CORROSION SCIENCE SECTION

Investigation of Carbon Dioxide Corrosion

of Mild Steel in the Presence of Acetic Acid—
Part 1: Basic Mechanisms

K.S. George* and S. Nes̆ić‡,*

ABSTRACT by organic acids is well recognized today, the analysis

of oilfield brines for the presence of organic acids is
The corrosion behavior of mild steel in the presence of acetic still not routinely done.
acid (CH3COOH) and carbon dioxide (CO2) has been investi- The following text will focus on the recent findings
gated using electrochemical techniques and weight-loss mea- related to carbon dioxide (CO2) corrosion of mild steel
surements. Electrochemical measurements have shown that in the presence of acetic acid ([CH3COOH] denoted as
the presence of acetic acid affects predominantly the cathodic
HAc in the text below), which is the most prevalent
reaction. The acetic acid effect is much more pronounced at
low molecular weight organic acid found in brines.
elevated temperatures when catastrophic corrosion rates may
be encountered at high concentrations. The undissociated However, a brief introduction to CO2 corrosion will be
form of acetic acid, present at lower pH, is responsible for the given first.
increases seen in the corrosion rate.
Carbon Dioxide Corrosion
KEY WORDS: acetic acid, carbon dioxide corrosion, electro- The corrosion mechanisms of CO2 and its effects
chemical techniques, mild steel, weight-loss measurements
on mild steel under varying conditions of pressure,
temperature, pH, and oil-water fractioning has been
INTRODUCTION a widely researched topic. Some of the key studies in-
clude work by de Waard and coworkers,2-5 Dugstad,
The presence of organic acids in oil and gas produc- et al.,6 and Nes̆ić and coworkers.7-10 Furthermore,
tion and transport lines was first discovered in 1944 these authors have proposed models to predict CO2
(for example, Crolet, et al.1). Classification of organic corrosion of mild steel based on the results of their
acids can be done on the basis of molecular weight, work. The following is a summary of the key processes
and it was found that the lower molecular weight or- occurring in CO2 corrosion.
ganic acids were primarily soluble in water and can Carbon dioxide gas dissolves in water and forms a
lead to corrosion of mild steel pipelines, as discussed “weak” carbonic acid through hydration by water:
below. Higher molecular weight organic acids are
typically soluble only in the oil phase and pose a cor-
CO2 ( g ) ⇔ CO2 (1)
rosion threat only at higher temperatures in the refin-
eries. Even if the threat to mild steel pipelines posed
CO2 + H2O ⇔ H2CO3 (2)
Submitted for publication October 2005; in revised form, July
2006. The carbonic acid (H2CO3) then partially dissociates to
‡
Corresponding author. E-mail: [email protected].
* Institute for Corrosion and Multiphase Flow Technology, 342 West form the bicarbonate ion, which can further dissociate
State St., Ohio University, Athens, OH 45701. to yield the carbonate ion:

0010-9312/07/000023/$5.00+$0.50/0
178 © 2007, NACE International CORROSION—FEBRUARY 2007
 CORROSION SCIENCE SECTION

H2CO3 ⇔ H+ + HCO3– (3) The effect of flow on CO2 corrosion, when no pro-
tective films are present, is through increased mass
HCO 3– ⇔ H+ + CO 23 – (4) transport of the corrosion species toward and away
from the metal surface.8 When the mass transport
It is widely known that solutions containing H2CO3 are rate of the species (e.g., hydrogen ion) is not high
more corrosive to mild steel than solutions of strong enough to support the electrochemical process at the
acids, such as hydrochloric (HCl) or sulfuric (H2SO4), metal surface, limiting reaction rates are reached. On
at the same pH. This was a topic of debate and specu- the other hand, species accumulation, supersatura-
lation in the past few decades. de Waard and Milliams2 tion, and film precipitation can occur at the metal
suggested that this is due to the reduction of the un- surface if the transport of the corrosion products
dissociated H2CO3 molecule, which occurs after it is (e.g., ferrous ions) away from the surface is not rapid
adsorbed onto the metal surface. According to them, enough. This is another mass-transfer effect of flow
this is the dominant and rate-determining step in the on CO2 corrosion. However, flow may also affect the
CO2 corrosion process, so the corrosion rate of the formation and survival of corrosion product films by
mild steel surface is directly related to the concentra- mechanical means via hydrodynamic stresses.
tion of the undissociated H2CO3 in solution and to the
CO2 partial pressure. Acetic Acid Corrosion
However, there are two possible cathodic reac- When HAc is present in the system it partitions
tions that can occur in the process of mild steel CO2 between the aqueous and the gas phases. The aque-
corrosion: the above-mentioned “direct” reduction of ous HAc then partly dissociates into hydrogen and
H2CO3 but also reduction of hydrogen ions: acetate ions:

