Effect of Precorrosion and Temperature on the Formation Rate of Iron Carbonate Film

7th Pipeline Technology Conference 2012 Effect of Precorrosion and Temperature on the Formation Rate of Iron Carbonate Film Winia Farida a,b, Tor Hemmingsen b, Tonje Berntsen b,c, Patrick Rabindran a a Wood Group Integrity Management, Løkkeveien 99, 4008 Stavanger, Norway b University of Stavanger, N-4036 Stavanger, Norway c Statoil ASA, Oslo, Norway Abstract Carbon dioxide (CO 2 ) corrosion is the main type of corrosion studied in oil and gas industry. The major concern with CO 2 corrosion is that it causes failure of equipment in the main downhole tubing and transmission pipelines, and thus may cause disruption in the production. Carbon dioxide gas dissolved in formation or injection water forms carbonic acid (H 2 CO 3 ). Carbonic acid is corrosive to carbon steel or low alloy steel. Corrosion rates in a CO 2 system can reach very high levels, but it can be effectively inhibited. One way of inhibition is by favoring the formation of the protective iron carbonate (siderite) scale. Magnetite scales are also formed in CO 2 systems, and corrosion product scales often consist of layers or mixtures of siderite and magnetite. Experiments were carried out in order to observe the iron carbonate scale formation on three common carbon steel materials of grades X65, St52 and St33. The materials were exposed to CO 2 saturated brine at 40 C and 80 C for precorrosion periods of twenty four or forty eight hours. The corrosion of the samples were monitored using Electrochemical Impedance Spectroscopy (EIS), Potentiodynamic Cathodic Scan (PSC), Rp/Ec trend and the samples were analysed by Scanning Electron Microscopy (SEM) with Energy Dispersed X-Ray System (EDS) analysis after experiments. In conclusion, the result of the experiments showed that the galvanostatic stimulated precorrosion and temperature had significantly contributed to the protective FeCO 3 formation on surface steel. Both for 40 C and 80 C, protective scales are formed on surface steel after few days to a week of exposure time. Keywords: CO 2 corrosion, FeCO 3 protective scale, Galvanostatic stimulated corrosion, EIS, Potentiodynamic cathodic scan, Rp/Ec trend and SEM with EDS analysis 1

1. Introduction The corrosive conditions, i.e. CO 2, H 2 S and free water in the well stream are transported in pipelines. Most fields at Norwegian Sector have low H 2 S concentration while CO 2 is the dominating species in corrosion. Therefore, corrosion of the inner walls of steel pipelines and process equipment is caused by CO 2 in form of general and pitting corrosion will often cause problems in the transportation of oil and natural gas [1-4]. One possible way to control CO 2 corrosion is to allow forming a protective layer of FeCO 3 scale on steel surface. FeCO 3 (iron carbonate) is the main corrosion product in aqueous CO 2 system which deposit on surface steel and act as protective scale [5]. However, the formation of a FeCO 3 film can be protective or non-protective dependant on the conditions which they are formed influence by temperature, ph, CO 2 partial pressure, water chemistry etc [6]. The objectives of this experiment were to promote protective FeCO 3 film by increasing the bicarbonate concentrations and to study the effect of temperature on the formation of FeCO 3 film on precorroded samples. In order to make a more rapid protective FeCO 3 formation film, the samples were precorroded by use of applied anodic current. During precorrosion exposed, Fe 3 C (carbide) phase was enriched on the surface which is expected to be necessary to promote FeCO 3 formation film [7]. The terms of precorrosion here is defined as applying 0.25 ma/cm 2 anodic current by galvanostatic anodic polarization in order to increase the corrosion rate and to achieve a uniformly corroded in the steel surface of a specimen. The carbon steels X65, St52 and St33 used in these experiments have some different chemical composition, especially in carbon (C), manganese (Mn) and silicon (Si). The differences in composition influence the amount of carbide, and thus also the corrosion rate. An anodic current was applied for 24 hours or 48 hours in the experiments which may cause different amount of enriched Fe 3 C on the surface and thus the free corrosion potential. The specific experimental conditions were chosen for galvanostatic stimulated precorrosion for periods of 24 hours or 48 hours in order to achieve a uniformly corroded surface with iron carbide (Fe 3 C). Temperatures of 40⁰C and 80⁰C were used in these experiments in order to study the effect of temperature on formation rate for FeCO 3 film. The protective properties of the film are improved when the temperature is increased. It is showed at temperatures below 60 C that the film is easily removable, while a stable protective film is formed above temperature 60 C [8]. Furthermore, with the variation of precorrosion time and temperature, different corrosion conditions were obtained. Electrochemical impedance spectroscopy, potentiodynamic cathodic sweeps, and Rp/Ec trend were methods used to follow the corrosion of these steels. In addition, SEM picture with EDS analyses were conducted to describe the electrode surface of Fe 3 C and FeCO 3 on the surface steels. Furthermore, the expected results from the experiments are that a stable protective film is formed and corrosion rate decreases. 2

