ML080880076

Slides from 3/27/2008 Meeting

ML080880076
Person / Time
Site:	Oconee
Issue date:	03/27/2008
From:	Duke Energy Carolinas
To:	Office of Nuclear Reactor Regulation
References
Download: ML080880076 (40)

Text

x Duke WEnergy Duke Energy - Oconee Nuclear Station Steam Generator Tube Wear Update with NRC March 27, 2008 1

P Duke dkEnergy@

Topics of Discussion

" Introductions

" Problem Statement

" Summary of Inspection Results

" Condition Monitoring/Operational Assessment

" Failure Analysis Investigation

-Investigational Approach Most Probable Cause Contributing Factors Metallurgical Examination

" Tube Plugging Projections

" SG Repair Option

" Future Direction & Conclusions 2

-Duke Problem Statement PSnergy.

Unexpected tube wear was identified during the first in-service inspection of the replacement once-through steam generators (ROTSG) on all three Oconee units.

o The wear on all steam generators is non-uniformly distributed over large regions.

o The Oconee Unit 1 steam generators have the most wear.

Most of the periphery region of the steam generators is affected by the wear, with the highest incidence occurring in the mid to upper spans of the steam generator between support plates 9 through 12.

o All wear is occurring at the intersection between the tubes and support plates.

o Numerous additional wear indications were identified during the second in-service inspections on each unit.

Some of the existing wear indications grew larger during the second cycle of operation.

3

Duke Energy 4

P Duke OEnergy@

Inspection Results Tube Integrity Assessment 5

Summary of Inspection Results PDuke EEnergy Oconee Nuclear Station Steam Generator Tube Wear Summary Unit 1 Unit 1 Unit 2 Unit 2 Unit 3 Unit 3 EOC 23 EOC 22 EOC 22 EOC 21 EOC 23 EOC 22 1A 1B 1A 1B 2A 2B 2A 2B 3A 3B 3A 3B

1. of wear indications 7114 5187 2439 1769 2169 2493 6274 902 3554 1540 1952 820

1. of tubes with indications 4488 3718 1798 1450 1587 1724 495 698 2559 1181 1563 673

% tubes with indications 29%

24%

12%

9%

10%

11%

3%

4%

16%

8%

10%

4%

Average wear depth 9%

9%

10%

10%

8%

9%

8%

8%

9%

9%

7%

7%

Maximum wear depth 49%

41%

42%

42%

30%

42%

22%

32%

40%

39%

23%

26%

1. indications > 40% TW 10 5

3 2

0 1

0 0

2 0

0 0

1. indications > 30% < 40% TW 48 40 17 13 4

10 0

3 5

2 0

0

1. indications > 20% < 30% TW 215 182 71 64 27 90 3

8 64 43 2

1 EFPY per cycle 1.37 1.37 1.24 1.24 1.37 1.37 1.31 1.31 1.38 1.38 1.27 1.27 Average growth rate per EFPY 2%

3%

8%

8%

2%

2%

6%

6%

1%

2%

6%

6%

95/50 growth rate perEFPY 7%

8%

15%

15%

7%

8%

11%

11%

6%

7%

9%

10%

Maximum growth rate per EFPY 27%

29%

34%

34%

22%

20%

17%

24%

18%

20%

18%

20%

1. tubes plugged (wear) *=tubes pulled 20 19 30 18 0

1 2*

3 2

0 0

0 Plugging criterion 35%

35%

28%

28%

35%

35%

28%

28%

40%

40%

28%

28%

6

Oconee Unit 1 Wear Distributions PDuke kEnergy Support ON1A 6-1**

Total Oct. 2006

%tw<=5 3

1 3

5 6

22 127 300 246 25 130 39 44 7

1168'

