ML071290563

From kanterella
Jump to navigation Jump to search
Calculation HC-31Q-301, Rev 0, RPV Steam Dome Dynamic Pressure Sensor Data Reduction, Including LCR H05-01, Rev 1, Updated Attachment 7, Steam Dryer Evaluation Provided in Attachment 5
ML071290563
Person / Time
Site: Hope Creek PSEG icon.png
Issue date: 02/16/2007
From: Szasz G, Trubelja M
Structural Integrity Associates
To:
Office of Nuclear Reactor Regulation
References
4500382987, LR-N07-0099 HC-31Q-301, Rev 0, LCR H05-01, Rev 1
Download: ML071290563 (32)
v • d • e


Text

Attachment 4 LR-N07-0099 LCR H05-01, Rev. 1 Hope Creek Generating Station Facility Operating License NPF-57 Docket No. 50-354 Extended Power Uprate RPV Steam Dome Dynamic Pressure Sensor Data Reduction Structural Integrity Associates Report HC-31Q-301, Rev. 0

StructuralIntegrity CALCULATION File No.: HC-31Q-301 Associates, Inc. PACKAGE Project No.: HC-31Q PROJECT NAME: Hope Creek RPV Dome Dynamic Pressure Data Reduction Contract No.: 4500382987 CLIENT: PSEG PLANT: Hope Creek CALCULATION TITLE: RPV Steam Dome Dynamic Pressure Sensor Data Reduction Project Mgr. Preparer(s) &

Document Affected Pages Revision Description Approval Checker(s)

Revision Signature & Signatures &

Date Date 0 1-4 Initial Issue M. Trubeija M. Trubelja Appendix 7>Z A1-A4 MT 2/16/07 MT 2/16/07 In Computer Files G. Szasz GSZ 2/16/07 Page 1 of 4 SI Form F200IR2a

Table of Contents 1 I INT R O D UC T ION ................................................................................................................................ 3 2 TECHNICAL APPROACH ......................................................................................................... 3 2.1 Data Acquisition Parameters and Filename Convention ..................................................... 3 2.2 D ata Reduction M ethodology .............................................................................................. 3 3 OB SE RV A T ION S ................................................................................................................................ 4 4 REF E RE N CE S ...................................................................................................................................... 4 APPENDIX A TIME HISTORY AND FREQUENCY SPECTRUM PLOTS .................................. Al List of Tables Table 1 Pressure RMS (psi), 0-Peak (psi) and Max-Min (psi) Values ................................................. 4 StructuralIntegrity File No.: HC-3 1Q-301 Revision: 0 Associates, Inc. Page 2 of 4

1 INTRODUCTION On February 8, 2007 a pressure transducer was installed on the Hope Creek Reactor Pressure Vessel (RPV) level sensing line to directly measure the dynamic pressure pulsations in the RPV dome region.

The data was collected following the January 2007 maintenance outage power ascension at 100%

power. The purpose of this calculation is to determine the amplitudes and the frequency content of the recorded signal.

2 TECHNICAL APPROACH 2.1 Data Acquisition Parameters and Filename Convention The pressure data was recorded on one channel using a PCB 113A26 dynamic pressure transducer and a portable SI-MiniDASTM data recorder. The pressure transducer was installed at Pressure Differential Transmitter N017 [1, 2] located outside the primary containment on the 77' elevation. One data set [3]

was recorded using a sample rate of 2048 samples per second (sps) for the duration of 120 seconds.

Incoming analog signals were low-pass filtered prior to the digitization with the cut-off filter frequency set automatically based on the sampling rate resulting in the recorded frequency band of approximately 1 - 800 Hz.

Once acquired, the data was uploaded to a web site, www.ibackup.com. The data was then downloaded from the web site by SI for analysis. The analysis of the data was done using Matlab [4].

The raw binary data contained it the file with extension * .dta was handled directly by a macro run using Matlab code.

2.2 Data Reduction Methodology No additional band-pass filtering was performed in post-processing. Recorded signals did not contain electrical noise spikes and, therefore, no notch filters were applied.

Time Histor RMS and Maximum/Minimum Values The time history signal was assessed for the root-mean-square (RMS), maxima, and minima over the entire duration of the acquired data. The RMS is calculated from the following equation:

x2 1n-1 2

= --n*-*X n i=0 X = (1) where x = discrete vibration signal reading n = number of readings per channel in each 120 second recording The RMS and Maximum-Minimum value was determined for the time history and then the frequency spectrum was computed. The Maximum-Minimum value was determined as the difference between StructuralIntegrity File No.: HC-31Q-301 Revision: 0 Associates, Inc. Page 3 of 4

the highest and lowest values in the data block. Table 1 lists the RMS, the 0-peak and the max-min values for the dynamic pressure at the 100% power level.

The time history and frequency spectra plots are contained in Appendix A for the dynamic pressure data collected at 100% power on February 8, 2007.

Table 1 Pressure RMS (psi), 0-Peak (psi) and Max-Min (psi) Values Description RMS 0-Peak Max-Min PDT N017 0.4895 2.2839 4.5583 3 OBSERVATIONS The recorded signals are valid. The frequency spectrum plots shown in Appendix A indicate that most of the pressure pulsation frequency peaks are below 300 Hz. Most of those peaks are relatively broad indicating no presence of acoustic resonances. In addition, all of the peaks are of the relatively low amplitude (below 0.12 psi RMS), with some as low as 0.03 psi RMS (80 and 85 Hz). This is noteworthy since the Acoustic Circuit Model (ACM) identifies - 80 Hz as a resonance frequency in reactors with 251-inch diameter steam dome diameters. The two peaks at 103.25 and 105 Hz are most likely due to the Reactor Recirculation (RR) pump vane passing frequency. The RR pump speed observed a couple of hours after the data was recorded was 1281 and 1262 rpm for Loop A and Loop B pumps, respectively [5] which corresponds to vane passing frequencies of 105.17 and 106.75 Hz, respectively. Since the frequencies do not match perfectly it is postulated that the pump speeds changed slightly from the time the data was taken.

4 REFERENCES

1. Structural Integrity Teleconference Record "Pressure Transducer Location," SI File No. HC-3 IQ-201.
2. PSEG Nuclear, L.L.C. P&ID Drawing No. M-42-1, "Hope Creek Generating Station Nuclear Boiler Vessel Instrumentation," Sheet 1, Revision 16, SI File No.-HC-31Q-201.
3. Test Data, "Feb 8 1OOpct power pressure," SI File No. HC-3 1Q-202.
4. MATLAB, Version 7.0.4.365, Release 14, Mathworks, January 2005.
5. Email from Karen Fujikawa (SI) to Milton Teske (CDI), "Hope Creek Main Steam Line Strain Gage Data - Feb 2007," 2/9/2007, SI File HC-31Q-203.

StructuralIntegrity File No.: HC-3 1Q-301 Revision: 0 Associates, Inc. Page 4 of 4

APPENDIX A TIME HISTORY AND FREQUENCY SPECTRUM PLOTS StructuralIntegrity File No.: HC-31Q-301 Revision: 0 Associates, Inc.

