1-s2.0-S1350630713003658-main

Accepted ManuscriptRoot Cause Analysis for 316L Stainless Steel Tube LeakagesS. Kaewkumsai, S. Auampan, K. Wongpinkaew, E. ViyanitPII: S1350-6307(13)00365-8DOI: http://dx.doi.org/10.1016/j.engfailanal.2013.11.008Reference: EFA 2198To appear in: Engineering Failure AnalysisReceived Date: 2 August 2013Revised Date: 19 November 2013Accepted Date: 19 November 2013

Please cite this article as: Kaewkumsai, S., Auampan, S., Wongpinkaew, K., Viyanit, E., Root Cause Analysis for316L Stainless Steel Tube Leakages, Engineering Failure Analysis (2013), doi: http://dx.doi.org/10.1016/j.engfailanal.2013.11.008

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customerswe are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, andreview of the resulting proof before it is published in its final form. Please note that during the production processerrors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1/12

Root Cause Analysis for 316L Stainless Steel Tube Leakages S. Kaewkumsai*, S. Auampan, K. Wongpinkaew, E. Viyanit

Failure Analysis and Materials Corrosion Laboratory (FAMC), National Metal and Materials Technology Center (MTEC), National Science and Technology Development Agency (NSTDA), 114 Thailand Science Park, Paholyothin Rd., Klong 1, Klong Luang, Pathumthani 12120, Thailand

ABSTRACT

Perforation of ASTM A270 TP316L stainless steel tube, used for transportation of ozonated high purity water in a pharmaceutical plant, was discovered after 3 months in actual service. The current investigation was conducted in order to explore the root causes of failure. Various techniques including on-site investigation, emission spectroscopy, ion chromatography, radioscopy, optical microscopy, scanning electron microscopy, energy dispersive spectroscopy, and intergranular corrosion testing were implemented for failure analysis of the tube components. The results revealed that the perforation of tube was initiated from the outer

wall and extended to the inner wall by pitting corrosion. The stagnant state of chloride-containing water was the main reason for inducing such corrosion attack. The weld metal was

the most susceptible to corrosion attack leading to perforation of the wall thickness, although initiation sites of pitting corrosion were also observed in the base metal. The dimensions of each pit mouth are very small, but enlarged subsurface cavities were observed. The selective dissolution of material due to galvanic effects between delta-ferrite and the austenite matrix occurred in the weld zone. It is suggested that failure prevention could be achieved by

controlling the quality of the insulation system. In addition, careful control of welding conditions must be implemented during fabrication.

Keywords: Pitting corrosion; Welding defect; Tubing failure; Stainless steel

------------------------------------------------------------------------------------------

* Corresponding author. Tel.: +66 2564 6500x4736; fax: +66 2564 6332. E-mail address: [email protected] (S. Kaewkumsai).

2/12

1. Background information A pharmaceutical plant has encountered the leakage of an ozonated high purity water transfer tube. The seamless tube was made of stainless steel grade ASTM A270 TP316L [1]. Each section of tube was connected by manual circumferential TIG welding with no filler metal. The tube has an outside diameter of 50 mm and wall thickness of 2 mm. The welding process was performed according to AWS D18.1 standard and each weld line was inspected by visual inspection according to ASME BPVC - Section IX. From the inspection, the qualities of welds were found to be within the acceptable range. Leak testing of the tubing system was also conducted at a pressure of 800 kPa for 6 hours using nitrogen gas. The pressure testing showed no indication of leakages.

The tubing system was horizontally installed with thermal insulation (EPDM non-halogenated synthetic elastomers) using chlorine-free glue for joining the insulator material together. The insulation was also externally wrapped by aluminum jacket with using silicone for bonding and sealing jacket materials firmly together. The insulation of tubing system was completely done ahead of schedule by 4 months. When it was time for plant commissioning test, plant personnel cleaned the inner wall of the tube system with hot de-ionized water and were

passivated. After the passivation, the tubing system has been running with the ozonated high purity water at a temperature of 297-299 K under the pressure of 500 kPa. However, after 3-month utilization, the leakage was found in the tubes that were installed outdoors on the plant roof. The perforation of the tube occurred in a short period of service relative to plant shutdown and replacement cost. Detailed investigation of the leakage of the tubes is presented

in this paper.

