【HaiChuan Pipeline Discussion】What are the consequences of welding dissimilar steels in the steam pressure sampling pipe of high-pressure control valves?
Thread Content
Breakage of the weld in the steam pressure sampling tube of the high-pressure control valve in a unit at a power plant in Guangdong. I. Course of events: At 17:27 on March 30, 2022, Unit 2 of a power plant in Guangdong was operating at a load of 530 MW. The operators monitoring the plant’s operation heard a sharp hissing sound of steam leakage; upon on-site inspection, it was found that high-pressure steam was leaking from the area around the turbine head of Unit 2. The load was immediately reduced, and an isolation barrier was set up at the site. Due to the large amount of leakage, it was not possible to approach the area to determine the exact location of the leak. To prevent the incident from spreading further, an application was made to the control center and, upon approval, the No. 2 turbine was shut down at 18:13, causing the generator to be disconnected from the system. II. Inspection and Handling Results At 19:03 on March 30, maintenance personnel conducted an on-site inspection and confirmed that the weld at the fitting and pipe seat of the steam pressure sampling tube of the high-pressure control valve in Unit 2 was broken, as shown in Figure 1. This measurement point is not related to protection; to restore Unit 2 to operation as soon as possible, a sealing solution was adopted. At 15:45 on March 31, the blocking was completed, as shown in Figure 2. At 01:39 on April 1, with the approval of the dispatch team, generator No. 2 was reconnected to the grid. Upon reviewing previous maintenance and inspection reports, no records of inspections for this weld were found, indicating either missed inspections or delayed testing. Paragraph 8.2.1 of the Metal Surveillance Procedure stipulates that during each Class A or Class B maintenance task, a random inspection should be carried out on pipe supports such as those for small-diameter pipes at a rate of not less than 20%; at least 5 such inspections shall be conducted, with 100% inspection to be completed after 100,000 hours of operation. III. Cause Analysis 1. The structural design was not standard: the material of the pipe socket was 12Cr2Mo1, the material of the valve body was 12Cr2Mo1+ZG15Cr1MoV, and the material of the tees was 304 stainless steel. The tees and the pipe socket were connected by insertion, with welding between different types of steel – a design that was clearly unreasonable. This led to defects such as incomplete welding at the root, which was the main cause of the accident that occurred. 2. The welding material for this dissimilar steel joint is 304 (Appendix B of the Welding Techniques Specification for Thermal Power Plants requires that nickel-based welding materials be used for such welds); the root depth of the weld is insufficient, and the thickness of the weld across its entire length is uneven, as shown in Figure 3. 3. Poor installation quality during the infrastructure construction period; the welds have incomplete penetration at their roots ； In recent years, the units have frequently undergone deep-depth load regulation as well as frequent startups and shutdowns, which has exacerbated vibration-induced fatigue cracking damage ； During operation, the pipe seat experiences fatigue damage due to vibration in conjunction with the main steam valve; prolonged vibration at high temperatures exacerbates this fatigue damage until fracture occurs.
Problems identified: 1. The inspection work was not thorough, resulting in gaps in the identification and resolution of potential hazards. In accordance with the requirements set forth in the group company’s notice regarding improvements to the structure of the connection fittings for high-temperature steam samplers in thermal power units, efforts should be made as soon as possible to modify those samplers that use an insert-type design. The relevant departments failed to identify potential hazards in this area and carry out necessary modifications; it has now been confirmed that all four units utilize an insert-type structure in this regard. 2. The coordination among specialized roles is not tight enough. Before the primary valve, the pipelines and welds for pressure sampling are under the management and maintenance of the mechanical engineering team, while those behind the primary valve are under the responsibility of the thermal engineering team. The mechanical engineering team does not have a comprehensive understanding of the layout of the sampling pipelines; moreover, the records related to the inspection of these pressure sampling pipes are incomplete, resulting in missed inspections. 3. The lessons learned from previous leaks in the external pipes and those outside the furnace were not taken into account; training was inadequate, and preventive measures were not properly implemented. On April 22, 2019, the weld at the front of the pressure sampling port of the fourth high-pressure steam conduit on Unit 1 broke down; this was also due to the fact that the weld was of a butt-weld structure, with incomplete welding at the weld root. Steam leakage led to a shutdown, and this hidden defect was not detected during the investigation. V. Rectification Measures 1. Conduct a thorough inspection of the structure, material, and welding techniques of all pressure sampling pipes, temperature sleeves, and drain pipes in the four units, by referring to the equipment drawings and the list of thermal measurement instruments; also conduct a comprehensive check of the pipe bases of sampling pipes of the same type and structure, and formulate plans for metal testing and rectification. 2. Formulate construction plans in strict accordance with process requirements, and strengthen the control of welding processes and quality management on site. 3. Implement a rolling inspection plan, employing various non-destructive testing methods to inspect the welds of the temperature sleeves. 4. Inspections revealed that similar structures still exist in the pipes connecting the turbine body and the valve bodies; further discussion is needed with the turbine manufacturer regarding improved designs for such structures. Unit 1 has used the unit adjustment window to cut out the weld of the sampling pipe there and temporarily seal it off ； Inspection of this area in Unit 3 showed welding using the same material, and no issues were detected during phased array inspection.