Engineering Failure Analysis 37(2014)42-52 Contents lists available at Science Direct VGINEFNING Engineering Failure Analysis ELSEVIER journalhomepagewww.elsevier.com/locate/engfailanal Failure analysis on abnormal wall thinning of heat-transfer CrossMark titanium tubes of condensers in nuclear power plant Part ll: erosion and cavitation corrosion Fei-Jun Chen, Cheng Yao, Zhen-Guo Yang epartment of Materials Science, Fudan University, Shanghai 200433, PR China ARTICLE INFO A BSTRACT In Part I of the failure analysis on abnormal wall thinning of heat-transfer titanium tubes eived 24 March 2013 sed in Accepted 19 Nover Available online 4 December 2013 abnormal thinning that commonly happened at the contact part between the tubes and the support plates. This kind of failure was the mainstream failure type in our case and ne main causes were found to be eccentric contact wear and three-body contact m tube Doted in processing defect of internal borings, corrosion products deposit and sagging, cles. however there still some individual failure tubes with different failure sites and modes and were located under the bypass pipes at the shoulder of the tube tower instead of in its lower part, obviously telling another failure story. In Part ll of the Cavitation corrosion failure analysis, material analysis, metallographic examination, mechanical performance tests, macro and microstructure analysis and composition analysis were conducted. The failure causes were found to be erosion and cavitation corrosion and the synergetic effect of them. Finally, corresponder ntermeasures uggested. e 2013 Elsevier Ltd. All rights reserved. 1 Introduction Since nuclear power was utilized, safety concerns have never stopped to bother us. from the disaster of Chernobyl in Russia decades ago, to the nuclear leak in Fukushima in the last two years, historical lessons written in blood have taught us that every detail in a nuclear power station is of critical importance and not a single tiniest potential peril can be ignored. The heat-transfer titanium tubes in condensers of the two 700 MW CANDU units in China- the first and only two pres- surized heavy water reactor(PHWR)units in the country-have encountered abnormal tube wall thinning with a design life of 40 years, the condensers were forced to temporarily stop operation after only 8 years in service because unexpected wall thinning problems were found on the heat-transfer titanium tubes, bringing about heavy economic loss and potential safety threat. Our team was asked to conduct failure analysis of the tubes. The tubes are made of industrial pure titanium in correspondence to Chinese brand TAl, with the length of 17370 mm, and specifications of 25. 4 mm x 0.5 mm(outside diameter x wall thickness). All these specifications have also been mentioned in Part I [ 1] of the failure analysis. Among the dozens of tube samples we got, most of them presented similar failure modes at similar positions. After detailed analysis by various techniques, we primarily ascribed most of the failure cases to eccentric contact wear al three-body contact wear rooted in processing defect of internal borings, corrosion products deposit and sagging, and foreig particles, discussed in Part I 1 of the failure analysis. However, we still found another kind of failure case with distinct appearance and failure positions. When we conducted inspection inside the condenser, we discovered the bypass pipes, onding author.Tel:+862165642523;fax:+862165103056 E-mailaddress:zgyang@fudan.edu.cn(Z-G.Yang). 1350-6307/S-see front 2013 Elsevier Ltd. All rights reserved
Failure analysis on abnormal wall thinning of heat-transfer titanium tubes of condensers in nuclear power plant Part II: Erosion and cavitation corrosion Fei-Jun Chen, Cheng Yao, Zhen-Guo Yang ⇑ Department of Materials Science, Fudan University, Shanghai 200433, PR China article info Article history: Received 24 March 2013 Accepted 19 November 2013 Available online 4 December 2013 Keywords: Titanium tube Wall thinning Failure analysis Erosion Cavitation corrosion abstract In Part I of the failure analysis on abnormal wall thinning of heat-transfer titanium tubes used in condensers in nuclear power plant, we analyzed the causes and mechanisms of abnormal thinning that commonly happened at the contact part between the tubes and the support plates. This kind of failure was the mainstream failure type in our case and the main causes were found to be eccentric contact wear and three-body contact wear rooted in processing defect of internal borings, corrosion products deposit and sagging, and foreign particles. However, there were still some individual failure tubes with different failure sites and modes and were located under the bypass pipes at the shoulder of the tube tower instead of in its lower part, obviously telling another failure story. In Part II of the failure analysis, material analysis, metallographic examination, mechanical performance tests, macro- and microstructure analysis and composition analysis were conducted. The failure causes were found to be erosion and cavitation corrosion and the synergetic effect of them. Finally, corresponding countermeasures were suggested. 