Stainless Steel Selection for Wet Flue Gas Desulfurization

STAINLESS STEEL SELECTION FOR FLUE GAS DESULFURIZATION EQUIPMENT . Gary M. Carinci William A. Pratt Outokumpu Stainless 3209 McKnight East Drive...

16 downloads 778 Views 4MB Size
S TAINLES S S TEEL S ELECTION FOR FLUE GAS DES ULFURIZATION EQUIP MENT Gary M. Carinci William A. Pratt Outokumpu Stainless 3209 McKnight East Drive Pittsburgh, Pennsylvania 15237 [email protected]

Abstract A flue gas desulphurization (FGD) unit is effectively a chemical processing plant incorporated into a power station with the purpose of removing sulfur dioxide from the flue gas. Corrosion conditions in a FGD plant utilizing the wet scrubbing process technology are a complex combination of chlorides and fluorides, acidity, temperature and other operating conditions details (deposits, crevices). This paper discusses different stainless steels that are used in FGD systems and the behavior of the alloys in artificial scrubber environments. The corrosion resistance of UNS S32205 duplex stainless steel is studied by electrochemical testing performed under laboratory conditions, which simulate process conditions. An overall table showing the limits of different grades with respect to chlorides, fluorides and temperature in simulated scrubber environments is presented.

In tro d u c tio n Methods for removing sulfur oxides from flue gases have been studied for many years. The effects of the industrial revolution first became evident in the mid 1800’s. This resulted in a series of ideas for cleaning the gases resulting from carbon based power generation. The problems became evident with the introduction of large-scale power plants in the 1920’s. Early commercial systems were operational in the UK during the 1930’s. These plants were built with mild steels with or without organic coatings. It was soon discovered that the corrosivity of the flue gases was much worse than expected. Rising environmental concerns led to the introduction of many commercial systems by the 1970’s. Japan and the US became market leaders in the application of FGD technology. Today, flue gas desulfurization is standard practice all over the world.

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

The environments in flue gas desulfurization plants are often very complex and vary widely. No plant is like the other and material selection has to be done for each plant individually, based on the corrosive media, location within the plant, operational conditions, plant design as well as the total economic aspect. The corrosivity also varies considerably from the inlet to the outlet section of the FGD unit (1).

Corrosion The main issues in FGD plants are uniform corrosion, caused when condensates of acids are formed, and pitting and crevice corrosion, which are the greatest threats caused by scrubbing solutions. Uniform Corrosion When the temperature drops below the dew point, concentrated acids will precipitate and increase the risk for uniform corrosion. These acids are often contaminated with halides such as chlorides and fluorides. If the condensate is formed at high temperatures with low pH and high contamination levels, it will be so aggressive that no structural metallic materials will have sufficient corrosion resistance to withstand long-term exposure. At these high temperatures, however, there will only be limited condensation. It is only when the steel temperature drops 2030 °C (68-86 °F) below the dew point, that the amount of condensates will be sufficient to cause severe corrosive attack (2). These problems can be avoided either by maintaining an operating temperature above the dew point or lowering the operating temperature to a value where the level of condensation is sufficiently low to keep the pH and halide levels at acceptable levels. It is also possible to spray the sensitive areas with water to further dilute the condensates. Generally, the uniform corrosion resistance of a stainless steel is enhanced with increasing contents of chromium, nickel and molybdenum. The corrosion rate is often given in mm/year and can be presented in isocorrosion diagrams showing the combination of acid concentration and temperature giving a corrosion rate of 0.1 mm/y (3.9 mpy-mils per year). Isocorrosion diagrams can be used for comparing the corrosion resistance of various grades in different media and also used as a tool for material selection. Figures 1 and 2 show isocorrosion diagrams, respectively, in pure sulfuric acid and in sulfuric acid with the addition of chlorides (3). It should be noted that the presence of chlorides markedly increases the corrosion rates for all grades.

