ML043440071
| ML043440071 | |
| Person / Time | |
|---|---|
| Site: | Dresden  |
| Issue date: | 12/02/2004 |
| From: | Exelon Nuclear |
| To: | NRC/FSME |
| References | |
| Download: ML043440071 (37) | |
Text
1 Generator Rotor and Steam Dryer Meeting Dresden Nuclear Power Station December 2, 2004
2 Opening Remarks Danny Bost Site Vice President Dresden Nuclear Power Station
3 Purpose
- Discuss cause of generator rotor cracking, corrective actions, and monitoring plans
- Discuss steam dryer inspections, modifications, and causal factor analysis
- Outline the basis for extended power uprate (EPU) operation
- Describe ongoing actions and proposed NRC interactions
4 Agenda
- Opening Remarks Danny Bost
- Generator Rotor Mark Kanavos
- Steam Dryer Jim Meister
- Background Information
- Inspection Results
- Quad Cities/Dresden Comparison
- Causal Factor Analysis
- Ongoing Actions
- Planned Interactions
- Bases for EPU Operation
- Closing Remarks Danny Bost
5 Generator Rotors
- We understand the causes for generator rotor cracking
- We have repaired the Dresden Unit 3 (D3) generator rotor and are completing repairs for Dresden Unit 2 (D2)
- We have developed a monitoring plan
6 Steam Dryers
- Recent inspections and analytical results confirm lower dryer loads at Dresden compared to Quad Cities (QC)
- Cover plate cracking on D3 dryer occurred in an area previously identified as a high stress location in analytical models, and was subject to cold spring
- Cover plate cracking for both units initiated at sites with evidence of lack of fusion (root cause ongoing) - corrective actions preclude future deficiencies
7 Steam Dryers (cont.)
- Structural improvements through the design modifications provide additional confidence for both Dresden units
- Substantial evaluation efforts continue - modeling capabilities continue to improve
- Both units can safely return to full power operations
- Propose ongoing meetings with NRC to discuss progress and results
- Fall 2005 meeting following the D2 outage will include discussion of impact on D3, including a potential mid-cycle inspection
8 Generator Rotor Mark Kanavos Site Engineering Director Dresden Nuclear Power Station
9 Generator Rotor Purpose
- Discuss rotor issue timeline and inspection findings
- Discuss root and contributing causes
- Oscillating torsional loads (cause)
- Fretting due to keyway design (contributing cause)
- Summarize corrective actions for Dresden
- Describe monitoring to determine source of oscillating loads
10 Generator Rotor Issue Timeline May 2004 - D2 bearing vibrations begin adverse trend June 2004 - Vibration monitoring equipment connected to D2 July 2004 - D3 vibration recognized as increasing since May August 2004
- X-Y proximity probes installed on D2 and D3
- Flux probe and soft foot tests performed on D2
- D2 shutdown to repair cracked foot support rail September 2004
- Thermal sensitivity test performed on D3
- Flux probe test performed on D3
- D2 and D3 foundation surveys conducted
- Vibration monitoring equipment connected to D3
- D2 shutdown to correct soft feet and align machine October 2004
- D3 shutdown for Fall 2004 refueling outage - 13 crack found at D coupling
- D2 subsequently shutdown and a similar 10 crack was found
11 Generator Rotor 2004 Rotor Inspection Findings
- Fretting at initiation site weakened area in coupling
- Original coupling design produced stress risers
- 45° angle of crack indicates torsional loads as the cause of crack growth
- Metallurgical evaluation indicates that crack growth started and stopped ~200+ times
- 6 months of increasing vibration suggests intermittent torsional loading
- Conclusion - crack propagation was caused by intermittent oscillating torsional loads above the material fatigue endurance limit
- Results to be shared with fleet and industry
12 Generator Rotor Intermittent Oscillating Torsional Loads
- D2 and D3 cracks show >200 beach marks
- Each beach mark indicates an occurrence that the fatigue limit was exceeded by torsional loads
- If the oscillating torsional loads were applied constantly, then the shaft would have failed in less than 2 days