2013 Elsevier Ltd. All rights reserved. 1. Introduction Since nuclear power was utilized, safety concerns have never stopped to bother us. From the disaster of Chernobyl in Russia decades ago, to the nuclear leak in Fukushima in the last two years, historical lessons written in blood have taught us that every detail in a nuclear power station is of critical importance and not a single tiniest potential peril can be ignored. The heat-transfer titanium tubes in condensers of the two 700 MW CANDU units in China – the first and only two pressurized heavy water reactor (PHWR) units in the country – have encountered abnormal tube wall thinning. With a design life of 40 years, the condensers were forced to temporarily stop operation after only 8 years in service because unexpected wall thinning problems were found on the heat-transfer titanium tubes, bringing about heavy economic loss and potential safety threat. Our team was asked to conduct failure analysis of the tubes. The tubes are made of industrial pure titanium in correspondence to Chinese brand TA1, with the length of 17370 mm, and specifications of 25.4 mm 0.5 mm (outside diameter wall thickness). All these specifications have also been mentioned in Part I [1] of the failure analysis. Among the dozens of tube samples we got, most of them presented similar failure modes at similar positions. After detailed analysis by various techniques, we primarily ascribed most of the failure cases to eccentric contact wear and three-body contact wear rooted in processing defect of internal borings, corrosion products deposit and sagging, and foreign particles, discussed in Part I [1] of the failure analysis. However, we still found another kind of failure case with distinct appearance and failure positions. When we conducted inspection inside the condenser, we discovered the bypass pipes, 1350-6307/$ - see front matter 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.engfailanal.2013.11.002 ⇑ Corresponding author. Tel.: +86 21 65642523; fax: +86 21 65103056. E-mail address: zgyang@fudan.edu.cn (Z.-G. Yang). Engineering Failure Analysis 37 (2014) 42–52 Contents lists available at ScienceDirect Engineering Failure Analysis journal homepage: www.elsevier.com/locate/engfailanal
F- Chen et aL/ Engineering Failure Analysis 37(2014)42-52 Fig. 1. Bypass pipes. which were designed for shock mitigation of high re steam during start and stop(Fig. 1). As illustrated by the tube rrangement diagram shown in Fig. 2, within t like structure of 9922 heat transfer titanium tubes in each con- denser, samples in Part II are located at the tower under the bypass pipes while the samples we analyzed in Part I 1 were all in the lower part of the tower. So we decided that these special tubes must tell another failure story, which discussed in the current Part ll of the failure analysis. After detailed characterization and analysis, the root cause of the failure was found to be erosion, cavitation corrosion and the synergetic effect of them. Previous work in our lab on the finite element modeling of erosion 5, 6 has been reported but in the current failure case, we are more concerned about the erosion mechanism in real engineering application. actually, avitation corrosion of pure titanium and titanium alloys in electrolyte solutions has been reported 2,3. And cavitation phe- nomenon of commercially pure titanium has been studied in the lab[4. But cavitation corrosion of pure titanium tubes industrially utilized has been rarely touched upon. what's more, erosion and cavitation corrosion interacted and aggravated the wall thinning of titanium tube in our case, which is a relatively novel discovery look fron the outlet: count fron left to right ow number number of tubes in a row failure tubes in PartⅡ 2s is found in 20 Fig. 2. Schematic illustration of the location of failure tubes in Part II in the condenser
which were designed for shock mitigation of high pressure steam during start and stop (Fig. 1). As illustrated by the tube arrangement diagram shown in Fig. 2, within the tower-like structure of 9922 heat transfer titanium tubes in each condenser, samples in Part II are located at the tower shoulder under the bypass pipes while the samples we analyzed in Part I [1] were all in the lower part of the tower. So we decided that these special tubes must tell another failure story, which was discussed in the current Part II of the failure analysis. After detailed characterization and analysis, the root cause of the failure was found to be erosion, cavitation corrosion and the synergetic effect of them. Previous work in our lab on the finite element modeling of erosion [5,6] has been reported but in the current failure case, we are more concerned about the erosion mechanism in real engineering application. Actually, cavitation corrosion of pure titanium and titanium alloys in electrolyte solutions has been reported [2,3]. And cavitation phenomenon of commercially pure titanium has been studied in the lab [4]. But cavitation corrosion of pure titanium tubes industrially utilized has been rarely touched upon. What’s more, erosion and cavitation corrosion interacted and aggravated the wall thinning of titanium tube in our case, which is a relatively novel discovery. Fig. 1. Bypass pipes. Fig. 2. Schematic illustration of the location of failure tubes in Part II in the condenser. F.-J. Chen et al. / Engineering Failure Analysis 37 (2014) 42–52 43
