materials

Article
Corrosion Failure Mechanism of Associated Gas
Transmission Pipeline
Weimin Zhao *, Timing Zhang, Yonglin Wang, Jianhua Qiao and Zerui Wang
School of Materials Science and Engineering, China University of Petroleum, Qingdao 266580, China;
zhangtm@upc.edu.cn (T.Z.); m13045013792@163.com (Y.W.); 15610036410@163.com (J.Q.);
m13455289561@163.com (Z.W.)
* Correspondence: zhaowm@upc.edu.cn; Tel.: +86-182-6663-9639




Received: 21 August 2018; Accepted: 1 October 2018; Published: 11 October 2018


Abstract: Corrosion has been responsible for several gas pipeline leakage accidents; thus, clarifying
its failure mechanisms is a precondition to prevent such accidents. On the basis of failure analysis
of corroded pipe sections, laboratory exposure tests were conducted by simulating three possible
corrosion environments inside a gas pipeline. The corrosion rate indicated by depth change was
adopted in this study. Scanning electron microscopy and X-ray diffraction were used to analyze
corrosion products. Results showed that the specimens completely immersed in condensate water
were generally corroded and that the specimens exposed to gas were locally corroded. However,
the corrosion rate of the latter was slightly lower; hence, no autocatalysis of occluded corrosion
cell occurred in the formation of corrosion pit, and uniform corrosion occurred in the precipitation
location of condensate water. The areas in the range of 5 mm below the waterline indicated severe
corrosion, and the rate could reach twice that of other areas. The corrosion products were mainly
FeO(OH) and FeCO3 , thereby proving that the corrosion failure of pipelines was caused by oxygen
absorption corrosion and CO2 corrosion. Suggestions were presented to control corrosion failure of
associated gas pipelines.

Keywords: associated gas; corrosion; failure mechanism; gas pipeline

1. Introduction
Associated gas is the gas that is obtained in crude oil production; the main ingredients of such
associated gas are methane, ethane, and hydrocarbons [1]. In the absence of a gas market, the traditional
means of managing associated gas are reinjection, flaring, and venting [2]. The flaring or venting of
associated gas not only causes energy waste but also adversely affects the environment. In contrast,
reinjection is a type of treatment method that is beneficial to sustainable development [3–5]. In some
oil fields, associated gas is used as a lift gas to raise oil recovery [6]. Gas lifting is the process of
injecting compressed natural gas (called lift gas) into the production tubing of an oil well, and the
lift gas increases the pressure gradient between the reservoir and well fluid that pushes the fluid to
the surface [7]. Usually, the compressed gas from a field processing facility is routed to oil gathering
manifold and wellhead via pipelines.
Associated gas in many cases has not been subjected to strict dehydration treatment, which can
be classified as wet gas. The wet gas contains liquid hydrocarbon and water, acid gases (such as CO2
and H2 S), and dissolved ions (such as Cl− and Ca2+ ) [8,9]. Under the combined action of temperature,
pressure, and stress, serious corrosion occurs in the inner wall of steel pipes [8,10,11]. Internal corrosion
is the main factor that affects the safety of natural gas pipelines [12–14]. It can reduce the thickness of
thick walls, reduce the strength of pipelines, and lead to leakage accidents. Once the gas transmission
pipeline leaks, it not only results in considerable economic loss but also causes environmental pollution.

Materials 2018, 11, 1935; doi:10.3390/ma11101935 www.mdpi.com/journal/materials

Materials 2018, 11, 1935 2 of 17

Therefore, studying the corrosion failure mechanism of gas transmission pipelines is necessary to
propose control measures.
Thus far, many studies have been conducted on corrosion and its control of gas pipelines.
Most researchers have deemed that the presence of water and dissolved gases, such as CO2 and
H2 S, introduce serious internal corrosion [12,15–18]. Many factors that promote liquid collection,
including gas quality, operating parameters, and pipeline features, affect the likelihood of internal
corrosion [19]. The water phase that wets the steel surface determines the initiation of corrosion [16].
CO2 corrosion is considered the main cause of corrosion in natural gas pipelines [20,21], and the
corrosion rate is accelerated by O2 [22], H2 S and organic acids [23,24], and salt content [25,26]. Various
laboratory tests have been conducted to study the corrosion of steel pipelines under full immersion
in the water or gas phase under water condensing conditions [17,26]. However, the liquid depth in
gas pipelines is relatively small in most cases, and the corrosion law of steel pipelines under thin
liquid film and in large-volume liquid should be different. No research has shown the corrosion of
the liquid–gas transition zone near a waterline. In addition, CO2 corrosion is not necessarily the root
cause of gas pipeline corrosion. The corrosion mechanism of gas pipelines changes with gas impurities.
At present, corrosion research on CO2 and O2 coexistence conditions remains insufficient because
corrosion products may change with changes in their partial pressure.
In this study, the corrosion mechanism of associated gas pipelines was developed for actual
engineering failure accidents. Gas lift facilities were commissioned in 2009–2010, and associated
gas was used as the lift gas. The associated gas contained water, CO2 and O2 . Several leaks were
observed in the associated gas lines in November 2014. Then, the situation became increasingly serious,
thereby seriously affecting the normal operation of pipelines. Relevant research is urgently needed
to determine the root cause of the leak and take remedial measures. Therefore, the corrosion failure
mechanism of the associated gas pipeline was clarified by simulating three corrosion environments
inside a gas pipeline on the basis of observing corrosion morphology and analyzing corrosion products.

2. Materials and Methods

2.1. Description of the Pipeline

The investigated pipelines are buried trunk lines (151.1 mm × 7.1 mm, 6” pipe) and flow lines
(102.3 mm × 6.0 mm, 4” pipe; 50.7 mm × 4.8 mm, 2” pipe) for conveying the associated gas. The pipe
material is ANSI B31.8 API 5L X42. The chemical compositions of the collected pipe materials were
determined by a Metallab 75/80 direct-reading spectrometer (GNR, Novara, Italy), and the results
are listed in Table 1. The tensile properties of the collected pipe materials were investigated by a
WDW-300E computer-controlled universal testing machine (Shijin Testing Equipment, Jinan, China),
and the test results are shown in Table 2. The corresponding acceptance criteria are also listed in
Tables 1 and 2 as Spec. It can be seen from Tables 1 and 2 that the chemical compositions and tensile
properties of the three pipes with different diameters are all acceptable based on the requirements
for PSL 2 seamless X42 pipe in API Spec 5L-2013; however, the content of sulfur in the 2” pipe is
more than the requirement. The microstructures of the 2”, 4” and 6” steel pipes in cross section,
longitudinal section, and inner surface layer were observed by a Leica DM2500 M optical microscope
(Leica Microsystems, Wetzlar, Germany). No essential difference was found between cross-sectional,
longitudinal-sectional, and inner surface microstructures; therefore, only the microstructures in the
cross-section are shown in Figure 1. It is obvious that the microstructures are normal and are composed
of ferrite and pearlite.
Materials 2018, 11, x FOR PEER REVIEW 3 of 17
Materials 2018, 11, 1935 3 of 17

Table 1. Chemical composition of collected pipe materials and that for PSL 2 seamless X42 pipe in
API Spec
Table 5L-2013.composition of collected pipe materials and that for PSL 2 seamless X42 pipe in API
1. Chemical
Spec 5L-2013.
C Si Mn P S V Nb Ti Cu Ni Cr Mo
2″ 0.167
C 0.245
Si 0.408
Mn 0.009
P 0.025
S 0.000
V 0.000
Nb 0.000
Ti 0.018
Cu 0.000
Ni 0.027
Cr 0.000
Mo
4″
2” 0.199
0.167 0.207
0.245 0.360 0.009
0.408 0.017 0.009
0.025 0.000
0.000 0.000
0.000 0.001
0.000 0.106
0.018 0.007
0.000 0.032 0.000
0.027 0.000
4”
6″ 0.199
0.175 0.207
0.251 0.360 0.017
0.499 0.010 0.009
0.010 0.000
0.000 0.000
0.000 0.001
0.000 0.106
0.100 0.007
0.000 0.032 0.000
0.036 0.000
6” 0.175 0.251 0.499 0.010 0.010 0.000 0.000 0.000 0.100 0.000 0.036 0.000
Spec
Spec ≤0.24
≤ 0.24 ≤≤0.40
0.40 ≤≤1.20
1.20 ≤≤0.025
0.025 ≤0.015
≤0.015 ≤≤0.06
0.06 ≤≤0.05
0.05 ≤≤0.04
0.04 ≤≤0.50
0.50 ≤≤0.30
0.30 ≤0.30
≤ 0.30 ≤0.15
≤ 0.15
Spec: acceptance
Spec: criteria
acceptance criteriaof
ofchemical composition
chemical composition forfor
PSLPSL 2 seamless
2 seamless X42 pipe.
X42 pipe.

