Metallogr. Microstruct. Anal. DOI 10.1007/s13632-016-0300-2
TECHNICAL ARTICLE
Microstructural Analysis of Corroded Carbon Steel Used in Process Equipment Handling Caustic Soda H. M. Tawancy1
Received: 30 May 2016 / Revised: 10 July 2016 / Accepted: 20 July 2016 Ó Springer Science+Business Media New York and ASM International 2016
Abstract Heat exchanger tubes and pipes made of carbon steel to handle aqueous solutions of spent caustic and crude caustic soda, respectively, at a petrochemical plant have prematurely sustained corrosion damage. Microstructural analysis of representative samples has been conducted using scanning electron microscopy combined with energyand wavelength-dispersive spectroscopy and X-ray diffraction. It is shown that the tubes were corroded due to the presence of sulfate ions in the spent caustic solution. The corrosion damage sustained by the pipes has been related to the presence of carry over chloride ions in the caustic solution. It is concluded that for better performance, both media require the use of Ni or some Ni alloys rather than carbon steels. Keywords Carbon steel Caustic soda Corrosion Electron microscopy
Introduction Sodium hydroxide (NaOH) also referred to as caustic soda is used in numerous industrial applications due to its ability to deteriorate organic matter [1]. For example, it is used in the petroleum industry as an additive to drilling mud to increase its viscosity and to neutralize acidic gases as well as to manufacture other chemicals such as sodium salts and detergents [1]. Caustic soda is also used as a scrubber of
& H. M. Tawancy
[email protected] 1
Center for Engineering Research, Research Institute, King Fahd University of Petroleum and Minerals, P.O. Box 1639, Dhahran 31261, Saudi Arabia
certain chemicals such as H2S [1]. In this application, NaOH is consumed or spent to form NaHS and H2O, and therefore, the resulting solution is commonly known as spent caustic. Crude NaOH is produced by electrolysis in electrolytic cells where concentrated aqueous solution of sodium chloride is allowed to react with water [1]. Chlorine gas is evolved at the anode, and hydrogen gas and aqueous NaOH are formed at the cathode according to the following reaction NaClðaqÞ þ H2 O ! NaOHðaqÞ þ 1=2 H2 ðgÞ þ 1=2 Cl2 ðgÞ Due to the scrubbing effect of caustic soda and the release of chlorine by the above reaction, caustic soda has been of particular importance to the petrochemical industry. First, it is used as a scrubber of the gas resulting from cracking of ethylene, which is usually contaminated with H2S and CO2. The resulting spent caustic solution becomes contaminated with sulfides and carbonates with smaller concentrations of hydrocarbons. Secondly, chlorine is used as a feedstock in the production of poly(vinyl) chloride (PVC) [2]. Therefore, production facilities of NaOH are typically located either in close proximity to petrochemical plants or near rock salt deposits. Two corrosion problems related to caustic soda have been encountered at a petrochemical plant which are the focus of this study. After being used as a scrubber of ethylene, the resulting spent caustic at a temperature of about 130 °C is cooled in a shell and tube heat exchanger where cooling water is passed through tubes of carbon steel as schematically illustrated in Fig. 1a. According to specifications, the tubes with 13 mm external diameter and 2.5 mm wall thickness are made of carbon steel grade API 5L B. The crude caustic soda produced at the same plant at
123
Metallogr. Microstruct. Anal. Fig. 1 Schematics illustrating the process equipment where the corrosion problems have been encountered. a Shell and tube heat exchanger used to cool the spent caustic solution. b Production of caustic soda by electrolysis; the pipe is used to transfer the crude caustic solution into the crude tank
a temperature of 110 °C is transferred into a crude tank by means of pipes with 407 mm external diameter and 13 mm wall thickness and made of the same steel grade as the tubes (Fig. 1b). According to plant records, the tubes and pipes were severely damaged by corrosion after a fraction of the expected service life. Items received for analysis to determine the most probable cause of corrosion damage included: (1) sections of corroded tube and pipe, (2) a sample of the corrosion product removed from the inner surface of the pipe, (3) sections of unused tube and pipe, and (4) compositions of the spent caustic and crude caustic listed in Table 1.
