Wear 404–405 (2018) 143–152

Contents lists available at ScienceDirect

Wear
journal homepage: www.elsevier.com/locate/wear

Evaluation of high shear inhibitor performance in CO2-containing flow- T

induced corrosion and erosion-corrosion environments in the presence and
absence of iron carbonate films
⁎
E.V. Senatorea, , W. Talebb, J. Owenb, Y. Huab, J.A.C. Ponciano Gomesa, R. Barkerb, A. Nevilleb
a
Federal University of Rio deJaneiro-LabCorr, Rio de Janeiro, Brazil
b
University of Leeds, Leeds, United Kingdom

A R T I C LE I N FO A B S T R A C T

Keywords: Carbon steel pipeline degradation occurs as a result of erosion-corrosion during oil and gas production. Sand
Jet impingement particles contribute to this effect when they are present in conjunction with a high flow velocity. In carbon
Carbon dioxide corrosion dioxide (CO2) environments, and under certain conditions, the corrosion rate of the steel can be reduced by the
Erosion-corrosion formation of a protective iron carbonate (FeCO3) layer. This work assesses the ability of FeCO3 to protect the
Iron carbonate
underlying steel in flow-induced corrosion and erosion-corrosion environments. Autoclave tests are performed at
Corrosion inhibitor
60 °C and 100 bar in a 1 wt%NaCl CO2-saturated solution for a duration of 48 h to develop 60 µm thick FeCO3
films. The film-covered samples were then transferred into a submerged impinging jet (SIJ) apparatus to assess
their ability to resist both flow-induced corrosion and erosion-corrosion environments at 25 °C and a flow ve-
locity of 15 m/s (both with and without 1000 mg/L sand). Tests were also conducted in the presence of a
commercially available corrosion inhibitor to evaluate the interaction. Results indicate that the FeCO3 layer is
able to considerably suppress corrosion of the carbon steel substrate. Experiments in the presence of both the
FeCO3 film and corrosion inhibitor demonstrated that there is a notable synergistic effect between these two
components in providing resistance to erosion-corrosion.

1. Introduction A number of studies have examined the erosion-corrosion resistance

of carbon steel in CO2-containing environments [4,5,7]. These studies
The corrosion of carbon steel pipelines in carbon dioxide (CO2)- typically focus on quantifying the resistance of carbon steel in condi-
containing environments is a common occurrence within the oil and gas tions where no corrosion product formation is anticipated, or consider
industry [1]. Dissolved CO2 within the brine component of production the potential of corrosion inhibitors in mitigating not only steel dis-
fluids permits the speciation of carbonic acid (H2CO3), which can result solution, but in alleviating the erosion component of material loss. A
in the electrochemical dissolution of carbon steel due to its low corro- number of studies have demonstrated that the action of the corrosion
sion resistance [2]. inhibitor can involve both the reduction of metal loss during mechan-
In addition to electrochemical dissolution, the presence of solid ical erosion from particle impingement and suppression of active cor-
particles in conjunction with the corrosive production fluid can result in rosion. Previous studies have also shown that corrosion inhibitors are
the pipeline material being subjected to erosive conditions as well as a able to adsorb onto the surface of sand particles. This can have the
corrosive environment [3]. This results in a degradation mechanism detrimental effect of reducing the bulk inhibitor concentration which
which is termed as ‘erosion-corrosion’. Erosion–corrosion is a form of influences the adsorption onto the steel surface [8]. However, it has
tribo-corrosion material loss mechanism caused by flowing fluid (ty- been argued that the adsorption of inhibitors onto sand may be bene-
pically in the presence of sand) degrading both the surface layer (e.g. ficial in terms of hindering their ability to cause erosion of the substrate
passive films or corrosion products) and the base metal through a [7,9]. The capability of these chemicals to lower electrochemical cor-
combination of mechanical removal and material dissolution. The sy- rosion reactions and the mechanical damage associated with particle
nergistic effects between erosion and corrosion are well documented impingement is well documented, but the underlying mechanisms have
and can significantly exceed degradation rates compared to the sum of been the subject of less attention [4].
the individual processes acting separately [4–6]. As previously stated, a considerable amount of research has been

⁎
Corresponding author.
E-mail address: [email protected] (E.V. Senatore).

https://doi.org/10.1016/j.wear.2018.03.014
Received 5 January 2018; Received in revised form 21 March 2018; Accepted 21 March 2018
Available online 22 March 2018
0043-1648/ © 2018 Elsevier B.V. All rights reserved.
E.V. Senatore et al. Wear 404–405 (2018) 143–152

