Rio Oil & Gas Expo and Conference 2020 ISSN 2525-7579 Conference Proceedings homepage: https://biblioteca.ibp.org.br/riooilegas/en/ Technical Paper Evaluation of the SCC-CO2 susceptibility of flexible pipe tensile armor wires Fabio Santos 1 Ingrid Poloponsky 2 Susana Modiano 3 Carlos Ribeiro 4 Eduardo Motta 5. 1. BAKER HUGHES, 2. BAKER HUGHES, 3. BAKER HUGHES, 4. BAKER HUGHES, 5. BAKER HUGHES, ENGENHARIA , ENGENHARIA DE MATERIAIS. RIO DE JANEIRO - RJ - BRASIL, fabio.santos@bakerhughes.com ENGENHARIA , ENGENHARIA DE MATERIAIS. RIO DE JANEIRO - RJ - BRASIL, ingrid.poloponsky@bakerhughes.com ENGENHARIA , ENGENHARIA DE MATERIAIS. RIO DE JANEIRO - RJ - BRASIL, susana.modiano@bakehughes.com ENGENHARIA , ENGENHARIA DE MATERIAIS. RIO DE JANEIRO - RJ - BRASIL, carlos.ribeiro@bakerhughes.com ENGENHARIA , ENGENHARIA DE MATERIAIS. RIO DE JANEIRO - RJ - BRASIL, eduardo.motta@bakehughes.com Abstract As part of Baker Hughes response to a flexible pipe failure caused by SCC-CO2 disclosed by a safety report issued by ANP in 2017, an extensive material testing program has been established to assess the susceptibility of the tensile armor wires to SCC triggered by the presence of CO2.The aim of the test program is to determine both stress and CO2 thresholds where the failure mechanism by SCC-CO2 could be triggered, creating a SCC risk free region to be used for design purposes.The test program is based on wires taken from manufactured pipes which have experienced the complete strain history of manufacturing and present representative residual stresses. The small-scale test protocol includes long term SCC, Corrosion and SRR tests carried out in virgin and pre exposed wires taken from a 12 months full scale corrosion test. This paper will present the results of tensile armor wire test campaign, discuss its findings and present the full-scale validation tests to be executed by Baker Hughes. Keywords: Flexible Pipe. Corrosion. Stress. SCC-CO2. Tensile Armor Received: February 27, 2020 | Accepted: Jun 06, 2020 | Available online: Dec 01, 2020 Article Code: 177 Cite as: Rio Oil & Gas Expo and Conference, Rio de Janeiro, RJ, Brazil, 2020 (20) DOI: https://doi.org/10.48072/2525-7579.rog.2020.177 © Copyright 2020. Brazilian Petroleum, Gas and Biofuels Institute - IBP. This Technical Paper was prepared for presentation at the Rio Oil & Gas Expo and Conference 2020, held between 21 and 24 of September 2020, in Rio de Janeiro. This Technical Paper was selected for presentation by the Technical Committee of the event according to the information contained in the final paper submitted by the author(s).The organizers are not supposed to translate or correct the submitted papers. The material as it is presented, does not necessarily represent Brazilian Petroleum, Gas and Biofuels Institute’ opinion, or that of its Members or Representatives. Authors consent to the publication of this Technical Paper in the Rio Oil & Gas Expo and Conference 2020 Proceedings. Evaluation of the SCC-CO2 susceptibility of flexible pipe tensile armor wires 1. Introduction Due to recent failures reported in the Pre-salt area by Oil & Gas operators, the corrosion mechanism identified as Stress Corrosion Cracking (SCC) due to CO2 presence (SCC-CO2) became a focus of study related to flexible pipes. The SCC-CO2 mechanism occurrence on armour wires of gas injection flexible pipes was implied in catastrophic failure of pipe in a short time period considering the service life for which the equipment was designed. As flexible pipes are widely applied for oil and gas transportation, especially for the ultra deep water applications of Brazilian Pre-salt, the concern with the integrity of installed pipes and with new flexible pipes design resulted in a large study around such SCC-CO2 corrosion mechanisms. The investigations identify the main parameters leading to its occurrence, characterization of cracks, and replication of the failure mechanisms in a laboratory. Different test methods have been used to evaluate the susceptibility of tensile armour materials to the SCC-CO2 corrosion mechanism. This paper presents the main results and observations obtained for representative material grades, comparing behaviors and delimiting envelopes to ensure safe operation under flooded annulus combined with service conditions where CO2 is present. 