CO2腐蚀文献2

FUNDAMENTAL ASPECTS OF CO 2 METAL LOSS CORROSION – PART II: INFLUENCE OF DIFFERENT PARAMETERS ON CO 2 CORROSION MECHANISMSGuenter SchmittLaboratory of Corrosion ProtectionIserlohn University for Applied SciencesFrauenstuhlweg 31, D-58644 IserlohnGermanyMichaela HörstemeierInstitute for Technical and Macromolecular ChemistryAachen University of TechnologyWorringerweg 1, D-52074 AachenGermanyABSTRACTThe literature has been reviewed with respect to information gained in the recent 20 years on CO 2 corrosion of materials used in the oil and gas industry. The paper discusses the effect of materials-related, medium-related and interface-related parameters on general (uniform) and localized corrosion.INTRODUCTIONCarbon dioxide corrosion of steels is influenced by materials-related, medium-related and interphase-related parameters. The paper will address the present day knowledge on the effect of these parameters on the corrosion of steels with respect to general corrosion and localized corrosion including pitting, preferential weld corrosion, flow induced localized corrosion and stress cracking. The rate of general corrosion and the susceptibility to localized corrosion is mainly dependent on the formation of protective, semi-protective or non-protective corrosion product scales which are affected by temperature, CO 2 partial pressure, medium composition (iron ion concentration, presence of certain chemicals), pH, flow, alloy composition and mechanical stress.06112Paper No.©2006 NACE International. Requests for permission to publish this manuscript in any form, in part or in whole must be in writing to NACE International, Conferences Division, 1440 South Creek Drive, Houston, Texas 77084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association. Printed in the U.S.A.CopyrightExtensive studies have been performed during the last four decades on the mechanism of CO2 corrosion of steels and the understanding of parameter effects has increased considerably. However, quite a few mechanistic questions are still unsolved or are a matter of debate and require further research. The first general review on CO2 corrosion of steel was performed in 1983 and summarised by Burke [1] and Schmitt [2] in 1984. Recently the key parameters affecting the intensity of carbon dioxide corrosion have been addressed by Kermani et al. [3]. In the same year Lopez et al. reviewed the influence of microstructure and chemical composition of steels on CO2 corrosion with respect to the influence of these parameters on the efficiency of corrosion inhibitors [5]. Several prediction models for carbon dioxide corrosion of oil and gas pipelines have been developed [5-13] and Nyborg [14] presented a complete overview of all relevant modelling approaches.FACTORS INFLUENCING CO2 CORROSIONThe properties of solid corrosion product layers formed on metal surfaces during corrosion reactions control the corrosion rate. This holds as well for the corrosion of steel in CO2 environments. Depending on the formation and the characteristics of iron carbonate scales, CO2 corrosion can be classified into general corrosion and localized corrosion. At low temperature ranges (< ca. 60°C) no protective films are formed on the surface due to the high solubility of iron carbonate and, therefore, general corrosion dominates. At temperatures at and above ca. 80°C the solubility of FeCO3 decreases considerably and high supersaturation leads to FeCO3 precipitation and very dense and protective iron carbonate films can form [14]. When discussing the protectivity of iron carbonate scales also the kinetics of iron carbonate precipitation has to be considered. At sites of nonuniform formation or localized destruction of carbonate films or scales severe localized corrosion can occur.General and localized CO2 corrosion is influenced by a number of factors which can be divided into interface-related, materials-related and medium-related parameters. Interface-related parameters include temperature, flow rate, condensation, and presence of scales. Materials-related parameters are alloy composition, microstructure and heat treatment. The influences of pH, CO2 partial pressure, solution chemistry, and presence of oxygen belong to medium-related parameters. All parameters are interdependent and influence the CO2 corrosion in different ways. The aim of the present work is to give an overview of the present day knowledge on the effect of these parameters on CO2 corrosion.GENERAL CORROSIONInfluence of surface filmsCO2 environments the main end product of carbon steel corrosion is iron carbonate which Informs on the steel surfaces if the supersaturation of FeCO3 in the near-surface solution is