H2CO3 + e – ⇒ H + HCO3– (5) HAc( g ) ⇔ HAc (8)

H+ + e – ⇒ H (6) HAc ⇔ H+ + Ac – (9)

While the rate of the former process is determined by Iron acetate salt can form in aqueous solutions; how-
the amount of CO2 in the system, the rate of the lat- ever, its solubility is much higher than that of iron
ter process is strongly pH-dependent. The electrons carbonate, and therefore precipitation and protective
required to keep the process going are provided by a film formation by iron acetate does not readily occur.
single anodic reaction, iron dissolution: In 1999, Crolet, et al.,1 described how low concen-
trations of HAc (6 ppm to 60 ppm) affect the corrosion
Fe ⇒ Fe 2+ + 2e – (7) rates of carbon steel. They argued that the increase
in the rate of corrosion in the presence of HAc occurs
Whether or not the direct reduction of H2CO3 (Re- due to an “inversion” in the bicarbonate/acetate ratio.
action [5]) actually occurs on the metal surface was At this inversion point, HAc is the predominant acid
and still is a topic of debate, since it could be argued compared to H2CO3 and is therefore the main source
that undissociated H2CO3 is merely a source of hy- of acidity. The work of Crolet, et al.,1 also suggested
drogen ions via Reaction (3) and would dissociate to that the presence of HAc inhibited the anodic (iron
give a hydrogen ion faster than it (carbonic acid) could dissolution) reaction.
diffuse to the surface of the steel. In this way H2CO3 Hedges and McVeigh published results on the
would act as an additional source of hydrogen ions role of acetate in CO2 corrosion.11 Experiments using
and lead to higher corrosion rates. Both pathways both HAc and sodium acetate as a source of acetate
(Reactions [5] and [6]) release hydrogen gas from ions in various media (3% NaCl and two synthetic oil-
water as a product and, nowadays, it is accepted that field brines) were performed using rotating cylinder
direct reduction of H2CO3 (Reaction [5]) dominates at electrodes. Both sources of acetate ions were shown to
high partial pressures of CO2 and high pH, while the increase the corrosion rate, but HAc decreased the pH
reduction of hydrogen ions dominates at low CO2 par- while sodium acetate increased it. The increased cor-
tial pressures and low pH. rosion rates were attributed to the formation of thinner
When steel corrodes in CO2-saturated water, iron carbonate films, since acetate ions have the ability
the solubility of iron carbonate salt (FeCO3) may be to form iron acetate and transport iron away from the
exceeded and precipitation sets in, which increases steel surface. However, no attempt was made to quan-
rapidly with the degree of supersaturation and an in- tify the thickness or morphology of the films formed in
crease in temperature. The iron carbonate precipitate their experiments. Little or no control of pH in their
may form a protective film depending on the solution experiments may have led to the ambiguous effect of
composition, pressure, and temperature of the sys- acetic species on corrosion. They also reported a mild
tem. Other solid corrosion products may form in the increase in the cathodic reaction rate in the presence
presence of chlorides, sulfides, oxygen, etc. of HAc, but their results were not fully conclusive.

