ML052550169

Presentation Material from August 2005 Quad Cities Steam Dryer Meeting

ML052550169
Person / Time
Site:	Quad Cities
Issue date:	08/29/2005
From:	Roddey T Exelon Nuclear
To:	Office of Nuclear Reactor Regulation
References
Download: ML052550169 (132)

Text

Exe kn.M Nuclear Quad Cities Replacement Steam Dryer Meeting August 29-September 1, 2005 1

Exe inSM Nuclear Introduction Thomas Roddey Licensing Engineer 2

Agenda Exe kn.M Nuclear

• Introduction

• Opening Remarks

• Quad Cities Unit 2 (QC2) Startup Test Report

• QC1 Startup Test Report

• Engineering Evaluation Summary

• Scale Model Test (SMT)

• Acoustic Circuit Model (ACM) and Refinements

• Main Steam Line (MSL) Strain Gauge Evaluation

• Uncertainty Evaluation

• QC2 and QC1 Load Definition

• Structural Analysis Summary

• QC2 and QC1 Dryer Structural Analysis

• Dresden Steam Dryer Replacement Overview

• Summary and Conclusions

• Open Items Closing Remarks 3

Exe On.M Nuclear Opening Remarks Roman Gesior Director - Asset Management 4

Opening Remarks Exek6 nSM Nuclear

• Exelon previously presented the basis for two design loads, startup test plans, startup test results, and load methodology validation for the QC replacement steam dryers

• Results for the steam path data collection during startup testing for both QC units demonstrated a need to perform a detailed load definition and finite element analysis (FEA) to validate structural adequacy at full extended power uprate (EPU) power levels 5

Opening Remarks (cont.)

Exekdn.

Nuclear

• QC2 Refinements to ACM and SMT methodologies are completed or in-progress For QC2 load definition, the modified 930 megawatts-electric (MWe)

ACM was used and provided conservative loads Analysis of the overall dryer and skirt demonstrate that the QC2 dryer is structurally adequate for the full range of operation, including EPU power levels Uncertainty analysis for data collection, load definition, and finite element analysis was completed

• Overall methodology uncertainty was determined to be <5% in the non-conservative direction, with the majority of the uncertainty being in the conservative direction 6

Opening Remarks (cont.)

Exek6 n Nuclear

• QC1 During data collection on QC1, strain gauge failures required data adjustments to produce load definition inputs; these adjustments were evaluated to be conservative and appropriate For the QC1 dryer load definition, the modified 930 MWe ACM was used and provided conservative loads; skirt load definition used the minimum error ACM, shown to be the best methodology for load definitions to date Analysis of the overall dryer and skirt demonstrate that the QC1 dryer is structurally adequate for the full range of operation, including EPU power levels 7

Opening Remarks Summary xe d5nsm Nuclear

• The engineering design and evaluation approach has been implemented as previously communicated to the NRC

• Results of engineering evaluations demonstrate adequate structural margin for both QC units

• The methodology and approach can also be implemented for the Dresden units 8

Exe knSM Nuclear QC2 Startup Test Report Brian Strub Design Engineer Quad Cities Nuclear Power Station 9

Instrumented Steam Path Exe onSM Nuclear Most heavily instrumented dryer in the industry 42 dryer sensors 56 MSL strain gauges 33 MSL accelerometers Uses of data Go/no-go criteria Develop loads Validate ACM actual dryer SMT and Validate design load Instrumented New Dryer Face 10

I QC2 Startup Test Strategy Exe k n M Nuclear 1

j 11

QC2 Data Collection Results ExeOns.

Nuclear TC41 2885 MWt 1000 900 800 700 LI 0

CD LU r.M R.,

M CD N

600 500 400 300 -

200 -

1 00 0

05116A05 05/17/05 05/1805 0511 9105 05/20105 05t21105 05/22fO5 05123/05 05/24f05 05125105 Date and Time 12

I QC2 Data Collection Results (cont.) Exe In.M Nuclear Test C ondition and Thermal Power

(% of CLTP) 90 30

!20 N

MP O

I 0 I'-

a V9 v

O C

E D

C" ti Ci l

l4 Is C o n t i. o el T

l Test C~ondition tITC 13

QC2 Data Collection Results (cont.) ExeI6nm Nuclear 2.000Q MX 0 5000 F -

SD P21 AutoPw fer 2B851rT093OJE TC4rl=

1--

0.3000 0.2000 0.1000 0.0_00 0.0800 CL 0.0300-0.0200 0.00 0.0070 0.0030 0.0)

-SO_

0.0020 0.0010 0.0D0 0

10 20 30 40 so 60 7080 901l0 110 120 130 14150 160 170 180 200 Ht 14 C-63

QC2 Startup Exek(n.m Nuclear Results in startup test report Strain gauge #7 exceeded Level 2 Criteria

• Full load definition and FEA completed for structural evaluation Pressure data trended Accelerometer data trended Moisture carryover - initially exceeded 0.09% then stabilized below 0.08%

Reactor water level - no anomalies noted MSL flow rates - no anomalies noted Feedwater flow - no anomalies noted Steam flow/feedwater flow/reactor power comparison - no anomalies noted 15

QC2 Data Collection Results (cont.) Exe On.

Nuclear Quad Cities Unit 2 Moisture Carryover C:

0.

it 0.1 0.09 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0

-Moist ure Carryo'~er

-- Thermal Power 3000 2900 2800 2700 2600 2500 2400 2300 2200 2100 2000 C

3:

X-E I-n U)

On U) kn0~

CD 01-c()

i04 LO kn CD (D

io koU) r'4

(!