-5<%tw<=10 7

13 14 4

12 3

31 257 766 938 90 556 338 282 19 4140 10<%tw<=15 5

2 2

2 13 159 354 277 177 112 93 3

1199 15<%tw<=20 2

32 119 85 60 46 34 3

381 20<%tw<=25 3

45 33 19 9

7 2

118 25<%tw<=30 2

24 25 5

56 30<%tw<=35 20 8

1 29 35<%tw<=40 9

9 40<%tw<=45 6

1 7

45<%tw<=50 3

3

%tw>50 0

Total 15 16 19 4

19 9

53 399 1262 1764 1564 947 544 461 34 7110 Support ONS1 B Total Oct. 20064

%tw<=5 5

1 3

163 194 214 372 122 5

79 94 11 1263 5<%tw<=1 0 7

16 16 5

10 6

130 147 491 1042 321 19 199 378 22 2809 10<%tw<=15 6

3 4

7 5

7 7

46 267 128 5

34 156 9

684 15<%tw<=20 2

2 1

7 94 62 1

5 53 4

231 20<%tw<=25 7

33 46 1

2 1_11 100 25<%tw<=30 1

26 24 2

1 53 30<%tw<=35 1

12 11 1 1 25 35<%tw<=40 6

8 14 40<%tw<=45 1

1 45<%tw<=50 0

%tw>50 0

Total 16 26 22 12 15 9

300 348 766 1853 722 31 319 694 47 5180 7

Duke ieEnergy@

Oconee Unit 1 Wear TSP Scatter Plots ONS1-A Fall 2006 %tw TSP Distributions Irto~Saturation oint Jz*r-*

50 -

45 4 0 - - -

t ----- -

35----

at 0 25---------------

20 1 0 -

Uto

-S 0

0 1

2 3

4 5

6 7

8 9

10 11 12 13 14 15 TSP ONS1-B Fall 2006 %tw TSP Distributions Saturation Point Lower Shroud Super Upper Shroud 55 50----------------------

40 152 0....

15 o

'_ I I"*:-*"..

U".........

10 10 0 -.

0 1

2 3

4 5

6 7

8 9

10 1 1 12 13 14 15 TSP 8

Oconee SG 1A Wear Radial Distribution P

Duke OEnergy Note: Wear 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0

Indications >10% TW Only ONS1-A Fall 2006 TSP (All)

• 11-15%

• 16-20%

o 21-25%

• 26-30%

o 31-35%

• 36-40%

o 41-45%

• >45%

X2 Y2 0 Y1 Xl Orientation 0

0 0

0 0

0 0

00 L0 o0 c

0aaa aa aa a

2 c\\

N 0N 0N N-N 9

Oconee SG 1 B Wear Radial Distribution MkDuke EEnergy Note: Wear Indications >10% TW Only 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0

ONS1-B Fall 2006 TSP (All)

• 11-15%

• 16-20%

o 21-25%

• 26-30%

o31-35%

. 36-40%

o 41-45%

0 >45%

X2 Y2 Y1 xl Orientation o

o 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

0 0

N M'

-It M

U CD N

)

0 CJC)~

U CO M-W C

0

-ý N~ m~ '-I in wD w

C CN C N N C

N C

N N N 10

B&Duke Condition Monitoring Results Iftnergy El Condition Monitoring Limit = 73% TW (NDE maximum depth)

- controlling structural integrity loading condition is 3AP = 4050 psi

- assumes tapered wear profile based on measurements with array probe

- axial load under accident conditions not an issue based on maximum width wear scars (= TSP land width of 0.2 inches)

- accident leakage not an issue at much lower MSLB pressure differential (leakage and burst of volumetric degradation are essentially coincident)

[]

Maximum Measured Wear Depth is Less Than CM Limit

- Oconee Unit 1 -49% TW

- Oconee Unit 2-42% TW

- Oconee Unit 3-40% TW

'3 Condition Monitoring Demonstrated at 95% Probability with 50% Confidence 11

Operational Assessment Results P

Duke oEnergy.

o]

End of Cycle Allowable Structural Limit = 78% TW

- EOC allowable limit based on 0.95 probability of meeting a minimum burst pressure 3AP value of 4050 psi at 50% confidence

[]

Considers material property, burst equation, and NDE uncertainties o]

Assumes maximum growth rate from previous cycle

[]

Calculated NDE repair limits for all cycles are shown below:

Cycle Oconee 1 Cycle 23 Oconee 1 Cycle 24 Oconee 2 Cycle 22 Oconee 2 Cycle 23 Oconee 3 Cycle 23 Oconee 3 Cycle 24 Maximum Growth Rate Calculated NDE Repair Limit 34% per EFPY 28%

29% per EFPY 35%

24% per EFPY 35%*

22% per EFPY 35%

20% per EFPY 36%*

20% per EFPY 40%

• 28% TW used as plugging limit for consistency with Unit 1 Cycle 23 12

Cycle-to-Cycle Wear Indication Comparison M Duke rEnergy Unit 1 Wear Growth in Relation to Original Depth for SG A 40r

--I Average Growth

-+

Individual Growth I

L) cc C)