CORROSION—Vol. 63, No. 2 179

CORROSION SCIENCE SECTION

TABLE 1 4. What is the best way to integrate the findings

Key Experimental Parameters about the HAc effect on CO2 corrosion into a
Steel Type X-65 corrosion prediction model?
The first point is the topic of the present paper.
Aqueous solution Oxygen-free, 3 wt% NaCl
Purged gas CO2, N2
Effects of HAc at high pressures and temperatures,
Total pressure Atmospheric protective film formation and modeling are subjects of
HAc concentration 0 to 5,000 ppm separate papers.
Temperature 22, 40, 60, 80°C
Rotation rate 100 rpm to 4,000 rpm
pH 4, 5, 6
EXPERIMENTAL PROCEDURES
Measurement techniques Potentiodynamic sweeps (PS),
linear polarization resistance (LPR), The test matrix described in Table 1 was executed
electrochemical impedance in order to seek answers to the first question posed
spectroscopy (EIS), above.
weight loss (WL) All of the permutations of the parameters in the
experimental matrix were not performed. When one
parameter was varied, the other parameters were set
Some limited evidence on the role of HAc in CO2 to their baseline values, which were selected to be:
corrosion comes from field experience as related to the 1 bar CO2, 22°C, 100 ppm total aqueous acetic spe-
so-called top-of-line corrosion,12 where HAc was said cies, pH 4, and 1,000 rpm. For example, in the HAc
to have a detrimental effect. concentration study, the HAc concentration was
Garsany, et al.,13 used voltammetry to study the varied from 0 to 5,000 ppm, but the temperature
effect of acetate ions on the rates and mechanisms remained constant at 22°C. The system pH was main-
of corrosion with a rotating disc electrode (RDE) and tained at 4 and the rotational velocity was kept at
film-free surfaces. Their voltammograms show two 1,000 rpm. In this way effects of various parameters
waves, which are attributed to the reduction of the were deconvoluted. This is particularly important for
hydrogen ion and “direct” HAc reduction on the steel parameters that are closely linked, such as HAc con-
surface: centration and pH.
It should be noted that in this part of the study
HAc + e – ⇒ H + Ac – (10) the key parameters (listed in Table 1), such as pH and
temperatures, have been chosen to yield conditions
They argue that, since HAc dissociation can oc- where corrosion product film formation is unlikely or
cur very quickly, it is not possible to distinguish the very slow. This enabled a more accurate study of the
reduction of hydrogen ions (Reaction [6]) from direct underlying electrochemical processes, mass-transfer,
HAc reduction (Reaction [10]) at the electrode surface. and water chemistry effects. The weight-loss experi-
Joosten, et al.,14 performed experiments using ments, for similar reasons, were performed over short
HAc, synthetic seawater, and an oil phase in glass periods of time (typically 24 h) to avoid any significant
cells. They found that HAc increased the corrosion film formation.
rate by decreasing the pH, but the system could be
inhibited very effectively (below 400 ppm HAc). Effec- EXPERIMENTAL SETUP
tive inhibition has also been reported by Hedges and
McVeigh.11 A schematic of the experimental cell is shown in
Generally speaking, very few systematic studies Figure 1. To begin, the experimental apparatus was
have been performed in the laboratory. Consequently, assembled. A salt solution was prepared, added to
little or no information exists regarding the basic ef- the cell, and then deoxygenated for at least 1 h us-
fect of HAc on the anodic and cathodic reactions as ing nitrogen or CO2 gas. The test temperature was
outlined above. set using a hot plate and controlled using a feedback
Based on previous research, some of the principal temperature probe. Once deoxygenation had occurred
questions still need to be answered: and the test temperature was reached, the appropri-
1. What are the main effects of HAc on the CO2 ate amount of HAc was then added to the cell and de-
corrosion process? Does HAc interfere with the oxygenation was continued for an additional 30 min.
anodic and cathodic reactions otherwise pres- Since HAc is somewhat volatile, bubbling pure CO2
ent in CO2 corrosion or does it lead to new ad- gas through the solution would strip the HAc out of
ditional electrochemical reactions at the steel the test cell. A gas “preconditioning” cell was used as
surface? How big is the effect? a remedy. The preconditioning cell was kept constant
2. Does anything change at high pressures and at the test temperature and contained the same fluid
temperatures? composition as the experimental cell. The precondi-
3. Is there an effect of HAc on the formation of tioning cell ensured the CO2 entering the experimental
protective iron carbonate films? cell was saturated with HAc and H2O vapor and that