U) 0D CD QD0B U) 00 (4

U) 0 0C0i I.-01 (0

U) 00N1 01 1'-

I.-

16 C -0c

QC2 Vibration Assessment Exe kbn.M Nuclear

• Acceptability of ERV components at EPU Four new QC2 ERVs have identical assemblies to QC1 Valve assembly hardened with X750 QC2 vibration levels are bounded - Wiley Labs test qualification data

• High Pressure Coolant Injection (HPCI) system -4 valve operator Vibration data collected shows unit 2 to be bounded based on original evaluation and laboratory qualification testing

• MSIVs Vibration data collected shows QC2 to be bounded, based on original evaluation and laboratory qualification testing

• Small bore piping/feedwater sample probes Meet Exelon Engineering Standard NES-MS-03-04 Feedwater probe inspection showed probe was intact New probe: 2" vs. 13" 17

Exe Iin.M Nuclear QCI Startup Test Report Brian Strub Design Engineer Quad Cities Nuclear Power Station 18

QCI Data Collection Exe inSM Nuclear

• MSL strain gauges were re-configured to more closely match the QC2 configuration

• For the first data collection during the QC1 startup following dryer replacement, a total of five strain gauges failed - S3, S6, S1 1A, S31, and S36 19

QC1 Data Collection Strain Gauge Configuration - Startup Exe lIn.

Nuclear S2 I sg S:34 57

_A S7 l

>LI LZ MSL/

MSLE 20 cycx0,,

I QC1 Data Collection Results Exe InS Nuclear U1 Start Up after Q1 M18 TC15a 2887 MWt TC115b 2900 MWt 1.000

0-U1 MW planned:.--*.

i.

t

• 0 Su ~U 1 M dI:t,W,-,.,..- ;

W E...........a,,;t;ua 0 V

<N

-i l

uEM gU-:

-00X S f 0 ;::--

a 700 ii7

/ *1'

'.'1:

• *-l, 600 I1.

