Crevice Corrosion of High-Grade Stainless Steels in Seawater: A Comparison Between Temperate and Tropical Locations Open Access

In natural seawater, microorganisms can fix, grow, and develop on practically any surface, including stainless steels.^1-3  The term biofilm is generally used for communities of microorganisms embedded in an organic polymer matrix (e.g., exopolysaccharides), produced by the microorganisms themselves and adhering to a surface, irrespective of the environment in which they develop.⁴  Stainless steels are widely used for different applications in seawater such as the oil and gas, desalination, and marine energy industries. The presence of a biofilm on passive alloys such as stainless steels or nickel-based alloys can strongly enhance the cathodic reactions,⁵  and shift their open-circuit potential (OCP) to the noble direction.^6-7  The resulting increase in OCP is also called cathodic depolarization or ennoblement. The ennoblement in aerated seawater is known to be one of the main factors affecting the risk of localized corrosion of stainless steel and nickel-based alloys, because their critical pitting or crevice potential for protective passive layer breakdown can be exceeded.⁸  Among the mechanisms that could explain the so-called ennoblement, some authors mention the role of microbial enzymes, the production of H₂O₂, and/or low pH inside the biofilm or at the biofilm/stainless steel interface.^9-12  Some recent studies also suggest the role of electroactive bacteria.^13-14  Cathodic depolarization also called electroactive biofilms has been observed in different seas around the world,^7,15-22  including in deep sea.^23-24 

However, a possible drawback of such a setup is that under polarization, the surface pH increases and associated species precipitation may occur, such as CaCO₃ and Mg(OH)₂, changing the expected service conditions.³⁵  Thus, the biofilm growth and activity depend upon the experimental setup which shall be carefully considered in the analysis of the data.

The biofilm-induced cathodic depolarization is known to be temperature dependent. For instance, no ennoblement has been reported in the North Sea at a temperature above 32°C, whereas a temperature limit of about 37°C has been reported in the Bay of Brest (North Atlantic Ocean), where no ennoblement was observed after several months of exposure.^7,19,36  In addition, the kinetic of ennoblement is temperature dependent with higher incubation time at low temperatures.³⁷  In a recent project, the potential ennoblement has been measured in tropical seawaters including Singapore, Saudia, Caribbean Sea, and Brazil.²¹  From these results, it was shown that (a) potential ennoblement occurred for all tested seawaters with rather similar results for all sites at 30°C, and (b) no ennoblement was observed in Brest (temperate seawater) above 37°C after 3 month of exposure while biofilm ennoblement was still measured at 45°C in Singapore. From the results of this study, the temperature and its effect on ennoblement were confirmed to be a major parameter to assess the localized corrosion of passive alloys because above 37°C large differences were highlighted in terms of potential ennoblement between temperate and tropical seawaters. Such temperatures can be easily reached in oil and gas offshore, desalination, and marine energy systems, where heat is produced. Thus, there is an interest to quantify the localized corrosion resistance of materials in heated tropical seawaters as data obtained in temperate seawater may not be representative. Indeed, in tropical seas above 40°C, ennoblement may still be present and consequently the corrosiveness is expected to be higher. However, the safe maximum temperature will also largely depend on the severity of the crevice geometry.³⁸  In the present study, electroactive biofilms and associated corrosion risks were investigated in different seawaters (tropical and temperate) and at different temperatures. The effect of crevice geometry was also studied, including for tube geometries.

The alloys, product forms, and geometries of the tested samples were given in Table 1. The chemical compositions were given in Table 2 together with their pitting-resistant equivalent numbers (PRE_Nw). For all materials, tube or bar geometry has been tested. Hot rolled (HR) plates made of superduplex UNS S32750⁽¹⁾ have been used for biofilm monitoring (cf. biofilm module). For corrosion exposure, UNS S39274, S32707, S33207, N06845, N06625, and S33205 seamless tubes were tested, while UNS S31266 was tested as a bar product. For the superduplex UNS S32750, rolled-welded (RW) tubes have also been tested. The duplex UNS S32205 was tested as low resistant reference for seawater applications. The nickel-based alloys UNS N06625 and N06845 were tested for comparison purposes with the stainless steel materials.