180 CORROSION—FEBRUARY 2007

 CORROSION SCIENCE SECTION

the HAc stripping occurred much slower and could

therefore be ignored.
The pH meter used in the experiments was cali-
brated at the test temperature by heating the buffer
solutions. The pH was monitored and adjusted before
and after the HAc addition to ensure the fluid compo-
sition was similar between test runs. To achieve the
desired system pH, minute adjustments were made
using droplets of hydrochloric acid or sodium bicar-
bonate (NaHCO3) solutions.
The working electrode had a 120-cm diameter
and an area of 5.4 cm2. The composition of the X-65
mild steel used in the experiments is shown in Table
2. The cylindrical electrode was polished using
600-grit sand paper, copiously washed with ethanol
(C2H6O), dried, immersed into the test solution, and
the electrode’s rotational velocity was set. After ap-
proximately 30 additional minutes, electrical connec-
tions were made and electrochemical measurements
started.
A concentric platinum wire ring was used as a
counter electrode and a silver/silver chloride (Ag/
AgCl; 3 M potassium chloride [KCl]) external electrode
connected with the cell via a salt bridge and a Luggin
capillary served as a reference electrode. The electro-
chemical measurements were typically conducted in
the same order:
—Corrosion rate was measured using linear po-
larization resistance (LPR).
FIGURE 1. Schematic of the test cell.
—Solution resistance was found using electro-
chemical impedance spectroscopy (EIS).
—Cathodic and then the anodic potentiodynamic
sweeps were performed. leads to an increase in the bulk pH. After the anodic
sweeps, the pH was then adjusted using HCl before
Electrochemical Measurements the beginning of the next sweep. Anodic sweeps were
All electrochemical measurements were made limited to polarization less than 200 mV above the
using a potentiostat. The potentiodynamic polariza- corrosion potential to limit buildup of excessive fer-
tion curves (sweeps) were conducted at a scan rate of rous concentrations in the test cell.
0.2 mV/s, and the potential was manually corrected The potentiodynamic sweeps were always con-
for solution resistance, which was measured using ducted starting from the corrosion potential. For
EIS. Potentiodynamic sweeps were conducted at con- example, a cathodic potentiodynamic sweep would
stant pH with the pH adjustment occurring after each scan from the corrosion potential to approximately
sweep. During the anodic sweep, for example, the –650 mV below the corrosion potential. The corrosion
system pH was set to 4 and the sweep started. Due potential would then be allowed to drift back to the
to iron dissolution, the pH of the system would rise starting corrosion potential before an anodic potentio-
to 4.08 by the end of the sweep. As ferrous ions are dynamic sweep was performed. The LPR measure-
rapidly generated on the working electrode, hydrogen ments were taken at ±5 mV around the corrosion
ions and water are reduced on the counter electrode, potential, by using a potentiodynamic scan at
releasing the hydroxide ions into the solution. This 0.1 mV/s.

TABLE 2
Chemical Composition of API 5LX65 Steel (wt%)
Al Cr Mo S V B Cu Nb Si
0.032 0.011 0.103 0.004 0.055 0.0002 0.010 0.030 0.240

C Fe Ni Sn Ca Mn P Ti
0.150 Balance 0.020 0.005 0.0032 1.340 0.011 0.001

CORROSION—Vol. 63, No. 2 181

CORROSION SCIENCE SECTION

to remove any corrosion product, and then weighed

after drying.

RESULTS AND DISCUSSION

Water Chemistry Calculations

The water chemistry of the experimental solutions
was found by solving the equilibrium expressions for
the reactions given above (1, 2, 3, 4, 8, and 9). Ex-
pressions for the equilibrium constants are the same
as those used in the mechanistic model of Nes̆ić, et
al.10 The concentrations of some of the key species at
FIGURE 2. The effect of pH on the concentration of species at 1 bar 1 bar CO2, 22°C, when 10 ppm HAc (total) is added,
CO2, 22°C, when 10 ppm HAc was added.
are shown in Figure 2. The concentration of dissolved
CO2 and H2CO3 is fixed with the partial pressure of
the purged gas and is not a function of pH. As most
of the experimental work was performed at pH 4, it is
evident that when 10 ppm HAc is added to the solu-
tion, HAc is the main source of acidity up to a pH of
approximately 4.7. When 100 ppm HAc is added, un-
der the same conditions, it is the main source of acid-
ity up to a pH of almost 6. This is shown in Figure 3.
In the results presented below, the concentra-
tion of HAc will be reported as both the total amount
of HAc added to the system as well as the amount of
aqueous “free” HAc that stayed in undissociated form.
For example, when 100 ppm HAc was added to the
FIGURE 3. The effect of pH on the concentration of species at 1 bar system and the pH adjusted to 4, the free concentra-
CO2, 22°C, when 100 ppm HAc was added. tion of HAc was 85 ppm, but when experiments were
performed at pH 6 the same addition of 100 ppm of
HAc to the system resulted in a free HAc concentra-
tion of only 6 ppm. The undissociated concentration
of HAc as a function of pH is shown in Figure 4.