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

Figure 1. Isocorrosion diagram (0.1 mm/y) in pure sulfuric acid

Figure 2. Isocorrosion diagram (0.1 mm/y) in sulfuric acid with addition of 2,000 ppm chlorides

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

Pitting and Crevice Corrosion Pitting and crevice corrosion are the most frequent modes of attack in FGD plants. They are most commonly observed on areas in contact with scrubbing solutions, and the risk increases with low pH values in combination with high halide concentrations. The resistance to pitting and crevice corrosion increases with the alloying elements chromium, molybdenum and nitrogen. An empirical formula for estimating the localized corrosion resistance is given by the Pitting Resistance Equivalent index (PRE). PRE is an empirically established index based on the levels of chromium, molybdenum and nitrogen: PRE = %Cr + 3.3(%Mo) + 16(%N) (4). PRE values based on this formula are shown in Table 1 for stainless steels commonly used for FGD applications. When comparing the corrosion resistance experimentally, common methods are to measure the lowest temperatures at which corrosion occurs. These temperatures are known as the Critical Pitting Temperatures (CPT) and Critical Crevice Temperatures (CCT), and are shown in Figures 3 and 4. CPTs and CCTs may be determined either in standard laboratory electrolytes or in more complex solutions simulating real service environments. These methods are useful in ranking materials for FGD applications, because the environments are typically acidic and high in chlorides. Crevice corrosion is especially important since it is difficult to avoid deposits in scrubber operations. It should be remembered that critical temperatures should only be used for ranking different alloys, not for estimating performance in real applications. For example, neither the CPT nor the CCT give any information on propagation rates.

Figure 3. Typical CPT values in 1 M NaCl for some stainless steels tested with a ground surface condition according to ASTM G 150

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

Figure 4

Typical CCT values in 6% FeCl3 + 1% HCl for some stainless steels tested in the ground surface condition according to ASTM G 48 Method F.

Stainless Steels As mentioned earlier, FGD plants can be divided into zones with different environments and consequently also different degrees of corrosivity. To find an optimal solution from performance and economical aspects, material selection has to be made for each zone individually. Some suggested grades are given in Table 1 together with the typical composition of these steels and their PRE values. Table 1. Stainless steels for FGD applications. Typical chemical composition, wt% UNS Microstructure S31603 S31703 S31726 N08904 S31254 S34565 S32205 S32750

Austenite Austenite Austenite Austenite Austenite Austenite Duplex Duplex

Cr

Ni

Mo

N

16.2 18.2 17.8 20 20 24 22 25

10.2 13.7 12.7 25 18 17 5.7 7

2.1 3.1 4.1 4.3 6.1 4.5 3.1 4

— — 0.14 — 0.2 0.45 0.17 0.27

Others

1.5Cu 0.7Cu 5.5Mn

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

PRE 23 28 34 34 43 46 35 43

A short description of the characteristics of each grade follows: 316L

Molybdenum alloyed austenitic stainless steel, S31603. To be used in mild environments, such as storage tanks.

317L type

Improved version of S31603 which includes S31703 and S31726. Higher level of molybdenum but also alloyed with nitrogen to improve the corrosion resistance. Application examples include spray lances and the upper part of the absorber and demister.

904L

N08904, high alloyed austenitic grade. Developed to perform well in dilute sulfuric acid. Application examples include spray lances and the upper part of the absorber and demister.

S31254

S31254, high alloyed 6% Mo austenitic grade. Developed to perform well in solutions containing chlorides. Examples of applications are condensers, parts of absorbers, stacks and flue gas ducts.

4565

S34565, high alloyed austenitic grade. Frequently used in FGD applications with good results. Can be used for absorber vessels, spray lances, demisters, ducts and the stack.

2205

S32205, duplex grade with corrosion resistance similar to N08904 in many environments. High mechanical strength. Can be used for absorbers and condensers.

2507

S32750, duplex grade with corrosion resistance comparable to S31254 but with higher mechanical strength. Applications include absorber and condenser components.

Corrosion Testing For practical reasons, evaluations of the corrosion performance of different alloys are often performed in the laboratory, where the corrosive environment is simulated as close to the real application as possible. However, laboratory tests are confined to relatively short times with a practical upper time limit on the order of a month. Normally, the environmental parameters (temperatures, concentrations) are kept constant to allow for a systematic evaluation. In practice, concentrations and temperatures within a FGD system can fluctuate within wide ranges over time. Field tests have the advantage of testing under actual conditions, but it can be more difficult to make systematic conclusions. In a typical field test, welded specimens of different grades are mounted on test racks, often with artificial crevice formers between each specimen. The test rack is welded in place on site where the test is performed and is often under exposure for at least half a year. By performing a field test, the effect of process fluctuations on the corrosion resistance of the stainless steels can be evaluated.

