ML20067E400
| ML20067E400 | |
| Person / Time | |
|---|---|
| Site: | Beaver Valley |
| Issue date: | 01/17/1991 |
| From: | Bain R, Deconto R, Testa M DUQUESNE LIGHT CO., STONE & WEBSTER ENGINEERING CORP. |
| To: | |
| Shared Package | |
| ML20067E390 | List: |
| References | |
| NUDOCS 9102140254 | |
| Download: ML20067E400 (161) | |
Text
s EVALUATION OT MAIN FEEDWATER PIPING MISALIGNMENT AND STEAM GENERATOR NOZZLE CRACKING BEAVER VALLEY POWER STATION UNIT 1 Principal Investigators M.F. Testa (DLC)
R.E. DeConto (SWEC)
R.A.
Bain (SWEC)
DU20ESNE LIGilT COMPANY NUCLEAR GROUP CORPORATE NUCLEAR SERVICES UNIT NUCLEAR ENGINEERING DEPARTMENT Final Report January 17, 1991 9102140254 910129 PDR ADOCK 05000334 P
4 4
1 TABLE OF CONTENT 1 SECTION TITLE EXECUTIVE
SUMMARY
l l
ACKNOWLEDGEMENTS INTRODUCTION 1.0 PURPOSE / SCOPE 2.0 DESCR1PTION/ OPERATION OF TEED;'ATER SYSTEM 2.1 MAIN FEEDWATER SYSTEM 2.1.1 MAIN FEEDWATER REGULATING VALVES 2.1.2 BYPASS REGULATING VALVES 2.1.3 FEEDWATER ISOLATION 2.2 AUXILIARY FEEDWATER SYSTEM 2.2.1 EVENTS TO INITIATE AUXILIARY FEEDWATER FLOW-2.2.2 DESIGN CONSIDERATIONS FOR AUXILIARY FEEDWATER FLOW 2.3 FEEDWATER PIPING LAYOUT / SUPPORTS 2.4 FEEDWATER FEEDRING DESIGN 2.5 PRESENT DESIGN METHODOLOGY
3.0 DESCRIPTION
OF 7R OBSERVED ANOMALIES E
I 3.1
'A' FEEDWATER LINE 3.2
'B' FEEDWATER LINE 3.3
'C' FEEDWATER LINE 4.0 ACTION PLAN (STEPS TAKEN PRIOR TO RESTART FROM 7Rl
4.1 INTRODUCTION
4.2 COMPILATION OF 7R INSPECTIONS 4.2.1
'A' FEEDWATER LINE 4.2.2
'B' FEEDWATER LINE i
i
o a
t i
SECTION TITLE i
1 4.2.3
'C' FEEDWATER LINE 4.3 EVALUATION / ANALYSIS OF AS-FOUND CONDITION 4.4 REPAIR / MODIFICATIONS i
4.5 IMPLEMENTATION OF INSTRUMENTATION PROGRAM 4.6 STATUS / DISCUSSION 5.0 ROOT CAUSE CONSIDERATIONS (INCORPORATING -PAST INDUSTRY EXPERIENCE) 5.1 LOCAL STRATIFICATION 5.1.1 NRC IE BULLETIN 79-13 5.2 GLOBAL STRATIFICATION 5.2.1 NRC AEOD-STUDY 5.3 WATER HAMMER i
5.3.1 PRESENT BV-1 DESIGN CONSIDERATIONS 5.3.2 FLOW CONTROL VALVE INSTABILITY 5.3.2.1 PAST BV-2 WATER HAMMER EVENTS 5.3.2.2 REVIEW OF NUREGS 5.3.3 STEAM GENERATOR WATER HAMMER 5.3.3.1 CREARE STUDY 5.3.3.2 TROJAN REPORT-
-6.0 RESULTS OF ANALYSIS (POST 7R) 6.1 CORRELATION OF ANALYTICAL MODEL TO PLANT i
DATA 6.1.1 NORMAL THERMAL MOVEMENT 6.1.2 INCORPORATION OF GLOBAL STRATIFICATION 6.1.2.1 AUXILIARY-FEEDWATER INJECTION ~AFTER 29% POWER A
REACTOR TRIP 6.1.2.2 INTRASYSTEM STRATIFICATION 11
i i
SECTIQH TITLE 6.2 NEW GLOBAL STRATIFICATION LOAD CASES (NOT PREVIOUSLY CONSIDERED) 6.3 PRESENT CONFIGURATION-GLOBAL STRATIFICATION AND CONSTRAINED MONOBALL SUPPORTS EXTRAPOLATED TO DESIGN BASIS TEMPERATURES 6.4 WATER HAMMER DISC *JSSION 6.4.1 STEAM GENERATOR WAT 2 HAMMER 6.4.1.1 LAWRENCE LIVERMORE NATIONAL LABORATORY REPORT STEAM GENERATOR WATER HAMMER 6.4.1.2 STATIC TIME NISTORY ANALYSIS 6.4.2 FEEDWATER REGULATING VALVE FLUID TRANSIENT EVENTS 6.5 LOCAL THERMAL STRATIFICATION 6.5.1 CORRELATION TO 7R PLANT DATA 6.5.1.1 CYCLE COUNTING 6.5.1.2 CORRELATION TO OPERATION WHEN SIGNIFICANT CYCLING OCCURRED 6.5.2 Ti?RMAL STRATIFICATION EVENTS 6.5,2.1 REVIEW OF SYSTEM STANDARD DESIGN CRITERIA 1.3 REV.2 6.5.2.2 ACTUAL NUMBER OF EVENTS SINCE 1979 6.5.3 ESTIMATE OF CYCLES (AND CALENDAR l
TIME) TO INDUCE CRACKING 6.5.4 SIGNIFICANCE OF LOCAL THERMAL STRATIFICATION VERSUS PIPE GLOBAL BENDING STRESSES TO INDUCE l
CRACKING l
l l
6.6 INSTALLATION O
111
l J
-SECTION-TITLE 6.6.1 DISPLACEMENT MODEL 6.G.2 CORRELATION WITH AS-FOUND DISPLACEMENT POSITION
7.0 CONCLUSION
S 7.1 PIPE IRREGULAR ALIGNMENT l
7.1.1 ROOT CAUSE FACTORS 7.1.1.1 UNCORRECTED ALIGNMENT FROM THE 1981 WATER HAMMER EVENT 7.1.1.2 PIPING MISALIGNMENT DUE TO THE 1979 AND 1988 ELBOW REPLACEMENT 7.1.2 NONFACTORS 7.2 STEAM GENERATOR N0ZZLE TAPER TRANSITION CRACKING 7.2.1 ROOT CAUSE FACTORS 7.2.2 NONFACTORS 7.3 GLOBAL STRATIFICATION 7.4 POTENTIAL FOR FUTURE WATER HAMMER 7.4.1 SIMILARITY TO TROJAN NUCLEAR POWER STATION 7.4.2 THERMAL SLEEVE EROSION 7.4.3 J-TUBE EROSION 4
1 7.' 4. 4 - AUXIL1ARY FEEDWATER IMPLICATIONS-
^8.0 RECOMMENDED CORRECTIVE ACTIONS 8.1 MANDATORY CORRECTIVE ACTIONS 8.1.1 IRREGULAR DEFORMATION PATTERN iv
4 i
l 1
SECTION TITLE o
B.1.2 GLOBAL STRATIFICATION AND POTENTIALLY LOCKED MONOBALL SUPPORTS 8.1.2.1 CONFIRM ANALYSIS ASSUMPTION THROUGH INSTRUMENTATION 8.1.2.2 REVISE PIPE RUPTURE CRITERIA (MEB 3-1) 8.1.2.3 MEASURE EXISTING ELBOW WALL THICKNESS 8.1.2.4 PIPE SUPPORT MODIFICATIONS 8.1.2.5 PIPE SUPPORT INSPECTIONS 8.1.2.6 INTERIM EVALUATION 8.1.2.7 REVISE PIPE SUPPORT CRITERIA 8.2 STRONGLY RECOMMEMDED CORRECTIVE ACTIONS 8.2.1 STEAM GENERATOR N0ZZLE CRACKING 8.2.1.1 MODIFICATION 8.2.1.2 PRE / POST INSTALLATION INSPECTIONS 8.2.2 FUTURE POSSIBLE WATER llAMMER EVENTS 8.2.2.1 FEEDRING/ THERMAL SLEEVE INSPECTION 8.2.2.2 J-TUBE EROSION INSPECTION 8.2.3 TACK WELD SHIM STACKS 8.3 RECOMMENDED CORRECTIVE ACTIONS 8.3.1 FUTURE POSSIBLE WATER HAMMER 8.3.1.1 PERIOD REVIEW OF FRV AND TURBINE BUILDING PIPING 8.3.1.2 LIMIT MAXIMUM AUXILIARY FEEDWATER FLOW 8.3.2 INSTALL MIN-K INSULATION AT ALL RUPTURE RESTRAINTS 8.3.3 OTHER CONCERNS 8.3.3.1 REVIEW GLOBAL STRATIFICATION EFFECTS IN THE MSVH AND TURBINE BUILDING 8.3.3.2 PREFORM SAMPLE NDE'S TO VERITY WALL THICKNESS B.3.3.3 NDE APW NOZZLE IN MSVH v
~.
. ~. -. -,. _.. - -
-~.. -._.. - _ -. ~. ~. - -. -...
e SECTION TITLE i
?
8.3.3.4 REPLACE REMAINING LOOP SEAL i
GRADE B ELBOWS WITH GRADE C MATERIAL l
B.3.3.5 INSPECT OTHER MONOBALL SUPPORTS
9.0 REFERENCES
10.0 APPENDICES 10.1 CHRONOLOGY OF EVl'NTS - FOST 7R HEATUP (APPENDIX A) 10.2 GRAPHS OF OCTOBER 6, 1990 MANUAL
- COOLDOWN (APPENDIX B) 10.3 PLANT EVENT DESCRIPTIONS (APPENDIX:C) 10.4 JUSTIFICATION FOR CONTINUED OPERATION (APPENDIX D) 10.5 CHRONOLOGY OF EVENTS WITH TEMPERATURE TIME ZONES (APPENDIX-E)
' 10.6; CHRONOLOGY OF EVENTS WITH STRATIFICATION RESULTS (APPENDIX F) 10.7 ASME SECTION III' APPENDIX F (APPENDIX G) 10.8
SUMMARY
OF SUPPORT LOADINGS AND STRESSES
= EVALUATED BY ALTERNATE CRITERIA (APPENDIX H)
D 9
4 vi' e,.
,v&wr e
,,w.n.e,,--w.,n
-,,wwm, I
w..
w--
-w~,-.
-~
.,,r.,-w,,
~r~-s,
e, LIST OF FIGURES FIGURE NUMBER TITLE 1
2.3-1 FEEDWATER LOOP A PIPE SUPPORT ARRANGEMENT 2.3-2 FEEDWATER LOOP B PIPE SUPPORT ARRANGEMENT 2.3-3 FEEDWATER LOOP C PIPE SUPPORT ARRANGEMENT 2.3-4
.FEEDWATER LOOP A, B AND C PIPE RUPTURE RESTRAINT ARRANGEMENT
]
3.1-1 7R AS-FOUND PIPING CONDITIONS 4.5-1 LOOP A POST 7R INSTRUMENTATION PLAN 6.1-1 STRUDL-SW LOOP A MODEL (2 SHEETS) 6.1-2 STRUDL-SW LOOP B MODEL (2 SHEETS) 6.1-3 STRUDL-SW LOOP C MODEL (2 SHEETS) 6.1-4 NON-STRATIFIED PIPE PROr'ILE 441' F (2 SHEETS) 6.1-5 NON-STRATIFIED PIPE PROFILE 32' F (2 SHEETS) 6.1.2.2-1 INTRASYSTEM STRATIFICATION GRAPH (RECORDED PLANT DATA) l 6.2-1 AFW INJECTION STRATIFICATION GRAPH (RECORDED PLANT DATA) 6.2.2 STRATIFIED PIPE PROFILE (2 SHEETS) 6.3-1~
CIRCUMFERENTIAL STRESS DISTRIBUTION DUE TO TOP TO BOTTOM TEMPERATURE GROWTH i
6.4.1.2-1 LOCATION OF STATIC WATER HAMMER INPUT LOADS LOOP A' 6.4.1.2-2.
LOCATION OF STATIC WATER HAMMER INPUT LOADS LOOP C 6.4.1.2-3 HORIZONTAL DISPLACEMENT RESULTS UNRESTRAINED MONOBALLS LOOP A 6.4.1.2 VERTICAL DISPLACEMENT RESULTS UNRESTRAINED MONOBALLS LOOP A 6.4.1.2-5 HORIZONTAL DISPLACEMENT RESULTS RESTRAINED MONOBALLS LOOP A 6.4.1.2-6 VERTICAL DISPLACEMENT RESULTS RESTRAINED-MONOBALLS LOOP A 6.4.1.2-7 HORIZONTALE DISPLACEMENT RESULTS UNRESTRAINED MONOBALLS LOOP C
+
6.4.1.2-8 VERTICAL DISPLACEMENT RESULTS UNRESTRAINED MONOBALLS LOOP C vii
. -.. ~ -
-. _. ~.... -.. - -.. -. _ - _ __, _
- f i
FIGURE NUMBFB TITLE 6.5-1
' LOCAL STRATIFICATION PROFILE AT THE LOOP A t
STEAM GENERATOR N0ZZLE 6.5.3-1 UP TRANSIENT-STEP FUNCTION 6.5.3-2 DOWN TRANSIENT-STEP FUNCTION i
6.6.1-1 PREDICTED HORIZONTAL INSTALLATION DISPLACEMENT
. PROFILE LOOP A
.i 6.6.1-2 PREDICTED HORIZONTAL INSTALLATION DISPLACEMENT PROFILE MONOBALLS LOCKED LOOP A 6.6.1-3 PREDICTED HORIZONTAL INSTALLATION DISPLACEMENT PROFILE LOOP C' i
8.1.2.1-1 PROPOSED INSTRUMENT PLAN LOOP A 8.1.2.1-2
' PROPOSED INSTRUMENT PLAN LOOP C 4
f
- b e
9 e
viii
_ __._..._.~,-.._ _.. _ _.. _..
.....~..-.m..
l 4
LIST or TJ4LLS TABLE NLHBER TITLE 2.5-1 TEEDWATER LINE POSTULATED RUPTURE POINTS 6.1.1-1 STRUDL-SW VS PLANT DISPLACEMENT DATA (LOOP A)- REACTOR POWER LEVEL AT 29% AND TEEDWATER TEMPERATURE OT 335' T 6.1.2.1-1 STRUDL-SW VS PLANT DISPLACEMENT DATA (LOOP A)
AUXILIARY TEEDWATER INJECTION 6.3-1 PIPE STRESS ANALYSIS RESULTS LOOP A 6.3-2 PIPE STRESS ANALYSIS RESULTS LOOP B 6.3-3 PIPE STRESS ANALYSIS RESULTS i.0')P C 6.3-4 PIPE SUPPORT RESULTS LOOP A 6.3-5 PIPE SUPPORT RESULTS LOOP B 6.3-6 PIPE SUPPORT RESULTS LOOP C 6.5.1.1-1
SUMMARY
OF T/C RANGE CYCLES FOR T/Cs 1-6 6.5.1.2-1
SUMMARY
OF CYCLES BY INTENSITY,
- MODE, AND PLANT EVENT TOR T/C No.1 (3 SHEETS) 6.5.2.1-1
SUMMARY
OF SSDC 1.3 REVISION 2 TRANSIENTS (REACTOR TRIP AND HEATUP ONLY) e Lx
o EXECUTIVE
SUMMARY
During the Beaver Valley Power Station-Unit 1 (BV-1) seventh refueling outage (7R), an ultrasonic examination of the first j
elbow outbound in the feedwater piping to steam generator RC-E-1A revealed an indication.
