Localized Corrosion of High-Grade Stainless Steels: Grade Selection in Chlorinated Seawater

It is generally admitted that a residual chlorine content of approximately 0.2 ppm can prevent the formation of biofilm and reduce the corrosion propagation compared to nonchlorinated seawater.^10-11  However, a control of 0.2 ppm of residual chlorine is hardly controlled in field operations which generally face values from 0.5 ppm to 1.0 ppm,¹²  increasing the risk of localized corrosion initiation on stainless steels.¹¹  The highest chlorine level which can be used depends on the stainless steel grade and on the temperature. NORSOK Standard M-001 rev. 5 (Norwegian petroleum industry standard¹³ ) suggests an upper temperature limit of 20°C for super-duplex (e.g., UNS S32750⁽¹⁾) and 6% Mo alloys (e.g., UNS S31254) with tight crevices (such as screw couplings) with 1.5 mg/L chlorine. Other experiences with these alloys show that the maximum temperature largely depends on chlorine concentration.^12,14  Above 40°C it was shown that even the highest alloyed stainless steels (i.e., with pitting resistant equivalent numbers, PRE_N, around 50) can initiate crevice corrosion with residual chlorine of 2 ppm and above.¹⁴  In the same reference it was shown that at 50°C, a chlorination at 0.5 ppm was sufficient to initiate crevice corrosion on high-grade stainless steels when tested under severe crevice configuration. At ambient temperature, the use of very high chlorine concentrations that can be encountered in biocide injection point (e.g., 500 ppm) was shown to promote crevice corrosion of super-duplex stainless steel.⁷  Even if data can be found, most results from the literature on localized corrosion of stainless steels in chlorinated seawater are limited in terms of tested free chlorine levels or tested number of alloys. Also, it shall be underlined that feedback from field experience frequently does not match the results of laboratory simulation. In particular, the successful use of super-duplex stainless steel is often reported in field use at relatively high residual chlorine content in natural seawater. For instance, Francis and Byrne¹²  recommend the following maximum chlorine contents for field use of super-duplex UNS S32760 in seawater: 200 ppm at 10°C, 5.0 ppm at 20°C, 1.0 ppm at 30°C, and 0.7 ppm at 40°C. Such differences between some field experiences and simulated environments may be attributed to differences in actual crevice geometries¹²  and in the metallurgical quality of the products. In addition, most corrosion results from laboratory studies are obtained from short-term laboratory electrochemical testing, not always representative of service conditions, and not always comparable to each other as they were not tested using the same geometry (crevice former, surface roughness, etc.). In the present study, a wide range of residual chloride contents was tested using mid to long-term exposure durations, involving several stainless steels in different simulated service conditions and tested using similar crevice geometries. The focus was made on severe crevice configurations to get expected conservative results. In addition to crevice corrosion coupons, specific service condition of tube heat exchangers was simulated using controlled heat flux to study the internal localized corrosion resistance of tubes. In parallel, the corrosion risk (both initiation and propagation) has been studied using a module initially developed to quantify the formation of electroactive biofilms.⁶  It was then used to quantify the cathodic depolarization as a function of residual chlorine, which can be correlated to the corrosion propagation rate. The aim of this study was thus to provide new engineering diagrams to make a risk assessment. This may help the end users to select adapted stainless steel grade for their applications in chlorinated seawater. Welded materials are not discussed in this paper.

The tested alloys, product form, and chemical compositions were given in Table 1 together with their PRE_N.¹⁵  The duplex UNS S32205 was used as lower alloying reference material to control the actual severity of test conditions. The duplex S32205 and super-duplex S32750 were tested with hot rolled plates and seamless tube geometries. The hyper duplexes S32707 and S33207 were tested as seamless tubes, while UNS S31266 was tested with hot rolled plates and bar geometries. All of the tested seamless tubes are solution-annealed final tube products. The PRE_N of the tested super-duplex UNS S32750 was 42 to 43, while hyper-duplexes S32707-S33207 and the high-grade austenitic UNS S31266 had PRE_N from 49 to 54. The microstructure was checked for all tested alloys and no defects were observed, with the expected ferrite/austenite balanced at 50/50. 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 and rods, the as-received surface roughness was similar with Ra = 0.3±0.05 µm (measured on 10 replicates per tested alloy). The surface roughness of all as-received tested plates was 2.2±0.5 µm. For crevice corrosion testing, the tested area below crevice formers was polished with SiC paper P1200 to get Ra of 0.3±0.1 µm, similar to the unpolished tested tube roughness (for comparison purpose).