sufficiently high [14,15]. FeCO3 precipitates when Fe2+ ions react with carbonate (CO32-) and bicarbonate (HCO3-) ions in the solution. Therefore a high supersaturation of Fe2+ and CO32-/HCO3- is necessary for the formation of protective films. Once the film is formed, it will remain protective at a much lower supersaturation [14]. To get a successful protection, the film must be adherent and cover the whole surface. Localized corrosion can occur if parts of the scale break down and cannot re-form. Nesic et al. described two mechanisms on the protective role of passive surface films [16]. They represent a physical barrier which retards the diffusive transfer of corrosive species and prevent further metal dissolution of the blocked steel surface.The protective properties of the surface scale depend on the characteristics of the material (metal composition, heat treatment/microstructure) and the environmental conditions (temperature, CO2 partial pressure, pH). The morphology and adherence of the corrosion product film has been often related to the presence of carbides. It is suggested that in case of ferritic-pearlitic microstructures the iron carbide phase strengthens the film and anchors it to the steel substrate [17]. This is in agreement with the fact that on carbon steels with a normalized (ferritic-pearlitic) microstructure the iron carbonate scale is better adherent to the metal surface than on quenched-and-tempered steels (martensitic-bainitic microstructure) [17-19]. Thus, the shape and distribution of the carbides plays a very important role.It is well recognized that the temperature strongly influences the conditions needed to form protective iron carbonate layers. At lower temperatures (ca. < 60 °C) the solubility of FeCO3 is high and the precipitation rate is slow and protective films will not form unless the pH is increased [15,20]. In this temperature range the corrosion rate increases with temperatures up to an intermediate range of ca. 60 – 80 °C. Above 60°C the protectiveness of the iron carbonate layer increases with temperature due to the decrease of iron carbonate solubility and, thus, the corrosion rate is reduced (Fig.1). At temperatures above 110°C magnetite (Fe3O4) may form through direct reaction between steel and water (Schikorr Reaction) [2] and at 130°C the steels are passivated [2]. On ferritic-pearlitic steels surface scales formed below 40°C mainly consist of iron carbide (Fe3C) with some FeCO3 and carbonates of alloying elements of the steel [20-21]. Fe3C is part of the original steel microstructure and it is suggested that iron carbides may become sites of the cathodic reaction [20]. In case of martensitic microstructure this model cannot be applied because the cementite is nearly homogenously dispersed and the grain size is very small. Furthermore, in-situ measurement of the conductivity of iron carbonate films revealed that iron carbonate films exhibit a very low electronic conductivity and act as an insulator [22]. Therefore, carbonate films as such cannot act as cathodic sites, e.g. they are not comparable with iron sulfides which exhibit electronic conductivity. This is also true when cementite is dispersed in the carbonate film.The precipitation rate of FeCO3 has been described as a slow and temperature dependent process and even under supersaturated conditions, high corrosion rates can maintain for weeks until protective iron carbonate layers are formed, specifically at low temperatures [23]. Furthermore, in flow systems corrosion films obviously can grow for months without giving protection unless the steel is exposed to stagnant or “wet” conditions [14]. During a few days stagnation, corrosion products can accumulate on the steel surface and form protective films [14]. Thus, the kinetics of FeCO3 precipitation seems to be a controlling factor for the protectiveness of the corrosion product layer. At higher temperatures the FeCO3 solubility is reduced and the precipitation rate is much faster, thus allowing the formation of iron carbonate films [20]. Protective carbonate scales can be recognized already by its morphology and crystallinity. At temperatures ≥ 90°C the scale is composed of well-defined and well-packed cubes, while at lower temperatures a flat grain-type appearance is found [24]. However, the morphology of iron carbonate scales strongly depends not only on the temperature, but on the pH and the CO2 partial pressure, as well. At higher pH values (e.g. 6.5), protective iron carbonate films can also form at room temperature [23-26,27].Effect of TemperatureThe temperature strongly influences the CO2 corrosion due to its effect on the rate of scale formation. As mentioned above, at low