Corrosion behavior of X65 steel in gaseous and liquid CO2 transport environments containing multiple impurities and different H2O content

The mechanism of H2O content in different CO2 phase states on the corrosion of pipeline steel was studied by using high-pressure corrosion simulation tests, scanning electron microscopy (SEM), energy dispersive X-ray spectrometry (EDS), X-ray diffraction (XRD), Raman spectroscopy, X-ray photoelectron spectroscopy (XPS), and water chemistry simulation calculations. The results show that the corrosion mechanism of X65 steel did not change significantly under different CO2 phase state systems, and the corrosion products were similar. As the H2O content of the system increased, the sulfur-containing products in the corrosion products increased and the degree of corrosion worsened. Additionally, because the aqueous phase formed by the liquid CO2 system contained more corrosive substances, which promote the electrochemical corrosion process of X65 steel, the corrosion degree of X65 steel in a liquid CO2 system was significantly higher than that of X65 steel in a gaseous CO2 system.


Introduction
Carbon capture, utilization, and storage (CCUS) technology is vital to achieving CO2 emission reduction [1,2].As an efficient mode of CO2 transport, pipeline transportation has been widely used in CCUS projects [3,4].However, there is inevitably a certain amount of corrosive impurity components (e.g., H2O, SO2, NO2, O2 and H2S) in the CO2 fluid transported by pipelines [5].The corrosion problems caused by them seriously threaten the service safety of carbon steel pipelines [6].Previous studies [4,7,8] have indicated that H2O content significantly affects the corrosion process of transportation pipelines.Especially when liquid water appears, the CO2 or other trace impurities in the fluid dissolved in the liquid water will significantly deteriorate the internal environment of the pipeline, and the corrosion degree of transportation pipelines will significantly increase [9,10].Virtually, during on-site CO2 pipeline transportation, the H2O content of the transported CO2 fluid is much lower than its saturated solubility in CO2 [11][12][13].However, it is worth noting that the various impurity gases contained in the transport fluid will change the solubility of H2O in CO2.Liu et al. [14] recently found that in a supercritical CO2-impurities transportation system, when the H2O content was only 20 ppmv, the X52 pipeline steel underwent significant corrosion.Additionally, the corrosion rate of X52 pipeline steel was close to 0.1 mm/y as the H2O content increased to 500 ppmv.The corrosion degree increased significantly.It follows that in a CO2 transportation system containing various impurities, liquid water still appears in the system even if the H2O content is far lower than its saturated solubility.The carbon steel pipelines still face severe corrosion problems [4,10,13,15].Furthermore, during long-distance pipeline transportation, the energy exchange between the fluid and the outside world, crossing different regions, pressurization, release, etc. will also lead to changes in the phase state of the CO2 fluid.It can cause the distribution of impurity components such as H2O and derivative products to change over time and space, affecting the corrosion of carbon steel pipelines and causing safety issues in pipeline transportation [10,16,17].Presently, the research on H2O content in transporting CO2 is mostly limited to a single and stable transport state system [18,19].There is a relative lack of research on the impact of H2O content in different phase CO2 transport environments on the corrosion of carbon steel pipelines.It is mostly a transportation environment without impurities [20].However, the types of impurity gases in CO2 delivery fluids will inevitably increase with the diversification of CO2 capture sources.Moreover, the transported CO2 fluid will inevitably undergo certain phase changes during long-distance pipeline transportation.Therefore, controlling the H2O content threshold in a complex transportation environment is imperative while considering corrosion and economic costs.This work aims to reveal the effect of H2O content and CO2 phase state on pipeline steel corrosion by using corrosion analysis technologies such as high-pressure corrosion simulation tests, SEM, EDS, Raman spectroscopy, XPS, and water chemistry simulation calculations.It is expected to provide a scientific basis for the limitation of H2O content when impurities such as O2, H2S, SO2, and NO2 coexist in the CO2 transport pipeline.

Materials and preparation
The tested steel in this study was X65 pipeline steel, the chemical composition is listed in Table 1, and the microstructure is shown in Figure 1.The X65 steel consisted of ferrite and granular bainite and belongs to low-carbon micro-alloy steel.The samples were machined into 40 mm × 15 mm × 3 mm, which were used for the corrosion simulation tests and product analysis.Besides, the samples needed to be marked and polished with sandpaper to 800 mesh, then rinsed with ethanol and dried with compressed air before the tests.Additionally, the weight of samples was measured by the electronic balance (the precision was 0.1 mg).