All of the exposure tests were performed in aerated natural seawater, continuously pumped from the bay of Brest, and showing standard characteristics of the Atlantic Ocean with pH 8.0 and salinity of 34±1‰. The desired residual chlorine contents were obtained with electrolysis of seawater between two mixed-metal-oxide-coated titanium electrodes, connected to a regulated power supply (electrolyzer). The continuous regulation was performed from redox potential measurements which were checked bi-weekly with manual measurements using a photometer. Exposure tanks were equipped with a regulated Teflon-coated thermo-heater and a continuous stirring was operated to ensure a homogenous environment in the exposure tanks at ±1°C. The temperature was continuously measured during exposures with Pt probes. The OCPs were measured using high input impedance (>10¹¹ Ω) data loggers, connected to AgAgCl/KCl-3M-gel electrodes. These electrodes were weekly checked with a certified saturated calomel electrode (SCE). All results are reported vs. the SCE scale.

The effect of residual chlorine level and temperature on the OCP of stainless steel have been investigated in aerated seawater. All coupons were exposed in nonmetallic exposure tanks of about 300 L, continuously renewed with seawater at a rate of about 300 L/d (i.e., one complete volume of the tank per day). Samples have been exposed for a minimum of 3 month and until stabilization of the OCP, arbitrarily considered when no variation of ±10 mV was measured over 48 h. A minimum of five replicates per tested conditions have been exposed to calculate standard deviations. The effect of chlorination level on cathodic polarization was investigated on super-duplex stainless steel UNS S32750 with the use of plate coupons (150 cm × 100 cm) connected to a zinc anode through different resistors. The system was designed to provide a potential range from OCP (i.e., no connection with anodic material) to approximately −700 mV_SCE, and the associated current was continuously measured. This method is detailed elsewhere¹⁶  and referred as “biofilm module” in this paper. It was initially successfully used to quantify the biofilm-induced cathodic depolarization.⁶  This technique appeared adapted to assess the efficiency of the biocide effect of chlorination (i.e., effect of residual chlorine level on the suppression of the biofilm ennoblement) and to quantify the effect of residual chlorine content on cathodic (de)polarization. In this study, the biofilm modules have been exposed in duplicate in natural seawater at 30°C, at chlorination levels from 0 ppm to 4 ppm.

All test conditions were collected in Table 3 together with the tested alloys and test durations. The selected heat exchanges were tested in duplicates. After exposure, a destructive method was used for visual, microscopic, and metallographic evaluations of localized corrosion on IDs by cutting and cross-section observations in the transverse directions.

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. Metallographic inspections were also performed on selected samples at corroded areas. The microstructure of super- and hyper-duplex stainless steels was reviewed on cross sections in the transverse direction after etching in 7 N NaOH with 2.5 V for 1 s to 5 s.

The internal tube pitting corrosion evaluations from the model tube heat exchangers tested in a controlled flow loop are given in Table 4. Details about the tested heat fluxes were provided in the experimental part. The results confirmed that duplex stainless steel UNS S32205 tested as a lower alloying reference material, cannot be used in chlorinated seawater under these conditions. The super-duplex UNS S32750 showed a restricted limit of use with pitting corrosion resistance for the heat 50/35 and pitting corrosion occurred for the heat 70/35. For all tested heat fluxes, the hyper-duplex tube UNS S32707 showed high pitting corrosion resistance, extending the limits of use for stainless steels in these demanding applications. This is consistent with results from the ranking test per ASTM G48,²⁹  indicating a critical pitting temperature of 80°C for UNS S32750 and >97°C for UNS S32707,³⁰  although it should be kept in mind that the test environment of this ranking test in 6 wt% FeCl₃ is not representative of seawater applications.