2. Experimental 2.1 Experimental Setting The terms of precorrosion here is defined as using an applied galvanostatic anodic current of 0.25 ma/cm 2. Electrochemical impedance spectroscopy, potentiodynamic cathodic sweeps, and Rp/Ec trend were methods used to follow the corrosion of these steels. 2.1.1 Base Solutions Base solutions in the experiments were; 1 g/kg NaCl, 50 wt% MEG in CO 2 purging under the precorrosion process. Furthermore, 100 mmol/kg NaHCO 3 was added just after galvanostatic exposure time. Table 2.1. Solutions and gas used in the experiments No Chemicals Description a 1 g/kg NaCl in distilled water NaCl used is Ph.Eur Sodium Chloride for analysis quality. b 50 wt% MEG (50% aqueous MEG in distilled water) Mono Ethylene Glycol (MEG) that used is 97% purity of MEG. c 1 atm CO 2 CO 2 in 30 kg gas cylinder from Yara Praxair AS. d 100 mmol/kg NaHCO 3 in distilled water NaHCO 3 which used is Merck for analysis quality. The experiments are conducted at 40 o C and 80 o C in a water bath. The water baths used are Julabo TW 20 and Yellow Line ET Basic. Gamry Software was used for measuring and controlling the corrosion of specimens during an experiment. Computers inserted a potentiostatic card using software Gamry Instruments Framework Version 5.61 (2010) and Gamry Instruments Framework Version 5.50 (2008) were used in the experiment. The chemical compositions of working electrodes are presented in Table 2. The electrodes were cylindrically formed as shown in Fig.2.1, with an exposed area of 3.14 cm 2. 3

Table 2.2. Chemical composition of specimen: X65 steel, St33 steel and St52 steel (from material certificate) Specimen Element C Si Mn S P Cr Ni V Mo Cu Al Sn Nb X65 0.08 0.25 1.54 0.001 0.019 0.04 0.03 0.045 0.01 0.02 0.038 0.001 0.043 St33 0.07 0.19 0.87 0.004 0.012 0.56 0.01 0.032 0.01 0.01 0.035 0.001 - St52 0.13 0.38 1.29 0.008 0.015 0.07 0.09 0.035 0.01 0.34 0.05 0.015 - The setup was according to standard Gamry methods, with a carbon steel is connected to working and working sense cell cables, a counter electrode (platinum) connected to counter and counter sense cell cables and a reference electrode (Saturated Calomel Electrode/SCE) connected to a reference cell cable. All electrodes are connected to a Gamry Potensiotat (PC3) set in Potentiostat/Galvanostat/ZRA modes and immersed into the electrolyte as shown in the Fig.2.1. Fig. 2.1 Corrosion cell potentiostat set-up. 2.1.2 Galvanostatic Anodic Polarization The methods to perform for precorrosion experiments were a) to measure the corrosion potential for 10 minutes, b) potentiodynamic sweep parameters, and c) setting precorrosion time (24 hour, 48 hour) with anodic current 0.25 ma. 2.1.3 Electrochemical Impedance Spectroscopy EIS characterization was carried out under following parameter: DC Voltage 0 V vs. Eoc, AC voltage 10 mv rms, frequency range from 500 Hz to 0.007 Hz with 10 points/decade. 2.1.4 Potentiodynamic Cathodic Scan Potentiodynamic cathodic scans were performed from +5 mv vs. Eoc, to -300 mv vs. Eoc with a scan rate of 0.2 mv/s. 4

