Thread Content
1. Overview of the situation before repair: The E-53 equipment at Tianhua Company has been in use since June 1995. The \"cone shell + cylinder ring weld\" on the shell side of this equipment has developed crack defects on four occasions (in 1999, transverse through-cracks appeared at points 2–3 along this weld; in 2002, a transverse through-crack occurred at point 5; in September 2006, another transverse through-crack appeared between points 2 and 3; and in the first half of 2007, a transverse through-crack appeared at points 3–4). The cracks that occurred in 1999 and 2002 were repaired by Qihua Construction, while the cracks that appeared later were sealed off using cover plates by Tianhua Company, after which the equipment continued to be used. The crack that appeared this time is still located on the shell side, at the \"cone shell + cylinder ring weld,\" and it is necessary to repair it in a location other than the original repair site; however, there is a longitudinal through-crack present. During the emergency repair in May 2007, UT testing conducted by the testing center of Lutianhua Hongxu Company initially revealed defect signals in a area measuring 300 mm in length and 100 mm in width. Moreover, irregularities in the detection waves were observed over an area of 700 mm in length and 40 mm in width, suggesting the presence of cracks. Due to the tight schedule for the emergency repair and high temperatures at that time, it was only possible to conduct inspections on one side at a time; as a result, there were many blind spots in the test results, and the data obtained could only be used as a reference. After the shutdown for major repairs, certain measures are taken, followed by a further comprehensive inspection. This equipment is a key device for ammonia synthesis. If the equipment does not operate properly, it will cause the system to stop, resulting in losses of over 2 million yuan per day. Each repair takes 6 to 7 days, which means losses of 12 to 14 million yuan. If new equipment is imported, it will cost 10 to 20 million yuan, without including the installation costs. Moreover, it takes at least 1 to 2 years from ordering to delivery. Therefore, the repair of this device is of great significance. 2 Equipment Overview 2.1 Process Parameters http://www.nmtech.com.cn/jishuwang/upload1/0806171503028613.jpg 2.2 Main Materials The materials used for the tube sheets and tube boxes on both sides are SA 336 F22 (equivalent to 2.25Cr1Mo), with σs≥310 MPa, σb=515～690 MPa, and δ (%)≥19 ； Its chemical composition is shown in Table 2. http://www.nmtech.com.cn/jishuwang/upload1/0806171503523545.jpg 2.3 Diagram of the ring seam locations that need repair: The diagram of the ring seam locations requiring repair is shown in Figure 1. http://www.nmtech.com.cn/jishuwang/upload1/0806171504469925.jpg 3 Comprehensive inspection of the original welds after shutdown 3.1 Macroscopic inspection and UT testing: After the system is shut down, the insulation layer is removed, and the excess height of the welds around the shell side cone and the cylinder is leveled off, followed by a comprehensive inspection. Using probes with different K values, 100% UT + 100% MT testing is carried out in 3 stages to determine the location and depth of any defects, with those areas and their depths being marked accordingly. Discuss the test results with Tianhua Company to determine the scope and methods of repair. UT and MT inspections revealed the presence of cracks in all 4 areas at points 1 to 5 of this ring seam. According to the flaw detection markings, use a grinder to grind away the defects, and confirm through flaw detection that they have been completely removed. The results showed that there were 4 defects with cracks of varying lengths, widths, depths, and orientations, and their depths, widths, and lengths were unusual. The test results are shown in the expanded view of E53 test status. After grinding open the 4 defects, and to prevent any missed defects, a further UT inspection was conducted; it was suspected that there might still be defects in the area connecting crack 2# and crack 3#. However, since the groove that has been dug is already quite large, if this groove were to be completed, a very large contraction force would be generated during welding, which could cause the unwelded areas to crack. Therefore, after consulting with Tianhua Company, it was decided as a precaution not to complete the groove at this time; instead, it will be inspected after welding, and if any cracks are found, the area will be dug open and repaired then. Based on the crack photos identified in the second UT inspection and grinding carried out by tracking the crack direction, the photos of crack propagation are shown in Figure 2. http://www.nmtech.com.cn/jishuwang/upload1/0806171511214898.jpg http://www.nmtech.com.cn/jishuwang/upload1/0806171511525089.jpg 3.2 Hardness and Metallographic Inspection 3.2.1 Hardness Inspection The hardness values of the welds, base metal, and heat-affected zone are shown in Table 3. http://www.nmtech.com.cn/jishuwang/upload1/0806171512241572.jpg 3.2.2 Metallographic examination: The microstructure before repair was ferrite + bainite (see Figure 5), while the microstructure of the weld was