i I

i C-1 ij

-,,j I:A c

I--

vc2 U

,,wI c1fdljI

-11yY UII IVIOL U LIIIIW, LUily I enni Olperation to 2410 MWI F-5 I

L EC 355874 Justifies

.2511 MWtI on U-1

-h based on fatigue v analysis. Power increase at 07:00 Ihrs on 061'08.'05 I

I-u ! I 4E~-4~

-~!~ ~

,~A~

~~~~

ir IS, M,

,~ '

f 400

7 I-I 4

.4-4IleI.

1.

300 200 100 Valve 1-3508B Won't Close to Less than 30i 0 -

..Leak

T Fails on X-2 (Personnel R At Tc est Hatch )

T e s t Faln-O 4 06102 0'0:00 06C'03 00:00 06104 00:00 0605 00:00 Date and Time 06Nt06 00:00 06.107 00:00 06/08 00:00 21 O-VG

QC1 Data Collection Results (cont.) ExeI OnSM Nuclear Quad Cities Unit 1 TC 15 - 0610512005 - MSL C 651 PSD 1.OOE+00 1.00E-01 1.00E-02 1.OOE-03 N

'19 I-1.OO-04 20 40 60 80 100 120 140 il60 180 80 Hz peak A

due to missing A

strain auges U2 100% MSL B S&s10

-1U2 100% MSL C S32iS34

-U 1 Test MSL C 651 S32iS34 Frequency [Hz]

1.OOE-05 1.002P-06 1.OOE-07 1.00 F-08 22 C -07

QCI Startup Exek(nnm Nuclear Results in startup test report MSL C strain gauge at 651 elevation exceeded Level 1 Criteria

• Fatigue evaluation completed to allow data collection at Full EPU Moisture carryover - no anomalies noted Reactor water level - no anomalies noted MSL flow rates - no anomalies noted Feedwater flow - no anomalies noted Steam flow/feedwater flow/reactor power comparison - no anomalies noted 23

QC1 Vibration Assessment Exe I e n SM Nuclear

• Acceptability of ERV Components at EPU 4 QC1 ERVs have identical assemblies to Unit 2 Valve assembly hardened with X750 Unit 1 vibration levels are bounded by testing - Wiley Labs test qualification data

• HPCI 4 Valve Operator Vibration data collected shows QC1 to be bounded based on original evaluation and laboratory qualification testing MSIVs Vibration data collected shows QC1 to be bounded based on original evaluation and laboratory qualification testing

• Small bore piping/feedwater sample probes Meet Exelon Engineering Standard NES-MS-03-04 Feedwater probe inspection showed probe was intact New probe: 2" vs. 13" 24

Exe InSM Nuclear Engineering Evaluation Summary Keith Moser Asset Management Engineer 25

New Dryer Design Strategy Exe kOnn Nuclear GE SIVT\\

Load Development 12 Resolve Structural Discrepanci

Mcdify Design Validate Loads and Design with I n-vessel QC2 Instrumented Dryer/Steam Path Independent Review QC2 Plant Data 26

Engineering Evaluation Summary (cont.)

Exekrn.m Nuclear

• New dryer design loads:

QC1 SMT QC2 In-plant loads - ACM Actual QC2 steam dryer pressure measurements had enough difference from design loads (see figure on next page) to require a new load definition based on QC2 startup data and detailed FEA SMT Comparisons show conservative correlation up to 135 Hz GE/Exelon pursuing QC2 model refinement

• ACM Model refinements completed based on QC2 instrumented steam path

• Detailed load cases have been created for QC2 and QC1 based on ACM

• FEA Modeling improvements made to better match as-built conditions - QC1 Results show both QC2 and QC1 are structurally acceptable for EPU operation 27

m QC2 Data Collection Results (cont.) Exel6n.m Nuclear I") I Spectral (verlay with SNIT and VIt 1. at T' 41 28 C,-cl

Exek nSM Nuclear SMT Daniel Sommerville General Electric 29

SMT Exe 1 n.M Nuclear This presentation contains information that is proprietary to General Electric 30

Exekd)n.s Nuclear ACM and Refinements Kevin Ramsden Senior Staff Engineer 31

ACM Exekdn.m Nuclear QC2 instrumented steam path provided a validation of load definition models As planned, Exelon provided the MSL strain gauge data to Continuum Dynamics, Inc. (CDI) at 790 MWe and 930 MWe without providing the 6 key pressure transducer data on the dryer until the first ACM predictions were made.

• At both power levels, a refinement to the model was made to better match the data and Exelon's acceptance criteria

• The end result was the Modified 930 MWe ACM that was used for QC2 dryer/skirt and QC1 dryer load definition

- After the initial ACM refinements were completed, the data from all 26 dryer pressure transducers were provided to CDI to develop the best ACM

• The first AC model was a least squares approach with the 26 pressure transducers

• The next refinement involved including the MSL strain gauge data, which resulted in the minimum error ACM

- This was used for the QC1 FEA of the dryer skirt 32

ACM - Refinements Exek nsm Nuclear

• First ACM - QC2 790 MWe evaluation Model parameters set to values used for QC2 new dryer design load

- Only in-plane MSL strain gauges were used for inputs

• Resulted in additional frequency content when compared to actual pressure transducer results on the dryer

• Second ACM - QC2 790 MWe evaluation Most significant change was to use both sets of MSL strain gauge pairs as input

- Additional changes to the model included the following:

• 60 Hz spike was filtered from the MSL strain gauge data

• Damping was adjusted at the steam-water interface, steam dome, and MSLs

• Acoustic speed of sound was adjust to 1484.33 feet per second (ftls) in accordance with ASME steam tables at 1000 pounds force per square inch absolute (psia) 33

ACM - Refinements (cont.)

Exen I K Iuclear

• Third ACM - QC2 930 MWe evaluation IN Same model as the 2nd QC2 790 MWe evaluation

• Under-predicted P-21

• Fourth ACM - QC2 930 MWe evaluation Damping in the MSLs was increased by 20%

• Resulted in the prediction for P-21 to be within the pre-defined acceptance criteria

• Skirt loads were over-predicted at the skirts natural frequency 930 MWe: P21 30 MWe: P24

.I I)

W N

'AI 4-,

0. I

().)

0.00)

(J.OO() I I oIY<

N

.4 C-A

(!)()

j.(!() I 0.0001 I0o'

()

50 I0()

150 Frequency (Hz) 200) 106 L 50 10 15c Frequency (Hz) 34 200 GY-N

ACM - Refinements (cont.)

xe xebnM Nuclear