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

In the following, laboratory tests in simulated scrubber solutions are described. Laboratory Tests in Artificial Scrubber Environments Specimens according to Figure 5, in size 60x60 mm, were mounted on test racks similar to the one shown in Figure 6. On both sides of each specimen, multiple crevice formers of PTFE were mounted. The test racks were then assembled with a torque of 1.58 Nm (14 lb-in). All specimens were insulated from each other and from the test rack.

Figure 5 - Welded test specimen

Figure 6 - Example of test rack

Corrosion tests were performed for 32 days in eight artificial scrubber environments prepared from sodium chloride and sodium fluoride. The pH was adjusted using sulfuric acid. The stainless steel grades tested, the composition of the artificial scrubber environments, pH and test temperatures, together with the test results are shown in Table 2 (1).

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

Table 2. Test conditions and test results showing maximum attack depth in microns. Test Solutions Cl-(ppm) F- (ppm) pH Temp, °C[°F]

Grade S31726 N08904 S31254 S34565 S32205 S32750 Not microns

A

B

C

D

E

F

G

H

33000 1000 4.0 55[131]

33000 1000 4.0 50 [122]

20000 200 4.0 70 [158]

20000 200 4.0 55 [131]

5000 200 4.5 50 [122]

2000 200 4.5 50 [122]

500 50 1.5 50[122]

500 50 1.0 50 [122]

0 0

0 0

0 0

0 0

310 210

113 0 0 0 87 0

0 0

0 0

0 0

0 0

0 0

64

0

TestedResistantNot resistant, max attack depth in

None of the alloys suffered from uniform or pitting corrosion, and the maximum corrosion rate measured was 0.003 mm/y (0.12 mpy). Crevice attack was found on some samples, and Table 2 indicates the maximum crevice depth. Both alloys S34565 and S31254 were resistant even in solution A, which contained the highest levels of chlorides and fluorides, at a temperature of 55°C (131°F). Both alloys also suffered crevice corrosion in the more diluted solution (C) at the higher temperature, 70°C (158°F). In solution A, grade S32750 was attacked by crevice corrosion at 55°C, but was not attacked at 50°C (122°F) in the same solution (solution B). Grade N08904 was slightly more resistant than grades S31726 and S32205. The results give the following ranking of the alloys with the least resistant alloys to the left:

S31726/S32205 < N08904 < S32750 < S34565/S31254

Effect of Chlorides, Sulfates and Passivation on Corrosion Resistance The corrosion resistance limitations of stainless steels are expanded for FGD applications when compared to those in natural waters containing similar levels of chlorides. This is likely due to the very high sulfate content in the scrubbing solutions, which shifts the critical pitting potential to more noble values and inhibits pitting corrosion. To assess the effect of sulfates on the corrosion resistance of UNS S32205 stainless steel, cyclic polarization studies were conducted in the laboratory using an electrochemical cell with an electrolyte capacity of 100 ml for the experiments. A diameter of 1 cm2 on a UNS S32205 stainless steel sample was exposed to the electrolyte during the test. Prior to testing, the 6 mm

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

thick hot rolled plate sample was mechanically polished with silicon carbide abrasive paper to a 600 grit finish, followed by cleaning with methanol. The electrochemical tests began at least 24 hours after surface preparation for the passive film to form naturally with the atmosphere. A graphite electrode was used as the counter electrode, and the reference electrode was a saturated calomel electrode (SCE) for all measurements. The electrolytes were prepared using distilled water and analytical grades of sodium chloride (NaCl) and sodium sulfate (Na2SO4). Two temperatures (55°C, 70°C), three chloride concentrations (30,000 ppm, 40,000 ppm and 50,000 ppm by weight) and three sulfate concentrations (0%, 0.2%, 2.0% by weight) were investigated at a pH of 4 (pH adjusted by H2SO4 addition). The tests consisted of plotting a cyclic polarization curve to assess the potential range of the passive area and the breakdown potential. The cyclic polarization scan began at -100 mV versus the open circuit potential and was scanned at a rate of 0.2 mV/s in the anodic direction until the current density reached 500 μA/cm2 or the potential reached 1 V in the forward scan. The scanning rate was increased to 1 mV/s in the reverse scan and was plotted again to -100 mV versus the open circuit potential. The breakdown potential was measured at a current density of 100μA/cm2 on samples where pitting was observed after testing.