Since this elbow had experienced fatigue cracking in the past a decision was made to replace the elbow.
After replacement of the elbow, upstream piping alignment was reviewed and it was found that line 16"-WFPD )
601-Q2 was bound against pipe rupture restraint FWR-38.
The purpose of this evaluation report is to identify and investigate the root cause-of the feedwater system piping misalignment and steam generator nozzle cracking and to provide recommendations which will prevent the reoccurrence of these phenomena.
A brief description of the feedwater system, a summary of the 1
7R anomalies and action plan implemented prior to the unit's restart is provided.
Root cause considerations incorporating past industry experience included local--and global thermal stratification effects, water hammer events, and installation practices.- An instrumentation program was implemented to measure and quantify the mechanism that resulted in the damage.
The
. instrumentation installed to record data included lanyard potentiometers, strain. gages', thermocouples, accelerometers, i
linear velocity _ displacement trensducers and pressure transducers.
These data were thoroughly evaluated and correlated to analytical'models.
]
During the root cause. evaluation it was determined that a phenomenon occurs under tertain plant conditions that was not included in the original plant design basis.
Global thermal stratification, not considered in the design.
basis qualification _of the~feedwater piping. system, was identified and quantified by data recorded during the plant restart following 7R.
Furthermore, correlation of the recorded data to analytical results-indicated that=two monoball restraints are potentially locked up and are not functioning as required.
Therefore the effects of global stratification and locked monoballs were considered' in the reanalysis' of the piping-
~,
system. _
Based on the' identification of the global stratification' existence, combined with potentially locked' monoball' supports, a JCO has been prepared.
The. results of the ' analysis show that the piping system-stresses are within code allowables.
Nozzle and penetration loadings meet design basis limits.
All supports are
= satisfactory except for the two nonoball supports under the assumption c'
bound monoballs.
Alternate criteria were x
i
.-..-..---- m.
._,--,,.._.m.
~.-
g i=
developed and demonstrated the acceptability of pipe supports and pipe break location selection until recommended activities can be completed during the next refueling outage.
The review of recorded data showed local stratification was occurring in the vicinity of the steam generator nozzle as anticipated..
It also revealed that the number of thermal stratification events was greater than the number of events experienced by the plant. A plant event is defined as either a plant heatup or a reactor trip.
It was shown that alternating stress levels were as high as 180 ksi and that crack ~ initiation could begin after only sixteen plant events.
'Thus it was concluded that local stratification in the steam generhtor' nozzle region is 'directly responsible for the cracking.-
Studies to determine pipe misalignment included evaluation of design; basis water hammer loads, time history loads generated-from both feedwater regulating control valve instability, and steam generator water hammer (SGWH) loads.
Design basis time history loads were determined not to be of sufficient I
magnitude to-cause the observed misalignment.
Although instability and SGWH can provide the required magnitude of loads (100-200 kips) and are addressed in this report, the
-piping misalignment'was not accompanied by support damage.
Therefore, it is ' concluded that the misalignment occurred during line positioning.
corrective actions were determined and categorized as mandatory, strongly recommended and recommended.
Mandatory corrective actions associated with the piping misalignment, q
steam generator nozzle cracking and unanalyzed conditions were identified.
There are no mandatory corrective ' actions required for future possible water hammer concerns.
The mandatory corrective action -for the irregular pipe i
deformation pattern is to realign the piping to its as-designed position.--This was-accomplished during the unit's 7th refueling outage.
The recommended corrective actions for the steam generator nozzle-taper transition, cracking aret (1) a modification t_o install thermal -sleeves to prevent reoccurrence of the j
cracking, and (2) pre / post installation inspection of the i
feedwater line.
The mandatory corrective actions for the global stratification
. phenomenon are to:
1)
Confirm analysis assumptions through an enhanced instrumentation program.
l xi l
I k
i 1
2)
Revise pipe rupture criteria to bc I
consistent ~
with Mechanical Engineering i
Branch. Technical Position 3-1 (MEB 3-1).
3)-Measure the existing elbow wall thickness.
- 4) Modify the monoball supports R-3, R-4 and R-11 which do not meet. plant design basis criteria.
5)
Inspect supports to verify input assumptions.
6)' Implement OBE seismic limits as determined during the-interim evaluation.
1
- 7).
Revise design basis-documentation to include ASME Boiler and Pressure Vessel Code L
Case.N-318.
Strongly. recommended' corrective. actions, primarily considered with future possible water hammer events, and recommended
- corrective actions-addressing other
- concerns, are also presented in this report.
r P
i e
T T
b 9
9 k
xii P
c-,v--c,-4-i r,w,- w M ~-.,
e r-r=44-ev--wmW+ -v
---v--~ee-e -ei-mvw-w4,e'w--~wm-w-w
<,r 4 4-w - r e
<--e's---
v-ew'--
7-t i
9
,r a
r + 4 t*
w + 's t-
-e
i e
ACYNOWLEDGMENTS It is recognized that others contributed significantly in the o
generation of this report.
Steve Collins developed the method of analysis and performed the qualification f or ef fects due to global thermal stratification.
Carol Allen performed the correlation studi demonstrating the consistency between ecmputer code methods to determine global stratification ef f ects. Bob Kimball reduced and interpreted recorded data and evaluated the effects of local stratification.
Yo Cho developed an instrumentation program to record, interpret and reduce feedwater system operating data.
O O
xiii
=.
4 INTRODUCTION During the Beaver Valley Power Station-Unit 1 (BV-1) seventh refueling outage (7R), an ultrasonic examination of the first elbow outbound in the f eedwater piping to steam generator RC-E-1A revealed an indication. Since this elbow had experienced fatigue cracking in the past a decision was made to replace the elbow.
After replacement of the elbow, upstream piping alignment was reviewed and it was found that line 16"-WTPD 601-Q2 was bound against pipe rupture restraint TWR-38.
Further review of rupture restraint TWR-38 revealed that shim pack clips were damaged and/or missing.
An extensive evaluation of the ' A' f eedwater line and supports followed.
Inspections were also initiated on the remaining feedwater lines
'B' and
'C'.
Damage similar to the
'A' line was found to exist on the
'C' line.
Inspection of the
'B' line revealed no damage.
A description of all observed anomalies and their respective dispositions are presented in Section 3.0 of this report.
Subsequently, it was determined that the feedwater lines for loops
'A' and
'C' had to be cut in order to reposition the lines properly within their respective rupture restraints.
A review of the rupture restraints' orientation and installation precluded movement of the restraints to accommodate the piping.
It was decided to restore both the ' A' and
'C' feedwater lines to their original configuration and correct all anomalies prior to restart of the unit.
In addition an extensive instrumentation program was implemented in order to gain an insight into the possible causes of the piping misalignments.
This report includos an in-depth review of the instrumented data taken, of records of past feedwater incidents both at BV-1 and in the industry, and of available literature associated with feedwater system problems in the area of nozzle fatigue cracking, local and global thermal stratification, and water hammer.
Conclusions stated herein are derived both from this review and from analyses conducted specifically for this effort.
xiv
1.0 PURPOSE / SCOPE The purpose - of this evaluation report is to identify and investigate the root cause of.the feedwater system piping misalignment and steam generator nozzle cracking discovered during the Beaver Valley power Station-Unit 1
seventh refueling outage (7R) and provide recommendations to minimize the reoccurrence of these phenomena.
The scope of-this evaluation includes.a brief description of the feedwater
- system, a
summary of the 7R. observed anomalies and a
description of the action plan implemented prior to the units restart. During the investigation of the root cause(s) of the piping-misalignment and nozzio cracking other potential feedwater system concerns which are not directly related to the root, cause of the misalignment or cracking, may be identified.
Recommendations will also be provided to correct any additional concerns beyond the piping. misalignment and piping cracking.
- )
l i
-i i.
1 l.
p
,.~~..._..__..,,__,__,,_c,.._.
2.0 DESCRIPTION
/OPERATIOd OF FEEDWATER SYSTEM The following fescription of the feedwater system pertains to the portion of the system from the main feedwater pumps to the steam generator. This is the primary area of interest for thic report.
2.1 MAIN FEEDWATER SYSTEM The main feedwater system is used to preheat and deliver condensate to three steam generators via a
common feedwater header. Two steam generator supply pumps, each rated at 15,200 gpm at 1900 feet total dynamic head (TDH) supply the common header.
Threo individual steam
_ generator feedwater lines are supplied by the feedwater header.
Each of these individual lines downstream of the feedwater header contains a flow measuring device, a main feedwater regulating valve controlled by a three-element steam generator water level control system, a bypass flow control valve controlled by a two-element level control
- system, flow control isolation valves and a motor operated containment isolation stop-check valve.
Techniques of providing feedwater flow to the steam generators are dependent upon plant condition.
For
- example, when the plant is either in hot standby operation or escalating in power up through low power operation, feedwater flow is being directed through the bypass feedwater regulating valve.
This feedwater flow could either be a continuous trickle flow-(approximately 200 gpm), intermittent plug flow (as in a feed and bleed mode),
or a
combination of the two.
Under these situations relatively cold water from the condenser hot well is introduced into the steam generator.
Once the main feedwater regulating valves are put in service, trickle flow or intermittent plug flow does not take place.
The specific components of interest are discussed below in more detail.
2.1.1 MAIN FEEDWATER REGULATING VALVES The valve is sized to control steam generator level from about 15% power to full power. Each valve is controlled automatically by a three-element steam generator level controller or manually from the control room bench board. The three-element control system continuously compares feedwater flow, steam flow and water level in order to maintain a 2
i l
l programmed water level in the secondary side of the
]
steam generator during normal operation.
The main feedwater regulating valves are Copes-Vulcan, 12 inch, 900 pound, cage and piston design with 16 inch inlets and outlets.
The operators are also supplied by Copes-Vulcan with their f ailure mode being spring to close.
A design change in 1982 installed hydraulic dampers on the actuators to provide better control over plug travel (Ref.
9.67).
2.1.2 BYPASS REGULATING VALVES A bypass flow control valve is located in parallel with the main feedwater regulating valve.
The bypass flow control valve is used for steam generator level control at power levels up to about 151 power.
The bypass flow control valves are left partially open at high power levels to provide cooling water flow to the feedwater heaters during a
The valve is controlled
-automatically by the two-element feedwater control system which senses steam generator level and reactor power level.
A controller continuously compares the two signals to regulate the bypass control valve and hence control steam generator secondary side level.
Manual local control of the valve can also be taken.
The bypass valves are Masonellan 4 inch, 900 pound rated valves.
2.1.3 FEEDWATER ISOLATION The-main feedwater regulating valves trip on the following:
a.
.Two out of three (2/3) high-high (Hi-H1) level signals in any steam generator, b.
Safety injection signal.
c.
Low average -temperature in 2/3 loops after a reactor trip.
The bypass flow control valves, the main turbine, the'. main feedwater-pumps, and main feedwater containment isolation valves trip on the following:
~
L a.
(2/3) Hi-Hi level signals in any steam-generator.
b.
Safety injection signal.
3 i
l -.
2.2 AUXILIARY FEEDWATER SYSTEM The auxiliary feedwater system is used to supply water to the steam generators in order to remove residual heat and cooldown the reactor coolant system when the main feedwater pumps are not available (i.e.
feedwater isolation). The auxiliary feedwater pumps transfer water to the steam generators from the 152,000 gallon primary grade demineralized water storage tank (WTaTK-10).
Auxiliary feedwater is supplied to each steam generator through two redundant supply headers, each containing a motor operated throttle valve.
The supply headers join downstream of the throttle valves and flow through a flow measuring device among other components prior to connecting into the main feedwater line. This connection is downstream of-the main feedwater containment isolation valves but just up-stream of the containment penetration in the main steam valve hriuse- (MSVH).
~
The auxiliary pumps _ include two 350 gpm electric driven pumps which auto start in tandem and one 700 gpm steam driven pump as a backup to the electric pumps.
2.2.1-EVENTS TO INITIATE AUXILIARY FEEDWATER FLOW The transients that cause an automatic start of the two electric auxiliary feedwater. pumps are as.
follows:
a.
Lo-Lo steam generator level signal in any-two steam generators, b.
Both nain feedwater pumps trip.
c.
' Safety-injection-signal.
LFor any of the above conditions to exist-they would have to be proceeded by 'one of -the following events:
a.
Loss of neraal feedwater.
b.
Loss of offsite power-followed by a reactor trip (results in a -loss of normal feedwater),
c.
Secondary side pipe rupture.
d.
Cooldown following a steam generator tube rupture.
4
.w.
2.2.2 DESIGN CONSIDERATIONS FOR AUXILIARY FTDWATER FLOW When a transient conditior as cited in Section 2.2.1 is presented, it take*, on the order of one to two ninutes before the aux iliary flow transmitters sensa auxiliary feedwater flow, Presently, the initial auxiliary f ee-jwate.- flow rate which can bs experienced is on the crdet of 400 gpm per stean generator until control is taken through the auxiliary feedwater throttle valves.
The primary grade demineralised water storage tank which supplies the auxiliary feedwater system is housed in an enclosure which is maintained at or near ambient conditions.
Freeze protection is provided in the form of two electric space heaters.
T'arefore the tank is not maintained muen above treezing during subfreezing ambient temperatures.
2.3 FEEDWATER PIPING LAYOUT /SUPPORTF The portion of th Icedwater piping evaluated for misalignment and nozzle cracking is located inside the reactor containment structure from the containment penetration to the steam generator nozzle connection.
Refer to Figures 2.3-1, 2 and 3 for the pipe support locations on loops A, B and C respectively.
See Figure 2.3-4 for the pipe rupture restraint locations on all three loops.
The piping is A106 Grade B and fittings are 16 inch, A234 Grade WPB.
For all three piping rans the pipe enters the containment at approximately elevation 758 and rises vertically about twenty feet.
Then the pipes run horizontally outside the crane wall before penetrating the crane wall and running into the steam generators.
Just prier to the steam generator nozzle connection in each loop is a loop seal comprised of four back to back short radius 90 degree elbows, Monoball supports are the only supports on the feedwater lines that provide rigid c7nstraint-(vertical only) for all loading conditions.
These supports are comprised of a
ball and socket joint to allow free rotational displacements and self lubricating (Lubrite) plates to allow both free lateral and axial r.ovement.
5 i
1 L
m i.
FEEIMfATER LOOP A PIPE SUPPORT ARRAK'3fEleT
~
+a.
P
+=.
e
.r.
e.
n t,
N'S k E-V HmRAuuc Tyrt BRACE -/
",f
- w.',
g y/
A77Acuto ya co w g w gi,g, s
r
- ( q, 56 **
, var con 5T
'i N,
/
t
^
i' c et e
EPRrnG.
'* Anne a hyn
"' m l M MTm gyg.
