides the schematic illustration of the experimental
system layout. The entire system consists of a 2.2 L capacity auto-
clave, a CO2 –SO2 –O2 mixture cylinder, a pump, a temperature and
pressure control system, and a tail gas treatment facility.
After deaerating the autoclave by CO2 for two hours, the spec-
imens were suspended in the middle part of the autoclave using
specimen holders. The deionized water, which was deaerated by
saturating the solution with CO2 for a minimum of 24 h prior to
testing was then injected into the autoclave using a pipette gun.
The amount of water for each experiment was selected to gen-
erate water-unsaturated conditions (i.e., water content below the
solubility limit) because actual CO2 pipelines implemented in the
industry generally contain unsaturated water to avoid rapid degra-
dation. After the addition of water, the CO2 –SO2 –O2 mixture with
a well-known concentration, which was provided by a local com-
pany, was pumped into the autoclave, then heated to the desired
temperature, and pressurized to the desired pressure. The rotation
rate of the rotor in the autoclave was set to 120 rpm to simu-
late the flow pattern of the CO2 stream in the practical pipelines.
Experiments were conducted at pressures of 8, 10, and 12 MPa, tem-
perature of 50 ◦ C, and various water contents of 1600–3000 ppm.
The test conditions are summarized in Table 3. The test period lasts
for 72 h for each test. The initial nominal SO2 and O2 concentrations
are 3000 and 1000 ppm (mol), respectively. It is mentioned that the
CO2 –SO2 –O2 is mixed in professional company to ensure the SO2
and O2 concentrations are highly accurate. Although the SO2 –O2
mixture is thermodynamic unstable, it is assumed herein that the
change in the actual gas concentration is minimal in the very short
storage time (usually less than 48 h), because the SO2 oxidation,
360 M. Xu et al. / International Journal of Greenhouse Gas Control 51 (2016) 357–368

Fig. 2. Corrosion test rig schematic diagram.

of which the reaction rate is kinetically controlled, is an extremely

slow process without the catalyst.
The solubility data of water in pure CO2 was calculated using the
model developed by Spycher et al. (2003). The solubility of water in
pure CO2 at 10 MPa is 5470 ppm, which is clearly higher than that
of 3470 ppm at 8 MPa, but similar to that of 5550 ppm at 12 MPa.
Given that the solubility of water in the CO2 –SO2 –O2 mixture has
not been reported in existing literature, the solubility of water in
the mixture is believed to be similar to that in pure CO2 , based on
the assumption that the influence of SO2 and O2 on the solubility
of water is small, as reported in previous studies (Xiang et al., 2011,
2012a). In addition, the degree of moisture in the CO2 stream is
also characterized by relative humidity, which is the ratio of the
practical water content to the water solubility in the CO2 stream.

2.3. Methods

The weight loss method is applied to determine the average CR.

Electrochemical measurements are relatively difficult to employ,
as the water film generated in the corrosion process would be too
thin for the correct operation of the electrochemical experimen-
tal apparatus. After corrosion, the corroded samples were exposed
to the Clarke solution (20 g Sb2 O3 , 50 g SnCl2 , and concentrated
hydrochloric acid to make 1 L) to remove the corrosion products
(ASTM Standard G1-03, 2011). Eq. (1) was used to calculate the
average CR, as follows:
 
  8.76 × 104 mm · h/cm · year × weight loss (g)
CR mm/year = . (1)
area (cm2 ) × density(g/cm3 ) × immertion time (h)

The product morphology of the samples was analyzed by scan-

ning electron microscopy (SEM). The elemental composition and
crystallization of the corrosion product layers were detected by
X-ray diffraction (XRD) and energy dispersive X-ray spectroscopy Fig. 3. Average corrosion rates variation with water contents for carbon steel sam-

(EDS). A 3D measuring laser microscope (LEXT OLS4100, OLYMPUS) ples exposed to pressure of (a) 8 MPa; (b) 10 MPa. , X60 steel; , X65 steel;
was also employed to examine the sample surface and determine
, X70 steel; , X80 steel.
the extent of localized corrosion after cleaning the product scale.