Page A I of A4

Sample Rate = 2048 sps Time History Plot Proc. Date: 12-Feb-2007 Time Duration = 120 sec File: 20070208082807.dta Dome Pressure, 100% Power, HCLocat.1, Ch 1 3

2 1

0_-

  • 0 0

-1

-2 0 20 40 60 80 100 120 Time [s]

Figure Al. Dynamic Dome Pressure at 100% Power, Time History StructuralInteogrity File No.: HC-31Q-301 Revision: 0 Associates, Inc.

Page A2 of A4

Sample Rate = 2048 sps Frequency Spectrum Plot Proc. Date: 12-Feb-2007 Time Duration = 120 sec File: 20070208082807.dta Dome Pressure, 100% Power, HCLocat.1, Ch 1 UFvS= 0.48951 AC Notch Filters Off iax-Mn= 4.5583 Band Filter Off U) 0~

ci) I I I I I I

~0 I I I I I I I I I I I I I I I I I I E

  • I I I I i1-.-. . . . .L - .- - . . .- I .

I - I I I I I I I

  • ,*,V

,-- . . . L-I i I Ii i Li i r 100 200 300 400 500 600 700 800 Frequency [Hz]

Figure A2a. Dynamic Dome Pressure at 100% Power, Full Spectrum StructuralIntegrity File No.: HC-3 1Q-301 Revision: 0 Associates, Inc.

Page A3 of A4

0.11 0 .1 -- I-- -- - --

0.09 - - i i i--

4- --- - ------- - _- -----

0 .068-- ----------------------- ----

0.05 -- --

-- - - - - - ------ -- -- - - - - - - - - T-- ----

E

< 0.04 -- - -- ---- -- - - - - - - - - -

0.03 - - i i 0.02 _ 7 -- - --- -- -

0 20 40 60 80 100 Freq [Hz]

Figure A2b. Dynamic Dome Pressure at 100% Power, Full Spectrum, 0 - 120 Hz Plot Structura IntegritY File No.: HC-3 1Q-301 Revision: 0 Associates, Inc. __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Page A4 of A4 LR-N07-0099 LCR H05-01, Rev. I Hope Creek Generating Station Facility Operating License NPF-57 Docket No. 50-354 Extended Power Uprate REQUEST FOR LICENSE AMENDMENT EXTENDED POWER UPRATE Steam Dryer Evaluation Rev. I April 2007 Rev 1 Dated April 2007 LR-N06-0286 LCR H05-01, Rev. I HOPE CREEK GENERATING STATION FACILITY OPERATING LICENSE NPF-57 DOCKET NO. 50-354 REQUEST FOR LICENSE AMENDMENT EXTENDED POWER UPRATE STEAM DRYER EVALUATION Rev I Dated April 2007 LR-N06-0286 LCR H05-01, Rev. 1 Table of Contents 1 Executive Summary ................................................................................................. 1 2 Background ..................................................................................................... 2 2.1 Post EPU Steam Dryer Failures ................................................................... 2 2.2 BW R Fleet Operating History ........................................................................ 3 2.3 BW ROG Recommendations ....................................................................... 3 3 HCGS Steam Dryer and Main Steam Line information ........................................ 3 3.1 HCGS Steam Dryer Operating History ......................................................... 3 3.2 HCGS Steam Dryer Design .......................................................................... 4 3.3 HCGS Steam Dryer Comparison to Dryers Operating at EPU ..................... 5 3.4 HCGS Main Steam Line and Safety Relief Valves ....................................... 8 4 HCGS Steam Dryer Evaluation Methodology .................................. ......................... 9 4.1 Evaluation Phases: ........................................................................................ 9 4.2 Analytical Tools ......................................................................................... 11 5 R e s u lts ................................................................................................................... 16 5.1 Baseline Conditions (100% CLTP) ............................................................. 16 5.2 Interim Analyses for EPU Conditions .......................................................... 17 6 R e fe re n ce s ......................... I................................................................................... 19 Page i Rev I Dated April 2007 LR-N06-0286 LCR H05-01, Rev. 1 HCGS STEAM DRYER EVALUATION 1 Executive Summary The Hope Creek Generating Station (HCGS) steam dryer is a curved hood design that was further upgraded on-site prior to commercial operation. It has been inspected on a recurring basis and has shown no fatigue damage as of the last inspection, spring 2006.

HCGS used the Acoustic Circuit Model (ACM) load transfer methodology to calculate steam dryer loads at 100% Current Licensed Thermal Power (CLTP) using plant measurements from strain gages installed on the main steam lines (MSL). The ACM methodology was benchmarked at Quad Cities 2 using an instrumented steam dryer to compare predicted and actual loads. The HCGS strain gage installation replicated the Quad Cities 2 strain gage locations. The loads calculated by the ACM were inputted into a finite element model (FEM) developed specifically for the HCGS steam dryer.

The CLTP FEM analysis shows that the most limiting component on the steam dryer is not overstressed, and has significant margin for operation at increased power.

HCGS is aware of industry concerns with the operation of steam dryers under Extended Power Uprates (EPU) conditions. To reduce risks, HCGS proactively undertook steps to minimize the unknowns associated with relief valve acoustic resonance, which has been identified as the primary loading that caused damage at Quad Cities and Dresden steam dryers. This effort included a 1 /8th scale model test (SMT) of the HCGS components to determine the % wave acoustic frequency for the HCGS safety relief valves (SRV) and obtain an estimate of the SRV loading as well as the overall increase between CLTP and EPU. HCGS reran the FEM using the predicted EPU loads that included a conservative estimate of the SRV loads. This is referred to as the interim EPU FEM. These-analyses demonstrate that-the-HCGS-maintains-a positive margin for EPU operation when considering conservative loading and shifts in the frequency of the loading up to plus and minus 10%.

The EPU power ascension test plan will incorporate predetermined hold points above CLTP to allow for review and confirmation that dryer loads remain below acceptable values. HCGS will rely on the strain gage readings to monitor the changes during this initial power ascension to EPU, and HCGS will validate during the power ascension that loading, including uncertainties, will not result in unacceptable steam dryer fatigue stresses.

This Attachment summarizes actions completed or currently planned to ensure the integrity of the steam dryer at the EPU condition.

Page 1 of 20

Attachment 7 Rev 1 Dated April 2007 LR-N06-0286 LCR H05-01, Rev. 1 2 Background 2.1 Post EPU Steam Dryer Failures In June 2002, Quad Cities 2 (a BWR 3) was operating at approximately 113% of original licensed thermal power (OLTP) when it experienced a failure of a steam dryer cover plate resulting in the generation of loose parts, which were ingested into a main steam line (MSL). In April 2003, the same plant experienced a second steam dryer failure.

This second failure occurred at a weld between the outer hood and an end plate.

In October 2003, a hood failure occurred at Quad Cities 1, a sister unit to the BWR 3 that had experienced the previously noted failures. This unit was also operating at EPU conditions. The observed hood damage was virtually the same as the 2003 failure described above.

In August 2002, General Electric Company (GE) issued a Services Information Letter (SIL) (Reference 1) that recommended monitoring steam moisture content and other reactor parameters for BWR 3-style steam dryers. Reference 1 also recommended inspection of the cover plates at the next refueling outage for those plants operating at greater than OLTP.