temperatures, corrosion rates increase because of high solubility of the FeCO3 film. As temperatures increase (around 60-80 °C) the iron carbonate layer becomes more adherent to the metal surface and more protective in nature resulting in a decrease of the corrosion rate[15,20,28-29]. This correlation between temperature and corrosion rate was reported by many researchers and Ueda et al. attribute the maximum corrosion rate to a critical temperature T max encountered both in carbon and chromium steels, however, differing in T max [30].Effect of FlowFlow affects the metal dissolution rate in different ways [31,32]. Below a critical flow intensity (Fig. 2) the corrosion rate increases only gradually with increasing flow intensity [29,33]. Above the critical flow intensity the interaction between the fluid and the wall becomes so intense that protective films or scales are destructed by near-wall turbulence elements which also prevent re-formation of the protective film [31,32]. In this case the dissolution of the unprotected (bare) metal surface is transport controlled. The typical apperance of flow-shaped surfaces with shallow pits and grooves is the result of chaotic actions of near-wall turbulences. The situation of a "bare" metal surface is encountered in CO2 corrosion of steel also under conditions leading to nonprotective films. Under such conditions the metal dissolution rate is transport controlled, as well, however, is significantly lower.In the presence of protective CO2 corrosion product films the flow intensity can influence to a certain extent the transport of cathodic species towards the steel surface yielding an increase of metal dissolution rates at high flow velocities [29,33]. At the same time flow may stimulate the removal of Fe2+ ions from the steel surface at the bottom of pores and may cause lower surface supersaturation and slower precipitation rates. This in turn yields less protective films and, hence, higher corrosion rates [29, 33,34]. At elevated temperatures critical flow intensities for destruction of protective films and initiation of flow-induced localized corrosion (FILC) are generally higher because films with higher protectivity are formed above certain temperature ranges discussed above. The influence of flow on localized corrosion will be discussed later in more detail.Influence of alloy compositionThe influence of alloy composition of low alloy steel has been investigated intensely in recent years. According to the reviews made in 1983/84 [1,2] the highest effect is encountered with additions of chromium. The corrosion rate is significantly decreased with increasing Cr content [1,2,15,28,30,35-40]. In recent years the interest in low alloy steels with increased Cr content in the order of 3 to 5% Cr has increased due to the pressure to reduce CAPEX and OPEX in oil and gas production. An intermediate alloyed steel between API 5CT grade L80 and 13Cr steel would be of interest which offers improved corrosion resistance but stays with its costs close to API 5CT grade. This prompted a number of investigations which were not only devoted to the chromium content but also to the influence of microalloying elements (V, Ti, Nb, Mo, Cu, Si) [18,35-44]. The idea behind this work was to avoid that chromium is tied up in carbides and then no more available for forming chromium oxide enriched "passive" layers. This should be achieved by reducing the carbon content [28,45] and adding elements that form more stable carbides (V, Ti, Nb, Mo).It appeared that microalloying influences the corrosion performance of 3% Cr steels significantly. Specifically microalloying with vanadium yielded excellent results. The corrosion resistance increases with increasing temperature. Compared to L80 steel the corrosion rate of 3% Cr is decreased by factors of 3 to 40 depending on the temperature (Fig. 3) and microalloying element. The effect of microalloying on the CO2 corrosion rate of steel is summarized in Fig. 4 [36,38]. However, for the effect of Cu in carbon steels controversial results were obtained by Kimura et al. [46] which indicated that increased Cu and Ni contents may accelerate mesa attack and additionally increase the general corrosion rate. Similar results were reported by Dugstad et al. in earlier investigation [47].Microalloyed 3% Cr steels have been announced as a new generation of low alloy steels meeting all requirements for downhole application. However, the benefit can only be used in those cases where corrosion rates are low enough that inhibition ist not necessary. Furthermore, in case of inhibition appropriate inhibitors have to be selected, because inhibitors that work perfectly at L80 steels may not