ADVANCED INSPECTION TECHNIQUE FROM INSIDE SURFACE FOR DISSIMILAR METAL WELD OF STEAM GENERATOR NOZZLE

Transcription

1 More Info at Open Access Database ADVANCED INSPECTION TECHNIQUE FROM INSIDE SURFACE FOR DISSIMILAR METAL WELD OF STEAM GENERATOR NOZZLE T. Yamaguchi, T. Shichida, T. Matsuura, Y. Tsuruta, S. Kawanami Mitsubishi Heavy Industries, Japan 1. Abstract In 2012, several axial flaws were not detected by Ultrasonic Testing (UT) from outside surface of nozzle. The flaws were located in Dissimilar Metal Weld (DMW) of an inlet (hot leg) Steam Generator (SG) nozzle at North Anna Power Station Unit 1. One of the causes of missing flaw seems to be particular geometry of nozzle outside surface with large taper angle (1). 100% coverage by UT from outside surface was very difficult, because there are some areas where ultrasonic beam cannot reach. On the other hand, the inspection from inside surface of nozzle has a merit uninfluenced by the nozzle outside surface geometry. Therefore, we have developed an advanced inspection technique from nozzle inside. This technique mainly consists of a new robot to access from inside surface and a new inspection technique with accurate depth sizing. The robot has high performance movement characteristic and easy installation structure which make it easy to apply not only UT but also Eddy Current Testing (ECT) and Visual Testing (VT) to DMW of SG nozzle remotely. This robot with state-of-art robotic technology has already been applied in the field service in Japan successfully. Up to date, accuracy in UT depth sizing from defect opening surface had been insufficient, so we developed both small Transmit-Receive type with L-wave Phased Array (TRL-PA) UT probe and matrices type with L-wave PA-UT probe. As results, it is verified that our advanced UT improves defect sizing capability remarkably even for a swallow defect (2) (3). In this report, we will introduce our new solution from nozzle inside surface and show experimental results and some field experiences. 2. Background To prevent the occurrence of PWSCC, Ultrasonic Shot Peening (USP) was applied to SG nozzle in Japan. Before application of peening, ECT for soundness check was applied. In these inspections, some axial PWSCCs were detected on the nickel alloy surface (Alloy 600) of SG nozzle in some plants (shown in Figure 1). After detection of PWSCC, Nuclear and Industrial Safety Agency (NISA) requested the application of ECT from inside surface for DMW of SG nozzle once in decade as In-Service Inspection (ISI). On the other hand, in 2012, 5 Primary Water Stress Corrosion Crackings (PWSCC) were detected in the DMW in an inlet (hot leg) SG nozzle at North Anna Power Station Unit

2 These PWSCCs were not detected by primary UT inspection from outside surface. However after outside surface machining in preparation for a full structural weld overlay, PWSCCs were founded by visually. One of the potential causes for the miss of the detection of these PWSCCs seems to be large taper angle of safe-end configuration (see Figure 2). 100% coverage by UT from outside surface was very difficult, because there are some areas where ultrasonic beam cannot reach. On the other hand, the inspection from inside surface of nozzle has a merit uninfluenced by the nozzle outside surface geometry. Therefore, Advanced inspection technique from inside surface for DMW of SG Nozzle have been developed to fulfill the ISI request in Japan. This technique is also effective for the case like North Anna Power Station Unit 1. This technique mainly consists of a new robot to access from inside surface and a new inspection technique with accurate depth sizing. (see Figure 3). Figure 1 Location of defect at DMWs of SG Buttering Welding DMW configuration consists of the 11-degree outside-surface-tapered safe-end Inlet Nozzle Steam Generator Cross-Section view of inlet nozzle (Ref. Pacific Northwest National Laboratory Web site) Figure 2 Safe-end configuration of North Anna Power Station Unit 1 SG 565

