effect of corrosion inhibitors

CORROSION SCIENCE SECTION

CORROSION—Vol. 65, No. 1 3ISSN 0010-9312 (print), 1938-159X (online)

08/000001/$5.00+$0.50/0 © 2009, NACE International

Submitted for publication December 2007; in revised form, August 2008.

‡ Corresponding author. E-mail: [email protected]. * Department of Chemical Engineering, Norwegian University of

Science and Technology, NO-7051 Trondheim, Norway. ** Institute for Energy Technology, Materials and Corrosion Technol-

ogy Department, NO-2027 Kjeller, Norway.

Effect of Corrosion Inhibitors and Oil on Carbon Dioxide Corrosion and Wetting of Carbon Steel with Ferrous Carbonate Deposits

M. Foss,‡,* E. Gulbrandsen,** and J. Sjöblom*

ABSTRACT

The wettability of the steel surface is an important factor co-determining the risk of corrosion in multiphase pipelines for transportation of oil and gas. The present paper deals with the effect of carbon dioxide (CO2) corrosion inhibitors on the wettability of carbon steel with ferrous carbonate (FeCO3) deposits. The wettability was studied by contact angle mea-surements at 25°C, 1 bar CO2, and 3 wt% sodium chloride (NaCl). Both oil-in-water and water-in-oil experiments were performed. Inhibitor performance was studied in CO2 corrosion tests at 60°C, 1 bar CO2, and 3 wt% NaCl, both in the absence of oil and in tests where the steel specimens were alternately exposed to oil and aqueous phases. Three inhibitors were investigated: two commercial, inhibitor-based chemicals (an oleic imidazoline salt [OI] and a phosphate ester [PE]), and cetyltrimethylammonium bromide (CTAB), a well-characterized quaternary ammonium compound. A refined, low-aromatic oil product was used in the tests. Both OI and PE substan-tially decreased the tendency of water droplets to spread on an initially oil-wet steel surface with FeCO3, but had a negli-gible effect on the wettability of initially water-wet surfaces. The inhibitor performance of the OI and PE was significantly improved by the presence of oil. Corrosion rates one order of magnitude lower than in the tests without oil exposure was obtained. The addition of inhibitor had a limited effect on the corrosion rate of steel with FeCO3 deposits in the absence of oil. The results indicated that the enhanced performance was caused by a modification of the inhibitor film. The water-in-oil

contact angle measurements indicated that CTAB significantly enhanced water wettability.

KEY WORDS: carbon dioxide corrosion, carbon steel, corrosion inhibitor, ferrous carbonate, wetting

INTRODUCTION

Carbon steel pipelines are commonly used in the transport of oil and gas. Carbon steel piping and pro-cess equipment are subject to corrosion caused by the presence of water and acidic gases such as carbon dioxide (CO2), hydrogen sulfide (H2S), and acetic acid (CH3COOH). The use of corrosion inhibitors and the manipulation of corrosion product films are two pos-sible ways of lowering the corrosion rate to acceptable levels.1-2

CO2 corrosion inhibitors typically consist of am-phiphilic, surface-active molecules with hydrocarbon chains in the range C12-C18.2-7 Amphiphilic mole-cules have a strong tendency for adsorption and self-assembly, driven by the hydrophobic interaction.2-3

The self-assembled hydrocarbon chains form struc-tures with hydrocarbon phase properties that may change electrochemical reaction rates, influence the mass transfer of reactants or reaction products, or simply block parts of the surface, and thus reduce the active area.2,4-7 The exact mechanisms of action for the inhibitors are, however, not fully understood. Some data on the kinetics of corrosion inhibition has been published recently. Bilkova, et al., found that the CO2 corrosion inhibition was composed of two processes: first, a rapid process (order of minutes) connected to hydrophobically driven adsorption of the inhibitor


4 CORROSION—JANUARY 2009

that leads to inhibition of the anodic part reaction, and a second slower process (order of hours) that leads to a reduction in the corrosion rate through in-hibition of the cathodic part reaction(s).8

Previous inhibitor testing work has demon-strated that the performance of CO2 corrosion inhibi-tors, in many cases, was reduced when the steel was corroded before inhibition.9 Based on these results, ferrous carbonate (FeCO3) corrosion product depos-its are expected to have an impact on the inhibitor performance. FeCO3 deposits typically form in sys-tems containing CO2 at elevated temperatures and may significantly reduce the corrosion rate.1 The scal-ing tendency model10-11 gives a description of FeCO3 film growth and protectiveness that is consistent with most of the experimental findings.10,12-13 This model focuses on the competition between FeCO3 film growth and corrosion. The scaling tendency (ST) is defined as the ratio between corrosion film growth and the cor-rosion rate. Thus, a ST factor greater than 1 (ST > 1) indicates that the film growth is faster than the corro-sion and a protective FeCO3 deposit will form until a steady-state situation is reached. Nesic and coworkers concluded from numeric modeling that the reduction in corrosion rates was mainly caused by FeCO3 crys-tallites blocking the steel surface, and not by a reduc-tion in the mass-transfer rate of CO2.

11,14 Limited data has yet been published on the interaction between this partly protective corrosion deposit and corrosion inhibitors.4,15 The corrosion deposit may reduce access of the inhibitor to the steel surface, and inhibitor per-formance might differ significantly from what is seen on bare carbon steel surfaces.