Effect of HAc Concentration

In the first series of experiments, potentiody-
namic sweeps were performed in 3 wt% NaCl solu-
tions purged with nitrogen for an undissociated HAc
concentration range from 0 to 850 ppm (obtained by
adding 0 to 1,000 ppm of HAc at pH 4). Nitrogen is in-
ert and was used to remove any interference from CO2
so that the effect of HAc could be seen more clearly.
Normally, addition of HAc reduces the pH; however,
the pH was held constant in these experiments (by
FIGURE 4. The concentration of undissociated aqueous HAc as a
adding HCl) to distinguish the effect of the acetic spe-
function of system pH at 22°C for various amounts of HAc added.
cies from the effect of hydrogen ion concentration on
the cathodic and anodic reactions. At a constant pH,
the concentration of hydrogen ions is fixed, and the
Weight-Loss Experiments effect of the acetic species on the cathodic and anodic
In a separate series of weight-loss experiments, reactions could be distinguished. The HAc concentra-
after the solution had come to the desired tempera- tion range was selected to represent situations typi-
ture and the pH was adjusted, a preweighed steel cally encountered in service.
sample was immersed into the solution. During the From Figure 5 it can be seen that at room tem-
24-h weight-loss experiments, the pH was adjusted perature there is a clear increase of the cathodic limit-
approximately every hour or two, which corresponded ing current density with increased concentrations of
with the procedure used in the LPR measurements. HAc. When the concentration of HAc increased from
After 24 h, the electrodes were taken out of the test 0 to 850 ppm, the limiting current density increased
solution, rinsed with alcohol, wiped with a soft cloth almost 30 times. In the absence of HAc, the limiting

182 CORROSION—FEBRUARY 2007

 CORROSION SCIENCE SECTION

FIGURE 6. The effect of HAc concentration on the corrosion rate of

FIGURE 5. Potentiodynamic polarization curves done on X65 steel X65 steel in aqueous solutions purged with N2 at pH 4, 1,000 rpm,
corroding in aqueous solutions purged with N2 at pH 4, 1,000 rpm, 22°C, as measured by the LPR and Tafel extrapolation techniques
22°C, containing various amounts of undissociated aqueous HAc (Figure 5). The error bars represent minimum and maximum values
(0, 100 ppm, and 1,000 ppm total acetic species, respectively). The obtained in repeated experiments.
dashed lines are estimated Tafel lines, which are extrapolated to
the corrosion potential to determine the corrosion rate. The shaded
areas represent a range of values of the corrosion current and
corrosion potential as determined by LPR (Figure 6).

current arises from diffusion of hydrogen ions. The

increase of the limiting current in the presence of HAc
was related to the diffusion of undissociated HAc as
discussed below. It may appear that under these con-
ditions the anodic reaction was retarded in the pres-
ence of HAc, as reported by Crolet, et al. However, this
conclusion must be taken with caution as the changes
in the anodic portion of the curves were overshadowed
by the large changes in the cathodic reaction, which
resulted in a large change of the corrosion potential. FIGURE 7. Potentiodynamic polarization curves done on X65 steel
corroding in aqueous solutions purged with CO2 at pH 4, 1,000 rpm,
The corrosion rate increased only by a factor of 2 to
22°C, containing various amounts of undissociated aqueous HAc
3 under the same conditions, as shown in Figure 6.
(0, 100 ppm, and 1,000 ppm total acetic species, respectively). The
Comparison of the corrosion rates measured by LPR dashed lines are estimated Tafel lines, which are extrapolated to
and estimated by Tafel extrapolation shown in Figure the corrosion potential to determine the corrosion rate. The shaded
6 suggests that the values obtained by the two tech- areas represent the range of values of the corrosion current and
niques are close, considering the inherent errors with corrosion potential as determined by LPR (Figure 8).
using these techniques. By assuming charge-transfer
control, a cathodic Tafel slope of 120 mV/decade and
an anodic Tafel slope of 40 mV/decade (determined
from Figure 5) were used in both methods. The values
of the estimated corrosion current by using LPR are
overlaid on the potentiodynamic curves in Figure 5. It
is clear that corrosion currents were in all cases much
lower than the cathodic limiting currents (by an order
of magnitude or more), confirming that the corrosion
process is charge-transfer-controlled under these con-
ditions.
In a subsequent series of experiments, the aque-
ous solution was purged with CO2 and the HAc
concentration was varied from 0 to 850 ppm of undis-
sociated HAc (corresponding to 1,000 ppm total acetic
species at pH 4). The resulting cathodic and anodic FIGURE 8. The effect of HAc concentration on the corrosion rate of
sweeps are shown in Figure 7. The effect of HAc con- X65 steel in aqueous solutions purged with CO2 at pH 4, 1,000 rpm,
centration on the corrosion rate is shown in Figure 22°C, as measured by the LPR and Tafel extrapolation techniques
(Figure 7). The error bars represent minimum and maximum values
8. As in the case described above without CO2, the
obtained in repeated experiments.
limiting current of the cathodic reaction in the pres-