Effect of Chlorides The cyclic polarization scans for the 40,000 and 50,000 ppm chloride condition (Figure 7) at a temperature of 55°C and pH 4 showed the anodic polarization behavior was significantly different in the forward scan. The sample exposed to 40,000 ppm chlorides was scanned to a potential of 1 V without a significant increase in the current density or breakdown in the passive film. The sample exposed to 50,000 ppm chlorides was scanned to a current density up to 500 μA/cm2 in the forward scan and showed a breakdown potential of 135 mV. After the completion of the electrochemical tests, the exposed areas of the stainless steel samples were examined using optical microscopy and it was found that the sample exposed to 40,000 ppm chlorides showed no pitting while the sample exposed to 50,000 ppm chlorides displayed surface pits.

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

Figure 7. Cyclic polarization scans (55°C) for UNS S32205 duplex stainless steel in 40,000 and 50,000 ppm chloride solutions at pH 4 without sulfate additions.

Effect of Sulfates The cyclic polarization scans for the UNS S32205 exposed to the 50,000 ppm chloride solution at a temperature of 55°C and pH 4 (Figure 8) showed that the polarization curves for 0% and 2% sulfates were significantly different in the anodic direction up to 500 μA/cm2 in the forward scan. The scan in the test solution with added sulfates showed the passive film required a significantly higher applied potential before the current density increased, and the sample exposed to the 2% sulfates displayed a larger passive domain than the solution without sulfates. In this investigation the breakdown potential is defined as the potential where the current density reaches a level of 100μA/cm2. The testing with 2% sulfates showed a breakdown potential of 411 mV versus 135 mV for the electrolyte with 0% sulfates. Microscopic examination after electrochemical testing showed pitting on both samples.

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

Figure 8. Cyclic polarization scans (55˚C) for UNS S32205 duplex stainless steel in a 50,000 ppm chloride solution at pH 4 with and without sulfates.

Additional electrochemical tests were conducted at 70°C to determine the effect of sulfates on the corrosion resistance of components operating at higher temperatures in the FGD scrubber. The cyclic polarization scans for the UNS S32205 stainless steel sample exposed to the 30,000 ppm chloride solution at a temperature of 70°C and pH 4 (Figure 9) showed the curves for 0.2% and 2% sulfates were similar over the entire range of scanned potentials. In this investigation the breakdown potential is defined as the potential where the current density reaches a level of 100μA/cm2. The testing with 0.2% and 2% sulfates showed a similar breakdown potential of approximately 115 mV for the UNS S32205 stainless steel. Microscopic examination after electrochemical testing showed pitting on both samples. In addition, the polarization curves for 0% and 2% sulfates for the 30,000 ppm chloride condition at a temperature of 70°C and pH 4 (Figure 10) were similar over the entire range of scanned potentials, and both test solutions showed a breakdown potential of approximately 120 mV for the UNS S32205 stainless steel. Both samples displayed pitting after electrochemical testing. At the higher exposure temperature of 70°C, sulfate levels up to 2% in the test solution had no significant effect on the pitting behavior of UNS S32205 stainless steel.

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

Figure 9. Cyclic polarization scans (70˚C) for UNS S32205 duplex stainless steel in a 30,000 ppm chloride solution at pH 4 with different sulfate contents.

Figure 10. Cyclic polarization scans (70˚C) for UNS S32205 duplex stainless steel in a 30,000 ppm chloride solution at pH 4 with and without sulfates.