Table 2.
Table 2. Tensile
Tensileresults
resultsof
ofcollected
collectedsteel
steelpipes
pipesand
andacceptance
acceptancecriteria
criteriafor
forPSL
PSL22X42
X42pipe.
pipe.

Steel
Steel Sample Rt0.5
Sample Rt0.5 (MPa)
(MPa) Rmm (MPa)
R (MPa) Rt0.5/Rm
Rt0.5 /Rm
2” 2″ 435435 500
500 0.870.87
4” 4″ 350350 498
498 0.700.70
6” 6″ 375375 515
515 0.730.73
SpecSpec 290–495
290–495 415–655
415–655 ≤
≤0.93 0.93
RRt0.5
t0.5,, yield
yieldstrength (0.5 %
strength total
(0.5 %extension); Rm , tensileRstrength;
total extension); m, tensileSpec: acceptance
strength; criteria
Spec: of tensile properties
acceptance criteria offortensile
PSL 2 seamless X42 pipe.
properties for PSL 2 seamless X42 pipe.

Figure1.1.Microstructures
Figure Microstructuresof
ofpipes
pipeswith
withdifferent
differentdiameters:
diameters:(a)
(a)2”,
2″,(b)
(b)4”,
4″, and
and (c)
(c)6”.
6″.

The
The compositions
compositionsofofrepresentative
representativegas gassamples
samples areare
listed in Table
listed in Table3. The main
3. The mainingredients
ingredientsof the
of
gas
the are
gashydrocarbon (methane,
are hydrocarbon ethane,
(methane, propane,
ethane, butane,butane,
propane, pentane, and >C5),
pentane, and nitrogen, CO2 , andCO
>C5), nitrogen, oxygen.
2, and

Furthermore, 1.2–2.5 mg/kg

oxygen. Furthermore, 1.2–2.5H2 mg/kg
S is present
H2S inis the gas. The
present in the ion gas.
concentration
The ion of representative
concentration of
condensate water is listed in Table + + 2+ + 2+
representative condensate water is 4. Thein
listed main
Tablecations
4. ThearemainNacations
, K , Ca are Na, Li , ,Kand
+ + , CaMg
2+ , Li ., The
+ and main
Mg2+.
anions areanions2-
SO4 areandSO -
Cl4 . and
TheClpH value
The main 2- -. The pH of condensate
value waterwater
of condensate is 6.65. TheThe
is 6.65. operating
operating pressure
pressureis
approximately 11 MPa, andandthe temperature is in the ◦ C. The normal flow is 3500, 38,400,
is approximately 11 MPa, the temperature is inrange 41–45 41–45
the range °C. The normal flow is 3500,
and 120,000
38,400, m3 /d inmthe
and 120,000 3/d 2”, 4”, 2″,
in the and4″,6”and
pipelines, with corresponding
6″ pipelines, with corresponding flow rates
flow ratesof 1.1,of5.9,
1.1,and
5.9,12.2
and
m/s, respectively.
12.2 m/s, respectively.

Table3.3.Compositions
Table Compositionsof
oftransmission
transmissiongas
gas(vol.%).
(vol.%).

Composition
Composition Percent
Percent Composition
Composition Percent
Percent Composition
Composition Percent
Percent Composition Percent
Composition Percent
Methane
Methane 72.72
72.72 Nitrogen
Nitrogen 19.89
19.89 Ethane
Ethane 3.66
3.66 CO2 2
CO 1.30
1.30
Propane
Propane 0.87
0.87 Oxygen
Oxygen 0.47
0.47 n-Butane
n-Butane 0.36
0.36 i-Butane
i-Butane 0.25
0.25
iso-Pentane
iso-Pentane 0.19
0.19 N-Pentane
N-Pentane 0.14
0.14 >C5
>C5 0.10
0.10 CO
CO 0.05
0.05

Table4.4.Ion
Table Ionconcentration
concentration[C]
[C]ofofcondensate
condensatewater
water(mg/L).
(mg/L).

Composition [C]
Composition [C] Composition
Composition [C]
[C] Composition
Composition [C]
[C] Composition
Composition [C]
[C]
Li Li
+ + 0.45 Na ++ 24.18 NH +4+ 0.43 K 2.3
+
0.45 Na 24.18 NH 4 0.43 K+ 2.3
MgMg2+ 2+ 0.14
0.14 Ca
Ca2+2+ 1.62
1.62 Cl- -
Cl 0.88
0.88 SO
SO442-
2- 72.21
72.21
Materials 2018, 11, 1935 4 of 17

2.2. Laboratory Studies of Corroded Pipe Sections

The information from the site investigation indicates that most leakage points are located at
the bottom of the pipeline, whereas some are located on the side; only a few are located at the top.
Representative corroded pipe sections were collected for laboratory studies. The internal corrosion
morphology was observed macroscopically by a camera and microscopically by using a JXA 8230
scanning electron microscope (SEM, JEOL, Tokyo, Japan). The chemical composition of corrosion
products on the pipe surface of the 2”, 4”, and 6” steel pipe was obtained by energy-dispersive
spectroscopy (EDS) analysis using the JXA 8230 SEM. The phase compositions of corrosion products
were analyzed using an X’ Pert PRO MPD X-ray diffraction (XRD, PANalytical B.V., Almelo,
The Netherlands).

2.3. Corrosion Simulation Tests

New pipe sections, which have the same pipe material with the leaking pipe, were collected.
Steel specimens were machined from the pipe sections. The dimensions of specimens from the 2” steel
pipe are 40 × 13 × 2 mm3 , and those from the 4” and 6” steel pipes are 40 × 13 × 3 mm3 . All the
specimens were ground to 2000-grade emery paper and cleaned with distilled water and absolute
ethyl alcohol.
Corrosion simulation tests were conducted following ASTM G111 “Standard guide for corrosion
tests in high-temperature or high-pressure environment, or both”. The tests were conducted in a kettle,
and the specimens were positioned in three different environments. One-third of the total specimens
was totally immersed in the water phase, which is a condition that corresponds to the so-called “bottom
of the line” situation in gas pipelines; another one-third of the total specimens was partly immersed in
the solution, which represents the working environment of the part of the pipe that corresponds to
the transfer zone from liquid to gas environment; and the final one-third of the total specimens was
exposed to high-pressure gas phase under water condensing conditions, which corresponds to the “top
of the line” situation. The gas and the solution were prepared based on Tables 3 and 4, and nitrogen
was used to replace the organic components in the simulated gas. Four specimens were investigated
for each pipe in each environment; one of them was used for SEM observation and XRD analysis,
and the other three were used to calculate the corrosion rate. The test temperature was 43 ◦ C, and the
test pressure was 11 MPa. The corrosion tests lasted 135 h.
The corrosion rates with different corrosion failure modes were determined in various ways.
The weight loss rate was used to evaluate the general corrosion rate, and the depth of local corrosion
pits was measured by Leica DM2500 M confocal microscopy. For comparison, the rate of depth change
was used to represent the corrosion rate. Corrosion products were analyzed by EDS and XRD.