- Potential causes include: switchyard events, breaker reclosures, line faults, and cycling of large loads
13 Generator Rotor EPU Impacts
- EPU did not cause the rotor cracking
- 10% torque increase from EPU
- Pre-vs. post-EPU torque capability of coupling shrink fit was acceptable
- Torque would not cause fretting by itself
- Steady state loads did not drive the crack
- Shaft would fail in <2 days if steady state loading were the cause
14 Generator Rotor 2004 Rotor Repairs
- Removed cracked end of the rotor shaft and welded in new stub-shaft
- Re-designed rotor shaft keyway to eliminate the stress risers
- Increased shaft torsional capacity with improved coupling shrink fit
15 Generator Rotor Going Forward Monitoring Plans Transient Torsional Vibration Monitoring System (TTVMS) was installed on both units
- Monitors each phase of generator current and voltage
- Monitors turbine speed at the front standard and D coupling
- If a torsional event occurs, TTVMS will record data
- Event data will be used to calculate individual and cumulative shaft life usage for each event for comparison to acceptance criteria High speed monitoring equipment will be installed in the switchyard to monitor generator output buses for feedback from the transmission system
- Work with transmission system operator to identify source(s) of events
- Eliminate source(s) of loading or modify plant to correct the imbalance Monitoring will allow Exelon to be proactive in addressing vibrations before cracking develops
- Technology has been successfully used elsewhere
16 Generator Rotor Summary
- Troubleshooting actions were thorough
- Cracked rotor was identified as a possible cause of vibration
- Root cause analysis is comprehensive
- Shaft and keyway design have been improved
- Cracks were caused by oscillating torsional loads above the fatigue endurance limit of the material
- Torsional and switchyard monitoring equipment is being installed to identify the source of the intermittent loads
17 Steam Dryer Jim Meister Vice President Nuclear Services
18 Background Information D3 Timeline Began EPU operation on November 4, 2002 During the Fall 2002 refueling outage, perforated plates were added to reduce moisture carryover Added 1/2 cover plate Steam dryer externals were inspected December 2003 steam dryer modification Added 1 partial height front hood face plate Added 3 gussets on both 90° and 270° sides that extended onto the 1 plate Interior and exterior examinations performed in accordance with General Electric (GE) Service Information Letter (SIL) 644 Insert/cover plate size increased to 1/2 in Fall 2002 1 inserts added December 2003 3 gussets that extend onto the 1 plate were added in December 2003
19 Background Information D2 Timeline Began EPU operation on December 26, 2001 During the Fall 2001 refueling outage, perforated plates were added to reduce moisture carryover Fall 2003 refueling outage (post 690 days continuous run at EPU power)
Added 1/2 cover plate Added 1 partial height front hood face plate Added 3 gussets on both 90° and 270° sides that did not extend to the 1 plate Interior and exterior examinations performed in accordance with GE SIL 644 Perforated plates installed December 2001 Cover plate replaced; 1 insert and 3 gussets (not extended to 1 plate) installed on front hood
20 Background Information Preplanned Modifications Replace front vertical plate and corresponding section of horizontal plate with 1 plate on both outer hoods Retain section on each end for attachment weld Vertical welds are 1/2 groove welds Top and bottom horizontal welds are 1/2 fillet welds Install 1/2 thick gusset to ~6 from top of vertical plate Taper to 1 at tip Groove welded to vertical plate in shop Weld continues around tip Round extension piece to connect gusset to lower cover plate Increase the cover plate to ring weld (R2) to 1/2 Dryer stresses reduced by approximately 2.5 times through modifications 1 Plate Gusset 1 Plate
21 Background Information Dryer Analytical Methods
- Analytical teams consist of experts in numerous technical fields from a variety of sources