3 Advanced Robot for DMWs of SG nozzle Inspection sensors SG nozzle Figure 3 Advanced Robot for DMW of SG nozzle 3. Robot Technique 3.1 Target and feature of Advanced Robot for DMWs of SG nozzle design The Advanced Robot was designed to achieve 3 functions for domestic inspection requirement such as quick installation, accurate coverage of inspection area and multiple inspection techniques. Details of these functions are shown in Table 1. Table 1 Target and feature of Advanced Robot design Quick Installation Target Installation through primary manhole Feature Module structure Special tools for installation from outside of manhole Self moving function on tubesheet Accurate Coverage of Inspection Area All circumference inspection High accuracy inspection 100% coverage of inspection area Accurate tracking with auto rectification of sensor position by laser measurement Multiple Inspection Technique Application of ECT, UT, and VT Retouchable sensor head 566

4 3.2 Quick Installation To achieve the quick installation, the Advanced Robot consists of four modules. Each module is lightweight, so easy assembly is possible. Moreover, by using of the special tools for installation from outside of manhole, working in channel head is unnecessary. Outline of assembly procedure is shown in Figure 4. After the Advanced Robot is attached on the installation position, the Robot has to move to the inspection position by itself. Therefore it has moving capability on tube-sheet remotely by using MHI's SG tube robot technology. The outline of Advanced Robot movement is shown in Figure 5. Manipulator Base Figure 4. Outline of assembly procedure of Advanced Robot (4) Manipulator (Root part) Sensor attachment Figure 5 Outline of Advanced Robot movements Manipulator (Top part) Move on tube-sheet Install from outside in channel head Inspection position 567

5 3.3 Accurate Coverage of Inspection Area To achieve 100% coverage of inspection area and accurate tracking, seven axes manipulator was adopted. This manipulator has the wide movable range and high flexibility so that it is possible to control the sensors smoothly like a human s arm. Moreover, by laser measurement instrument, the distance between nozzle inside surface and the sensor attachment is kept constant and the forcing load to sensors is also controlled even for slight distortion on nozzle with auto rectification. The apparatus of the seven axes manipulator is shown in Figure 6. Laser measurement instrument Sensor attachment Figure 6 Apparatus of seven axes manipulator 3.4 Multiple Inspection Technique In inspection, ECT is first applied to detect flaws. When flaws are detected by ECT, VT will be applied to check the characteristics of them and finally UT will be applied for depth sizing. Thus, the Advanced Robot can carry the ECT, UT and VT sensor head respectively. Attachments of each inspection technique are shown in Figure 7. Sensor exchange can be performed without removing a robot from channel head. The time for sensor exchange is approximately 3 minutes. Camera Tip Probes Attachment for ECT and UT Attachment for ECT and UT Attachment for VT Attachment for VT Figure 7 Attachment for each inspection technique 568

6 4. ECT Detection Technique ECT is applied to check the soundness of nozzle weld. Both normal and magnetic biased ECT probes with cross-coil are used (see Figure 8). The magnetic biased ECT probe can decrease noises due to permeability change between safe-end weld and stainless overlay. The results of EDM notch mock-up test are shown in Figure 9 and 10. All EDM notches including EDM notches with 0.5 mm depth are detected. Moreover, SCC mock-up test was performed. The result showed 0.5mm depth SCC is detectable (see Figure 11) (5). Normal ECT probe Magnetic biased ECT probe Figure 8 Normal and magnetic biased ECT probes with cross-coil SUS EDM notches dimension Depth / Length / Width TT600 weld material Buttering Overlay Figure 9 Layout of EDM notch on mock-up Notch depth is 0.5mm Notches on buttering TT600 weld material Buttering Notch depth is 0.5mm Overlay Results of Normal ECT probe Results of Magnetic biased ECT Figure 10 ECT Results of mock-up test 569