3. Results

3.1. Macroscopic Observation of Corroded Pipe Sections

Figure 2 shows the typical corrosion morphology on the internal walls of the 2”, 4”, and 6” steel
pipe sections with internal corrosion. The upper and lower positions of the pipe section in Figure 2a,b
are consistent with the working status. Corrosion products are distributed across all the exposed
surfaces. However, the corrosion products on the bottom part of pipe sections, especially in pipes with
a small diameter, are thick and nonuniform. A perforation occurs on the bottom part of the 6” pipe
section (Figure 2c), and thickness reduction can be clearly observed around the perforation, thereby
indicating that the perforation is induced by internal corrosion. The corrosion environment for the
bottom part of the pipe sections is different from those for the side and top parts, and the bottom parts
are likely covered with condensate liquid.
Materials 2018, 11, 1935
Materials x FOR PEER REVIEW 55 of
of 17
17
Materials 2018, 11,
11, x FOR PEER REVIEW

perforatio
perforatio

Figure 2. Corrosion morphology on the internal walls of steel

steel pipe
pipe sections:
sections: (a) 2″
2″ pipe, (b) 4″
4″ pipe
Figure 2. Corrosion morphology on the internal walls of steel pipe sections: (a) 2” pipe, (b) 4” pipe
and (c) 6″ pipe.
6″ pipe.
and (c) 6”

3.2. Analysis of
3.2. Analysis of Corrosion
Corrosion Products
Products
An
An SEMSEMequipped
SEM equipped
equipped with
with EDSEDS
with EDS was applied
was applied
was applied to investigate
to investigate
to investigate the cross-sectional
cross-sectional
the cross-sectional
the and surface and and surface
morphology
surface
morphology
and chemical and chemical
compositions compositions
of corrosion of corrosion
products on products
the pipe on the
surface
morphology and chemical compositions of corrosion products on the pipe surface of the 2″, 4″, pipe
of the surface
2”, 4”, of
andthe
6” 2″, 4″,pipe
steel and
and
6″ steel
sections. pipe
The sections.
approximate The approximate
observed observed
positions are positions
marked are
in marked
Figure 2 in
with
6″ steel pipe sections. The approximate observed positions are marked in Figure 2 with red boxes. Figure
red 2
boxes.withNored boxes.
obvious
No obvious
No obvious micromorphology
micromorphology micromorphology
differences occur differences
between occur
differences occur between
smoothbetween smooth
scale (Figure
smooth scaleraised
3a) and
scale (Figure
(Figure 3a) and
corrosion
3a) and raised
products
raised
corrosion
(Figure products
3b). (Figure
Subsequently, the3b). Subsequently,
chemical composition the chemical
of composition
corrosion products
corrosion products (Figure 3b). Subsequently, the chemical composition of corrosion products of
on corrosion
the 2”, 4”, products
and 6” on
steel
on
the 2″,
pipes 4″,
were and 6″
analyzed steel
by pipes
EDS. were
The analyzed
results show by EDS.
that noThe results
essential show
differencethatis
the 2″, 4″, and 6″ steel pipes were analyzed by EDS. The results show that no essential difference is no
found essential
in the difference
composition is
found
of
found in along
scales
in the composition
the composition
its thickness ofdirection
of scales along
scales along its thickness
thickness
for different
its pipes.direction
Therefore,
direction for only
for different pipes. Therefore,
the analysis
different pipes. Therefore, only
results ononly the
the
6” analysis
steel pipe results
are on the
presented 6″
in steel
Figure pipe
4. are
The presented
main in
elements Figure
include4. The
Fe, Cmain
and
the analysis results on the 6″ steel pipe are presented in Figure 4. The main elements include Fe, C elements
O. Then, include
the Fe,
corrosion C
and O.
and O. Then,
products Then, the corrosion
corrosion
are speculated
the products
to beproducts
oxides and arecarbonates
are speculatedfrom
speculated to be
to bethe
oxides andbetween
reaction
oxides and carbonates
carbonates from
steelfrom
and Othe reaction
2 and
the CO2
reaction
between
in the steel
environment. and O 2 and CO2 in
Furthermore, fewtheS environment.
is found in Furthermore,
corrosion products,
between steel and O2 and CO2 in the environment. Furthermore, few S is found in corrosion fewand S is
the found
materialin corrosion
should be
products,
sulfide of and
Fe, the
which material
is the should
reaction be sulfide
product of of Fe,
steel which
with H is
S.the reaction product
products, and the material should be sulfide of Fe, which 2is the reaction product of steel with H2S. of steel with H 2 S.

Figure 3.
Figure 3. Top view
view of
of corrosion
corrosion scales
scales on
on the
the 2″
2″ pipe:
pipe: (a)
(a) flat area
area and
and (b)
(b) raised
raised area.
area.
Figure 3. Top
Top view of corrosion scales on the 2” pipe: (a) flat
flat area and (b) raised area.
Materials 2018, 11, 1935 6 of 17
Materials 2018, 11, x FOR PEER REVIEW 6 of 17

Materials 2018, 11, x FOR PEER REVIEW 6 of 17

Figure
Figure 4. 4. EDS
EDS analysisresults
analysis resultsof
of corrosion
corrosion products
productson
onthe
the6″pipe: (a) (a)
6”pipe: surface, (b) inner
surface, layerlayer
(b) inner and (c)
and
outer
(c) outer layer.
layer.
Figure 4. EDS analysis results of corrosion products on the 6″pipe: (a) surface, (b) inner layer and (c)
outer layer.
The The phase
phase compositionsofofcorrosion
compositions corrosionproducts
products on on the
the 2”,
2″, 4″,
4”, and
and 6″6”steel
steelpipes
pipeswere
wereanalyzed
analyzed
using
using XRD.XRD.
The The corrosion
corrosion products
products on on
the the surfaces
surfaces of of
the the
three three
pipes pipes
are are identical,
identical,
The phase compositions of corrosion products on the 2″, 4″, and 6″ steel pipes were analyzed and and
FigureFigure 5
5 shows
shows
the using the
XRD spectra XRD
XRD. The spectra
of corrosion of corrosion
corrosionproducts
productsfrom products
on the from
thesurfaces the
6” steel of
pipe.6″ steel
the The pipe.
threecorrosionThe corrosion
pipes areproducts
identical, products
areand
composed are
Figure 5of
Fe3composed
O
shows O3 ,ofXRD
4 , Fe2the
Fe3OFeCO
FeO, 4, Fe2O,3,of
spectra 3 and,
FeO, FeCO3,products
FeO(OH).
corrosion and,
TheFeO(OH).
counts
from perThesecond
the 6″counts per second
steel (CPS)
pipe. value (CPS) value
in Figure
The corrosion in Figure
5 shows
products that
are
5composed
FeO(OH)showsand that
of FeO(OH)
Fe3O34,have
FeCO Fe2Oand FeCO
, FeO,
3large 3 have
FeCO
shares 3,in large
and, shares
FeO(OH).
corrosion inThe
scale.corrosion
counts per
Element scale. Element
S issecond
found (CPS)S isvalue
in the found in the
in Figure
corrosion scale
(seecorrosion
5Figure
shows 4). scale
that (see Figure
FeO(OH)
However, and 4).
sulfide FeCOHowever,
3 have
cannot sulfide
be large cannot
shares
detected bebecause
detected
in corrosion
by XRD ofby
scale. XRD
itsElement because of itsin
S is found
low content; thus, low
the
sulfide
content; thus, sulfide has a minimal effect on the corrosion failure of steel pipes.
hascorrosion
a minimalscale effect(see
on Figure 4). However,
the corrosion failure ofsulfide cannot be detected by XRD because of its low
steel pipes.
content; thus, sulfide has a minimal effect on the corrosion failure of steel pipes.

Figure 5. XRD analysis results of corrosion products.

Figure XRD
5. 5.
Figure XRDanalysis
analysisresults
resultsof
ofcorrosion
corrosion products.
products.
Materials 2018, 11, 1935
x FOR PEER REVIEW 7 of 17