CD, C3 35'-

÷

-I-30 F 25 20 15 10 5

0

-5 K

+

+

+

+t

+

+

+t

+

++

+

+++

--I

+-

+

÷

÷

+

-I-

+

+

4 I-

+

+

+

+++ +

+

+-t-

÷

+

+1

++/-

+++

• jA+*4-4.+ 4

+4 i*l++_-

+-+

+

+

_:+

++-+

++

+

+

+

-10 0 I

i 5

10 15 20 Fret Depth in Cylce 1 (%TW) 25 30 13

Duke Summary of Inspection Findings ODuknergy.

n Wear is predominantly occurring in periphery region of steam generator

(-88% of indications are beyond outermost tie rod)

[]

Wear is predominantly occurring in the steam space (-90% of indications are at the 9 th tube support plate or higher) oTubes with multiple indications range from 10% to 30% per steam generator, with the highest incidence in the worst steam generator (1A) and the lowest incidence in the best steam generator (3B) o]

Wear rates are declining (in terms of depth)

[]

EOC wear depths do not correlate well with BOC depths

[]

While the number of wear indications is high, the wear depth is not inconsistent with structural wear experienced in other operating steam generators

[]

Use of the maximum growth rate to determine the plugging limit is very conservative 14

P Duke U Energy Failure Analysis Investigation 15

p Duke Failure Analysis Investigation kSnergy@

Investigative Approach

[]

Failure analysis investigation performed by Babcock & Wilcox Canada with review and support from Duke Energy personnel.

o]

Investigative approach combined analytical methods, laboratory testing, and field data collection to either support or rule out potential causes and contributors o]

Industry experts used to perform some analyses and testing, and to peer review results o]

Joint meetings were periodically held between Babcock & Wilcox Canada, Duke Energy, and the consultants/industry experts to review results, substantiate or rule out causes and contributors, and determine next steps o]

Most probable cause and key contributing factors determined using the above approach based on the weight of the evidence 16

PkDuke Failure Analysis Investigation Avenues of Investigation

1. Development of comprehensive 3-dimensional CFD model for the ROTSG

2. Detailed review of plant operating conditions (flow, pressure, water level, heat transfer)

3. Side-by-side comparison of ROTSG design features with original OTSG

4. Analytical modeling of support plate deflection

5. Analytical testing to determine material wear coefficients

6. In-situ measurement of tube tension in Oconee Unit 2 steam generators

7. In-situ measurement of tube tension and tube alignment in Midland steam generators (original OTSG)

8. Visual inspection of Oconee SG secondary side to verify design tolerances and assess steam quality/flow conditions based on deposit profiles

9. Metallurgical examination of two tubes pulled from Oconee Unit 2 steam generators

10. Field monitoring on all three Oconee units for structural vibration, acoustics, and pressure pulsations

11. Mockup testing to measure axial flow instability using various tube support geometries

12. Mockup testing to measure tube vibration under simulated plant conditions 17

Oconee Tube Wear Probable Cause Dukery Most Probable Cause Statement Oconee ROTSG tube wear is due to the precise alignment of the tube supports that allows small excitation forces (axial flow hydraulic forces) to cause tube vibration, since the tubes are minimally damped and consequently very responsive over their lengths. This 'perfect' alignment reduces support effectiveness and allows the tubes to vibrate and impact supports with sufficient energy to cause wear since there is little lateral interference with the supports and little damping.

Supporting Evidence 1.

Mockup testing with prototypical design produced significant tube vibration under simulated flow conditions that could be mitigated upon inducing a lateral load

2. Metallurgical. examination of pulled tubes showed that the wear resulted from low impact, high cycle vibration

3. Measurements of tube to support plate alignment (lack thereof) on an original OTSG (Midland).

18

B&Duke Contributing Factors PSnergy@

Contributing Factor #1 - Annular Flow Instability

• The hour-glassed tube support plate geometry decreases the stability of the tubes and increases tube vibration when subjected to axial flow hydraulic forces

" Although the energy associated with annular flow is believed to be small, the minimal lateral support of the tubes due to alignment allows the vibratory response Supporting Evidence

1. Mockup testing of various support plate geometries (Note: The hour-glass geometry produced a significantly larger response than the support plate holes in the original OTSG under the same axial flow) 19

Duke illustration of Hour-Glass Geometry PDEnergy.