Table 1 Chemical compositions of spent caustic and crude caustic (vol%) Spent caustic
Crude caustic
1–5 NaOH
11–12 NaOH
1–2 Na2SO4
12–15 NaCl
0.5
Na2SO3a
1–2 Na2CO3 a
Sodium thiosulfate
b
Sodium hypochlorite
123
0.4–0.8 Na2SO4 5–50 ppm NaOClb
Experimental Procedure Representative metallographic samples parallel to the surface and along the cross section were removed from the asreceived tube and pipe. Inductively coupled plasma-atomic emission spectroscopy (ICP-AES) was used to measure the chemical composition of the steel used in the application with the exception of the carbon concentration, which was measured by combustion calorimetry. The grain structure of selected specimens was revealed by etching in 10 % aqueous solution of oxalic acid. Detailed microstructural characterization was carried out using (1) scanning electron microscopy (SEM) combined with energy-dispersive spectroscopy (EDS) and wavelength-dispersive spectroscopy (WDS) at an accelerating voltage of 20 keV and (2) X-ray diffraction with CuKa radiation.
Experimental Results and Discussions Characterization of Unused Carbon Tube and Pipe Table 2 summarizes the measured chemical compositions of the steel used in both applications in comparison with the nominal composition of grade API 5L B carbon steel.
Metallogr. Microstruct. Anal. Table 2 Chemical composition of API 5L B carbon steel Element
Fe
Mn
Si
C
P
S
Nominal
Bal.
0.5–1.1
0.45a
0.22a
0.04a
0.035a
Tube
98.51
0.93
0.32
0.18
0.03
0.03
Pipe
98.70
0.69
0.35
0.19
0.03
0.02
a
Maximum
The compositions of both the tube and pipe fall within that of the steel grade specified for the applications. Figure 2 shows characteristic grain structures of the unused tube and pipe. The microstructure of the respective pieces was observed to consist of ferrite matrix containing pearlite noting that the steel of the tube appears to contain a slightly higher amount of pearlite consistent with its slightly higher C content in comparison with that of the pipe. Various grades of carbon steels are frequently used in process equipment handling of NaOH although their respective noting that their respective corrosion resistance is rated as satisfactory [3, 4]. Generally, carbon steels can handle all concentrations of caustic soda at room temperature; however, they become susceptible to corrosion attack at higher temperatures with rate depending the exact caustic concentration and temperature [3, 4]. For example, it has been reported that carbon steels can safely be used up to a temperature of 65 °C with 50 % caustic solution [3]. On the other hand, carbon steels are known to have poor corrosion resistance in acidic environments such as nitric, phosphoric, sulfuric (low concentrations), hydrochloric as well as sulfur- and chlorine-contaminated environments [3, 5]. Table 1 shows that the tube has been exposed to spent caustic with concentration of 1–5 % at a temperature of
130 °C, while the pipe has handled 11–12 % caustic soda at a temperature of 110 °C. Although no data are available in the open literature on such combinations of relatively low caustic concentrations and high temperatures, it has recently been shown that the corrosion resistance and resistance to stress corrosion cracking of carbon steels are functions of the C content [6]. It is found that C contents B0.23 wt% accelerate the anodic dissolution of Fe and that both the corrosion resistance and resistance to stress corrosion cracking are improved with higher C contents [6]. Reference to Table 2 shows that the actual C contents of the tube and pipe are 0.18 and 0.19 wt%, respectively. Although it is possible that the relatively low C content of the carbon steel used in the application has partially contributed to the accelerated corrosion rate, the results presented below suggest that the corrosive nature of the spent caustic and crude caustic has played the key role in the corrosion sustained by the tubes and pipes. Analysis of the Corroded Tube A secondary electron SEM macrograph illustrating the general appearance of the tube surface in as-received condition is shown in Fig. 3a. It is observed that the tube has been thinned to the extent of forming large hole with corrosion product covering the surface of the remaining thickness. The corresponding EDS spectrum of Fig. 3b shows the average elemental composition with Fe and O as the major constituents with smaller concentration of S and trace amounts of Si and Mn. It is noted here that most impurities in carbon steels have no adverse effect on their corrosion resistance [3, 5]. In the presence of caustic environments, corrosion has been shown to be controlled by diffusion of oxygen
Fig. 2 Secondary electron SEM images illustrating the microstructures of unused tube (a) and pipe (b); the insets are higher magnification images of the enclosed areas in (a, b)
123
Metallogr. Microstruct. Anal.