Corrosion simulation tests
The corrosion simulation tests were performed in a C-276 autoclave (as shown in Figure 2) under the test conditions in Table 2, where the H2O content of 3815 ppmmol and 3242 ppmmol was the saturated solubility of water in gaseous CO2 (5 MPa-50 ℃) and liquid CO2 (10 MPa-25 ℃), respectively.(when the H2O content does not exceed 3815 ppmmol and 3242 ppmmol, water will completely dissolve in CO2, meaning there is no free water phase in the initial corrosion environment).Besides, considering the limitations of the European DYNAMIS project [11] and ISO 29713 standard [21] on the content of impurity components in pipeline CO2 fluids, various impurity gas contents were selected as 200 ppmmol.Before the test, setting up 4 parallel samples which were placed on a fixture made by PTFE, and then adding the water for the test (deoxidized and deionized water).Removing residual air from the reactor during installation, closing the reactor and continuously injecting high-purity CO2 into the reactor for 2 hours.Heating the reactor to the setting temperature, then adding O2, H2S, SO2, and NO2 to the desired concentration, and finally adding CO2 to the setting pressure.The corrosion experiment period was 72 hours.The temperature indicator in the control cabinet had an accuracy of 0.1 ºC and the pressure indicator had an accuracy of 0.01 MPa.
After finishing the test, the samples were taken out from the autoclave, dehydrated with absolute ethanol, dried with cold air, and photographed with a digital camera.Besides, one of the samples remained for the corrosion product analysis, and the others were used to calculate the average corrosion rates by Eq. ( 1) [22,23].Moreover, the corrosion product film was removed by the acid pickle, which consisted of 0.5 L HCl (the density is 1.19 g/mL), 3.5 g Hexamethylenetetramine, and 0.5 L deionized water [23,24].
Where, CR is corrosion rate, mm/y; ΔW is the lost weight of the sample, g; S is the corrosion area of the sample, cm 2 ; ρ is the steel density, 7.85g/cm 3 ; t is the time, h.

Corrosion characterization and water chemistry analysis
The micro morphology of the corrosion products was observed by the JEOL JSM-7200F (SEM), and the OXFORD X-Max50 (EDS) was used to analyze the elemental composition of the corrosion products.Besides, Raman spectroscopy (Renishaw Qontor), with a laser excitation wavelength of 532 nm, was used to analyze the phase composition of the corrosion products.The Thermo Scientific K-Alpha (XPS) was used to analyze the chemical valence states of the corrosion products, which operated at a photon energy of 1486.6 eV and an Al target.The XPS patterns were charge-corrected concerning the C 1s peak binding energy of 284.8 eV.Additionally, OLI Analyzer software was used to analyze aqueous phase chemistry in CO2 pipeline corrosion systems.

Corrosion rate and the macro morphology
Figure 3 presents the corrosion rate (CR) of X65 steel in different phase states of CO2 systems.In the liquid CO2 system, when the H2O content increased from 100 ppmmol to the water-saturated solubility of 3242 ppmmol, the CR of X65 steel increased from 0.0609 mm/y to 0.1883 mm/y.Besides, the CR of X65 steel was only 0.0981 mm/y when the H2O content increased to the water-saturated solubility of 3815 ppmmol in the gaseous CO2 system.Regardless of the system, the CR of X65 steel gradually rised as the H2O content increased.However, it is worth noting that the corrosion degree of X65 steel in the liquid CO2 system was significantly higher than that in the gaseous CO2 system.Figure 4 is the macro morphologies of X65 steel in gaseous and liquid CO2 systems under different H2O content.The surface of X65 steel in the gaseous CO2 system was covered by an incomplete layer of brown corrosion products, and a silver-white substrate was found locally at the H2O content from 100 ppmmol to 1000 ppmmol (Figure 4a1-a3).As the H2O content was over 1000 ppmmol, the corrosion product formed on the steel changed from brown to gray (Figure 4a4-a5).The corrosion product film completely covered the surface of the steel.However, that in the liquid CO2 system was noticeably different.When the H2O content was below 1000 ppmmol, there was a complete brown corrosion product film on the steel surface.As the H2O content reached 1000 ppmmol, the corrosion product changed into taupe.Afterward, the colour of the corrosion products of X65 steel turned to gray with the H2O content increased to saturation.It follows that the composition of corrosion products may change with the development of H2O content from the evolution of the macroscopic morphology on X65 steel surface in two CO2 systems.Additionally, there are differences in the change nodes of corrosion mechanisms under different CO2 systems.Therefore, based on the evolution differences of macro morphology in different CO2 systems, 500 ppmmol, 1000 ppmmol, and saturated water content were selected for further analysis.