2. Experimental On-site investigation was conducted to observe the service conditions to investigate the

leakage site and to consider collection of samples including the accumulated trap water on the insulator at the bottom of the tube in the leaked site as well as the corroded specimens for laboratory testing. After receiving the samples, bulk composition of one of the failed tubes was determined using a spark emission spectrometer for comparison with the standard specification. In order to determine the corrosive species contained in the sample trapped water, ion chromatography (IC) was used. To determine the nature of metal loss through the tube thickness, radioscopy was conducted before taking specimens for microstructure analysis. Surface analyses were performed both on the outer and inner surface of the corroded

3/12

tube samples. The physical appearance of the tubes was captured using a low magnification digital camera for damage analyses. High magnification observation with a scanning electron microscope (SEM) was conducted to study the surface morphology. Energy dispersive spectrometry (EDS) was used for chemical composition analyses of the substances or particles deposited at the surface. The samples were then cross-sectioned and prepared for

microstructural analyses following these steps; cold mounting in resin, grinding with silicon carbide paper, polishing with diamond suspension, and etching in a glyceregia solution (one part of HNO3 and four parts of HCl with H2O2 as a catalyst). The microstructural analyses were performed using a reflected light microscope. Finally, the susceptibility of the weldment

to intergranular attack was conducted in accordance to ASTM A262 [2]. The samples were polished with silicon carbide papers and subsequently etched in oxalic acid (10%) for 1.5 min after polishing. The current density was controlled at 1 A/cm2. The etched surfaces of each sample were studied by optical microscopy.

3. Results 3.1 Visual examination During on-site investigation, it was found that the tubing system had 5 parallel lines consisting of two steam transfer tubes, one waste water transfer tube and two ozonated water transfer tubes as shown in Fig. 1. The perforation problem was only found on both ozonated water transfer tubes. Close observation of the perforated tubing revealed that the aluminum jacket was defective in that the silicone used for sealing joints in the jacket was degraded (Fig. 2a), resulting in gaps between the adjacent sections (Fig. 2b). After removing the jacket and insulator from the tube, it was found that accumulated stagnant water was trapped in the insulation as shown in Fig. 2c. It is believed that the defects in the insulation system allowed the entry of rain water and condensed moisture which accumulated in the insulation at the bottom of the tube. The temperature outside the tube monitored using an infrared thermometer

was 303.6 K (Fig. 2d). The trapped liquid was collected for evaluation of anions by Ion Chromatography in the laboratory. Observation on the outer surface of tube revealed the corrosion attack in many locations. Strongly adhered of black insulator was also observed on the outer surface. Then, two samples of stainless steel tubes containing the corrosion pits were cut and removed for metallurgical examination in the laboratory. They can be divided into two groups as follows: 1) tube sample with corrosion in the base metal (as labeled with SC-1) as illustrated in Fig. 3a and 2) tube sample with corrosion in the weld (as labeled with WD-1) as shown in Fig. 3b.

4/12

3.2 Chemical analyses Bulk composition of the failed tube sample using spark emission spectrometer is presented in Table 1. The results were compared with the specification of ASTM A270 TP316L according to ASTM standard practice for Seamless and Welded Austenitic Stainless Steel Sanitary Tubing, A270-03. Based on the analysis results, they are in agreement with the standard

specification.

Table 1 Chemical composition of the tube sample and standard grade (wt.%). Element C Mn Si P S Cr Ni Mo Nb Cu N ASTM A270 TP 316L

0.035a 2.0a 0.75a 0.04a 0.03a 16-18 10-15 2-3 - - -

Failed tube 0.03 1.43 0.49 0.03 0.008 16.32 10.07 2.31 0.056 0.319 0.033 a Maximum.

In order to determine the corrosive ions contained in the accumulated stagnant water, Ion

Chromatography was used to determine the anions in solution. From this test, the concentration of chloride (Cl-) was 138.73 ppm, sulfate (SO42-) 78.65 ppm and nitrate (NO3-) 61.96 ppm as presented in Fig. 4. The presence of chloride ions can result in pitting attack [3]. Furthermore, the simultaneous presence of sulfate in the stagnant water could accelerate the

corrosion attack [4].