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

Effect of Passivation The effect of a passivation treatment on the pitting resistance of UNS S32205 stainless steel was also assessed by electrochemical testing a sample with a 600 grit ground finish and then testing another site on the same sample after a passivation treatment. The electrochemical tests were initiated seven days after surface preparation for the passive film to form naturally with the atmosphere. The passivation treatment consisted of exposing a 1 cm2 test area of the ground sample to a 30% nitric acid solution for thirty minutes at a temperature of 22°C. The cyclic polarization scans in the 50,000 ppm chloride solution (Figure 11) at a temperature of 55°C and pH 4 showed the anodic polarization behavior was significantly different in the forward scan. The ground and passivated sample exposed to 50,000 ppm chlorides was scanned to a potential of 1 V without a significant increase in the current density or breakdown in the passive film. The ground sample exposed to 50,000 ppm chlorides was scanned to a current density up to 500 μA/cm2 in the forward scan and showed a breakdown potential of 135 mV when measured at a current density of 100 μA/cm2. After the completion of the electrochemical tests, the exposed areas of the stainless steel sample was examined using optical microscopy and the ground and passivated site exposed to 50,000 ppm chlorides showed no pitting while the ground site exposed to 50,000 ppm chlorides displayed surface pits. The difference in the polarization scans is due entirely to the passivation treatment, as the same 600 grit ground sample of UNS S32205 stainless steel was used in both electrochemical tests. The corrosion resistance of a ground stainless steel surface is improved significantly by a passivation treatment. Passivation treatments of stainless steel with nitric or mild organic acids are useful to enhance the protective nature of the natural, air-formed passive film. Nitric acid treatment enhances the level of chromium in the protective film on stainless steels (5).Consideration should be given to using some type of surface passivation treatment after final fabrication of air pollution control equipment to enhance the corrosion resistance of the stainless steel components. Passivation treatments are not designed to remove weld heat tint, embedded iron particles, weld spatter, and other surface defects produced during fabrication, since nitric acid does not attack or remove the surface layer with heat tint or embedded defects. Elimination of these defects requires removal of the protective oxide layer on stainless steels and 25 to 40 microns of the substrate metal by pickling the surface with nitric-hydrofluoric acids. The protective film then reforms in air over the freshly pickled surface. This oxide film is uniform and leaves the stainless steel surface in its normal passive condition. While pickling is not strictly a passivating treatment, it provides many of the same benefits. Nitric-hydrofluoric acid pickling pastes and gels are most useful for localized cleaning of weld areas, but also can be used to improve the corrosion resistance of ground, blasted, and brushed surfaces.

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

Figure 11. Cyclic polarization scans (55°C) for UNS S32205 duplex stainless steel in a 50,000 ppm chloride solution at pH 4 without sulfates show the effect of a passivation treatment on the pitting resistance.

Laboratory Tests in an Absorber Solution from a Coal Fired Power Plant To assess the corrosion resistance of UNS S32205 stainless steel exposed to an operating absorber solution with 15,700 ppm chlorides, cyclic polarization studies were conducted in the laboratory using an electrochemical cell with a solution capacity of 100 ml for the experiments. A diameter of 1 cm2 on a UNS S32205 mill pickled stainless steel sample was exposed to the filtered absorber solution during the test. A graphite electrode was used as the counter electrode, and the reference electrode was a saturated calomel electrode (SCE) for all measurements. The absorber test solution was filtered to remove gypsum and other solids before testing in the electrochemical cell at 50°C. The as-received absorber solution at a pH of 6.6 was investigated as well as lower pH scans (pH adjusted by H2SO4 addition) were conducted at pH values of 3.4 and 1.5. The tests consisted of plotting a cyclic polarization curve to assess the potential range of the passive area and the breakdown potential. The cyclic polarization scan began at -50 mV versus the open circuit potential and was scanned at a rate of 0.2 mV/s in the anodic direction until the current density reached 500 μA/cm2 or the potential reached 1.1 V in the forward scan. The scanning rate was increased to 1 mV/s in the reverse scan and was plotted to -60 mV versus the open circuit potential. The cyclic polarization scans of the UNS S32205 in the as-received absorber solution (Figure 12) at a temperature of 50°C and pH 6.6 showed the anodic polarization behavior was significantly different in the forward scan. The sample was scanned to a potential of 1.1 V without a breakdown in the passive film. After the completion of the electrochemical test, the exposed area of the stainless steel sample was examined using optical microscopy and the sample did not

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

exhibit any indications of pitting corrosion. A thick, black surface deposit developed on the surface of the UNS S32205 sample during the cyclic polarization testing, and Scanning Electron Microscopy (SEM) in combination with Energy Dispersive Spectroscopy (EDS) was used to investigate the morphology and composition of the surface film (Figure 13). Based on the EDS analysis, the black surface film was comprised mainly of manganese oxide. The manganese (Mn2+) in the absorber slurry is likely from manganese bearing residuals in the limestone used to react with the sulfur dioxide in the absorber reaction tank.

