Austenitic and duplex stainless steels with 3 mm thickness have been used in the current work. The type, EN designations, and typical chemical composition of the steels are given in Table 1. The corresponding PREN (%Cr+3.3x%Mo+16x%N) is also included to provide a rough ranking of their relative resistance to localized corrosion. All the steels were taken from standard commercial production at Outokumpu, either in Sweden or Finland. All steels had typical microstructures with very low levels of non-metallic inclusions.

The results from five electrochemical methods were investigated to develop a robust method for constructing pitting engineering diagrams:

Method 1: Potentiodynamic polarization technique (Pitting potential measurements, E_pit) at a constant temperature.

Method 2: Potentiostatic critical pitting temperature (CPT).

Method 3: Potentiostatic measurement at a constant temperature for 2 h.

Method 4: Sweep potential (from E_OCP to E_app) followed by potentiostatic polarization at constant temperature for 2 h.

Method 5: Sweep potential (from E_OCP to E_app) followed by potentiostatic polarization at constant temperature for 30 days.

In method 5, two selected potentials (E_app) were set for defining the pitting no pitting boundaries in this study: 1) 150 mV vs. SCE for simulating sterile tap water and 2) 400 mV vs. SCE for simulating slightly chlorinated water or water with some bacterial activity (Mameng et al., 2017; Outokumpu, 2023). Measurements 2–4 were performed at 400 mV vs. SCE.

Potentiodynamic anodic polarization curves were used to determine the E_pit for EN 1.4404 at different chloride concentrations and temperatures. The typical curve for the evaluation of E_pit is illustrated in Figure 2A. E_pit was defined at the point where the current exceeds 100 μA·cm². The reason for using this current density and not 10 μA·cm⁻² is that the so-called metastable pits could also result in current peaks above 10 μA·cm⁻². Hence, clear identification of stable pitting required a higher threshold (Mameng et al., 2017). Figure 3 shows the effect of chloride concentration and temperature on E_pit and illustrates that the E_pit decreases when the chloride concentration and/or temperature increase. The breakdown potential plot vs. a wide range of temperatures is typically observed as a reversed S-shaped curve (Meguid et al., 1998; Mameng et al., 2017). As can be seen from this figure, at the lowest chloride level investigated the breakdown potential is high (above 1,200 mV vs. SCE). The electrochemical reaction taking place here is transpassive corrosion and not pitting corrosion. At higher chloride concentrations and temperatures, the breakthrough potential is the result of pitting corrosion.

The E_pit results obtained from the potentiodynamic polarization curves were used to define the limits in terms of chloride concentration and temperature where the risk for pitting corrosion exists. The idea of pitting engineering diagrams based on E_pit measurements has been presented earlier (Fielder and Johns, 1989; Lu et al., 1993; Laycock and Newman, 1998). In the present study, the diagrams were obtained by mapping the regions where pitting occurs at selected potentials, as shown in Figure 4. The solid points indicate that E_pit is equal to or lower than 400 mV vs. SCE, whereas open points indicate E_pit at higher potentials. The results in Figure 4 show that E_pit data gives sufficiently small standard deviations for evaluating a broad range of chloride concentrations and temperatures. The pitting diagram obtained from Method 1 could potentially be useful for material selection in slightly chlorinated water or the presence of biofilm in the water system. It can also serve as a basis for assessing corrosion risks in service, particularly if the effect of additional environmental components can be considered.

Raising the temperature at a constant potential to determine the critical pitting temperature (CPT) is well established and included in the standard test method ASTM G150 (ASTM G150-18, 2018). The pitting engineering diagram for EN 1.4404 was constructed using the CPT results. Method 2 follows the ASTM G150 standard except that the measurements were performed at 400 mV rather than 700 mV vs. SCE level described in the standard. Moreover, NaCl concentrations other than 1 M were used to construct the pitting diagram based on method 2. The current density vs. temperature curve and the corresponding pitting engineering diagram constructed based on the CPT data at the applied potential of 400 mV vs. SCE for EN 1.4404 are illustrated in Figure 5. The results show that CPT data give sufficiently small standard deviations for tests performed at 1,000 ppm or higher chloride concentrations. Method 2 has experimental challenges at low chloride concentrations (<1,000 ppm Cl⁻), as unwanted crevice corrosion is easily initiated, especially for the low alloyed stainless steel tested grade.

In method 3 a constant potential (400 mV vs. SCE) was applied to the specimen, and the solution was kept at a constant desired temperature. The current density was monitored as a function of time during the whole experiment. The test was repeated for solutions with different chloride ion concentrations, temperatures, and steel grades, as seen in Figure 6A and the corresponding pitting engineering diagram was obtained (Figure 6B). The advantage of this method compared to method 2 is that it can also be used at low chloride levels (<1,000 ppm Cl⁻).