ferrite + granular bainite strands (see Figure 6). Microcracks could be seen in the coarse-grained region of the heat-affected zone, extending along the original austenite grain boundaries (see Figure 7) ； Microcracks were also found in the heat-affected zone of the stainless steel repair weld; these cracks extended along the original austenite grain boundaries, appearing as multiple parallel cracks with branches (see Figure 8) ； During the pre-welding inspection for the major repair in 2007, microcracks were also found in the base metal between the first and second repair welds on the cone; corrosion products could be observed within these cracks, and the cracks exhibited short-distance clustered propagation (see Figure 9). http://www.nmtech.com.cn/jishuwang/upload1/0806171513111549.jpg Based on the metallographic morphology, distribution, and intergranular propagation characteristics of the crack, this is a reheat crack; no abnormalities were found in the structure of the base material. 4 Crack analysis: The E53 equipment had been in use for only 3 years; leaks and fires were detected as early as 1999, and to date 4–5 cracks have appeared, indicating that the equipment has inherent defects. According to the article \"Cracking in Welds of 2‑1/4 Cr‑1Mo Ammonia Synthesis Towers\" by C. R. PRESCOTT, when welding low-alloy steel (2‑1/4Cr‑1Mo), transverse cracks often occur in the welds; these cracks are caused by hydrogen being absorbed by the alloy during the welding process. The excessive diffusion hydrogen content during the manufacturing process led to the formation of crack zones in the center of the weld. Since the microcracks that form in advance along the fusion line cannot be detected, during equipment operation these microcracks gradually merge with each other to form visible cracks, which eventually spread through the wall thickness. In the areas where weld toe cracks intersect and where pre-existing microcracks are present, the crack propagation accelerates. The nature of the cracks is a combination of cold cracks (hydrogen-induced cracks, weld toe cracks) and reheat cracks. To this end, the repair plan is formulated around the measures to be taken to prevent the formation of these cracks. During the previous crack repair attempts, the inspection methods were insufficient and not thorough enough; the heat treatment plan was not developed based on the actual conditions. As a result, the hardness of the welded joint exceeded the maximum value allowed for Cr-Mo steel. After the repairs, no effective inspections were carried out, and the cause of these weld defects was not identified. These defects persisted over time and became increasingly severe, resulting in rare defects that were large in area, depth, width, and length. 5 Tests and preparatory work before repair Before determining the repair plan, we carried out the following tasks in order to select the welding materials, preheating temperature, and heating method, as well as to understand external wall heating, the temperature difference between the inner and outer walls, and to train welders. 5.1 Selection of welding materials for repair: Tianhua Company uses E9015-B3 for its E53 material, while Lutianhua uses E9015-B3L for repairing H423 (the material of which is the same as that of E53). The welding materials provided by the Italian company BOSCO are also E9015-B3L. E9015-B3L has a carbon content of 0.05%, whereas E9015-B3 has a carbon content of 0.05% to 0.12%; as a result, E9015-B3L possesses better plasticity and toughness, as well as improved crack resistance. The strength of the E9015‑B3L weld metal is ≥550 MPa, which is comparable to that of E9015‑B3. Since the ASME 2001 standard has replaced E9015‑B3L with E8015‑B3L, and to prevent cold cracking, it is considered more appropriate to use E8015‑B3L. 5.2 Simulation tests: Tianhua Company provided the 10CrM0910 material from West Germany, in the form of thick-walled pipes with dimensions of φ410×60, for use in these simulation tests. (1) In an environment preheated to 250–300°C, base welding for adaptation training in welding 2–1/4Cr–1Mo steel was carried out using E8015–B3L electrodes. Meanwhile, methods such as TIG welding to build up material in gaps of about 20 mm and the use of backing plates were tested to determine the best approach for achieving a smooth transition of the weld bead on the inner wall. The specimen diagram for the simulation test is shown in Figure 10. http://www.nmtech.com.cn/jishuwang/upload1/0806171520554750.jpg (2) Conduct simulation tests on the heating process during heat treatment. According to the article \"Formation of Weld Cracks in the Outlet Pipeline of Ammonia Synthesis Towers,\" they conducted heat treatment tests on 10CrM0910 steel pipes with dimensions of φ329.9×31 using external electric heating plates. Its heat treatment specifications are as follows: a. The two ends of the tube are open, allowing air to circulate naturally; the outer wall temperature is 700 ℃, and the inner wall temperature is 640 ℃, resulting in a temperature difference of 60 ℃. b. Closed at both ends (repeated 3 times), outer wall at 700 ℃, inner wall at 670 ℃, temperature difference of 30 ℃. c. Insulation for internal surface welds (repeat 3 times): 700 ℃ for the outer wall, 