Figure 12. Cyclic polarization scan (50°C) for UNS S32205 duplex stainless steel in an absorber solution with 15,700 ppm chloride solution at pH 6.6 showed no indications of pitting

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

Elt. O Mg Al Si P S Cl K Ca Cr Mn Fe Ni Mo

Conc Units 14.210 wt.% 4.009 wt.% 0.643 wt.% 1.396 wt.% 0.350 wt.% 0.512 wt.% 0.205 wt.% 0.349 wt.% 1.126 wt.% 12.315 wt.% 22.534 wt.% 36.662 wt.% 3.357 wt.% 2.333 wt.% 100.000 wt.%

Total

Figure 13. SEM micrograph of the surface of the stainless steel after cyclic polarizationscan inthe as received operating absorber solution (pH 6.6) with 15,700 ppmchlorides. The EDS composition analysis (wt. %) for the circled area in the micrograph is also shown.

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

The cyclic polarization scans of the UNS S32205 in the pH adjusted absorber solution (Figure 14) at a temperature of 50°C and pH 3.4 showed the anodic polarization behavior was significantly different in the forward scan. The sample was scanned to a potential of 1.1 V without a breakdown in the passive film. After the completion of the electrochemical test, the exposed area of the stainless steel sample was examined using optical microscopy and the sample did not exhibit any indications of pitting corrosion. A thin, brown deposit developed on the surface of the UNS S32205 sample during the cyclic polarization testing, and Scanning Electron Microscopy (SEM) in combination with Energy Dispersive Spectroscopy (EDS) was used to investigate the morphology and composition of the surface film (Figure 15). Based on the EDS analysis, the surface film was slightly enriched in manganese and oxygen. The manganese (Mn2+) in the absorber slurry is likely from manganese bearing residuals in the limestone used to react with the sulfur dioxide in the absorber reaction tank. Pitting attack of UNS S32205 due to surface ennoblement from the deposit was not evident at a potential of 1.1 V.

Figure 14. Cyclic polarization scan (50°C) for UNS S32205 duplex stainless steel in an absorber solution with 15,700 ppm chloride solution at adjusted pH 3.4 showed no indications of pitting.

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

Elt. O Mg Si S Cl K Ca Cr Mn Fe Ni Mo

Figure 15.

Conc 2.759 0.043 0.421 0.242 0.000 0.099 0.108 18.664 2.114 67.284 5.505 2.762 100.000

Units wt.% wt.% wt.% wt.% wt.% wt.% wt.% wt.% wt.% wt.% wt.% wt.% wt.% Total

SEM micrograph of the surface of the stainless steel after cyclic polarization scan in the pH adjusted (pH 3.4) absorber solution with 15,700 ppm chlorides. The EDS composition analysis for the encircled area in the micrograph is also shown.

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

The cyclic polarization scans of the UNS S32205 in the pH adjusted absorber solution (Figure 16) at a temperature of 50°C and pH 1.5 showed the anodic polarization behavior was significantly different in the forward scan. The sample was scanned to without a breakdown in the passive film. After the completion of the electrochemical test, the exposed area of the stainless steel sample was examined using optical microscopy and the sample did not exhibit any indications of pitting corrosion. A deposit did not develop on the surface of the UNS S32205 sample during the low pH cyclic polarization testing, and Scanning Electron Microscopy (SEM) in combination with Energy Dispersive Spectroscopy (EDS) was used to investigate the surface morphology and composition of the scanned surface (Figure 17). Based on the EDS analysis, the surface composition was typical for UNS S32205 stainless steel. The UNS S32205 stainless steel did not show any indications of pitting corrosion when tested in the absorber solutions (15,700 ppm chlorides) that ranged in pH from 6.6 to 1.5 even after scanning to potentials of 1.1 V. This grade of stainless steel should be resistant to pitting attack when exposed to similar absorber solutions, but crevice corrosion as a result of under deposit corrosion from various scales and surface deposits in the absorber reaction tank is still a concern in operating units. The appearances in the surface deposits were significantly different after cyclic polarization testing at the three pH levels (Figure 18), and can be explained by reviewing the Pourbaix diagram for manganese (Figure 19). At higher more alkaline pH values, Mn2+ is less stable and results in manganese oxide precipitation, while at lower more acidic pH values the Mn2+ is stable up to higher potentials and remains soluble in the absorber solution. Manganese deposition was most pronounced at pH 6.6 after scanning to high potentials but was less pronounced at pH 3.4 and was nonexistent after scanning in the pH 1.5 solution. In all cases, UNS S32205 was not susceptible to manganese oxide related corrosion due to surface ennoblement even after high potential electrochemical scanning. This result was consistent with the absorber inspection results where manganese oxide deposits were lifted from the walls of the UNS S32205 absorber reaction tank but corrosion indications were not evident.