directed towards understanding erosion-corrosion synergies for carbon Samples used for SIJ mass loss measurements or FeCO3 film-for-
steel in CO2 environments in the absence and presence of corrosion mation experiments in the autoclave were stored in a vacuum de-
inhibitors [4,6–10]. This research typically focuses on conditions where siccator until required and weighed immediately before use on an
corrosion product formation does not occur on carbon steel. However, electronic balance to within an accuracy of 0.01 mg. To obtain elec-
the formation of iron carbonate (FeCO3) is a common observation in trochemical measurements from carbon steel samples within the SIJ,
CO2-containing environments and readily precipitates onto the inner wires were soldered to the back of the steel samples before mounting
pipeline if a critical level of supersaturation is exceeded. The FeCO3 them in a non-conducting resin. In the case of the FeCO3 filmed sam-
layer can afford considerable corrosion protection to carbon steel by ples, the back of each disc was carefully ground in the absence of water
blocking active sites on the substrate surface, and developing a diffu- before attaching the wire using conductive glue. The sample was then
sion barrier to electrochemically active species [11]. Considering that inverted in the mold on a holder and resin was injected carefully around
this layer can develop to be the order of 10's or 100's of μm thick, there the back and sides of the specimen until it reached a height just below
is the potential for it to mitigate the erosive effect generated by sand the top of the FeCO3 filmed surface. This resulted in one face of the
particle impingement. This paper focuses on addressing this knowledge sample being exposed to the jet, consistent with the mass loss samples
gap by evaluating the effect of a 60 µm thick FeCO3 layer in its ability to which were placed in holders to expose only one face of the sample to
withstand flow-induced corrosion and erosion-corrosion environments, the jet with the aid of an O-ring seal.
both in the absence and presence of a commercially available corrosion
inhibitor. The work initially assesses the flow-induced corrosion and 2.2. Autoclave film-formation procedure
erosion-corrosion performance of API 5L X65 carbon steel in the ab-
sence of an FeCO3 layer, both with and without corrosion inhibitor A selection of steel specimens were pre-filmed with FeCO3 prior to
using a submerged impinging jet (SIJ). Test conditions selected were a insertion in the SIJ. The FeCO3 layers developed on the surface were
3.5 wt%NaCl CO2-saturated solution at 25 °C, pH 4.7 and 15 m/s, with created by placing five of the 25 mm diameter samples within the au-
and without the presence of 1000 mg/L sand to create a flow-induced toclave, generating a total surface area of ~ 24.5 cm2 exposed to 200 ml
corrosion and erosion-corrosion environment, respectively. The ex- of solution.
periments are then repeated under the exact same experimental con- With regards to the autoclave procedure, the prepared 1 wt%NaCl
ditions for a steel sample whereby an FeCO3 film has been generated in solution used in each experiment was de-aerated with CO2 in a separate
advance on the steel surface at 100 bar and 60 °C for 48 h in a 1 wt% container for a minimum of 12 h prior to use. Specimens were then
NaCl CO2-saturated solution using an autoclave system. suspended inside the autoclave on non-conducting wires whilst en-
In addition, the experiments are performed with and without the suring they were not in contact with one another or the walls of the
presence of corrosion inhibitor to determine the interaction with a wet- autoclave to eliminate galvanic effects. The prepared, CO2-saturated
ground surface and a FeCO3 filmed surface in each environment. solution was then delivered into the autoclave at ambient temperature
Although testing of corrosion inhibitors is usually carried out so to and pressure before sealing. All lines to the vessel, as well as the vessel
avoid or delay the FeCO3 protective layer, the inhibitor potency in the itself were then purged with CO2 and evacuated multiple times to en-
presence of an FeCO3 partially or fully covering the steel also deserves sure removal of CO2 within the system. Additional CO2 was then
attention and is mainly relevant for parts of pipelines where rapid transferred into the autoclave before it was heated and pressurised to
formation of FeCO3 is expected and/or there is intermittency in the the correct temperature and pressure (100 bar and 60 °C in this in-
injection of inhibitor which allows the film to develop [12]. Usually, it stance). The starting point of the experiment was taken from the point
is assumed that the corrosion protection is achieved when the FeCO3 is at which the operating conditions were reached in the autoclave, with
not yet developed while the corrosion inhibitor effect is negligible if the experiments running for a total of 48 h.
FeCO3 film is already fully covering the steel surface, but contradictory
conclusions at various conditions have already been uncovered [13]. 2.3. Flow-induced corrosion and erosion-corrosion experiments
Moreover the protection provided by the FeCO3 when sand is produced
is still not well defined. The SIJ used for erosion-corrosion testing shown in Fig. 1 consisted
of a reservoir with a 50L capacity, which was filled with the test so-
lution and recirculated through a dual nozzle arrangement before im-
2. Experimental setup and procedure pinging onto two specimens set at a distance of 5 mm below the exit of
the nozzles at a velocity of 15 m/s. The flow velocity was measured at
2.1. Materials and sample preparation the exit of the 4 mm diameter nozzles and calibrated prior to starting
the test. The temperature of the solution was maintained at 25 °C
The material used in this study is an API 5 L X65 carbon steel which throughout the entire experiment.
is commonly employed in the oil and gas industry due to its adequate Two types of experiments were performed in this study; flow-in-
mechanical properties and low cost [14]. The steel possesses a ferritic- duced corrosion (in the absence of sand) and erosion-corrosion (in the
pearlitic microstructure and the elemental composition is provided in presence of sand particles). For both experiments, CO2 was bubbled into
Table 1. The specimen geometry consists of discs with an exposed area the 50L, 3.5 wt%NaCl solution for a minimum of 12 h to reduce the
of 4.9 cm2 (25 mm diameter) to the electrolyte in all experiments. dissolved oxygen concentration. Both mass loss samples and electro-
Surface preparation prior to SIJ or autoclave experiments consisted of chemical samples were placed in customised holders which ensured
wet-grinding the sample using 1200 grit silicon carbide abrasive paper that only one face of the sample (with an exposed area of 4.9 cm2) was
before degreasing with acetone, followed by high purity ethanol and exposed to the electrolyte within the rig. Prior to starting the SIJ ex-
drying with compressed air. periment, samples were placed into the holders and lowered into the
rig. The system was then sealed with the exception of an outlet hole
Table 1 through which CO2 could escape as the system was continuously purged
Elemental composition of API 5L X65 steel (wt%).
with CO2 throughout the experiment to prevent oxygen ingress. For
C Si Mn P S Cr Mo Ni erosion-corrosion experiments, 1000 mg/L of sand was added to the
system. The sand particles possessed a mean diameter of ~ 250 µm,
.12 0.18 1.27 0.008 0.002 0.11 0.17 0.07
with an image of the particles provided in Fig. 2. For all erosion-cor-
Cu Sn Al B Nb Ti V Fe
0.12 0.008 0.022 0.0005 0.054 0.001 0.057 Balance rosion experiments, the sand was added to the system after the pump
was started to ensure an even distribution of sand particles throughout