Reference 2 broadened the earlier recommendations for BWR 3-style steam dryer plants and provided additional recommendations for BWR 4 and later steam dryer design plants planning to or already operating at greater than OLTP. Following this revised guidance, inspections were performed on plants operating at OLTP, stretch uprate (5%), and extended power uprate (EPU) conditions. These inspections indicated that steam dryer fatigue cracking could also occur in plants operating at OLTP.

Reference 2 described additional significant fatigue cracking that has been observed in steam dryer hoods and provided inspection and monitoring recommendations for all BWR plants.

Significant industry efforts were expended in finding a methodology to measure plant loads since it became apparent that generic loads were not appropriate. This resulted in the development of the Acoustic Circuit Method, which uses either strain gages or pressure sensors on the MSL to measure plant data and then calculate the differential pressure across the steam dryer.

In addition, efforts were expended to determine the source of the significant loading.

The remedial actions in the spring of 2006 that installed acoustic side branches at the Quad Cities relief valve standpipes confirmed that the majority of the Quad Cities steam dryer loading was from the relief valve 4 wave acoustic resonance. Prior to installation of the acoustic side branches, Quad Cities had been operating at essentially the peak SRV acoustic resonance condition.

Page 2 of 20

Attachment 7 Rev 1 Dated April 2007 LR-N06-0286 LCR H05-01, Rev. 1 2.2 BWR Fleet Operating History In addition to the instances described in section 2.1 for Quad Cities, steam dryer cracking has been observed throughout the BWR fleet operating history. Steam dryer cracking has been observed in the following components at several BWRs: dryer hoods, dryer hood end plates, drain channels, support rings, skirts, tie bars, and lifting rods.

These crack experiences have predominately occurred during OLTP conditions, and are described in References 2 and 4. Except for the Quad Cities events, there have not been any loose parts generated.

The operating environment has a significant influence on the susceptibility of the dryer to cracking. Most of the steam dryer is located in the steam space with the lower half of the skirt immersed in reactor water at saturation temperature. These environments are highly oxidizing and increase the susceptibility to inter-granular stress corrosion cracking (IGSCC).

2.3 BWROG Recommendations The BWR Owners Group (OG) in September 2004 issued Reference 3. Sections 3.5 and 3.6 of that document address steam dryer loads and inspections/evaluations, respectively. The two recommendations specific to steam dryers are cited below.

  • An evaluation of steam dryer loads for EPU conditions should be made prior to implementation of EPU. Modifications of the dryer or bases for not making modifications should be made based on the results of this evaluation.
  • Follow the inspection and monitoring recommendations made by the GE SIL (Reference 2) and by the EPRI BWR vessel internal project (VIP) steam dryer inspection guidelines (Reference 4).

HCGS has performed an evaluation of the steam dryer with estimated EPU loads. This evaluation is summarized later. HCGS has performed inspections in RF12 and RF13 following guidance of References 2 and 4. The Power Ascension Test Plan (PATP) addresses post EPU steam dryer inspections.

3 HCGS Steam Dryer and Main Steam Line information 3.1 HCGS Steam Dryer Operating History The HCGS reactor steam dryer went into service with the startup of the plant in 1986.

Since start-up, HCGS has concentrated on maintaining good water chemistry in the reactor coolant system, which contributes to reducing the occurrence of IGSCC.

IGSCC-type indications have been observed on the HCGS dryer as noted below.

HCGS has performed the baseline visual inspections of its steam dryer per BWRVIP guidelines (Reference 4). The baseline inspection was completed during the latest refueling outage (RF13 in spring 2006). The HCGS steam dryer indications observed during inspections performed prior to and through RF13 are listed below:

Page 3 of 20

Attachment 7 Rev 1 Dated April 2007 LR-N06-0286 LCR H05-01, Rev. 1

1. Support ring, 2050, horizontal crack, found in RF07. Measured every outage since.

RF07 through RF1 0 measured at 2.25 inch. RF1 1 measured at 2.87 inch. RF1 2 measured at 2.25 inch. The RF1 1 measurement was discounted. This indication has been dispositioned as IGSCC due to residual stress from cold forming of the support ring and its proximity to the upper weld Heat Affected Zone (HAZ). No growth observed since discovery.

2. Skirt, 50, horizontal weld below the dryer support lug, found in RF10. Horizontal crack in the HAZ below the weld. Measured RF10, 11, and 12. All measurements 0.75-inch. This indication has been dispositioned as IGSCC due to residual stress from welding. No growth observed since discovery.
3. Lifting lug, 220', upper support bracket, found broken on one side in RF1 1. During RF12 the upper bracket was removed. Left less than one-inch stub. Justification for removal on file.
4. Support ring, 200, on top, radial 0.625 inch (from edge to hood weld) and down side vertical 0.75 inches. Crack thought to be started on top and side shows depth.

Identified during RF12. This indication has been dispositioned as IGSCC due to residual stress from cold forming of the support ring and its proximity to the weld HAZ.

5. Inner curved hood, crack above a repair strip added prior to commercial operation.

Identified during RF13. Dispositioned as IGSCC. Insufficient crack opening to result in any steam bypass.

6. Lifting lug, 140', cracks near the tack welds that prevent rotation of the lifting rod eye on the lift rod. Identified during RF13. Dispositioned as IGSCC.
7. Steam outlet end plate for outer hood, crack at the bottom of the plate, about 1-1 inches long. Identified during RF13. Dispositioned as IGSCC.
8. Support ring at 2300 top surface, approximately 5-inches in length. Identified during RF13. Dispositioned as IGSCC.

None of these indications approach the critical flaw size. They will be reinspected periodically as required by their current flaw evaluations.

A key finding is that no indications have been found indicating FIV damage on any of the HCGS dryer areas. This includes the hoods, cover plates, tie bars, drain channels, and steam outlet end plates. As discussed in section 3.2, Hope Creek has a curved hood steam dryer, which was reinforced in the areas of greatest FIV concern.

3.2 HCGS Steam Dryer Design The HCGS steam dryer is typical of the late BWR 4/5 curved hood design. Per Reference 4, the HCGS steam dryer is essentially the same as Limerick 1&2, LaSalle 1&2, Susquehanna 1&2, Fermi 2, Tokai 2, 1 Fukishima 6, Nine Mile Point 2, and Columbia (12 total). The differences are primarily the extent of field modifications.

Page 4 of 20

Attachment 7 Rev 1 Dated April 2007 LR-N06-0286 LCR H05-01, Rev. I Prior to operation, the HCGS steam dryer was modified on-site with the following recommended GE upgrades to improve its structural integrity:

  • The 0.125-inch thick outer hoods were replaced with 0.5-inch hoods. In addition, the weld attaching the outer hoods to their internal, vertical hood supports was strengthened.
  • The 0.1875-inch thick central steam outlet end plates, on the outlet of the inner hoods, were replaced with 0.5-inch plates.
  • The 0.5 by 1-inch tie bars, spanning across the top of the vane assemblies, were replaced with an increased number of 2 by 2-inch tie bars. Seven (7) bars tie the outer vane assembly to its middle vane assembly. Nine (9) bars tie the middle vane assembly to its inner vane assembly. Five (5) bars tie the two inner vane banks to each other.
  • 0.1 87-inch thick reinforcing strips were added at the outer edges of the middle and inner 0.125-inch thick hoods where they are welded to their 0.250-inch end plates.