Method 4 is the modified version of Method 3 in which the potential is swept slowly from the E_OCP to 400 mV vs. SCE after which the potential is kept unchanged for 2 h. The current density is monitored as a function of time during the whole experiment. The potentiostatic polarization curves obtained from Method 4 and the corresponding pitting engineering diagrams are shown in Figure 7.

The test procedure of Method 5 is similar to Method 4, but there is a difference in the testing duration during which the potential is kept constant at either 150 mV or 400 mV vs. SCE for 30 days. Figures 8A–C show the current vs. time curves obtained from Method 5 tested in 100 ppm Cl⁻ at different temperatures for EN 1.4404. The occurrence of pitting corrosion is seen as an increase in the current density, as illustrated in Figures 8B, C. The tests showed an increase in current density after about 230 and 65 h, respectively. In contrast to the tests at higher temperatures, no corrosion was seen at 60°C, maintaining a low current density throughout the test. The pitting engineering diagrams based on Method 5 is shown in Figure 8D. After testing, the visual and microscopy examination indeed confirmed that pitting had occurred at 70°C and 80°C (Figures 9B, C), whereas the sample tested at 60°C (Figure 9A) shows no pitting corrosion attack. Figure 9C indicates some pitting close to the edge. Nevertheless, examination of the pit initiation region (using a needle) indicated that the pit indeed was initiated from the surface.

A summary of the results for EN 1.4404 stainless steel based on Methods 1–5 at different chloride concentrations is shown in Table 3.

To assess the relevance of the pitting engineering diagrams to long-term service performance, long-term laboratory testing (Method 5) was performed on a wide range of stainless steel grades to simulate the corrosive environments. In these tests, two oxidation potentials of the water system were selected based on those presented earlier (Fielder and Johns, 1989; Mameng et al., 2014; Mameng et al., 2017). In the current work, a primary consideration was that 150 mV vs. SCE corresponds to simulating sterile tap water. In comparison, 400 mV vs. SCE corresponds to slightly chlorinated water or water with some biological activity. Nine different stainless steel grades were tested in different chloride ion concentrations and temperatures at two oxidation potentials of the water system.

The long-term test results on austenitic and duplex grade stainless steels are summarized in Tables 4, 5, respectively. Samples were examined after exposure to the chloride solutions with different temperatures for 30 days. For each test condition and material, it is investigated whether the material suffered pitting corrosion on the tested surface. The pitting corrosion on the edges of the tested specimens was disregarded. Where corrosion occurs, the cells are filled red (X); whereas no corrosion is indicated as green cells (O).

Based on the result, the obtained pitting engineering diagrams from different test methods show similarities and differences regarding the limiting condition for pitting corrosion (i.e., the border for pitting and no pitting) of stainless steel. Five electrochemical methods were performed on samples of the same materials in identical test solutions to construct and assess the diagrams’ accuracy. The most notable feature from the diagrams obtained from the five methods is the curve-shaped form (see Figure 10) compared with the straight lines or S-shape curves usually seen in other studies (Mameng et al., 2017; Mameng and Pettersson, 2014). The fundamental trends are very similar for the five methods. An explanation for the curved shape boundary between pitting and no pitting regions in the engineering diagrams, shown in Figures 10–12, is that the pitting initiation process is impeded by the lower availability of chloride ions at sufficiently low chloride contents. An increase in temperature is to be expected to have several impacts on pitting initiation and growth. Firstly, the rate of chemical reactions, including metal dissolution increases with temperature, as does the pit growth rate (Sedriks, 1986). Additional effects of temperature include faster diffusion of various species through the oxide film as well as into and out of the pit (Leckie and Uhlig, 1966; Sedriks, 1986; Moayed et al., 2003). Once the critical level of chloride ion and temperature is exceeded, the fast dissolution rate will occur, leading to pitting initiation and propagation (ASTM G61-86, 2018; ASTM G150-18, 2018).

Even though the different methods result in a similar shape of the borderlines between pitting and no pitting regions, it is evident that they provide quite different results (see Figure 10). The borderlines (pitting limit) obtained from methods 1–5 differ due to the protocols of the electrochemical method used; i.e., the procedure used for applied potential and incubation time (testing duration). These parameters may affect the passive film’s stability and performance (Olsson and Landolt, 2003; Wegrelius and Olefjord, 1993). It is well known that the passive film breakdown depends on the induction time for pit initiation. It appears that the existence of an incubation time means that the relatively high rate of potential change (dE/dt) for the potentiodynamic technique at constant temperature (Method 1) shifts the E_pit to higher values compared to an applied constant potential at a given temperature (Methods 3, 4, and 5). Accordingly, Method 1 seems to give too optimistic data due to the high sweep rate (20 mV/min) of the potential (almost no induction time), and it is very time-consuming as it requires many tests and specimens.