144
E.V. Senatore et al. Wear 404–405 (2018) 143–152

Fig. 1. The submerged impingement jet (SIJ) apparatus used for flow-induced corrosion and erosion-corrosion testing.

2.4. Corrosion rate determination using mass loss and electrochemical

measurements

The evaluation of the degradation rate was conducted using a

combination of mass loss and electrochemical measurements. Mass loss
measurements were conducted on erosion-corrosion samples while
electrochemical measurements were performed on both erosion-corro-
sion and flow-induced corrosion samples. After 6 h of exposure, the
mass loss samples were removed from the rig, cleaned with Clarke's
solution (20 g antimony trioxide + 50 g stannous chloride + 1000 ml
38% hydrochloric acid) in accordance with ASTM Standard
G1–03 [15], rinsed with distilled water, dried with compressed air and
weighed to determine the mass loss.
Electrochemical measurements involved the use of the Linear
Polarisation Resistance (LPR) technique which was implemented to
provide a general corrosion rate across the entire steel surface. The SIJ
Fig. 2. SEM images of HST60 sand particles used in this study; mean diameter is
was integrated with a three electrode cell which comprised of a working
~ 250 µm. electrode, a platinum auxiliary electrode and a Ag/AgCl reference
electrode. The LPR measurements were performed over the first 4.5 h of
the 6 h experiment. This consisted of scanning from 15 mV more ne-
the rig. A fresh supply of sand was used in every experiment to ensure
gative than the open circuit potential (OCP) of the steel sample to
consistent results.
15 mV more positive than OCP at a scan rate of 0.333 mV/s. After LPR
Both flow-induced corrosion and erosion-corrosion experiments
measurements were complete, potentiodynamic polarisation sweeps
were performed in a 3.5 wt%NaCl solution set at a pH of 4.7 and
were conducted by performing anodic or cathodic scans. This consisted
temperature of 25 °C, with a flow velocity of 15 m/s exiting the nozzle.
of polarising each sample from OCP to either + 300 mV or −500 mV vs
Flow-induced corrosion tests were conducted on wet-ground and FeCO3
the OCP at a scan rate of 0.333 mV/s, completing the 6 h test duration.
pre-filmed samples in the absence and presence of 100 ppm corrosion
AC impedance was used in a frequency range of 20,000–0.1 Hz purely
inhibitor for a total of 6 h. The same experimental matrix was utilised in
to measure the solution resistance (Rs) which was used in conjunction
erosion-corrosion environments.
with the polarisation resistance determined from the LPR method to
In relation to the application of corrosion inhibitor, a commercial
enable the charge transfer resistance (Rct) to be determined using Eq.
high shear CO2 corrosion inhibitor was used in this study at a con-
(1):
centration of 100 ppm. The chemical package is based on a combination
of 2-butoxyethanol, quaternary ammonium compounds and amines. R ct = Rp − Rs (1)
Corrosion inhibitor was added once the pump started to ensure thor-
ough mixing of the inhibitor in the system. The open circuit potential The corrosion current density, icorr, could then be calculated using
(OCP) of the steel was then monitored and allowed to stabilise for Eq. (2), taking into account the measured anodic and cathodic Tafel
5 mins before electrochemical measurements commenced. constant determined at the end of each experiment:
The full set of experiments is provided in Table 2 for clarity to
B 1 βa βc
summarise the entire test matrix. icorr = =
R ct R ct 2.303(βc − βa) (2)