3. Results and discussion
3.1. Effect of pressure on the general corrosion behavior of carbon
steels at pressures of 8–12 MPa with various water contents
The dependence of the general corrosion behavior of X60, X65,
X70 and X80 carbon steels on pressure, water content and steel
type in condition of pressures of 8–12 MPa and water contents of
1600–3000 ppm is discussed in detail as follows. The general cor-
rosion behavior is characterized by calculating average corrosion
rates, analyzing corroded surface morphology and product compo-
sition.
3.1.1. Average corrosion rates of carbon steels at pressures of 8
and 10 MPa with various water contents
Fig. 3 shows the average corrosion rates of X60, X65, X70, and
X80 carbon steels exposed to water-unsaturated dense phase CO2
at pressures of 8 and 10 MPa and a temperature of 50 ◦ C.
Fig. 3a shows the average corrosion rates of carbon steels
exposed to 8 MPa. The corrosion rates of X60, X65, X70, and
 M. Xu et al. / International Journal of Greenhouse Gas Control 51 (2016) 357–368
X80 carbon steels collectively range between 0.02 mm/year and
0.31 mm/year. At a water content of 1600 ppm, the corrosion rates
of X60, X65, X70, and X80 carbon steels vary from 0.025 mm/year
to 0.06 mm/year. The corrosion rates of X80 and X65 carbon steels
exceed 0.1 mm/year at a water content of 2000 ppm or more. Mean-
while, the corrosion rates of X60 and X70 carbon steels surpass
0.1 mm/year at a water content of 2600 ppm or more. In contrast to
the previous research conducted by Hua et al. (2014a), the authors
illustrated that no corrosion was observed on the X65 steel at a
water content of 1600 ppm, whereas a corrosion rate of approx-
imately 0.014 mm/year was detected on the X65 steel at a water
content of 2650 ppm when exposed to pure supercritical CO2 at
8 MPa and 50 ◦ C. Therefore, the existence of SO2 and O2 is con-
firmed to decrease the endurance of water content for carbon steel
pipeline networks.
Fig. 3a shows that the corrosion rates of X60 and X65 steels
monotonously increase with the water contents. However, the
corrosion rates of X70 steel slightly decrease from 0.04 mm/year
to 0.02 mm/year with water contents varying from 1600 ppm to
2000 ppm and then gradually increase with water contents ranging
from 2000 ppm to 3000 ppm. Meanwhile, the corrosion rates of X80
steel generally increase with water contents less than 2600 ppm
and slightly decrease from 0.31 mm/year to 0.30 mm/year with
water contents varying from 2600 ppm to 3000 ppm. In spite of
the special corrosion behavior of X70 steel at a water content of
2000 ppm and X80 steel at a water content of 3000 ppm, which
may be partly due to contingency factors, such as measurement
errors, the general corrosion behavior of X60, X65, X70, and X80
carbon steels can be concluded to be typically enhanced by water
contents. Such water feature, which accelerates the average corro-
sion rates, is also revealed by previous studies (Hua et al., 2014a;
Sim et al., 2013b; Xiang et al., 2012a).
Fig. 3b shows that the average corrosion rates of carbon steels
at 10 MPa collectively range from 0.01 mm/year to 0.94 mm/year
with the water contents. At water contents of 1600–2600 ppm,
the corrosion rates of X60, X65, X70, and X80 steels are all
less than 0.1 mm/year. However, the corrosion rates of carbon
steels vary from 0.81 mm/year to 0.94 mm/year at a water con-
tent of 3000 ppm. For all carbon steels, the corrosion rates show
a monotonous increasing trend with the water contents, which
strongly agrees with previous research (Xiang et al., 2012a).
3.1.2. Comparison of the general corrosion behavior of carbon
steels at pressures of 8 and 10 MPa
Fig. 3 shows that the general corrosion behavior of X60, X65,
X70, and X80 carbon steels at 8 MPa vary widely at water contents
greater than 2000 ppm, with X80 steel exhibiting the most severe
degradation, followed by X65, X60, and X70 steels. However, X60,
X65, X70, and X80 carbon steels exhibit a similar corrosion behavior
at 10 MPa, which is supported by the fact that the average corrosion
rates are similar upon exposure to the same water contents.
Fig. 4 shows a separate comparison of the general corrosion
behavior of X60, X65, X70, and X80 steels exposed to the dense
phase CO2 mixtures at 8 and 10 MPa. The corrosion rates of X60,
X65, and X80 steels at 8 MPa are higher than those at 10 MPa
exposed to water contents of 1600, 2000, and 2600 ppm and are
considerably lower when exposed to water content of 3000 ppm.
Meanwhile, the corrosion rates of X70 steel at 8 MPa are greater
than those at 10 MPa with water contents of 1600 and 2600 ppm,
are similar at a water content of 2000 ppm, and are significantly
lower at a water content of 3000 ppm. One explanation for the
fact that the corrosion rates of X60, X65, X70, and X80 steels at
8 MPa are generally higher than those at 10 MPa with water con-
tents less than 2600 ppm, is the higher relative humidity as well as
the higher degree of moisture for the corrosive medium at 8 MPa
361
than that at 10 MPa, resulting from the lower water solubility in the
CO2 mixtures at 8 MPa than that at 10 MPa.
Moreover, the trend of the variation of the corrosion rates with
the water contents at 8 and 10 MPa exhibits an evident difference.
Generally, the corrosion rates at 8 MPa gradually increase with the
water contents. By contrast, the corrosion rates at 10 MPa slowly
increase when the water content is less than 2600 ppm and signif-
icantly increase when the water content is greater than 2600 ppm.
At 10 MPa, the abrupt change of corrosion rate exists at the water
content point of 2600 ppm, with the corrosion rate increasing sig-
nificantly above the said point. However, the corrosion rates at
8 MPa do not reveal the existence of transition, which is charac-
terized by critical relative humidity by previous researchers (Xiang
et al., 2012a), in the water content point in the present study.
3.1.3. Effect of pressure on the corrosion behavior of carbon steels
at a water content of 2000 ppm
Further investigation into the effect of pressure on the corrosion
behavior of carbon steels at a water content of 2000 ppm is con-
ducted with another experiment exposed to the same CO2 mixtures
at 12 MPa. The effect of pressure on the corrosion behavior of car-
bon steels at a water content of 2000 ppm is presented in Fig. 5. The
average corrosion rates of X60, X65, and X80 steels are shown to be
the highest at 8 MPa, but the average corrosion rates are at 10 and
12 MPa similar, partly because of the fact that the water solubility
in the CO2 mixtures is the lowest at 8 MPa, and the water solubility
is similar at 10 and 12 MPa. These result in the highest degree of
moisture for the corrosive medium at 8 MPa and the similar degrees
of moisture at 10 and 12 MPa.
In contrast to the study of Zhang et al. (2012), the average cor-
rosion rate of X65 steel under a pure supercritical CO2 condition
(9.5 MPa) was determined to be higher than that at a low CO2 par-
tial pressure (1 MPa) in a CO2 -saturated water solution, because
the CO2 solubility in water is much higher under a supercritical CO2
condition, which consequently enhances the concentrations of car-
bonic species (H2 CO3 , HCO3 − , and CO3 2− ) in the solution. However,
such feature of pressure that enhances the corrosion rates does not
apply to the water-unsaturated CO2 condition in the present study,
because the dependence of corrosion rates on pressure seems to be
more determined by the relative humidity as well as the degree of
moisture instead of the solubility of CO2 in water phase.
3.2.4. Comparison of X65 steel corrosion behavior exposed to 8
and 10 MPa by means of photographic investigation
Photographic investigation is employed to compare the carbon
steel corrosion behavior at 8 and 10 MPa, and X65 steel is taken for
an example to present the typical corrosion behavior of all carbon
steels. Fig. 6 shows the photographed examples of the corroded
surfaces of the X65 steel samples exposed to different water con-
tents at 8 MPa (Fig. 6a) and 10 MPa (Fig. 6b). In contrast to the
non-uniformly corroded surfaces shown in Fig. 6a, the surfaces for
X65 steel samples exposed to water contents of 1600 and 2000 ppm
at 10 MPa shown in Fig. 6b are noted to be uniformly covered by a
layer of product and with dim metal luster. The surfaces exposed to
a water content of 2600 ppm at 10 MPa are composed of an area of
dark-colored surfaces and a large area of surfaces with dim metal-
lic luster, with the existence of dark-colored surfaces indicating
the accelerated degradation compared with the samples exposed
to 1600 and 2000 ppm. The surfaces exposed to a water content of
3000 ppm at 10 MPa are significantly corroded with a large amount
of product formation, which results in abrupt high corrosion rates
of approximately 0.9 mm/year.
In contrast to the surfaces of X65 steel samples exposed to
10 MPa, those exposed to 8 MPa are presented in Fig. 6a. These
surfaces are composed of dark-colored surfaces and surfaces with
metallic luster. The dark-colored surfaces are visibly covered by
362 M. Xu et al. / International Journal of Greenhouse Gas Control 51 (2016) 357–368
Fig. 4. Comparison of general corrosion behavior of carbon steels exposed to pressure of
a product film, whereas the surfaces with metallic luster need to
be detected by a microscope to determine whether the products
exist on the surfaces. Nevertheless, the dark-colored surfaces are
assumed to contribute more in promoting the degradation of the
entire surface than the surfaces with metallic luster. Meanwhile,
the area of dark-colored surfaces on the samples increases with
water contents at 8 MPa, which is consistent with the gradually
increasing trend of the corrosion rate with the water contents.
The phenomenon wherein the corroded surfaces are composed
of dark-colored and discolored sections was also observed by Hua
et al. (2015d). Corrosion tests were conducted for the X65 steel
samples exposed to the water-saturated CO2 phase at 35 ◦ C and
8 MPa. Hua et al. (2015d) reported that a distinct difference lies in
the morphology, size, and distribution of FeCO3 crystals on the steel
surface in both regions, thus resulting in two contrasting mecha-
nisms of corrosion attack and product formation behavior. They
argued that one way of explaining the difference in FeCO3 crys-
tals can be related to the extent of water, which condensed onto
the steel in each location. In a dark-colored region, a significant
amount of water condensed onto the surface, whereas relatively
smaller water droplets have condensed onto the steel and left the
surface discolored in the region with dim metallic luster.
Comparing the corrosion behavior and product formation of the
surfaces for the samples exposed to 8 and 10 MPa in the present
study, not only the dependence of the average corrosion rates on the
water contents differs, but also the behavior of product formation
on the surfaces, although the test conditions in the present study
, 8 MPa; and , 10 MPa. (a) X60 steel; (b) X65 steel; (c) X70 steel; (d) X80 steel.
are all in a supercritical CO2 phase at 8 and 10 MPa. The macro mor-
phology difference at water contents less than 2600 ppm between
8 MPa and 10 MPa are speculated to originate from the difference
in water condensation behavior. Water condensed non-uniformly
onto the steel surfaces at 8 MPa. By contrast, water condensed uni-
formly onto the steel surfaces at 10 MPa at water contents less than
2600 ppm.
3.1.5. Analysis of corrosion product morphology and composition
Fig. 7 shows the SEM images of the corroded surface exposed to
varying water contents and pressures. The labels “A–I” represent
the area where EDS tests are conducted to analyze the chemical
composition of the product. The EDS tests results are presented in
Table 4. The product composition mainly consisted of the elements
Fe, S, C, and Mn and was enriched with the element O. The ele-
ment of C might indicate the existence of Fe3 C and FeCO3 , and the
element of Mn came from the steel substrate. The surfaces of sam-
ples exposed to 10 MPa at water contents of 1600 and 2000 ppm
shown in Fig. 7a and b are covered by a flat and smooth product
layer, although the surfaces are with dim metallic luster. Similar
to the photographic investigation of the corroded samples exposed
to 8 MPa, the surface morphology exhibits at least two patterns in
Fig. 6c and d. The surface morphology shown in Fig. 7c consists of a
thick and dense product layer, a thin and sparse product layer, and
a stacking pattern of the product, whereas the surface morphology
shown in Fig. 7d consists of a thick layer and a thin layer. The thin
 M. Xu et al. / International Journal of Greenhouse Gas Control 51 (2016) 357–368 363