necessarily perform well on surfaces enriched with chromium.Increasing the chromium content up to 5% Cr generally reduces the corrosion resistance of steel to uniform corrosion but localized corrosion like pitting or SSC could occur. In a study with 3 and 5% Cr steels the corrosion rate of the Cr steels was reduced by a factor of two or more compared to chromium free carbon steel, but 5% chromium steel suffered deeper localized attack than 3% chromium steels [35]. Different results were reported earlier in a 5 months field test exposure where 5% chromium steels showed superior corrosion resistance to general and localized corrosion than steels with 0.2- 3%Cr [48].Chromium steels with more than 12 % Cr are very resistant in severe sweet and mildly sour production conditions and have been used for effective corrosion control [40,49]. But it is also widely recognized that stainless steel exhibit less resistance to localized corrosion at elevated temperature and may be susceptible to stress corrosion cracking [40,49,50]. The influence of alloying elements on localized corrosion will be discussed later.Microstructure and Heat TreatmentThe microstructure plays an important role in corrosion of carbon steel in CO2-containing environments. Several authors have studied the influence of steel microstructure on the corrosion process in aqueous solutions containing CO2, although there is no general agreement on this issue [17, 19,30,44,51-54]. The majority of the investigations show that carbon steels with ferritic-pearlitic microstructure exhibit a better corrosion resistance than compared with martensitic or martensitic-bainitic microstructured steels [17,19, 44,53] but also the opposite has been reported [51,52].The effect of microstructures is attributed to different roles of different phases in acting as anodic or cathodic sites. The cementite (Fe3C) phase is generally thought to act cathodically and to enhance the corrosion of ferrite galvanically [30,54]. This, however, holds only for ferritic-pearlitic steels. Furthermore, lamellar pearlite seems to provide better grip for protective scales than particulate pearlite due to the cementite lamellae [17,19,30]. In this context, the influence of heat treatment on the shape and distribution of the cementite gains importance [51]. It appears as kind of a rule that the more uniform the distribution and the smaller the ferrite grain size, the lower the corrosion rate [55].Palacios and Shadley [19]postulated that the structural characteristics of iron carbonate corrosion-product (e.g. the adherence or thickness) depend on the microstructure of the material. It is claimed that iron carbonate is less tenacious and less crystalline on quenched-and-tempered (QT-) steels with martensitic microstructure, and that the corrosion product scale forms more dense and larger crystals on normalised steel with ferritic-pearlitic microstructure [19].This agrees with observations by Ueda et al. [51] who compared the morphology of the corrosion product films of three different carbon steels. Ferritic-pearlitic steel yielded homogeneous FeCO3 films, while carbon steels with martensitic-bainitic microstructure gave porous iron carbonate films [51]. Thus, the better corrosion resistance of ferritic-pearlitic C-Mn steels could be attributed to differences in the morphology of corrosion products. This is explained in more detail in [17] and [19]. It was also reported that in the initial state J55 C-Mn steel with ferritic-pearlitic microstructure showed higher mass loss rate than N80 with martensiticmicrostructure [53, 125]. The rapid initial mass loss of J55 indicates that the film formation occurs earlier and thus corrosion protective films form faster on J55 than on N80 steel. Therefore, J55 exhibits a better corrosion resistance than N80 steel [53].Attempts to profit from both Cr addition (1% Cr) and ferritic-pearlitic microstructure were not successful [48]. It appeared (Fig.5) that the corrosion rate of Cr bearing steel with martensitic microstructure was lower than the corrosion rate of the steel with ferritic-pearlitic microstructure. Furthermore, the corrosion resistance of Cr bearing steel with martensitic microstructure was superior to the corrosion rate of a C-Mn steel with martensitic microstructure. According to Mishra et al [55] the microstructural effect due to different heat treatment becomes vanishingly small at temperatures above 60°C. At the end, the microstructure of materials has to be selected according to the needs of the technical application. If no acceptable compromise is possible with low alloy steel, high alloy steels like 9Cr1Mo or 13Cr should be looked at.Influence of CO2 partial