680 ℃ for the inner wall, with a temperature difference of 20 ℃. d. The position of the pipeline, the insulation time, and changes in heating rate are all ineffective. The hardness of the cracked weld was measured: the hardness of the inner wall was 360–370 HV, while that of the outer wall was 220–280 HV. According to available data, when the hardness is ≥250 HV, the material becomes highly sensitive to hydrogen-induced cracking. To this end, we conducted a heat treatment test by maintaining the temperature at 730 ℃ for 4 hours, and measured the temperature difference between the inner and outer walls; it should be less than 40 ℃. The test employed external surface heating of the annular seam for a temperature difference experiment: (1) The thick-walled tube was heated on one side while the back side was exposed to atmospheric conditions, and the temperature difference between the two sides was measured ； (2) Seal both ends of the tube with plates to create a dead air layer, and measure the temperature difference between the two sides ； (3) Seal the inside of the tube with silica cotton and measure the temperature difference between the front and back sides. The test results for the temperature difference between the inner and outer walls of the pipe after heat treatment are shown in Table 4. The hardness measurement results of the specimens after post-weld heat treatment are shown in Table 5. http://www.nmtech.com.cn/jishuwang/upload1/0806171521372392.jpg As can be seen from Table 4, in the test where the inner wall is in direct contact with the atmosphere, the temperature difference between the inner and outer walls is too large; this affects the softening effect achieved through heat treatment of the inner wall, and may result in the hardness of the joint being higher than the specified value of ≤HV225. When the hydrogen content in the weld exceeds the critical level, the likelihood of cold cracks and reheat cracks occurring at stress concentration areas and the weld root increases. It can be seen that the temperature difference between the inner and outer walls varies under different conditions. Moreover, the tests were conducted on pipes with small diameters; the equipment in question has a larger diameter and thicker walls, which results in a greater temperature difference. The temperature of the inner wall is too low, and heat treatment alone is not sufficient to meet the requirements for softening the welded joints. To this end, it is strictly required that Tianhua Company isolate the equipment from the system (cut off the nitrogen) before starting to repair this weld, in order to create a dead air layer and ensure that the temperature difference between the inner and outer walls of the equipment is less than 30°C. The hardness test results showed that the HB value in all areas was ≤225. It is normal for the hardness of the weld metal to be higher than that of the base material; the E8015B3L grade selected is appropriate, and the heat treatment parameters are also suitable. However, when welding using the growth method, the weld develops an inward depression, with a severity of about 7–8 mm. The inner wall is rough, with excessive unevenness ； By using shims, the welding quality is good and it is easy for welders to operate, making it suitable for construction at maintenance sites. 5.3 Determination of heat treatment temperature http://www.nmtech.com.cn/jishuwang/upload1/0806171522288341.jpg In previous maintenance plans, the heat treatment temperature was set at 690°C. Based on our tests regarding weld heat treatment, the temperature inside the shell at this level could fall within the critical range of 500–650°C for reheat cracking, making it prone to such cracks. Therefore, it is necessary to change both the heat treatment temperature and duration in order to prevent reheat cracking. Based on the above test results: 730℃±14℃×4h is reasonable, and the various properties of the welded joint are fairly satisfactory. To this end, we originally recommended 730°C ± 14°C for 4 hours; only when both ends are properly sealed during weld heat treatment can it be possible to raise the inner wall temperature above the critical temperature range for reheat cracking, thereby achieving the goal of softening the joint. 6 Defect repair: The repair steps are shown in Figures 11 and 12. http://www.nmtech.com.cn/jishuwang/upload1/0806171529392096.jpg The welding groove should be shaped according to the inspection markings; rust and scale within 20 mm of both sides of the groove must be removed thoroughly. Color penetrant testing is performed on the groove of the repaired area; the acceptable standard is PTI grade as specified in JB4730–2005. Welding is carried out according to Figure 12, with 5 heat cycles. After welding, grind the surface of the weld flat. During the entire welding process, the following should be noted: (1) Preheating before welding, with a preheating temperature of 250–300°C. (2) Heating and insulation method: It is wrapped with electric furnace wire on the outside for external heating; the heating plate covers the entire seam area as well as a 1-meter width on each side of it, with insulating cotton applied on the outside. (3) Due to the large extent of the defect, the total welding