These reinforcing strips extended the full length of the weld. In addition, the inside corners of the 0.125-inch hoods to 0.250-inch end plate welds were inspected, prepped, and back-welded to a minimum height of 50-inches above the support ring.

In addition, to prevent "rocking" of the steam dryer on its reactor vessel support lugs, HCGS did tests prior to commercial operation to determine if the top of the four lugs were at equal heights. These tests resulted in machining three of the four support points to ensure that the steam dryer was supported evenly on the four lugs.

3.3 HCGS Steam Dryer Comparison to Dryers Operating at EPU Dryer design, steam velocities, and relief valve standpipe acoustic loads have been identified as contributors to dryer failures and the subsequent generation of loose parts.

The HCGS steam dryer curved hood design is an upgrade to the earlier, square hood dryers that failed at Quad Cities. These upgrades include improved flow characteristics and improved structural strength in the upper part of the dryer.

The square hood design has 4-foot high dryer vanes, sharp 900 corners at various flow points, and includes a steam dam (raised plate perpendicular to the top of the dryer).

This design inherently causes flow separation/turbulence as the steam flows through the steam dryer into the reactor steam dome. Furthermore, the square hood design results in flow separation/turbulence as the steam flows from the steam dome towards the MS nozzles since the steam encounters the outside 900 corners of the outer hoods.

The curved hood design has 6-foot high vanes, eliminates the 900 corners, and eliminates the steam dam, all of which provide a distinct advantage in reducing turbulence. Average steam flow velocities through the dryer vanes at EPU conditions will remain relatively modest (- 4 feet per second). The average velocity entering the bottom of the hoods and exiting the outlet of the hoods is approximately 15 fps.

Page 5 of 20 Rev 1 Dated April 2007 LR-N06-0286 LCR H05-01, Rev. 1 The significant structural failures in the Quad Cities dryers were at the outer hood, facing the MS nozzles. The square hood design initially used internal bracing, which provided support to the hoods only at the upper corner of the hood, resulting in a high stress condition. The curved hood dryer uses interior, vertical support plates, which provide continuous support along the entire height of the hood and eliminate the need for external gusset plates. The HCGS outer hoods consist of 0.5-inch thick bent plate welded to 0.375-inch thick end plates and, at the bottom, to a 0.375-inch thick horizontal cover plate.

Another advantage of the curved hood design is that it has a total of four wide drain channels welded to the outside of the dyer skirt. Each drain channel spans approximately 45 degrees along the circumference of the skirt and spans nearly the full height of the skirt (from the bottom of the upper support ring to just above the bottom ring). These four wide drain channels provide added stiffness to the skirt.

The table below summarizes steam dryer design and main steam line (MSL) velocities for several BWR plants that have received extended power uprates, and it provides post EPU steam dryer experience. HCGS information is included for comparison. Section 2.1 of Reference 3 provides a more complete listing of BWR uprates.

Reactor Station/Plant MSL Velocities EPU Comments Type Dryer Design fps Operation 251BWR 3 Dresden 2, 3 OLTP 168 117% RV 1/4 wave Square hood EPU 202 OLTP acoustic resonance peaked below OLTP.

Failure in the outer hood area.

251BWR 3 Quad Cities 1, 2 OLTP 168 117% RV 1/4 wave Square hood EPU 202 OLTP acoustic resonance peaked at EPU.

Failure in the outer hood area.

205BWR3/4 Vermont Yankee OLTP 140 120% Steam dryer Square hood EPU 168 OLTP strengthened before EPU.

Onset of RV 1/4 wave acoustic resonance between OLTP and EPU.

Page 6 of 20

Attachment 7 Rev 1 Dated April 2007 LR-N06-0286 LCR H05-01, Rev. 1 Reactor Station/Plant MSL Velocities EPU Comments Type Dryer Design fps Operation 218BWR 4 Brunswick 1, 2 OLTP 129 120% No cover Slanted hood EPU 149 OLTP plate/hood fatigue failures.

218BWR 4 Hatch 1 OLTP 119 115% No cover Slanted hood EPU 134 OLTP plate/hood fatigue failures Hatch 2 OLTP 121 Curved hood EPU 140 251BWR 4 HCGS CLTP 145 115% CLTP=

Curved hood EPU 167 CLTP 101.4% OLTP requested The Brunswick 1, 2 and Hatch 1 slanted hood design and the Hatch 2 curved hood have not had any failures that are attributable to EPU based on the information presented in Table 2-2 of Reference 4, Steam Dryer Inspection Results. Hatch 2, like HCGS, has 0.125" inner hoods.

The curved hood design at HCGS was the next stage of steam dryer design. It was used in later BWR 4 units and in BWR 5s and 6s. Similarities with the slanted hood design include 6-foot high dryer vanes, internal, vertical support plates, and elimination of the upper dam. The primary difference is that the curved hood uses a single bent plate rather than four straight plates in forming the hood.

One design drawback of the earlier curved hood design, which applies to the HCGS steam dryer, is that the inner hoods are 0.125-inch thick, whereas the inner hoods on the previous square and slanted hood designs were 0.5-inches thick. This has led to two types of curved hood failures described below.

Per Reference 2, weld joint cracking was found in the weld between the 0.125-inch middle hood and its 0.250-inch end plate at five plants. Cracks at four plants occurred early in plant life, within the first three or four cycles of operation. The weld joints were subsequently reinforced, either by adding reinforcing strips or by adding an additional weld on the inside of the joint inside the hood. The fifth occurrence occurred after about 16 years of operation, the last 9 years at 5% stretch power. This dryer had been reinforced in that weld joint with additional, inside welding except at the upper part of the weld. The cracking occurred in the upper part of the hood where the joint was not reinforced. The fix implemented was to add external reinforcement strips. As stated earlier, this area was reinforced at HCGS on the middle and inner hoods prior to the start of commercial operation by adding the external strips along the entire length of the weld and by back-welding on the inside to a minimum height of 50-inches. No failures Page 7 of 20

Attachment 7 Rev 1 Dated April 2007 LR-N06-0286 LCR H05-01, Rev. 1 have occurred at HCGS, and this detail received careful modeling when the HCGS finite element model (FEM) was prepared.

Per Reference 4, weld joint cracking has also been reported on the curved hood steam outlet end plates. These plates serve as a steam dam that bridges the gap between the vane bank outlet end plate and the next hood. They ensure that the steam exiting the vane banks flows upward into the steam dome rather than laterally. On one side, they overlap the vane bank end plates and are attached to it by a fillet weld. At the other side, they are fillet welded to the outside of the adjacent hood. Since these plates are on the steam outlet, weld cracking would not result in steam bypassing the vane bank.

No failures have occurred at HCGS.