In the case of CPT measurements (Method 2), the time for the temperature to reach CPT gives the passive film a chance to adjust its properties. The passive film is modified during the time spent at temperatures below the CPT. Hence, the trend is a more stable passive layer that breaks down at higher temperatures than those obtained in Methods 3 and 4. Nevertheless, the relatively fast rate of the temperature change most probably increases the CPT to slightly higher values. The diagram obtained from Method 2 seems to give data that is too optimistic due to the high-temperature ramp rate (almost no induction time). Comparing the border lines between Method 1 and Method 2 (Figure 10) shows a general agreement between the results of the two tests for grade EN 1.4404. Nevertheless, there is less than a 5°C difference between Method 1 and Method 2 (see Table 3). As described earlier, the advantage of the potentiostatic CPT (Method 2) is that it gives results in a short period and with a limited number of tests.

The results obtained from Methods 3 and 4 give more conservative data than Methods 1 and 2 and can be utilized even at low chloride concentrations (<1,000 ppm Cl⁻). There is a relatively good agreement between the boundary lines developed from Methods 3 and 4. The diagrams from Method 3 are slightly more conservative than Method 4, as the specimens from Method 4 are subject to a gradual potential increase from the E_OCP to E_app (with a scan rate of 20 mV·min⁻¹). This sweep potential procedure strengthens the passive film compared to Method 3. Hence, the breakdown of a more stable passive layer (formed in Method 4) occurs at higher temperatures than that in Method 3. However, for EN 1.4404, the borderline from Method 3 was quite similar to Method 4 at tested low and high chloride levels. At the intermediate chloride levels, the borderline was similar to Methods 1 or 2. Comparing the borderline between Method 5 and other Methods confirms the general agreement between these methods, but as seen in Figure 10, the most conservative borderline is obtained from Method 5. Method 5 clearly shows the importance of an induction time for pit initiation (more than 2 h), giving the most conservative yet most realistic data. Even though Method 5 is time-consuming, it provides more confidence when predicting in-service corrosion performance.

A long-term electrochemical method was used to determine the pitting corrosion resistance of a range of stainless steels in aqueous solutions. It represents a matrix of the critical environmental parameters that strongly influence pitting corrosion performance, i.e., the system’s chloride ion concentration, temperature, and oxidizing power of the medium. The data generated by this method have been used to construct pitting engineering diagrams (see Figures 11, 12). The idea of engineering diagrams based on long-term laboratory testing has been presented earlier (Fielder and Johns, 1989; Mameng et al., 2014; Mameng et al., 2017). The result from the two potentials is included in the same diagram, one for each stainless steel grade, as shown in Figure 11. The diagrams show a significant relationship between pitting corrosion behavior, environmental conditions, and material parameters. The dashed area between the lines indicates the uncertainty of the borders between pitting and no pitting. The differences in the borderline of the different oxidizing potentials of the water system are noticeable (Figure 11). The lines in the diagrams for 400 mV vs. SCE have a similar trend to 150 mV_SCE but are more conservative. The borderline (pitting limit) differs due to the oxidizing effect in these parameters, which may affect the passive film’s stability and performance (Bardwell et al., 1992; Wegrelius and Olefjord, 1993; Olsson and Landolt, 2003; Fredriksson et al., 2010). In Figure 12, the results obtained on all investigated stainless steel grades are shown in the same diagram, but only one potential per diagram. Stainless steel grades EN 1.4547 and EN 1.4410 are only represented in Figure 12B, as they are too corrosion-resistant at a lower potential of 150 mV vs. SCE (i.e., Figure 12A). It is seen that the alloying composition of the different stainless steels affects the limits for the pitting corrosion resistance. Two types of pitting engineering diagrams have been constructed based on the water system’s oxidation potential (E_OCP). As was mentioned earlier, 150 mV vs. SCE corresponds to the upper limit of the potential exerted on stainless steel when exposed to sterile tap water. In comparison, whereas 400 mV vs. SCE represents slightly chlorinated water or water with some biological activity. In practice, The maximum potential is typically in the range 200–250 mV vs. SCE for a drinking water system (Mameng and Pettersson, 2014; Mameng and Pettersson, 2011) and about 300 mV vs. SCE for a water heater system (Johansson and Mameng, 2014). For chlorinated water, the potential of the system depends on the chlorine level: it has been reported that the maximum potential is about 400 mV vs. SCE for 0.2 ppm chlorine and about 500 mV vs. SCE for 0.5 ppm chlorine (Larché et al., 2013; Mameng and Alfonsson, 2012). However, the main point to consider when using the diagrams is that the oxidation potential of the water system may not always be known. It should be noted that these diagrams only give approximate guidance as to the suitability of stainless steel. The final choice of stainless steel, however, will depend on several other factors, such as different surface conditions and the presence of other environmental species or crevice design or weld. It is important to remember that the actual service conditions may differ considerably from those used to design the diagram. One example is if crevices or weld oxides are present in the construction. Crevice corrosion typically initiates more easily than pitting corrosion, so the lines in the diagram are shifted towards more conservative values. Another example is if some contaminants are present in the water, as certain substances could activate, and others could inhibit the corrosion process. The extension of engineering diagrams in which these factors are considered is out of the scope of this work but is ongoing in our laboratory.