I5~-MPO - 22
- 60s or
-[
&l9 j
r t.,
M-SA E-V
--i
/
ti eemcp tv tf%4TTACHtb HvotaAouc vv%
l CI3 To ctAnt w^u 4 1
/
VEM co*43r p
/
m
?
,/
e
/
125 Cl4 vitT. conST
. t 1,(*- ~
Tas -
__ N""[.
/
.w f
a i
att g
'b.
Ci5 i
H-54 E-W uioRhut:t T m p atRF ATTACd[D,,, Q ujf g t
/f 2", ;
e MD%,3C. APG Mougg Au.f Typs y
[s-~
~
~
i i CdDIUG SUPPOQT3.120UMTED OM 't tsBRtrF' rC l
r j
3
- 3,
- 2. M,5,otNot ts noR T** -SouTP
/
E-W OfNOTES EAST-w/EST
\\ ((
.# /l - - +s e t
- mum LJc 6 f YtrL BPA't at r,
Lt is C-if
(
i
-.m Af
-~4r a m%- Cwt STM.GEu Rc-E-ar i
4 FICURE 2.3-1 i.
1
FEEDWATER LOOP B PIPE SUPPORT ARRANCEMENT I
j'Y,u.
.ra Ny t4 NceTH - south 1 EAST-wtST WYdeAut<. MT. ERAct ATTscutb to cQANE to*'.L "2
16*- w&D 601 - C2 h
g,\\,[2+*
ces b
~
u
& see <c-
,\\
a7 u Air e a (y y o) f f
.~
4*N S}
een l
x
,1 s
I eos -
i
-pEtJE TG ATIOM " 77 y
I l
9 i
- A 7; FOGt COMTNu Af to*3
(
l
.e4
?
, SEE ELEVATION l
y
^Q eo noo
/
I c,
/_._
pel"* l6* QE D
~
c._...,,4 a
p.
n C6
/,i j N-s a t-w uven Autic
\\
('
/
TYPE B A?J,E At i
l 2( -
}-y tit va t ees 74&- fo
- ~
Al N 1t 7s si re* g 4 3T M. GE t1 D C -G -IB FIGURE 2.3-2
'~~'-
~
~ ~ ~ ~ ~ '~
' " ~ ~ ~ ~ ~ ~ ~ ~ ~
~~
FEEDWATER LOOP C FIPE ' SUPPORT ARRAIGCEMEllT-
- 'a*. '
7
.w.
- 's ea E-W 4 N 5 HYDRAUuc
' 1' TYPE ~ BRACE (TYP) f,5T,N,G, (yy,n A7yAcuto To (nANE Cl7 IS A. 'MONOB ALL' TYPE SLtDING '
- SUPPORT MOUNT ED _ON.*LUBRITE,*.
% watt f i 5*-wF?D-t* - &Cf-0 2 est :
' Cl7
-7 ese
/
N-5 t E-k/ HyoeAuuc TvPE BRACE
\\
f' psyttkAy,og
- 73 ATTACHED TO CRANE k/ALL -
- 4. VERricAL. coNSTRAtNT
\\
\\
To d,#
- its l
/
)
i
,f*
N
\\
%, /
~
=
~,
^*.s
- 3. I 6*
41 i
n,
~'
M'y ~x Nv lf' ceAcreo f
v[
"N x.nc.
M:'
' /
l N-S 4 E-W HY DR AULIC YYPE BRAtt AT it 7n' k
/
l-C. >
E CB
/
[
A,-
,t,,c c 4
- 4. STu. c s u.-
-cc-e - ec FICURE 2.3-3 2
e D
7
-4 a
4 4
4, 4
a~
r"--
gb 7h -
f.t.
7 5. _
.Og g s!-
!?
f.
A ii i
t
.~
' 9j :
s
-l l
-p.
l f
- jj o
=
iff i
1: -
-g
-,11 n,
.D, f,f
-[#+,.
r ri!!!
1 e,
-3
. Q.
ks
\\
g 5
g i
k. _..~
k 1
N f m _y _._4
~
77 7,
i
/.
E
~
1 0
IJ
/
'~f>.g
/
~"
, / --
/
'g 3
t l ' _" W j!
-g 3
3-M 2 $
LW 1 ;-
t 7,.
/. e
}CW
- f
. g
/
3 y
j x-2 Mr
!.- / f L
l(-
9 e
is.-
p :/
y N
j Dg pg[, j.
/
k
-W P
2 j
.8 e.
,.=
c+
j
.I g,:
.; f l 1
=s 2 g M
- s :
1
,f
- N L-
/
1 y...
\\ g%
II 3
_ hl.Q.
.7
- )
..i.
N P
g
- N., N _.
- ex_ '4x
-\\
v-
/.
it 3
I
~
'o t 7,
s.
- w s.
I
.4 9
Rupture restraints are used to absorb energy and restrain pipe movement following a postulated pipe rupture.
These restraints are nearly concentric to the pipe with shim stacks installed to provide the final gaps between the pipe and the restraint.
These gaps allow sufficient cicarance for free thermal and seismic movements.
Unexpected closure of a pipo rupture gap will result in an ununalyzed condition and potentially increased stress levels and support loadings.
Maximum allowable pipe gaps are calculated and design gaps are set equal to or less than this value to keep the energy imparted to the restraint within the design capacity of the restraint.
l 2.4 FEEDWATER FEEDRING DESIGN The Westinghouse Model 51 steam generator utilizes a 12 inch distribution feedring inside the steam generator which lays horizontally at the same elevation as - the incoming feedwater piping.
The feedring to shell nozzle is connected via a thermal sleeve.
This thermal sleeve is basically a slip fit into the back of the nozzle.
Specified gaps for this slip fit are on the order of.025 inches.
j The feedwater feedring employs J-tubes (2 inch schedule 80 A-106 Grade. B) to discharge feedwater from the feedring. The purpose of the J-tubes is to prevent steam voids from developing in the feedring.
The rapid collapse of a void in the feedring can result in a severe water hammer.
2.5 PRESENT DESIGN METHODOLOGY The feedwater piping from the containment penetration to the steam generator is designed to Category I class Q2 requirements.
The design incorporates normal thermal conditions and anticipated upset conditions such as seismic and time history events.
These time history loads are associated with postulated water hammer events such as main feedwater regulating valve closure due to feedwater isolation, main feedwater containment isolation check valve' slam and a main feedwater pump trip.
Feeduater temperature is a function of flow rate and reactor power level. The thermal analysis utilizes the design condition of 441' F at 100% power to envelop all intermediate feedwater temperatures.
The system, including the auxiliary feedwater nozzle and steam generator nozzle, is designed to withstand a 10 I
\\
1 l.
thermal shock incident for incoming ambient auxiliary feedwater flow while the system is at normal operating temperatures and pressures.
In order to minimize steam generator water hammer events, a loop seal adjacent to the steam generator nozzle is employed to minimize the horizontal length of piping prior to the feedring to reduce the potential for bubble formation and the resulting water hammer. The loop seals also prevent steam, should it enter the feedring,-from traveling back up the horizontal feedwater piping and causing more severe voiding.
J-tubes are installed on the feedring to prevent steam from entering the feedring.
Current postulated break locations are at the terminal ends and at points along the vertical risers.
The specific points are shown in Table 5.2-16 of UFSAR Section 5.2.6.3 and is reproduced here as Table 2.5-1.
Refer to Figures 2.3-1, 2 and 3 for the location of the point numbers specified.
W 4
11 l
l l
I
mum-__m2,__ _ _ _ _ _. _ _ _. _., _.
- w,
. s 1 TABLE 2.5-1 i
FEEDWATER LINE POSTUIJlIED RUPTURE POINTS
~
'i
~Feedwater Lines Location
'16-WFPD-22 16-WFPD-23 16-WFPD-24
?.
Terminal: Points 199 98-140 244 128
--18 8 '
l Point of Maximum 307 102
.144 Primary.+~ Secondary
' Stress Point Where None.
- None None.-
P +.S >. 8 (S, +' S }
h 2 0p 2 ).-
102 r a l-
-Point Where
- 202(2)
P>.8 Allowable =
101 f'I 144 f2I
.8(1.2).S-
'307 110' 143
- 1 f
a DJ Point Where None' None None S>.8'S a
Total Points-5f 5
4 j
1 Total Areas' 4(2)
.4(2) 3 (2)
Where:
P = Primary Stress I
S = Secondary. Stress Sh' S are defined'in ASME III NC3611 a
Note:
With.the exception of.Pt 180 oni16-WFPD-24 all points. listed above are I8i at' elbows.
Becauseiof the. proximity of two points, the area'between the two points-
'rzl is considered'one break area.
J!
'I m ~ ~ ~ - -.
-i l'
3.0-DESCRIPTION OF 7R OBSERVED ANOMALIES The following conditions were observed on the containment feedwater piping during the seventh refueling outage (7R). An extensive inspection / evaluation of all three feedwater lines was conducted once anomalies on the
'A' line became apparent.
3.1
'A' FEEDWATER LINE The specific anomalies found included:
a.
Pipe treak restraint, FWR-38, had the pipe blocked up against the restraint, shims dislodgod on two shim packs and shim retaining c.11ps bent, b.
Pipe break restraint, FWR-19, had the top wear plate shifted at one shim pack only, c.
Hydraulic
- snubber, HSS-201, had surface cracking along the edges of the embedded plate.
d.
Vertical monoball
- support, R-5, had loose bolts on the support frame, concrete spalling at crane wall,and evidence that free lateral and axial translation was not taking place as designed.
e.
Spring can, SH-1,.
Was topped out and not supporting any load, f.
-Hydraulic snubber, HSS-206, had a loose nut on the pipe clamp.
g.
Monoball supports, R-3 and R-4 had the pipe bound up against the crane wall.
h.
No other pipe support or pipe discrepancies other than those listed in a through g were observed.
The effect of the above observed anomalies on the qualification of the feedwater line included:
a.
Additional thermal stresses generated by restraining the line with the rupture restraint and monoball supports in the undesigned restrained direction.
13
b.
Effect on the dynamic analyses by.' changing the : modal. response due to support function changes. (i.'e.
rupture restraint gap closure and monoball lateral / axial direction restraint.
c.
Reduced flexibility (increase in earthquake anchor loads).
See Figure -3.1-1 for a summary of as-found ' 7R piping conditions.
3.2
'B' FEEDWATER LINE-
- Extensive _ inspections were also conducted on the ' ' B' feedwater line and only one minor anomaly was observed.
a.
FWR-14 had one slide plate move approximately 0.6 inches.
3.3
'C' FEEDWATER LINE a.-
Pipe break restraint, FWR-1, had the pipe blocked. up against the restraint, outboard shim brackets bent, and shims missing and ;
= dislodged. Four of 8 shim stack locations were.
affected.-
b.
Vertical ;monoball _ support,
- R11, had loose bolting on the angle iron.
c.
Pipe break restraint, FWR-13, had a wear plate loose, shims shifted down, and the-pipe hard against the inside of the restraint.
Contact.of the line with the; rupture: restraint had the effect of elevating thermal stresses-in the line since it-
.was. not free to = move in the vertical direction as designed.
This-also would have an impact on the-dynamic analyses since : individual mode shapes would be af fected.
In addition an outboard. shim pack was dislodged thereby increasing the rupture restraint gap in that directi'on-
.beyond design.
During a design basis terminal end break at. the steam generator, energy input to the rupture restraint may have. exceeded. the restraint's energy absorption = capacity.
This problem. no longer exists' because-the pipe' was relocated to its as-designed-location _during 7R.
f
\\
14
.e
,I
\\
l
-.,,.,n' -o<
7., " ' JM 6'*
l
(
1 1
},
c *j
- [r.f I
!1 e
s yr
.i j' g
- @J,
\\f \\
g ti
.9 e
it 9
m,s x,...s a t.
- - g f,.e.wi. -
,5 g.
t o f w r
! (F 4 a! !!
v uo
,a tt r
a
> :- ef
--y s
./
q..
k e'....
E P! hs py(--___I!_/,---O--( rG : c:
- ?
.c. 'k 3 M j ;} l 'd t T t rg - - - " " - " " ~ ~ [g m f { . / - s,. u '. c i '.i - / We
- ,,fe-
.e '} b ,f ' / 5 ./ /- g-i %:(- i }6. m ,4-4 _ w y e .I.C g ~ e- $i /,g-rg ss (%a g u i 8 'i Jr ~ . \\ -}.f V, e y x e 8 1 8 a ni,. y w ,.a. y u, e m l' 'I' 7 C 7-s. s a a / P i
- g th 1
/ ea e.: ,Y /- .i- ,1 'g i _.,:..- r_ ' \\. r rr 3~ w, 3 a ax y / t 7. s,, - _ s, c *t g s! ,1 ) } ,. )- A s -xgs- .....,. A s 3 i t :i t- ~ w. 5 t \\ i ; y ---- y 5 , 2 3, .g p S ? ~ _______k [3 '\\ i 7 -~~ ( p
- /\\ i N L ;f $
41 7.N,4 ., j i I!i! l/: it , :- t i u; . - e V.
- g 1
.,,rd t iC 3. s,, .g .k ' *. -*/- I_% (; g 7,
- ' ~
.%d.a sa ye g }>( g 't 1 j
- l..
9,~ M Ei. Q'4 f/, % I[ ! t*I5 ', \\ 'N li i7
- j'i 1 $
it s\\ (4s$4 .t ye t,,. ~, rn - + 3 s, .e oo s./ t - ) er m, J 44 i {j~ $ ** k ~
- 1. -
6' 15
4.0. ACTION PLAN = (STEPS TAKEN PRIOR TO RESTART FROM 7R) 4.1. INTRODUCTION This.section of the report delineates the course of action taken during the seventh refueling outage (7R) to document any anomalies that existed, and to evaluate these ' anomalies analytically _to ensure that no other piping or support damage may have resulted _ (e.g., thermal overstress). All the supports including all rupture restraints,_ spring hangers and snubbers, were inspected. If a reference is not nade to a specific support under Section 4.2 then no discrepancies were found. Any other-areas of concern that required inspections are also noted-in Section 4.2. In addition once this review was conducted some additional inspections resulted.- All of the observed pipe and support discrepancies were corrected prior to --startup, except the pipe up against the crane wall at R-3 and R-4. The feedwater lines were restored to their = original design position. .The instrumentation program was implemented to' determine
- the mechanism that led to the pipe misalignment.