3.4 HCGS Main Steam Line and Safety Relief Valves HCGS has four 26-inch nominal diameter steam lines in containment. The lines increase to a nominal 28-inch diameter in the Main Steam tunnel outside of the containment. At EPU, the velocity is estimated at 167 fps in the 26-inch line and under 140 fps in the 28-inch line.

HCGS has a total of fourteen (14) relief valves on the MSLs. Two MS lines have 4 valves each; two MS lines have 3 valves each. As opposed to earlier plants, the HCGS design has only one type of relief valve. This is the Target Rock 7567F design, which combines the function of a safety relief valve (SRV) to prevent overpressurization and a power operated relief valve to provide controlled depressurization and cooldown.

HCGS has collected the required information for detailed modeling of the SRV standpipe and valve in the scale model tests. The relief valve standpipe and valve, including heights, are identical at all 14 locations. The standpipe configuration is a 26-inch to 8-inch sweepolet fitting, an 8-inch nominal diameter schedule 160 pipe stub, and a flange that bolts up the bottom of the relief valve. The flange serves a second function. The inside diameter (ID) of the reducing flange is tapered to transition from the 6.8-inch to 5.2-inch at the entrance of the SRV. The ID on the inlet of the relief valve is 6.0-inch. The uppermost portion of the SRV chamber is rounded. In addition, there is a 1 5 th standpipe location that is a spare. The standpipe is identical, but instead of a valve, it has a blank flange. It is on the "B" MSL line.

The "A" and "D" MSLs are mirror images of each other. The "B" and "C" MSLs are also mirror images with one exception. The "B" line has a fifth standpipe location that is blanked off. The spacing of the standpipes with SRVs is essentially identical except for minor fabrication differences, which result in a maximum of 0.04-feet (0.5-inch) difference in the spacing. Refer to the table below. All SRVs are mounted in sections of MS lines that have flow thru the lines. The HCGS design does not have dead-headed MS branch lines.

Page 8 of 20

Attachment 7 Rev I Dated April 2007 LR-N06-0286 LCR H05-01, Rev. 1 "A"and "D" lines Standpipes Distance From Upstream Elbow (ft) (respectively) 1 st Standpipe with Target Rock 5.13, 5.09 2nd Standpipe with Target Rock 8.14, 8.15 3 rd Standpipe with Target Rock 11.16, 11.17 "B" and "C" lines Standpipes Distance From Upstream Elbow (ft) (respectively) 1 st Standpipe with Target Rock 6.95, 6.95 2 nd Standpipe with Target Rock 10.11, 10.07 3 rd Standpipe with Target Rock 17.90, 17.86 Spare Standpipe 21.06 (Only in the "B" line) 4 th Standpipe with Target Rock 24.21, 24.17 4 HCGS Steam Dryer Evaluation Methodology 4.1 Evaluation Phases:

The HCGS process for demonstrating that the steam dryer is adequate for operation at a higher power level consists of three distinct phases. These are: (1) establish that there are adequate margins at CLTP (2) perform a SMT to identify and estimate any significant, new loads above CLTP, and (3) perform a power ascension test plant (PATP) to verify that steam dryer stresses are not exceeded. These are discussed in 4.1.1 thru 4.1.3 below.

The analytical models that are used for one or more of these phases are discussed in section 4.2.

4.1.1 Phase 1: Establish margins at CLTP HCGS used in-plant measurements to calculate the flow induced vibration (FIV) stresses at CLTP. The key reports that support this phase are:

a) Reference 16, Continuum Dynamics Incorporated (CDI) Report 07-01 RO:

HCGS gathered in-plant data in mid-2006 and again in early 2007. The mid-2006 data was based on readings from only two of the 4 MSLs, and therefore used an algorithm to conservatively bound the in-plant loads. The 2007 data had readings from all 4 MSLs.

Sections 3.1 and 3.2 of Reference 16 provides the information for both CLTP in-plant readings as well as a comparison of the 2006 and 2007 loads.

Page 9 of 20

Attachment 7 Rev 1 Dated April 2007 LR-N06-0286 LCR H05-01, Rev. 1 b) Reference 10, CDI Report 06-24 R3:

The mid-2006 in-plant data was used to generate high resolution loads which were inputted into the CLTP finite element model (FEM) analysis, Reference 10. Note that Reference 16 demonstrates that the more accurate in-plant data gathered in early 2007, predicts lower loads than those used for Reference 10.

The results of the CLTP FEM analysis are summarized in section 5.1.

4.1.2 Phase 2: Identify any significant, new loads at EPU conditions.

Based on industry experience, the primary concern with a power uprate is a high magnitude load at a discrete frequency that is either not present at CLTP or else would not be scalable from data taken at CLTP. The specific concern for HCGS is pressure pulsations due to flow instabilities across SRV standpipes, which occur only over a specific steam flow velocity range. Thus they are not scalable or predictable using CLTP data. Dresden experienced the peak of this phenomenon below CLTP, and Quad Cites experienced the peak at EPU.

HCGS reviewed the discrete frequencies for the CLTP in-plant data and determined that relief valve acoustic response is not present at or below 100% CLTP. This included a review of the May 2006 strain gage readings at 95.9%, 97.5%, and 100% CLTP power. There is no appreciable change in the overall magnitude and no indication of SRV 4 wave acoustic resonance. The 2007 strain gage data does not indicate SRV acoustic resonance. The MSL accelerometers likewise did not detect any discrete frequencies other than those corresponding to the 5x recirculation pump speed.

However, available literature indicates that based on the Strouhal number calculation, the onset of SRV 1/4 wave acoustic resonance could be anticipated at HCGS at or above CLTP. The available information indicates that the loading at the onset conditions is much lower than at peak conditions and that it is dependent on a number of other factors. The experience gained on FEM analyses demonstrated that the resulting stresses are dependent not only on the magnitude of the loading, but also on the frequency.

HCGS elected to use scale model testing (SMT) of the HCGS steam delivery system to predict the SRV loads. This does not substitute for verification during EPU power ascension (Phase 3, section 4.1.3). PSEG will perform similar verification steps during its power ascension to ensure steam dryer integrity as was done by the Vermont Yankee program, which did not have the benefit of a SMT.

It is stressed that the SRV loading is the key output of the SMT effort. The broad-band, non discrete loads, are far less critical since these are already present at CLTP, and phase I shows that there is significant margin for the broad band loads. These are the key reports from the SMT phase:

a) Reference 13, CDI Report 06-16 R2:

This report summarizes the 1 /8th SMT testing done over a wide range of reactor powers. Revision 2 updated the report following the re-benchmarking effort Page 10 of 20

Attachment 7 Rev 1 Dated April 2007 LR-N06-0286 LCR H05-01, Rev. 1 completed early 2007. Although a key part of the re-benchmarking effort was to remove excess conservatism in the SMT, the on-set of SRV acoustic resonance was treated conservatively. Specifically, the in-plant data (CDI Report 07-01) shows no SRV acoustic resonance at CLTP whereas the re-benchmarked SMT shows some SRV acoustic resonance at CLTP.

b) Reference 16, CDI Report 07-01 RO:

Section 3.3 of this report provides a benchmark at CLTP between the 2007 in-plant data and the SMT. This report shows that the SRV loading in the SMT at CLTP is conservative with respect to the in-plant data.

c) Reference 15, CDI Technical Memo 06-23:

This Technical Memo provides a comparison on the predicted HCGS loads at EPU against the Quad Cities 2 EPU loads (before the Quad Cities 2 load mitigation). It demonstrates that the predicted HCGS loads are small in comparison with the Quad Cities load.

d) Reference 11, CDI Report 06-17 R3:

Reference 11 provides the MSL power spectral density (PSD) figures for the SMT at CLTP and EPU and it provides the low-resolution comparison of the SMT loads between CLTP and EPU.

e) Reference 14, CDI Report 06-27 R2:

The SMT EPU data was used to generate high resolution loads which were inputted into the EPU "interim" finite element model (FEM) analysis.