pressureThe CO2 partial pressure plays an important role in CO2 corrosion both for film-free conditions (formation of non-protective films) and for film-forming conditions. In most publications the authors established a relationship between CO2 partial pressure and corrosion rate [29, 56-60]. However, higher CO2 partial pressure means not necessarily also higher corrosion rates. This is a matter of environmental conditions.Generally, under film-free conditions higher CO2 partial pressures results in higher corrosion rates [28,56,58,60] by reducing the pH [61] and by increasing the rate of carbonic acid reduction [62]. This agrees with the finding [63] that the anodic reaction is practically unaffected when the CO2 partial pressure is increased from 3 to 20 bar while the cathodic limiting current density is strongly increased due to a higher reservoir of carbonic acid. This again indicates that in CO2 corrosion the pH is only an orientating parameter for assessing the corrosivity. What really matters is the availability of carbonic acid and the surface sites for its reduction.Under conditions favourable for protective film formation (high supersaturation at the interface with subsequent carbonate precipitation) high CO2 partial pressure can also reduce the corrosion attack due to a lower availability of cathodic sites. It falls within this picture that increasing the CO2 partial pressure from 4 to 18 bars under film-free conditions in a horizontal wet gas flow yields an increase of the corrosion rate from about 3 mm/y to about 8 mm/y [56]. On the other hand, an increase of the CO2 partial pressure in the same flow system from 3.8 to 10.6 bar reduces the max. corrosion rates from about 15 to 0.2 mm/y (Fig. 6) [56] under conditions when semiprotective films are formed, e.g. in the pH range below pH 5.2. In another case, an increase of the CO2 partial pressure from 1 to 30 bar in an unbuffered 0.1 % NaCl solution yielded an increase of the corrosion rate of rotated low alloy steel probes (average flow rate at the probes: 1m/s) from 2 mm/y to 4 mm/y at 40°C and from 1.5 mm/y to 2.5 mm/y at 90°C, respectively [28]. Thus, the correlation between the corrosion rate of low alloy steels and the CO2 partial pressure is quite complicated and by no means linear. So far there is no model which can predict this correlation satisfactorily. What model would e.g. predict that in 1% unbuffered NaCl at 40 °C an increase of the CO2 pressure from 10 to 95 bar would decrease the corrosion rate of low alloy steels from about 5.5 (10 bar) to 0.5 mm/y (95 bar) [64]. It appears that at high CO2 partial pressures (e.g. >20 bar) the protectivity of iron carbonate films is considerably higher than expected from existing models. This was also found in a series of pH-static experiments with CO2 pressures up to 50 bar [27].Effect of pHThe pH affects the CO2 corrosion of carbon steels by different mechanism. Increased pH values generally lead to a reduction of the corrosion rate by influencing the electrochemical mechanisms and the formation of protective iron carbonate films [29,47,17,56,65]. By an increase of pH the cathodic reduction of H+ is slowed down which decreases the anodic dissolution rate of iron. Furthermore at very high pH values protective carbonate scales are formed on the surface that reduces the corrosion rate significantly [25-26,29]. This is due to the effect of decreasing the solubility of iron carbonate in the solution [29]. In flow loop tests with carbon steels protective scales could only be observed for pH’s greater than 5.0 [26]. In another study it was found that at lower temperatures protective scales are only formed when the pH is increased to values of at least 6 while above 80°C protective scales are easily formed [25]. In multi-phase flow conditions the increase of pH leads to a significant decrease of the corrosion rate of carbon steel due to the formation of protective corrosion product films above pH 6.2 [56]. Another study revealed a decrease in corrosion rate by a factor of 3 with an increase of pH from 3.5 to 5.5 depending on the flow velocity (Fig.7) [29].Brine composition and ion concentrationThe iron content in the corrosive medium is dependent on the anodic reaction mechanisms in CO2 corrosion of steel [66] and the solubility varies with the CO2 concentration, temperature, pH and the concentration of other ions in the solution [67]. The Fe2+ concentration has a large influence on the corrosion rate in CO2 environment given high corrosion rates at low Fe2+ levels [7,65,67]. An increase in iron ion concentration results in higher supersaturation which leads to the precipitation of protective corrosion passive films [34]. In a