145
E.V. Senatore et al. Wear 404–405 (2018) 143–152

Table 2
Summary of test conditions evaluated using the submerged impinging jet (SIJ).
Material and operating conditions Corrosion inhibition Flow regime Sand concentration Initial surface conditions of
sample

API 5L X65 steel, 25 °C, 15 m/s pH 4.7 3.5 wt.% NaCl 90° No corrosion inhibitor Flow-induced 0 mg/L Wet-ground surface
impact angle 6 h duration corrosion FeCO3 pre-filmed surface
Erosion-corrosion 1000 mg/L Wet-ground surface
FeCO3 pre-filmed surface
100 ppm corrosion Flow-induced 0 mg/L Wet-ground surface
inhibitor corrosion FeCO3 pre-filmed surface
Erosion-corrosion 1000 mg/L Wet-ground surface
FeCO3 pre-filmed surface

where B is the Stern-Geary coefficient, βa is the anodic Tafel constant

and βc is the cathodic Tafel constant. The general corrosion rate is then
calculated from Eq. (3):
EW .icorr
CR = 3.27
ρ (3)

Where CR is the corrosion rate in mm/year, EW is the equivalent weight

of steel (which is equal to the molecular weight of steel (55.847 g/mol)
divided by the number of electrons involved in the corrosion process
(2)), ρ is the density of the steel in g/cm3 and 3.27 is a conversion
factor.

2.5. Surface analysis

SEM was performed on samples using a Carl Zeiss EVO MA15 SEM
to evaluate the morphology and coverage of corrosion products on the
sample surfaces. All images were collected at a working distance of
approximately 8 mm with an accelerating voltage of 20 kV.
Samples were profiled using a Bruker NPFLEX white light inter-
ferometer. 3D profiles of the surfaces after each test were measured
using this technique to compare the material loss as a result of the
impacts from sand particles.
X-ray diffraction (XRD) measurements were performed on a
Panalytical X′pert multipurpose diffractometer with a voltage of 40 kV
and an intensity of 40 mA using dual Cu Kα1+2 radiations with
10 × 10 mm programmable divergence slits.

3. Results and discussion

3.1. Flow-induced corrosion and erosion-corrosion behaviour of wet ground

samples Fig. 3. LPR corrosion rates of API 5L X65 as a function of time for wet-ground
samples exposed to (a) flow-induced corrosion and (b) erosion-corrosion con-
3.1.1. Electrochemical and mass loss results of wet-ground samples ditions; experiments are in the presence and absence of 100 ppm corrosion
Initial tests were conducted in the absence of the pre-formedFeCO3 inhibitor (CI) and 1000 mg/L sand; test conditions are 15 m/s, 25 °C and pH
film in flow-induced corrosion and erosion corrosion environments, 4.7in a 3.5 wt%NaCl solution CO2-saturated at 1 bar total pressure.
both with and without the addition of corrosion inhibitor to enable
comparison with the tests when theFeCO3 corrosion product was pre- Table 3
sent. From these full results, it is possible to assess the role of the FeCO3 Stern-Geary coefficients used in blank and inhibited flow-induced corrosion and
film and its synergistic or antagonistic behaviour with the corrosion erosion-corrosion tests.
inhibitor.
0 mg/L sand 0 mg/L sand 1000 mg/L 1000 mg/L
Fig. 3 compares the corrosion rates calculated from LPR measure- 0 ppm 100 ppm sand 0 ppm sand 100 ppm
ments in the absence and presence of sand and corrosion inhibitor both corrosion corrosion corrosion corrosion
individually and collectively. Table 3 presents the Stern-Geary coeffi- inhibitor inhibitor inhibitor inhibitor
cients (B) applied for the calculation of corrosion rate using the LPR
Stern-Geary 24.92 23.78 25.17 23.33
data. It should be stressed here that the analysis of the LPR technique coefficient
and its representation in Fig. 3 assumes a uniform dissolution rate (B) mV/
across the entire surface of the steel sample. Based on profilometry decade
analysis of samples subjected to flow-induced corrosion conditions, no
wear profile existed, indicating no significant accentuation of dissolu-
tion rate at the centre of the sample due to the differing flow char- marginally greater than in flow-induced corrosion, this suggests that
acteristics across the diameter of the sample. Given that the in situ the plastic deformation created at the centre of the sample has little
corrosion rate in uninhibited erosion-corrosion conditions was only effect, if any, in terms of enhancing the corrosion rate locally. As for