The ' instrumentation was used to continuously.rucord during? the complete heatup of the unit from cold shutdown. This . instrumentation will be used periodically during the course of the next fuel cycle as warranted. -4.2 COMPIIJ. TION OF 7R INSPECTIONS .4.2.1 'A' FEEDWATER LINE a. Spring can, SH-1,- was topped out and not supporting:any_ load, b. Vertical monoball
- support, R-5, had loose bolts on the support frame, concrete spalling at crane wall,and evidence that' free-lateral-and axial translation-was not taking place as-designed.
c. Hydraulic
- snubber, HSS-201, had surface cracking along ~ the edges of the embedment plate.
d. Hydraulic snubber, HSS-206, had a loose nut on the pipe clamp. e. Pipe break restraint, FWR-38, had the pipe 16 ~
l blocked up against the restraint, shims were dislodged on two shim packs and shim retaining 4 clips bent, f. Pipe break restraint, FWR-19, had the top wear plate _ shifted at one shim pack only. g. Ultrasonic Testing (UT) examination of the complete SD pipe bend in the vicinity of.FWR-38 showed no indications._ An ovality check demonstra.ted ovality was_ acceptable. h. Surface examination of pipe wall at monoball-restraint R5 showed no. indications. i. Radiograph-and UT examination of the elbow'to steam generator nozzle weld after fitting replacement showed acceptability. j 4.2.2 'B' FEEDWATER LINE I a. Pipe break restraint, FWR-14, had one wear plate slide approximately 0.6 inches, u 4.2.3 'C' FEEDWATER LINE a a. Pipe -break restraint, FWR-1,- had the pipe - blocked up against the restraint-outboard shim brackets bent, and shims missing and dislodged. Four.of 8 shim stack locations were affected. I b. Vertical monoball-support,
- R11, had - loose
^ -bolting on the angle iron. c. Pipe break restraint, FWR-13,.had a wear plate. loose, shims shifted down,- and the pipe hard against the-inside of the restraint. d. UT examination conducted on the complete SD pipe _ An evality check demonstrated bend in area of FWR-1 showed no indications-the ovality to be acceptable. e. UT examination of the steam generator nozzle to elbow weld and counterbore area was performed and no indications were found. -4.3 EVALUATION / ANALYSIS OF AS-FOUND CONDITION-Prior to instrumenting the feedwater lines, analyses were conducted on both the 'A' and 'C' feedwater lines in an 17
i ..I attempt to understand the cause of the anomalies.and to determine if any other detr!.nental ef fects might have. -been imparted to the piping system. Various sample thermal analyses were run to qualify the-observed deficiencies. For example, the 'A' line translational displacements were restrained, individually - and concurrently, at pipe support and pipe rupture restraints suspected of restricting pipe movement to derive the worst scenario. l Under these thermal cases the worst case derived was restraining the vertical and lateral directions of the pipe at FWR-38 and concurrently restraining all translation displacements - at monoball R5. This case yielded an overstress at the first elbow of f of the steam generator nozzle. The stress level shown was-on the order of 27 ksi which was only slightly over the 25 ksi-allowable _(Ref. 9.11). In addition,-a-local pipe wall overstress was shown to have occurred at monoball R5. Due to these results additional -inservice inspection (ISI) on those ~ analytical code overstress regions was conducted as defined above.- The pipe against the crane wall at monoballs R3 and R4 was -judged to have an insignificant ef fect for the worst thermal case because the thermal deflection moves the pipe away from the crane wall. On the 'C' line,-analyses were conducted by. restraining the line-at FWR-1 since this is the only location.where the line;was observed-to be-constrained. A slight code-overstressi was shown at.the steam generator nozzle to elbow transition region (Ref. 9.11).. Additional ISI as-detailed.above_was then conducted. For the seismic scenario, the pipe was restrained as.in the thermal case run but included disabling snubber HSS-4 201 on the 'A' line because of the. observed surf ace cracking at its embedded plate. For bcth the ' A' and 'C' lines the dynamic analyses demonstrated acceptability to code requirements. - 4.4 REPAIR / MODIFICATIONS l The following is a compilation of all the repairs conducted during the seventh refueling outage prior to startup in order to return both 'A' and 'C' line stress . levels to within code acceptability. 18 l h_
a. Replaced first elbow t!f the steam generator nozzle in loop A due to observed indication under UT examination. Further destructive evaluation revealed that this may have been a machining mark or possibly the onset of crack initiation. b. Restored rupture restraints FWR-38 and FWR-1 to within design limits by providing adequate gaps to allow f ree normal thermal movement and dynamic clearances. Gaps were also set to account for energy input into the restraint. Min-K insulation was utilized at restraint locations to facilitate inspection of gap settings in the future. c. Reset spring hangers SH-1 and SH-10. d. Restored function of snubber HSS-201 by utilizing alternate embedded plate to maintain the same direction of constraint by the snubber. Replaced monoball R5 with a new support design e. to retain same support function. f. Verified acceptability of gap settings on all rupture restraints. g. Both 'A' and 'C' feedwater lines had to be cut and m4tored to allow correct repositioning through the;vupture rostraints. 4.5 IMPLEMENTATION'OF INSTR'd54ENTATION PROGRAM An instrumentation pr.ograe on the ' A' feedwater line was implemented prior to start 9p 'ct the unit to aid in determining root cause of-the' piping misalignments and the steam generator nozzac to olbow counterbore cracking. The-instrumentation layout is depicted in Figure 4.5-1 and' summarized below; .a. Full bridge strain gages to capture horizontal and vertical plane bending and torsion at the elbow counterbore location were installed. L 1 i 1, 19 l l
'-a,- ' LOOP A POST' 7R INSTRUMENTATION PLAN-Steam - 5 N L8 9,eg,; -L6 L7 % Monoboll o P . hts e / eg*fe,- L1 /'O L4 - L2 e. 4' es M _Therr.ocouples 8 - T1 thru ' T6 0 L3 Stroin Gages ~ S1 -thru S4 Thermocouples 6 T7 thru T12 - ACO2 NOMENCLATURE _ - T = TilERMOCUPLE S = STRAIN CACE A-* ACCELEROMETER L = LANYARD 6 - 3 q -b ~& r FICURE 4.5-1 n S I g g s-o g-w g-e p-g 3- ,w w w ~i,v- -r
b. Lanyards to measure both stacic a id dynamic displacements were installed at the following locations: 1. Three directional displacement s at the steam generator nozzle location (lanyards 4-6). 2. Three directional displacements at the pipe in the vicinity of rupture restraint FWR-38 (lanyards 1-3). 3. Lateral and axial lanyards at vertical constraint RS (lanyards 7-8). c. Accelerometers, one at each location where the three directional lanyards were applied. This provides local acceleration levels and modal information. d. Six thermocouples were installed circumferentially on the elbow counterbore region to measure local thermal cycling ef fects. Six thermocouples were also installed circumferentially on the piping several diameters upstream of the loop seals to determine if the elbow loop seals were effective in preventing back leakage from the feedring and to determine if global thermal stratification existed. These thermocouples were not sufficient to completely describe global stratification profiles, but were only used to determine if global stratification existed. Based upon prevailing industry
- beliefs, global.
stratification was not initially thought to exist in this pipe due to both the location of r l the auxiliary feedwater connection and the vertical runs of feedwater piping where mixing. l would have been expected to occur. i e. A pressure transducer was installed downstream of main feedwater regulating valves, FCV-FW-478, 488 and 498. This was used to determine if the main feedwater regulating valve was L inducing pressure oscillations in the system (i.e., control valve instability). f. A linear velocity displacement transducer (LVDT) was placed on the stem of FCV-FW-478 to determine if stem oscillations were occurring. 21
Oscillations could occur as a result of valve resonance with the piping system at a specific flow rate. Specific flow rates can be correlated to precise plant power-level. Monitoring of the. above parameters was performed continuously - during the ' complete heatup of the unit following the refueling.outege. This allowed data to be extracted during all modes of interest including hot standby operation, low power ranges and power escalation. Plant recorded data (e.g., feedwater flow rate, auxiliary feedwater flow rate, steam generator level, -feedwater
- pressuro, reactor coolant loop temperature, etc.)
gathered by permanently installed devices were collected along with the date from the temporary instrumentation. 1 This was performed to provide the basis for data correlation to specific plant-operation. The temporary instrumentation has remained in place after 100% power was_ achieved but is presently not recording continuously. Additional data will be recorded under _ plant transient conditions since these are the time periods where more data are needed. ALsoftware program has been ' written to allow automatic. recording of data under-a- transient = - condition.- Specifically when thermocouple T-11 (refer to Figure 4.5-1) experiences a. temperature ' drop of - 10'. F over less than.120 seconds recording begins. Thermocouple T-11 is on the bottom of the pipe. ._The-instrumentation program has been documented in Reference 9.41. Manual cooldown (transient condition) data were collected for the~two cooldowns that'have.been experienced by the-plant since the temporary instrumentation was installed. These events: occurred on January 19,-1990 and October 6, 1990.- 4.6: STATUS / DISCUSSION Both feedwater: lines 'A' and ?C' were returned to their original = design. basis through the work conducted in 7R. ThisL. included complete. correction of all observed deficiencies considered to be significant. Consideration of-the< instrumentation' recorded data was required to aid- -in root cause. evaluation of both the elbow-counterbore . cracking ' and : the irregular ; deformation pattern _ of -the , piping. l With current design basis considerations being evaluated (i.;e.,_ time history loads _for current postulated water 22
hammer Levents), it-was concluded that the current-. analyzed. postulated events-couldn't generate ; loads of . suf ficient magnitude to move or' deform the pipe into its - 7R as-found condition. It became evident that a review of existing industry ' literature of-feedwater-perturbations was necessary to determine sources of possible significant loads outside of current design basis considerations. The significant-events reviewed were steam generator water hammer (SGWH) and feedwater control valve instability. Global' stratification-at this point was not documented or its -ef fects quantif ed on pressur zed Water Reactor feedwater i i piping as evident through NRCs AEOD study published in March 1990 (Ref. 9.21). - When instrumentation mounted in 7R demonstrated that- - global stratification was_present, steps were taken to incorporate - = its effectc into the design calculations. Subsequent review determined that global stratification, although present, wasn't the source or root cause of the i observed 7R misalignment. Water hammer events and past-installation practices of the feedwater elbows became suspect.- - The remainder of this_ report addresses 1these f actors for si root cause ef fects, investigates the potential for these u ~ occur in the future at BV-1 and identifies . events' to steps-that can be taken to minimize their reoccurrence. l q l, I 4 l.. 23 11 j;
~ 5. 0 ~ ROOT CAUSE CONSIDERATIONS (INCORPORATING PAST INDUSTRY EXPERIENCE) Historically -feedwater lines at-both Pressurized Water Reactors (PWRs) and' Boiling Water Reactors- (BWRs) have been associated with several concerns.. Some of these concerns include internal surface-cracking (local thermal stratification), global thermal stratification, and severe water hammer. Several organizations have studied many of these problems in' great detail. Many of these ' studies were obtained and reviewed. Their results and conclusions were incorporated into this evaluation to aid in identifying the root cause(s) of the taper transition cracking and the irregular deformation pattern of the feedwater piping. 5.1 LOCAL STRATIFICATION ~ ~ Local stratification has been identified as a source of pipe line cracking in feedwater lines at both BWRs and-PWRs (Refs. 9.16, 9.17 and _9.18]. The local stratification phenomenon occurs in the vicinity of the nozzle'to the reactor in a BWR and in the vicinity of the. nozzle-to the steam generator-in a PWR.' This' phenomenon occurs during plant heatup when low temperature feedwater l 'is-mixed with high temperature fluid from the reactor in. l a BWR or steam generator in a PWR. . ~Since' this L investigation reviews taper transition nozzle cracking at the Beaver Valley Power Station-Unit i loop A steam generator, a detailed review of NRC IE Bulletin 79-13 I " Cracking in'Feedwater System' Piping" (Ref. 9.12] was H performed and the results of the review follow, f', 5.1.1 NRC IE BULLETIN 79-13 on June 25, 1979, the NRC issued'IE Bulletin No. I 79-13.to address cracking in PWR feedwater system L piping.' Cracking was_first identified in May 1979 at:DC Cook' Unit No. 2.-- Leaking circumferential' L -L' - cracks' were identified in the 16 inch feedwater -elbows adjacent to two steam generator nozzle elbow L welds.. In-. order to further explore the nature and-L magnitude - of the cracking : problem industry-wide, 1 the office of Inspection and Enforcement requested licensees.of-PWR plants to' conduct volumetric' ~ examination of certain ~~feedwater piping ~_ welds. l H_ , Beaver ' Valley Power Station-Unit 1 was' one of several units-that identified cracking in the-feedwater piping.-to vessel nozzle weld. region.- 1 A Westinghouse Owners Group (WOG) was formed. Westinghouse, ' acting as the agent for the WOG, -investigated the feedwater line cracking 24
t phenomenon. Their results were published in several ASME papers (Refs. 9.16, 9.17 and 9.18). The investigation concluded that the cracking occurred as a result of corrosion assisted fatigue. Westinghouse recommended the installation of thermal sleeves to reduce the heat transfer rate in order to protect the feedwater line taper transition nozzle weld to the steam generator. 5.2 GLOBAL STRATIFICATION General stratification of water in piping was first reported by the NRC in Information Notice 84-87 (Ref. 9.59), which cited a major event in feedwater piping at WNP-2 that resulted in damage -to supports and some leakage at the flanged connection. This type of stratification is due to the slow flow of cold water into a region of hot water, or vice versa, which causes the pipe to bend, leading to excessive deflection in long horizontal runs of pipe. The event at WNP-2 resulted from the slow admission of cold feedwater (about 100* F) into a horizontal run of pipe previously heated to about 400' r. The difference in temperature between the top and bottom of the pipe due to stratified flow caused the pipe to deflect and damage several feedwater pipe hangers and snubbers, and also loosen a flange allowing a small leak. Global stratification has been a subject of increased visibility since the issue of NRC Bulletin 88-11 (Ref. 9.42). This bulletin describes global thermal stratification in the pressurizer surge line. Global thermal stratification is defined as a top to bottom pipe temperature differential which can occur over a portion or entire. horizontal length of piping under consideration. The potential exists for global thermal stratification any time a low flow injection of fluid occurs in a line sized for much larger flow and a temperature differential exists between the injected flow and the fluid already in the line. This potential exists on the feedwater line after a reactor trip when the feedwater line at a-maximum possible temperature of 441' F experiences low flow injection of auxiliary feedwater at a temperature which can be as low as 45' F. Data collection at Beaver Valley Power Station-Unit 1 also indicates that significant stratification can occur when feedwater. flow is stopped. This is due to the tendency for the more buoyant hot water to rise and cold water to fall (natural convection). The top to bottom temperature differential causes the 25 I l 1
~ piping to' deflect in an unexpected direction and with a larger magnitude ' than plant design basis calculations predict. This-was also true with the pressurizer surge line as described in NRC Bulletin 88-11. The AEOD study. summarized in the following section is another industry investigation of global thermal stratification. 5.2.1 NRC AEOD STUDY The-
- study, entitled
' Review of Thermal Stratification Operating Experience', issued in March 1990 by the Office for Analysis and ' Evaluation of Operational Data.(AEOD) (Ref. 9.21), delineates all reported thermal stratification events to-date (including both local and global stratification). The report was reviewed to determine. if feedwater. global thermal' stratification has been report'ed at other plants, Other than the pressurizer surge line [Ref. 9.42), the only; specific recorded instance of global stratification that resulted in support damage was at; Washington' Nuclear Plant Unit 2 (WNP-2) on the feedwater system._in 1984. WNP-2 is a BWR, This. 3 event occurred when plant personnel slowly admitted cold feedwater into a pipe filled-with high temperature water. It also occurred when reactor water clean-up (RWCU) (hot water) was admitted to cold piping which caused back flow-and stratification. Global 1 stratification has not been associated with PWR feedwater piping to-date in the industry.- However global stratification was still reviewed as a potential source of the piping misalignment since injection of colder fluid into the feedwater system ~ takes place under auxiliary feedwater < injection. The id"ntification.of this as a mechanism to cause global thermal stratification in PWR feedwate:/ piping could become a new industry issue that would' have to be addressed. 5.3 WATER HAMMER-Water hammer in PWR feedwater systems has been identified R as the phenomenon which has plastically deformed piping and _ broken pipe supports. Several detailed-investigations have been performed-(Refs. 9.24 - through~ 9.38). These studies investigated the dif ferent aspects -of flow. ' control valve instability and-steam generator -water hammer due to bubble collapse. A detailed review of these phenomena follows the present Beaver Valley 26 e
Power Station-Unit I water hammer design considerations discussion which includes rapid check valve closure (VCW-90 AV) due to pump trip, flow control valve closure (FCV-FW-478, 488 and 498) and isolation valve (MOV-FW-154 A, B, and C) closure. Design modifications were instituted which eliminated the need to evaluate flow control valve instability and bubble collapse. 5.3.1 PRESENT BEAVER VALLEY POWER STATION-UNIT 1 DESIGN CONSIDERATIONS Beaver Valley Power Station-Unit 1 feedwater lines have been analyzed to withstand the effects of the following water hammer events (Ref. 9.43). n. Feedwater pump P-1A or P-1B tripping resulting in a closure of check valve VCW-90AV downstream of the pumps at 100% and 75% power levels. b. The closure of flow control valves FCV-FW-478, 488 and 498 at 100% power level. c. The closure of isolation valves MOV-FW-154A, B and C. No other significant water hammer events were identified and specified in the licensing documents. The analysis of the above water hammer transients provides the following significant conclusions: a. The 100% power condition provides higher peak loads than the 75% power condition when both pumps are tripped. b. Peak loads are higher on the piping. inside containment when both pumps are tripped. c. The pump trip loads are significantly greater than those from the flow control valve closure and are generally bounding for all three water hammer events considered above. d. The analysis of feedwater piping inside containment (Refs, 9.7, 9.8 and 9.9] for design basis water hammer transient events results in stresses well within + ANSI B31.1 1967 code allowable limits. 27 l 1
w e. The feedwater pipe supports insite containment are designed to withstand the water hammer loads. f. The maximum water hammer unbalanced force was determined to be less than 15 kips. 5.3.2 FLOW CONTROL VALVE INSTABILITY Flow control valve instability has been identified in initiating water hammer events. This mechanism was reviewed to determine if it could have led to } the observed piping misalignment. As reported in NUREG 0582 (Ref. 9.24), by 1979 twenty-two water hammer events had been attributed to main flow control valve opening, closing or instability. The term " flow control valve instability" has been used generically to encompass several contributing factors, all of which could Icad to a water hammer event. These factors include unbalanced hydraulic forces on the valve plug which may cause the valve to override the force applied by the actuator or, improper valve flow characteristics (ratio of flow to plug travel) or improperly adjusted control circuitry which may cause a rapid change in valve opening or closing. Another factor which could lead to a water hammer event is possible resonances set up by the dynamic characteristics of-the valve and valve actuator. Resonance is the amplifying effect caused when an input or forcing function frequency and the natural frequency of a system /sub-system coincide. Two sources associated with flow control valve resonance are identified below: a. Standing wave frequencies, defined as pressure oscillations in the piping system, are a function of both the speed of' sound in the medium at the specified temperature and length of the piping. b. Valve plug / stem frequencics, defined as oscillations of valve plug movements, are a function of both mass and stiffness of the valve plug and valve body. Resonance established by a and b above, requires that the valve be dynamically coupled to the fluid system. This dynamic coupling effect would result in valve oscillations feeding or driving the fluid i system standing waves and producing pressure 28
oscillations. The result of the pressure oscillations in turn would have the effect of trying to move the valve plug in and out of the valve body. Instability results as valve stem oscillations amplify. This resonance phenomenon " called impedance coupled valve / piping instability", could result in severe chattering or vibrations locally at the valve, and can induce large sinusoidal pressure oscillations resulting in significant forces being introduced into the piping system. Local valve vibrations could be responsible fce fatigue failure of items such as the valve
- stem, yoke
- studs, feedback linkage, etc.