CORROSION—Vol. 63, No. 2 183

CORROSION SCIENCE SECTION

Effect of pH
The effect of pH on the potentiodynamic polariza-
tion curves was studied in solutions saturated with
CO2 with the addition of 100 ppm HAc adjusted in the
pH range from 4 to 6. The results are shown in Figure
9. As the pH is increased from 4 to 5, the anodic reac-
tion rate is increased, consistent with Bockris’s iron
dissolution mechanism.15 However, a further increase
in pH from 5 to 6 did not result in a further increase
in the anodic reaction rate. On the other hand, the
cathodic reaction and particularly the limiting cur-
rent was retarded by an increase in the pH. This is
FIGURE 9. The effect of pH on potentiodynamic sweeps done on expected, since with the increase of the pH the con-
X65 steel corroding in aqueous solutions purged with CO2 containing
centration of the hydrogen ion decreases, but also
100 ppm total acetic species at 1,000 rpm, 22°C.
because the concentration of undissociated HAc de-
creases with pH increase (Figure 3).

Effect of Rotational Velocity

The effect of velocity was studied by using poten-
tiodynamic sweeps performed in 3% NaCl solution,
saturated with CO2 with the addition of 100 ppm HAc
(85 ppm undissociated HAc at pH 4), 22°C. The re-
sults are shown in Figure 10. The increase in the lim-
iting current is measured to be approximately a factor
of 1.8 to 2 for every doubling of velocity, which is in
excellent agreement with the mass-transfer theory.
One can conclude that in HAc solution the reduc-
tion process is limited by the mass transfer of acetic
FIGURE 10. The effect of rotational velocity on potentiodynamic species (probably undissociated HAc). In contrast,
polarization curves done on X65 steel corroding in aqueous solutions the limiting currents seen in pure CO2 corrosion are
purged with CO2 containing 100 ppm total acetic species (85 ppm chemical reaction-controlled (arising from a slow CO2
undissociated HAc at pH 4), 22°C. hydration step, Reaction [2]) and exhibit a weak flow
dependence.16-17 The corrosion potential, as well as
the anodic reaction, do not change with velocity, sug-
ence of CO2 was greatly increased with increasing gesting pure charge-transfer corrosion process under
concentrations of HAc. On the other hand, the corro- these conditions and a constant corrosion rate unper-
sion current/rate approximately doubled when HAc turbed by velocity.
was present. In pure CO2 corrosion (0 ppm HAc) the
corrosion current was of a similar order of magnitude Effect of Temperature
as the limiting current, suggesting mixed charge/ The effect of temperature on the potentiodynamic
chemical control. However, in the presence of HAc sweeps was studied in 3% NaCl solutions saturated
the corrosion current was at least an order of mag- with CO2 with the addition of 100 ppm HAc (85 ppm
nitude smaller than the limiting current, suggesting undissociated HAc at pH 4), 1,000 rpm. The results
pure charge-transfer control. The anodic Tafel slope are shown in Figure 11. There is a clear acceleration
was found to be 60 mV/decade to 80 mV/decade of both the cathodic and anodic reaction rates with an
when CO2 was present, suggesting a slight change in increase in temperature, as is expected. The corrosion
mechanism for iron dissolution. The cathodic Tafel rate, which is under charge-transfer control at room
slope remained 120 mV/decade. The presence of HAc temperature (22°C), becomes mass-transfer limiting
did not change the situation. Using these Tafel slopes current controlled at higher temperatures. The corro-
the corrosion rates calculated by LPR and the Tafel sion rates measured by LPR and calculated from Tafel
extrapolation agreed within the error of measurement. slopes were in good agreement.
When comparing the case without CO2 (Figures 5
and 6) to the case that includes CO2 (Figures 7 and 8), Repeatability of the Potentiodynamic Sweeps
it is seen that both the limiting current and the cor- To put the results presented above into a proper
rosion current/rate are larger for the latter case with perspective, a single baseline experiment was per-
CO2. Clearly, the presence of CO2 introduces an ad- formed multiple times to estimate the repeatability.
ditional corrosive species: undissociated carbonic acid The repeatability was rather good, and the changes
(Reaction [5]). seen on the plots presented above are genuine and do