The present work presents data on the perfor-mance of corrosion inhibitors on carbon steel with partly protective FeCO3 deposits in the presence of oil. Alterations in wettability were also investigated by contact angle measurements. The main objective for

the work was to obtain a better understanding of how inhibitors interact with an iron carbonate-covered steel surface in the presence of oil. As a starting point for our investigations, we tested two CO2 corrosion, inhibitor-based chemicals: a phosphate ester (PE) and an oleic imidazoline salt (OI). One widely studied sur-factant (cetyltrimethyl ammonium bromide [CTAB]16) was also tested (Table 1). The choices of inhibitors were based on applicability to real systems and previ-ous knowledge of the inhibitors.17

CTAB is a preferentially water-soluble quaternary ammonium compound with an aliphatic C16 chain that has been studied thoroughly as a surfactant16 as well as a corrosion inhibitor.8,18-21 Recent research has shown that CTAB molecules adsorbed on hydro-philic silica and mica form discrete structures, i.e., admicelles.16 The surface coverage was therefore poor. A similar conclusion was reached in quartz crystal microbalance studies of CTAB adsorption onto iron and gold.20-21

The OI contains an imidazoline group and a C17 hydrocarbon chain with a double bond; it is a reac-tion product of diethylenetriamine and a fatty acid.22 OI has been reported to partition preferentially to water. The adsorption of OI onto an iron surface was found to be a fast process. Studies of OI by atomis-tic simulation methods suggest that a self-assembled monolayer forms.2 Moon and Horsup found that the maximum of the inhibition efficiency was reached at about the same concentration as the minimum of the oil-water interface tension was reached.7

The detailed structure of the present PE is unknown. However, the general structure (Table 1) is a phosphate group connected to one or more hydro-carbon chains. Several authors have studied the cor-rosion-inhibiting properties of PE.23-25 The compounds are very effective, especially at moderate temperatures or in the presence of trace amounts of oxygen.24-26

TABLE 1Inhibitor Compounds Used in the Experiments

Inhibitor Compound Type Structure

Cetyltrimethyl-ammonium bromide (CTAB) Cationic

Phosphate ester (PE) Anionic

Oleic imidazoline salt (OI) Cationic


CORROSION—Vol. 65, No. 1 5

Surfactants, such as corrosion inhibitors, might change the wetting behavior of steel and FeCO3 sur-faces due to the formation of molecular structures at the surface. If the FeCO3 surface could be modified to become hydrophobic, it might attract an oil film to the surface of the steel and reduce the corrosion rate. Wettability has been widely studied on dielec-tric materials such as calcite (CaCO3) and some noble metals, but to date, no data has been published on the wetting behavior of FeCO3 in oxygen-free environ-ments.27-32 The influence of organic surfactants on wetting was investigated for some reservoir materials, but to our knowledge, no publication has discussed the impact on corrosion.30,33

Contact angle measurements have been used to determine the effect of liquids containing surface-active molecules, such as crude oil.14,34-35 In surface chemistry, the term “wetting” refers to the extent of contact established between a liquid and a solid sur-face when the two are brought together. The wetting properties of a surface/liquid system are determined by the balance of forces at the line of contact between the liquids and the solid.36 The wetting of an FeCO3 surface in an oil/water system is thus a function of the interfacial tension forces between oil and water (γow), water and FeCO3 (γsw), and oil and FeCO3 (γso).

36 The contact angle (θ) and the surface energies of the phases involved are related by Young’s Equation (1):

γ γ γ θso sw ow= + cos (1)

A sketch showing the direction of the forces in Equation (1) is seen in Figure 1. The reported contact angles were measured in the aqueous phase. A con-tact angle in the aqueous phase of 90° or greater gen-erally characterizes a surface as hydrophobic, while an angle of less than 90° means that the surface is hydrophilic.

EXPERIMENTAL PROCEDURES

Chemicals and MaterialsA refined, low-aromatic oil product was used as

model oil in the tests. It has a density of 0.788 kg/L and a boiling point range from 193°C to 245°C, cor-responding to kerosene.37 Analytical reagent (AR)-grade CTAB (99+ %) was used as-received. The OI was supplied as technical grade (67.5 wt% in a solvent), formed by the reaction of diethylene triamine and tall-oil. The PE was also supplied as a technical-grade product; details of its composition are not known. The reported concentrations are referred to the sup-plied product. An outline of the inhibitor structures is given in Table 1. The test brine consisted of distilled water with 3 wt% technical-grade NaCl added. CO2

gas was supplied as a 4.0 grade gas with a maximum O2 content of 10 ppm (by volume). The critical micelle concentration (CMC) of CTAB is 7 ppm in brine with 3 wt% NaCl.22 The CMC of the other products could not be determined, since no clear break points were found in the surface tension vs. concentration curves. Water-to-oil partitioning has been investigated in-house.16 Preferential partitioning to water was found for all three inhibitors.

The test specimens used in both the contact angle measurements and the corrosion inhibition experi-ments were machined from API X-65(1) ferritic-pearl-itic low-carbon pipeline steel (element analysis [wt%]: 0.08 C, 1.54 Mn, 0.03 Ni, 0.25 Si, 0.04 Cr, 0.019 P, 0.001 S, 0.045 V, 0.01 Mo, 0.038 Al, 0.04 Nb, 0.001 Sn).