146
E.V. Senatore et al. Wear 404–405 (2018) 143–152

erosion-corrosion experiments in the presence of inhibitor, it cannot be

disputed that the corrosion rate at the centre of the sample is likely to
be enhanced due to the successive sand particle impacts which can
remove the inhibitor film, so the average LPR response for this sample
in Fig. 3 should be interpreted with caution. This is also for the case
with the interpretation of results later in this paper with the use of
FeCO3 pre-filmed samples.
Nonetheless, referring to Fig. 3, it is clear that the inhibitor is ef-
fective in reducing corrosion rate regardless of whether the sand erosive
counterpart is present or not. Application of the corrosion inhibitor
reduces the LPR corrosion rate from approximately
4.9 ± 0.01–0.08 mm/year when no sand is present and from
5.5 ± 0.01–0.2 mm/year when the abrasive particles are added. Thus, Fig. 5. 2D profiles for API 5L X65 erosion-corrosion samples after cleaning.
the damage associated to corrosion of API 5 L X65 steel can be de- Tests are with and without the application of 100 ppm corrosion inhibitor on
creased by 98% when the corrosion inhibitor is injected at 100 ppm initially wet-ground X65 steel samples. Test conditions: 15 m/s, 25 °C, pH 4.7,
with no sand and 96% when sand particles are present. 1000 mg/L sand in a CO2-saturated 3.5 wt%NaCl solution at 1 bar total pres-
This reduction in inhibitor efficiency with regards to the corrosion sure.
component of material loss is likely to be associated with impingement
of sand onto the steel surface causing periodic removal of the inhibitor (C+dCE) makes up ~70% of the total degradation rate. The addition of
film so as to expose the specimen's surface and allow further dissolution inhibitor was able to reduce the damage attributed to corrosion from
before the re-forming of the protective layer. Furthermore, from pre- 70% to 21% at 25 °C while the percentage contribution from the erosion
vious work it is clear that some of the active compounds within the component increased because of the suppression of the active corrosion
inhibitor are able to adhere to the sand particles themselves, reducing process. The application of the inhibitor is able to reduce the total
the available inhibitor concentration in the bulk solution [8]. erosion-corrosion rate by 86%, while the corrosion component is re-
The total mass loss values expressed for the erosion-corrosion tests duced by 96%. The inhibitor also has a role in reducing the erosion
in the presence and absence of inhibitor for initially wet-ground steel component of material loss. Several theories have been identified to
samples are provided in Fig. 4. This total degradation rate can be at- explain this, including the role of inhibitor adsorption onto sand par-
tributed to the summation of four different material loss components ticles [8,16] and surface which influences impact velocity/energy, but
[9] as shown in Eq.(4): also the fact that the lack of corrosion produces a work-hardened layer
at the impingement location which provides more resistance to the
TML = E + C + dEC + dCE (4)
erosion processes at shallow impact angles [17].
where TML is the total mass loss, E is pure erosion in the absence of The surface profile of the two erosion-corrosion samples is provided
corrosion and C is the corrosion in the absence of erosion, dCE is the in Fig. 5. The two profiles show similar penetration depths, with the
effect of erosion on corrosion and dEC is the effect of corrosion on inhibited test exhibiting greater loss at the very centre of the sample
erosion [9]. The combination of dEC and dCE is termed as the sy- compared to the surrounding area. With regards to each sample, the
nergistic effect which is the factor responsible for producing degrada- original height of the entire surface needs to be considered as this will
tion rates greater than the summation of the erosion and corrosion rates be greater for the uninhibited sample as a result of the significant
acting individually. corrosion process. Based on the LPR corrosion rates, the difference in
From the data collected in erosion-corrosion conditions, it is pos- corrosion rates of ~ 5–5.3 mm/year equates to a thickness loss of the
sible to determine the contribution of the corrosion component (C + surrounding area of ~ 4 µm, meaning that the penetration depth of
dCE) through the application of in situ electrochemistry and the erosion each sample are very similar. This is important as it indicates that for
component (E + dEc) using the total mass loss in conjunction with the direct impingement, the addition of inhibitor fails to alleviate the pe-
corrosion component. The contribution of these components is reflected netration depth at the maximum point, despite considerably reducing
in Fig. 4 in the absence and presence of 100 ppm corrosion inhibitor. the overall level of degradation.
Considering the components of material loss in Fig. 4, in blank tests According to Barker et al. [4], a potential increase in penetration
there is a corrosion dominant environment since the corrosion damage depth in the presence of inhibitor could be associated with the role of
the inhibitor on the corrosion process, as a certain degree of plastic
deformation would be necessary to facilitate or encourage material
removal at high impact angles at the centre of the sample. It is expected
that the high corrosion rate in the blank tests could reduce the extent of
plastic deformation on the surface, thus reducing the level of material
removal from erosion. Using inhibitor results in a reduction in corrosion
rate at the centre of the specimen, leaving the material exposed to more
plastic deformation, so more material could potentially be removed at
the sample centre. However, based on these results the depths of pe-
netration are comparable and in this scenario, the effect had not been
observed to any great extent.