Table 4 lar boundary. Upon exposure to a water content of 3000 ppm, the
Elemental composition (mol.%) of products in region A–I labeled in Fig. 7.a
surfaces are covered by a larger amount of product and an irregu-
Region Fe S C O Mn lar pattern of product at 10 MPa, as shown in Fig. 7f. By contrast,
A 9.80 17.17 9.51 63.35 – the surfaces of the samples exposed to 8 MPa exhibited a small
B 41.09 3.06 33.12 22.73 – patch pattern of product and a thick product layer with porous
C 12.71 6.00 15.57 66.02 – appearance, as shown in Fig. 7h.
D 19.48 16.70 – 63.82 – XRD is utilized to determine the composition and crystallogra-
E 7.99 9.77 10.02 72.77 –
phy of the corrosion product. Fig. 8 shows the XRD results of the
F 10.33 11.27 9.57 68.63 0.20
G 28.69 6.57 9.16 55.17 0.41 product layers on X65 steel surfaces of samples exposed to varying
H 34.60 6.85 – 58.55 – water contents and pressures. Apparently, hydrated ferrous sul-
I 6.24 14.97 26.39 52.40 – fates (FeSO4 ·xH2 O) are the main components of the product films.
a
The regions ‘A–I’ labeled in Fig. 7 represent the area where the EDS tests are However, as the XRD was done using Cu-source for the X-ray and
conducted to analyze the chemical composition (mol.%) of the product. Bragg-Brentano geometry instead of low angle technique, the full
phase information (for example, FeCO3 , Fe3 C phase) in thin layers
might be missing.