Contact AngleThe contact angles (CA) were measured with a

commercially available drop tensiometer instrument, using a liquid measuring chamber. Continuous gas purging of the measurement chamber was done to avoid oxygen contamination in the tests. Two different types of contact angle measurements were made:

—water droplet on FeCO3-covered steel surface immersed in oil (denoted water-in-oil experi-ments)

—oil droplet on an inverted FeCO3-covered steel immersed in the aqueous brine (denoted oil-in-water experiments)

The experiments were conducted at different inhibitor concentrations. The fluids and corrosion coupons used in the experiments were prepared in a separate 3-L precorrosion cell. Formation of FeCO3 was achieved by corroding several test specimens for more than 200 h at 75°C. The electrolyte was bub-bled with CO2 at ambient pressure at least 3 h prior to immersion of the coupons to saturate with CO2 and deoxygenate the solution. The CO2 partial pressure

(1) American Petroleum Institute (API), 1220 L St. NW, Washington DC 20005.

FIGURE 1. Sketch showing the direction of the forces acting on the line of contact between the liquids and the solid, and the contact angle (θ) of the droplet.



was about 0.6 bar (water vapor pressure: 0.4 bar). Flat steel coupons (25 mm by 25 mm, 3 mm thick) were used in the tests. They were prepared by grind-ing with 1000-mesh silicon carbide (SiC) paper fol-lowed by cleaning with technical acetone (CH3COCH3) in an ultrasonic bath. Shortly before immersion in the test solution, the samples were cleaned with isopropa-nol (CH3CH[OH]CH3). All samples were mounted on a horizontal sample holder placed in the glass cells. The precorrosion stage was done in 2,400 mL brine fol-lowed by addition of 600 mL oil after 24 h, to ensure oil saturation of the brine. The inhibitor partition-ing between the oil and the brine was therefore equal to the partitioning in the corrosion experiments. The first inhibitor dose was added 1 h after the oil, and the concentration was increased in steps with inter-vals of 1 h. One sample was removed before each new inhibitor concentration increase. The samples were rapidly lifted through the oil layer and dried by blow-ing pressurized CO2 onto the surface of the steel to rid the steel surface of the liquid film covering the sur-face. The dry sample was then immersed in the con-tact angle cell, which was filled with either oil or brine sampled from the precorrosion cell. The oil-in-water experiments were done with an inverted surface where the droplet was deposited on the surface using the droplet flotation. The specimens were examined using scanning electron microscopy (SEM) after termination of the tests. This was done to ensure that the FeCO3 films were comparable for all the tests. Figure 2 shows a typical FeCO3 layer used in the CA experiments.

Corrosion ExperimentsCorrosion experiments were carried out on sta-

tionary cylinder (10 mm diameter, 10 mm length, and 3.14 cm2 exposed area) specimens in a 3-L thermo-statted glass cell at ambient pressure. The steel speci-mens were ground with 1000-mesh SiC paper wetted

with isopropanol, cleaned with technical acetone in an ultrasonic bath, and flushed with ethanol (CH3CH2OH) before immersion in the test solution. The electrolyte brine solution was identical to that of the contact angle measurements. The solution was continuously purged with CO2 at ambient pressure throughout the experiments. In the test with CTAB and oil, only 15 mL of oil was used. In the tests where the samples were immersed in the oil phase, 20 wt% of oil was used. The pH of the system was initially around pH 3.9. The pH increased to a maximum pH of about 6.0. The increased pH is caused by formation of HCO3

– in the cathodic reaction. After reaching the maximum, the pH decreased to 5.7 because the concentration of bicarbonate ions (HCO3

–) in the solution was lowered due to FeCO3 precipitation. The corrosion experiment can be divided into two stages:

—Formation of the FeCO3 layer. Formation of FeCO3 was achieved by the same technique as for the contact angle measurements. The tem-perature was 80°C during FeCO3 formation. The total corroding surface area in the corrosion tests corresponded to approximately 20 cm2/L. Corrosion rates were measured on one or more of the specimens. A partly protective FeCO3 layer started to form after approximately 100 h. The FeCO3 layer formation was stopped after about 200 h. The corrosion had dropped to approximately 0.5 mm/y after the FeCO3 forma-tion stage.

—Inhibitor testing on the FeCO3-covered speci-mens. Corrosion testing was performed at 60°C, to facilitate comparison with previous results. The inhibitor was added from a 1% inhibitor stock solution.

Two types of experiments were conducted: —Corrosion in brine without oil added. Inhibi-

tor was added in increasing concentrations throughout the tests.

—Corrosion in brine with oil. This was done to saturate the water phase with the oil compo-nents (mainly aliphatic hydrocarbons). Inhibi-tor was added stepwise in the test with CTAB. In the test with OI and PE, the inhibitor con-centration was held constant throughout the tests. The samples in the tests with OI and PE were also periodically immersed in the oil to determine the effect of oil on the corrosion rate. The oil immersion step consisted of a proce-dure where the samples were lifted out of the brine up into the oil layer covering the aqueous phase in the corrosion cell.