Sinusoidal resonance oscillations, if produced, would have the same frequency as the valve oscillations with pressure variations as high as 600 psi (Ref. 9.35). This amount of pressure variation could be responsible for loads on the order of 100 kips (Ref. 9.35). Loads on this order of magnitude would cause significant pipe support and piping damage. However hydraulic dampers, which have been implemented at_ several plants including' Beaver Valley. Power Station-Unit 1, severely limit large valve oscillations and hence large pressure variations. Instrumentation data recorded during 7R at all power levels. confirmed that large valve plug oscillations are-not occurring and that pressure oscillations are-not being induced into the system. !~
- 5. 3. 2.1 - PAST BEAVER VALLEY. POWER STATION-UNIT 1 WATER HAMMER EVENTS There have been four documented water hammer events at Beaver Valley Power Station-Unit i since the plant became operational in 1976.
-The specific details of each event have been l previously reported to the NRC under Licensee i l Event Reports (LERs) which have been accompanied by formal evaluation reports. H These .are included under the list of references and provide additional detail. The root cause of the three water hammer events occurring in late 1976 and early 1977 s have been associated with plug / trim 29
instability (consistent with improper - valve characteristics). These events . were associated mainly with the 'B' feedwater line. A replacement of the original plug-type trim with a cylindrical trim was implemented after the' January 5, 1977 event. This change was implemented to reduce cavitation and unbalanced forces in the valve which had been associated with the unstable operation of the valve. The only other documented water hammer event occurred in 1981 on the 'A' feedwater line. This event was preceded by a mechanical failure of the' feedback' linkage whereby valve control was lost. It would appear, which was not-evident at the time, that resonance of the valve with the fluid system caused excessive vibration levels in the valve itself to a -point where fatigue failure of the linkage occurred. Currently, due to the plug / trim change and the addition of hydraulic dampers in 1982,-large valve stem movaments and the resulting large induced pressure oscillations can-not occur. However, as evidenced by recent failures of valve component studs (stem breakage), vibration induced fatigu'. is still occurring locally at _the' valve, Therefore', it is possible with the : current arrangement that mechanical fatigue of'a valve component could render the valve uncontrollable and initiate a water hammer , ant' as was experienced in 1981. The date-'of the events and ; contributing-factors are summarised here.
- a. November 5, 1976 Contrcl valve instability; uncontrolled increase in feedwater.
flow rate-either due to-- a. malfunction in the feedwater regulator _ control or ' unbalanced forces in the valve itself. Damage was mainly limited to-the 'B' line outside containment and included instrument tubing and a valve actuator. Two snubber brackets attached to feedwater piping were found bent in the containment. 30
j ..? ; ' -a l
- b. Dec?mber 27, 1976 control valve instability; valve instcbility causcd by the plug design and mismatch of the valve system amplified the response to the system *;ransient and the valve moved independent of the control signal.
Pressure oscillations were observed to be on the order of 400 psi. Damage was. limited to mainly instrument tuoing on ' the outside containment 'B' line piping. f
- c. January 5, 1977 control valve. instability; valve again moved independently of control signal' (opened despite signal to close).
-Damage--again was mostly limited to the 'B' outside containment piping which included instrument tubing and a damaged motor operator. An instrumentation program was put J into place. Valve plug and trim changes were . implemented on all three feedwater regulating valves.
- d. May 6, 1981 Initiating-cause - of event-was a disconnected feedback linkage' on the 'A'-feedwater regulating valve.
- This resulted =in a large-increase in f eedwater ' flow to ~ the -'A' steam- . generator-followed by flow : spikes 'in the 'B' and 'C'-loops. Damaged-resulted.to both the'inside-and outside containment piping on i the 'A'- feedwater' line. This included. damage; to. outside containment instrument-tubing and a motor operator. ~.Inside-containment damage-included-. bent snubber . brackets and dislodged rupture restraint shims. 5.3.2.2 REVIEW OF NUREGS Several NUREGS have been issued, mainly in the late. 197 0 ' s, to compile' and classify past water hammer events (i.e. NUREG-2059) (Ref. 31 l l l
f 3l 9 9.25) and tot. define methods of evaluaticu ~ (i. e. NUREG-2 7 81) '( Re f. 9.26). As specifically related to feedwater control I valve' instability, water hammer can occur.if- 'the feedwater control valve-ic improperly
- sized, the control circuitry. is f uprcperly
- adjusted, closing and opening time is 1mproperly adjusted or-the system isn't filled and vented properly' prior to startup.
These i factors were evaluated for their effect on BV-1. 'Of these, the most ~ significant water hammer event can be attributed to -improper VGve-sizing. Unbalanced forces in =the ' valve itself, created possibly due co cavitation, can be of_ sufficient magnitude on the plug and-stem to overcome the force being applied L by the-actuator. This.is.the phenomenon considered to -' be associated with the; water hammer: events.of-1976 and 1977. -A design change to replace the trim in 1977 Cn 3 all three main feedwater regu~ ' ting' valves was instituted. _ Subsequent moni coring' - through instrumentation installed-both on the valves and -feedwater piping showed tnat: the trim : change was effective in reducing the likelihood of: reoccurrence "of this. event. - Results of this' study and evaluation of-the feedwater. regulating _ valve sizing to _ meet system: requirements vere submitted previously_ to the NRC (Ref. 9.22). ' Subsequent-changes at BV-1, : including t.dditj on '. hydraulic - dampers -further minimizes the. likelihood-of feedwater' control valve instability water--hemmer. Water hammer events associated with rapid open or. closure times-or. possibly improperly _ adjusted control circuitry are-less severe.- Analyses performed following the 1981 water hammer event in which a mechanical failure-of 1 the linkage was cited, showed that with a' rapid closure of the valve-(approx. .1 sec), loads'of_only approximately-5 kips axially on the horizontal inside. containment piping could be produced- (Ref 9. 52). Although bent anubber 'l -brackets were.found, loads of this-magnitude-are insignificant in terms of pipe strens. 32 L.. 1 e
__m,_ ~ . ci; -'g, [% Therefore, d'ue to resonance of-the valve with-N, the_-third standing wave (17 Hz) at 65% power ~ on the 'A!-line, _ future - water hammer events may occur through a mechanical. failure at the -valve-induced by this -localized vibration. Future design considerations should take into account decoupling the-valve-stem / plug frequency from the third standing wave. [ 1 5.3.3 -STEAM GENERATOR WATER FAMMER Steam generator Was'r hammer (SGWH) is a term used
- to describe a wat9tr.3ammEY, rent initiating in the feedring of the steam generat.". This event occurs -
when the f aedring is partiall-voided, for example, following a feedwater ieolat on signal. Under this scenario _ steam is then drawn into the feedring . establishing-a steam / water interface. It is then ,1 possible'that voids or pockets of steam can form. Subsequent'co3d water. injection through auxiliary feedwater.can cause this void to collapse such that a - water slug : moves rapidly into the void. When - this slut impacts the incoming water filling-the header,-_large hydraulic pressure waves can be created. This phenomenon has 'been studud extensively by - Creare and documented under NUREG-0291 [Ref. 9.31). -A stuam generator water hammer event at Trojan will be cited-due to its similarity to BV-1. 5.3.3.1 CREARE STUDY Steam generator _ water hammer events occurred-in the ~ ear 1*] iS'10 's :- with the most notable event: being at Indian Point. At that time bottom-hole discharge in-the feedring was used (: L which meant that-when-the. steam generator L Lleveli dropped below-the feedring.as under_a~ feedwater. isolation signal,. steam could be drawn into'the line. Water hammer _was-ther ' initiated. As'.a. result of these-early events, four h modifications were proposed to minimize the l' potential for occurrence and are listesd hero N. 'a. Top discharge-through J-tubes. Thio entailed plugging the botton discharge holes and installing J-
- tubes, l
33 i i i-
b. Minimization of horizontal length of 4 piping from the feedring T section-A, by installing a loop seal.- The' loop i seal was.to be installed as close to the steam generator as possible so ea -to minimize - the. length of horizontal run the slug could i, traverse thereby minimizing slug-impact foress, r c.- Minimization of time to initiate n auxiliary feedw:ater injection. The effect of thin is : to minimize the amount of-draining that could occur in the xing from the time main feedwater was isolated. y y d. Institution of - an - upper limit on maximum auxiliary _ feedwater flow' rates following a main feedwater. isolation. This'would minimize the-likelihood of generating-a slug and. collapsing a steam void, of - the. above' f our items, 'a, b and c were j implemented' - at BV-1. The fourth was not instituted.- The justification. for this vas that-testing was performed ~ at Trojan.[Refs. = 9.- 6 8 - and 9.69). This testing.- was conducted. with7theza through c' modifications in place at Trojan. The testing -involved varying ' the j auxiliary -feedwater -flow rates-from: approximately 200 to 440 gpm. The results of thir testing:showed that with these auxiliary. fe.edwater.. flow rates ~ steam-generator water hammer-was.not induced. It-'must.; be stated that provided 'feedring integrity is-ensured (i.'c.,~~no source of out-leakage) and the' ring. remains water solid,
- l highLauxiliary feedwater flow rates should,be i
of no; consequence. However, should the ring: be 1 allowed to drain through either cracked 1 p' lugs in -the bottom. of the feedring or erosion 1 7 of the: thermal liner-at the slip fit region in= [ the steam. generator nozzle, then = the ' safety - analysis performed by-Trojan would be i invalidated.- -Through a review of the NUREG, the effect of a -leakage hole in the feedring on the vater level in the ring is quantified. For. a 34 l
Wer.tinghouse feedring and with feedwater isolated f or only one minute (time it takes for auxiliary feedwater to initiate), should a leakage hole arca equivalent to one hole be s postulated it is shown that the water level height in the ring drops to approximately 81%. This alone would allow for a significant amount of steam to be drawn into the ring. In addition, without an established upper bound as is presently the case at BV-1, auxiliary feedwater flow can be as high as 460 gpm at 100% power. This is far above the 150 to 200 gpm range cited for cold water injection in this system co minimize the potential f or wat',r hammer (Ref.
- 9. 58).
Since significant loads, possibly on the order of 200 kips [Ref. 9.13), could propagate as a result of this phenomenon, steps should be taken to either periodically check the integrity of the feedring or institute an upper bound on auxiliary feedwater flow. r-t 4 35 a.- -
l I 6.0 RESULTS OF ANALYSIS (POST 7R) Global stratification was determined not to be a contributing mechanism fcr the T irregular deformation pattern, therefore, other pote.' ' al root causes such as water hamer and past installation practices were roviewed. Tnts entailed a review of existing industry information. During the root cause evaluation unanalyzed condicions were identified (i.e., nonoballo locked and global stratification etf acts) ead their effects were incorporated into analyses. The results of the various analyses are summarized in this section of the report. These analyses were all performed af ter the seventh Beaver Valley power Station-Unit i refueling outage (7R). Section 6.1 discusses the steps undertaken to ensure that the computer model agreed with recorded plant data. Sections 6.2 and 6.3 discuss the recently identified i glob-stratification phenomenon. Section 6.4 investigates the effects of water hammer. Section 6.S investigates the effects of local thermal stratification and Section 6.6 reviews possible pipe misalignment during the replacement of the cracked elbows adjacent to the stean generator. 6.1 CORRELATION OF ANALYTICAL MODEL TO PLANT DATA The correlation of the analytical model to the plant data was performed in soveral steps. The first step was to develop a STRUDL-SW (Ref.9.44) piping model based on the NUPIPE-SW (Ref. 9.45) piping model of record. See the attached STRUDL-SW piping models (Figures 6.1-1, 2 and 3 for loops A, B and C respectively). The piping geometr:f is modeled on STRUDL-SW because this program has the capp ',ty to analyze a top to bottom temperature differential over some or all of the elements in the model. The next step was to benchmark the STRUDL-SW model to the existing NUPIPE-SW run of record model for the 100% power level operation mode. The displacement agreement between the two models was excellent. See Figures 6.1-4 and 6.1-5 for the displacement profile f rom the non-stratified thermal cases. The remaining steps s involved correlacion of the recorded plant data (gathered in Action Plan, Section 4.5, Implemew
- tion of Instrumentation Program) with the results from t;.RUDL-SW analysis. This was done to verify the base model before global stratification analysis could be performed.