On the basis of the analysis results of EDS and XRD, the corrosion products on different pipes
shareOn thethesamebasischemical
of the analysis
and phase results of EDS and however,
compositions; XRD, the corrosion
only the 6″ products
pipe is on different in
perforated pipes
the
share the same chemical and phase compositions; however, only
collected pipe sections. Given that the pipe materials and the transmission medium are the same, the 6” pipe is perforated in the
collected pipe sections.
the only difference is theGiven
flow ratethatofthe pipe
the materials
medium in the andpipe.theUnder
transmission
the given medium
working areconditions,
the same,
the only difference is the flow rate of the medium in the pipe. Under
the medium flow rate increases with the pipe diameter. Therefore, the influence of flow rate change the given working conditions,
the medium
on the flow rate increases
microstructure of corrosion withproducts
the pipe is diameter. Therefore, the influence of flow rate change on
further observed.
the microstructure
The surface and of corrosion products
cross-sectional is further observed.
morphology of the corrosion scale on the 2″, 4″ and 6″ steel
pipes is shown in Figure 6. The scale on the 2″ pipeof
The surface and cross-sectional morphology is the
dense corrosion
(Figure scale on the 2”,
6a); however, the4”bond
and strength
6” steel
pipes
betweenis shown
the scale in Figure
and steel6. The scale on
substrate is the 2” pipe
weak, is dense (Figure
and cracking tends to6a); however,
occur at thisthe bond strength
interface (Figure
between the scale and steel substrate is weak, and cracking tends
6b). The surface of scale on the 4″ pipe seems dense, and no pores can be found; however, to occur at this interface (Figure
cracks6b).
in
The
scalesurface of scale on
are observed the 4”6c).
(Figure pipeThe seems dense, and no
cross-sectional pores can be
observation found;
also shows however,
cracks cracks
and poresin scale are
in the
observed
corrosion(Figure 6c). The 6d),
scale (Figure cross-sectional
and the innerobservation
layer also shows cracks
is stronger than and the pores
outerinlayer.
the corrosion
The surfacescale
(Figure 6d), and the inner layer is stronger than the outer layer. The
morphology of the corrosion scales on the 6″ pipe is shown in Figure 6e. Cracks and many pores aresurface morphology of the corrosion
scales
found on in the
the6”corrosion
pipe is shown in Figure
scales; 6e. Cracks
therefore, these and scales many canporeseasilyarepeelfound in the
off. Thecorrosion scales;
cross-sectional
therefore,
morphology theseon scales
the 6″ can
pipeeasily
is shownpeel off. The cross-sectional
in Figure 6f, and the corrosion morphology scaleson thethe
near 6” pipe
steelissurface
shownare in
Figure 6f, and the corrosion scales near the steel surface are
denser than the outer parts. As shown in Figure 6, the corrosion products are porous at thedenser than the outer parts. As shown
in Figure 6,formation
beginning the corrosion andproducts
will become are porous
denser at with
the beginning
time. The formation
dense inner and willlayer become denser
can prevent
with time. The dense inner layer can prevent penetration of the
penetration of the outer corrosive medium. However, the flow rate of the transmission medium willouter corrosive medium. However,
the flow rate
influence theofstructure
the transmission medium
of corrosion willand
scale, influence
high the flow structure
rate can of corrosion
damage the scale, and high
integrity offlow
the
rate can damage the integrity of the corrosion scale, thereby resulting
corrosion scale, thereby resulting in many pores and cracks in the scale. Intact and dense corrosion in many pores and cracks in the
scale. Intact and dense corrosion scales with strong adhesion
scales with strong adhesion can reduce the uniform corrosion rate, whereas defects and scale can reduce the uniform corrosion rate,
whereas
sheddingdefects
can induceand scale
severe shedding can induce
local corrosion severe
[27,28]. Thelocal corrosion
corrosive [27,28].can
medium Thepenetrate
corrosivealongmedium the
can penetrate along the paths formed by cracks and porosities [29],
paths formed by cracks and porosities [29], and corrosion can occur beneath corrosion products. and corrosion can occur beneath
corrosion
Meanwhile, products. Meanwhile,
the peeling the peeling
of the corrosion of the
scale willcorrosion
lead to scale will lead to
the exposure of the
freshexposure
metal, of fresh
thereby
metal, thereby
accelerating theaccelerating
corrosion here.the corrosion
Evidently, here.
the 6″Evidently,
pipe with theloose
6” pipe with loose
corrosion corrosion
products is mostproducts
prone to is
most prone to
corrosion failure. corrosion failure.

Figure 6. Cont.
Materials 2018, 11, 1935 8 of 17
Materials 2018, 11, x FOR PEER REVIEW 8 of 17
Materials 2018, 11, x FOR PEER REVIEW 8 of 17

Figure
Figure 6.
6. Top
Figure 6. view
Top view
Top (a,c,e)
view (a,c,e) and
(a,c,e) and cross-sectional
and cross-sectional view
cross-sectional view (b,d,f)
view (b,d,f) of
(b,d,f) of corrosion
of corrosion scales
corrosion scales on
scales on steel
steel pipes:
on steel pipes: (a,b)
pipes: (a,b) 2″
(a,b) 2”
2″
pipe,
pipe, (c,d)
pipe, (c,d) 4″
(c,d) 4” pipe and
pipe and
4″ pipe (e,f)
and (e,f) 6″
(e,f) 6” pipe.
pipe.
6″ pipe.

3.3. Corrosion
3.3. Corrosion Simulation Test
Test Results
3.3. Corrosion Simulation
Simulation Test Results
Results
3.3.1. Failure
3.3.1. Failure Mode and and Corrosion Products
Products
3.3.1. Failure Mode
Mode and Corrosion
Corrosion Products
After 135
After h inina asimulated corrosion environment, the simulated gas was released from the kettle,
After 135
135 hh in a simulated
simulated corrosion
corrosion environment,
environment, the the simulated
simulated gasgas was
was released
released from
from the
the
and the
kettle, specimens were removed. The specimens immersed in solution exhibits a general corrosion
kettle, and the specimens were removed. The specimens immersed in solution exhibits a general
and the specimens were removed. The specimens immersed in solution exhibits a general
behavior, whereas
corrosion the specimens exposed to gas presents gas
a localized corrosion behavior. In comparison
corrosion behavior,
behavior, whereas
whereas the
the specimens
specimens exposed
exposed to to gas presents
presents aa localized
localized corrosion
corrosion behavior.
behavior.
with
In other regions, more corrosion products are piled up near the waterline for the waterline
specimens partly
In comparison with other regions, more corrosion products are piled up near the waterline for
comparison with other regions, more corrosion products are piled up near the for the
the
immersed in
specimens the solution.
partly immersed Figure
in the7solution.
shows each specimen
Figure 7 showsin each
detail.
specimen in detail.
specimens partly immersed in the solution. Figure 7 shows each specimen in detail.

7. Morphologyof
Figure 7.
Figure of specimensafter
after corrosiontest:
test: (a) full immersion, (b) half immersion and
Figure 7. Morphology
Morphology of specimens
specimens after corrosion
corrosion test: (a)
(a) full
full immersion,
immersion, (b)
(b) half
half immersion
immersion and
and (c)
(c)
(c) high-pressure
high-pressure gasgas exposure.
exposure.
high-pressure gas exposure.
To compare
To the chemical and phase compositions of corrosion products obtained obtained
in the simulation
To compare
compare the the chemical
chemical and and phase
phase compositions
compositions of of corrosion
corrosion products
products obtained in in the
the
tests with
simulation those collected from a leaking pipeline, the corroded specimens covered with corrosion
simulation teststests with
with those
those collected
collected from
from aa leaking
leaking pipeline,
pipeline, the
the corroded
corroded specimens
specimens covered
covered with
with
products were
corrosion investigated by EDS and XRD. Figure 8 shows the EDS analysis results. From the
corrosion products
products were were investigated
investigated by by EDS
EDS andand XRD.
XRD. Figure
Figure 88 shows
shows thethe EDS
EDS analysis
analysis results.
results.
figure,the
From the main elements include Fe and O for all the specimens, and the chemical compositions of
From the figure,
figure, the the main
main elements
elements include
include Fe Fe and
and O O for
for all
all the
the specimens,
specimens, and and the
the chemical
chemical
corrosion
compositions products are the same as those of the corrosion products collected from the leaking pipeline
compositions of of corrosion
corrosion products
products areare the
the same
same as as those
those of
of the
the corrosion
corrosion products
products collected
collected from
from
(Figure
the 5). Figure 9 presents the XRD spectra of corrosion products, namely, Fe3 O4 , Fe 2 O3 , FeO, FeCO3
the leaking pipeline (Figure 5). Figure 9 presents the XRD spectra of corrosion products, namely,
leaking pipeline (Figure 5). Figure 9 presents the XRD spectra of corrosion products, namely,
and
Fe FeO(OH). FeO(OH) 3has the largest share in the corrosion scale.
Fe33O
O44,, Fe
Fe22O
O33,, FeO,
FeO, FeCO
FeCO3 and
and FeO(OH).
FeO(OH). FeO(OH)
FeO(OH) hashas the
the largest
largest share
share in
in the
the corrosion
corrosion scale.
scale.
Materials 2018,11,
Materials2018, 11,1935
x FOR PEER REVIEW 99of
of17
17
Materials 2018, 11, x FOR PEER REVIEW 9 of 17

Figure 8. EDS analysis results of corrosion products: (a) full immersion, (b) half immersion and (c)
Figure8.8. EDSgas
Figure
high-pressure analysis results
results of
exposure. of corrosion
corrosionproducts:
products:(a)
(a)full
fullimmersion,
immersion,(b)(b)
half immersion
half and
immersion (c)
and
high-pressure
(c) gas
high-pressure exposure.
gas exposure.