2.1.5 Rp/Ec Trend Rp/Ec trend were performed from -5 mv vs. Eoc to +5 mv vs. Eoc with a scan rate of 0.05 mv/s and repeat time for 60 minutes. 2.1.6 Scanning Electron Microscope with EDS Analysis SEM analysis were done in order to show the presence of both FeCO 3 and carbide structure on the steel surface. In SEM analysis, an ultra-high resolution Hitachi S- 4800 equipped with Noran System Six energy dispersive spectrometer (EDS) for element analysis was used. SEM picture with EDS analysis were obtained by using 15.0 kv accelerating voltage of secondary electron image resolution, backscattered electron image resolution amount 3.0 nm guaranteed (at 15 kv YAG detector). 2.2 Experimental procedures Base solution was purged with CO 2 for 2 hours before the electrode was exposed to the solution and the ph was checked to be within the range 4.0-4.2. The E corr was measured and a potentiodynamic cathodic sweep was performed to -300 mv vs E corr with sweep rate 0.5 mv/s in order to activate the surface of the electrode. The specimen was galvanostatically polarized in the anodic direction at 0.25 ma for 24 hours or 48 hours. EIS measurement scan, Potentiodynamic sweep, and Rp/Ec trend were performed at intervals 2 hours, 10 hours, 24 hours, 48 hours, 96 hours, 192 hours and 216 hours. 3 Results and discussion 3.1 Galvanostatic anodic polarization A longer precorrosion time results in higher amount of Fe 3 C on the surface, and also a rougher surface which gives an increased exposed surface [9]. Theoretically the corrosion potentials after 48 hours precorrosion time should become more positive than after 24 hours precorrosion time. Furthermore, at 80⁰C, the value of potentials at the end of precorrosion period is should be more positive compared to 40⁰C. At 80⁰C, steel is corrodes faster and is considered to enhance Fe 3 C concentration compared to 40⁰C at same exposure time. The increasing of potentials during precorrosion period was caused by Fe 3 C formation. The condition was supported in theory that increase in corrosion rate during the precorrosion period were caused by several factors as (1) removal of a protective oxide films, (2) galvanic coupling to the uncorroded iron carbide (cementite) film, (3) increase in the true specimen surface area, and (4) acidification of the solution inside the corrosion product film. The results of galvanostatic anodic polarization are illustrated in Fig.3.1 and Table 3.1. 5

Fig. 3.1 illustrated the potentials of 3 steels with different precorrosion times and temperatures; (a) Temperature 40 C, (b) Temperature 80 C Table 3.1. The experiment results of the three steels with variation of precorrosion time and temperature Specimen T ( C) t Precorrosion (hours ) E corr (mv) I -300 mv (µa) E G (mv) X65 40 24-598 -252-552 48-660 - 302-601 St52 24-397 -5-533 48-668 -247-592 St33 24-640 -276-581 48-676 -262-620 X65 80 24-720 -810-655 48-666 -681-654 St52 24-728 -773-665 48-676 -772-683 St33 24-678 -1-688 48-650 -1-693 6

A summary of the results with variation of precorrosion time and temperature is given on Table 3.1. The force corrosion potential (Ecorr) is measured in the start. A current of -300 mv is applied for 24 or 48 hours and the potential (E G ) and current (I -300 mv) is measured in the end of the precorrosion period. 3.2 Electrochemical Impedance Spectroscopy A typical Nyquist plot of CO 2 corrosion is given in Fig. 3.2a, where the solution resistance, (Rs) is found at high frequencies and the polarization resistance (Rp) is found at low frequencies. For longer exposure times, a diffusion controlled reactions is observed at 80 o C, see Figs 3.2b, 3.3b and 3.4b [10]. The Nyquist plot for X65, St52 and St33 steels showed the highest polarization resistance mostly was obtained at the end of 216 hours as shown in Figures below. An increase in the polarization resistance at low frequencies indicated that a protective film was formed and a lower corrosion rate was observed. The larger diameter was obtained at the end of 216 hours indicated the corrosion rate decreased more rapidly at the end of 216 hours of EIS measurement. The high impedance values at low frequency shows that the protective FeCO 3 film has formed on surface steel and lead to a decrease in corrosion rate. The Nyquist plots are presented in the following Figs. 3.2-3.5. Fig.3.2 Nyquist plots result of St52 steel after 216 hours of immersion in solution (a) in 48 hours precorroded sample at 40 C, (b) in 24 hours precorroded sample at 80 C 7

Fig. 3.3 Nyquist plots result of St33 steel (a) in 24 hours precorroded sample at 40 C after 192 hours of immersion in solution, (b) in 24 hours precorroded sample at 80 C after 216 hours of immersion in solution Fig.3.4 Nyquist plots result of St33 steel after 216 hours of immersion in solution (a) in 48 hours precorroded sample at 40 C, (b) in 48 hours precorroded sample at 80 C 8