184 CORROSION—FEBRUARY 2007

 CORROSION SCIENCE SECTION

FIGURE 11. The effect of temperature on potentiodynamic polarization

curves done on X65 steel corroding in aqueous solutions purged with FIGURE 12. The effect of temperature and HAc concentration on
CO2 containing 100 ppm total acetic species (approximately 85 ppm the corrosion rate of X65 steel in aqueous solutions purged with CO2
undissociated HAc at pH 4), 1,000 rpm. The shaded areas represent at pH 4, 1,000 rpm, as measured by the LPR and weight-loss (WL)
the range of values of the corrosion current and corrosion potential techniques. The error bars represent minimum and maximum values
as determined by LPR. obtained in repeated experiments and the figure above the bars is
the number of repeats.

reflect influences of various parameters beyond the

experimental error.

Verification—Weight-Loss Experiments
A series of weight-loss experiments was con-
ducted to verify the effect of HAc on the CO2 corrosion
of carbon steel in 3% NaCl solutions. The results are
shown in Figure 12. The average value of the corro-
sion rate obtained by the weight-loss method is pre-
sented in the form of solid bars, while the error bars
represent the maximum and minimum experimental
values obtained in “repeated” experiments. The num-
ber above the error bars represents the number of
repeated experiments used to calculate the average
FIGURE 13. The effect of HAc concentration on the corrosion rate
value. The values presented for the LPR method are
of X65 steel in aqueous solutions purged with CO2 at 60°C, pH 4,
time-averaged over the course of the experiment. In 1,000 rpm, as measured by the LPR and weight-loss (WL) techniques.
most experiments weight loss and LPR were done on The error bars represent minimum and maximum values obtained in
the same specimen; however, in a few experiments repeated experiments and the figure above the bars is the number
no electrochemical experimentation was done to see of repeats.
if, by measuring the corrosion rate using LPR, the
system was disturbed enough to change the corrosion
rate measured by weight loss. No interference was concentration doubles the corrosion rate. If one con-
found. verts the corrosion rates obtained at 40°C, shown in
The effect of increasing temperature on the cor- Figure 12, into corrosion currents (approximately
rosion rate as measured by weight loss and LPR in 8 A/m2 to 10 A/m2) and compares them with the
3% NaCl solutions without HAc and solutions where cathodic limiting current observed under the same
100 ppm HAc was added (85 ppm undissociated HAc conditions, shown in Figure 11 (which are approxi-
at pH 4, 22°C) is shown in Figure 12. It is evident mately 10 A/m2), this confirms that at elevated
that the corrosion rates measured using LPR and by temperatures the corrosion process is under mass-
weight loss are not in perfect agreement, even when transfer control.
HAc is not present. Nevertheless, the changes in cor- The temperature effect was investigated further
rosion rate observed with HAc concentration and tem- by looking at the effect of adding various amounts of
perature are significant when compared to the error HAc at 60°C to a 3% NaCl solution at pH 4 (Figure
measured for each technique individually. It appears 13). It is evident that even a 10-ppm HAc addition to
that as the temperature increases, the influence of the solution affects the corrosion rate at 60°C, while
HAc is more pronounced. For example, at 22°C, add- adding 1,000 ppm leads to catastrophic corrosion
ing 100 ppm HAc increases the corrosion rate approx- rates. It is worth noting that the 1,000-ppm HAc, 24-h
imately 30%, while at 60°C, the same increase in HAc weight-loss experiment was performed four times. In