7 0.5mm Figure 11 Cross-section of SC on mock-up 5. Field Application of Robot Technique with ECT Since 2010, MHI has applied the Advanced Robot successfully in ISI of Alloy 600 DMWs on SG nozzles of 2 PWR plants as shown in Table 2. ECT was performed in ISI and no indication was verified. In future, the Advanced Robot will be applied for about 4 plants per year. Table 3 shows the typical schedule of ECT application in ISI. Installation of Advanced robot can be carried out in only one day. Then, critical schedule can be finished in about 3 and half days. The work in the inside of channel head is not required by introduction of the Advanced Robot. Therefore, as compared with the case using old type equipment for USP, numbers of total workers have been reduced by about 50%. In addition, amount of radiation exposure reduced by about 30% compared with the case using old type equipment for USP. Table 2 Experiences of ISI of SG nozzles Unit Type Data Plant A Plant B PWR 2 - Loops PWR 2 - Loops 09/ /2011 (a) (b) Table 3 Typical schedule of ECT by Advanced Robot (days) Work Installation of NDT Robot to channel head Inspection (Data acquisition / Analysis) (c) (3)(d) (4) Tear down Estimated Condition (a) Schedule is daytime shift (b) Equipment conveyance is excepted (c) Inlet nozzle only (d) When there is a problem in data, re-inspection will be carried out 570

8 6. UT Depth Sizing Technique In order to improve depth sizing of shallow and deep defects, the PA-UT technique was developed. A large size matrix PA-UT probe was developed for depth sizing of deep defects. This probe can utilize 3-Dimensional beam scan. Small size TRL PA-UT probe was developed for depth sizing of shallow defects. This probe can utilize a large refraction angle. Figure 12 shows these two PA-UT probes and two conventional probes with different refraction angles used in ISI for the comparison of depth sizing test. The depth sizing test was performed using artificial SCCs. In this test, a total of 17 test pieces were made. After ultrasonic inspection, all test pieces were cut and the depth size and shape of cracks was measured. Figure 13 shows a representative artificial SCC. Results of PA UT and conventional UT probes are shown in figures 14 (a) and 14(b) respectively. The conventional UT showed a good sizing performance in depth region less than 15mm, but it became extremely degraded in around 20mm depth region. On the other hand, the PA-UT technique showed a better accurate sizing result up to around 40mm depth region. The Root Mean Squire Error (RMSE) of the conventional UT was 9.35mm, the RMSE of the PA-UT with small TRL PA and Matrix PA is 2.54 mm. From these results, it is verified that the newly developed PA-UT achieved conquest of the weak point of the conventional UT and has adequate sizing capability up to approximately 40mm depth. The example of Advanced PA UT sizing result is shown in Figure 15. Just for information, volumetric inspection by normal UT probe and angle UT probe is also applicable. T70, R50 T50, R40 Matrix PA probe Specs: 2MHz-96ch Apply to: Middle to deep defects Small type TRL PA probe Specs: 5MHz-40ch Apply to: Shallow defects Conventional TRL probe Specs: 3MHz-TR Apply to: All defects Figure 12 UT probes LAS 低合金鋼側 Y=0 ステンレス側 Y+ 方向 SUS 最大深さ :23.1mm Max depth: 23.1mm 位置 :X=0 Y=0 Figure 13. Representative artificial SCC 571