• Fifth ACM - QC2 930 MWe evaluation

- 26 steam dryer pressure sensors used as input Previous model parameter for Helmholtz solution from fourth ACM evaluation remained the same

- The approach was to minimize the error by using a least squares methodology as outlined below 26 Error(co) = I,[Ppred (0O) - Pdata ((o) ]2 Where: X = Frequency

• This methodology resulted in over-predictions at frequencies which create dynamic structural responses on dryer components 35

ACM - Refinements (cont.)

Exekrbn.m Nuclear Minimum error - QC1 skirt evaluation

- Sixth ACM - QC2 930 MWe evaluation

• To improve the QC2 least squares approach using 26 steam dryer pressure transducers, the sixth model used eight MSL strain gauge locations and 26 steam dryer pressure transducers as drivers for the model

• The steam dome Helmholtz solution was refined to more accurately compute the skirt loads fl 27 Error.= 1 W [I pred Pd

)

0-0 \\-I 36

ACM - Refinements (cont.)

Exe O n SM Nuclear

• Minimum error ACM comparison to modified 930 MWe ACM and dryer data for the skirt P24 sensor 930 MWe: P24 0.1 N

• -1

=-4 0.01 0.() I 0.0001 I 0-4 10 o 0 5o0 100 150 Frequency (Hz) 200 37 0-f 2

I ACM Minimum Error (cont.)

Exe InSM Nuclear P20 Pressure distribution Pressure distribution 700 Performed histograms of 600 pressures, irrespective of 500 time 5

400 __Gaussian distributions HI 300 I

• See Figure for P20 200 _-

Determined that pressure distributions can be 100

\\!z represented by a normal distribution

,-3

-2

-I 0

1 2

3 prsessurdeHuH2r Used statistical evaluation measured predicted

• Data population capture

• Confidence levels 38 C-(f

ACM - Minimum Error (cont.)

Exek(bnSM Nuclear Pressure distribution results

- Two-sided 95-95 tolerance limit equals 1.96 x standard deviation

• 95% of the total population is captured with a 95% confidence level 18 of the 26 pressure locations meet 95-95 tolerance limit

• P7 isslightly below the 95-95 tolerance limit

• P14/P26 are located inside the dryer and under-predictions produce conservative results as this yields a higher pressure differential on the dryer face

• P16/P26 and P27 are in low pressure areas with very low pressure oscillations

- not significant to the dryer structural response

• P2/P11 meet 95-90 tolerance limit 39

ACM - Minimum Error (cont.)

Exek nSM Nuclear Conclusions for the minimum error ACM The pressure distribution can be represented by Gaussian distribution functions With two exceptions, 95-95 confidence limits were met at significant locations

- The two exceptions meet a 95-90 confidence limit

- A more representative skirt loading has been achieved with this ACM

- This load methodology is acceptable for the entire dryer 0 Used to confirm stresses on QC1 skirt are acceptable

- The load methodology with resulting damping values for MSLs, steam dome, and steam-water interface developed on the instrumented steam path for QC2 can be used for QC1, D2, and D3 40

ACM Methodology Typical Uncertainties Exe 6nSM Nuclear Main steam line Ref leg AP2 Uncertainty Source Typical Uncertainty Typical Sensitivity to Steam Dryer Load Strain gauge measurement 5-10%

0.6 Reference leg/venturi DP 5%

measurement Instrumental line transfer 177% or less function Pressure uncertainty due to 25% or less compliance Instrument location Up to 50%

41

ACM Validation Exe In.M Parameter Changes: QC2 Data Nuclear Data Set for Evaluation Steam/Water Interface Ratio Steam Dome Damping Ratio Steam Line Damping Ratio Acoustic Speed Ratio Input Data Average Error on RSD of RMS Comments 790 MWe 0.93 20.0 0.0 1.08 8 strain gages:

4 pressure After receipt of Blind in-plane pairs sensors: 0.136 dryer data: P3, only P12, P20, P21 790 MWe 1.0 4.0 1.5 1.0 8 strain gages:

4 pressure Trial and error Modified average all sensors: 0.359 to meet pairs, filter 60 acceptance Hz criteria on P3, P12, P20, P21 930 MWe 1.0 4.0 1.5 1.0 8 strain gages:

4 pressure After receipt of Blind average all sensors: 0.113 dryer data: P3, pairs, filter 60 P12, P20, P21 Hz 930 MWe 1.0 4.0 1.8 1.0 8 strain gages:

4 pressure Trial and error Modified average all sensors: 0.130 to meet pairs, filter 60 acceptance Hz criteria on P3, P12,P20,P21 Least 1.0 4.0 Not 1.0 27 pressure 4 pressure Singular Value Squares Applicable sensors sensors: 0.095 Decomposition Dryer Data (SVD) to minimize error*

Least 1.0 1.0 1.0 1.0 8 strain gages:

4 pressure SVD to Squares All average all sensors: 0.152 minimize error*

Data pairs, filter 60 Hz +27 pressure sensors

• W. H. Press, S. A. Teukolsky, W. T. Vetterling and B. P. Flannery. 1992. Numerical Recipes in FORTRAN: The Art of Scientific Computing (Second Edition). Cambridge University Press.

42

Exek1nSm Nuclear MSL Strain Gauge Evaluation Karen Fujikawa Structural Integrity Associates 43

Overview Exe ki n.M Nuclear

• Data collection for the QC1 startup following the steam dryer replacement had five strain gauges fail

- S3, S6, S11A, S31, and S36

• Each strain gauge location consisted of four strain gauges installed 90° apart

• Each pair of strain gauges located 1800 apart were installed in opposite arms of a Wheatstone bridge (1/2 bridge configuration)

• The Wheatstone bridge was reconfigured into a 1/4 bridge for each location with a failed strain gauge 44

Overview (cont.)

Exek(n.m Nuclear To account for local shell effects, both 1/2 bridges at each MSL location were averaged together and the resultant strain measurement was converted to a dynamic pressure

• For locations where a strain gauge failed, the resultant strain measurement was determined by combining the 1/4 and 1/2 bridge data

• An evaluation was performed to assess the effect of losing five strain gauges at QC1 45

Evaluation Exek(5 nM Nuclear