3.3 Radioscopic inspection From visual inspection of the failed samples at the site, pitting corrosion was observed in all

samples. Therefore, radioscopy was used to determine the shape of metal loss through the tube thickness. The results were useful for considering the location for the samples for microstructural analyses. Radioscopy of the failed tube sample in the pitted area reveals large internal pores as shown in Fig. 5. It clearly shows the evidence of pit centralizing is found to be a group together of small pits as seen on the outer surface of the weldment.

3.4 Corrosion analyses 3.4.1 Corrosion at the base metal Visual examination of the SC-1 sample revealed that corrosion pits were only observed on the outer surface. Corrosion attack did not penetrate through the wall thickness. Corrosion pits were found in several areas but specify at the bottom half of the tube.

5/12

Based on the corrosion analysis on the SC-1 sample, only localized corrosion was found in a form of pitting corrosion which is distributed around on the outer surface of the tube in the base metal as shown in Fig. 6a. Strong adherence of black insulator was also observed (Fig. 6a-b). The results from EDS showed that the black insulator contained significant concentration of chlorine and sulfur (Fig. 6c). After removing the black insulator, corrosion attack was found on the surface of the tube (Fig. 6d). It is believed that the adhered insulator acts as the preferential adsorption of the corrosive species from the accumulated stagnant water. High magnification observations with SEM on the pits show corrosion characteristics as shown in Fig. 7a-b. The results from EDS showed that the residue present within these pits

contained significant concentrations of chlorine and sulfur (Fig. 7c). These elements were not found in the base metal area (Fig. 7d). Cross-section through the pits (in Fig. 7a-b) indicated that corrosion attack was initiated from the outer wall and extended into the material (Fig. 8). The depths of the pitting corrosion were measured to be approximately 200 and 260 micrometers (along cross-section A-A and B-B), respectively.

From the corrosion analyses at the base metal, it can be concluded that pitting had started from the outer surface and penetrated into the wall thickness at several areas but did penetrate

the wall. The corrosive species deposited in the pits mainly contained chlorine and sulfur. The sources of chlorine and sulfur should be attributed to the rain water or condensed moisture.

3.4.2 Corrosion at the weld Visual examination showed that the tube sections of the WD-1 sample had been manually circumferential welded. Rough surfaces of weldment were observed. Corrosion pits were mostly found at the interface between the weld metal and heat affected zone (HAZ). Examination the inner surface of the tube revealed that severe discoloration was noticed in the HAZ region (Fig. 9), indicating that the region might be sensitized during welding.

Close examination of the WD-1 sample indicated corrosion attack at the weld metal adjacent to the HAZ. The outer surface of the perforation is shown in Fig. 10a. High magnification examination on the outer surface with SEM found grain boundaries as shown in Fig. 10b probably caused by acid pickling during tube welding. The inner surface of the perforation is shown in Fig. 10c, no intergranular corrosion was detected. Comparing the corrosion damage between the outer and inner surfaces of the tube around the perforation location, there was more severe corrosion on the outer wall. High magnification of area 2 in Fig. 10b show

6/12

micro-galvanic corrosion at the interface between the -ferrite and austenite (Fig. 10d). The small hole and black precipitates deposited on surface of the weldment were further examined as shown in Fig. 11a. High magnification of the small hole shows shrinkage defect (Fig. 11b). High magnification observation with SEM of the black precipitates deposited area indicated the presence of skeletal dendritic structure [5] (Fig. 11c). From chemical composition analysis of the black precipitates by EDS, contaminants including sulfur and manganese were found (Fig. 11d). They could segregate to shrinkage area in weld. Cross-section through the hole indicated that the pitting corrosion occurred at the interface between the weld metal and HAZ, and extended into the weld metal in the transverse direction (through thickness), then extended further into the base metal in the longitudinal direction with irregular shape like tunnel (along tubing) as shown in Fig. 12a. The microstructure of the fusion line as shown in Fig. 12a resembles the cast structure consisting of -ferrite interdendritic structure in an austenitic matrix. High magnification observation at the fusion zone revealed general corrosion, which was caused by the micro-galvanic effect between -ferrite and austenite (Fig. 12b). This general attack could initiate the preferred sites for growth of pits [5-6]. This observation indicated that corrosion was initiated from the outside and propagated to the inner wall. As the corrosion progresses, the wall thickness will decrease, and finally, leading to

perforation.