through a partially protective layer of FeO(OH) [7, 8]. However, this layer can be degraded by the presence of sulfur compounds present in the spent caustic (Table 1). This accelerates the inward diffusion of oxygen and in turn
the anodic dissolution of Fe. A more detailed analysis of the corrosion product observed in Fig. 3 is given below. Figure 4 summarizes the results of analyzing the elemental composition of the corrosion product at the tube surface in the vicinity of the hole shown in Fig. 3a. The morphology of the corrosion product is shown in the secondary electron SEM image of Fig. 4a. Figure 4b–d shows EDS spectra illustrating the elemental compositions of the regions marked 1, 2 and 3 in Fig. 4a. As shown in Fig. 4b and c, Fe and O are the major elemental constituents of regions 1 and 2, respectively, with smaller concentration of S and trace amounts of Si and Mn. In contrast, the outer discontinuous layer marked 3 contains less Fe and more S (Fig. 4d) in comparison with regions 1 and 2. Table 3 summarizes the results of quantifying the spectral data by wavelength-dispersive spectroscopy. It was observed that the composition of the corrosion product in regions 1 and 2 approaches that of FeO with some S contamination. The composition of region 3 appears to be consistent with that of iron sulfate (FeSO4). Observation of iron sulfate appears to be consistent with the results of earlier studies showing that the corrosion rate of carbon steels in sulfide/sulfate solutions is controlled by the rate of mass transfer of ferrous ions from a surface layer of ferrous sulfate (corrosion
Table 3 Chemical composition of the corrosion product formed on the tube surface (Fig. 4a) as determined by WDS (at%)
Fig. 3 Secondary electron SEM macrograph a illustrating the general appearance of the shell side of the tube (outer surface) in the asreceived condition and EDS spectrum, b showing the corresponding elemental composition of the corrosion product
Fig. 4 Analysis of the corrosion product formed on the shell side of the tube. a Secondary electron SEM image showing the morphology of the corrosion product near the hole shown in Fig. 3a. b–d are EDS spectra illustrating the elemental compositions of the regions marked 1, 2 and 3 in (a), respectively
123
Element
Fe
O
Region 1
48.63
45.81
Region 2
45.38
50.40
Region 3
18.26
65.11
S
Mn
Si
Na
P
4.77
0.41
0.25
0.11
0.02
3.42
0.42
0.22
0.13
0.03
15.86
0.39
0.27
0.09
0.02
Metallogr. Microstruct. Anal. Fig. 5 Secondary electron SEM images along cross section of the tube illustrating the microstructures near the outer and inner surface. a Outer surface; specimen is lightly polished. b Inner surface; specimen is polished and etched
expected. To summarize, the above observations indicate that the tube has been corroded by the presence of sulfate ions in the spent caustic solution, which accelerates the oxidation rate. Analysis of the Corroded Pipe
Fig. 6 X-ray diffraction pattern derived from the sample of corrosion product removed from the inner surface of the pipe; standard patterns of FeCl2, FeCl3 and FeO(OH) are also shown
product)–liquid interface into the bulk of the respective solution [9, 10]. When viewed along the cross section, the grain structure of the tube near the outer surface was revealed by light polishing as shown in the example of Fig. 5a. It was noted that the structure resembles that of a heavily etched material reflecting the extent of corrosion attack along and through the grains near the outer tube surface. In contrast, the grain structure near the inner surface shown in Fig. 5b could only be revealed by polishing and etching as
Figure 6 summarizes the results of X-ray diffraction analysis of the corrosion product removed from the inner surface of the pipe. Most of the observed Bragg diffraction maxima were found to match a combination of chloride and oxide phases as shown in the standard patterns of FeCl2, FeCl3 and FeO(OH). Figure 7 illustrates the morphology and elemental compositions of the corrosion product formed on the internal surface of the pipe. It is evident from the secondary electron SEM image of Fig. 7a and the corresponding elemental compositions shown in the EDS spectra of Fig. 7b, c that the corrosion product consists of an outer layer of iron chloride and an inner layer of iron