• Due to similarities between QC1 and QC2, the strain gauge measurements can be compared and used to validate combining 1/4 and 1/2 bridge

• Structural characteristics between the two units are comparable

- MSL pipe characteristics Frequency content and magnitude

- Time history characteristics (root mean square (RMS), max, min)

- Relationship between orthogonal planes (i.e., cross spectral density (CSD))

• 1/4 bridge data obtained for QC2 was compared to equivalent 1/2 bridge data 46

Strain Gauge Locations ExeI O n M Lsi1 S9

LjJ S7 LL1 LEJEIJ LiI1 L1 LJ II IS39

K1 S12A

_-M.

LE L1 LI1 I-- S41A I VIIM EJM6 MSL Z

MSL L1=3 MSL

.MSL Legend:

lI

- - Failed SG I -Half Bridge 47

MSL Pipe Characteristics Exeki5nM Nuclear

• MSL piping for both units are essentially identical except for some valve locations and the HPCI connection

• Pipe size, material, configuration, and relationship to the reactor pressure vessel are the same

• Dynamic response will be similar; specifically, for a similar excitation, the piping will respond the same and provide similar vibration and acoustic measurements 48

QC1 and QC2 Spectra Comparison Exekin.1 Nuclear Profiles of the spectra are similar

- Overall amplitudes across the spectrum are the same except

• QC1 A651, QC1 C651, QC1 D651 - where each has relatively higher amplitudes at 78.6 Hz and 157.7 Hz

- Predominant frequencies are similar between the units

• QC1 frequencies - 23, 78.6, 138.7, and 157.7 Hz

• QC2 frequencies - 23, 139.2, 150.9, and 154.8 Hz 49

QC1 and QC2 Spectra - Typical Exe k5n.

Nuclear QC1 D624 QC2 D624 Unit 2 MSL D Combined 624 1.OOE01 20 40 60 30 100 120 140 160 Fnq..n.y [H.]

160 0

20 40 60 80 100 Frequenc.y [Ml]

120 140 160 160 Comparison of QC1 and QC2 frequency spectra shows similar frequency content and magnitude 50 C7-L

QC1 and QC2 RMS Comparison Exe 6nSM Nuclear QC1 QC2 Description

'RMS RMSavg Description RMS RMSavg S1 A651

-0.678 l

ISi/S3A651 0.305 S2/S4 A651 0.459 0.49 S2/S4 A651 0.422 0.572 S5/S5A A624 l

0.3

-S5S5A A624 094 S6A A624 0.885 0.876 S6/S6A A624 1.223 1.151 S7/S9 B651

-0.242 S7/S9 B651 0.323 S8/S1O B651 0.361 70.216 WS8S10 B651 0.416 0.333 S11B624 0.314 SI-I 1/S11A B624 0.319 S1V2S12A B624 0.353 T

S12/S12A B624 0.337 0.399 S33 C651 0.4011 IS31/S33 C651 0.25

-3 S32/S34 C651 1.11 0.69 S32TS34 C651 0.593 0.593 S35/S35A C624 0.371 S35/S35A C624 0.272 S36A C624 0.444 0.33 7336/S36A C624 0.399 0.319 37/S39 D651 0.256 S37/S39 D651

.449 S38/S40 D651 0.397 0.237 S38/S40 D651 0.572 0.344 S41/S41A D624 T38 S41/S41A D624 1.151 S42/S42A D624 1.036 0.325 S42/S42A D624 1.512 0.427 RMSavg 0.438 RMSavg 0.517 Total nns 8.993 Total

-ns 9.456 RMS Avg 0.562 RMS Avg 0.591 RMS IP avg 0.493 RMS IP avg 0.498 RMS OP avg 0.631 RMS OP avg 0.684 Comparison of QC1 and QC2 shows the difference is less than 18%

51

QC1 and QC2 RMS Comparison (cont.)

Exe sn.SM Nuclear Comparison of QCI to QC2, RMS Strain 1.600 E

0 aI

a.

tM L

0 0

L XL M

a-a-

X)

CL aL 0

CL O-

0)

a.

L X L L X L 0

O, C

(N T

em e

E

( N 0

N). c O

e N

CD C

D C

D C

D C

Location U

• U2

• U1avg U2avgl QC1 RMS values are comparable to QC2 52 C<lh

RMS Comparison by Elevation xek 5nSM Nuclear Ratio of RMS Averages for 651 to 624 Elevation 2.500 2.000 1.500 4-

.2 1.000 --

0.500 -

0.000 A651/A624 B651/B624 Pipe lUit 1 *Uni2l C651/C624 D651/D624 Ratio of RMS values are similar for QC1 and QC2 at each elevation J 53

CSD Comparison Exe bn.M Nuclear

• The CSDs were calculated for both units

• A CSD was calculated from the power spectral density (PSD) of each orthogonal bridge pair and graphed as magnitude and phase versus frequency

• Results show that, with the exception of A651, the phase is relatively close in amplitude in the 10 400 range 54

CSD Evaluation Exe knM Nuclear QCI TCISa MSiLB 651 Cross Spectm S7/9,S8/IO 0.02 0.018 0.016

-g 0.014 0.012 o

B 0.01 B 0.008 i.

0.006 UL 0.004 0.002 0

X 23.93 Y: 0.0037S 0

20

,I- _--

I 71 I

I i

I i

I I

I IJ9.

I ml I

I r ~

~ ~~ ~t

-r X IS7.7

___Y: 0.017 l

--I 40 60 80 100 120 140 160 180 2 00 L4

(.

150 100 50 0

-50

-100

-I50 I

_ I Y; 6.237 r

i

__t_

20 40 60 80 100 120 140 160 ISO 200 Pw4quency, Itr Typical QC1 Phase Relationship 55

CSD Evaluation (cont.)

Exe IbnSM Nuclear 0.03 0.025 0.02

.r5 0.015 i

0.01 uz J 0.005 r

1 I

I I

I1 X 2 1.97 r-0

- -~ --

r-20 40 60 QC2 Tr,41 MSLC 651 Ciusa Spectma S31/33.S32/34 I-X 154.8 Y

10.928 Y:0.0 1 162 Y 001336 - --

L - -

-- L L

- -I 80 100 120 140 160 ISO 200 150 8 oo 50 I

I.

-50 1.100 J

J,,X21.9 Y

I 7.017-9

~

~ I

\\

I FI'quency, H.

Typical QC2 Phase Relationship 56

Phase Comparison Exe I5nSM Phase for 157.7 (QCI) and 154.8 (QC2) Hz Nuclear 8

0a 0i 180 160 140 120 100 80 60 40 20 0

I B624 C651 Location E Unit 1

• Unit 2 1 I

A651 A624 B651 C624 D651 D624 Phase relationship is consistent between the two units 57 Q7I& 1

/4 and / Bridge Comparison Exeknm Nuclear

• 1/4 bridge data was obtained for QC2 MSL B and C

• Review of the data shows that all strain gauges worked for the 1/4 bridge test

• Comparison of 1/4 and 1/2 bridge combination to a two 1/2 bridge combination confirms that the combination with a 1/4 bridge is almost identical to the combination with only 1/2 bridges 58

/4 and /2 Bridge Comparison (cont.) Exekn.M Nuclear QC2 C651 Avg( /4,1/2/2)

QC2 C651 Avg( /2,/2)

Sample Rate - 2000 s ps Tlie Moration - 200.2 sec PowerSpectral Deasity UL2 NML QB Test, (2VS33+S32+S34)/4. Ch 28 Date:

13-Jui2005 Fr: U2 Quar Bridge 7-7-0 Soupic Rotc

- 2000 sps Tsne Durativn - 200.2 sec PowerSpectra l Desoity U2 NML QB Test,(S31+S33+S32+S34)/4. Cb 27 Dotc: 13-Jui2005 Fi: 2 QuarBridge 7-7-0 10 f IO Fr _

1

-I-1 I -- --

-2 I-104 II I

I I

I I

iI i

K4\\

I I

I I

, I l I

I I!

'I N

I Ij*'

I r

--1 2.98 Notch Filer I

I I

sn'1

' r 20 40 60 00 1OO 120 140 I "

.i2 Y.

i 10_

10 0