F-f. Chen et aL/ Engineering Failure Analysis 37(2014)42-52 able 1 Chemical composition of TAl in-service titanium tube and the standard values(wt). Fe N H Other elements Total of the rest elements Measured values International values 03 ≤0.1 Note: International values are from GB/T 3620. 1-2007titanium and titanium alloy brand and their chemical composition"[8 Fig 3. Metallographic structure of one in-service titanium tube 200x kN Force-displacement curve 12 4. Force-d Table 2 Mechanical properties of in-service titanium tubes. Tensile strength(ab/MPa) Elongation percentage ( Average 327 379 Note: Sample 1 and sample 2 were taken from tubes 2A032012-l and 2A032012-5, 6 in condenser 2A of unit 1
Table 1 Chemical composition of TA1 in-service titanium tube and the standard values (wt%). Chemical element Fe C N O H Other elements Total of the rest elements Measured values 0.058 0.012 0.004 0.1 0.0026 – – SB338 values 0.30 0.08 0.03 0.25 0.15 60.1 60.4 International values 0.20 0.08 0.03 0.18 0.015 60.1 60.4 Note: International values are from GB/T 3620.1-2007 ‘‘titanium and titanium alloy brand and their chemical composition’’ [8]. Fig. 3. Metallographic structure of one in-service titanium tube 200. Fig. 4. Force–displacement curve of tensile test sample. Table 2 Mechanical properties of in-service titanium tubes. Item Yielding strength (r0.2/MPa) Tensile strength (rb/MPa) Elongation percentage (d5/%) Sample 1 365 440 37.0 Sample 2 290 450 38.8 Average 327 445 37.9 Note: Sample 1 and sample 2 were taken from tubes 2A032012-1 and 2A032012-5, 6 in condenser 2A of unit 1. 44 F.-J. Chen et al. / Engineering Failure Analysis 37 (2014) 42–52
F-A. Chen et aL/ Engineering Failure Analysis 37(2014)42-52 3 Comparison of mechanical properties of in-service tube and the standard values. Yielding strength(oo.2/MPa) Tensile strength(ob/MPa) Elongation percentage(6s/6) 445 B338 value International value 370-530 Note: International values are from Chinese standard GB/T 3624-1995 110]. Fig. 5. Macro-and microscopic morphologies of tensile fracture of titanium tube(a)macroscopic appearance of tensile fracture(b)magnified fracture edge. 2. Experiments and results 1. Material analysis of titanium tubes 1. 1 Chemical composition To better analyze this individual case, we started from its material analysis and tried to decide whether the failure came from material pre We analyzed the material of one failure tube of condenser 1B in unit 1 as a sample ICP-AES, infrared carbon sulfur analyzer and nitrogen hydrogen oxygen gas chromatograph were applied to measure the content of Fe, C, and n, H, o respec- tively. The results are listed in Table 1. It presents that the content of impurity element in the in-service titanium tube is strictly controlled, far below the upper limit of requirements of ASME SB338 titanium specification [7(equals to the TAl industrial pure titanium in GB/T 3620 1-2007 standard of China 8), indicating that the material of the in-service titanium ubes is qualified
2. Experiments and results 2.1. Material analysis of titanium tubes 2.1.1. Chemical compositions To better analyze this individual case, we started from its material analysis and tried to decide whether the failure came from material problems. We analyzed the material of one failure tube of condenser 1B in unit 1 as a sample. ICP-AES, infrared carbon sulfur analyzer and nitrogen hydrogen oxygen gas chromatograph were applied to measure the content of Fe, C, and N, H, O respectively. The results are listed in Table 1. It presents that the content of impurity element in the in-service titanium tube is strictly controlled, far below the upper limit of requirements of ASME SB338 titanium specification [7] (equals to the TA1 industrial pure titanium in GB/T 3620.1-2007 standard of China[8]), indicating that the material of the in-service titanium tubes is qualified. Table 3 Comparison of mechanical properties of in-service tube and the standard values. Item Yielding strength (r0.2/MPa) Tensile strength (rb/MPa) Elongation percentage (d5/%) Ti tube sample 327 445 37.9 SB338 value 275–450 P345 P20 International value P250 370–530 P20 Note: International values are from Chinese standard GB/T 3624-1995 [10]. Fig. 5. Macro- and microscopic morphologies of tensile fracture of titanium tube (a) macroscopic appearance of tensile fracture (b) magnified fracture edge. F.-J. Chen et al. / Engineering Failure Analysis 37 (2014) 42–52 45