3.2. Flow-induced corrosion and erosion-corrosion behaviour of FeCO3 pre-

filmed samples
Fig. 4. Erosion-corrosion degradation rates of wet-ground X65 carbon steel
expressed as erosion and corrosion components with and without the applica- 3.2.1. Development and analysis of FeCO3 layer
tion of 100 ppm corrosion inhibitor. Test conditions: 15 m/s, 25 °C, pH 4.7, Fig. 6 indicates the top view and cross-section of the developed
1000 mg/L sand in a CO2-saturated 3.5 wt%NaCl solution at 1 bar total pressure FeCO3 layer after exposure to the CO2-saturated 1 wt%NaCl solution at
for 6 h. 60 °C and 100 bar for 48 h. The corrosion product layer is homogeneous

147
E.V. Senatore et al. Wear 404–405 (2018) 143–152

Fig. 6. SEM images indicating (a) top view and (b) cross-section view of the developed FeCO3 films on top of an API 5L X65 steel sample after exposure to a CO2-
saturated 1 wt%NaCl brine at 60 °C and 100 bar for 48 h.

in nature, covering the entire steel surface with a relatively uniform that in a blank system without inhibition (from 4.9 to 0.2 mm/year),
thickness of 60 µm. These conditions were chosen to ensure rapid de- the addition of the corrosion inhibitor when the FeCO3 is already pre-
velopment of a thick layer in a short time frame which still possessed sent reduces the corrosion rate further to 0.07 mm/year. This is mar-
the visual characteristic of FeCO3 films which precipitate at lower ginally, yet not significantly lower than the corrosion rate with in-
pressure over extended periods of time [18]. In addition, the generation hibition in the absence of the FeCO3 layer (0.08 mm/year shown in
of the corrosion product film under these conditions is highly re- Fig. 3). However, the results demonstrate the inhibitor is able to pe-
producible. The XRD pattern provided in Fig. 7 confirms that the netrate through and work in conjunction with the pre-formed FeCO3
crystalline corrosion product observed is FeCO3. layer. The protection is provided by the film acting as both a diffusion
barrier for cathodic species and also by covering portions of the steel
3.2.2. Electrochemical and mass loss results of FeCO3 pre-filmed samples surface and blocking the iron dissolution reaction. It is clearly de-
After the pre-formation of FeCO3 on the metal surface, specimens monstrated in these experiments, that in a flow-induced environment,
were evaluated in flow-induced corrosion and erosion-corrosion en- the FeCO3 layer developed in this study expresses a similar level of
vironments. Fig. 8 shows the LPR corrosion rates obtained for each efficiency in terms of suppressing general corrosion compared with
environment, both in the absence and presence of corrosion inhibitor. chemical inhibition.
In the flow-induced corrosion environment (Fig. 8(a)), the sole presence Fig. 9 shows SEM images of the steel surface in each of the flow-
of the FeCO3 layer assists in reducing the corrosion rate by 96% from induced corrosion tests with the pre-filmed sample (both with and

148
E.V. Senatore et al. Wear 404–405 (2018) 143–152

without inhibitor). At the centre of the sample in the absence of in-

hibitor (Fig. 9(a)), signs of FeCO3 removal were evident. It is not clear
whether this was chemical dissolution or as a result of the hydro-
dynamic forces present due to turbulent effects from the fluid flow.
Towards the outer edge of the sample in the absence of inhibitor
(Fig. 9(b)), no significant disruption to the FeCO3 layer was observed.
Addition of 100 ppm inhibitor appeared to protect the FeCO3 from re-
moval, with no noticeable difference being evident between the centre
of the sample under direct impingement (Fig. 9(c) showing a typical
image), and the regions further towards the outer edge of the sample.
This suggests that the inhibitor is able to improve the resistance of the
FeCO3 layer to either hydrodynamic removal, chemical dissolution or
both.
Fig. 8(b) shows the in situ corrosion rate as a function of time for
erosion-corrosion tests on the pre-formed FeCO3 film, both in the pre-
sence and absence of 100 ppm inhibitor. Here, a dramatic difference is
observed in the corrosion component of mass loss for the FeCO3 layer in
the absence of inhibitor compared to the flow-induced corrosion en-
vironment. The pre-formed layer is still able to suppress corrosion,
however, it is much less efficient in the presence of sand impingement,
with the corrosion rate averaging ~ 0.7 mm/year, a reduction of 87%.
This value is still much lower than the corrosion rate in the absence of
FeCO3 (5.5 mm/year), however, not as low as in the same environments
with no sand present (0.2 mm/year). The combination of inhibitor and
Fig. 7. XRD pattern of API 5L X65 sample shown in Fig. 6 confirming the FeCO3 layer generates adequate resistance to the corrosion component
crystalline phase as FeCO3. of degradation in the erosion-corrosion environment, reaching an in situ
corrosion rate of 0.1 mm/year, which is lower than the corrosion rate of
the wet-ground sample subjected to erosion-corrosion conditions with
inhibitor (0.2 mm/year).
In the erosion-corrosion environment, the effect of sand particle
impingement can produce significant levels of erosion in conjunction
with corrosion. Fig. 10 shows the erosion-corrosion resistance of the
FeCO3 layered surface in the presence and absence of inhibitor. As
before, these have been divided into erosion and corrosion components
based on consideration of total mass loss and the LPR response. The sole
FeCO3 layer is able to considerably suppress the corrosion component
of degradation in the erosion-corrosion environment but the erosion
component is relatively high and it results in a total erosion-corrosion
rate similar to tests in the absence of inhibitor and FeCO3 layer (Fig. 4).
In presence of 100 ppm inhibitor, the tendency remains the same.
Only the corrosion component is reduced when it is compared with
erosion-corrosion rate in blank tests in the absence of the FeCO3 layer
(Fig. 4). However, it is important to note that the erosion component of
damage here includes removal of the FeCO3 layer (either by chemical or
mechanical effects) in addition to damage to the substrate.
Considering the results in Fig. 10, the inhibitor in conjoint with
FeCO3 layer plays a role in reducing not only the corrosion component
of material loss, but also the erosion component from the blank system
with the pre-filmed sample.
The surface morphologies of filmed samples after erosion-corrosion
tests were observed by scanning electron microscopy (SEM). Fig. 11
shows the centre (Fig. 11(a)), transition region (11(b)) and edge (11(c))
of the sample subjected to erosion-corrosion in the absence of inhibitor,
while Fig. 11(d) to (f) depict the surface of the erosion-corrosion sam-
ples in the presence of inhibitor. Fig. 11(a) and (d) indicate full removal
of the film at the centre of impingement. However, the transition region
between the centre and outside of the sample indicates that the in-
hibitor is able to provide added protection to the FeCO3 layer as there
Fig. 8. LPR corrosion rates of API 5L X65 as a function of time for FeCO3 pre-
are less regions of local removal of the film.
filmed samples exposed to (a) flow-induced corrosion and (b) erosion-corrosion
conditions; experiments are in the presence and absence of 100 ppm corrosion The SEM images indicate that the degradation of the sample is not
inhibitor (CI) and 1000 mg/L sand; test conditions are 15 m/s, 25 °C and pH uniform across the entire steel surface. Consequently, the assumed
4.7in a 3.5 wt%NaCl solution CO2-saturated at 1 bar total pressure. uniform degradation rates discussed previously could be somewhat
misleading in terms of identifying the extent of damage caused to the
steel surface. What effectively needs to be considered is the penetration
depth and geometry of the wear scar generated from the impingement
process. To address this issue, profilometry measurements were