The crevice corrosion results obtained from adapted ISO 18070:2015 crevice former at the severe gasket condition of 20 N/mm² are given in Table 5. Whatever the tested product geometry (plate, tube, and bar) the use of similar tested surfaces, similar gasket pressure, and similar roughness (0.3 µm) allows a direct comparison for all testing conditions. The low alloying reference material UNS S32205 was confirmed to be highly sensitive to crevice corrosion in 0.5 ppm chlorinated seawater, even at 20°C. For the super-duplex UNS S32750 it was observed that under severe crevice configurations, the limits of safe use in chlorinated seawater were mainly affected by the temperature. For the simulated severe crevice configuration, the limits of safe use of UNS S32750 in 0.5 ppm to 15 ppm of residual chlorine was 20°C. This temperature limit is in good line with Norsok recommendations¹³  given for natural seawater. However, it shall be mentioned that some field experience indicates upper limits of use of super-duplex stainless steels³⁰  and this can be attributed to the significant effect of crevice geometry (gasket nature, surface roughness, etc.) on the crevice corrosion risk³⁰  and/or to the exact metallurgy of tested products. Table 5 shows that all tested high-grade alloys (S31266, S32707, and S33207) have higher limits of safe application compared to the super-duplex UNS S32750. For the tested conditions, similar results have been obtained for S31266, S32707, and S33207 in terms of the risk of initiation of crevice corrosion. For these three alloys, a synergistic effect of residual chlorine and temperature was observed, with the following critical conditions: 40°C for chlorine contents from 2 ppm to 15 ppm, 45°C at 1 ppm chlorine content, and 50°C at 0.5 ppm chlorine content. When the temperature decreased to 35°C, no crevice corrosion initiated from 0.5 ppm to 15 ppm of chlorine content, suggesting safe conditions below 35°C in the range of residual chlorine.

The electrochemical effect of residual chlorine content in natural seawater (Atlantic Ocean, Brest, France) has been investigated on high-grade stainless steels base materials UNS S31266, S32707, and S33207 at different temperatures and compared to UNS S32750. UNS S32205 was included as a lower alloying reference material. Localized corrosion exposure tests have been performed, simulating different applications (heat exchangers, controlled crevice configuration) and involving different product shapes (plates, tubes, and bars). The main results are given below:

• The OCP of corrosion-resistant stainless steels is increasing with the chlorine residual content with a maximum plateau of about +900 mV_SCE with 30 ppm at 30°C (very high residual chlorine content that might be encountered near biocide injection points). At 30°C, the OCPs are around +600±50 mV_SCE in seawater chlorinated between 0.5 ppm and 2 ppm.

• At a given chlorine content, the OCP of high-grade stainless steels increases with temperature, indicating a synergistic effect of residual chlorine and temperature. UNS S31266, S32707, and S33207 show extended environmental limits for use in chlorinated seawater compared to super-duplex UNS S32750.

• For UNS S32750 the most important parameter for crevice corrosion in chlorinated seawater (from 0.5 ppm to 15 ppm) was the temperature, with no corrosion at 20°C and crevice corrosion at 30°C when tested with the simulated severe crevice configuration. The difference with some field back indicating higher limits of use can be attributed to the differences in crevice geometries.

• For UNS S31266, S32707, and S33207, the environmental limits for the safe use of these alloys in chlorinated seawater with the presence of severe crevice geometry (20 N/mm²) are ≤15 ppm chlorine at 35°C, ≤1 ppm chlorine at 40°C, and ≤0.5 ppm chlorine at 45°C.

• For UNS S31266, S32707, and S33207, no crevice corrosion was initiated below 35°C in seawater chlorinated from 0.5 ppm to 15 ppm of chlorine content.

• In the tested conditions simulating severe heat flux in the tube heat exchanger, the hyper-duplex UNS S32707 was not susceptible to pitting corrosion in any tested conditions for up to 18 month.

• Based on this study, the lower alloying reference material UNS S32205 is not suitable for use in chlorinated seawater in any tested conditions.

The following sponsors of the projects used in this study are gratefully acknowledged: Flávia Maciel at PETROBRAS, Thierry Cassagne at TOTAL ENERGIES, Stéphane Trottier at VEOLIA, Lars Mehus at AKER SOLUTIONS, Yves Denos at EDF, Xiaoxue An at TechnipFMC, Viktor Räftegård at VOLVO PENTA, Josefin Eidhagen and Ulf Kivisäkk at Alleima (supply of SAF 2205, SAF 2507, SAF 2707, and SAF 3207 seamless tubes tested in this study), Jean-Marc Lardon at ERAMET AUBERT & DUVAL (supply of UNS S31266 bars tested in this study), INDUSTEEL (supply of UNS S31266 plates tested in this study), Tadashi Kawakami at NIPPON STEEL, Sophie Delettrez, Luciana Lima, and Jérôme Peultier at VALLOUREC, and Valérie Noël at NAVAL GROUP. Pascal Moullec from Institut de la Corrosion is acknowledged for experimental setups and control of the experiments. Dominique Thierry from RISE is also particularly acknowledged for his input and review of the paper.