2. SCC-CO2 corrosion mechanism overview The Stress Corrosion Cracking mechanism is a type of Environmental Assited Cracking which occurs due to the interaction of three factors: corrosive environment; material susceptibility and total stress as explained in Hanninnen (2003). SCC cracks nucleation and growth is a results of the balance between active stress and electrochemical reactions. The stresses required to trigger SCC-CO2 occurrence are tensile in nature and below the material’s yield, both applied load and residual stress contribute to crack appearance. Crack growth is perpendicular to primary loading direction, depending on the balance in these driving forces environment/material cracks can be characterized as intergranular, transgranular or present a mixed mode as observed by Newman (2010). The presence of multiple cracks and cracks branching are common characteristics of SCC occurrence. Jones (2017) explains that SCC occurrence is generally associated with an environment/material combination that results in a corrosion film formation on metals surface leading to a semi passive state, which reduces uniform corrosion rate but can result in localized corrosion and a preferential site for SCC cracks nucleation. Thus, an SCC fracture surface presents characteristics of brittle behavior indicating little or no material loss associated with the failure. Newman (2010) states that it is not always clear if crack propagation occurs only due to anodic dissolution, per hydrogen embrittlement or due other mechanism influence. In a flexible pipe the annulus region is determined by the space between pressure barrier layer and external outer sheath, both polymeric layers. The pressure and tensile armour layers manufactured in carbon steel are located in the annulus region as can be seen in Figure 1. During oil and gas transmission, contaminant gases presented in fluid composition, like CO2 and H2S, permeate through the inner pressure barrier reaching annulus region. At the same time water ingress into annulus region may occur due to water vapor permeation also from pipe bore, due to a damage on outer sheath layer or even as a result of seal failure in an end fitting port. The association of contaminant gases and water transforms the once dry annulus region into a corrosive environment for the layers of metallic pressure and tensile armour wires. The Brazilian Pre-salt reservoirs present high content of CO2, and as a result a new failure mode for flexible pipe armours was observed in operation. After gas injection pipes tensile armour wires rupture, the failure analyses carried out verified that wire breaking occurred due to Stress Corrosion ________________________________________________________________________________________________ Rio Oil & Gas Expo And Conference, 2020. | ISSN 2525-7579 2 Fabio Santos, Ingrid Poloponsky, Susana Modiano, Carlos Ribeiro, Eduardo Motta Cracking due to CO2 presence (SCC-CO2). Figure 2 shows a representative image of SCC-CO2 cracks initiating from locallised corrosion in laboratory replication experiments. Figure 1 – Flexible pipe annulus region.. Source: Baker Hughes. Figura 2 –SCC-CO2 cracks generated in laboratory. Source: Baker Hughes. In the specific case of the SCC-CO2 mechanism observed in carbon steel wires, the unique environment conditions of annulus region results in the formation of a porous and non-protective iron carbonate film (FeCO3). Ferreira, Klok, Ponte and Farelas (2016) explain that iron carbonate film precipitation and morphology are directly related to Fe2+ iron íons supersaturation, associated with a CO2 partial pressure which exceeds the solubility limit of CO32- resulting in a supersaturation of this specie. Annulus region of a flexible pipe exhibits low confinement ratio, i.e. the relationship between free volume and steel exposed surface area (ml/cm²), resulting in iron ions supersaturation. Ferreira et al. (2016) add that parameters like temperature and CO2 partial pressure have an important role on iron carbonate film formation. CO2 solubility in aqueous solution together with carbonic acid ionization followed by pH reduction are directly related to the temperature. The CO 2 partial pressure influences other environment parameters like pH and íons concentration in aqueous solution, which affects the precipitation of corrosion products. Based on this literature review, it is clear that the SCC-CO2 mechanism is a result of the interaction of different parameters and that the threshould for each parameter and material shall be determined to avoid cracking occurrence. In this way, an extensive test campaign including different tests configuration and environment parameters is being conducted by Baker Hughes aiming to evaluate tensile armour wires material susceptibility to the SCC-CO2 mechanism and establish a safe envelope for flexible pipe operation. The material strength, temperature, CO2 pressure, applied load and residual stress, and iron íons saturation among other parameters are part of this evaluation, to ________________________________________________________________________________________________ Rio Oil & Gas Expo And Conference, 2020. | ISSN 2525-7579 3 Evaluation of the SCC-CO2 susceptibility of flexible pipe tensile armor wires provide reliable data for a better understanding of the boundaries for the SCC-CO2 corrosion mechanism occurrence. 