3.2. Effect of pressure on the localized corrosion behavior of

carbon steels at pressures of 8 and 10 MPa with various water
contents

Localized corrosion was measured using a 3D measuring laser

Fig. 5. Effect of pressure on average corrosion rates of carbon steels exposed to
water content of 2000 ppm.
product layer shown in Fig. 7d sparkled with a particle pattern of
the product.
With the water content increasing to 2600 ppm at 10 MPa, the
corroded surfaces shown in Fig. 7e are not only covered by a smooth
product layer, but also contained the radial pattern with circular
boundary, which may indicate the water condensation occurrence.
By contrast, the surfaces of samples exposed to 8 MPa shown in
Fig. 7g exhibited stacking patch and radial patterns with a circu-
microscope (LEXT OLS4100, OLYMPUS) with specimens after
removing the product films. Multiple measurements are conducted
on randomly selected areas of the corroded surface to determine
the 10 deepest pits depth in accordance with the American Society
of Testing Materials (ASTM) Standard G46-94 (ASTM Standard G46-
94, 2005). The penetration rates are calculated from the average
penetration rates of the 10 deepest pit depth.
Figs. 9 and 10 present the evaluation of the localized corrosion
of X60, X65, X70, and X80 steel samples exposed to 8 and 10 MPa,
respectively. Fig. 9a shows the penetration rate calculated from the
extrapolation of the 10 deepest pit depth within the 72 h test period
for samples exposed to CO2 mixtures at 8 MPa. Penetration rate is
calculated herein under the assumptions that pitting initiates at
the beginning of the experiment and its growth follows the linear
law (Sim et al., 2013a). Fig. 9a shows that almost all of the pene-
tration rates exceed 0.1 mm/year and the penetration rate of X60
steel at a water content of 3000 ppm reaches 3.25 mm/year, cor-
responding to a 27 ␮m deep pit, the image of which is shown in
Fig. 11a. In terms of the type of pipeline steel, the penetration rate

Fig. 6. Photographic images of corroded X65 steel samples exposed to varying water contents at pressure of 8 MPa and 10 MPa. (a) 8 MPa; (b) 10 MPa.
364 M. Xu et al. / International Journal of Greenhouse Gas Control 51 (2016) 357–368