H2O content of 500 ppmmol
Figure 5 shows the micro corrosion morphology and cross-section EDS mapping spectra of X65 steel in the two CO2 systems with an H2O content of 500 ppmmol.The corrosion products formed on the surface of X65 steel had similar shapes, mostly spherical (Figure 5a1 and 5b1).Additionally, it can be

Results and discussion
seen from the EDS mapping analysis results of the cross-section (Figure 5a2, 5b2) that the corrosion products were uniform and similar in composition, mainly composed of Fe and O elements, with a small amount of S. From the perspective of corrosion degree, that in liquid CO2 system is higher.The surface of X65 steel was almost covered by spherical corrosion products, and a complete corrosion product film formed.As shown in Figure 6, the corrosion products generated by X65 steel in the gaseous and liquid CO2 systems were similar, and the corresponding positions of the Raman scattering peaks were the same: all at 217, 274, 387, 586 and 1300 cm -1 place.According to previous research, the corresponding corrosion products were mainly Fe2O3/FeOOH [23,25].Additionally, the XPS was used to further analyze the composition of corrosion products.Figure 7 exhibits the XPS spectra of the two systems.Among them, the C 1s, O 1s, and Fe 2p high-resolution XPS spectra of the corrosion products in the two systems are similar.From Figure 7a1 and 7b1, it can be seen that the peaks of Fe 2p3/2 and Fe 2p1/2 are close to the binding energies of 711 eV and 725 eV respectively, which indicates that the Fe element is in an oxidized state [14].As shown in Figure 7a2 and 7b2, the peak at 288.6 eV in the C 1s spectrum corresponds to carbonate and the peak at 284.6 eV is derived from foreign carbon [26].While the O 1s spectra in both systems can be decomposed into three peaks (Figure 7a3 and 7b3), located at 530.1/530.2eV, 531.5 eV, and 532.2 eV binding energies respectively, corresponding to oxyhydroxides, carbonates/oxyhydroxides and sulfates [14,26].However, there is a difference in the S 2p spectra of corrosion products in gaseous and liquid CO2 systems (Figure 7a4 and 7b4).The S 2p spectrum in gaseous CO2 system can be decomposed into three peaks (Figure 7a4): 163.7 eV, 167.1 eV, and 168.7 eV.They respectively correspond to S, sulfite, and sulfate [14].Moreover, that in the liquid CO2 system only has a peak at the binding energy of 168.6 eV, corresponding to sulfate [14].Based on the results of EDS, Raman, and XPS analyses, it can be inferred that the corrosion products formed in the gaseous CO2 system are mainly Fe2O3/FeOOH with small amounts of FeCO3, FeSO3, and FeSO4.While the corrosion products in the liquid CO2 system are mainly Fe2O3/FeOOH and small amounts of FeCO3 and FeSO4.

H2O content of 1000 ppmmol
As shown in Figure 8, there were obvious differences in the micro characteristics of corrosion products on X65 steel under the two systems.Among them, the corrosion products in the gaseous CO2 system were similar to those at 500 ppmmol H2O content.They were all granular corrosion products (Figure 8a1).The cross-sectional EDS mapping of the corrosion products indicates that the main elements of the corrosion products were still Fe, O, and a small amount of S (Figure 8a2).While the corrosion products formed under the liquid CO2 system mainly presented a clay-like deposition form (Figure 8a2).Compared with the gaseous CO2 system at the same H2O content, the thickness of the product film significantly increased in the liquid CO2 system and the intensity of the O and S elements in the corrosion product was enhanced (Figure 8b2).In addition to the changes in morphology and elemental compositions of corrosion products, the increased H2O content also has an influence on the corrosion product types of X65 steel.As seen in Figure 9a, the Raman scattering peaks of corrosion products in gaseous CO2 were basically the same as those in 500 ppmmol H2O content.However, the scattering peaks of corrosion products in the liquid CO2 system changed.The peaks at 713, 975, and 1049cm -1 were also found (Figure 9b).Among them, and 1049cm -1 correspond to FeOOH and ferrihydrites [23], and 975cm -1 corresponds to FeSO4 or FeSO4 hydrates [6].Through XPS spectra of corrosion products in the two systems (Figure 10), it can be found that the peaks of Fe 2p, C 1s, and O 2p were similar to those in 500 ppmmol H2O content, which respectively correspond to the same corrosion product.Nevertheless, there were certain differences in the S 2p spectra under the two systems (Figure 10a4 and b4).The S 2p spectrum under the gaseous CO2 system did not change.While that in the liquid CO2 system was decomposed into four peaks, located at 161.5 eV, 163.6 eV, 167.1 eV, and 167.6 eV, corresponding to FeS, elemental sulfur (S), FeSO3, and FeSO4 respectively [14,26].It follows that the corrosion products remained unchanged in the gaseous CO2 system when the H2O content reached 1000 ppmmol.However, those in the liquid CO2 system changed and mainly consisted of Fe2O3/FeOOH and FeSO4.