The results of the interim F EM analysis are discussed below, in section 5.2.

4.1.3 Phase 3: Power Ascension Test Plan (PATP).

Because of uncertainties in any predictive SMT used for estimating new loads, HCGS will do a slow power ascension above CLTP to gather in-plant measurements at specific power steps. Specifically, PSEG will follow similar key verification steps during the power ascension as was done by Vermont Yankee, which did not have the benefit of predictive information from a SMT. The PATP data will be used to validate that the FIV stresses remain below the code allowable.

The PATP is provided separately as attachment 23 to the EPU LCR.

4.2 Analytical Tools 4.2.1 Acoustic Circuit Model The method used to determine the loading across the HCGS steam dryer is called the Acoustic Circuit Model (ACM) developed by CDI. The ACM requires sensors mounted on the MSLs to detect and measure plant specific, pressure pulsation loads. The ACM uses a mathematical model of the HCGS main steam system including the steam dome, Page 11 of 20

Attachment 7 Rev 1 Dated April 2007 LR-N06-0286 LCR H05-01, Rev. 1 steam dryer in the reactor pressure vessel, and the MSLs including the sensor locations to calculate steam dryer loads from the measured MSL pressure pulsations.

Specifically, the ACM model provides a high-resolution load over a grid mesh of three inch spacing across all surfaces of the steam dryer. The ACM is applicable to in-plant data and SMT data. In both cases the loads are obtained from MSL sensors. The difference is that in-plant data uses strain gages whereas the SMT uses pressure transducers.

The strength of the ACM is that it does not need to identify or model the source of the MSL pressure pulsations. This is crucial because analyses of main steam line pressure data at BWRs shows that there is broad band loading due to turbulence, but it also showed the presence of pressure pulsations at discrete frequencies, which suggests that deterministic mechanisms are also active in the MS system. Since the ACM uses plant measurements, the dryer loading reflects all sources and does not rely on an analytical approach to determine what the sources are. The ACM methodology was validated through benchmarking the predicted results against an instrumented steam dryer at Quad Cities 2.

The specific ACM load transfer used for HCGS is the "Bounding Pressure" method described in CDI Report 05-28, Reference 6. This method is the same that was used for Vermont Yankee during its EPU power ascension testing. All CDI ACM efforts for HCGS were done under CDI's QA program, which conforms to 10CFR50 Appendix B requirements. HCGS reviewed piping configuration, as modeled in the CDI design record for HCGS. HCGS verified that the appropriate drawings were used and that the correct piping geometry and dimensions were obtained from those drawings.

Note that the ACM is the methodology for estimating steam dryer loads from plant data used by all other pending EPU submittals and is in line with the approach endorsed by the BWR Owners Group in late 2006.

The in-plant input to the ACM for HCGS are strain gages at eight locations on vertical runs of the MSLs in the drywell approximating the benchmarked Quad Cities 2 locations. This consists of two locations per MSL approximately 36 feet apart. Each location presently consists of four strain gages located at 900 intervals. The two strain gages located 1800 apart were wired together to obtain two channels per location. The two channels were then averaged together to obtain an average reading for that location. HCGS is planning to provide added redundancy to support the EPU power ascension. The plan is to double the number of strain gages by placing strain gages at 450 spacing from the existing strain gages. The strain gages 1800 apart will be wired together to obtain two additional channels per location. When combined with the existing strain gages channels, there will be a total of 4 channels per location.

Page 12 of 20 Rev 1 Dated April 2007 LR-N06-0286 LCR H05-01, Rev. 1 4.2.1.1 ACM Uncertainty/Bias Evaluation The uncertainty/bias evaluation for the steam dryer loading considers the uncertainty/bias of the strain gage data inputted into ACM. This is broken down into two items.

a) The strain gage conversion of micro-strain reading to dP:

This was obtained from Reference 7. Note that it conservatively neglects any pipe bending that is not balanced out by the other strain gages.

b) The strain gage location with respect to the benchmarked QC2 location:

Reference 6 section 7.2 provides the formula to calculate the error. This is small for HCGS since the strain gage locations are very similar to QC2.

It also considers the uncertainty/bias of the ACM itself. This is broken down into three items.

c) QC2 steam dryer pressure data measurements:

Reference 8 Table 10 provides the values for the QC2 steam dryer pressure data measurements bias and uncertainty. Note that the reference states a minus 3 to minus 8% range for the bias. The minimal bias is assumed herein for conservatism.

d) ACM bias due to low frequency limitations (0-20Hz):

Reference 8 Table 10 provides the values for the ACM bias due to low frequency limitations (0-20Hz).

e) ACM bias and uncertainty on predicted versus measured QC2 data:

Reference 9 compares the QC2 measured root mean square (RMS) pressure against the QC2 predicted root mean square (RMS) pressure for the 15 pressure measurements on the QC2 outer hoods. It determines the bias as well as the standard deviation (one sigma). This comparison used the range of frequencies predicted for HCGS, 120Hz and less. The higher frequencies were not used since these are not predicted at HCGS. Also, the evaluation used the PRMS value for comparison since it provides the energy for the signal (i.e., the area under the peak).

The above uncertainties are summarized in the table below. As used herein, a minus bias over predicts the loads and a positive bias under predicts the loads. The individual uncertainties are combined by SRSS. Individual biases are algebraically added.

Term Item Bias Uncertainty Strain Gage Conversion a 0 +/- 7.2%

Strain Gage Location b 0 +/- 4.2%

QC2 dryer pressure measurement c -3% +/- 2.9%

Page 13 of 20

Attachment 7 Rev 1 Dated April 2007 LR-N06-0286 LCR H05-01, Rev. I ACM low frequency limitation d 3% 0 ACM accuracy e -10.8% +/- 25.7%

Subtotal a-e -10.8% +/- 27.2%

For conservatism, when combining the bias and uncertainty for the 0-120 Hz range, HCGS will not credit the 10.8% bias under prediction, which results in a +/- 27.2%

loading uncertainty. When evaluating individual peaks, a minus bias will be used to offset the uncertainty.

4.2.2 Scale Model Testing As stated in an ASME Journal of Pressure Vessel Technology article (Reference 5):

"High velocity flow past a cavity such as the stub of a closed SRV creates vortices which, under the right conditions, can couple with the acoustic resonance of the stub. Thus relatively small vortex pulsations can be amplified...."