3. Materials and methods Test methodologies for SCC evaluation are intended to determine the susceptibility of a certain material in a specific environment providing information for a proper materials selection in accordance with required service. The test conditions shall be representative of the worst case scenario predicted for service operation. SCC tests can be static or dynamic. The test campaign presented in this document considers two small scale test methods: constant displacement in four-point bending configuration (4PB) and slow strain rate testing (SSRT). Two carbon steel materials commonly used as tensile armour layers for Flexible pipes in Brazilian pre-salt fields were tested in different temperatures, pressures and strain for SCC-CO2 suscepbility. The test campaign included evaluation of samples from manufacturing process to consider the influence of residual stress and accumulated inelastic strain history. 3.1. Materials The two tensile armour wires materials investigated in this study are carbon steel grades with high strength, being both with UTS higher than 1000MPa and one heat treated. Steel A is used for sweet applications, i. e., with low H2S content or without H2S. Steel B is a material with lower mechanical strength but with better corrosion resistance used for H2S content up to partial pressures of 10mbar in the annulus region. The tensile armour wire full cross section was tested, in this way no machining was performed in the test specimens. Samples from vendor coils and taken from flexible pipe were used. 3.2. Constant displacement tests using 4PB devices There are several configurations for SCC evaluation under static conditions, in this investigation the test method chosen uses a four point bending device to apply an specific deformation in tensile armour wire specimens, as presented in Figure 3. The applied strain represents a loading state equal or higher than material yield at the outer bend surface of the samples. Various exposure periods were considered for SCC-CO2 evaluation, from 90 days up to 180 days. Compared with common corrosion tests, 180 days is a long exposure period but it is understood that the SCC-CO2 corrosion mechanism occurs as a function of slower electrochemical processes when compared with other failure mechanisms such as the ones related to H2S dissolution. Figure 3 – Four point bending test device Source: produced by the author. ________________________________________________________________________________________________ Rio Oil & Gas Expo And Conference, 2020. | ISSN 2525-7579 4 Fabio Santos, Ingrid Poloponsky, Susana Modiano, Carlos Ribeiro, Eduardo Motta A total amount of three samples per each test setup were tested using armour samples retrieved from an actual flexible pipe to consider the accumulated inelastic manufacturing strain level. Different temperatures were tested but for the purpose of this paper only the results for 40°C will be reported considering that this temperature was identified as being critical for SCC-CO2 corrosion mechanism occurrence. Tables 2 and 3 present the test matrix for Steels A and B, reespectively. Table 2 – Constant displacement test matrix for Steel A Material Internal Setup # pCO2 (bara) Temperature Duration (°C) (days) Samples origin Steel A 8 30 40 180 From manufacturing (formed) Steel A 10 10 40 180 From manufacturing (formed) Steel A 16 30 40 90 From manufacturing (formed) Steel A 22 10 40 90 From manufacturing (formed) Steel A 27 10 40 90 From manufacturing (formed) Steel A 28 10 40 180 From manufacturing (formed) Steel A 30 5 40 180 From manufacturing (formed) Steel A 31 5 40 90 From manufacturing (formed) Steel A 40 5 40 90 From manufacturing (formed) Steel A 41 10 40 90 From manufacturing (formed) Steel A 42 10 40 90 From manufacturing (formed) Steel A 43 62 40 90 From manufacturing (formed) Steel A 48 10 40 90 From manufacturing (formed) Steel A 49 10 40 90 From manufacturing (formed) Steel A 67 10 40 180 From manufacturing (formed) Steel A 70 30 40 180 From manufacturing (formed) Source: produced by the author. After samples loading, the 4PB device is placed in the test vessel; in sequence this system is filled with test solution and deaerated with nitrogen for 2 hours. Test solution considered on all setups consists of synthetic sea water in accordance with ASTM D1141. In addition, steel microspheres were used in order to achieve iron ions supersaturation. Before increasing the pressure, the solution was bubbled with CO2 for 15 minutes. Subsequently pressure was increased to achieve the required pressure for each test setup. ________________________________________________________________________________________________ Rio Oil & Gas Expo And Conference, 2020. | ISSN 2525-7579 5 Evaluation of the SCC-CO2 susceptibility of flexible pipe tensile armor wires Table 3 – Constant displacement test matrix for Steel B Material Internal Setup # pCO2 (bara) Temperature Duration (°C) (days) Samples origin Steel B 26 80 40 180 From manufacturing (formed) Steel B 33 10 40 180 From manufacturing (formed) Steel B 34 5 40 180 From manufacturing (formed) Steel B 35 5 40 90 From manufacturing (formed) Steel B 37 10 40 90 From manufacturing (formed) Steel B 38 30 40 180 From manufacturing (formed) Steel B 39 50 40 180 From manufacturing (formed) Steel B 55 10 40 180 From manufacturing (formed) Steel B 56 50 40 180 From manufacturing (formed) Steel B 57 5 40 180 From manufacturing (formed) Steel B 58 30 40 180 From manufacturing (formed) Steel B 59 10 40 180 From manufacturing (formed) Steel B 61 5 40 180 From manufacturing (formed) Steel B 62 30 40 180 From manufacturing (formed) Source: produced by the author. 3.3. Slow strain rate testing (SSRT) The slow strain rate test (SSRT) is used to explore potential cracking mechanisms by application of a slow strain rate to failure thus allowing a chance for SCC mechanisms to develop, Henthorne (2016). Generally, this methodology is widely used to study all alloy systems known to be susceptible to SCC, providing a method to evaluate the influence of various parameters. In contrast to fixed displacement/load tests, it may provide a faster way to evaluate the most critical parameters affecting the SCC-CO2. Test execution is similar to a standard uniaxial tensile test, except for the adopted strain rate of 1 x 10-6. SSRT shall always be conducted until fracture of test specimen. Figure 4 presents a device specially manufactured for SSRT fixed in an electromechanical universal test machine . The SSRT method enables the comparison of several CO2 levels criticality and also the comparison and ranking of materials in a given environment. A total of four CO2 levels were considered to evaluate materials behavior together with one control environment, that consists in reference environment where the material under evaluation is not susceptible to SCC. This work focus on the results for 40°C considering that this temperature was identified as being critical for SCC-CO2 corrosion mechanism occurrence. The test environment consists of synthetic seawater prepared in ________________________________________________________________________________________________ Rio Oil & Gas Expo And Conference, 2020. | ISSN 2525-7579 6 Fabio Santos, Ingrid Poloponsky, Susana Modiano, Carlos Ribeiro, Eduardo Motta accordance with ASTM D1141 with CO2 (99,9900% of purity) under different pressures as presented in Table 4. Figure 4 – Slow strain rate test device Source: produced by the author. Table 4 – Slow strain rate test matrix Material Setup ppCO2 (bara) Temperature Samples origin (°C) Steel A 7 Control environment 40 vendor coil Steel A 8 80 40 vendor coil Steel A 9 10 40 vendor coil Steel A 10 5 40 vendor coil Steel A 12 65 40 vendor coil Steel B 1 Control environment 40 vendor coil Steel B 2 5 40 vendor coil Steel B 3 10 40 vendor coil Steel B 4 80 40 vendor coil Steel B 6 65 40 vendor coil Source: produced by the author. ________________________________________________________________________________________________ Rio Oil & Gas Expo And Conference, 2020. | ISSN 2525-7579 7 Evaluation of the SCC-CO2 susceptibility of flexible pipe tensile armor wires 4. Results After the conclusion of 4PB test exposure period, a series of post-tests were performed to verify SCC-CO2 cracks occurrence, these being: magnetic particle inspection via wet florescent technique; stereoscopic optical evaluation was performed after testing before and after specimens cleaning and microscope evaluation. In the SSRT, material susceptibility to SCC is primarily evaluated by direct comparison of the stress on the load-crosshead