The correlations performed and results are discussed in the i 36 (
9 e 6 i M N C O 2 3 OD = x o-s - w o E w O O L, e. O. O. / =. C e o c z -
- =
'/ 4". I CD - to e g g,- g L-O D CE C t *. si .is. J Z r = c,. a ; y g C C % tS 0 w E g i, WW q g h. g 6 >N C 3 L' !~ p x
- . 33<
= m g t V" ^ t ~ C S n 7 il e
- r%
st e, s N N g ?( Ef L L h x w (t O Z g w a n C e \\ C N U g k w e 1 o y j 4 8 \\, BO a:H x 4 to w C 3 C s w W 6 \\ e. e C 9 s L m l r l l I 37 I
4 5 4 -.m h hge enn go b' E A b x o c-c = e "e"C.C.o.O 2 C W W O O L. UD Z g C C to u.- e" o / 2 e e 0 mr 6 O D E = m O o i o C 2 Z (L s c 3 C % 0 4*, w E Ww w y K >N aD C O m p yE r o + * - V' g K 4 %D7 >g "
- m 5
/, N D 2 W# 'N
- q U
Q + '\\ m O \\. gW o OE f' O .4 g i H b 2: a 40 C 4 2 W O P i to W O t H Q w ine Q C b pf, [ ~ m H 2 to w w C 2 O w W L d a q~' m,' 4 I l 38
l e D* C M C_ Z O 0A x g /
- e. _
Z $e f.; e bE 5CC9~ ~ 2l S e _ _ _ o. / o o = - = _c t e e n e o E .O D 1 O W ,,O \\ J C L e n . r c. .s O 2 \\ W a8 C C C % 0 W W C w, y % 4 @i J W w _a i + u El O 3 D w Y k } Y C! n + A W W O ? O [ 7 w/ = ~ w e De 6 g-t* W ~ $ 4 C N W ~ D 4 E Q \\ o w \\ v ti. C w D a 1 E b m to Z 4 m e w C N 2 P C e w W L Y./ ~ 8 E wf_ E L i e g s \\ \\g \\ e\\ w l l 1 39
e 0 e 6.= L*. 1 O 2 0 O.h x Q_ c
- ")
a 2* g U / O C DC e _, e, c. o [ t, t x w O O 4 x =. 0 z - e t;. e a .- t a k
- s x C C s
E u. ~. n L , C 2 W
- J
D 4 g C O k % 0 0 W E w w c" h m x p % ] C U E s. O O t xl?g}3e o CC M x N C Oo \\ e w 6 G. 3 Y A \\ tm) O
- ~
O .a-E N R y t r ~ d b / 1% C e a e e W\\ Z o u U g I .\\ D m w o i Q g 4 g to AD 2 E H x to w C2 Cw J O w X \\ O %D 9 40
T 4 f S F 0 i r G X, 0 <3 0s D a t J i i t
- 0. g; t
R E G O r ~ 0 6 i N R R - n n4 Y 1 l i 3 a F f t f O U D t T 0 nl 0 O. y T S H D i S 5 P r Y I F M f G U R O S t L R P L o W l 4 I O i t i LL O H R S S E r T E r f x Y 2 L t G I Og v i Y C o r l l J 4 - Wn 3 F M X - F
- n' g (
4 3,, " D 8 e A "' $ 7 r I x N B O R P LE D P O 0 M n 3 I C -1 C P O 6 O T J E L t E R 1 U W -a G B t i I L n F i D n U o R C TS E D Is 1 R E inW D E G E /Ft F 5 \\ 1 s V j B Np e ) A 2 ~ s-
4 9 9 e he c b k y >I Er C-n $Eo 5E$E 5 r o. e. E / O 8-w= c=-en j o u.. e z. a + u ys=w N e a z x a8gg 2 6w e b 5e += X = w b hw J D c= ur m am e O ~T }5e N k a. e X 4 f
- 7. CN f
's, x w k;& 6 x =_ - -h ~. / b a / O Co / w L 0 w c Q fM C M e o W 0 3: v 8 M y i U H As C 8 rE ~ 3r c o O z 3 H o b ~ W a w N Q to O i w >= C 2Q )d W w W / .,/ ew \\, ,I 5 b C .A h L' .w>/ 42
i v k, 9 U I,0 C X brg ~5 1 7 U O u u. W c. - O C 2 ,e 0" O - % 6 4 ZOCO m I i j Q _J U L g ug C - 5_ p 'C I D' u i E 'S E L Z CO-O t- , h3 -C 2 QC C O 't H !! ,9 y O O_ g a p: 9_
- c. O w%
- u. t 2
U m a w x >- N 1 OZ QW E W.) Q: 5 O uW ) D y > r m u u O a LO _J h c,- ,o .s s. I 64 / e.d e 4 4 M 4O LA.' 4 la) .p. D***D. ,p. ~ .g*ct,S "p..'p, 6 + A Z o, [ ,c g - -~% o, L 'u. Z la.1 h ,U' . e/ N, g ~ U ,o m .c ~ q' M s' M g,f 'k IC W b
- r/
C, s g D/ y t b a' e ey
- a C
2 r ? k LC /. W 4 t; aC I C C.C C 9 4 3 43
l 0 g 4 0 C b W 9 E g W Du ps b, 9 Z h0 g C 2OCO p i u Q3 C C u e m _W i- >w o. u"N 0-CO J W 1 C> q t .. b -C Z QZ 9 it ti li OO-kN M 3 U 'J g 4 a O W WX>N M 1 0 T W ' - + W J g uu D p 8 x u u o e > r n e m u o r 48 8 F, = .I 4 d et N O b E 5 L g w .I m L. w b p'. { S D i: e c 8 n E C. I ZO C z ~ t: Q ~ '. I, to C 5 'l-4 g b i: 9 9l: l C cll .I lI k m 44
-.4& a 4 6 6 S O L9 C X b b") h' e 5> uOu n D mC ,n Z N$ h ZOCO n gn Q - O, u t 6-E O o, C 5'W yx u i O
- u L3 uZ CO-O N
m N E C ^ - > +- # U , b3 2 g2 c O,$ n,i a 3 I M C C W"WN +d k b U LX>N O d- ,E f 't 1 O b L U E (O D r LD a: U; ) u u O d y x rp c vo r "9 d C A e Nm -k O g E e g * * % N.....,..,' o 4 a'- (. e m \\ 'w C v, W c' - T: C a p -4 0 b y .. g e ..c m R r / c' to f f - I g L W I: b t N 84 14 a 9 C S C O l 9 ' b. 45 I l
9 0 4 e R [*) ab y is ( o=D z~ C-cp- ,w O U 99k C$ (D a
- ~
C ,E Q 0 -* hjj0 OW N F-CD C >.. h'4 m , h' -c ~ o -, v OO O2Og ghg 3 g g w-
- u. x >- N a o c O
F.4-k dJ q a ea~ ~ h"C g O y y 3 $ g' u O i / m g L O t U E V 5 2 0 Nm w e l L., i c 1 o 4 6 N 3 A W M 8 w W C ~ a x las a 9 6 kJ g 2 M fr C 4 E ii N h m I h G d s O n h g he
- 1 CI N
- t W
.o' n;; A7 a. s, C C - 7 C 9, i ii 46 1 \\ o 46
-i following two sections entitled Normal Thermal Movements I + -- and Incorporation of Global Stratification. Only loop A j d2splacement-data were recorded during the power ascension from 7R. All conclusions drawn from the evaluation of loop-A data are applied to the analysis of loops B and C where applicable. j In addition to the licensing basis STRUDL-SW finite element'model results, a personal computer version of ANSYS (Ref. 9.46) was utilized. Excellent agreement between the two conputer codec_was obtained for all the postulated thermal stratification scenarios (Ref. 9.47). l 6.1.1 NORMAL THErsiAL MOVEMENT Initial attempts at correlating the recorde'd normal i thermal displacements with the STRUDL-SW model were i unsuccessful considering the au-designed support functions. The monoball supports are designed as-vertical only supports.= The single trunion-ball and socket design on sliding plates is intended to provide freedom of movement in the other five ' degrees-of freedom. A second series ' of computer j runs were 'perf ormed with the monoballs locked in 4 both'the lateral and axial n rections and allowing { free rotational movement. 'e correlation-between . the STRUDL-SW model with'tne monoballs locked and. I the plant recorded data was very good (see Table 6.1.1-1). Therefore the interim qualification of the feedwater piping analysis was ' performed considering the monoballs locked (acting as'3-way constraints) as well as -unlocked (acting as vertical constraints -_ only). For-the long term solution the monoballs will be addressed as stated in Section 8.1.2.4. 6.1.2 INCORPORATION OF GLOBAL STRATIFICATION Two ' separate cases of global stratification were considered in this analysis. Both of these scenarios were' experienced during the recent post 7R heatup. The first global stratification _ case is 4 auxiliary feedwater injection after a. reactor trip at 29%: power. The second case,is described as-significant intrasystem = stratification 3 (stratification caused by two different temperature' ~ fluids within the feedwater system). This occurs o . when feedwater flow is reduced. These two cases are discussed-in detail in Sections 6.1.2.1 and 6.1.2.2 respectively. 1 47 e+, ..,.. m a m.,__m m,=.s-m., ,,.----.m, -,,,,,.,..J.. ,-,....mm., ,y,
TABI F 6.1.1-1 i STRUDL SW VS PLANT DISPLACEMENT DATA (LOOP A) REACTOR POWER LEVEL AT 29% AND FEEDWATER TEMPERATURE OF 335 F Lanyards STRUDL Measured Global STRUDL Global Joint Displacements Displacements 1, 2 & 3 45 X 1.41" X 1.90" Y 1.34" Y 1.00" Z 1.34" Z 1.43" 7&8 125 X 0.09" X 0.21" -0.66* Z 0.59" The approximations made in the STRUDL-8W evaluation and the inherent variations in field measured data make it unreasonable to expect an exact correlation between recorded and calculated-displacements. The displacements shown above indicate as good a correlation as should be expected. l 48
i l t 6.1.2.1 AUXILIARY FEEDWATER INJECTION APTEA 29% POWER REACTOR TRIP After the reactor trip the feedwater line was at approximately 335' F. Auxiliary feedwater at about 45' F was injected into the system and the line thermally stratified. ? Initial attempts at correlating the recorded-auxiliary feedwater injection global stratification movements with the monoballs free to move as designed (see Section 6.1.1) with recorded data were unsuccessful. The-pipe moved opposite to the predicted direction. This problem was also experienced in the Section 6.1.1 (normal thermal movement) correlation. The best correlation of data occurs when this monoballs are locked in the axial and lateral directions-(see Table 6.1.2.1-1). 6.1.2.2 INTRASYSTEM STRATIFICATION Intrasystem global stratification can occur anytime during plant operation, but significant intrasystem thermal stratification - has only been observed during power reduction (and subsequent.feedwater flow reduction) at power-levels.-less than or equal to 29%. The -feedwater regulating-valves are bypassed at low power. operation (normally 15% to 30%) and condensate is obtained directly from the condenser hotwell. When the feedwater line is hot ='and. flow to the-steam generators is reduced the feedwater lines will stratify (see t Figure 6.1.2.2-1). .The maximum intrasystem global stratification is : considered to be enveloped by the. maximum auxiliary feedwater injection case because a larger stratification -potential exists for Ehe auxiliary feedwater case. Stratification potential is defined as the maximum feedwater temperature minus.the minimum auxiliary feedwater temperature- (see -Section '6.2). -The maximum ' potential - f or intrasystem stratification is 255' F. This is =a result of a feedwater' temperature of 335' F at 29% power and a condenser hotwell temperature of 80' F.. All conclusions drawn from the displacement correlation in loop A data are applied in the remaining two loops. i' Instrumentation was not installed in. loops B and C. 49 h -.4 - %,.,. - --,~ -m m - - --,_ _,,,,--, ,m,.,m.
T.ARI E 6.1 J2.1-1 STRUDL.SW VS PLANT DISPLACEMENT DATA (LOOP A) AUXILIARY FEEDWATER INJECTION Lanyards STRUDL Measured Global STRUDL Global Joint Displacements Displacements 1, 2 & 3 45 X 1.19" X 1.72" Y 0.04" Y -0.66 Z 0.89" Z 0.82" 7&8 125 X 0.22' X 0.38" Z 0.51" Z -0.47" The approximations made in the STRUDL-8W evaluation _ and the inherent variations in field measured data make it unreasonable to expect an. exact correlation between recorded and calculated displaconents. The displacements shown above indicate as good a correlation as should be expected. i l 50 1 i .. ~.