In a fluid system with many junctions and branch lines of various lengths and diameters, a strictly analytical approach cannot be relied on to determine if the vortexing across the various branch lines will create acoustic resonance at that branch line, and furthermore, if the acoustic resonance in a branch line is amplified or attenuated by the piping system. Thus, scale model testing is desirable. PSEG believes that the CDI 1/8th SMT is the best technology presently available for this purpose. PSEG selected this option after the CDI SMT was able to replicate the QC2 in-plant relief valve acoustic resonance, which allowed CDI to design and test a fix on the 1/8th SMT that successfully mitigated the QC2 relief valve acoustic resonance. Using the CDI SMT is a proactive step that PSEG undertook to reduce the power ascension unknowns.

The HCGS 1 /8th SMT is described in Reference 13.

4.2.3 Finite Element Model In order to determine the steam dryer stresses due to the pressure pulsations calculated by the ACM, a HCGS plant specific finite element model (FEM) was developed. The FEM was developed by CDI using the ANSYS 10.0 program. The entire steam dryer was modeled including the skirt and the water at the lower portion of the skirt. All CDI FEM activities were done under CDI's QA program, which conforms to 10CFR50 Appendix B requirements.

The HCGS FEM development benefited significantly from the availability on-site of the abandoned steam dryer (intended for HCGS Unit 2). HCGS verified that the abandoned dryer was identical in design and fabrication to the one in use with the exception of the field modifications which were made only to the Unit 1 steam dryer. Detailed measurements were made by CDI of the abandoned HCGS Unit 2 steam dryer supplemented by the available, detailed information on the field modifications. HCGS Page 14 of 20

Attachment 7 Rev 1 Dated April 2007 LR-N06-0286 LCR H05-01, Rev. I engineers, knowledgeable on the steam dryer, supported the CDI effort by determining the applicable drawings and field modification instructions that were applicable to the HCGS steam dryer.

The HCGS FEM is discussed in References 10 and 14. They discuss the modeling refinements and studies undertaken to optimize the modeling to obtain accurate results while maintaining a reasonable model size. The FEM model received third party review by MPR Associates. The FEMs differ slightly in that Reference 14, .the interim EPU FEM, incorporated additional modeling refinements; however, it was judged that the changes were not that significant to warrant rerunning the CLTP FEM from plant data.

4.2.3.1 Uncertainties on the FEM The calculated stress values are compared against an allowable alternating stress limit of 13,600 psi, which is the lowest value for stainless steel (Reference 12).

To conservatively estimate the stresses resulting from the inputted loads, the calculated weld stresses were increased by an appropriate factor to account for weld uncertainty.

In addition, a Raleigh damping ratio of 1% was used. This is considered the minimum damping for a welded structure and it results in predicting higher fatigue stresses. For example, assuming only 1% damping results in alternating stresses that are approximately 1.4 times (f2k/Il) those that would be produced with a credible 2%

damping.

The uncertainty on the finite element frequency modeling of the steam dryer components is addressed by varying the measured frequency of the steam dryer loads by up to plus and minus 10% at 2.5% intervals. This also captures any ACM frequency uncertainty. This is done solely for EPU FEMs. It was not done for the CLTP baseline case since satisfactory operation at CLTP has been demonstrated by plant operation and inspections.

4.2.4 Computation Fluid Dynamic (CFD) Calculations PSEG pursued in mid-2005 an alternate approach to load definition. PSEG undertook an exploratory investigation with a vendor (not CDI) into the possibilities that CFD could be used to generate steam dryer loads. This effort was abandoned before the end of 2005. Towards late 2005, PSEG had discussions with its primary vendor, CDI, who is also knowledgeable on CFD calculations. CDI provided PSEG with the opinion that it would be difficult to validate a CFD analysis for this application. The major obstacles included requirements for small mesh sizes which would make the computational requirements prohibitive and the lack of benchmarking. Since late 2005, no further effort was expended on CFD calculations. The 2005 CFD information is not used in the HCGS steam dryer qualification program.

Page 15 of 20 Rev 1 Dated April 2007 LR-N06-0286 LCR H05-01, Rev. 1 5 Results 5.1 Baseline Conditions (100% CLTP)

The ACM loads for the baseline condition (100% CLTP power) were initially taken in 2006. Since the "C" and "D" strain gage channels failed, a conservative algorithm was used to bound the loads. This conservative load was inputted into the CLTP FEM (Reference 10). In early 2007 the strain gages were repaired and data was taken from all 4 MSLs. As discussed in Reference 16, a comparison shows that the original, 2006 data is conservative with respect to the later, 2007 data.

The predicted loads from the ACM using plant data at CLTP included a significant load spike at 80 Hz. CDI reviewed their model and determined that the ACM with the HCGS parameters would generate a significant 80 Hz load even when the only input to the HCGS ACM was white noise. PSEG and CDI compared the upper and lower strain gage data and determined that the 80 Hz signal is uncorrelated. PSEG confirmed with Quad Cities that they had similarly had an 80 Hz prediction that they confirmed was not present using plant measurements, including the steam dryer instrumentation. HCGS in February 2007 installed a pressure transducer on an instrument line off the top of the reactor vessel head. The data was decomposed to provide a magnitude versus frequency graph. It showed that there was an 80 Hz signal, but that it was relatively small in comparison with other frequency signals. Accordingly, the large 80 Hz spike load was eliminated.

Reference 10 is the CLTP FEM analysis using the conservative 2006 strain gage data.

The table below tabulates the limiting components (stress ratio less than 2.5), as reported in Table 6.3 of Reference 10. The FEM report includes figures that clarify the locations.

Components (all are welds) Peak Alternating stress ratio stress ratio Drain trough vertical plate to drain trough bottom 1.54 >2.5 plate Inner hood to steam outlet end plate 1.60 >2.5 Inner hood to center support (stiffener), junction at 1.79 2.26 bottom Outer hood to its end plates 2.33 >2.5 The stress ratio is obtained by dividing the allowable value by the adjusted calculated stress value. The adjustments to the calculated stress value include a 1.1 factor to account for differences in the Young's modulus of elasticity, and for welds, an appropriate multiplier to account for stress intensification in the weld.

Page 16 of 20

Attachment 7 Rev 1 Dated April 2007 LR-N06-0286 LCR H05-01, Rev. 1 5.1.1 Consideration of Uncertainties - Baseline Conditions The peak stress ratio includes dead weight. Since dead weight has minimal uncertainty and is not impacted by changes in power, the margins on peak stress ratios are less of a concern for a power uprate. The alternating stress ratios are more impacted by the load uncertainties and power changes.

In order to calculate the alternating stress ratios assuming the loads are increased by the strain gage and ACM uncertainty of +27.2%, the reported alternating stress ratio values should be divided by (1 + .272). This omits the conservatism in the 2006 load demonstrated by Reference 16.

Note that the dryer has operated over 20 years, including approximately 3 years at CLTP, with no indication of any fatigue induced cracks.

5.2 Interim Analyses for EPU Conditions The term "interim" is used to designate the evaluations done to predict the steam dryer stresses based on the EPU SMT results. This differentiates this analysis from the "final" EPU analysis, which will be based on the measured plant loads following power ascension.