A three-electrode setup was used for the electro-chemical measurements. A silver/silver chloride (Ag/AgCl; 3 M potassium chloride [KCl]) electrode and a titanium ring surrounding the working electrode were used as reference and auxiliary electrodes, respectively. The polarization resistance (PR) tech-

FIGURE 2. SEM picture showing a cross section of a steel specimen with a FeCO3 film used in the contact angle measurements.



nique 38 was used to monitor the corrosion rate. This technique is also commonly referred to as the linear polarization resistance (LPR) technique. Measure-ments were done at 30-min intervals throughout the test. The potential was scanned from –5 mV to 5 mV vs. the open-circuit potential with a rate of 0.1 mV/s. The polarization resistance (Rp) was corrected for uncompensated resistance measured by electrochemi-cal impedance spectroscopy. The corrosion current density (Icorr) was calculated from Rp using Equation (2):

I B A Rcor p= / ( ) (2)

where A is the exposed surface area of the test elec-trode and B is a constant. In the case of simple acti-vation-controlled reactions, the value of the B can be determined from the Tafel slopes of the part reac-tions38 by Equation (3):

B

b bb b

a c

a c

=+2 303. ( )

(3)

where ba and bc are the Tafel slopes of the anodic and cathodic polarization curves, respectively. Empirical B values are commonly applied in the CO2 corrosion field due to the complexity of the polarization curves.

This complexity is caused by the different cathodic reactions taking place in CO2 corrosion.39-42 Anodic Tafel slopes in the range from 40 mV to 60 mV are reported.37 B values of 20 ± 5 mV are typically obtained for uninhibited CO2 corrosion by calibration to mass loss following the procedure of Macrides as described in Reference 43.41-42 This is consistent with ba values in the range from 40 mV to 60 mV and bc being infinite. In most cases of inhibited CO2 corro-sion, the cathodic polarization curves normally show Tafel behavior (bc about 120 mV), but the anodic polarization curves often deviate from Tafel behavior at higher polarization rates. A more detailed discus-sion on the determination of the value of the B con-stant can be found elsewhere.41-42. An empirical value of 20 mV was used for B in all the reported data. In this way, the reported results are also proportional to 1/Rp and thus reflect the changes in the measured parameter. The corrosion rate is reported in mm/y, and is obtained from the corrosion current using Faraday’s law calculation according to the anodic dis-solution reaction Fe = Fe2+ + 2 e–, and using 7.9 g/cm3 as the density of the steel.

RESULTS

Contact Angle—Water-in-OilA plot of contact angles vs. inhibitor concentra-

tion for tests containing OI, PE, and CTAB are plot-ted in Figure 3. The contact angle of the inhibitor-free water phase was about 40° in this system. Additions

of OI or PE lead to a transition from a hydrophilic to a hydrophobic surface. The tendency was seen for all concentrations of OI and PE. The contact angle for OI increased to 140° for concentrations above 20 ppm, at which point the curve seemed to reach a stable value. A contact angle of more than 155° was found for PE at concentrations of 20 ppm and above. When CTAB was used to complete water wetting, a contact angle <5° was found for all inhibitor concentrations. No signifi-cant change in contact angle with time was observed.

An illustration of the large difference in contact angle obtained in the PE and the CTAB tests at high concentrations is seen in Figure 4. When the cal-culated contact angles exceeded 160° or more, the droplet rolled off the surface if a tilt was introduced. This indicates a super-hydrophobic behavior where the contact angle approaches 180°.44 Contact angles approaching the complete wetting limits (0° or 180°) could not be determined precisely by the instrument.

Contact Angle—Oil-in-WaterFigure 5 shows contact angles measured in the

water phase for tests where the FeCO3 surface was exposed to the inhibited brine. The contact angle in the uninhibited water phase was approximately 17°. Addition of inhibitor caused some variations in the recorded contact angles for all three inhibitors, but the surface remained hydrophilic at all tested concen-trations.

The small change in contact angle obtained at intermediate concentrations of PE is illustrated in Figure 6. Some FeCO3 particles are dispersed in the aqueous phase, giving the images a cloudy appearance.

Corrosion Inhibition ExperimentsCetyltrimethylammonium Bromide — An introduc-

tory corrosion test in the pure aqueous phase showed

FIGURE 3. Contact angle for water-in-oil experiments steel with FeCO3 deposit. Experimental conditions: 1 bar CO2, 3% NaCl, and ambient temperature.



that this compound was not able to inhibit CO2 cor-rosion on bare carbon steel. The same result was expected for specimens with FeCO3, and tests without oil therefore were not performed.

Figure 7 shows the results for a test with CTAB in the presence of a small amount of oil. A partly protec-tive film of FeCO3 was formed after approximately 190 h of the test, lowering the corrosion rate to less than 0.5 mm/y. The temperature thereafter was low-ered from 80°C to 60°C before inhibitor and oil was added. The corrosion testing was performed at 60°C, to facilitate comparison with previous results. The change in temperature did not have a significant impact on the corrosion rate in this test. After 198 h of testing, 1.5 mL of oil was added to the test cell to saturate the brine with oil. The addition of oil did not affect the corrosion rate significantly. Stepwise addi-tion of inhibitor was done starting after 215 h. No reduction in the corrosion rate was observed for con-centrations lower than 100 ppm. When the inhibitor

(a) (b)

FIGURE 4. Photos of the water droplet in tests with 60 ppm PE (left) and 21 ppm CTAB (right). Both pictures were captured approximately 50 s after deposition of the droplet.

FIGURE 5. Contact angle for oil-in-water experiments on steel with FeCO3 deposit. Experimental conditions: 1 bar CO2, 3% NaCl, ambient temperature.