Base on the corrosion analyses at the welds, it can be concluded that the perforated hole had started from the outer surface and propagated to the inner surface at the interface between the

weld metal and HAZ. Although the openings of the corrosion pits are very small, the large subsurface cavities are formed. Grain boundary corrosion on the outer surface at the HAZ and the black precipitates containing the elemental sulfur and manganese were also observed. The microstructure of the base metal was normal austenitic structure, while the dendritic structure

of -ferrite and austenite was present in the weld. No evidence of chromium carbide precipitate at the grain boundaries was observed.

3.4.3 Intergranular corrosion test Due to the presence of grain boundary corrosion observed in the HAZ of the WD-1 sample, specimens of this sample were prepared for intergranular corrosion tests. They were cut from

the base metal and near the weldment areas. The susceptibility to intergranular attack was conducted in accordance to ASTM A262, Practice A. It is used to assess the intergranular corrosion resistance of stainless steel through the three basic morphologies: step, dual and

7/12

ditch [6]. Based on the test results (Fig. 13), the HAZ and base material showed no degree of sensitization and were classified as step. The presence of steps between grains is an acceptable structure after oxalic acid etching for stainless steel components [7]. It also confirmed that the stainless steel samples were not likely to be susceptible to intergranular corrosion attack. Evidently, chromium carbides were not precipitated at the grain boundary.

4. Discussion From the experimental results, it is evident that the tube failed from pitting corrosion mechanism. It is well known that the susceptibility of austenitic stainless steel to pitting

corrosion depends on environmental factors [8, 9], their chemical compositions and metallurgical features. For the environmental factor, localized corrosion of austenitic stainless steel; the severity of which chloride can induce depends on the temperature and pH [10]. Furthermore, the simultaneous presence of chloride and sulfate in the stagnant water could accelerate the corrosion attack [4]. Therefore, the presence of sulfates in the trapped liquid that accumulated in the insulation at the bottom of the tube as shown in Fig. 4 increase the susceptibility of stainless steels to pitting corrosion.

In stagnant waters with the presence of corrosive species, the corrosion resistance of stainless steels could be decreased due to lower protective film stability, which leads to pit initiation. In the present study, the stagnant trapped water (Fig. 2c) could result from the defective insulation system as observed during onsite investigation (see Fig. 2a-b). It is believed that the initial source of accumulated water could come from the diffusion and/or the penetration of condensed moisture/rain water through the defective joints in the insulation. Most likely, leaked water from the tube interior could increase the volume of accumulated water. Furthermore, chemical analysis indicates that the trapped water contains Cl- (138.73 ppm), SO42- (78.65 ppm) and NO3- (61.96 ppm). These results are consistent with the EDS analyses of the deposit present within the corroded area, which contains significant concentrations of chlorine and sulfur (see Fig 7c). The adhered insulator could facilitate the adsorption of corrosive species as suggested by EDS results (Fig. 6c) since the thermal insulation was made of non-halogenated synthetic elastomers and the glue for joining the insulating material together was chlorine-free, the source of chloride and sulfate must have come mainly from the rain water or condensed moisture.