ROTSG Tube Support Plate Hour-Glassed Hole Geometry

)

20

pDuke Contributing Factors kSnergy Contributing Factor #2 - Dishing of Support Plates The position of the support plates is fixed by tie rods in the center of the bundle and by blocks attached to the shroud on the periphery The support plates become slightly bowed at temperature due to differences in thermal expansion between the 410 SS tied rods and carbon steel shroud oThis reduces contact area by creating an angle of impact, thereby increasing initial through-wall wear rates Supporting Evidence

1. Analytical modeling of support plate deflection

2.

Eddy current profiling and metallurgical examination showing a tapered wear geometry 21

Illustration of Support Plate Dishing P

Duke PEnergy Note: Displacements Exaggerated I

Upper Shroud Tie-Rods TSP-10

- Blocks Lower Shroud 22

Contributing Factors I Duke Contributing Factor #3 - Tube Tension

" Tubes in the periphery region of the ROTSG are under tension at hot conditions

" Tension increases the straightness of the tubes, making the precisely aligned support plates even less effective in suppressing vibration due to annular flow or other low energy sources Supporting Evidence

1. Observation of radial location of wear indications

2. Tube axial stress calculations by an independent consultant

3.

In-situ tube tension measurements on an Oconee Unit 2 ROTSG 23

F-Duke Other Factors Energy Possible Contributing Factors oAcoustic Energy - Low level acoustic energy (either standing acoustic waves or acoustic resonance) might be sufficient to cause excitation of the tubes with minimal lateral support. Different acoustic levels were observed on each unit by field instrumentation. However, calculated loads on the tubes were determined not to be sufficient to cause wear.

[]

Wear Couple - Material combination between the Alloy 690 tubes and 410 SS support plates. Considerable variation exits in the wear couple data from various sources. However, there is no conclusive evidence this couple is any worse than the Alloy 600/carbon steel wear couple on the original OTSG's.

Factors Ruled Out o]

Cross-Flow Induced Vibration - Classic FIV due to high cross-flow velocities, including effects of bleed steam flow. Eliminated based on validation of CFD model, and linear and non-linear FIV analysis.

24

PDuke Other Factors (continued)

CSnergy@

Factors Ruled Out o Fluid Elastic Instability - Occurs when flow across a tube bundle reaches a critical velocity and tube amplitudes increase. Eliminated based on analysis, wear location, and behavior.

o Structural Vibration - Steam generator structural vibration due to RCS flow perturbations, excitation by RCP's, excitation by steam nozzle flow restrictor, or ineffective lateral restraint. Eliminated based on field accelerometer measurements.

o Feedwater Induced Pressure Perturbations - Includes excitation of tubes or shroud by feedwater flow entering through FW spray nozzles.

Eliminated based on analysis. Also, no conclusive evidence of significant pressure perturbations from field instrumentation.

o Plant Operational Excursions - Eliminated based on review of plant operating conditions. Operating conditions on all three Oconee units are within expected design range.

o3 Fabrication Non-Conformances - Eliminated based on review of manufacturing records. No significant non-conformances were identified.

25

pDuke Metallurgical Exam Results

/CSnergy.

Key Findings

[]

The wear scars are confined to TSP broached land contact areas o]

The wear scars have a tapered geometry o] The amount of wear varied at different support plate elevations o]

Evidence of wear at support plate intersections along the length of the tube indicates that tube motion is occurring in both the steam and liquid phases o]

There was an absence of sub-surface cold work on wear scar surfaces o]

There was no evidence of smearing or galling on wear scar surfaces o]

There was an absence of grain boundary distortion, indicative of low impact metal to metal contact Conclusion The observed wear phenomenon is a result of low load, high cycle impact of the tubes with the support plates which produces wear scars on the tubes.

26

Wear Scar on Oconee 2 Pulled Tube I

Duke EEnergy 53-114, Pce 32,TSP 14, 27

Wear Scar on Oconee 2 Pulled Tube P Duke OEnergy 1 in Top 53-114, Pce 28,TSP 12, 120 Deg.

28

Duke

~Energy@

Tube Plugging Projections SG Repair Option 29

Duke Tube Wear Growth Model ISnergyo Overview of Model

• ] Tube wear growth model developed by Babcock & Wilcox Canada based on concept of constant volumetric metal loss

• 'Wear data from first two cycles of operation was used to develop the model Wear data follows a log normal distribution Results of the growth model are compared to observed growth Future wear distributions can be projected from the model Future tube plugging can be predicted from the model Duke is working through independent means to validate the model assumptions and approach 30

Duke Tube Wear Growth Model Pok~nergyo Key Assumptions The volumetric growth distribution does not change from cycle to cycle The volumetric growth behavior is random A fret angle of 0.6 degrees is used to covert wear depth to volume Each tube is represented by the largest wear scar on it A log-normal distribution is used to fit the observed wear depth distribution of Cycles 1 & 2. It is also used for the volumetric growth model.