9 50 40 縦波前後分割 Conv. TRL (T70, 70 R50) R Small 小型 RL PA type TRL PA Matrix マトリックス PA (for PA middle 浅い欠陥用 defect) マトリックス Matrix PA (for PA large 深い欠陥用 defect) Evaluation 評価深さ depth m (mm) Evaluation 評価深さ depth mm (mm) Evaluation using 面エコーによる評価 surface echo Actual 実深さ depth mm (mm) Outline mark means conventional SCC Actual 実深さ depth mm (mm) Conv. TRL T70, R50 Advanced method combined Small TRL PA with Matrix PA RMSE 9.35 ( 1.36 at <20mm) RMSE 2.54 (a) Conventional method (b) Advanced method Figure 14 Sizing accuracy of SCC Example of Depth Sizing / Actual depth: 22.9 mm Conventional Method Advanced Method Probe scanning position Probe scanning position Wall thickness Impossible depth sizing (Interim value : 18.8mm) Wall thickness Evaluated depth : 22.3mm Figure 15 Example of Advanced PA-UT sizing result 7. Field Application of UT The PA-UT method is designed to be used with the Advanced Robot. However, we had field experience of this method for Reactor Vessel (RV) outlet nozzle inspection in 2011 introduced below. Two defects were detected and the PA-UT method which is described in section 6 was carried out for depth sizing. Figure 16 shows the location of the detected defects. Estimated depths of defects are less than 3mm and 4.7mm respectively. UT results (B-scope) for each defect is shown in Figure 17. The actual depths of defects estimated in 2.7mm and 4.8mm respectively as the grinding process of the nozzle for repair. This shows that deviation between estimated and actual depth of defects is very small. Thus, effectiveness of the PA-UT method in power plant inspection was proven. 572

10 View from R/V side (Top shall be 0 degree) R/V R/V Stainless overlay Outside Outside Buttering portion Outlet nozzle Inside Inside (low alloy steel) approx.75mm Circ weld portion (Weld portion : 600 series Ni base alloy) Safe-end end MCP MCP Figure 16 Location of SCC indications in Outlet nozzle of RV Comparison with Estimated depth and Actual depth No.1 Indication No.2 Indication Advanced method (Small TRL-PA) Probe scanning position Probe scanning position Wall thickness Wall thickness No indication No (( < 3mm ) ) Actual Depth : 2.7mm Evaluated depth : : 4.7mm Actual Depth : 4.8mm Figure 17 UT Sizing Result of Outlet DMWs 8. Conclusion Advanced Robot for DMWs of SG nozzle and PA-UT (inspection technique from inside surface) have been developed. Advanced Robot with state-of-art robotic technology has been applied in the field service in Japan successfully. The PA-UT has been also applied to the RV outlet nozzle and effectiveness was proven by field experiences. These techniques will be effective ISI method for particular geometry of nozzle outside surface with large taper angle. MHI will intend to introduce these advanced techniques to contribute to the maintenance of power plants hereafter. 573

11 Reference 1) MT Anderson, AA Diaz, SR Doctor, Evaluation of Manual Ultrasonic Examinations Applied to Detect Flaws in Primary System Dissimilar Metal Welds at North Anna Power Station, PNNL-21546, Pacific Northwest National Laboratory, ) T. Yamaguchi, T. Kawashima, M. Takatsugu, K. Unate, Advanced NDT Robot for Dissimilar Metal Weld of Steam Generator Nozzles, Proceeding of the 9 th International Conference on NDE in Relation to Structural Integrity for Nuclear and Pressurized Components, pp , May ) S. Kawanami, T. Kimura, M. Kurokawa, J. Nishida, T. Yamaguchi, T. Matsuura, T. Tsuruta, Advanced Phased Array UT Sizing Technique on Stress Corrosion Cracking in Dissimilar Metal Welds of Nozzle, Proceeding of the 9 th International Conference on NDE in Relation to Structural Integrity for Nuclear and Pressurized Components, pp , May ) Y. Kohata, J. Fujita, K, Onishi, H. Tsuhari, F. Hosoe, The Application of Manipulator Robot for Nuclear Plant Maintenance, Preprints of 7th Annual Conference of the Japan Society of Maintenology, Omaezaki, pp , July (in Japanese) 5) Y. Asada, K. Tokuhisa, M. Takatsugu, K. Kurokawa, K. Kawata, N. Hirano, T. Sera Development of Eddy Current Testing Method for PWR Vessel s Dissimilar Metal Weld, Japan Society of Maintenology, Maintenology, Vol.6-No.4, pp.38-43, (in Japanese) 574