Metallographic inspections of tested alloys were performed. For all of these materials, the expected microstructure was observed (i.e., expected ferrite/austenite balance of 50/50 for duplex and superduplexes) without defects.

The roughness of test specimens (tubes, bars, and plates) was quantified with a 3D-optical profiler system, using the interferometry technique. For all tested tubes (seamless and RW) and rods, the roughness was similar with R_a = 0.3±0.05 µm (measured on 10 replicates per tested alloy). The roughness of all as-received tested plates was 2.2±0.5 µm.

One of the main objectives of the study was to define for each test site (cf. tropical and temperate seawaters) the maximum temperature for which biofilm-enhanced ennoblement on stainless steel is still observed. For this purpose, the OCP of superduplex stainless steel tubes (S32750) were measured at increasing temperatures from the ambient and until ennoblement is not detected anymore (i.e., no ennoblement measured after 2 weeks for each tested temperature). The exposures were performed in nonmetallic exposure tanks of about 300 L containing continuously renewed seawater at a rate of about 300 L/d (i.e., one complete volume of the tank per day), and equipped for temperature regulation with an accuracy of ±1°C. For all systems, regulated Teflon-coated thermo-heater have been used together with continuous stirring (pumps), ensuring a homogenous environment in the exposure tanks. The potentials were measured with AgAgCl/KCl-gel electrodes, calibrated with a saturated calomel electrode (SCE). All results are reported vs. the SCE scale. The potentials were recorded using high-impedance (>10¹¹ Ω) data loggers, suitable for electrochemical measurements.

Biofilm ennoblement was also characterized by cathodic current and corrosion potential measurements of superduplex stainless steel UNS S32750. For this purpose, six superduplex stainless steel plates (150 cm × 100 cm) were connected to a zinc anode through different resistors designed to provide a potential range from OCP to approximately −700 mV_SCE. This method is detailed elsewhere.²³  In such a system (referred to as “biofilm module” in this paper), the biofilm-induced depolarization is detected and quantified from the associated increase of the cathodic current. As mentioned in the Introduction section, this technique appears very interesting to characterize the electrochemical effect of biofilms and to assess at the same time the corrosion risk (regarding the existing correlation between cathodic efficiency and localized corrosion rate). The biofilm modules have been deployed in France (Brest, Atlantic Ocean) and Malaysia (Kelatan/Bachok, Tropical China Meridional Sea).

For each tested configuration (tube, bare, and plate), six replicates were used. For RW specimens, one of the two crevice gaskets was placed on the longitudinal weld to determine if this is a preferential corrosion area.

After exposure, a complete evaluation of the corrosion was performed for each tested specimen including visual evaluations, binocular, and microscopic evaluations of the corrosion using optical microscopy (corroded area, maximum corrosion attacks evaluated with optical focalization connected to micrometric gauge). Metallographic inspections were also performed on selected samples in corroded areas.

The biofilm ennoblement and corrosion investigations were conducted at three selected locations, covering different kinds of seawaters in terms of temperature levels and seasonal variations. The main characteristics of the test sites were measured during the year of exposure are reported in Table 3. France (Brest) and Brazil (Arraial do Cabo) sites are facing season-induced changes of temperature, while the Malaysian site (Kelatan/Bachok), close to the equatorial line, has almost constant temperature during the whole year (around 30°C). Mean salinities ranged from 31% to 36‰, with the lower value for the Malaysian site. The mean pH was quite similar for all test sites, from 8.1 to 8.3. The dissolved oxygen (DO) contents are in good line with saturation values at mean temperatures (i.e., DO around 8 ppm at 12°C and 6 ppm at 30°C).