displacement plot at which the test environment curve deviates from the control environment curve. Additionally, SSRT evaluation is also quantified using the ratio between obtained result in test environment and the results in control environment. When the calculated SSRT ratios decrease in value from unity material susceptibility to the SCC mechanism is verified. As the closer the SSRT ratios are from unity, the better is the material resistance to SCC. 4.1. Steel A – 4PB This section summarizes the main findings of the post-tests evaluations performed for Steel A. At first, all samples were investigated to verify SCC-CO2 surface cracks using wet fluorescent magnetic particle technique (MPI). All positive indications positions were marked and observed via metallographic techniques in order to confirm the presence of SCC-CO2 cracking. SCC cracking consistent evidences were found on the three samples of setup 22 (10bar/ 40°C/ 90days). It shall be highlighted that similar test environment with longer exposure time, setup 10 (10bar/ 40°C/ 180days) has not developed the same pattern of indications. Figure 5(a) presents one MPI result for setup 22, linear indications along the wire width can be observed. Figure 5(b) presents the MPI evaluation for Setup 10, showing no indications. Figure 5 – (a) setup 22 SCC-CO2 indication; (b) setup 10 without indication (a) Source: Baker Hughes property. (b) After magnetic particle inspections, stereoscopic optical evaluation was performed in the region between internal rollers (the test-span region) and outside this region before and after cleaning for removing scale formation; images from one of the samples with SCC-CO2 cracks can be seen in Figure 6. Following stereoscopic optical evaluation, samples were cut for defect characterization using a microscope. Figure 7 presents the images obtained for setup 22 where all the samples presented SCCCO2 cracks with the same pattern. It is important to highlight that these cracks were observed in samples from manufaturing process and for a loading equivalent to the material’s yield. ________________________________________________________________________________________________ Rio Oil & Gas Expo And Conference, 2020. | ISSN 2525-7579 8 Fabio Santos, Ingrid Poloponsky, Susana Modiano, Carlos Ribeiro, Eduardo Motta Figure 6 – Setup 22 stereoscope evalution (a) before cleaning ; (b) after cleaning (a) Source: Baker Hughes property. (b) Figure 7 – Microscope evaluation (a) setup 22 ; (b) setup 30 Source: Baker Hughes property. 4.2. Steel B – 4PB For Steel B, no SCC-CO2 cracking evidence or fracture was found for any tested setups, unlike the material Steel A. MPI results showed one indication for test setup 56 and two indications for setup 59, all the other specimens presented no indication of cracks or surface flaws. In the first instance, samples surface was evaluated before cleaning using a stereoscope, after that all samples were cleaned and subjected to a second evaluation in a stereoscope to confirm the presence of SCC-CO2 surface cracks. It was verified that none of the tested samples presented surface cracks, Figure 8 presents one representative image of these evaluation before and after cleaning. For this material grade only localized corrosion in the form of pitting was found; for all tested setups no SCC-CO2 cracking was observed, even considering that the material was subjected to loading equivalent to the material’s yield. ________________________________________________________________________________________________ Rio Oil & Gas Expo And Conference, 2020. | ISSN 2525-7579 9 Evaluation of the SCC-CO2 susceptibility of flexible pipe tensile armor wires Figure 8 – Setup 57 stereoscope evalution (a) before cleaning ; (b) after cleaning Source: Baker Hughes property. 4.3. Materials direct comparison – Slow stress rate test (SSRT) In general, for both materials it was clear from the stress versus load-crosshead displacement plot obtained for SSRT that the increase of CO2 pressure implies in a decrease of material properties. However, this effect is only observed in the plastic region of the curve, i. e. the amount of plastic deformation withstood by the material is reduced as a function of CO2 level. Figure 9 shows the time to failure ratios for each CO2 pressure considering both materials evaluated. It is possible to observe the SSRT ratio decreases from unity due to increase in CO 2 pressure and this effect is more accentuated for Steel A for which the ratio at 80bar of CO 2 reaches 0.78. For Steel B, a lighter reduction was observed. Together