Fig. 7. SEM images of corroded carbon samples exposed to varying water contents. The labels ‘A–I’ represent the area where the EDS tests are conducted to analyze the
chemical composition of the product and the EDS tests results are presented in Table 4. (a) X70 steel at water content of 1600 ppm exposed to 10 MPa; (b) X80 steel at water
content of 2000 ppm exposed to 10 MPa; (c) X70 steel at water content of 1600 ppm exposed to 8 MPa; (d) X80 steel at water content of 2000 ppm exposed to 8 MPa; (e) X60
steel at water content of 2600 ppm exposed to 10 MPa; (f) X65 steel at water content of 3000 ppm exposed to 10 MPa; (g) X60 steel at water content of 2600 ppm exposed
to 8 MPa; (h) X65 steel at water content of 3000 ppm exposed to 8 MPa.

of X60 steel increases with the water contents, exhibiting a trend
similar to the average corrosion rate variation. By contrast, the pen-
etration rate of X65 steel generally increases with water contents
greater than 2000 ppm. For X80 steel, the penetration rate varia-
tion with the water contents presents a similar trend comparable
to the average corrosion rate variation with the water contents. In
other words, the penetration rate for X80 steel initially increases
with water contents up to 2600 ppm and then decreases with water
contents varying from 2600 ppm to 3000 ppm. In contrast, the pen-
etration rate of X70 steel initially decreases with water contents
ranging from 1600 ppm to 2000 ppm and then increases with water
contents from 2000 ppm to 2600 ppm, followed by a stable value
with water contents from 2600 ppm to 3000 ppm. Overall, the X80
steel suffers the most severe mass loss degradation (as shown in
Fig. 3), whereas the X60 steel endures the harshest localized attack
 M. Xu et al. / International Journal of Greenhouse Gas Control 51 (2016) 357–368 365

(as shown in Fig. 9a) in spite of its relatively low average corrosion
rate (as shown in Fig. 3).
In addition, the increment of water is concluded to not only
enhance the average corrosion behavior, but also promote the local-
ized attack. Mass loss is also attributed to the localized corrosion,
as the penetration rate variation exhibits a high similarity to the
average corrosion rate variation for almost all carbon steels.
In order to investigate the extent of localized corrosion com-
prehensively, the pitting factor is shown in Fig. 9b for the samples
exposed to 8 MPa. Pitting factor is defined as the ratio of the pene-
tration rate to the average corrosion rate, shown in Eq. (2) (ASTM
Standard G46-94, 2005). Localized corrosion is reported to occur
if the pitting factor exceeds 5, shown in Inequality 3 (Singer et al.,
2013). The pit morphology is either deep and narrow, or hemispher-
ical and cup-shaped (Singer et al., 2013).
 
maximum penetration rate mm/year
pitting factor =   (2)
general corrosion rate mm/year

if pitting factor ≥ 5, localized corrosion occurred (3)

Apparently, almost all the pitting factors in Fig. 9b exceed 5,
indicating that the localized corrosion occurred and dominated the
corrosion behavior at 8 MPa. In terms of the type of pipeline steel,
the pitting factors of X60 and X65 steels evidently decrease with
the water contents, whereas the pitting factor of X80 steel gradu-
ally decreases with the water contents. This phenomenon indicates
that the degree of water increment enhancing uniform corrosion is
higher than that of water promoting localized attack. For X70 steel,
the pitting factor shows an irregular relationship with the water
contents. As shown in Fig. 9b, the pitting factors of X60 steel are
the highest among the carbon steels, indicating that the X60 steel
suffers the most severe localized degradation, which is consistent
with the highest penetration rate of X60 steel shown in Fig. 9a. The
pitting factors of X65 and X80 steels are lower than those of X60
steel because of higher average corrosion rates.
Overall, the samples exposed to water-unsaturated CO2 mix-
tures at 8 MPa particularly suffer a severe localized attack. Water
increment not only enhances the average corrosion rates, but also
promotes the localized corrosion behavior. However, the degree
of enhancing the average corrosion behavior is higher than that of
promoting the localized attack, thus decreasing the pitting factors.
It is noted that localized attack is more detrimental to the durabil-
ity of pipelines than the uniform corrosion at low water contents
as long as the localized attack occurs.
Fig. 10a and b present the penetration rate and pitting fac-
tor with varying water contents for the steel samples exposed
to CO2 mixtures at 10 MPa, respectively. Fig. 10a shows that the
penetration rates of X60, X65, X70, and X80 steels increase from
0.04 mm/year to 6.02 mm/year with the water contents. Water
increment also enhances the localized attack for the steel sam-
ples exposed to 10 MPa, which is similar to that of samples
exposed to 8 MPa. The penetration rates of carbon steels are all Fig. 8. XRD results of product layers for corroded X60 steel samples exposed to
pressure of (a) 8 MPa; and (b) 10 MPa with water content varying from 1600 to
less than 0.1 mm/year at a water content of 1600 ppm and exceed
3000 ppm.
0.1 mm/year at water contents of 2000 ppm or more. In addition,
the maximum pit depth, which is 50 ␮m deep and corresponds to
the penetration rate of 6 mm/year, occurs on the X80 steel sample corrosion. However, the pitting factors generally exceed 5 at water
exposed to a water content of 3000 ppm at 10 MPa, as shown in contents of 2000 and 2600 ppm, which indicates that localized cor-
Fig. 11b. rosion occurs. The pitting factors of X60, X70, and X80 steels initially
The penetration rates of carbon steel samples exposed to 10 MPa show a monotonous increasing trend with water contents up to
are lower than those for samples exposed to 8 MPa at water con- 2600 ppm and then decrease with water contents varying from
tents of 1600, 2000, and 2600 ppm, but are higher at a water content 2600 ppm to 3000 ppm. Meanwhile, the pitting factors of X65 steel
of 3000 ppm. This phenomenon is consistent with the comparison achieve the maximum pitting factor at a water content of 2000 ppm.
of the average corrosion rates of samples exposed to pressures of 8 Comparing the pitting factors between the samples exposed to
and 10 MPa, as shown in Fig. 4. 8 and 10 MPa leads to the conclusion that the degree of localized
The pitting factors of carbon steels are less than 5 at water con- attack for the samples exposed to 8 MPa is higher at water contents
tents of 1600 and 3000 ppm, indicating the occurrence of uniform of 1600–3000 ppm.
366 M. Xu et al. / International Journal of Greenhouse Gas Control 51 (2016) 357–368