r\\I Li I

- -i-= -

~F-t--1 S

I I I

I I

I C

I I

I

!I I

I l

I Ii l

I

-s v\\,Ji r v l

l a

j t 200 2

60 I

I 120 140 1

0 0

20 20 40 60 0

O 2

4 6

0 0

lo'

76 s On (60&IO01k) 200 Ok 10to 0

160 180 200 Frequency [1k Frequency [zk]

Comparison of 1/4 and 1/2 bridge shows similar I

frequency content and magnitude 59

Conclusions Exekinsm Nuclear Based on similarities between QC1 and QC2 MSLs and the strain gauge data, both units appear to be similar in both the pressure excitation of the piping and the response to the loading

• The consistency provides a measure of the quality of the data for both units

• Consistencies between the MSL strain data shows that, even with some structural differences, both units appear to respond the same due to the pressure excitation of the piping 60

Conclusions (cont.)

Exe I n.M Nuclear

• Review of the QC2 1/4 bridge data confirms the combination of a 1/4 bridge and a 1/2 bridge produces results that are almost identical to the averaged two equivalent A bridge results 61

Strain Gauge Uncertainty Exek5nSM Nuclear

• Uncertainty analysis consists of the uncertainty of the strain-to-pressure calculation and the measurement system

• Overall uncertainty is determined by a square-root-of-the-sum-of-the-squares (SRSS) method, used from random error

- Nominal pressure is calculated

- New pressure is recalculated by changing each variable by the amount of potential error, one variable at a time

- Difference between the nominal pressure and the new pressure is squared, summed, and the square root taken of the result 62

Strain Gauge Uncertainty (cont.)

Exe hbn.M Nuclear

• Actual pipe thickness were obtained via ultrasonic testing (UT) measurement

• Uncertainty calculated as 5.03%

63

Exe OnSM Nuclear Uncertainty Evaluation Kevin Ramsden Senior Staff Engineer 64

Uncertainty Evaluation Exekt nM Nuclear

• Sources of uncertainty

- Measurement accuracy of the strain gauges

- Accuracy of the ACM

- Measurement accuracy of the QC2 pressure measurements used for validation of the ACM

- Accuracy for the finite element model (FEM) 65

Uncertainty Evaluation (cont.)

Exe oIn.

• Strain gauge measurements Wall thickness of the pipe

• UT measurements made at QC2/QC1 for most accurate results Strain gauge measurements - two functioning pairs

• Assessed to be 5.03%

MSL pipe structural response

• Two functioning pairs of strain gauges yield more accurate results

• Loss of in-plane strain gauges results in conservative input for load definition Nuclear S34 I

l S32 1

I Outof-plane I A 77

anc

anc Unit 2 TC 41a MSL "C" El 651 S31_S33 -

Combined S32_S34 10 w

0 I-5 0

-5 MSL 66

-10 Time - Seconds C,_)

Uncertainty Evaluation (cont.)

Exek6nsm Nuclear

• Measurement accuracy of dryer pressure instruments 3.9% absolute measurement uncertainty 2.9% relative measurement uncertainty

• Relative measurement uncertainty is the most appropriate value to apply, since variations form the mean are the measurements of interest

• Dome vs. flush mounted pressure instruments Dome mounted sensors tend to over-predict by 3% to 8%

Wind tunnel testing showed that dome mounted sensors contain the appropriate frequency content Since the trend is toward over-prediction, no additional uncertainty correction is needed for dome mounted sensors 67

Uncertainty Evaluation (cont.)

Exekrn.M Nuclear

• ACM uncertainties Minimum error ACM

• Uncertainty evaluated at 3.6% for the applied pressure on the dryer face Modified 930 MWe ACM

• Tends to over-predict pressures, especially at the skirt natural frequency Statistical Comparison of Min Error and Modified 93OMWe ACM Sensor RMS pressure psi Min pressure psi Max pressure psi P-3 min err 0.603

-1.894 1.971 P-3 mod 930 0.682

-2.262 2.193 P-12 min err 0.625

-1.94 1.906 P-12 mod 930 0.659

-1.751 1.848 P-20 min err 0.541

-1.648 1.738 P-20 mod 930 0.605

-1.977 1.994 P-21 min err 0.638

-1.974 2.026 P-21 mod 930 0.804

-2.289 2.337 P-24 min err 0.288

-1.129 1.016 P-24 mod930 0.251

-1.034 0.986 68

Uncertainty Evaluation (cont.)

Exekn.m Nuclear FEM uncertainties

- FEM is based on detailed as-built dryer geometries

• Specific uncertainty associated with the FEA analysis calculation has not been determined due to the complexity of the model

• Sensitivity analysis of performing varying time step by +/- 10%

and using the most conservative results covers the potential for uncertainties

- Comparison of the measured strain gauge response to the predicted strains based on the ACM load and FEA shows results are generally bounding 69

Uncertainty Evaluation (cont.)

ExeI6nsm Uncertainty Terms in Unit 2 Dryer Analysis Nuclear Uncertainty Absolute Effect on Analysis Term Effect %

Strain Gage Measurement 5.02

+/- 3.6% based on minimum error model sensitivity Strain Gage Failure Impact N/A N/A Pressure Sensor 3.9 Absolute

+/- 2.9%

Measurement 2.9 Relative Pressure Sensor N/A

+3 to +8%

Phenomenological ACM Uncertainty Modified 930 MWe ACM used for Unit 2 analysis is conservative compared to minimum error model, particularly in skirt region and front hood Structural FEA Bounding values selected based on +/-10% time step sensitivity cases, plus other attributes of FEA noted in section 2.4 Net Effect Range: underpredict by 3.5% to overpredict by 14.5%

• Plus conservatism introduced by use of modified 930 MWe ACM QC2 Dryer/Skirt 70

• QC1 Dryer

Exe kbn.M Nuclear QC2 and QCI Load Definition Guy DeBoo Asset Management Engineer 71

QC2 Load Definition ExekrnSM Nuclear

• Steam path data collected at Test Condition 41 (TC41)

- 930 MWe

- 2885 MW-thermal (MWt)

• ACM used for load definition

- Modified 930 MWe ACM

• As noted previously, this ACM tends to over-predict load especially at the skirts natural frequency 72