A 4 i-.._..-.er. 4. .4 C tI o ) O I l f N 1 i l e \\ 'l l I ~u k 2
- /
1 o ,s ~2 syog-t -r 8-88 1 n08 p 8 om ~ on 5; i ga. gys ( I 9 P-s N j s e s e pC4 58 l e>k - = W N mga \\ 8 O gu? se e s b H m g \\j w e g \\ m, k g .t t e to C 1 + O o O on I L03o] 3an.'.vaseW2.t l i I i 51
6.2 NEW GLOBAL STRATIFICATION LOAD CASES (NOT PREVIOUSLY CON'4IDERED) New global stratification cases have been developed. These new cases have never before been considered as part of the design basis for PWR feedwater system piping in any plant. These cases envelope the effects of postulated-i thermal stratification. See Figure 6.2-2 for sample - ) displacement profiles caused by thermal stratification effects. These cases.are based on the extrapolated case of :the - recorded auxiliary feedwater injection.at 29% power and the maximum potential te:nperature dif ference. The maximum feedwater temperature of 441' F is at the 100% power level. If a reactor trip occurred in the winter, the. temperature of the fluid in the safety related demineralized water tank WT-TK-10 located in the yard, heated only by space heaters, could be as low as 45' F. The auxiliary feedwater system would inject'the r 45' F water into the 441' F feedwater system. -The temperature difference of 396* F would not be realized based on ' two recorded. temperature differential cases-(Ref. 9.41 and Appendix 10.1). -During the more significant case of the two (Ref.9.41), the feedwater line was at 335' F and the auxiliary feedwater line was considered to be 45' F (actual water temperature was later determined to be 50* F) The potential for 290'- ( LF' stratification existed, however, only 180* F (250' F -(top of--pipe).- 70' F (bottom of pipe)) stratification was realized. This resulted in.a 62% scale. f actor (180/290) (see Figure' 6.2-1). This scale f actor is applied to the full range potential-stratification or. O. 62 x _ (441-4 5) = L 245' F. All horizontal - piping was conservatively considered to have the potential to be stratified at 245' F. Piping with-vertical components (e.g., risers) are considered to be unstratified in the-model. See References 9.7, 9.8 and 9.9 for the= specific global thermal stratification profiles that constituto the new global stratification cases. 6.3-PRESENT CONFIGURATION-GLOBAL-STRATIFICATION AND CONSTRAINED MONOBALL SUPPORTS EXTRAPOLATED TO DESIGN ~ BASIS TEMPERATURES After the magnitude-of the' global. stratification-temperature was determined, it was considered in the three STRUDL-SW models. Loops A and C both have monoballs. Since the monoballs in loop A were postulated to be acting as 3-way supports (axial / lateral restraints as well' as their as-designed vertical function) loop A -1 52 fwee -9 g,iw y w--g-9w77-y- ry--ygy-ge,,vy-g-wcyai s g,,-- y- ,w <--yi y g-w y y
m '( \\ f'1 / ~ y -,( Lt _e ' I .S = f. L n. o 53 m o p z-t 03E mN ? dag l / g8 s gue / -*g$
- d-hh5 ff up*
e n H O ge O H 1 g h I yE 4 = E W F -N
- g. 5~
y \\ I p-o 8 8 8 8 8 8 o e c n m [3 69p] DJM%DJedW9j, 53 ___=-----_---_---:z-.-
~ ~ bu9 n US 9 b-C x U O u E a K v .O Z r,M - O E C C O, k. ~ r o ,l _2 O. C C. c 9 99i u i O._ _ e w, D .J u$ uZ uN E. O ~ O N A a m g w ~ ,3 [ 2 2 d C ll >l il D
- =e 3
OC u~ U 3 'A rn 5
- N C
1 b' ~ ff CC D O_ CC LD ff L7 .J C O ww O D p h x U U C y >r 7 c .I 3 4 J e r d smo 04 L b n. ..rg r> > + - Q *.o...g*s.y8-g n M w. -m N A 6 w ..e twN,. 8. m . v V,, e - .*s, su u LC w a .e / xU 8 w M s ,U g H c' / 6 '/ e, t' H ,0 l E 44 / c' O i 2 I e =_ 9 c. i W Ce 4 g. i O' p w es= 54 I l l
9 0 0 e -) d h C (+ . p -t e 3 t o i v 5 E,, .a o u a + h ~ P. o _. m r O'? $ g I] O * * *. O o '*9 ) O 4 mm n s > ~ y i,' O ~W p-CO a J L. + w.* w ssV e u a C ~C 2 O 2 'l C O rt u ll W ,9 C C-rd Z W in H$b ? yt d,,. X*N mw c O c .O t 1 ,,. O ~ n p:1 vi J w w -g ww, ) a r-U u O f. E N > r n o 0 .J 7 g i f s' ?B m O Y, U-a i -k O N. N. 6 A M Le, j w w t p: A a 4 e w 5 m k A o m h W h 4 -E ill s a a >= q k v. I x G tg ?.= l + 1 F c F b -C og t l 55
was analyzed twice (considering the monoballs once as a three way and once as a vertical). In the absence of recorded data, loop C was also analyzed twice (once with the monoball as a three way and once as a ve).ical). The results of pipe stress analyses (Refs. 9.7, ' 8 and 9.9) for all three loops were determined to be within the original plant design basis allowables in accordance with ANSI B31.1 - 1967. The stress analysis of the feedwater piping considers both the global thermal stratification moments generated along the axis of the pipe as well as the circumferential stress generated in the piping due to the hotter top of the piping expanding more than the cooler bottom (see Figure 6.3-1). The circumferential stress is an internal stress and does not produce external loads. Tables 6.3-1, 6.3-2 and 6.3-3 contain the pertinent pipe stress analytical results and load combinations for loops A, B and C respectively. Global thermal stratification is the only loading condition where the feedwater piping closes the gaps at pipe rupture restraints. This does not occur under any other loading condition (i.e., design basis fluid transient or seismic event) unless that loading conditions occurs concurrently with global stratification. The loadings generated at the restraints by stratification are significantly less than the design basis loads developed by pipe break scenarios. Stress levels used to evaluate pipe break locations were impacted by the consideration of global stratification effects. No additional break locations or restraint requirements were identified utilizing the guidelines provided in Reference 9.6. Pipe support analyses (Refs. 9.48 and 9.56) indicate that all of the existing pipe supports, except for monoball supports (H-3 and H-4 (loop A)) and one sprir, hanger (SH-6 (loop A)) passed their design basis code eq.2ations. See Tables 6.3-4, 6.3-5 and 6.3-6 for the pipe support results for loops A, B and C respectively. ASME Boiler and Pressure Vessel Code case N-318 was used to analyze the local pipe wall stresses for spring hanger SH-6. ASME Section III Appendir F rules were used to show interim acceptability of the supports H-3 and H-4. Several inspections of snubbers and hanger frames are also required to confirm assumptions (see Tables 6.3-4, 6.3-5 and 6.3-6). Snubbers to be as-built are HSS-201, HSS-202, HSS-203, HSS-204, HSS-205, HSS-206, HSS-212 and HSS-212A. Frames to be as-built are H-3, H-4, H-5 and H-11. 56
4 1 CIRCUMFERENTIAL STRESS DISTRIBUTION DUE TO TOP TO BOTTOM TEMPERATURE GROWTH n ~. V t i HOT i, i D INTERFACE .t ~. i COLD y ~ /\\,. / .y ^ * ,._.,4y.,# p FIGURE 6.3-1 ( e i t m
~. - TABLE 6.301 PIPE STRESS ANALYSIS RESULTS LOOP A i LOAD COMBINATION MAXIMUM ANSI B31.1 FACTOR OF MONOBALL CALCULATED ALLOWABLE SAFETY *"*" CONDITION STRESS *" STRESS (JOINTP 7 LP + DL 7052 PSI 15000 PSI 2.13 FREE"" (75) LP + DL + OBEA + SRSS(OBEl.TH) 15903 PSI 1B000 PSI 1.13 FREE"" (10) LP + DL + SRSS(SSEl.TH) 11482 PSI 36000 PSI 3.13-LOCKED" (160) 24764 PSI 22500 PSI 0.91 ""
- LOCKED"
- **** THER (205) 30908 PSI 37500 PSI 1.21 LOCKED"
..r** LP + DL + THER (205) p. 17846 PSI 37500 PSI 2.10 LOCKED" CTD + H .(30) The joint listed is the point where the maximum stress occurred. All joints were checked for all loadings. " Locked" refers to the analysis performed considering moncballs R 3 and R - 4 as three way restraints (vertical, lateral & axial). Maximum stress from either the " Free" or
- Locked" monoball case.
" Free" refers to the analysis performed considering the monoballs R 3 and R-4 as vertical restraints only (as designed). ^ Either cr.4eria may be satisfied. Therefore, the overstress of the THER cnteria ' is acceptable based on the LP+DL+THER stress level. - (Allowable Stress)/(Calculated Stress) Loadino Conditions Analyzed Deadioad_. (DL) Operational Basis Earthquake inertia (OBEl) Operational Basis Earthquake Anchor Movements (OBEA) Design Basis Earthquake Inertia (DBEI) Circumferential Temperature Distribution (CTD) Time History (TH) Thermal Expansion (including stratification effects) (THER) Longitudinal Pressure (LP) Hoop Stress (H) 58 e c.g y%.ne-s- er- + e9 w 4 ,1 N ~ ~eNU!*=r n v t--m tu-b h
TABLE 6.3,2 PIPE STRESS ANALYSIS RESULTS LOOP B ANSI D31.1 LOAD COMBINATION MAXIMUM FACTOR OF JOINT
- ALLOWABLE CALCULATED
$AF m *** STRESS.... STRESS LP + DL 6433 PSI 15000 PSI 2.33 75 LP + DL + OBEA + SRSS(OBEl,TH) 14095 PSI 18000 PSI 1.28 10 LP + DL + SRSS(SSEl,TH) 10167 PSI 36000 PSI 2 06 75 22791 PSI 22500 PSI 0 91 " 75 .* THER 29225 PSI 37500 PSI 1.28 75 .* LP + DL + THER 16310 PSI 37s00 PSI 2.30 10 CTD + H
- The joint listed is the point where the maximum stress occurred. All joints were checked for all loadings.
- Either criteria may be satisfied. Therefore, the overstress of the THER criteria is acceptable based on the LP+ DL+THER stress level.
- (Allowable Stress)/(Calculated Stress)
- Maximum stress from either the " Free" or " Locked" monoball case.
- (Allowable Stress)/(Calculated Stress)
- Either criteria may be satisfied. Therefore, the overstress of the THER criteria is acceptable based on the LP+ DL+THER stress level.
Loadino Conditions Analyzed Deadload (DL) Operational Basis Earthquake inertia (OBEI) Operational Basis Earthquake Anchor Movements (OBEA) Design Basis Earthqua'KO Inertia (DBEI) Circumferential Temperature Distribution (CTD) Time History (TH) Thermal Expansion (including stratification effects) (THER) Longitudinal Pressure (uP) Hoop Stress (H) 59 I mis m'iu iu-
_ TABLE 6.S 3 e PIPE STRESS ANALYSIS RESULTS LOOP O AN8I B31'1 FACTOR OF MONOBACL LOAD COMB'. NATION MAXIMUM CALCULATED ALLOWABLE SAFETY "'" CONDITION STRESS "*". aTRESS (JOINT)* LP + DL 10508 PSI 15000 PSI 1.90 FREE"' (10) LP + DL + OBEA + SRSS(OBEl,TH) 15347 PSI 19000 PSI 1.17 FREE"' (10) 129s5 PSI 36J00 PSI 2.00 FREE" i LP 4 DL + SRSS(GSEl,TH) (10) 22026 PSI 22500 PSI 1.02 LOCKED" ..*THER (125) 28716 PS! 37500 PSI 1.31 LOCKED" ...* LP + Di. THER (125) 16310 PSI 37500 PSI 2 19 CTO + H The joint listed is the point where the maximum stress occurred. All joints were checked for all loadings. " Locked" refers to the analysis performed considering monoball R 11 as a three way restraint (vertical, lateral & axial). " Free" refers to the analysis performed considering the monoball R 11 as a vertical restraint only. Either criteria may be satisfied. (Allowable Stress)/(Calculated Stress) Maximum stress from either the " Free" or " Locked" monoball case. Loadina Conditions Analyzed Deadload (DL) Operational Basis Earthquake Inertia (OBEl) Operational Basis Earthquake Anchor Movements (OBEA) Design Basis Earthquake Inertia (DBEI) Circumferential Temperature Distribution (CTO) Tirne History (TH) Thermal Expansion (including stratification effects) (THER) Longitudinal Pressure (LP) Hoop Stress (H) 60
TABLE 6.3-4 PIPE SUPPORT RESULTS - LOOS A SUDPORT NUMBER QUAUFIED FOR CONFIRMATION NOTES ACTIOfJ REOUIRED REQUIRED ) WFPD-SH-1 DESIGN BASIS NO WFPD-SH-2 DESIGN BASIS NO i WFPD-R-3 INTERIM NO Apperwix F for bolt allowables JCO Appenjix F for rnember shear allowables Walkdown required IWA - OBE loads compared to DBE allowables WFPD-R-4 INTERIM YES IWA - OBE loads compared to DBE allowables JCO Con Reqd for use of 60% CTD stress Results of data coRection Waikdown required DESIGN BASIS NO Walkdown required WFPD-R-5 WFPD-SHE 318 W im MA ad@ [ b WFPD-SH-7 T N ^* DESIGN BASIS YES 6.2' swing angle [ WFPD-HSS-201 Con Reqd for piston setting Waikdown required i ) M8IG" 8^8I8 WFPD-HSS-202 ' Con Reqd for piston setting Walkdown required i 5.2* swing angle WFPD-HSS-203 DESIGN BASIS YES Con Reqd for piston setting Walkdown required i DESIGN BASIS YES Con Reqd for piston setting Walkdown required WFPD-HSS-204 WFPD-HSS-205 DESIGN BASIS YES Con Read for Diston setting Walkdown reouired WFPD-HSS-20G DESIGN BASIS YES Con Reqd for piston setting Walkdown required Note: The loads include the effects r of global thermal stratification i-and postulated locked monoballa. - - - ~ - - - - - - - - ~ ~ - ~ ~ - - -
- TABLE 6.3-5 ~ PIPE SUPPORT RESULTS - LOOP B - SUPPORT NUMBER OOAUFIED FOR - CONRRMATION ' NOTES ACTION REOUIRED REQUIRED WFPD-SH4 DES!GN BASIS ' ~NO YvFPD-SH_g DESIGN BASIS NO WFPD-HSS 207 ' DESIGN BASIS 'NO Primaryloads dMded by 2 for PfTRIFE i DESIGN BASIS NO Primary loads dMded by 2 for FiTRIFE WFPD-HSS-208 DESIGN BASIS NO WFPD-HSS-208A l i Note: The loads include the effects og global thermal stratification O O
/--
e TABLE 6.3-6 PIPE SUPPORT RESULTS - LOOP C SUPPORT NUMBER QUALIFIED FOR CONFIRMATION NOTES ACTION REQUIRED I PEQUIRED A WFPD-SH-10 DESIGN BASIS NO WFPD-R-11 DESIGN SASIS YES Con Reqd for use of 60% CTD stress Wa!kdown required Restits of data coIIection DESIGN BASIS NO Spring 1/8" out of range WFPD-SH-12 DESIGN BASIS NO WFPD-SH-13 DESIGN BASIS NO WFFD-HSS-209 DESIGN BASIS NO WFPD HSS-210 DESIGN BASIS YES (same cale as HSS-212 & HSS-212A) V. '~PD-HSS-211 g '" ^' WFPD-HSS-212 DESIGN BASIS YES Con Reqd for piston seuing Walkdown required WFPD-HSS-212A Note: The loads include the effects of global thermal stratification and postulated locked monoballs. i a a ~
i 6.4 WATER HAMMER DlSNSSION A Ptudy was-performed to investigate magnitude of loads to cause_ plastic deformation by first determining the j yield stress of the piping under. consideration. The ^ yield stress of the piping material A-106 Grade-B is 35 ksi per ANSI B31.1 1967. Plastic deformation occurs when this ~ threshold value-c* stress is exceeded. The corresponding moment for m straight run of 16 inch-OD, 0.843_ wall' thickness pipe is the yield stress multiplied by the section modulus (35000 psi times 170.6 inches cubed-which equals approximately 6 million in-lbs). The short radius elbows are associated with a-stress intensification factor (stress riser). This factor is-3 based on: the elbow geometry and is equal to 2.36 i t (dimensionless). This reduces the corresponding plastic moments from 6 million in-lbs to 2.5 million in-lbs. Water hammer is the only event that could be identified which could generate loadings of this magnitude. 6.4.1 STEAM GENERATOR WATER HAMMER i The Lawrence'Livermore National Laboratory Report (Ref. 9.13] was reviewed to determine the magnitude of loads which could be expected for a steam generator bubble collapse water hammer event. A detailed finite rilement analysis program (NUPIPE-F'f) (Ref. 9.45) was-used to determine the movement q , d i r e c t i o n o f the feedwater line during a water hammer event. This information leads to the
- i conclusion that a _ ' steam generator water hammer
_l ,could_plastically deform the piping into the as-found shape in the beginning of 7R. m 6.4.1.1 LAWRENCE LIVERMO"P. NATIONAL LABORATORY REPORT - STEAM GI J.RATOR WATER HAMMER The Lawrence-Livermore National Laboratory -(LLNL) Report. (Ref. 9.13] documents the results of a ' generic. investigation for the ef fect of hydraulic shock or water hammer on feedwater piping of pressurized-water reactors. The most severe water hammer event studied by LLNL 1 ~ occurred-at ' Indian Point Unit 2 Nuclear i o Power Station - where -the water hammer caused local deformation of the feedwater 1 _ pipe near the stehm generator, and 3 failure of the pipe at the containment wall. Evidence indicated that the water hammer was caused by the formation and subsequent sudden collapse of = a steam 64 1 w.-
bubble in the feedwater line near its ~
- inlet-to the steam generator.