31

Log Normal Wear Depth Distribution

f. Duke L Energy Log-Normal Fit of Wear Depth Data for Oconee Unit 1 Cycle 2 LL I-I E

z 1000 900 800 700 600 500 400 300 200 100 0.2 0.25 0.3 Fraction of Tube Wall Thickness 0.5 32

Wear Volumetric Growth Distribution P Duke WEnergy Wear Volumetric Growth Distribution within Susceptible Region for Oconee Unit 1 Cycles 1 & 2, and Simulated Cycle 2 0.9

/

0.8 0.7 5 0.6 C_ 0.5 E 0.4 01 0.3 0.4 0.5 0.6 0.ý Fraction of Volume Loss of 40%tw Fret 33

Future Wear Distribution P Duke EEnergy Wear Volumetric Distribution within Susceptible Region for Oconee Unit 1 Cycle 20 I

m E

0.3 0.4 0.5 0.6 0.7 Fraction of Volume Loss of 40%tw Fret 34

Tube Plugging Projections M Duke VEnergy Plugging Prediction for ONS 1, 2 & 3 C,)

.0 I-.

"0 01 0

01) 01 a-0 2

4 6

8 10 Cycle 12 14 16 18 20 35

PDuke Steam Generator Repair Snergy@

Conceptual Repair Option SBabcock & Wilcox Canada has developed a conceptual repair to mitigate the tube wear oRepair involves shifting support plates to pin the tubes using an externally mounted spring which forces support plate movement during operation Repair requires physically removing a portion of the anti-rotation blocks to allow the support plate to move oRepair requires drilling small holes in steam generator at each repair elevation to insert spring-loaded rod o]

Duke is working with Babcock & Wilcox Canada to validate the feasibility of the repair and finalize the conceptual design 36

Duke WEnergy Future Direction and Conclusions 37

f&

pDuke Upcoming Activities PSnergy@

Unit 1 Inspection Critical data to be obtained from Unit 1 EOC24 inspection in April, 2008 (third ROTG inspection) data to validate plugging projection model data to analyze feasibility of 24 month fuel cycle data to determine if less conservative plugging limits are appropriate (i.e., plug at 40% technical specification limit)

Technical Issues Previous fatigue analysis to be re-visited based on current understanding of problem and data from mockup testing Repair Option Duke technical personnel plan to make recommendation to Management by August 31 on whether to pursue repair implementation or continue with status quo (i.e., inspect and plug) validation of plugging projection model and feasibility of making repair will be key factors in making decision how to proceed going forward 38

7Duke Longer Term Issues COEnergy Tube Plugging Limits o]

Current tube repair limits are based on very conservative operational assessment approach using maximum wear rate.

o This approach is justified until wear behavior is well understood. However, it results in premature tube plugging (small number of tubes now, but much greater in the future).

o If justified based on wear behavior, we plan to use the 95 th percentile wear rate to determine repair limits - will essentially mean plugging at 40% TW oLonger term, we may pursue an alternate repair criterion for Oconee tube wear (45% or 50%) based on wear behavior and available margin Inspection Frequency o

Oconee steam generator inspections are being performed every 18 months (100% scope) o Oconee may transition to 24 month fuel cycle if justified based on wear behavior o]

Ultimately, we would like to skip inspections if justified based on wear behavior 39

F-1Duke Conclusions CEnergy.

ol The Oconee ROTSG's have unexpected tube wear occurring predominantly in the upper peripheral area within the super-heated steam region ol The tube wear is occurring due to minimal lateral support of the tubes that allows low energy forces to excite the tubes as a result of alignment and minimal damping o

Contributing factors are related to certain design differences made to the ROTSG along with the precise installation of the tubes during manufacturing o]

Although the amount of wear is unusual, the depth of wear being experienced is not inconsistent with structural wear in other steam generators o

Current tube plugging projections indicate the Oconee steam generators can meet their design life, albeit with significant tube plugging

[]

The structural integrity of the tubing is not being challenged by the rate of through-wall penetration being experienced when inspecting on an 18-month frequency

[]

Future inspection frequency (24 or 36-month) depends on future wear behavior 40