F-f. Chen et aL/ Engineering Failure Analysis 37(2014)42-52 a Fig. 6. Appearance of tubes 052003-17, 18(a) top of the tube (b)bottom of the tube. 2.1.2. Metallographic structure Fig. 3 is the metallographic structure of one in-service titanium tube. It is the typical metallographic structure of pure a-Ti -equiaxed polygonal grains that are evenly distributed with uniform size and clear grain boundary with no visible 2.1.3. Mechanical tests of titanium tubes The outer wall and inner wall of the in-service titanium tubes were exposed to high purity water steam and sea water espectively. Tests should be done to judge whether their mechanical performance had deteriorated after long time service The sample preparation and tensile testing of titanium tubes were done according to the standard of ASME SA370 [9) Fig 4 is the force-displacement curve. Table 2 shows the measured value of mechanical properties of in-service titanium tubes in unit 1 Table 3 further displays the comparison between the measured values and the standard values of AsME SB338 2]and Chinese GB T 3624-95 10 It can be learned from Fig 4 that the force-displacement curve of the titanium tube basically conforms to the tensile behavior of metal. Meanwhile, the comparison of several mechanical property indexes in Table 3 presents that the yielding strength, tensile strength and elongation percentage do not show any obvious deterioration and that all the performances have met the ASME SB338 standard [7] and Chinese GB/T 3624 standard [10. Therefore, the performance of the tube mate- rial is qualified and has not deteriorated Fig 5 is a group of pictures of microscopic and macroscopic fracture morphologies of an in-service titanium tube after tensile test. The fracture takes on a look of indention( Fig. 5(a))and the fracture surface is not smooth with tangles. also duc tile shear zones and dimples( fig. 5())are seen from the microscopic morphology of fracture edge, which is characteristic of ductile fracture. So the sample must have undergone sufficient plastic deformation before fracture At high magnification, there are hardly any inclusions in the dimples, indicating that the microscopic structure of the material is fine( Fig. 5(b)). Therefore, we can safely conclude that the mechanical performance of the in-service tube is good without indications of Till now, we have ruled out the possibility that the failure was caused by material problems
2.1.2. Metallographic structure Fig. 3 is the metallographic structure of one in-service titanium tube. It is the typical metallographic structure of pure a-Ti -equiaxed polygonal grains that are evenly distributed, with uniform size and clear grain boundary with no visible inclusions. 2.1.3. Mechanical tests of titanium tubes The outer wall and inner wall of the in-service titanium tubes were exposed to high purity water steam and sea water respectively. Tests should be done to judge whether their mechanical performance had deteriorated after long time service. The sample preparation and tensile testing of titanium tubes were done according to the standard of ASME SA370 [9]. Fig. 4 is the force–displacement curve. Table 2 shows the measured value of mechanical properties of in-service titanium tubes in unit 1. Table 3 further displays the comparison between the measured values and the standard values of ASME SB338 [2] and Chinese GB/T 3624-95 [10]. It can be learned from Fig. 4 that the force–displacement curve of the titanium tube basically conforms to the tensile behavior of metal. Meanwhile, the comparison of several mechanical property indexes in Table 3 presents that the yielding strength, tensile strength and elongation percentage do not show any obvious deterioration and that all the performances have met the ASME SB338 standard [7] and Chinese GB/T 3624 standard [10]. Therefore, the performance of the tube material is qualified and has not deteriorated. Fig. 5 is a group of pictures of microscopic and macroscopic fracture morphologies of an in-service titanium tube after tensile test. The fracture takes on a look of indention (Fig. 5(a)) and the fracture surface is not smooth with tangles. Also ductile shear zones and dimples (Fig. 5(b)) are seen from the microscopic morphology of fracture edge, which is characteristic of ductile fracture. So the sample must have undergone sufficient plastic deformation before fracture. At high magnification, there are hardly any inclusions in the dimples, indicating that the microscopic structure of the material is fine (Fig. 5(b)). Therefore, we can safely conclude that the mechanical performance of the in-service tube is good without indications of deterioration. Till now, we have ruled out the possibility that the failure was caused by material problems. Fig. 6. Appearance of tubes 052003-17, 18 (a) top of the tube (b) bottom of the tube. 46 F.-J. Chen et al. / Engineering Failure Analysis 37 (2014) 42–52