8/12

Pitting morphology was clearly observed on both the base metal and the weld metal. The most severe attack, complete penetration through wall thickness, was found at the weld metal whereas the base metal suffered only from pitting not complete perforation. In many cases, welding decreases corrosion resistance of metal [11-12]. In this case, the austenitic base metal, which has higher corrosion resistance, was changed to dual austenite-ferrite

microstructure in the weld area. When the welded tube is exposed in the electrolyte, the weld metal therefore acts as an anode whereas base metal acts as a cathode leading to preferential attack at the weld. The rate of localized attack is very fast in a system with a large cathode (base metal) and a small anode (weldment) [13] as in this case. Welding can also alter the stability of the passive layer and its corrosion behavior [12, 14]. When the affected area is exposed to the corrosive environment, the defect sites on passive film could promote the localized corrosion such as pitting corrosion [11, 15]. The residual stresses introduced in weld metals could enhance the contribution factor of corrosion attack in the weld [2]. However, correct weld anneal along with the carefully surface cleaning after welding could prevent this problem [16]. In the case of stainless steel, welds and HAZ are most susceptible to pitting corrosion [17]. Base metal attack is less common, although it has been observed as presented in this case.

Examination of the weld metal at the leaked site revealed the grain boundaries corrosion surface as shown in Fig. 10b. Such evidence indicated that the protective film was imperfect and inhomogeneous. Onsite welds are usually welded from the exterior of the tube using

shielding gas inside. Normally, it is difficult or even impossible to clean the root of the weld after welding properly. If the shielding gas contains even a relatively small amount of oxygen, heat tint layers (discoloration) will be formed close to the weld as seen on the inner surface of the tube in Fig. 9. The outer surface probably experienced the same degree of discoloration, but it has been removed by abrasive treatment or acid pickling. These heat tint layers could

promote the initiation and growth of corrosion pits even in seemingly harmless environments. Furthermore, improper cleaning of the heat tint could generate intergranular corrosion as seen in Fig. 10b although the results from evaluation of the corrosion behavior of the weldment show that there was no susceptibility to intergranular corrosion (Fig. 13). However, the presences of inhomogeneous surface at the weld metal are susceptible areas for chloride induced pitting corrosion [2, 18].

9/12

Cross-sectional analyses through pits at the welds indicated that the initiation of pitting started from the outer surface of the tube and propagated toward the inner surface. Surface analyses of the failed area revealed that the openings of the corrosion pits are very small but cross-section of the pit showed networks of interconnecting tunnels through the tube wall (see Fig. 12a). Corrosion propagation tends to be migrating downward and slightly to the sides rather than straight down through wall thickness. Such evidence could further confirm by the results from radioscopy as shown in Fig. 5. It differs from the general pitting morphology, which is normally spherical in nature, but, in this finding, pits are of irregular shape. From the results, it is suggested that the pitting corrosion at the weld is very dangerous for engineering

structure because sometimes there were no clear evidence of corrosion attack on the outer surface, but inside the material was very severe attack.

Autogenous welding, without adding filler metal, of the austenitic stainless steel produced a fusion zone with a dendritic structure of chromium-rich delta ferrite in a Ni-rich austenite matrix [16]. Such structure is generated from the high cooling rate of fusion line after welding. From chemical composition of the failed tube as shown in Table 1, the chromium and nickel equivalents can be calculated to predict the ferrite content in fusion zone according

to the WRC-1992 [15] as follows;

Creq = Cr + Mo + 0.7Nb = 16.32 + 2.31 + 0.7(0.056) = 18.67 Nieq = Ni + 35C + 0.25Cu + 20N = 10.07 + 35(0.03) + 0.25 (0.32) + 20(0.033) = 11.86

From this calculation the amount of delta ferrite in fusion zone is about 7%. The presence of 3-8% delta ferrite is optimal to reduce hot cracking [19]. However, this amount of Cr-rich delta ferrite could contribute to general corrosion and pitting corrosion susceptibility of the weld [17]. Microstructural analysis through the weld metal indicated the general corrosion of the fusion zone, as seen in Fig. 12b, results from micro-galvanic effects between delta ferrite and austenite matrix. The high content of chromium in delta ferrite, in turn causes Cr depletion in the adjacent austenite matrix, making the material very sensitive to localized corrosion. Therefore, corrosion at the fusion zone is localized and acts as the initiation mechanism for pitting.

From the discussion above, it is clear that the defective insulation and the presence of stagnant chloride containing solution caused to the perforation of the ASTM A270 TP316L stainless

10/12

steel tube by pitting corrosion. The presence of dual delta ferrite/austenite phases and welding defects in the weld metal could lead to the fast rate of tube leakage. Thus, the leakage of tube in this case occurred in a short service life.