CORROSION—Vol. 63, No. 2 185

CORROSION SCIENCE SECTION

two of the tests, weight-loss measurements on the ACKNOWLEDGMENTS

order of approximately 45 mm/y were observed, and
these results are presented in Figure 13. In the other During this work, K.S. George has been sup-
two experiments, the samples experienced some pit- ported by a grant from the Ohio University’s Stocker
ting corrosion. No discernable difference between the Fund. The authors would like to acknowledge the
four experiments can be found to identify the trigger contribution of a consortium of companies whose con-
for the pitting corrosion. tinuous financial support and technical guidance en-
abled this work. They are BP, Champion Technologies,
Summary of Findings Clariant, ConocoPhillips, ENI, MI Technologies, Ondeo
From all of the above it appears that the key ef- Nalco, Saudi Aramco, and Total.
fect of HAc on the CO2 corrosion of carbon steel is
to introduce an additional cathodic species into the REFERENCES
solution. All the results point to the same conclusion:
1. J.-L. Crolet, N. Thevenot, A. Dugstad, “Role of Free Acetic Acid on
this additional corrosive species is the undissociated the CO2 Corrosion of Steels,” CORROSION/1999, paper no. 24
HAc. The additional cathodic reaction (Reaction [10]) (Houston, TX: NACE International, 1999).
2. C. de Waard, D.E. Milliams, Corrosion 31, 5 (1975): p. 175.
appears to be under charge-transfer control at room 3. C. de Waard, U. Lotz, D.E. Milliams, Corrosion 47, 12 (1991): p.
temperature, while at higher temperatures it becomes 976.
mass-transfer-controlled. Consequently, the presence 4. C. de Waard, U. Lotz, “Prediction of CO2 Corrosion in Carbon
Steel,” CORROSION/1993, paper no. 69 (Houston, TX: NACE,
of HAc is a problem predominantly at high tempera- 1993).
tures and at low pH, where the concentration of un- 5. C. de Waard, U. Lotz, A. Dugstad, “Influence of Liquid Flow Ve-
dissociated HAc is high. locity on CO2 Corrosion: A Semi-Empirical Model,” CORROSION/
1995, paper no. 128 (Houston, TX: NACE, 1995).
6. A. Dugstad, L. Lunde, K. Videm, “Parametric Study of CO2
CONCLUSIONS Corrosion of Carbon Steel,” CORROSION/1994, paper no. 14
(Houston, TX: NACE, 1994).
7. S. Nes̆ić, G.T. Solvi, J. Enerhaug, Corrosion 51, 10 (1995): p. 773.
❖ Electrochemical measurements have shown that 8. S. Nes̆ić, J. Postlethwaite, S. Olsen, Corros. Sci. 52, 4 (1996): p.
the presence of HAc strongly affects the cathodic lim- 280.
9. S. Nes̆ić, B.F.M. Pots, J. Postlethwaite, N. Thevenot, J. Corros.
iting current, which is mass-transfer-controlled. The Sci. Eng. 1, paper no. 3 (1995).
anodic reaction (iron dissolution) was unaffected or 10. S. Nes̆ić, M. Novdsveen, R. Nyborg, A. Stangeland, “A Mechanistic
mildly retarded with increasing HAc concentration at Model for CO2 Corrosion with Protective Iron Carbonate Films,”
CORROSION/2001, paper no. 01040 (Houston TX: NACE, 2001).
room temperature. 11. B. Hedges, L. McVeigh, “The Role of Acetate in CO2 Corrosion:
❖ At room temperature the CO2 corrosion rate of car- The Double Whammy,” CORROSION/1999, paper no. 21
bon steel is only mildly affected by the presence of (Houston TX: NACE, 1999).
12. Y.M. Gunaltun, D. Larrey, “Correlation of Cases of Top of Line
HAc. At 22°C the corrosion process seems to be under Corrosion with Calculated Water Condensation Rates,” CORRO-
charge-transfer control and unaffected by flow. SION/2000, paper no. 71 (Houston, TX: NACE, 2000).
13. Y. Garsany, D. Pletcher, B. Hedges, “The Role of Acetate in CO2
❖ As the temperature increases, the HAc effect is
Corrosion of Carbon Steel: Has the Chemistry Been Forgotten?”
much more pronounced and much higher corrosion CORROSION/2002, paper no. 02273 (Houston, TX: NACE, 2002).
rates are obtained. It seems that HAc corrosion be- 14. M.W. Joosten, J. Kolts, J.W. Hembree, M. Achour, “Organic Acid
Corrosion in Oil and Gas Production,” CORROSION/2002, paper
comes mass-transfer-controlled at elevated tempera- no. 02294 (Houston, TX: NACE, 2002).
tures, making it more prone to flow effects. 15. J.O’M. Bockris, A.K.N. Reddy, Modern Electrochemistry (New
❖ It appears that the undissociated form of HAc is the York, NY: Plenum Press, 1973).
16. G. Schmitt, B. Rothman, Werkst. Korros. 28 (1977): p. 816.
key species primarily responsible for changes seen in 17. S. Nes̆ić, B.F.M. Pots, J. Postlethwaite, N. Thevenot, J. Corros.
the mild steel corrosion process. Sci. Eng. (1995).

186 CORROSION—FEBRUARY 2007