Analysis of Corrosion Rate Difference
In gaseous and liquid CO2 systems, the composition of the corrosion products generated by X65 steel was similar.The phase change of CO2 has little impact on the corrosion mechanism of X65 steel.The corrosion process is mainly controlled by impurity components.However, it is worth noting that there was a significant difference in the corrosion degree of X65 steel under the two systems, which is closely related to the change in aqueous phase chemical properties caused by the change in the CO2 phase state.
In the special corrosion system of CO2 transportation environment, whether pipeline steel corrodes depends on whether a free aqueous phase can be formed on the steel surface.Relevant studies have shown that as the H2O content in impurity-containing CO2 environments increased, the corrosion rate of pipeline steel and the thickness of the corrosion film formed also increased accordingly due to the increase in the amount of corrosive aqueous phase that can be formed on the steel surface [14].In gaseous and liquid CO2 systems, the difference in the thickness of the film formed on the X65 steel indirectly indicates the difference in the amount of corrosive aqueous phase that can be deposited on the surface of the steel.Once the aqueous phase is deposited on the steel, CO2 and impurity components are dissolved in the aqueous phase film, which will change the chemical environment of the aqueous phase film, thereby causing corrosion of the steel.
To further analyze the impact of the CO2 phase state on aqueous phase chemistry, the Stream Analyzer module of OLI Analyzer studio software was used to perform water chemistry simulation calculations and analyze the differences between corrosive substances in water-saturated gaseous and liquid CO2 systems.The corrosion medium composition of impurities used in the calculation is listed in Table 2. Besides, the mass of H2O was selected as 0.6 g (approximately the saturated dissolved water amount of gaseous CO2 system in this experiment) and 3.3 g (approximately the saturated dissolved water amount of liquid CO2 system in this experiment).And the mass of CO2 was 2478.00 g (Mass of CO2 in a 3L high-pressure autoclave at 10 MPa-25 ℃) and 313.79 g (Mass of CO2 in a 3L high-pressure autoclave at 5 MPa-50 ℃).The calculation results of aqueous phase chemical analysis are shown in Table 3.It can be seen that compared with the gaseous CO2 system, the liquid CO2 system has a more precipitated water phase, and the corrosive medium contained in the water phase is much higher than that of the gaseous CO2 system.The liquid phase water precipitated from a liquid CO2 system has a lower pH and higher ionic strength.This causes the corrosion environment faced by X65 steel in the liquid CO2 system to be more severe, and its electrochemical corrosion process is more intense.As a result, the corrosion degree of X65 steel in a liquid CO2 system is always higher than that of a gaseous CO2 system.In brief, the reason for the difference in the CR of X65 steel in the gaseous and liquid CO2 systems is closely related to the difference in the amount of aqueous phase formation and the chemical environment of aqueous phase caused by the change in the CO2 phase state.

Conclusion 1.
As the H2O content increases from 100 ppmmol to the saturated solubility, the corrosion rate of X65 steel changes from 0.0178 mm/y to 0.0981 mm/y in a gaseous CO2 system, while that increases from 0.0609 mm/y to 0.01883 mm/y in liquid CO2 system.Under the same H2O content, the corrosion rate and degree of X65 steel in a liquid CO2 system are higher than that in a gaseous CO2 system.2. The corrosion film formation of X65 steel is mainly controlled by the impurities and the reaction products between impurities, independent of the change of the CO2 phase state environment.The main components of the corrosion products on the steel in gaseous and liquid CO2 systems containing the same water content are similar.As the H2O content increases the corrosion products of X65 steel are mainly Fe2O3/FeOOH and gradually transform into an oxide-sulfate mixed film mainly Fe2O3/FeOOH and FeSO4.
3. In gaseous and liquid CO2 systems, O2, H2S, SO2, and NO2 impurities and their interaction promote the formation of the corrosive liquid phase and intensify the corrosion of X65 steel.Compared with a gaseous CO2 system, a liquid CO2 system easily forms more aqueous phase and contains more corrosive media, which causes more serious corrosion hazards to pipeline steel.

Table 2 .
Corrosion test conditions

Table 3 .
Aqueous phase chemistry calculation results by OLI Analyzer Studio software.