(a) (b)

FIGURE 6. Photos of the oil droplets in the tests containing PE at two different concentrations: 30 ppm (left) and at 42 ppm (right). Both pictures were captured approximately 50 s after deposition of the droplet, and are rotated 180°.



concentration was increased to 200 ppm toward the end of the test, the corrosion rate started decreasing after an initial increase following the inhibitor addi-tion. Addition of 15 mL oil did not have any additional effect on this behavior. The rate was ca. 0.2 mm/y at the end of the test, with a slowly decreasing trend. The corrosion potential increased from –620 mV to –590 mV during the course of the test.

Figure 8 shows the appearance of the FeCO3 film after the film formation stage at 80°C. Precipitation is seen on all parts of the surface.

Oleic Imidazoline Salt — Figure 9 shows corrosion rate vs. time for a test with OI in the absence of oil. An FeCO3 layer was grown at 80°C for more than 300 h until a stable corrosion rate had been reached. The corrosion rate was then 0.2 mm/y. Following the temperature change from 80°C to 60°C, an increase in the corrosion rate was seen. The corrosion rate increased from 0.2 mm/y to 1 mm/y before it started decreasing again. This might be caused by cracking of the FeCO3 layer. A sudden drop in corrosion poten-tial, from –580 mV to –620 mV, was also seen after the temperature change. The potential then increased toward the value it was before the temperature change. After 300 h the corrosion rate had dropped to below 0.3 mm/y and a more stable corrosion rate was observed. Consecutive additions of OI up to a total concentration of 20 ppm did not have any imme-diate effect on the corrosion rate. The corrosion rate had decreased to between 0.1 mm/y and 0.2 mm/y after more than 800 h of testing. The corrosion rate was still decreasing at this point.

Figure 10 shows corrosion rate vs. time for a test with OI where the sample was periodically exposed to

oil. FeCO3 was formed for 300 h at 80°C. The corro-sion rate stabilized around 0.45 mm/y before the tem-perature was changed to 60°C. No significant change in corrosion rate was seen after the change in tem-perature. Addition of 20 wt% oil to the test cell did not affect the corrosion rate. Addition of OI salt up to a concentration of 10 ppm led to a slight decrease in the corrosion rate. Periodic exposure to oil did not have any significant impact on the corrosion rate below this concentration. When the concentration was increased to 20 ppm, a significant drop in the corrosion rate could be seen; the rate had dropped to 0.03 mm/y after 400 h. Exposure to oil at this concentration led

FIGURE 7. Polarization resistance corrosion rate and Ecor vs. time. FeCO3 film formation for 190 h at 80°C. Experimental conditions: CTAB as inhibitor, 15 mL oil, 60°C (0.8 bar CO2).

FIGURE 8. SEM image of a cross section of a steel specimen with FeCO3. The image is captured after FeCO3 precipitation was finished at 80°C in the test with CTAB.



to a drop in the corrosion rate, but no lasting effect was seen. The decrease in corrosion rate was accom-panied by a large increase in the corrosion potential.

Phosphate Ester Compound — Figure 11 shows the corrosion rate vs. time for a test with PE in the absence of oil. The corrosion rate stabilized below 0.2 mm/y after the carbonate layer buildup, but increased to 1 mm/y after the change from 80°C to 60°C. This behavior is similar to what was seen in Fig-ure 9. After this increase in corrosion rate, a steady

decrease toward the corrosion rate after FeCO3 growth was seen. This decrease seemed to be affected by the stepwise addition of PE. The corrosion potential also showed a stepwise change depending on the PE addi-tion, increasing from –640 mV at the beginning of the test to –590 mV at the end of the test. After more than 600 h the corrosion rate was almost stable and had dropped to below 0.2 mm/y. No further effect on the corrosion rate and potential was found for concentra-tions above 70 ppm PE.

FIGURE 9. Polarization resistance corrosion rate and Ecor vs. time. FeCO3 film formation for 290 h at 80°C. Experimental conditions: OI as inhibitor, no oil, 60°C (0.8 bar CO2).

FIGURE 10. Polarization resistance corrosion rate and Ecor as a function of time. FeCO3 was formed for 300 h at 80°C. Oil exposures are indicated by gray lines. Experimental conditions: OI as inhibitor, 20 wt% oil, 60°C (0.8 bar CO2).



Figure 12 shows corrosion rate vs. time for a test with PE where the sample was exposed periodically to oil. FeCO3 was formed for 300 h at 80°C. The corro-sion rate stabilized around 0.35 mm/y before the tem-perature was changed to 60°C. No significant change in corrosion rate was seen after the change in tem-perature. Addition of 20 wt% oil to the test cell did not affect the corrosion rate. Increasing the concen-tration of PE to 10 ppm and then to 20 ppm lead to a reduction in corrosion rate to 0.05 mm/y. No last-ing effect of the direct exposure to the oil was seen. When the concentration was increased to 35 ppm, a

significant drop in the corrosion rate was seen; the rate had dropped to 0.005 mm/y after 380 h. Fur-ther increasing the concentration beyond 35 ppm did not have any significant effect on the corrosion rate. The decrease in corrosion rate was accompanied by a large increase in the corrosion potential, which had increased to –550 mV by the end of the test.

Electrochemical Impedance Spectroscopy Mea-surements — Figures 13 and 14 show moduli of the impedance vs. frequency EIS measured in the tests with OI and PE, respectively. The uncompensated resistance (Ru) and the sum of the uncompensated

FIGURE 11. Polarization resistance corrosion rate and Ecor vs. time. FeCO3 film formation for 280 h at 80°C. Experimental conditions: PE as inhibitor, no oil, 60°C (0.8 bar CO2).