The loads from the ACM using SMT transducer data on the MS lines also predicted a significant load at 80 Hz. To supplement the earlier discussion on why this is not a valid plant load, the test data for the single pressure transducer on each SMT outer hood, near the MS nozzle connection, was reviewed. It confirmed that there was no 80 Hz load in the SMT steam dome.

The tables below tabulate the weld locations with peak or alternating stress ratios of 2.0 or less. All non-weld locations maintained stress ratios greater than 2.0. The stress ratios were calculated for the nominal frequency case and for eight frequency shifts,

-10%, -7.5%, -5%, -2.5%, +2.5%, +5%, +7.5% and +10%. As with the discussion for the CLTP FEM, the reported stress ratios include a 1.1 factor to account for differences in the Young's modulus of elasticity, and for welds, it includes an appropriate multiplier to account for stress intensification in the weld.

Table 1 Peak Stress Ratio at Predicted EPU Conditions Weld Ratio at nominal Ratio at limiting Limiting frequency frequency frequency Inner hood to steam 1.53 1.53 0%

outlet end plate Skirt to upper support ring 1.54 1.46 +2.5%

Outer hood to cover plate 1.96 1.77 +2.5%

Page 17 of 20

Attachment 7 Rev 1 Dated April 2007 LR-N06-0286 LCR H05-01, Rev. 1 Table 2 Alternating Stress Ratio at Predicted EPU Conditions Weld Ratio at nominal Ratio at limiting Limiting frequency frequency frequency Middle hood / >2.0 > 1.90 -10% to reinforcement strip +7.5%

1.33 +10%

Drain channel to skirt (at 1.96 1.62 +2.5%

bottom)

Outer hood to cover plate 2.00 1.80 +2.5%

Middle hood to steam >2.0 1.76 -10%

outlet end plate Inner hood to steam >2.0 1.86 -10%

outlet end plate The nominal frequency and limiting frequency values are extracted from Tables 7a and 7b, respectively of Reference 14. The locations are shown in Figures 19 and 20 of Reference 14.

The alternating stress ratios at the nominal frequency case have significant margin at 115% CLTP. The lowest value is 1.96. When considering the frequency variations, all locations maintain a stress ratio > 1.62 with one exception. The one exception is a length of the middle hood to end plate reinforcement strip. This location maintains a stress ratio > 1.90 between -10% and +7.5%, but the stress ratio drops to 1.33 at +10%.

5.2.1 Consideration of Uncertainties - Interim EPU Analysis Loadina Uncertainty at nominal freauencv The load magnitude uncertainty is discussed in section 4.2.1.1. Since only the uncertainty that increases the load is considered, this results in increasing the calculated stress by a factor of 1.27 (27.2% increase).

Freauencv Uncertainties The impact due to frequency uncertainty was determined by changing the frequency of the loads.

The stresses for the two limiting alternating stress locations at the nine frequency cases are provided on Table 7 and Figure 27 of Reference 14. They show that the stress at the middle hood to reinforcement strip increases from 2200 psi at nominal frequency to 5147 psi at the +10% frequency shift. This is a factor of 2.34.(134% increase).

Similarly, for the drain channel, the increase is from 3496 psi at nominal frequency to 4246 psi at +2.5% frequency shift. This is a factor of 1.21 (21 % increase).

Page 18 of 20 Rev 1 Dated April 2007 LR-N06-0286 LCR H05-01, Rev. 1 Frequency and Loadinq Uncertainties When considering the uncertainty due to frequency and load, the square root of the sum of the square (SRSS) is used to combine the two uncertainties since the uncertainties are random. The limiting two locations are evaluated.

For the middle hood to reinforcement strip weld, the increase in stress due to the load uncertainty at the nominal frequency is 27.2%. When it is combined with frequency uncertainty, 134%, by SRSS, the combined uncertainty is 137%, only slightly greater than the frequency uncertainty. The stress ratio at the +10% frequency shift when combined with the load uncertainty is 1.31.

For the drain channel, the SRSS of 27.2% and 21% is 34%. The stress ratio at the

+2.5% frequency shift when combined with the load uncertainty is 1.46.

6 References

1. General Electric Service Information Letter 644, "BWR Steam Dryer Integrity,"

August 2002, Revision 0.

2. General Electric Service Information Letter 644, Revision 2 "BWR Steam Dryer Integrity," August 30, 2006
3. BWROG NEDO-33159 "EPU Lessons Learned and Recommendations,"

November 2004

4. EPRI Technical Report (TR) 1011463, "BWR Vessel and Internals Project, Steam Dryer Inspection and Flaw Evaluation Guidelines (BWRVIP-139)"
5. ASME Journal of Pressure Vessel Technology "Flow-induced Vibration in Safety Relief Valves," August 1986, R.M. Baldwin and H.R. Simmons
6. CDI Report 05-28 "Bounding Method to Predict Full Scale Steam Dryer Loads from In-Plant Measurements (CDI Proprietary)" Revision 1, dated May 2006.

VTD 430117

7. SIA Calculation HC-28Q-301, "Hope Creek Strain Gage Uncertainty Evaluation and Pressure Conversion Factors," dated May 30, 2006. VTD 430118
8. Exelon Report AM-2005-012, Revision 1, "An Assessment of the Uncertainty in the Application of the Modified 930MWe Acoustic Circuit Model Predictions for the Replacement QC Units 1 and 2 Steam Dryers," dated 11/28/05.
9. CDI Technical Note 06-16 Revision 0, "Hope Creek Unit 1 Uncertainty Analysis,"

undated. VTD 430119

10. CDI Report No. 06-24 "Stress Analysis of the Hope Creek Unit 1 Steam Dryer for CLTP," Revision 3, September 2006. VTD 430120
11. CDI Report 06-17 "Hydrodynamic Loads on Hope Creek Unit 1 Steam Dryer to 200 Hz," Revision 3, April 2007. VTD 430121.

Page 19 of 20 Rev I Dated April 2007 LR-N06-0286 LCR H05-01, Rev. 1

12. ASME Section III, Division 1 Appendixes, Appendix I, Table 1-9.2.2, 1986 Edition, Curve "C" for 1 El1 cycles.

13.CDI Report 06-16 "Estimating High Frequency Flow Induced Vibration in the Main Steam Lines at Hope Creek Unit 1; A Subscale Four Line Investigation of Standpipe Behavior" Revision 2, dated April 2007. VTD 430113.

14. CDI Report No. 06-27 "Stress Analysis of the Hope Creek Unit 1 Steam Dryer Using 1 /8th Scale Model Pressure Measurement Data," Revision 2, dated February 2007.
15. CDI Technical Memo 06-23P (Proprietary), "Comparison of The Hope Creek and Quad Cities Steam Dryer Loads at EPU Conditions," Revision 1, dated January 2007.
16. CDI Report 07-01, Revised Hydrodynamic Loads on Hope Creek Unit I Steam Dryer to 200 Hz Page 20 of 20
Retrieved from "https://kanterella.com/w/index.php?title=ML071290563&oldid=2577202"