Fig.3.5 Nyquist plots result of X65 steel in 48 hours precorroded sample at 40 C after 216 hours of immersion in solution The Nyquist plots at 40 C and 80 C in Figs. 3.2-3.4 showed that with the increased of temperature gave significantly more a rapid protective film. For example in Fig.3.4, both experiments at 40 C and 80 C gave a Rp value of 500Ω with no film while at 192 hours, the experiment 80 C has higher impedance values at low frequency (1500 Ω) than the experiment at 40 C (600Ω). This indicated that a protective film has formed earlier at 80 C than for 40 C. 3.3 Potentiodynamic Cathodic Scan In general, the results of potentiodynamic cathodic scan from X65, St52 and St33 steels showed that the cathodic currents decreases with increasing time. The decreasing in cathodic reaction in Fig 3.6 (a) and 3.7 (a) indicate the scale is formed on the steels. Mostly the results of potentiodynamic cathodic scan have supported the EIS scan results which showed that the protective film has began to form at 192 hours. Therefore, it is slow down the corrosion process which proven by the results of potentiodynamic cathodic scan. The potentiodynamic cathodic scan is shown in Figs 3.6-3.11. 9

Fig.3.6 Potentiodynamic cathodic sweeps of X65 steel (a) 24 hours precorroded sample at 40⁰C after 2-216 hours (b) 24 hours precorroded sample at 80⁰C after 2192 hours Fig.3.7 Potentiodynamic cathodic sweeps of St52 steel (a) in 24 hours precorroded sample at 40⁰C after 2-192 hours, (b) 24 hours precorroded sample at 80⁰C after 2-96 hours 10

Fig.3.8 Potentiodynamic cathodic sweeps of St33 steel (a) in 24 hours precorroded sample at 40⁰C after 2-96 hours, (b) 24 hours precorroded sample at 80⁰C after 2-216 hours Fig.3.9 Potentiodynamic cathodic sweeps of X65 steel after 2-216 hours (a) in 48 hours precorroded sample at 40⁰C, (b) 48 hours precorroded sample at 80⁰C 11

Fig.3.10 Potentiodynamic cathodic sweeps of St52 steel after 2-216 hours (a) in 48 hours precorroded sample at 40⁰C, (b) 48 hours precorroded sample at 80⁰C Fig.3.11 Potentiodynamic cathodic sweeps of St33 steel after 2-216 hours (a) in 48 hours precorroded sample at 40⁰C, (b) 48 hours precorroded sample at 80⁰C 3.4 Rp/Ec Trend Rp/Ec trend is a method where the Rp values, and thus the corrosion rate, and the corrosion potential can continuously be followed. In general, corrosion rate decreased significantly after 95 hours which indicated that the protective scale FeCO3 was formed. This trend is also been confirmed by EIS scan data and potentiodynamic cathodic scan data. A summary of corrosion rate data at the end of exposure period with variation of precorrosion time and temperature is presented in Table 3.2. 12

Table 3.2 Corrosion rate of X65, St52 and St33 steels with variation of precorrosion time and temperature T (⁰C) 40 80 Precorrosion time (h) 24 48 24 48 Exposure time (h) Corrosion rates at various exposure time (mm/year) X65 St52 St33 2 0.31 0.17 0.21 47 0.3 0.08 0.09 95 0.28 0.07 n.a 191 0.43 n.a n.a 2 0.24 0.28 0.18 47 0.24 0.26 0.17 95 0.16 0.23 0.17 191 0.05 0.13 n.a 2 0.83 0.23 0.15 47 0.01 0.15 0.13 95 0.01 0.01 0.08 191 0.01 n.a 0.01 2 0.87 0.25 0.18 47 0.04 0.19 0.12 95 0.03 0.07 0.07 191 0.03 0.02 0.01 As shown in Table 3.2, the corrosion rate for all steels decreased with time, and the lowest corrosion rate is obtained at the end of exposure time as expected. However, the higher corrosion values for X65, 24 hours precorrosion at 40⁰C is probably not correct and caused by too low exposure temperature by a failure in the water bath (approx. 20⁰C) after 95 hours. The corrosion rate results in Table 3.2 show that St33 steel has the lowest corrosion rate compared to others steel. The low corrosion rate can be caused by the chemical composition of St33 which has highest Cr content and act as a good corrosion resistant. 3.5 Surface Morphology The specimens were removed from the electrolyte after the experiments and examined in an optical microscope to see the differences of specimen before and after the experiments. Figure 3.12 shows typical microscope pictures of a specimen before and after anodic galvanostatic exposure and Rp/Ec trend where a typical FeCO 3 film is formed. 13