with ratios evaluation, it is important to verify secondary cracking since its occurrence can be an indicator of susceptibility to SCC. As highlighted per Henthorne (2016), even a high SSRT ratio might be critical if secondary cracking is confirmed; on the other hand, lack of SCC crack morphology and absence of secondary crack might imply acceptability of a lower ratio. There are cases where corrosion mechanisms cause ductility and toughness losses without SCC occurrence. These investigations are currently being performed in the tested samples, after that it will be possible to confirm if the ratio reduction observed is due to SCC-CO2. The aim here is to compare both tensile armour materials behavior and susceptibility, to understand the strength influence. Figure 9 – Time to failure ratios versus CO2 pressure Source: produced by the author. ________________________________________________________________________________________________ Rio Oil & Gas Expo And Conference, 2020. | ISSN 2525-7579 10 Fabio Santos, Ingrid Poloponsky, Susana Modiano, Carlos Ribeiro, Eduardo Motta 5. Conclusions Two tensile armours wires materials susceptibility were investigated adopting complementary test methodologies. The results presented herein provide information to establish an initial safe region for flexible pipe operation considering conveyed fluid composition and operational parameters. Constant displacement tests results show a difference of material susceptibility when comparing Steel A and Steel B; the material with lower strength presented better results with absence of SCCCO2 cracking for all 14 tested setups even considering conditions as severe as 80bar of CO2 pressure and straining up to material’s yield. For Steel A, 16 setups were concluded and cracks were found in one of them, identified as setup 22. The test methodology itself was proved to be acceptable to investigate the SCC-CO2 corrosion mechanism since one of the setups presented consistent cracking for all specimens. It is important to highlight that cracks observed on Steel A on setup 22 were generated for a stress condition equivalent to material’s yield, this loading state being far more aggressive than any service scenario even for extreme conditions. Additionally, considering that a total of nine setups were completed with the same strain conditions and environment parameters, 10bar at 40°C, and only in one setup SCC-CO2 cracking was triggered, it is assumed that this should be a material susceptibility boundary. Residual stress and accumulated deformation history might also play a role on SCC-CO2 cracks triggering, and both are being addressed by measuring samples’ residual stress and mapping tensile armour layer manufacturing process. These two variables alone may explain why SCC-CO2 cracks are not observed for all manufacturers. The results obtained in SSR testing reinforces the main findings of 4PB tests. It is clear that material identified as Steel A is more affected per environment, i. e. CO2 pressure. SCC-CO2 occurrence in this test should be confirmed after fracture surface observation and secondary cracking confirmation. However, the direct comparison provided for the materials in this study is considered enough to support the 4PB test campaign observations. Additionally, the differences in the sample mechanical properties subjected to different CO2 levels are only perceived after the material’s yield stress, i.e., in the inelastic domain. It is important to enphasises that the entire Flexible pipe project is made within the elastic domain, thereby CO2 level appears to do not imply a consistent influence in the pipe structural integrity capability for the two tensile armour materials discussed herein.. For higher levels of CO2 no cracks were observed and the corrosion morphology was for large and round pits. Pitting is also associated with the mechanical partial removal of previously formed scale and chemical dissolution of exposed areas by solutions undersaturated with dissolved Fe2+. It is understood that there are still uncertainties related to the SCC-CO2 failure mechanism occurrence. However, the program summarized herein addresses most of them and is being updated continuously based on the main findings. All the different test methods adopted together with the wide range of variables considered, should cover and be able to establish a definitive safe area for the flexible pipe operation. ________________________________________________________________________________________________ Rio Oil & Gas Expo And Conference, 2020. | ISSN 2525-7579 11