The three stations were equipped with the same heating regulation systems (regulated Teflon-coated thermo-heater), mounted in similar seawater exposure tanks of about 300 L. In France (Brest marine station), the seawater tanks were also equipped with a cooling system, allowing the performing of some exposure tests at 5°C (to study the kinetics of biofilm formation on a larger range of temperatures). For all sites, the seawater was continuously renewed at a rate of about 300 L/d in exposure tanks. Exposure tests were conducted from ambient temperature to about 55°C in which the following investigations have been performed (all detailed above): (a) OCP vs. temperature (determination of critical temperatures for biofilm ennoblement for all sites), (b) biofilm-induced current measurements as function of temperature, with the used of biofilm module on two sites (France and Malaysia), and (c) crevice corrosion testing of materials listed in Table 2 for all sites.

The most significant difference between natural tropical sites (Malaysia) and Atlantic Ocean sites such as Brest is thus observed when seawater is heated at about 40°C. A comparison of Figures 5(b) and 6(b) shows that at 40°C and at −500 mV_SCE, the biofilm-induced currents are about 50 times higher in this tropical seawater.

Despite the large impact of the ennoblement, no unifying mechanisms have been described as responsible for the phenomenon. In a study published elsewhere¹⁴  we performed DNA analyses of biofilms using last-generation techniques for sequencing and data treatment. The analyses were performed in Brest at different temperatures ranging from 30°C to 40°C, i.e., covering the critical temperature for which ennoblement is observed or not. The strict electrotroph bacterium Candidatus Tenderia electrophaga was clearly detected as an ennoblement biomarker and was only present at temperatures at which ennoblement was observed. Our results suggested that Candidatus Tenderia electrophaga, and its extracellular electron transfer metabolism coupled with oxygen reduction activity, could play a central role in modulating stainless steel OCP and consequently mediating ennoblement.¹⁴  It is suggested that electrotroph bacteria similar to Candidatus Tenderia electrophaga can still be active above 40°C in tropical seawaters where the average ambient temperature is much higher than in temperate seawater. However, for logistical reasons that these analyses were not performed in tropical seawater and investigations in other seawaters would be required to validate the exact role of this electrotroph bacteria in the ennoblement process of stainless steels.

The crevice corrosion results obtained at ambient local temperatures and at critical temperatures for maximal biofilm electroactivity on each test site (determined from the above OCP measurements) are given in Tables 4 and 5, respectively. At local ambient temperatures, the most corrosive site was the Meridional China Sea (Kelatan-Bachok, Malaysia), followed by the South Atlantic Ocean (Araial do Cabo, Brazil). As expected regarding the much cooler averaged ambient temperature, the North Atlantic Ocean (France, Brest) was the less corrosive of these test sites, with crevice corrosion of only duplex stainless steel UNS S32205 tested with the severe crevice configuration (i.e., 20 N/mm²). The results clearly confirm the very significant influence of the crevice geometry on the localized corrosion results.^38,40  Using the lowest tested gasket pressure (i.e., 3 N/mm², which is the one defined in ISO standard 18070:2015) only the duplex UNS S32205 showed crevice corrosion in the two hottest test sites, while most of the tested alloys initiated crevice corrosion using a 20 N/mm² gasket pressure. It shall be noticed that roughness also influences crevice geometries. The R_a of all tested tubes was 0.3 µm, which is known to be much more severe for the crevice corrosion risk of stainless steels than R_a 2 µm to 3 µm (i.e., typical R_a for HR plate products).⁴⁰  This feature must be carefully considered for engineering diagrams constructed from experimental results and to compare the corrosion performance of materials. In the most severe crevice configuration (20 N/mm²) and most severe conditions in terms of temperature regarding the biofilm activity, only the hyperduplex UNS S32707 and the UNS S31266 resisted crevice corrosion (see Table 5). The next more resistant alloy was the hyperduplex UNS S33207, followed by the nickel-based alloys and superduplexes which globally showed rather similar performance at tested sites. The Ni-based alloy UNS N06625 showed better performance than superduplex stainless steels only in the Brest site heated at 30°C. The typical aspect of some corroded samples (UNS S32205, S32750, and N06625) after 12 month of exposure in Brest, Brazil, and Malaysia seawaters, all heated at temperature for their maximal biofilm electroactivity, are given in Figure 7.