FIGURE 12. Polarization resistance corrosion rate and Ecor vs. time. FeCO3 film formation for 300 h at 80°C. Oil exposures are indicated by gray lines. Experimental conditions: PE as inhibitor, 20 wt% oil, 60°C (0.8 bar CO2).



resistance and the polarization resistance (Rp + Ru) can be found at the high- and low-frequency plateaus of the modulus curves, respectively. Ru may comprise the resistances of the electrolyte, the FeCO3 film, and a possible oil layer at the surface or in the film. In the test with OI (Figure 13), Rp increased from an initial 15 Ω to more than 1,900 Ω at the end of the experi-ment. Ru increased from an initial value of 1 Ω to 25 Ω. In the test with PE, Rp increased from 1 Ω to 20 Ω. The polarization resistance increased from 10 Ω to more than 8,000 Ω. A low-frequency plateau was not reached at 10 mHz for any of the highly inhibited systems; therefore, a definite value for Rp was not measured.

DISCUSSION

The corrosion rate experiments clearly demon-strated a significant effect of oil on the performance of both the PE and the OI salt. The addition of oil led to a corrosion rate one order of magnitude lower than what was found in the absence of oil for these two inhibitors. The observed stabilization of the hydropho-bic surface in the water-in-oil experiments and the improved effect of the inhibitors in the presence of oil will be the main focus of the discussion.

The focus of the contact angle measurements was to obtain results that may be relevant to a field situ-

FIGURE 13. Plot showing the modulus of the impedance (Zmod) vs. frequency at different stages of the experiment with OI and oil exposure. The PR corrosion rate curve for this experiment is given in Figure 10.

FIGURE 14. Plot showing the modulus of the impedance (Zmod) vs. frequency at different stages of the experiment with PE and oil exposure. The PR corrosion rate curve for this experiment is given in Figure 12.



ation. The oil-in-water experiments resemble the wet-ting behavior at a surface that is initially water wet, such as the bottom of a pipe with stratified flow or stagnant pools of water at low points in a pipeline. Oil dispersed in the water phase might, in stratified flow, hit the surface and interact with it, and a potentially beneficial transition from a water-wet to an oil-wet surface could be achieved. The water-in-oil experi-ments resemble the opposite situation where water droplets entrained in the oil-phase interact with an initially oil-wet surface. In this case, the beneficial effect would be to avoid water droplets attaching to the steel wall and spread out, and thus wet the steel.

Figure 3 shows a significant difference between the different inhibitors. The addition of CTAB increased water wetting and concentrations as low as 1 ppm led to a highly hydrophilic surface. CTAB is known to emulsify oil45 and may have a deter-gent effect. Such an effect where oil is washed off the steel surface could explain the very low contact angles found for CTAB. The development of hydro-philic inhibitor aggregates of CTAB on the steel sur-face might also explain the hydrophilic behavior. For low concentrations of inhibitor, the monomers are expected to adsorb individually. At higher con-centrations (CMC and above) the development of a structured CTAB layer might be expected. Atkin, et al., reported hemicelles and worm-like structures of CTAB on silica and mica substrates for concentra-tions around CMC and above.16 The change in the wetting properties of the steel might be explained by this effect but, to our knowledge, no studies on the structure of CTAB on corroding steel has been pub-lished to date. An ordered surfactant structure of this kind is also likely for both OI and PE, but the results for the contact angle of water in oil suggest a different structure. OI and PE both increased the hydrophobic-ity dramatically of the oil-wet surface. In the field-rel-evant situation this effect might lower the steel pipe wall’s affinity to water and thus prevent water drop-lets from spreading on an already oil-wet surface.

The tested inhibitors had little effect on the wet-ting of the initially water-wet surface, and the surface remained hydrophilic (Figure 5). This indicates that it is difficult to form an oil film on a water-wet FeCO3 surface.

The corrosion tests were conducted to investigate the impact of oil on the corrosion rate in inhibited systems of FeCO3-covered steel. When determining the polarization resistance, and thus the corrosion rate for systems with efficient inhibition, some con-siderations regarding the scan rate used in the PR measurements should be made. If the scan rate is too fast, the measured corrosion rate can be overesti-mated due to the contribution of capacitive current.46 It is therefore necessary to determine the maximum scan rate for the system. An evaluation of the maxi-mum scan rate based on a Bode plot can be found

elsewhere,35 where it is concluded that a scan rate of 0.1 mV/s underestimates corrosion rates of systems with a break-point frequency of 0.025 Hz or lower.

The introduction of oil into the solution improved the performance of OI and PE. In the absence of oil, the inhibitor performance was poor on steel with FeCO3. It is clear that the presence of oil leads to a change in the way the inhibitor interacts with the FeCO3-covered steel surface. The reduction in corro-sion rate may be caused by various mechanisms.

Three tentative mechanisms are:—a modification of the FeCO3 layer caused by the

presence of oil—inhibition of oil in the FeCO3 film or the forma-

tion of an oil film at the solid surface—a modification of the inhibitor film by oilFeCO3 growth is a slow process (Figures 7 and

9), and modification of the FeCO3 film is therefore unlikely to cause the rapid drop in corrosion rate observed after the oil exposures.