- Scale model test
- Load set for finite element analysis (FEA) of dryer modification
- Identify acoustic source(s) of dryer loads
- Acoustic circuit analysis
- Used scale model test to validate acoustic circuit analysis
- Demonstrated correlation of acoustic circuit analysis to scale model test
- QC2 pre-EPU and post-EPU load definition provided for FEA
- Developed scaling of loads between units
22 Background Information Dryer Analytical Methods (cont.)
- FEA (time history from QC2 plant data)
- Original design pre-EPU - high stress locations identified - below material endurance
- 2003 dryer modification - high stress locations identified - above material endurance
- Empirically established load limits
- Identified locations where cracking has occurred at D/QC units
- Evaluated stresses at these locations based on shell element model with computational fluid dynamics loadings
- Established design limit based on stresses at locations that have not experienced cracking
- Used methodology to evaluate modification design
23 Inspection Results 2004 D3 Inspection Scope/Results
- Best effort VT-1 and VT-3
- Approximately 1100 dryer inspection points
- Satisfies recommendations described in GE SIL 644, Revision 1
- Includes areas of dryer previously inspected on QC2
- 16 total indication notification reports (INRs) identified
- 12 were dispositioned as acceptable for additional service or stop-drilled
- No measurable crack growth from previously identified indications
- Two required minor repair
- Remaining two INRs were not previously seen in the industry
- Missing startup instrumentation pipe - subsequently retrieved from moisture separator
- Cover plate to dryer support ring crack
- No concerns for structural integrity or loose parts within reactor vessel
24 Inspection Results 2004 D2 Inspection Scope/Results Based on D3 inspection results, an inspection of D2 was performed Inspected areas where loose parts could be generated or the structural integrity of the dryer could be threatened
- All high stress welds identified in GE SIL 644, Revision 1
- Welds in areas modified in 2003
- Outer tie bars
- Startup instrumentation piping
- Interior inspections of the illustrated welds One INR identified - cover plate to dryer support ring crack 1 Insert Plate Horizontal Welds H2a, H2b, H2c Vertical Welds V5, V6 Gusset Tips G5, G6, G7 Ring to Cover Plate R2 Weld Cover Plate to Front Hood H3 Weld Vertical welds V7, V8 Outer Tie Bars
25 Quad Cities/Dresden Comparison Unit Inspection Comparison
- Severity and magnitude of cracking identified on the Dresden steam dryers has been considerably less significant than that previously identified at QC
- Number of observed indications has increased over time due to the expansion of areas examined on the steam dryers and improved inspection techniques
- Inspection results demonstrate the efficacy of previous repairs
- Previously identified indications have shown no measurable crack growth
- Inspections and analytical work continue to support our conclusion that Dresden is different from QC in both the loading and its effect on the dryers
26 Causal Factor Analysis D3 Flaw Evaluation
- Boat samples from four symmetrical locations were taken and analyzed
- Only one of four locations was cracked
- Other locations did not show any weld abnormalities
- Fillet welds were undersized at all four locations
- Cracking occurred in a location confirmed to be in a high stress region by the FEA
- Fatigue cracking occurred only at this one location and was not seen in the other boat samples
- Root cause and contributing factors
- Margin for load uncertainty was not adequate
- Fillet weld was undersized with a small lack of fusion defect
- Higher than nominal residual stresses (i.e., cold spring) reduced fatigue life
27 Causal Factor Analysis D2 Preliminary Flaw Evaluation
- One boat sample was taken and analyzed
- Crack propagated into cover plate base metal near gusset
- Crack in cover plate removed during boat sample removal
- Fillet weld leg lengths were adequate
- Fatigue cracking occurred at only this one location
- Preliminary root cause
- Lack of fusion defect at the crack initiation site
28 Causal Factor Analysis Corrective Actions
- Stress intensities have been decreased through the modifications
- Ongoing steps to further refine steam dryer loads
- Increased R2 weld size to 1/2
- Improved welding process
- Enhanced briefing with welders stressing importance of weld quality and size
- Revised weld sequence to minimize fit-up stresses
- Welder inspection of every pass
- Periodic supervisory checks of in-progress welds
- Verified correct weld profile through measurements
29 Ongoing Actions
- FEA of modified dryer using QC loads
- Performed scale model tests
- Validated acoustic circuit analysis
- Convergence is being reached between actual dryer loads and analytical predictions
- Further refinement will be obtained through main steam line (MSL) data Plot of Filter SMT data (Red) compared to CDI predicted data (Blue) -