length required was 1340 mm, which includes a penetration crack that is 800 mm long, 80 mm wide, and 80 mm deep, plus the grinding lengths before and after the crack. To this end, a 4 mm thick 10CrM0910 gasket is first placed at the bottom of the groove, followed by surfacing welding on both sides of the groove. While performing surfacing welding, strike with a hammer to gradually reduce the gap at the groove; once the gap is less than 8 mm, use TIG welding to weld the two gaskets together thoroughly. The stress distribution during surfacing is relatively simple, being plane stress. Hamming causes the weld metal to elongate, which can eliminate 40% to 60% of the internal stress. Although the weld coefficient for the gasket plate weld is 0.9. But using it on circumferential seams will not affect the strength. Five heat cycles are required because the base material of this type has a large thickness and a large amount of filler metal, so stress relief treatment must be carried out in stages. (4) If welding is suspended for some reason during the welding process, the temperature should be raised to 350°C for dehydrogenation; when welding resumes, the interpass temperature should be reduced to 300°C. (5) Welding shall not be carried out until the preheating temperature is reached; moreover, the interpass temperature should be maintained at 250–300°C. During welding, dedicated inspectors should be assigned to conduct supervision and inspection using infrared thermometers. (6) The welding rod drying temperature is 400–420°C, with a holding time of 2 hours. The electrode cylinder should first be dried by electric heating; once the welding materials are taken out of the oven, they should be immediately placed into a pre-dried and warm electrode holder, after which the lid of the holder should be closed. Upon arriving at the welding site, weld one electrode; once one electrode is used, immediately close the cylinder lid. Prevent atmospheric moisture from entering the cylinder; keep the heating element powered on to maintain warmth in the cylinder whenever welding is paused. (7) The weld bead should be formed through multiple layers and passes of welding; minimize lateral movement of the welding rod, and control the welding heat input as much as possible while ensuring good fusion, in order to reduce welding stress. (8) After welding each rod, immediately use a small air gun to strike the weld metal until small pit-like defects are formed; however, no striking is performed on the root coat or the top coat. (9) Welders must be highly responsible; they need to thoroughly clean the slag from each weld seam and check for any defects on their own. Only after confirming that there are no issues can they proceed to weld the next seam. (10) Welding is generally not permitted in rainy weather or when the environmental humidity is above 80%. If welding must be carried out under such conditions, measures must be taken, such as setting up a shelter or using heating to keep the humidity inside the shelter below 80%, before welding can proceed. After the first heat cycle of patch welding, a UT re-inspection was conducted, and cracks with a depth of 45–50 mm and a length of approximately 60 mm were found at 2–3 locations where the defects had been repaired (at the originally suspected defect sites). To this end, the above process was repeated to repair this area; in order to eliminate the weld defects that had been repaired in 1999, the repair area was expanded to a width of 80 mm, a depth of 80 mm, and a length of 360 mm. Perform a second patch weld; the location is shown in Figure 13. http://www.nmtech.com.cn/jishuwang/upload1/0806171532591838.jpg When performing patch welding on defects, the above process requirements shall still be followed (only the heat treatment step is omitted when the welding thickness reaches 20 mm) for both patch welding and heat treatment. 7 Precautions for heat treatment: During the heat treatment process, attention should be paid to the temperature, and it is essential to avoid staying in the range of 500–650°C. This temperature range is the critical temperature range for reheat cracking (where reheat cracks are likely to occur). For this reason, as a precaution, it is advisable to use two layers of insulation when using a coiling furnace for partial heat treatment. If it is found that the temperature does not rise, the power can be turned off, the insulation removed, and once the temperature drops below 500°C, the issue can be resolved before heating up again. Since it is not possible to apply heating measures inside the equipment, it is necessary to seal the equipment and find ways to measure the temperature in the weld areas on the inner walls, in order to control changes in the hardness of those inner walls. 