QC2 Load Definition Exe'I oknSM Nuclear

• Examples of modified 930 MWe ACM margins 93() NIN\\Ve: P3i

-r-

,--- _r -

r--r V

01)I F

-I

-1 e,

. r'l 0).01 01)10 I

( O(M(R I

I Data i

i

--- Modified Preedictioli

___.I

--- -- r---

-- ---- r -

I I1 t~

.\\;

I I

1 I 4- %

I

.^

I I

, 4 i,

.~i, e l

rI

,-1.1 I, ;

V :

.1,'

_j I

I i(

((

10 C)

Frequency (Hz) 9 M\\Y N e: P1 '

N

-1 J"

I'l

0. I r -----------

. l Data

-Mlodified Prediction 000 IF-

-I, k

~~......

tI 0 01 1()-

---

i--

- I l-0

50) 100(

Freclulencv (Hz I I'I ';

4 I

I i v

p. -

i...

%A I "

I -O I 2100 73

QCI MSL Data Collection Exek5n.m Nuclear

• Additional MSL strain gauges were installed to more closely match the QC2 configuration

• For data collection during the QC1 startup following dryer replacement, a total of five strain gauges failed - S3, S6, S1 1A, S31, and S36

• Startup test maximum power achieved

- TC15a: 2887 MWt

- TC15b: 2901 MWt

- Minimal difference in maximum-minimum strains

• QC1 input to ACM used TC1 5a data 74

QC1 AC MSL Data Input Exek6n.m Nuclear

• All available MSL strain measurements at each location were used to define the pressure

• As discussed previously, due to reduced number of strain gauges, data was corrected for non-relevant signals at 80 Hz Results of QC2 first acoustic circuit assessment at 790 MWe

• 80 Hz frequency amplitude was higher using a strain gauge pair In-vessel instrumentation on QC1

• High speed pressure data was recorded for the vessel level instrument taps (59A and 59B) located in the dryer skirt region This data provided an understanding of the acoustic pressure field in the vessel 80 Hz was not a predominate frequency MSL instrumentation on QC1

• MSL C elevation 624

• Strain gauge data showed little 80 Hz contribution

• Accelerometers were installed on the MSL C electromatic relief valve (ERV)

Accelerometers had little 80 Hz amplitude 80 Hz + 4Hz reduced to a level consistent with MSL D 75

QC1 AC Load Definition Exel6n.M Nuclear

• First load case (TC1 5a) was based on minimum error ACM The only adjustment was to combine S33 with strain gauge pair S32/S34 at the MSL C 651 elevation This load case was used to structurally evaluate the skirt on QC1

• Second load case (TC15a 2) was based on the modified 930 MWe ACM Combined functional individual strain gauges with strain gauge pairs for those locations with damaged strain gauges Adjusted the remaining 80 Hz amplitude to equal MSL D This load case was used to structurally evaluate the dryer and vessel brackets

• Third load case (TC1 5a_3) was based on minimum error ACM Combined functional individual strain gauges with strain gauge pairs for those locations with non-functioning strain gauges Adjusted the remaining 80 Hz amplitude to equal MSL D This load case was compared to the first load case to demonstrate the TC15a load was bounding 76

QC1 AC Loads Minimum Error ACM Exe knSM

().1 N

f r_

a-0.1:)1 (00 1 (0001

10) '

K

!Ž-v Nuclear 10 11-.

.L.

C)

~

A tW 10(1) 1 0 Frequency (Hz) 20,1)

PSD col0x1arisonl betmveen TC15a ('black cIIA-e an md T( I a S preaslre sensor number P.,.

4 blue cur-v-e for I v,

I I

r s

I Z,,

e-

-r "I

(.1

(.01) 0.(( 1 0(0(01 Io I I

'V 1A/'1 7/

--3 t 9 F

~-l

... I.... I........,

I j Ii

)

_f I 0())

I S,()_(

10(

(,

Frequency (Hz)

PSD comparison between TC15a (black curve) and TC15a_3 (blue cuirve) for pressure sensor numiber P12.

77 0-19

QC1 AC Loads Minimum Error ACM (cont.)

Exekln.

1

I T

Nuclear 0.001 10-6 0

50 100 150 200 Frequency (Hz)

PSD comlparisoln between TC15a 4 black cuiA-e) and T( I1a -

(blue cti-e I fori pre sure sesor nubner P'O.

l 0.1 0. 00 1

---

..............A.........

0.00 1 C16 0

50 100 150 200 Frequency (Hz)

PSD co-miparisoii between TC15a (black curv-e) and TC I5a 3 WIbe cun-e) 1olr 78 pressiure sensor number P' I.

Cy1'c

Exe knSM Nuclear Structural Analysis Summary Guy DeBoo Asset Management Engineer 79

Time History Analysis Methodology Exek6n.

Summary Nuclear

• FEA methodology used was generally the same as the methodology used for dryer design analyses

• Three time history analyses are run for each load case:

nominal, +10%, and -10% frequency shifts

• Fatigue analysis performed using weld factors applied to time history analysis results

• Disposition of high stress locations using Local solid finite element models with forces extracted from the full shell model Increased damping to 4% for skirt and vane banks

• Minor shell model changes for QC1 FEAs 80

FEA Presentation Exekrnsm Nuclear QC2A FEA Flow induced vibration (FIV) pressure time histories from TC41 at 2885 MWt

- Frequency content: load and structural response

- Strain gauge comparison: measured vs. FEA results

- Analysis results: fatigue, ASME cases, vessel bracket

• QC1 B and QC1 D FEA

- FIV pressure time histories from TC15a at 2887 MWt

- Model differences from QC2 and April evaluations

- Frequency content: load and structural response

- Analysis results: fatigue, ASME cases, and vessel bracket 81

Exek nSM Nuclear QCI and QC2 Dryer Structural Analysis Leslie Wellstein General Electric 82

Exekon.m Nuclear The next 13 slides contain information that is proprietary to General Electric 83

Dryer Model Full Model without Super Elements Exe On.M Nuclear 84

FEM Boundary Conditions Exe InSM Nuclear 85

Full Model with Super Elements Exek nSM Nuclear 86

QC1 and QC2 Replacement Dryer Time History Analyses Exekr5n..i Nuclear 87

QC2 FEA Maximum Differential Pressure =

Exel nSM Nuclear 88

QC2 Loads Frequency Content Exe k nSM Nuclear 89

QC2 Frequency Domain Response xekrn.

Nominal Loads: Outer Hood Nuclear 90

QC2 Frequency Domain Response Nominal Loads: Skirt, Vane Bank Top and End Lxe fl SM Nuclear 91

Strain Gauge Locations Exe On..,

Nuclear 92