149
E.V. Senatore et al. Wear 404–405 (2018) 143–152

Fig. 9. SEM images of (a) centre of sample under direct impingement in flow-induced corrosion environment with 0 ppm inhibitor, (b) outer region of sample in flow-
induced corrosion conditions with 0 ppm inhibitor and (c) typical image of centre/outside of sample in flow-induced corrosion environment with 100 ppm inhibitor
(no noticeable difference between the two regions); test conditions are 15 m/s, 25 °C and pH 4.7in a CO2-saturated 3.5 wt% NaCl solution at 1 bar total pressure.

performed on the steel surface after removal of the corrosion product.

The profiles are provided in Fig. 12 and illustrate that the FeCO3 layer is
able to suppress the penetration depth from ~ 20 µm to less than ~
10 µm. This indicates that the FeCO3 layer does provide some initial
protection. However, the 60 µm layer was completely removed by di-
rect impingement in the centre of the same after 6 h of exposure, lim-
iting its effectiveness.

4. Conclusions

This work focused on the evaluation of FeCO3 corrosion product to

provide protection to carbon steel substrates in CO2-containing flow-
induced corrosion and erosion-corrosion environments, both in the
Fig. 10. Erosion-corrosion degradation rates of X65 carbon steel samples pre-
absence and presence of a commercial corrosion inhibitor. As a result of
filmed with FeCO3 expressed as erosion and corrosion components with and
this work, the main findings were:
without the application of 100 ppm corrosion inhibitor. Test Conditions: 15 m/
s, 25 °C, pH 4.7, 1000 mg/L sand in a CO2-saturated 3.5 wt%NaCl solution at
1 bar total pressure for 6 h. • The sole presence of the FeCO 3 layer on carbon (i.e. in the absence of
chemical inhibition) can assist in the reduction of the corrosion
component of degradation in both the flow-induced and erosion-
corrosion environments considered in this work.
• Although the erosion component recorded under uninhibited

150
E.V. Senatore et al. Wear 404–405 (2018) 143–152

Fig. 11. SEM images of (a) centre (b) transition region, (c) edge of the sample in erosion-corrosion environment with 0 ppm inhibitor and (d) centre (e) transition
region and (f) edge of the sample in erosion-corrosion environment with 100 ppm inhibitor; test conditions are 15 m/s, 25 °C and pH 4.7 in a CO2-saturated 3.5 wt%
NaCl solution at 1 bar total pressure.

erosion-corrosion conditions was relatively high for pre-filmed the sample from ~ 20 to ~ 10 µm based on 6 h experiments in the
carbon steel samples (as it comprises of the removal of the FeCO3 absence of inhibitor.
layer as well as damage to the steel substrate), profilometry analysis • Full removal of FeCO3 was observed at the centre of the carbon steel
revealed that the layer does afford some erosion protection to the sample in erosion-corrosion conditions with no inhibitor within 6 h,
steel substrate, reducing the total wear scar depth at the centre of indicating that although protection was afforded, it was not