Figure 9. XRD spectra of corrosion products.

Figure 9. XRD spectra of corrosion products.
Figure 9. XRD spectra of corrosion products.
3.3.2. Corrosion Rate
3.3.2. Corrosion Rate
3.3.2.The
Corrosion
specimens Rateused for corrosion rate analysis were washed in an acid solution to remove the
The specimens
corrosion products. Theused forsolution
acid corrosion wasrate analysisfollowing
prepared were washed in an
ISO 8407 acid500
using solution
mL ofto remove the
hydrochloric
The specimens
corrosion products. used
The for corrosion
acid solution ratewasanalysis were washed
prepared following in an
ISOacid solution
8407 usingto 500
remove
mL theof
acid (HCl, ρ = 1.19 g/mL), 3.5 g of hexamethylenetetramine, and diluted to 1000 mL with distilled
corrosion
hydrochloric products.
acid (HCl,The acid
ρ =then solution
1.19cleaned
g/mL), with was prepared
3.5 gethanol following
of hexamethylenetetramine, ISO 8407 using 500 mL of
water. The specimens were and dried with argon and dilutedas
gas stream, toshown
1000 mLin
hydrochloric
with distilled acid (HCl, ρ = 1.19 g/mL), 3.5 g of hexamethylenetetramine, and diluted to 1000 mL
Figure 10. Afterwater. Thethe
removing specimens
corrosionwere then the
products, cleaned with ethanol
specimens immersed and dried with water
in condensate argon and
gas
with
stream, distilled
as shown water. The
in were
Figure specimens
10. After were then cleaned with ethanol and dried with
immersedgas
argon
those exposed to gas corroded in removing
a mode ofthe corrosion
general products,
and localized the specimens
corrosion, respectively. in
stream, as shown in Figure 10. After removing the corrosion products,
condensate water and those exposed to gas were corroded in a mode of general and the specimens immersed in
localized
condensate water
corrosion, respectively.and those exposed to gas were corroded in a mode of general and localized
corrosion, respectively.
Materials 2018, 11, 1935 10 of 17
Materials 2018, 11, x FOR PEER REVIEW 10 of 17

Figure 10. Morphology of specimens after removing corrosion products: (a) full immersion, (b) half
immersion and (c) high-pressure gas exposure.
exposure.

The corrosion
The rates with
corrosion rates with different
different corrosion
corrosion failure
failure modes
modes were
were determined
determined by
by using
using different
different
approaches, as
approaches, as follows.
follows.

• General corrosion rate
The weight loss
The weight lossrate
ratecan
canbebeused
usedtoto evaluate
evaluate thethe general
general corrosion
corrosion rate.rate.
TheThe specimens
specimens before
before and
and
after the corrosion test were weighed three times, and the relevant weight loss data are listed in Tablein
after the corrosion test were weighed three times, and the relevant weight loss data are listed 5.
Table 5. Theloss
The weight weight lossconverted
rate was rate was intoconverted into loss
a thickness a thickness loss ratethe
rate to compare to general
compare the general
corrosion rate
corrosion
with rate with
the latter local the latter local
corrosion rate. corrosion
It is clear rate. It isgeneral
that the clear that the general
corrosion corrosion
rate of 2” pipe rate ofslowest,
is the 2″ pipe
is the slowest,
while whilerate
the corrosion the of
corrosion
4”and 6”rate of 4″and
pipes 6″ and
is faster pipes is faster andthe
substantially substantially
same. the same.

Table 5. Corrosion
Table 5. test results
Corrosion test results of
of full
full immersion.
immersion.

Surface Area Weight

Surface Area
Loss
Weight Loss
Corrosion Rate
Corrosion Rate
Specimen
Specimen No.
No. (mm 22)
(mm ) (g)
(g) (mm/y)
(mm/y)
11 1207.83
1207.83 0.1637
0.1637 1.12
1.12
2″
2” 2 1211.13
1211.13 0.1787
0.1787 1.13
1.13 1.15
1.15
33 1208.51
1208.51 0.1745
0.1745 1.19
1.19
5 1332.04
1332.04 0.2210
0.2210 1.37
1.37
4”
4″ 66 1341.24
1341.24 0.2203
0.2203 1.36
1.36 1.35
1.35
7 1335.49 0.2117 1.31
7 1335.49 0.2117 1.31
99 1298.14
1298.14 0.2198
0.2198 1.40
1.40
6” 10 1350.60 0.2164 1.32 1.34
6″ 10
11 1350.60
1333.63 0.2164
0.2075 1.32
1.29 1.34
11 1333.63 0.2075 1.29

• Local corrosion rate in a gaseous environment

 Local corrosion rate in a gaseous environment
For the specimens exposed to a simulated transmission gas environment, the condensation of
water droplets results in local corrosion. The depth of the corrosion pits was measured by a confocal
microscope to determine the local corrosion rate. Figure 11a presents the surface morphology of the
2” steel
2″ steel pipeline
pipeline specimen,
specimen, andand the
the depth
depth ofof the corrosion
corrosion pit was measured based on the profile
(Figure 11b) along the line in Figure 11a. The obtained 16.12 μm µm is the mean depth of the pit between
lines AA and
andBBininFigure
Figure11b.11b.With
Withthis method,
this method,10 corrosion
10 corrosionpitspits
werewere
measured for the
measured forspecimens from
the specimens
the 2” pipe section. The average local corrosion rate of these pits and the highest local
from the 2″ pipe section. The average local corrosion rate of these pits and the highest local corrosion corrosion rate
are listed
rate in Table
are listed 6. The
in Table 6. local corrosion
The local rates rates
corrosion for theforspecimens from from
the specimens the 4”the
and4″6” pipe
and 6″ sections were
pipe sections
also determined
were using using
also determined the samethe method, and theand
same method, results are alsoare
the results listed
alsoinlisted
Tablein6.Table
The 2”6. pipe haspipe
The 2″ the
slowest corrosion rate, while the 4”and 6” pipes have a faster and substantially identical
has the slowest corrosion rate, while the 4″and 6″ pipes have a faster and substantially identical corrosion rate.
This is similar
corrosion rate.to theisuniform
This similarcorrosion rate ofcorrosion
to the uniform the three rate
pipesofinthe
thethree
full immersion
pipes in thezone,
full except that
immersion
the corrosion rate of the material in the gas phase is slower than that in the full immersion
zone, except that the corrosion rate of the material in the gas phase is slower than that in the full zone.
immersion zone.
Materials 2018, 11,
Materials 2018, 11, 1935
x FOR PEER REVIEW 11 of
11 of 17
17

Figure 11. Surface morphology and profile of 2” morphology and (b)

2″ steel pipe specimen: (a) surface morphology
profile along the
the line.
line.

Table 6. Localized corrosion rate for the three steel pipelines.

Table 6. Localized corrosion rate for the three steel pipelines.
Corrosion
Corrosion Pits
Pits Depth
Depth CorrosionRate
Corrosion Rate
Specimen (µm) (mm/y)
Specimen (μm) (mm/y)
11 22 33 44 55 66 77 88 99 10
10 Ave.
Ave. Max.
Max.
2”2″ 15.46
15.46 16.11
16.11 15.46
15.46 16.08
16.08 15.88
15.88 16.12
16.12 15.65
15.65 16.09
16.09 16.22
16.22 16.12
16.12 1.0330
1.0330 1.0460
1.0460
4”4″ 19.36
19.36 19.64
19.64 18.98
18.98 19.17
19.17 19.55
19.55 18.87
18.87 19.86
19.86 19.52
19.52 18.64
18.64 19.06
19.06 1.2501
1.2501 1.2744
1.2744
6”6″ 18.69
18.69 19.38
19.38 19.93
19.93 18.76
18.76 19.46
19.46 18.66
18.66 19.86
19.86 19.84
19.84 19.37
19.37 18.99
18.99 1.2520
1.2520 1.2887
1.2887