Therefore, to prevent such failure from occurring in the future, it is necessary to keep the

thermal insulation free from gaps at the joints to prevent the surrounding corrosive media from reacting with the tube surface. High quality welding to ensure the delta ferrite content as well as prevention of trapped liquid is also highly advisable. Post-weld annealing will increase the intergranular corrosion resistance as well as decrease the susceptibility to general

corrosion at the fusion zone [11, 16]. To ensure that the tube and weld are free from defects i.e. porosity, cavity, and etc., the non-destructive testing technique such as radiography, etc., is recommended before implementing the components in service.

5. Conclusions The ASTM A270 TP316L stainless steel tube was leaked due to pitting corrosion. Leakage of the tube was initiated from the outside and propagated toward the inside. The rapid rate of wall penetration resulted from defects and discontinuities in the weld metal. It is recommend

that the thermal insulation be kept free from gaps at the joint and the quality of welding be closely monitored to prevent this failure from occurring in the future.

Acknowledgment

The authors are grateful to Dr. John T. Harry Pearce and Dr. Amnuaysak Chainpairot for valuable comments and Failure Analysis and Materials Corrosion Laboratory, MTEC for support the testing instruments.

11/12

References [1] Standard Specification for Seamless and Welded Austenitic Stainless Steel Sanitary Tubing (ASTM A270),

ASTM Standards, 2003. [2] Standard Specification for Detecting Susceptibility to Intergranular Attack in Austenitic Stainless Steels

(ASTM A262, ASTM Standards, 1998. [3] Wu W, Cheng G, Zhu W, Xu C, Cai H. Corrosion failure analysis of stainless steel components for maleic

anhydride plant. Eng Fail Anal 2010; 17: 90612. [4] Laitinen T. Localized corrosion of stainless steel in chloride, sulfate and thiosulfate containing

environments. Corros Sci 2000; 42: 421-41. [5] Mohammadi Zahrani E, Saatchi A, Alfantazi A. Pitting of 316L stainless steel in flare piping of a

petrochemical plant. Eng Fail Anal 2010; 17: 810817. [6] Garcia C, de Tiedra MP, Blanco Y, Martin O, Martin F. Intergranular corrosion of welded joints of

austenitic stainless steels studied by using an electrochemical minicell. Corros Sci 2008; 50: 23907. [7] Lu BT, Chen ZK, Luo JL, Patchett BM, Xu ZH. Pitting and stress corrosion cracking behavior in welded

austenitic stainless steel. Electrochim Acta 2005; 50(6): 1391-403. [8] Ramana KVS, Anita T, Mandal S, Kaliappan S, Shaikh H. Effect of different environmental parameters on

pitting behavior of AISI type 316L stainless steel: Experimental studies and neural network modeling. Mater Design 2009; 30: 37705.

[9] Pardo A, Merino MC, Coy AE, Viejo F, Arrabal R, Matykina E. Pitting corrosion behaviour of austenitic stainless steels combining effects of Mn and Mo additions. Corros Sci 2008; 50: 17961806.

[10] Starosvetskya J, Starosvetsky D, Armon R. Identification of microbiologically influenced corrosion (MIC) in industrial equipment failures. Eng Fail Anal 2007; 14: 150011.

[11] Werner SE, Johnson CA, Laycock NJ, Wilson PT, Webster BJ. Pitting corrosion of 304 stainless steel in the presence of a biofilm containing sulphate reducing bacteria. Corros Sci 1998; 40: 465-80.

[12] Garcia C, Martin F, de Tiedra P, Blanco Y, Lopez M. Pitting corrosion of welded joints of austenitic stainless steels studied by using an electrochemical minicell. Corros Sci 2008; 50: 1184-94.

[13] Cuia L, Xiaoganga L, Chaofang D. Pitting and galvanic corrosion behavior of stainless steel with weld in wet-dry environment containing CI. J. Univ. Sci. Technol. Beijing 2007; 14(6): 517-22.