151
E.V. Senatore et al. Wear 404–405 (2018) 143–152

of CO2 in the mechanism of carbon steel corrosion, Corros. Sci. 120 (2017)
239–250.
[2] H. Marchebois, J. Leyer, B. Orlans-Joliet, Ssc Performance of A Super 13% Cr
Martensitic Stainless Steel for Octg: Three-Dimensional Fitness-for-Purpose
Mapping according to PH2S, Ph and Chloride Content, NACE Corros (Paper ID:
NACE-07090), 2007.
[3] K.D. Efird, E.J. Wright, J.A. Boros, T.G. Hailey, Correlation of steel corrosion in pipe
flow with jet impingement and rotating cylinder tests, Corrosion 49 (12) (1993)
992–1003.
[4] R. Barker, A. Neville, X. Hu, S. Cushnaghan, Evaluating inhibitor performance in
CO2-saturated erosion-corrosion environments, Corrosion 71 (1) (2015) 14–29.
[5] X. Hu, R. Barker, A. Neville, A. Gnanavelu, Case study on erosion-corrosion de-
gradation of pipework located on an offshore oil and gas facility, Wear 271 (2011)
1295–1301.
[6] A. Akbar, X. Hu, C. Wang, A. Neville, The Influence of Flow Rate, Sand and Inhibitor
on Iron Carbonate Scales under Erosion-corrosion Conditions using a Submerged
Impingement Jet, NACE Corros (PAPER ID: NACE-2012–1396), 2012.
[7] S. Hassani, K.P. Roberts, S. Shirazi, J.R. Shadley, E.F. Rybicki, C.J.B. Joia, A. New,
Fig. 12. 2D profiles for API 5L X65 erosion-corrosion samples after cleaning. Approach for Predicting Inhibited Erosion-Corrosion in CO2- saturated oil/brine
Tests are with and without the application of 100 ppm corrosion inhibitor on Flow Condition, SPE (PAPER ID: SPE-1551136-MS), 2012.
initially FeCO3 pre-filmed X65 steel samples. Test Conditions: 15 m/s, 25 °C, pH [8] S. Ramachandran, Y.S. Ahn, V. Jovancicevic, J. Basset, Further Advances in the
Development of Erosion Corrosion Inhibitors, NACE Corros (PAPER ID: NACE-
4.7, 1000 mg/L sand in a CO2-saturated 3.5 wt%NaCl solution at 1 bar total 05292), 2005.
pressure. [9] A. Neville, C. Wang, Erosion-corrosion mitigation by corrosion inhibitors – an as-
sessment of mechanisms, Wear 267 (2009) 195–203.
[10] R. Barker, X. Hu, A. Neville, S. Cushnaghan, Inhibition of flow-induced corrosion
sustained throughout the entire test. and erosion-corrosion for carbon steel pipe work from an offshore oil and gas fa-
• The application of a commercial corrosion inhibitor was shown to
[11]
cility, Corrosion 69 (2) (2013) 193–203.
A. Dugstad, Mechanism of Protective Film Formation During CO2 Corrosion of
work synergistically with the FeCO3 layer to reduce the corrosion
Carbon Steel, NACE Corros (PAPER ID: NACE-98031), 1998.
component of degradation in flow-induced corrosion and erosion- [12] A. Palencsár, E. Gulbrandsen, K. Kosorú, Corrosion Inhibition under FeCO3-forming
corrosion environments. Conditions at Elevated Temperatures, SPE (PAPER ID: SPE-169614-MS), 2014.

• The use of the inhibitor improved the resilience of the FeCO3 layer [13] S.M. Yaakob, M.C. Ismail, Corrosion inhibitor performance with presence of FeCO3
film in CO2 corrosion environment under fluid flow effect, Adv. Mat. Res. 789
to either chemical or hydrodynamic removal (or both) in flow-in- (2013) 507–510.
duced corrosion environments. [14] O.O. Ige, L.E. Umoru, Effects of shear stress on the erosion‐corrosion behaviour of
X-65 carbon steel: a combined mass-loss and profilometry study, Tribol. Int. 94
(2016) 155–164.
Acknowledgements [15] ASTM, STANDARD G1−03, Standard Practice for Preparing, Cleaning, and
Evaluating Corrosion Test Specimens, ASTM International, 2003.
The authors would like to acknowledge the support provided by [16] A.J. McMahon, J.W. Martin, L. Harris, Effects of Sand and Interfacial Adsorption
Loss on Corrosion Inhibitor Efficiency, NACE Corros (PAPER ID: NACE-05274),
CNPq, ANP and Shell. 2005.
[17] I. Finnie, Some observations on the erosion of ductile metals, Wear 19 (1972)
References 81–90.
[18] F. Pessu, R. Barker, A. Neville, The influence of pH on localized corrosion behavior
of X65 carbon steel in CO2-saturated brines, Corrosion 71 (12) (2015) 1452–1466.
[1] T.C. Almeida, M.C.E. Bandeira, R.M. Moreira, O.R. Mattos, New insights on the role

152