Fig. 9. Evaluation of localized attack for carbon steel samples exposed to 8 MPa and

varying water contents. , X60 steel; , X65 steel; , X70 steel; , X80
steel. (a) The penetration rate for carbon steel samples exposed to 8 MPa and varying
water contents. (b) The pitting factor for steel samples exposed to 8 MPa and varying
Fig. 10. Evaluation of localized attack for steel samples exposed to 10 MPa and vary-
water contents.
ing water contents. , X60 steel; , X65 steel; , X70 steel; , X80 steel. (a)
The penetration rate for carbon steel samples exposed to 10 MPa and varying water
contents. (b) The pitting factor for steel samples exposed to 10 MPa and varying
water contents.
3.3. Additional discussion

3.3.1. The safe limit of water content for carbon steels exposed to (a) Dissolution of SO2 in thin condensed water on the steel sur-
the CO2 mixture at 50 ◦ C, 10 MPa face and ionization:
As stated, uniform corrosion is indicated to occur on carbon
SO2(g) ↔ SO2(aq) (4)
steels exposed to 10 MPa at a water content of 1600 ppm. Notably,
a corrosion rate of 0.1 mm/year is industrially acceptable in the SO2(aq) + H2 O(l) ↔ H2 SO3(aq) (5)
context of oil and gas production pipelines (Hua et al., 2014b).
+ −
The average corrosion rates and penetration rates for carbon steels H2 SO3(aq) ↔ H + HSO3 (6)
exposed to 10 MPa at a water content of 1600 ppm are less than
0.1 mm/year, which indicates that the safe limit of the water con- HSO3 − ↔ H+ + SO3 2− (7)
tent for carbon steels exposed to CO2 –SO2 –O2 mixture in present
H2 SO3 may be oxidized by oxygen to form H2 SO4 , which is more
study at 10 MPa, 50 ◦ C is 1600 ppm.
aggressive. The formation of sulfate is expressed as follows:
Meanwhile, localized corrosion occurred on the carbon steel
samples exposed to 8 MPa at water contents of 1600–3000 ppm, 2SO3 2− + O2 ↔ 2SO4 2− (8)
with the average corrosion rates and penetration rates surpassing
0.1 mm/year. Therefore, the safe limit of water content for carbon (b) Cathodic reaction can occur either by the direct reduction of
steels exposed to CO2 –SO2 –O2 mixture in present study at 8 MPa, hydrogen ions or the reduction of oxygen:
50 ◦ C must be less than 1600 ppm. However, the safe limit of water
contents for carbon steels exposed to different temperature, pres- 2H+ + 2e− → H2 (9)
sure and CO2 mixtures need to be further studied.
O2 + 4H+ + 4e− → 2H2 O (10)

(c) The anodic dissolution of iron is expressed as follows:

3.3.2. Reaction mechanism
Fe → Fe2+ + 2e− (11)
The reaction mechanisms associated with the formation of
FeSO4 observed in this study have been reported by Xiang et al. (d) The precipitation of FeSO4 occurs if solubility is exceeded:
(2011, 2012a) and Choi et al. (2010). The possible reaction mecha-
nisms that include anodic and cathodic reactions are as follows: Fe2+ + SO4 2− → FeSO4 (12)
 M. Xu et al. / International Journal of Greenhouse Gas Control 51 (2016) 357–368 367

4. Conclusions

The corrosion behavior of X60, X65, X70, and X80 carbon steels
exposed to water-unsaturated supercritical CO2 mixtures are sim-
ulated by using an autoclave with rotor. The effect of pressure and
water content on general corrosion and localized attack are investi-
gated. A systemic analysis method, which includes the weight loss
method, SEM, EDS, XRD, and 3D measuring laser microscope, is
employed to detect the extent of corrosion attack. Several conclu-
sions of this work are listed as follows:

(1) The average corrosion rates of carbon steels generally increase

Fig. 11. 3D surface morphology of localized corrosion on the cleaned samples after
removing products from corroded carbon steel samples (a) the image of localized
corrosion for cleaned X60 steel sample exposed to 8 MPa and water content of
3000 ppm; the maximum pit depth is 27 ␮m; (b) the image of localized corrosion
for cleaned X80 steel sample exposed to 10 MPa and water content of 3000 ppm;
the maximum pit depth is 50 ␮m.
with the water contents, with the average corrosion rates rang-
ing from 0.02 mm/year to 0.31 mm/year at 8 MPa and from
0.01 mm/year to 0.94 mm/year at 10 MPa. However, the trend
of the variation of the average corrosion rates with the water
contents at 8 MPa differs from that at 10 MPa, although both
conditions are in the supercritical CO2 phase.
(2) The degree of general corrosion for steel samples exposed to
8 MPa is higher than that for steel samples exposed to 10 MPa
at water contents of 1600, 2000, and 2600 ppm, but lower at a
water content of 3000 ppm.
(3) Localized corrosion is observed and determined by measuring
the penetration rate and pitting factor. The penetration rates
range from 0.2 mm/year to 3.25 mm/year at 8 MPa and from
0.04 mm/year to 6.02 mm/year at 10 MPa. The pitting factors
vary from 4 to 70 at 8 MPa and from 3 to 43 at 10 MPa. The
penetration rates for samples exposed to 8 MPa are higher than
those for samples exposed to 10 MPa at water contents of 1600,
2000, and 2600 ppm, but lower at a water content of 3000 ppm.
(4) The average corrosion rates and penetration rates of samples
exposed to 10 MPa at a water content of 1600 ppm are less than
0.1 mm/year, which indicates that the safe limit of water con-
tent for carbon steel samples exposed to CO2 mixtures in this
study at 10 MPa is 1600 ppm. However, this safe limit does not
apply to the conditions with a pressure of 8 MPa, in which the
safe limit of the water content is lower than 1600 ppm.
Acknowledgements