 Uneven general corrosion near the waterline

• Uneven general corrosion near the waterline
General corrosion occurs below the waterline. However, the corrosion rate near the waterline is
General
different fromcorrosion occursfrom
that far away below thethe waterline.
waterline However,
(Figure the corrosion
12). The rate near the
residual thicknesses in waterline
the areas
is
exposed to gas, 1 mm below the waterline, 5 mm below the waterline, and near the bottom,inwere
different from that far away from the waterline (Figure 12). The residual thicknesses the
areas exposed to gas, 1 mm below the waterline, 5 mm below the waterline,
measured and recorded to determine the change rule of the corrosion rate near the waterline. Figureand near the bottom,
were measured
12a shows and recorded
the measuring to determine
positions. Here, thethewaterline
change rule of the corrosion
is considered rate near
the critical the waterline.
position at which
Figure
the overall thickness of the specimen begins to change. The region exposed to gas phase but position
12a shows the measuring positions. Here, the waterline is considered the critical without
at which thepit
a corrosion overall
was thickness
selected as of athe specimen
baseline, andbegins to change.
the corrosion The
rate region
was exposed
set at 0 mm/y. to gas
Thus,phase
the
but without a corrosion pit was selected as a baseline, and the corrosion rate
corrosion rate from the waterline to the bottom of the specimen can be determined, and the relevant was set at 0 mm/y.
Thus, the listed
data are corrosion rate from
in Table the the
7. For waterline
areas to the bottom
from of the specimen
5 mm below can betodetermined,
the waterline the bottomand the
of the
relevant data are listed in Table 7. For the areas from 5 mm below the waterline to
specimen, the corrosion rate of the three specimens is approximately the same as the results in Table the bottom of the
specimen, the corrosion
5. This means rateoutside
that the area of the three
5 mmspecimens
below theiswaterline
approximately
worksthe same
in the as the
same results in Table
environment 5.
as the
full immersion zone. Table 7 shows that the area 1 mm below the waterline exhibits a higher
Materials 2018, 11, 1935 12 of 17

This means
Materials 2018, that the area
11, x FOR PEERoutside
REVIEW 5mm below the waterline works in the same environment as the full
12 of 17
immersion zone. Table 7 shows that the area 1 mm below the waterline exhibits a higher corrosion
rate, whichrate,
corrosion is approximately two times higher
which is approximately than higher
two times that of than
the area
thatoutside
of the 5area
mmoutside
below the waterline.
5 mm below
It is waterline.
the believed that thebelieved
It is area in the
thatrange 5 mminbelow
the area the waterline
the range 5 mm belowshowsthe
severe corrosion,
waterline especially
shows severe
the 1 mm area
corrosion, below the
especially the 1waterline.
mm area below the waterline.

C B A

Figure 12. Morphology

Morphology of of specimen
specimen partly
partly immersed
immersed in simulated condensate:
condensate: (a)
(a) key
key positions,
positions,
A—area with
A—area withcorrosion
corrosion rate
rate of 0ofmm/y,
0 mm/y, B—waterline,
B—waterline, C—area
C—area near
near the the bottom;
bottom; (b) change(b)inchange in
thickness
thickness near the waterline.
near the waterline.

Table 7. Corrosion rate for specimens partly immersed in simulated condensate water (mm/y).
Table 7. Corrosion rate for specimens partly immersed in simulated condensate water (mm/y).
Location
Location
Specimen
Specimen
1 mm
1 mm Below
Below Waterline 5 5mm
Waterline mmBelow
Below Waterline
Waterline Near Bottom
Near Bottom
2″2” 2.3661
2.3661 1.2016
1.2016 1.2146
1.2146
4″4” 2.7675
2.7675 1.3642
1.3642 1.3578
1.3578
6” 2.7864 1.3489 1.3562
6″ 2.7864 1.3489 1.3562

4. Discussion
4. Discussion
The corrosion
The corrosionproducts
productsare arecomposed
composedofof FeFe
3O34O 4 , 2O
, Fe Fe32, O 3 , FeO,
FeO, FeCO FeCO and FeO(OH).
3 and3 FeO(OH). The CPSThevalue
CPS
value in Figures 6 and 9 show that FeO(OH) accounts for the largest
in Figures 6 and 9 show that FeO(OH) accounts for the largest proportion of corrosion products, proportion of corrosion
products, by
followed followed
FeCO3; by FeCO
thus, the3 ;corrosion
thus, the ofcorrosion
steel pipeof steel pipe is controlled
is controlled by oxygenby oxygen absorption
absorption corrosion
corrosion
[30–32] [30–32]
and CO2 and CO2 corrosion
corrosion [33–35]. The [33–35]. The of
presence presence
H2O is of H2 O is a necessary
a necessary condition condition that
that produces
produces FeO(OH) and FeCO
FeO(OH) and FeCO3 [36]. H2O [36]. H O and O together with
3 and O22 together2with CO2 are the key CO 2 are the key factors that induce
factors that induce pipeline
pipeline corrosion.
corrosion.
Oxygen absorption
Oxygen absorptioncorrosion
corrosionoccursoccurswhen
whensteel
steel is exposed
is exposed to environment
to an an environmentwithwith water
water and
and oxygen
oxygen [32]. [32].
Anodic reaction:
Anodic reaction:
Fe − 2e− →Fe2+ (1)
Fe − 2e−→Fe2+ (1)
Cathodic reaction:
Cathodic reaction:
O2 + 2H2 O + 4e− →4OH- (2)
O2 + 2H2O + 4e−→4OH‒ (2)
Fe2+ can be further oxidized to Fe3+:
4Fe2+ + O2→4Fe3+ + 2O2− (3)
Materials 2018, 11, 1935 13 of 17

Fe2+ can be further oxidized to Fe3+ :

4Fe2+ + O2 →4Fe3+ + 2O2− (3)

Then, the following multistep reactions and dehydration equilibria affect the formation of
corrosion product.
Fe2+ + 2OH− →Fe(OH)2 (4)

Fe3+ + 3OH− →Fe(OH)3 (5)

Fe(OH)2 →FeO + H2 O (6)

Fe(OH)3 →FeO(OH) + H2 O (7)

2FeO(OH)→Fe2 O3 + H2 O (8)

With limited dissolved oxygen, Fe2+ -containing materials, including FeO and Fe3 O4 , are favored.
High oxygen concentrations favor the formation of Fe3+ materials [37]. Given that FeO(OH) has a
large share in the corrosion scale, the oxygen concentration in the associated gas is high.
Oxygen concentration cell can be used to explain why the corrosion rate in the area of 1 mm below
the waterline is the highest among different corrosion environments. Great variation in the oxygen
concentration gradient occurs near the waterline [38]. The oxygen concentration decreases with water
depth and finally reaches a stable value [39]. When the different areas of the same metal are in contact
with a solution with different oxygen concentrations, the area with high oxygen concentration is good
for the oxygen absorption reaction, whereas the metal dissolution reaction in this area is inhibited to
a certain extent. By contrast, areas with low oxygen concentrations are subject to metal dissolution
as anodes. It can be speculated from Figure 12 that the oxygen concentration changes within 1 mm
below the waterline and then reaches a low stable value. The anodic reaction near the waterline is
suppressed; thus, the degree of corrosion is lighter than that in large water depth. After leaving the
waterline, the potential of the metal should not change after the oxygen concentration is stabilized.
However, the area 1 mm below the waterline, where the oxygen concentration has only reached a
stable low value, has a shorter conductive path than the areas farther from the waterline. Therefore,
the anodic current concentrates flow through this narrow area and results in severe corrosion.
The mechanism of waterline corrosion can explain why corrosion perforation mostly occurs at
the bottom of the actual gas pipeline. Figure 2 shows that numerous corrosion products are found at
the bottom of the pipe. Notably, the amount of corrosion products in the gas pipeline indicates the
degree of corrosion. The more products are present, the more severe the corrosion because although
the wall shear stress caused by the gas flow in the pipeline can influence the structure of corrosion
scale (Figure 4), this stress is insufficiently high to carry the corrosion product away [40]. Therefore,
the corrosion product is easily deposited in the position where corrosion occurs. Figure 2 shows that
the amount of corrosion products at the bottom of the pipe is considerably more than that in other
areas, thereby indicating severe corrosion at the bottom of the pipe. On-site sampling proves the
presence of condensate water at the bottom of the pipeline. When the depth of water is shallow, severe
corrosion will occur at the bottom (Figure 2a). When the depth exceeds 5 mm, severe corrosion will
occur at the symmetrical position on both sides of the bottom of the pipe (Figure 2b), and the severe
corrosion sites correspond to the areas at a short distance below the waterline. Water with a depth of
less than 5 mm is most likely to cause severe corrosion at the bottom of the pipe.
Corrosion of steel pipe in a gaseous environment only occurs in spots in which condensate water
droplets exist. Therefore, local corrosion occurs. Tables 5 and 6 show that the localized corrosion rate of the
steel pipeline in a gaseous environment is slightly lower than the general corrosion rate of the specimen
immersed in condensate. Analysis indicates two reasons for this situation. First, no autocatalysis of
occluded corrosion cell occurs in the formation of a corrosion pit because the area around the pit
without condensate cannot form a corrosion couple with the area with condensate droplets. Without
Materials 2018, 11, 1935 14 of 17