Inhibition of oil in the FeCO3 film or the creation of an oil film on the surface would also lead to an increase in the Ru values. To evaluate the probabil-ity of an oil film it is possible to estimate the change in Ru following the creation of an oil film on the sur-face. The electrical resistivity (ρ) for the oil used in the experiments at 20°C is ρ >1012 Ω·m.39 The increase in uncompensated resistance due to an oil film of thickness L can then be calculated as: Ru = ρ (L/A), where A is the surface area (3.14 cm2). An increase in uncompensated resistance (Ru) of more than 10 Ω would be easily detectable by the impedance measure-ments. The thickness corresponding to 10 Ω of resis-tance is only L < 3 nm, i.e., of the same thickness as an inhibitor film. The presence of an oil film is there-fore not likely.

The results therefore support the idea that the oil enhanced the protectiveness of the inhibitor film. This might be caused by coadsorption of inhibitor mole-cules and hydrocarbons from the oil phase, causing a modified inhibitor film structure.

The results obtained in the present study provide important information for inhibitor testing. It is clear that the addition of hydrocarbon might have a signifi-cant effect on the performance of some inhibitors. The ability of some surfactants, such as CTAB, to function as a detergent may also be important to future inhibi-tor testing.

CONCLUSIONS

The effect of CO2 corrosion inhibitors on the wet-tability of FeCO3-covered carbon steel was investi-gated. Also investigated was the effect of oil on the performance of these inhibitors.v Both OI and PE decreased the tendency of water droplets to spread on an initially oil-wet surface with FeCO3 and may thereby reduce the risk of corrosion



caused by dropout of water droplets entrained in the oil.v CTAB increased the water wettability of initially oil-wet steel even at very low concentrations. This effect can be explained by a detergent effect and/or the for-mation of hydrophilic inhibitor films. CTAB improved the ability of water droplets to spread out on the oil-wetted FeCO3 surface.v The presence of dissolved hydrocarbons in the brine improved the inhibitor performance of both the OI salt and the PE significantly. The performance of the OI salt and the PE improved further after a direct exposure of the test specimen to the oil phase.v The results indicate that oil influences the struc-ture of the inhibitor film to improve the protective-ness. The evidence suggests that the improved performance is caused by a coadsorption effect rather than the formation of a microscopic and stable oil film.v Addition of oil may strongly influence the inhibitor performance and is therefore an important factor in inhibitor performance testing.

ACKNOWLEDGMENTS

The work is part of a joint project between the Institute for Energy Technology and the Norwegian University of Science and Technology, sponsored by the Research Council of Norway (Project no. 158913/I30).

REFERENCES

1. A. Dugstad, “Mechanism of Protective Film Formation During CO2 Corrosion of Carbon Steel,” CORROSION/98, paper no. 31 (Houston, TX: NACE International, 1998).

2. S. Ramachandran, B. Tsai, M. Blanco, H. Chen, Y. Tang, W.A. Goddard, Langmuir 12, 26 (1996): p. 6,419.

3. D.F. Evans, H. Wennerström, The Colloidal Domain, 2nd ed. (New York: Wiley-VCH, 1999), p. 41.

4. I.L. Rozenfield, Corrosion Inhibitors (New York, NY: McGraw-Hill, 1981), p. 67.

5. P. Li, J.Y. Lin, K.L. Tan, J.Y. Lee, Electrochim. Acta 42, 4 (1997): p. 605.

6. S. Meng, E.G. Wang, S.W. Gao, J. Chem. Phys. 119 (2003): p. 15. 7. T. Moon, D. Horsup, “Relating Corrosion Inhibitor Surface Active

Properties to Field Performance Requirements,” CORROSION/ 2002, paper no. 02298 (Houston, TX: NACE, 2002).

8. K. Bilkova, E. Gulbrandsen, M. Knag, J. Sjöblom, “Kinetic Study of Model Corrosion CO2 Inhibitors,” Proc. 10th European Symp. on Corrosion and Scale Inhibitors, no. 12 (Ferrara, Italy: Ann. Univ. Ferrara, N.S., 2005), p. 585.

9. E. Gulbrandsen, R. Nyborg, T. Løland, K. Nisancioglu, “Effect of Steel Microstructure and Composition on Inhibition of CO2 Cor-rosion,” CORROSION/2000, paper no. 23 (Houston, TX: NACE, 2000).

10. E.W.J. van Hunnik, B.F.M. Pots, E.L.J.A. Hendriksen, “The For-mation of Protective FeCO3 Corrosion Product Layers in CO2 Cor-rosion,” CORROSION/96, paper no. 6 (Houston, TX: NACE, 1996).

11. S. Nesic, K.-L.J. Lee, V. Ruzic, “A Mechanistic Model of Iron Car-bonate Film Growth and the Effect on CO2 Corrosion of Mild Steel,” CORROSION/2002, paper no. 02237 (Houston, TX: NACE, 2002).

12. K. Chokshi, W. Sun, S. Nesic, “Iron Carbonate Film Growth and the Effect of Inhibition of CO2 Corrosion of Mild Steel,” CORRO-SION/2005, paper no. 05285 (Houston, TX: NACE, 2005).

13. M.L. Johnson, M.B. Tomson, “Ferrous Carbonate Precipitation Kinetics and Its Impact on CO2 Corrosion,” CORROSION/91, paper no. 268 (Houston, TX: NACE, 1991).