Consequently, FeSO4 might be further oxidized to form FeOOH The research leading to these results has received funding from
through an acid regeneration cycle with sufficient O2 (Chio et al., the European Community’s Seventh Framework Programme (FP7-
2007; Nishimura et al., 2004): ENERGY-20121-1-2STAGE) under grant agreement N◦ 308809 (The
IMPACTS project). The authors acknowledge the project partners
4FeSO4 + 6H2 O + O2 → 4FeOOH + 4H2 SO4 (13) and the following funding partners for their contributions: Statoil
Meanwhile, another alternative reaction path for a direct reac- Petroleum AS, Lundin Norway AS, Gas Natural Fenosa, MAN Diesel
tion between the iron and H2 SO4 gas molecules to form FeSO4 was & Turbo SE and Vattenfall AB.
suggested by several researchers (Kranzmann et al., 2011; Ruhl
and Kranzmann, 2013). The possible reaction mechanisms are as
References
follows:
(a) Vaporized water might react with sulfur dioxides in the ASTM Standard G1-03, 2011. Standard Practice for Preparing, Cleaning, and
gaseous phase to form acids, which subsequently condensate on Evaluating Corrosion Test Specimens. ASTM International, West
the sample surface. Conshohocken, PA.
ASTM Standard G46-94, 2005. Standard Guide for Examination and Evaluation of
2SO2(g) + O2(g) → 2SO3(g) (allinthegasphase) (14) Pitting Corrosion. ASTM International, West Conshohocken, PA.
Chio, C.H., Sharma, S.K., Muenow, D.W., 2007. The hydrates and deuterates of
SO2(g) + H2 O(g) → H2 SO3 (g) (15) ferrous sulfate (FeSO4 ): a Raman spectroscopic study. J. Raman Spectrosc. 38,
87–99.
Choi, Y.-S., Nešić, S., 2011. Determining the corrosive potential of CO2 transport
H2 SO3 + 2H2 O + (1/2)O2 → SO4 2− + 2H3 O+ (16) pipeline in high pCO2 -water environments. Int. J. Greenh. Gas Control 5,
788–797.
SO3(g) + H2 O(g) → H2 SO4(g) (17) Choi, Y.-S., Nešić, S., Young, D., 2010. Effect of impurities on the corrosion behavior
of CO2 transmission pipeline steel in supercritical CO2 -water envirionments.
(b) The resulting acids lead to the dissolution of ferrous iron. Environ. Sci. Technol. 44, 9233–9238.
Choi, Y.-S., Farelas, F., Nešić, S., Magalhães, A.A.O., de Azevedo Andrade, C., 2013.
Fe + H2 SO3 → Fe2+ + H2 + SO3 2−
0
(18) Corrosion behavior of deep water oil production tubing material under
0 2+ 2−
supercritical CO2 environment: part 1—effect of pressure and temperature.
Fe + H2 SO4 → Fe + H2 + SO4 (19) Corrosion 70, 38–47.
368 M. Xu et al. / International Journal of Greenhouse Gas Control 51 (2016) 357–368
de Visser, E., Hendriks, C., Barrio, M., Mølnvik, M.J., de Koeijer, G., Liljemark, S., Le
Gallo, Y., 2008. Dynamis CO2 quality recommendations. Int. J. Greenh. Gas
Control 2, 478–484.
Gale, J., Davison, J., 2004. Transmission of CO2 —safety and economic
considerations. Energy 29, 1319–1328.
Hua, Y., Barker, R., Neville, A., 2014a. Effect of temperature on the critical water
content for general and localised corrosion of X65 carbon steel in the transport
of supercritical CO2 . Int. J. Greenh. Gas Control 31, 48–60.
Hua, Y., Barker, R., Neville, A., 2014b. Understanding the influence of SO2 and O2 on
the corrosion of carbon steel in water-saturated supercritical CO2 . Corrosion
71, 667–683.
Hua, Y., Barker, R., Neville, A., 2015a. Comparison of corrosion behaviour for X65
carbon steel in supercritical CO2 -saturated water and
water-saturated/unsaturated supercritical CO2 . J. Supercrit. Fluids 97, 224–237.
Hua, Y., Barker, R., Neville, A., 2015b. The influence of SO2 on the tolerable water
content to avoid pipeline corrosion during the transportation of supercritical
CO2 . Int. J. Greenh. Gas Control 37, 412–423.
Hua, Y., Barker, R., Neville, A., 2015c. The effect of O2 content on the corrosion
behaviour of X65 and 5Cr in water-containing supercritical CO2 environments.
Appl. Surf. Sci. 356, 499–511.
Hua, Y., Barker, R., Charpentier, T., Ward, M., Neville, A., 2015d. Relating iron
carbonate morphology to corrosion characteristics for water-saturated
supercritical CO2 systems. J. Supercrit. Fluids 98, 183–193.
Kranzmann, A., Neddemeyer, T., Ruhl, A.S., Huenert, D., Bettge, D., Oder, G.,
Neumann, R.S., 2011. The challenge in understanding the corrosion
mechanisms under oxyfuel combustion conditions. Int. J. Greenh. Gas Control
5S, S168–S178.
Nishimura, R., Shiraishi, D., Maeda, Y., 2004. Hydrogen permeation and corrosion
behavior of high strength steel MCM 430 in cyclic wet–dry SO2 environment.
Corros. Sci. 46, 225–243.
Ruhl, A.S., Kranzmann, A., 2013. Investigation of corrosive effects of sulphur
dioxide, oxygen and water vapour on pipeline steels. Int. J. Greenh. Gas Control
13, 9–16.
Sim, S., Cole, I.S., Bocher, F., Corrigan, P., Gamage, R.P., Ukwattage, N., Birbilis, N.,
2013a. Investigating the effect of salt and acid impurities in supercritical CO2
as relevant to the corrosion of carbon capture and storage pipelines. Int. J.
Greenh. Gas Control 17, 534–541.
Sim, S., Bocher, F., Cole, I.S., Chen, X.B., Birbilis, N., 2013b. Investigating the effect of
water content in supercritical CO2 as relevant to the corrosion of carbon
capture and storage pipelines. Corrosion 70, 185–195.
Singer, M., Hinkson, D., Zhang, Z., Wang, H., Nešić, S., 2013. CO2 top-of-the-line
corrosion in presence of acetic acid: a parametric study. Corrosion 69, 719–735.
Spycher, N., Pruess, K., Ennis-King, J., 2003. CO2 -H2 O mixtures in the geological
sequestration of CO2 . I. Assessment and calculation of mutual solubilites from
12 to 100 degrees and up to 600 bar. Geochim. Cosmochim. Acta 67,
3015–3031.
Xiang, Y., Wang, Z., Xu, C., Zhou, C.C., Li, Z., Ni, W.D., 2011. Impact of SO2
concentration on the corrosion rate of X70 steel and iron in water-saturated
supercritical CO2 mixed with SO2 . J. Supercrit. Fluids 58, 286–294.
Xiang, Y., Wang, Z., Yang, X.X., Li, Z., Ni, W.D., 2012a. The upper limit of moisture
content for supercritical CO2 pipeline transport. J. Supercrit. Fluids 67, 14–21.
Xiang, Y., Wang, Z., Li, Z., Ni, W.D., 2012b. Effect of exposure time on the corrosion
rates of X70 steel in supercritical CO2 /SO2 /O2 /H2 O environments. Corrosion
69, 251–258.
Xiang, Y., Wang, Z., Li, Z., Ni, W.D., 2013a. Effect of temperature on corrosion
behavior of X70 steel in high pressure CO2 /SO2 /O2 /H2 O environments. Corros.
Eng. Sci. Technol. 48, 121–129.
Xiang, Y., Wang, Z., Li, Z., Ni, W.D., 2013b. Long term corrosion of X70 steel and iron
in humid supercritical CO2 with SO2 and O2 impurities. Corros. Eng. Sci.
Technol. 48, 395–398.
Xiang, Y., Wang, Z., Xu, M.H., Li, Z., Ni, W.D., 2013c. A mechanistic model for
pipeline steel corrosion in supercritical CO2 –SO2 –O2 –H2 O environments. J.
Supercrit. Fluids 82, 1–12.
Zhang, Z.X., Wang, G.X., Massarotto, P., Rudolph, V., 2006. Optimization of pipeline
transport for CO2 sequestration. Energy Convers. Manag. 47, 702–715.
Zhang, Y.C., Pang, X.L., Qu, S.P., Li, X., Gao, K.W., 2012. Discussion of the CO2
corrosion mechanism between low partial pressure and supercritical
condition. Corros. Sci. 59, 186–197.