Strain Comparison Exe Onm Measured vs. Analytical (QC2A FEA, % Damping)

Nuclear 93

Strain Comparison Hood with % Damping Exe I nM Nuclear 94

QC2 Time History Analysis Design Margins Exek n.M Nuclear 95

QC2 ASME Code Case Stress Margins Exe lIinSM Nuclear 96

QC2 Structural Analysis Conclusions n M Nuclear

• Replacement dryer and support bracket meet the design fatigue limits for EPU conditions

• Replacement dryer and support bracket meet the ASME Code limits for all service levels (normal, upset, and faulted)

• QC2 replacement dryer is structurally adequate for EPU conditions 97

QCI Analysis Model Changes Exekmn.

Nuclear

• Skirt Super element water mass was reduced to account for the steam separators

• Mounting blocks Shell elements were added to more accurately represent the mounting blocks, and the stiffness was increased for some existing shell elements at the mounting block locations

• Trough attachment to the outer hood

- Trough attachment to the outer hood closure plate at the mounting block locations was modified to more accurately represent the connection between the outer hood closure plate and the drain trough

• Closure plate

- Angle attachment (flange) on the closure plate was added to the model 98

QC1 FEA ExekdnM Nuclear

• QC1 D acoustic load used for the dryer

- Loads from the same ACM as TC1 5a_2, which predicts skirt strains an order of magnitude higher than those measured on QC2

• QC1 B acoustic load used for the skirt

- Loads from minimum error ACM (i.e., load case TC15a) 99

Exe In.SM Nuclear The next 10 slides contain information that is proprietary to General Electric 100

Closure Plate Flange Exe InSM Nuclear 101

Mounting Block Exe OnM Nuclear 102

Trough Attachment to Outer Hood Closure Plate ExekOnSM Nuclear 103

QC1 D FEA Load Input Maximum Differential Pressure =

Exe k6n.M Nuclear 104

QC1B FEA Load Input Maximum Differential Pressure =

Exe In.M Nuclear 105

QC1 Load Frequency Content ExekrnSm Nuclear 106

QC1 Frequency Domain Response xekrns.

Nominal Loads: Outer Hood Nuclear 107

QCI Frequency Domain Response xekn QC1 +10% Loads: Skirt Nuclear 108

QC1 Time History Design Margins Exe Onm Nuclear 109

QCI ASME Code Case Stress Margins Exe k6n.M Nuclear 110

QC1 Structural Analysis Conclusions Exek 6nM Nuclear

• Replacement dryer and support bracket meet the design fatigue limits for EPU conditions

• Replacement dryer and support bracket meet the ASME Code limits for all service levels (normal, upset, and faulted)

• The QC1 replacement dryer is structurally adequate for EPU conditions 111

Exe l n M Nuclear Dresden Steam Dryer Replacement Overview Dan Pappone General Electric 112

Going Forward Plan Exe 6nSM Nuclear

• Leverage QC experience

- QC2 loads and instrumented dryer results bound Dresden generally

• No significant changes to replacement dryer design for Dresden

- Remove modifications that were not structurally significant

- Perform structural analyses to confirm Dresden configuration remains acceptable 113

Exe I n.M Nuclear The next four slides contain GE proprietary information 114

Components Considered for Removal Exe I nM Nuclear 115

Components Considered for Removal (cont.)

Exe OnSM Nuclear 116

Components Chosen for Analysis Exe lOnm Nuclear 117

Modified Center Reinforcement Plates (Due to Frame Removal)

Exer n.M Nuclear 118

Other Modifications Exek(5nsm Nuclear

• Outer tee to vane cap doubler plates

• Fillet full penetration welds on the closure plate weld between the vane bank end plate and trough

• Increase load capacity of jacking bolts (under evaluation)

• Modify vane bank perforated plates 119

Load Definitions for Dresden ExekOnm Nuclear

• FIV Use QC2A load definition for FIV during normal operation

• ASME load combinations per dryer design specification

- Use QC loads for

• FIV (normal, upset)

• Static differential pressure (normal, upset, faulted)

• Acoustic, flow impact (upset, faulted)

- Use Dresden loads for seismic evaluation

• Operational Basis Earthquake (OBE)

• Safe Shutdown Earthquake (SSE) 120

FIV Load Exekdlln Nuclear

• QC2A FIV load case is applicable to Dresden

- Plant operating conditions, vessel geometry, dryer geometry are the same

• Minor internal modifications to dryer do not affect dryer pressure loading

- Difference in load definition due to differences in MSL configuration 121

Exe In.M Nuclear The next six slides contain GE proprietary information 122

Source Identification from SMT Exe IbnM Nuclear 123

MSL Configuration Comparison IExeIn.M Nuclear 124

MSL Configuration Comparison (cont.)

ExeI1n.M Nuclear 125

MSL Configuration Comparison (cont.)

Exekrn.M Nuclear 126

Branch Line Comparison Exe On.M Nuclear 127

Branch Line Comparison (cont)

ExernOm Nuclear 128

Branch Line Comparison (cont)

Exe I n.

Nuclear

• Upper steamline strain gauge comparison Red: QC2 at EPU Black: Dresden Unit 3 at 1795 MWt (-60%)

PSD of D3 to QC2 on B MSL Upper SG 1

1 0.1 0.01 PSDDmk N PSD Dk1.-oI3 I.io-5 1 10-6 1.16 0

50 100 150 Freqk frequency hz QC2 EPU D3 1795 200 200 129

FIV Load Conclusion Exe I6 n SM Nuclear

• QC and Dresden FIV load on dryer expected to be the same in low and middle frequency ranges

• QC EPU RV resonance bounds Dresden RV resonance during power ascension

• QC FIV load is applicable to Dresden 130

ASME Load Combinations ExekanSM Nuclear

• Use same loads as QC for

- FIV (normal and upset)

- Static differential pressure (normal, upset, and faulted)

• Same geometry and operating conditions

- Acoustic/flow impact (upset and faulted)

• Same geometry and operating conditions

• Same turbine stop valve closing time

• Use Dresden loads for seismic evaluations

- OBE

- SSE 131

Exe krnSM Nuclear Summary and Conclusions Roman Gesior Director - Asset Management 132