14. M. Nordsveen, S. Nesic, R. Nyborg, A. Stangeland, Corrosion 59 (2003): p. 443.

15. M. Foss, K. Bilkova, E. Gulbrandsen, M. Knag, J. Sjöblom, Proc. 10th European Symp. on Corrosion Inhibitors, no. 12 (Ferrara, Italy: Ann. Univ. Ferrara, N.S., 1995), p. 601.

16. R. Atkin, V.S.J. Craig, E.J. Wanless, S. Biggs, Adv. Colloid Inter-face Sci. 103 (2003), p. 219.

17. M. Foss, E. Gulbrandsen, J. Sjöblom, “Alteration of Wettability of Corroding Carbon Steel Surface by Carbon Dioxide Corrosion In-hibitors—The Effect on Carbon Dioxide Corrosion Rate and Con-tact Angle,” Corrosion (2009): in press.

18. G. Schmitt, B.N. Labus, H. Sun, N. Stradmann, “Synergisms and Antagonisms in CO2 Corrosion Inhibition,” Proc. 8th European Symp. on Corrosion Inhibitors, no. 10 (Ferrara, Italy: Ann. Univ. Ferrara, 1995).

19. G. Schmitt, B.N. Labus, “Effect of Surfactants on Inhibitor Per-formance in CO2 Corrosion of Steel under Highly Turbulent Flow Conditions,” CORROSION/94, paper no. 37 (Houston, TX: NACE, 1994).

20. M. Knag, J. Sjöblom, G. Øye, E. Gulbrandsen, Colloids Surf., A 250 (2004): p. 269.

21. M. Knag, J. Sjöblom, E. Gulbrandsen, J. Dispersion Sci. Technol. 26 (2005): p. 207.

22. A.J. McMahon, Colloids Surf. 59 (1991): p. 187.23. B. Alink, B. Outlaw, V. Jovancicevic, S. Ramanchandran, S.

Campbell, “Mechanism of CO2 Corrosion Inhibition by the Phos-phate Esters,” CORROSION/99, paper no. 37 (Houston, TX: NACE, 1999).

24. R.L. Martin, R. Annad, Corrosion 37, 5 (1981): p. 297.25. R.L. Martin, Mater. Perform. 22, 9 (1983): p. 33.26. Y. Wu, K.E. McSperitt, G.D. Harris, “Corrosion Inhibition and

Monitoring in Seagas Pipeline System,” CORROSION/88, paper no. 365 (Houston, TX: NACE, 1988).

27. D.C. Standnes, T. Austad, J. Petrol. Sci. Eng. 39 (2003): p. 431.28. D.C. Standnes, T. Austad, Colloids Surf. A 216, 1-3 (2003): p.

243.29. S. Strand, E. Hognesen, T. Austad, Colloids Surf. A 275 (2006): p.

1-10.30. K.A. Rezaei Gomari, A.A. Hamouda, J. Petrol. Sci. Eng. 50 (2006):

p. 14,030. 31. C. Drummond, J. Israelachvili, J. Petrol. Sci. Eng. 33 (2002): p.

243-259.32. D.Y. Wang, H.W. Duan, H. Mohwald, Soft Matter 1, 6 (2005): p.

31.33. C. Legens, T. Palermo, H. Toulhoat, A. Fafet, P. Koutsoukos, J.

Petrol. Sci. Eng. 20 (1998): p. 277.34. S. Hoeiland, T. Barth, A.M. Blokhus, A. Skauge, J. Petrol. Sci.

Eng. 30, 10 (2001): p. 91.35. A. Horsthemke, J.J. Schroder, Chem. Eng. Process. 19, 5 (1985).36. P. Atkins, J. de Paula, Atkins’ Physical Chemistry, 7th ed. (Ox-

ford, U.K.: Oxford University Press, 2002), p. 154.37. “Shellsol D-70” data sheet, www.shellchemicals.com, May 27,

2005.38. J.R. Scully, “Electrochemical Tests,” in Corrosion Tests and Stan-

dards, ed. R. Baboian (West Conshohocken, PA: ASTM Interna-tional, 1995).

39. S. Nesic, J. Postlethwaite, S. Olsen, Corrosion 52 (1996): p. 280.40. W.J. Lorenz, F. Mansfeld, Corros. Sci. 21 (1981): p. 647.41. E. Gulbrandsen, J. Kvarekvål, H. Miland, Corrosion 61, 11 (2005):

p. 1,086.42. E. Gulbrandsen, A. Granå, Corrosion 63 (2007): p. 1,009.43. A.C. Macrides, Corrosion 29 (1973): p. 148.44. L. Lacroix, M. Lejeune, L. Ceriotti, M. Kormunda, T. Meziani, P.

Colpo, F. Rossi, Surf. Sci. 592 (2005): p. 182.45. G. Gutiérrez, Á. Cambiella, J.M. Benito, C. Pazos, J. Coca, J.

Hazard. Mater. 144, 3 (2007): p. 649.46. R.G. Kelly, J.R. Scully, D.W. Shoesmith, R.G. Buchheit, Electro-

chemical Techniques in Corrosion Science and Engineering (Boca Raton, FL: CRC Press, 2002), p. 140.

effect of corrosion inhibitors

Documents

carbon dioxide corrosion

corrosion rate of steel

risk of corrosion

corrosion rates

effect of corrosion

co2 corrosion tests

presence of oil

absence of oil