Figure 3.12. Picture of the specimen (a) before the corrosion exposure; after precorroded 48 hours at 80 C (b) X65, (c) St52, (d) St33 steels (a) (b) (c) (d) 3.6 SEM with EDS Analysis The SEM with EDS analysis revealed that Fe 3 C was formed on the steel surface which increased corrosion process after 95 hours. FeCO 3 was deposited and accumulated on surface steel and acted as a protective scale. The protective scale is useful as it decreases the corrosion rate as shown by Rp/Ec trend data. Fig. 3.13. SEM picture with EDS analysis at forced precorrosion time 24 hours and temperature 40⁰C (a) X65 after 216 hours exposure time, (b) St52 after 192 hours exposure time (a) (b) As shown in the EDS analysis in Fig. 3.13 (b), FeCO 3 and Fe 3 C were formed on St52 steel and X65 steel, after 24 hours precorrosion at temperature 40⁰C. Temperature increases significantly affects the FeCO 3 formation on a steel substrate as shown by SEM with EDS analysis in Fig. 3.14. At a temperature of 80 C, it much more formed on steel surface than at a temperature of 40 C. Fig. 3.14 shows SEM with EDS analysis of St33 and St52 at forced 48 hours precorrosion time at 80 C. 14

Fig. 3.14 SEM picture with EDS analysis of (a) St33 at forced precorrosion time 24 hours and temperature 80⁰C after 192 hours exposure time, (b) St52 at forced precorrosion time 48 hours and temperature 80⁰C after 216 hours exposure time. FeCO 3 FeCO 3 Steel Substrate (a) (b) 4 Conclusions The potentials of steels increased with time of applied current (precorrosion). The increasing of potentials was caused by Fe 3 C formation during precorrosion period which stimulated corrosion process. At 80 C, a more rapid protective FeCO 3 film is formed on surface than at 40 C. EIS measurements showed that the protective film is earlier to form at 80 C than at 40 C. Diffusion reactions are found at a temperature of 80 C in EIS measurement but the species involved is not further quantified. The decreasing potential from potentiodynamic cathodic scan after 216 hours of immersion in solution indicates protective scale formation on surface steel. At 40 C and at 80 C, the protective scales seemed to be formed on surface steel after 95 hours exposure time. St33 has the lowest corrosion rate compared to X65 and St52 steels due to its high Cr content. 15

References [1] C. De Waard, D.E. Milliams: Prediction of Carbonic Acid in Natural Gas Pipelines. First International Conference on the Internal and External Protection of Pipes. (1975) Paper F-1. [2] C. De Waard, U. Lotz, D.E. Milliams: Predictive Model for CO 2 Corrosion Engineering in Wet Natural Gas Pipelines. Corrosion 47 (1991) 976-985. [3] C. De Waard, U. Lotz: Prediction of CO 2 Corrosion of Carbon Steel, in Prediction of CO 2 Corrosion in the Oil and Gas Industry. Institute of Materials Publisher, UK (1994) pp. 30-49. [4] C.A. Palacios, J.R. Shadley: Characteristics of Corrosion Scales on Steel in a CO 2 -Saturated NaCl Brine. Corrosion 47 (1991) 122-127. [5] S. Slåtten: CO 2 Corrosion of Carbon Steel with Low Temperature. NTNU, Trondheim (2007) [6] S. Nesic: Key Issues Related to Modelling of Internal Corrosion of Oil and Gas Pipelines A Review. Corrosion Science 49 (2007) page 4308-4338 [7] A. Dugstad: Mechanism of Protective Film Formation During CO 2 Corrosion of Carbon Steel. NACE CORROSION/98, paper no.31 [8] N.O. Ågotnes, T. Hemmingsen et.al: Comparison of Corrosion Measurements by Use of AC-impedance, LPR and Polarization Methods on Carbon Steel in CO 2 purged NaCl Electrolytes. NACE Corrosion 99, Paper No.27. [9] J.L. Mora Mendoza, S. Turgoose: FeCO 3 Influence on the Corrosion Rate of Mild Steel in Aqueous CO 2 System Under Turbulent Flow Conditions, Corrosion Science 44 (2002) 1223-1246 [10] Z.Q. Bai, C.F. Chen, M.X. Lu, and J.B. Li, Analysis of EIS characteristics of CO2 corrosion of well tube steels with corrosion scales, Applied Surface Science 252 (2006) 7578 7584. 16