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

Figure 16. Cyclic polarization scan (50°C) for UNS S32205 duplex stainless steel in an absorber solution with 15,700 ppm chloride solution at adjusted pH 1.5 showed no indications of pitting.

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

Elt. O Mg Si S Cl K Ca Cr Mn Fe Ni Mo

Figure 17.

Conc 0.813 0.065 0.304 0.000 0.043 0.067 0.019 19.910 0.687 69.433 5.927 2.730 100.000

Units wt.% wt.% wt.% wt.% wt.% wt.% wt.% wt.% wt.% wt.% wt.% wt.% wt.% Total

SEM micrograph of the surface of the stainless steel after cyclic polarization scan in the pH adjusted (pH 1.5) absorber solution with 15,700 ppm chlorides. The EDS composition analysis for the encircled area in the micrograph is also shown.

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

Figure 18.

Photo of the stainless steel surface after cyclic polarizations scans at thethree different pH levels using theabsorber test solution.

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

Figure 19. Pourbaix diagram for manganese shows the relationship between the potential (SHE) versus the solution pH - the dashed line shows the boundaryfor water stability/water dissociation. Summary The environments in wet FGD systems can be very corrosive. There are large variations in environments between different plants and material selection has to be done for each plant individually based on the corrosive media, i.e. type of plant, location within the plant, operating conditions, plant design and economics. The main corrosion risks for stainless steels are crevice corrosion and pitting. Laboratory corrosion tests in simulated scrubber environments, and especially field corrosion tests in already existing plants, are good tools to estimate the performance of different alloys in different flue gas cleaning applications. The corrosion resistance limitations of stainless steels are expanded for FGD applications when compared to those in natural waters containing similar levels of chlorides. This is due to the high sulfate content in the scrubbing solutions and to the lower oxidizing conditions. There are clear differences between the corrosion resistance of different stainless steels, and the following ranking is valid when crevice corrosion and pitting are the service limiting corrosion forms

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved

S31603
The UNS S32205 stainless steel did not show any indications of pitting corrosion when cyclic polarization tested in absorber solutions (15,700 ppm chlorides) that ranged in pH from 6.6 to 1.5 even after scanning to potentials of 1.1 V. This grade of stainless steel should be resistant to pitting attack when exposed to similar absorber solutions, but crevice corrosion as a result of under deposit corrosion from various scales and surface deposits in the absorber reaction tank is still a concern in operating units. In all cases, UNS S32205 was not susceptible to manganese oxide related corrosion due to surface ennoblement even after high potential electrochemical scanning in an absorber solution. This result was consistent with the absorber inspection results where manganese oxide deposits were lifted from the walls of the UNS S32205 absorber reaction tank but corrosion indications were not evident. Acknowledgments The authors gratefully acknowledge Outokumpu for funding this work and for providing the stainless steel samples. The author also gratefully acknowledges Dr. James Fritz for his assistance with the electrochemical experiments.

References 1.

B. Beckers, A. Bergquist, M. Snis, E. Torsner, “Stainless Steels for Flue Gas Cleaning: Lab trials, field tests and service experience,” Paper No 07351, presented at NACE Corrosion 2007, Nashville, TN (March 2007).

2.

Avesta Sheffield, North American Division. “Avesta Sheffield Stainless Steels for FGD Scrubber Systems”

3.

Corrosion Handbook, 9th Edition, Outokumpu Stainless, Espoo Finland,2004,pp. II:60

4.

S. Szklarska-Smialowska.Pitting and Crevice Corrosion, Houston,Texas: NACE International, 2004.

5.

K. Asami, K. Hashimoto, “An X-Ray Photo-electron Spectroscopic Study of Surface Treatments of Stainless Steels,” Corrosion Science, Vol. 19, 1979, p.1007-1017

Copyright © 2012 Electric Power Research Institute, Inc. All rights reserved