First Benchmark Data Point M21 70 70.5 71 71.5 72 72.5 73 100 50 0
50 100 zfi zf1i ti
30 Ongoing Actions (cont.)
- Strain gauges provide a direct measurement
- Strain gauges provide high frequency content up to 200 Hz
- Measurements will eliminate nozzle phasing assumption for the acoustic circuit analysis
- Improved accuracy of load definition
- Data will be collected for both units during power ascension and steady state operation
- Acoustic circuit analysis will generate a load definition at EPU power
- Dynamic FEA will be completed for the as-modified dryers
- Will share analysis results with NRC in early 2005
- Ongoing steam dryer performance monitoring will continue
- Includes monitoring during power ascension
31 Ongoing Actions (cont.)
- In Spring 2005, the QC1 instrumented steam dryer and MSLs will provide the data correlation for acoustic circuit dryer load prediction
- Data collected on the MSL strain gauges will be provided to CDI to predict the dryer pressure measurements at instrumented locations
- In-plant validation for acoustic circuit analysis will be completed shortly after data collection on QC1
- Will meet with NRC in mid-2005 to present results of analytical work
- Will meet with NRC in Fall 2005 to share D2 inspection results, provide update on analytical results based on latest QC1 instrumented dryer and D2/D3 strain gauge data, and discuss impact on D3, including a potential mid-cycle inspection
32 Planned Interactions Vibration/Vulnerability Meeting Staff Mgmt Dresden Rotor and Steam Dryer Inspection Meeting QC Steam Dryer Loadings Meeting Staff Mgmt QC Steam Dryer Design Meeting Staff QC EPU Management Meeting Mgmt D/QC Instrumented Dryer Preliminary Results Meeting Staff Scale model test for source identification Dryer vane fabrication Fabricate QC1 dryer Install and test dryer instrumentation Ship dryer to station Install QC1 dryer QC1 startup testing Install QC2 dryer SEP O4 OCT 04 NOV 04 DEC 04 JAN 05 FEB 05 MAR 05 APR 05 MAY 05 JUN 05 JUL 05 AUG 05 Fall 05 Dryer analysis (final)
Staff Mgmt Mgmt QC1 Final Results Meeting Staff Mgmt Dresden Steam Dryer Meeting Install strain gauges on D2 and D3 Begin strain gauge data collection on D2 and D3 D2 refueling outage/dryer inspection Staff/Mgmt Dresden Results Meeting Staff
33 Bases for EPU Operation
- D2 and D3 steam dryer loading is less than either QC unit
- Inspections confirmed that the 2003 repairs to the outer hood maintained structural integrity
- Causes of cover plate cracks are understood and resolved
- Modification improves outer hood structural capacity
- 2004 outer hood modification will be validated using
- Acoustic circuit analysis time histories generated for QC2 at EPU conditions
- Scale model test time histories for QC1 model at EPU conditions
- Dresden in-plant data from strain gauges on MSLs
34 Conclusions Inspections and analytical work continue to support our conclusion that Dresden is different from QC in both the loading and its effect on the dryers Causes of the identified cracking are understood and corrective actions have been implemented or are in progress Structural improvements through the dryer modifications provide additional confidence Monitoring is in place that provides the capability to identify loss of structural integrity and to take appropriate actions Additional instrumentation on Dresden MSLs and QC1 replacement dryer will increase our understanding of loads and stresses We continue to enhance our analytical tools to refine our understanding of the loads
35 Closing Remarks Danny Bost Site Vice President Dresden Nuclear Power Station
36 Closing Remarks
- Generator rotors
- Exelon repaired the generator rotors with redesigned keyways and improved torsional capacity
- Exelon will monitor for torsional loading
- Steam dryers
- Steam dryer inspections were performed on both units and thorough repairs have been performed
- D2 and D3 have the new steam dryer outer hood modification which lowers weld stresses by a factor of 2.5
- As new insights are gained, appropriate actions will be taken as necessary
37 Closing Remarks (cont.)
- Both units can safely return to full power operations
- Monitoring is in place that provides the capability to identify loss of dryer structural integrity and to take appropriate actions
- Update April 2, 2004, letter detailing bases for EPU operation
- Propose ongoing meetings with NRC to discuss progress and results
- Fall 2005 meeting following the D2 outage will include discussion of impact on D3, including a potential mid-cycle inspection