electrolytes, no electrochemical reactions can occur. Localized corrosion in the gas phase is actually
uniform corrosion that occurs at the precipitation location of condensate water. The other reason why
the corrosion rate in a gaseous environment is lower than that in full immersion zones is that corrosion
products are not easily washed away in gas, and the anodic process will be impeded by the formation
of corrosion products.
The pH value of collected condensate water is 6.74; thus, acid gas corrosion should not be serious.
Carbonic acid formation by CO2 dissolution and hydration is the major reaction during CO2 corrosion.
At pH < 4, the hydrogen reduction is the dominant cathodic reaction, whereas direct reduction of
carbonic acid becomes important at pH > 4 [41]. High pH results in a decreased solubility of iron
carbonate and leads to an increased precipitation rate and high scaling tendency, and the higher pH of
6.6 results in the formation of several protective scales, which are reflected by a rapid decrease of the
corrosion rate with time [42]. The pH value of the medium in the studied gas pipeline is as high as 6.74;
however, corrosion perforation proves that the corrosion product does not have effective protection.
This result also shows that CO2 does not play a leading role in the corrosion of the studied pipeline.
Generally, the dissolution of O2 and CO2 together with mineral ions let the condensate water be a
strong corrosive medium, thereby resulting in steel pipe corrosion. The transport medium cannot carry
the corrosion products away; hence, the corrosion products will increase with time, although their
compactness is affected by the medium flow rate. Pores or cracks occur in the corrosion scale when the
flow rate of the transmission medium is high (Figure 4e,f). The outer corrosive medium can penetrate
through the corrosion scale along the paths formed by pores and cracks, thereby causing corrosion
beneath the corrosion products. Furthermore, the oxygen concentration beneath the corrosion products
is considerably lower than the open zones. The difference in oxygen concentration can accelerate the
corrosion beneath the corrosion products, thereby promoting local corrosion.
H2 O and O2 together with CO2 are the key factors that induce pipeline corrosion; thus, three
measures can be taken to control the corrosion failure of gas pipeline, namely, dehydration,
deoxygenation, and application of inhibitors. However, no standard is used to specify the oxygen
content in natural gas, and deoxidization is not a common method in controlling gas pipeline corrosion.
Here, we recommend deoxygenation and application of inhibitors. The corrosion rate of gas pipelines
is directly related to the water content in natural gas [43]; therefore, strict dehydration is one of the
most effective ways to control corrosion of gas transmission steel pipelines [44]. Given the lack of
mature application experience of inhibitors at vapor/gas atmosphere as a reference, the injection of
inhibitors can only be used as an auxiliary method.

5. Conclusions
(1) The corrosion products are mainly composed of FeO(OH) and FeCO3 , and the corrosion of steel
pipe is controlled by oxygen absorption corrosion and CO2 corrosion. Flow rates of the transmission
medium affect the structure of the corrosion scale, and defects in scale promote the occurrence of
medium penetration and local corrosion.
(2) The specimens immersed in condensate are generally corroded, and the specimens exposed
to a gaseous environment are locally corroded. The 1 mm area below the waterline has the highest
corrosion rate, which is approximately two times higher than that of the fully immersed area far away
from the waterline. Condensate with a depth of less than 5 mm is most likely to cause severe corrosion.
(3) The localized corrosion rate in a gaseous environment is slightly lower than the uniform
corrosion rate in full immersion zone. Therefore, no autocatalysis of the occluded corrosion cell occurs
in the formation of the corrosion pit, and a uniform corrosion essentially takes place in the precipitation
location of the condensate water.
(4) The presence of water is a necessary condition for oxygen absorption corrosion and hydrogen
evolution corrosion. Dehydration treatment of the transmission gas is suggested to be the main
measure to control the corrosion failure of the associated gas pipeline.
Materials 2018, 11, 1935 15 of 17

Author Contributions: Conceptualization, W.Z.; methodology, W.Z. and T.Z.; validation, T.Z.; formal analysis,
W.Z. and T.Z.; investigation, T.Z., Y.W., J.Q., and Z.W.; data curation, T.Z.; writing, W.Z.; funding acquisition, W.Z.
Funding: This research was funded by Sinopec Energy Conservation and Environmental Protection Engineering
Technology Corporation.
Acknowledgments: Thanks to Sheng Wan, who is the project manager from Sinopec, for collecting pipe sections
and providing the compositions of transmission gas and condensate water, and for agreeing to publish this article.
Thanks to Tingting Li, Qiushi Deng, and Yujiao Zhao for participating in some of the experimental work.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Fan, L.; Zhang, F. The study on effects of associated gas generations in oilfield grid. In Environmental
Protection and Resources Exploitation, PTS 1–3, Proceedings of International Conference on Advances in Energy and
Environmental Science, Guangzhou, China, 30–31 July 2013; Trans Tech Publications Ltd.: Guangzhou, China,
2013; pp. 2589–2594.
2. Lawal, K.A.; Ovuru, M.I.; Eyitayo, S.I.; Matemilola, S.; Adeniyi, A.T. Underground storage as a solution for
stranded associated gas in oil fields. J. Petrol. Sci. Eng. 2017, 150, 366–375. [CrossRef]
3. Moreno, J.; Badawy, A.; Kartoatmodjo, G.; AlShuraiqi, H.; Zulkhifly, F.; Tan, L.; Friedel, T. Flaring,
gas injection and reservoir management optimization: Preserving reservoir energy maximizes recovery
and prolong the life of an ageing brown field. In Proceedings of the Asia Pacific Oil and Gas Conference &
Exhibition, Jakarta, Indonesia, 4–6 August 2009; Society of Petroleum Engineers: Jakarta, Indonesia, 2009;
pp. 262–270.
4. Kaya, E.; Zarrouk, S.J. Reinjection of greenhouse gases into geothermal reservoirs. Int. J. Greenh. Gas Con.
2017, 67, 111–129. [CrossRef]
5. Zheng, D.; Xu, H.; Wang, J.; Sun, J.; Zhao, K.; Li, C.; Shi, L.; Tang, L. Key evaluation techniques in the
process of gas reservoir being converted into underground gas storage. Petrol. Explor. Dev. 2017, 44, 840–849.
[CrossRef]
6. Blann, J.R.; Laville, G.M. Gas lifting a major oil field in Argentina with high CO2 content associated gas.
In Proceedings of the 1995 SPE Annual Technical Conference and Exhibition, Dallas, TX, USA, 22–25 October
1995; Society of Petroleum Engineers (SPE): Dallas, TX, USA, 1995; pp. 83–91.
7. Misener, R.; Gounaris, C.E.; Floudas, C.A. Global optimization of gas lifting operations: A comparative
study of piecewise linear formulations. Ind. Eng. Chem. Res. 2009, 48, 6098–6104. [CrossRef]
8. Javidi, M.; Bekhrad, S. Failure analysis of a wet gas pipeline due to localised CO2 corrosion. Eng. Fail. Anal.
2018, 89, 46–56. [CrossRef]
9. Mansoori, H.; Mirzaee, R.; Esmaeilzadeh, F.; Vojood, A.; Dowrani, A.S. Pitting corrosion failure analysis of a
wet gas pipeline. Eng. Fail. Anal. 2017, 82, 16–25. [CrossRef]
10. Laumb, J.D.; Glazewski, K.A.; Hamling, J.A.; Azenkeng, A.; Kalenze, N.; Watson, T.L. Corrosion and failure
assessment for CO2 EOR and associated storage in the Weyburn Field. Eng. Procedia 2017, 114, 5173–5181.
[CrossRef]
11. Ilman, M.N. Analysis of internal corrosion in subsea oil pipeline. Case Stud. Eng. Fail. Anal. 2014, 2, 1–8.
[CrossRef]
12. Pouraria, H.; Seo, J.K.; Paik, J.K. A numerical study on water wetting associated with the internal corrosion
of oil pipelines. Ocean Eng. 2016, 122, 105–117. [CrossRef]
13. Ryder, J.; Tilekar, J.; Srinivasan, S.; Yap, K.M. Advanced internal corrosion direct assessment methodology
for wet gas/dry gas pipelines. In Proceedings of the Corrosion 2014: Collaborate Educate Innovate Mitigate,
San Antonio, TX, USA, 9–13 March 2014; National Assoc. of Corrosion Engineers International: San Antonio,
TX, USA, 2014.
14. Ren, L.; Jiang, T.; Jia, Z.G.; Li, D.S.; Yuan, C.L.; Li, H.N. Pipeline corrosion and leakage monitoring based on
the distributed optical fiber sensing technology. Measurement 2018, 122, 57–65. [CrossRef]
15. Elhoud, A.; Jewilli, F.; Abouswa, K.; Rageai, O. Failure analysis: Internal corrosion rupture of a 6-in gas line.
Mater. Perform. 2012, 51, 70–73.
16. Zhang, H.; Lan, H.Q. A review of internal corrosion mechanism and experimental study for pipelines based
on multiphase flow. Corros. Rev. 2017, 35, 425–444. [CrossRef]
Materials 2018, 11, 1935 16 of 17