8 Flaw detection results: (1) Ultrasonic testing: Ultrasonic testing is carried out after the ring joints have been rewelded and heat-treated; the acceptable grade is Level I according to JB4730–2005. (2) Magnetic particle testing: After welding repairs and heat treatment, the welds are subjected to magnetic particle testing to detect surface and near-surface cracks (including areas beyond the detection range of other testing methods). The required quality standard is Grade I according to JB4730–2005. (3) Hardness testing and metallographic examination: The results of the hardness testing are shown in Table 7. http://www.nmtech.com.cn/jishuwang/upload1/0806171533311500.jpg After several previous weld repairs, the hardness in Area 2 was too high, reaching up to HB=333; this was one of the main reasons for the continuous expansion of cracks. This indicates that the heat treatment did not achieve its purpose. During this repair process, the hardness measurement values (of the weld, heat-affected zone, and base metal) after the first repair were quite satisfactory. However, after the second repair, although the hardness value remained within the required range, the hardness of the base material had dropped to near the lower limit. (4) The areas inspected for metallographic coating include the weld zone, heat-affected zone, and base metal of the repair welding area. Metallographic photo of the final state after this repair: http://www.nmtech.com.cn/jishuwang/upload1/0806171534236436.jpg Inspection of the metallographic structure after re-coating revealed no hardened areas. The conclusion is that the repair quality is satisfactory. 9 Review and Explanation (1) Due to the special structural location of this weld, it is not possible to determine the nature of defects using RT inspection. Using UT for inspection allows only single-sided testing; there are large blind spots and the inspection is not thorough. In this repair process, only those defects that can be detected can be fixed. (2) Since this equipment has undergone multiple incomplete repairs, there may be many microcracks on the inner surface of the weld joints (as indicated by the preliminary UT inspection results). This repair involved thorough treatment of all the defects detected; large through-holes with a width of 80 mm, a depth of 80 mm, and a length of 800 mm were created. The total length of the repair welds was 1340 mm, and the amount of weld filler used was approximately 60 kg (120 kg of welding rod was utilized), which is unprecedented in terms of manual repair work for such defects. Although the weld met the acceptance criteria in both UT and MT inspections, it was also subjected to stress relief and softening treatment. However, the weld metal has a cast microstructure, which always represents a weak point in the welded joint. (3) Post-weld heat treatment can only be carried out as localized treatment on one side, which easily leads to uneven temperatures between the inner and outer walls. When the temperature of the inner wall falls within the dangerous range of 500–650°C and the holding time is long, microcracks may appear at the stress concentration points on the inner wall weld toe. During this repair process, during the dehydrogenation and denitrogenation treatment, the temperature could not be increased; the outer wall temperature reached 540°C while the inner wall temperature was only 250°C. During preheating before welding, the outer wall temperature reached 450°C and the inner wall temperature remained at 250°C. During heat treatment after surfacing, the temperature of the outer wall reached 650°C but failed to rise any further. After searching for the cause, it was discovered that there was condensate water in the system, which prevented the temperature from increasing; the wall remained in this zone at risk of reheat cracking for 8 to 9 hours. The fact that it is not possible to detect the presence of reheat cracks under current conditions is a matter of concern. Fortunately, our preheating temperature is high, which significantly reduces the tendency for reheat cracks; generally, a preheating temperature in the range of 200–450°C is quite effective in reducing reheat cracks. According to the information provided, exceeding a certain preheating temperature can prevent the occurrence of reheat cracks. For this repair, we kept the preheating temperature above 250°C. The hardness value obtained after this repair is much lower than the required HB≤225 for Cr–Mo steel. As the hardness of the weld decreases, the tendency for reheat cracking also decreases. The use of a low-strength electrode E8015-B3L during welding improves the plasticity of the weld metal, thereby reducing its sensitivity to reheat cracking. (4) For the above reasons, this repair can only ensure that the repaired parts meet the requirements under the current testing conditions. It is hoped that inspections of the equipment during operation will be intensified, and UT and MT tests should be conducted on this area again during each major overhaul to ensure safe operation. (5) After the final heat treatment, the hardness value was checked and found to be HB=132 on the base material side of the cylinder. When repairing this circumferential crack in the future, in order to prevent the hardness of the base material from dropping too much, the holding time during heat treatment can be determined based on the thickness of the weld metal. This reduces the heat retention time, thereby slowing down the rate of decrease in the hardness of the base material. Furthermore, the temperature of the intermediate heat treatment can be reduced to the stress-relief temperature, which is beneficial for improving the strength and temper brittleness of the base material.