oxide. Quantification of the spectral data by wavelengthdispersive spectroscopy shows that the composition of the chloride approaches that of FeCl2 and that of the oxide is consistent with that of FeO as summarized in Table 4. It was concluded from the above observations that the protective oxide layer developed by the steel has been broken down by chloride ions. Pitting has also been detected along cross sections of the pipe as shown in the secondary electron SEM image of Fig. 8. The attack appears to have preferentially occurred along the grain boundaries and is expected to occur as a result of carry over chlorides noting that the crude caustic contains 12–15 % NaCl. Although carbon steels are commonly used as structural materials in applications involving exposure to caustic media as pointed out earlier, the present study demonstrates that they may not be suited in certain cases despite their attractive low cost. The initial lower material cost may not
123
Metallogr. Microstruct. Anal. Fig. 7 Analysis of the corrosion product formed on the inner surface of the pipe. a Secondary electron SEM image illustrating the morphology of the corrosion product. b, c EDS spectra illustrating the elemental compositions of the regions marked 1 and 2 in (a), respectively
Table 4 Chemical composition of the corrosion product formed on the inner surface of the pipe (Fig. 7a) as determined by WDS (at%) Element
Fe
O
Cl
Mn
Si
Na
P
Region 1
28.37
13.02
58.21
0.16
0.12
0.09
0.03
Region 2
47.19
42.26
10.12
0.14
0.16
0.11
0.02
be justified if the long-term maintenance and repair costs as well as unscheduled shutdowns of plants are taken into account. In this regard, it well known that Ni can significantly increase the resistance to caustic solutions, and therefore, Ni and some Ni alloys are recommended for those applications where carbon steels are not the most suitable choice [3].
Conclusions Tubes of a heat exchanger used to cool spent caustic solution and pipes handling crude caustic solution at a petrochemical plant have been prematurely corroded. Both the tubes and pipes have been made from the same grade of carbon steel. The tubes are found to corrode due to the presence of sulfate ions in the spent caustic solution. Corrosion of the pipes was correlated with carry over chloride ions in the crude caustic. It is concluded that in such cases high Ni alloys can better handle the respective caustic media particularly Ni 200 and Ni 201 grades. Fig. 8 Secondary electron SEM image along cross section of the pipe showing a pit at the inner surface of the pipe as indicated by the arrows
123
Acknowledgments The author is grateful for the continued support provided by King Fahd University of Petroleum and Minerals.
Metallogr. Microstruct. Anal.
References 1. Caustic Soda Handbook, Occidental Chemical Corporation, Dallas, TX 2013. http://www.oxy.com/OurBusinesses/Chemi cals/Products/Documents/CausticSoda/caustic.pdf 2. H.M. Tawancy, Failure analysis of process equipment used in the production of polyvinyl chloride. Metallogr. Microstruct. Anal. 3, 310–322 (2014) 3. B.D. Craig, D.B. Anderson, Handbook of Corrosion Data (ASM International, Materials Park, 1995) 4. L. Garverick, Corrosion in the Petrochemical Industry (ASM International, Materials Park, 1994) 5. P.A. Schweitzer, Carbon steels, in Corrosion and Corrosion Protection Handbook, ed. by P.A. Schweitzer (Marcel Dekker, New York, 1983), p. 27 6. J. Flis, M. Ziomek-Moroz, I. Flis-Kabulska, Effect of carbon on corrosion and passivation in hot concentrated NaOH solution in
7.
8.
9.
10.
relation to caustic stress corrosion cracking. Corros. Sci. 51, 1696–1701 (2009) W. Xu, K. Daub, X. Zhang, J.J. Noel, D.W. Shoesmith, J.C. Wren, Oxide formation and conversion on carbon steel in mildly basic solutions. Electrochim. Acta 54, 5727–5738 (2009) D.J. Duquette, Aqueous and stress corrosion resistance, in Alloy and Microstructural Design, ed. by J.K. Tien, G.S. Ansell (Academic Press, New York, 1976), p. 251 B.W.A. Sherar, I.M. Power, P.G. Keech, S. Mitlin, G. Southam, D.W. Shoesmith, Characterizing the effect of carbon steel exposure in sulfide containing solutions to microbially induced corrosion. Corros. Sci. 53, 955–960 (2011) B.T. Ellison, W.R. Schmeal, Corrosion of steel in concentrated sulfuric acid. J. Electrochem. Soc. 125, 524–531 (1978)
123