17. Belarbi, Z.; Farelas, F.; Singer, M.; Nei, S. Role of amines in the mitigation of CO2 top of the line corrosion.
Corrosion 2016, 72, 1300–1310. [CrossRef]
18. Ifezue, D. Root cause failure analysis of corrosion in wet gas piping. J. Fail. Anal. Prev. 2017, 17, 971–978.
[CrossRef]
19. Eckert, R.B. Internal corrosion health check advised for liquid and gas pipeline operators. Mater. Perform.
2012, 51, 48–53.
20. Wang, B.; Liu, X.L.; Xiong, Z.X.; Cheng, J.C.; Yang, B.; Yu, C. Corrosion reasons and control measures of a
natural gas pipeline. Surf. Technol. 2018, 47, 89–94. [CrossRef]
21. Qiao, Q.; Cheng, G.; Wu, W.; Li, Y.; Huang, H.; Wei, Z. Failure analysis of corrosion at an inhomogeneous
welded joint in a natural gas gathering pipeline considering the combined action of multiple factors.
Eng. Fail. Anal. 2016, 64, 126–143. [CrossRef]
22. Choi, Y.S.; Nesic, S.; Young, D. Effect of impurities on the corrosion behavior of CO2 transmission pipeline
steel in supercritical CO2 -water environments. Environ. Sci. Technol. 2010, 44, 9233–9238. [CrossRef]
[PubMed]
23. Singer, M.; Al-Khamis, J.; Nesic, S. Experimental study of sour top-of-the-line corrosion using a novel
experimental setup. Corrosion 2013, 69, 624–638. [CrossRef]
24. Singer, M.; Camacho, A.; Brown, B.; Neic, S. Sour top-of-The-line corrosion in the presence of acetic acid.
Corrosion 2011, 67, 0850031–08500316. [CrossRef]
25. Kovalenko, S.Y.; Rybakov, A.O.; Klymenko, A.V.; Shytova, L.H. Corrosion of the internal wall of a field gas
pipeline. Mater. Sci. 2012, 48, 225–230. [CrossRef]
26. Wan, H.X.; Yang, X.J.; Liu, Z.Y.; Song, D.D.; Du, C.W.; Li, X.G. Pitting behavior of L415 pipeline steel in
simulated leaching liquid environment. J. Mater. Eng. Perform. 2017, 26, 715–722. [CrossRef]
27. Sun, C.; Sun, J.; Wang, Y.; Sui, P.; Lin, X.; Liu, H.; Cheng, X.; Zhou, M. Effect of impurity interaction
on the corrosion film characteristics and corrosion morphology evolution of X65 steel in water-saturated
supercritical CO2 system. Int. J. Greenh. Gas Con. 2017, 65, 117–127. [CrossRef]
28. Sun, C.; Wang, Y.; Sun, J.; Lin, X.; Li, X.; Liu, H.; Cheng, X. Effect of impurity on the corrosion behavior of
X65 steel in water-saturated supercritical CO2 system. J. Supercrit. Fluid 2016, 116, 70–82. [CrossRef]
29. Zhao, W.-M.; Wang, Y.; Dong, L.-X.; Wu, K.-Y.; Xue, J. Corrosion mechanism of NiCrBSi coatings deposited
by HVOF. Surf. Coat. Technol. 2005, 190, 293–298. [CrossRef]
30. De Oliveira, L.A.; Correa, O.V.; dos Santos, D.J.; Páez, A.A.Z.; de Oliveira, M.C.L.; Antunes, R.A. Effect of
silicate-based films on the corrosion behavior of the API 5L X80 pipeline steel. Corros. Sci. 2018, 139, 21–34.
[CrossRef]
31. Wang, S.; Du, C.; Li, X.; Liu, Z.; Zhu, M.; Zhang, D. Field corrosion characterization of soil corrosion of X70
pipeline steel in a red clay soil. Prog. Nat. Sci. Mater. Int. 2015, 25, 242–250. [CrossRef]
32. Liu, H.; Xu, D.; Dao, A.Q.; Zhang, G.; Lv, Y.; Liu, H. Study of corrosion behavior and mechanism of carbon
steel in the presence of Chlorella vulgaris. Corros. Sci. 2015, 101, 84–93. [CrossRef]
33. Yu, J.; Wang, H.; Yu, Y.; Luo, Z.; Liu, W.; Wang, C. Corrosion behavior of X65 pipeline steel: Comparison of
wet–Dry cycle and full immersion. Corros. Sci. 2018, 133, 276–287. [CrossRef]
34. De Motte, R.A.; Barker, R.; Burkle, D.; Vargas, S.M.; Neville, A. The early stages of FeCO3 scale formation
kinetics in CO2 corrosion. Mater. Chem. Phys. 2018, 216, 102–111. [CrossRef]
35. Ma, Z.; Yang, Y.; Brown, B.; Nesic, S.; Singer, M. Investigation of precipitation kinetics of FeCO3 by EQCM.
Corros. Sci. 2018, 141, 195–202. [CrossRef]
36. Pfennig, A.; Wolthusen, H.; Kranzmann, A. Unusual corrosion behavior of 1.4542 exposed a laboratory saline
aquifer water CCS-environment. Eng. Procedia 2017, 114, 5229–5240. [CrossRef]
37. Speight, J.G. Oil and Gas Corrosion Prevention: From Surface Facilities to Refineries; Elsevier Inc.: Waltham, MA, USA,
2014; pp. 1–151.
38. Wu, J.; Wang, P.; Gao, J.; Tan, F.; Zhang, D.; Cheng, Y.; Chen, S. Comparison of water-line corrosion processes
in natural and artificial seawater: The role of microbes. Electrochem. Commun. 2017, 80, 9–15. [CrossRef]
39. Jeffrey, R.; Melchers, R.E. Corrosion of vertical mild steel strips in seawater. Corros. Sci. 2009, 51, 2291–2297.
[CrossRef]
40. Li, W.; Pots, B.F.M.; Brown, B.; Kee, K.E.; Nesic, S. A direct measurement of wall shear stress in multiphase
flow-Is it an important parameter in CO2 corrosion of carbon steel pipelines? Corros. Sci. 2016, 110, 35–45.
[CrossRef]
Materials 2018, 11, 1935 17 of 17

41. Hatami, S.; Ghaderi-Ardakani, A.; Niknejad-Khomami, M.; Karimi-Malekabadi, F.; Rasaei, M.R.;
Mohammadi, A.H. On the prediction of CO2 corrosion in petroleum industry. J. Supercrit. Fluid 2016,
117, 108–112. [CrossRef]
42. Nesic, S. Key issues related to modelling of internal corrosion of oil and gas pipelines–A review. Corros. Sci.
2007, 49, 4308–4338. [CrossRef]
43. Kong, Z.Y.; Mahmoud, A.; Liu, S.; Sunarso, J. Revamping existing glycol technologies in natural gas
dehydration to improve the purity and absorption efficiency: Available methods and recent developments.
J. Nat. Gas Sci. Eng. 2018, 56, 486–503. [CrossRef]
44. Sakheta, A.; Zahid, U. Process simulation of dehydration unit for the comparative analysis of natural gas
processing and carbon capture application. Chem. Eng. Res. Des. 2018, 137, 75–88. [CrossRef]

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).