In order to confirm that the biofilm-induced depolarization is one of the main factors affecting crevice corrosion risk of passive alloys, corrosion exposures were performed in Brest at 40°C for comparison with the results in Malaysia at 40°C. At this temperature, the biofilm activity (in terms of electrochemical activity) was shown to be very different for the two sites (see Figures 4 through 6). The corrosion results given in Table 5 clearly confirm that localized corrosion risk is not at all the same in the tested temperate and tropical seawaters when tested at 40°C, with almost no corrosion in North Atlantic Ocean.

• In the tested locations, the effect of temperature on biofilm ennoblement is different between the Meridional China Sea (tropical seawater) on one hand and the Atlantic Ocean (North and South). The critical temperature for which biofilm ennoblement is not measured was between 35°C and 40°C in Atlantic Ocean (France-Brest and Brazil-Arraial do Cabo), and between 46°C and 50°C in the Meridional China Sea (Malaysia-Kelatan, tropical seawater).

• No significant differences have been observed between North and South Atlantic Oceans at tested locations in terms of biofilm activity (regarding electrochemical aspects). When heated at 30°C, the North and the South Atlantic Ocean sites showed similar results, with biofilm activity detected during the first day of exposure.

• Among all of the tested sites and temperatures in the study, the most active seawater in terms of biofilm-induced currents was the Meridional China Sea (tropical seawater).

• Above 37°C, the biofilm activity (cathodic current reduction on biofilmed surfaces) is much more pronounced in tropical seawater (Meridional China Sea, Malaysia-Kelatan) compared to Atlantic Ocean sites (France-Brest), with an optimal activity around 40°C in tropical seawater.

• At 30°C, the exposures in the South Atlantic Ocean (Brazil-Araial do Cabo) and Meridional China Sea (Malaysia-Kelatan) were more severe than in the North Atlantic Ocean (France-Brest) for UNS N06625 and the hyperduplex UNS S33207. For all of the other tested alloys, the corrosion results were rather similar for all test sites at 30°C. In Malaysia (tropical site), the corrosion results were rather similar at 30°C and 40°C, with slightly more corrosion occurrence at 40°C.

• In tropical conditions, the high-grade alloys UNS S32707 (hyperduplex 2707) and superaustenitic S31266 showed better performance than commonly used superduplex (S32750 and S39274).

The following sponsors of this project are gratefully acknowledged: Flávia Maciel from PETROBRAS, Thierry Cassagne from TOTAL ENERGIES, Stéphane Trottier from VEOLIA, Lars Mehus from AKER SOLUTIONS, Yves Denos from EDF, Xiaoxue An from TechnipFMC, Viktor Räftegård from VOLVO PENTA, Wenle He & Ulf Kivisäkk from ALLEIMA, Jean-Marc Lardon from ERAMET AUBERT & DUVAL, Sandra Le Manchet from INDUSTEEL, Tadashi Kawakami from NIPPON STEEL, Sophie Delettrez, Luciana Lima, & Jérôme Peultier from VALLOUREC, Valérie Noel from NAVAL GROUP. The test sites are operated respectively by the French Corrosion Institute (Ste Anne du Portzic Marine Station in Brest, France), the Institute of Ocean and Earth Science (Bachok Marine Station, Kelatan, Malaysia), and the Brazilian Navy’s Institute of Marine Studies Admiral Paulo Moreira (IEAPM) Marine Station in Arraial do Cabo, Brazil. A great thanks goes to Ricardo Coutinho and Luciana V. R. de Messano from IEAPM/Brazil, and to Po Teen Lim and Chui Pin Leaw from IOES/Malaysia for their collaboration and help in the project. Marie Roustan and Pascal Moullec, both from Institut de la Corrosion, are gratefully acknowledged for their support in the experimental setups.