study at 70°C, 50 ppm Fe2+ could reduce the corrosion rate of carbon steels from 5 mm/y to 2 mm/y at pH 4 by the formation of more protective corrosion films [65].Cations such as Ca2+ and Mg2+ react with carbonic acid and deposit calcium and magnesium carbonates. This codeposition of earth alkali and iron carbonates enhances the scale formation and, hence, reduces the corrosion rates [17]. Ueda et al. found that the corrosion rates of carbon steels in CaCl2 solutions were smaller than those in 5% NaCl + 2,5meq/l NaHCO3 solution with same pH, but the steels showed higher localized attack [17]. Thus, it is not only the pH that commands the protectivity of scales but also its composition. Earlier investigations in the author's lab had shown, that even minor concentrations (e.g. 10-2 mole/L) of earthalkali salts or salts of Zn, Cd, Sn, Pb, and other heavy metal ions would be found accumulated in the corrosion product scale [68].The formation and stability of protective corrosion layers depend heavily on the supersaturation of the bulk solution and hence, any variation in this level of supersaturation could affect the severity of the corrosion [3]. High supersaturation of anions and cations leads to precipitation of a corrosion layer which would reduce the corrosion rate through several kinds of effects [3].Effect of oxygenOxygen in CO2 systems exhibits a strong effect on the corrosion rate and facilitates the formation of localized attack. While this is known for some time only a few papers have been devoted to this problem [69-71]. Martin studied the corrosion consequences of oxygen entry into both sweet and sour systems [71]. He found that the corrosion rate in CO2 system is accelerated in the presence of oxygen by 0.5 mm/y per ppm of oxygen at medium velocity and ambient temperature (Fig. 8). He suggested that the mechanism is consistent with a change in surface corrosion product, which accelerates H2CO3reduction. Investigations of Lyle and Schutt revealed the same relationship between the concentration of O 2 and the general corrosion rate while also pitting was observed [70].Effect of H 2SThe effect of H 2S in CO 2 systems is particularly important because low concentrations of H 2S may arise in otherwise sweet environments, e.g. by microbial influenced reservoir souring. Depending on the concentration, hydrogen sulfide significantly affects CO 2 corrosion by influencing the type and properties of the corrosion film or acting as a promoter of anodic dissolution through sulfide adsorption. At high concentrations H 2S may also affect the pH [3].Below the NACE sour limit of 3 mbar H 2S [72] the composition of the corrosion films depends on the CO 2 /H 2S ratio, the corrosion rates and the temperature [73]. Dunlop et al. [74] suggested ratios which for room temperature allow the estimation of the carbonate and sulfide regime. According to this suggestion the carbonate regime exists when the ratio of the partial pressure of CO 2 and H 2S is above 500, while the sulfide regime is reached when this ratio of partial pressures is below 500. New investigations from Smith and Pacheo [75] resulted in a more precise equation for the calculation of critical constants assuming coexistence of FeCO 3 and all forms of Fe X S Y (Equ. 1 and 2).()()Svap H Svap H aq CO H vap CO vap CO aq CO O H FeCO FeS a K K a K a K 223222223//⋅⋅⋅⋅= (1) Svap H vap CO FeCO FeS a a C K 223/⋅= (2)In Equ. 1 the a values stand for activities or fugacities of the various species in the aqueous or vapor phase (the fugacities often equal corresponding partial pressures) and the K values for equilibrium constants. Equ. 2 is a simplification of Equ. 1 by combining the equilibrium constants into a single constant and making the simplifying assumption that the activity of water is equal to 1.0, to produce an equation that is very similar to the one reported by Dunlop [74].Calculations based on solubility data lead to theoretical CO 2 /H 2S ratios for the transition between FeS and FeCO 3 formation in the range of 105 -107 [73]. Observed CO 2 /H 2S ratios of 1000-5000 have been reported in literature [73,76,77].For worst case calculation of the corrosion rate (CR) of low alloy steel in slightly sour conditions the following Equ. 3 was suggested [75]:17925510848.510]/[⋅⋅⋅=−−pH T e y mm CR (3)The actual corrosion rates will be lower than the rate calculated from Equ. 3 because the equation assumes infinite flow rates. At the transition point between sweet and slightly sour corrosion, the corrosion rate drops as the mechanism shifts from the CO 2-controlled to the slightly sour corrosion mechanism. This has been verified experimentally by Smith and Pacheco when testing the steel。