[14] Timofeev BT, Karzov GP, Gorbakony AA, Nikolaev YK. Corrosion and mechanical strength of welded joints of downcomers for RBMK reactors, Int J Pres Ves Piping 1999; 76: 299307.

[15] Kou S. Welding metallurgy. 2nd ed. John Wiley & Sons; 2003. [16] Shi X, Avci R, Geiser M, Lewandowski Z. Comparative study in chemistry of microbially and

electrochemically induced pitting of 316L stainless steel. Corros Sci 2003; 45: 257795. [17] Olszewski AM. Avoidable MIC-Related Failures. JFAP 2007; 7: 23946. [18] Geesey GG, Gillis RJ, Avci R, Daly D, Hamilton M, Shope P, Harkin G. The influence of surface features

on bacterial colonization and subsequent substratum chemical changes of 316L stainless steel. Corros Sci 1996; 38: 73-95.

[19] Davis JR (Ed.). Corrosion of Weldments. Davis & Associates. ASM International. 2006: 5.

12/12

Captions for Illustrations Fig. 1. The tubing systems have 5 parallel lines consisting of two steam transfer tubes, one waste water transfer tube and two ozonated water transfer tubes.

Fig. 2. (a) The degradation of silicone used for joining and sealing sections of insulation jacket (b) the gap between the overlapped insulation jackets and (c) the stagnant trapped water under the insulator and (d) surface temperature of the tube. Fig. 3. Stainless steel tube samples for analysis: (a) samples with corrosion in the base metal and (b) samples with corrosion in the welds. Fig. 4. Chromatograph showing the anions of the trapped water. Fig. 5. Radioscopic image of the pitting area of the failed tube sample showing a large internal defect (white spots). Fig. 6. Corrosion of the SC-1 sample: (a) pitting corrosion in the base metal on the outer surface of the tube (b) SEM photograph showing pitting corrosion on the strongly adhered insulation (c) EDS spectrum of the black insulator and (d) corrosion attack under the insulation. Fig. 7. Microanalyses of the SC-1 sample: (a-b) SEM photographs showing pitting on the outer surface and small particles depositing in the holes (c) EDS spectrum of the residue within the pit showing significant concentrations of chlorine, sulfur, silicon and aluminum and (d) EDS spectrum of the normal area.

Fig. 8. Cross-section through the pits of the SC-1 sample showing morphologies of the pits initiated from the outer surface: (a) along cross-section A-A in Fig. 7a and (b) along cross-section B-B in Fig. 7b.

Fig. 9. Severe discoloration on the inner surface of the HAZ of WD-1 sample. Fig. 10. Corrosion characteristic of WD-1 sample: (a) The perforation was initiated on the outer surface at the interface between the weld metal and HAZ (b) grain boundary corrosion was found near the perforation (c) the characteristics of the inner surfaces around the perforation and (d) high magnification image of area 2 in Fig. 10b showing micro-galvanic corrosion at the interface between the -ferrite and austenite structures.

Fig. 11. (a) the small hole and black precipitates deposited on surface of the weldment (b) High magnification image of the small hole show a shrinkage (c) high magnification image of the black precipitates deposited area indicated the presence of skeletal dendritic structure and (d) chemical analysis of the black precipitates showing contaminants including sulfur and manganese.

Fig. 12. Cross-section through the pit of WD-1 sample: (a) the network of interconnecting tunnels within the tube wall and (b) the general corrosion caused by the micro-galvanic effect between -ferrite and austenite structures.

Fig. 13. The HAZ and base material showed zero degree of sensitization and were classified as step structure according to ASTM A262.

Fig. 1.

Fig. 2.

Fig. 3.

Fig. 4.

Fig. 5.

Fig. 6.

Fig. 7.

Fig. 8.

Fig. 9.

Fig. 10.

Fig. 11.

Fig. 12.

Fig. 13.

13/12

The highlight in this article are consists of;

1. The initiation of pitting was propagated anti-gravity and only found at the bottom half of the tubes.

2. The rapid rate of wall penetration resulted from defects and discontinuities in the weld metal.

3. The perforation of the tube occurred in a short service time relative to downtime and replacement cost.