The. loads generated by similar steam ~ generator steam-bubble collapse events were postulated to be in excess of 100,000 lbs by LLNL. -6.4.1.2 STATIC TIME HISTORY ANALYSIS Three NUPIPE-SW static analyses were performed to determine the. direction of displacement of loop A and C feedwater piping. Two analyses were performed on loop A and one on loop C. ~ The analyses of loop A considered the monoballs locked and unlocked. Loop A considered two raodels in order to reflect the results of Section 6.1 (postulated-locked monoballs). It is considered possible to free the monoballs with large water hammer loads. A. ten kip load was statically applied at selected points along the p! ping (see Figures. 6. 4.1. 2-1 and 6.4.1.2-2 for loops A and C respectively). The loading was applied in the direction of the steam generator. This -is consistent with a water hammer event originating in the steam. enerator. The g results of the ten kip load case can be easily increased to obtain results which correspond to the magnitude of loads in the LLNL Report discussed in Section - 6.4.1.1. Loop A analyses indicate that the resulting displacements plotted by NUDRAW-SW are -both up and in towards-the reactor'for-monobal'Is ' restrained or, unrestrained (see-Figures
- 6. 4.1. 2-3 and
.6.4.1.2-4 .for unrestrained monoballs and 4 L 6 '. 4.1. 2-5 and 6.4.1.2-6 'for ' restrained monoba11s). The. loop 'C-displacement ~ profile (Figures 6.4.1.2-7-and 6.4.1.2-8) also show the loop C piping displacement up ; and in.towards the i reactor. The maximum -calculated stress -(including stress intensification f actors)' for the 10 kip load on loop A.was ' 3532 psi for unrestrained monoball case and 3525 psi for the restrained case.- Theistress en loop C was 6581' psi. It is obvious-f rom-these stress values-that a 100 kip or greater load:could result in stress values which would exceed our yield stress value of 35000 psi. Therefore it 65
- I
- j l13 j!!Il
\\,lli
- i o o >.,i s,,c. *. l,.,, -
.e s s , sr ,..,. a l .Q; s. s.,,p,~c, a. ; '.. :,. e. s 3 u ,4 s r u ': f,c L A_ 7 a.%';rp,3, .'s ~k ~ ~, J,, * ~ r ,f', ~ d a -.u' o t&. p, , _. A ~j. , - [', ' l .c L._ r
- s I s e
n ,o.e. *w. m..**'h 'r. m ^, l' -v ( e ~rf a a h 8 r s I*! 4 e n,. s .s - u' . r . y I' A t z s a 9 s y e l P w u, m.f N . h'., # ' "N=- ,c O
- r D
p .~ i L k m ~, / w m. t1 \\ 0 r, S s 1 / y,,, '3 ' E
- 3 D
=." A 3 c 2
- /, W_W 1)e D
) i s 3 r t t I s e 1 y a 5 p T Ng t U g 4 li s dk - h P N e 1 8 I g t p " 2 _ R f -. E a 1 M M &. & 'y,., a m. '*.""., r n4 A g __7 = * ' } _o H o t 6 i t V<- R E E g a R. T J.- c U A o a q. V l s .I F. C 9 -s d / n I n s T o i. vQ'y ' a c. A o T f n 2 S o 1 i F t o O c r e S a c r N w i 'n O d' I a T t u A C p D-n I i .u N s / e x f t a c .n 7," i 4 d' C R'). t' n' fo I d' ;. a
- a. y*
= . s )4 u
- 1. -
w e s (r llll:
..; a IDCATIONS OF STATIC VATER flAMMER INPUT LOADS - Th0P C 1 .. s. s e v. g. 4 -g r s=. .stg #_ ) ~ s.s m 3 .41 sa ~* ,0 ,.3so a j-0~... g s. 1-q -,-~~~- ~ ' .s. -/..sn 4o i < ~ .r 'I'. ,..,,N., jj >- m ' ,e n e,=. / 3, . u. 1 5 '"* * * *.. nk ,,,,.,, / '... % .z. / r/ a.rio. s.. 1 .'. t... r, a t 4-<......~..^ '* ., /,- c... : ,, foi - . r ** !
- d..
,,,yn t osteEW'W s*** "s7 ' f5. igo u s:sto** 5" ,, y.y... g' l.. o J 4 6 v g /~;. 6, - .y u, =.., - C w u..... e.- .,3p' G.g e u s s ...-.i, m pwy r 3.. u ' ,..A,....,,( ..a 3 ; ; a; n..... ..u,_ a.- t ~ -n -AI/. .. J ..j;g(....... _ e. __ t... 4 --- - Indicates liiput direction and location of the static 10 kip water hammer load. FIGURE 6.4.1 2-2 m 1
l! I,l R x G E D 0 0 O l ,0 7 I O R A P O D I SL Cg LA } BONO o M D E L':: N IAR 3. T S 2 ER 1 N U
- 4.,
S 6 TL E U f} R SE G R I F TNEM EC A IP S I D LA T NOZ I ROH 0 9 llllll!
VERTICAL DISPLACEMENT RESULTS UNRESTRAINED HONOBALLS - IMOP A o- ~~ JS ~_40, 10 ~ ~ ' ' 2 s / V I Ell. P0 l tli X : 1.0 Y 0 Z : 0.0 RU ~ 0-0 DEGR
- )5 "15 Y
W 1 \\w FIGURE 6.4.1.2-4
i! ,li!iIIll!-',ill4( R X G I E N 0 0 0. D I O 0 01 0 f P l 0 u E X Y Z I v I O R A P OD rb, I x SL = L A S BON O M D EN I 3 A RT 2 S 1 E R 4, S 6 TL E U R S U E G ~ R _~ IF TN E M O-E N CA L P S I 5" D LATNOZ 0 I R0 9 3' 5 1 ) ( 8 ll lll!
R G T L J 000 U r I O 00 0 1 Y, P 0 W E X Y I I V 0 Z. R A P OO L 0 A-s LLA 5 B l 0 o N O M D EN 6 IA 2 R T 1 SE 4 S R ' s. 6 S L T E R 4 4 U U s S G E I R 0' F T 3' N E M E C A L P S I D LA C I T RE V 5) aw
.ii1l a ( R ) G T E J D G 000 r l 01 0 0 P 0 Z O P D 0 I i l E X Y Z I S V I L O LA R B0NOM D E N I A 7 R T 0 ?. SE ix 1 h R N 4 U .6 ST E L 1 R U 6' U SE 6 n I R F -)/'. TN
- 5:'
E .~ M o '* E ~ ] -[ CA J Z-Q'.- lP S I D LA T NO Z IR 0 l 9 '. I ?',f .;*7*.'.'.' ) f 4 l IIlI1 1
~ s VERTICAL DISP 1ACEMENT RESULTS UNRESTRAINED MON 0 BALLS - N vlEu POINT .40 -,3 7 X : -1 6 g y.. Y - 0.0 '~ ^f? d ,_g 3 aor - 0 0 DEGR ,N[ db Y c EOd' .Lx 6 FIGURE 6.4 1 2-0
) a 4 is concluded that a steam senerator bubble collapse event could generate loads of a significant magnitude to plastically deform the piping into the irregular deformation pattern observed at the beginning of 7R. 6.4.2 FEEDWATER REGULATING VALVE FLUID TRANSIENT EVENTS The main feedwater regulating valves (FRVs) at Beaver Valley Power Station-Unit i have been particularly prone to cavitation and resonant vibration problems. This current problem by itself can not generate loadings of 50 kips (See Section 5.3.2). Fifty kips was identified in Section 6.4.1.1 as the load which could begin to plastically deform the loop C piping. Based on review of the snubber design (1 1/2 inch and 2 1/2 inch bore) (Ref. 9.48) if significant loads (assume 15 kips is significant) were developed by this transient event which occurs only during specific f power
- levels, inside containment pipe supports would have been damaged.
Since the piping is much more durable than the pipe supports (the pipe can withstand almost 50 kips before yielding) and no pipe supports have been
- damaged, the current cavitation / resonant vibration problem by itself could not p]3stically deform the piping inside containment.
This is also supported by the recorded data of Reference 9.41 where only local vibration and acceleration of the valve were reported. Additionally damage due to this type of water hammer would have been observed outside containment, consistent with these events in the past (See Section 5.3.2.1). 6.5 LOCAL THERMAL STRATIFICATION Local thermal stratification in the vicinity of the feedwater. nozzle was studied by Westinghouse during their investigation of NRCB 79-13 " Cracking in PWR Feedwater Lines" (Ref. 9.32). Results from the instrumentation program described in Section 4.5 identified rapid temperature cycling between 557' F (steam generator hot stand-by temperature) and 80* F (the temperature of the fluid drawn from the condenser hotwell during start up). The following discussion examines the recorded plant data and correlates it to the cracking discovered at the beginning of 7R (see Figure 6.5-1 for recorded plant data). Note these data have been modified per the direction of Reference 9.41. 3 l 74 j
a .g N JX l _q s (> Q> ( t \\ _e { \\ N Y g 3 -3 A // ~ l _N g a n i y 23 i ce ,2 .E m e w e N [ g (
- 2 E
\\ t _e I f Np-N Nh 5 X3 ? 3 h W -r 0 - F _q e Q. n 'N N <f s ( p 8 8 ,8 8 8-8 n n e-o [J.6ep] e;ngo;odwo,t 75
e 1s Spurious-- temperature readings of less than 80* F have-been. deleted. 6.5.1 CORRELATION TO 7R PLANT DATA Local thermal stratification was only considered to be a -possible contributor to the cracking discovered on the loop A steam generator nozzle taper transition. No possible mechanical link could-be identified between the local-stratification and the irregular pipe deformation pcttern. Since the nozzle cracking was identified 'as fatigue cracking by metallurgical examination (Ref. 9.49), the data recorded'during the post 7R plant restart were reduced into meaningful fatigue data. The reduction of these data is described in the following section of this report. 6.5.1.1 CYCLE COUNTING The plant data recorded during the post 7R restart _ of the. Beaver Valley Power Station-Unit 1 were reduced. The thermocouple data were graphed for all twelve thermocouples (see q Section 4.5 for_ instrumentation details).-. The -number-of cycles, magnitude and mode number of operation were recorded. A data sort - was performed and the result-of the sort is shown -in Table 6.5.1.1-1. This table indicates that thermocouple number-1 (the thermocouple on top of the pipe, almost on top of the-taper transition) experiences the most severe thermal' cycling. This is confirmed by metallurgical. examination (Ref. 9.49) which-determined that the largest' cracks occurred at the top inside surface of-the pipe and were-- due to corrosion assisted' fatigue cracking,: 6.5.1.2 CORRELATION TO OPERATION WHEN SIGNIFICANT CYCLING OCCURRED The data were further - reduced ' by intensity, mode, and plant' event'for thermocouple number The results.of this-review are summarized' in Table 6.5.1.2-1 for the 7R recorded data, a significant cycles were observed to be present in Modes 1,'2 and 3 (reactor at 29% power to hot standby). during' normal heatup and post-7' reactor trip /AFW injection heatup. Significant cycles were not observed. during the-manual cooldowns on January 19, 1990 and October 6, 1990. More specific correlation to 76 4 4
TABLE 6.!L1.1-1
SUMMARY
OF T/C RANGE CYCLES FOR T/Cs 1 THROUGH 6 1 NUMBER OF TEMPERATURE CYCLES FOR TEMPERATURE RANGE (F) T/C No. 477 400 300 200 100 50 1 7.5 ' 22 18 '39 100 86 i 2 7 29 53 107 79 64 1 3 10= 23 66 90 128 46 l 4 13 22 34 53 59 35 i 5 4 14 15 31 19 14 -6 13 35 68 117 111 279-j - Note:-1)'477 F transients include all' transients greater thansor equal ~ to 400 F. -2)-400 F transients-include all transients. greater than or equal to 300 F but less than 400lf.
- 13) 300,F transients' include all_ transients greater than or equal-to=200;F but less than 300 F.
L 4):200'F transients include all' transients greater than or equal L ~ to:1001F but less than 200 F. l 5)h100 F transients include all transients greater than or equal to 50 F but-less than 100 F. M-
- 6) ISO F transients include all' transients greater than or equal 11 to 20 Frbut lessethan 50 F.
l 'i I 9 ) 24-77
g TABLE 6.5.1.21
SUMMARY
OF CYCLES BY INTENSITY, MODE, AND PLANT EVENT FOR T/C NO.1 i i 477 F INTENSITY NUMBER OF CYCLES PLANT-EVENT MODE 1 MODE 2 MODE 3 MODE 4 MODE 5 TOTAL l
6 0
0 33 HEAT ~UP 19-8 REACTOR TRIP 21 19 2 0 0 42 400 F INTENSITY j i i NUMBER OF CYCLES 3 PLANT ~ EVENT MODE 1 MODE 2 MODE 3 MODE 4 MODE 5 TOTAL ) 1
5 0
0 14 HEAT UP 8 1 REACTOR TRIP-8- 0 0 0 0 8 - p i
- 300 F INTENSITY
'f I NUMBER OF CYCLES MODE 2 MODE 3 MODE 4 MODE 5 TOTAL PLANT EVENT MODE..1 HEAT UP-8 1 5 3 0 17 REACTOR TRIP-1 0 0 0 0 1 a 78
,_.m.- ... _ _. ~... ?,' I ' ' -,1 - ( - TABLE 6.5.1.21 (cont.)
SUMMARY
OF CYCLES BY INTENSITY, MODE, AND PLANT EVENT FOR T/C No.1 ' i, a e 200 F INTENSITY s.- = NUMBER OF CYCLES r y PLANT EVENT MODE 1 MODE 2 MODE.3 MODE 4 MODE 5 TOTAL HEAT:UP- -13 1 16 5 - 0 35 1 j REACTOR-TRIP-0- 0 0 0-0 0 q y 4 =100 F-INTENSITY-1
- ~
NUMBER'OF CYCLES. PLANT EVENT-MODEl-MODE 2 -MODE'3 MODE'4 MODE 5 TOTAL i HEAT-UP~ 17-17~ - 35 6 1 76 -i REACTOR TRIP 'l -6 'O O 7 -50 F INTENSITY. s, e
- NUMBER OF-CYCLES.
MODE. MODE-5 TOTAL ..____4 W -PLANT EVENT- - MODE: 1 MODE 2 MODE 3. n HEAT UP -~ 13 1 13 5 10 42 -i i 0-12 0 -' O 12 REACTOR;T.; P..0 r:- 4
- .,.r L
n .[ z 79 r l! If r e n,- .,.in. n ~., -,-..v e- ~., - w
TABLE 6.5.1.21 (cont.)
SUMMARY
OF CYCLES BY INTENSITY, MODE, AND PLANT EVENT FOR T/C No.1 Mode descriptions are provided below: Mode Descriotion 1 Reactor-critical at greater than 5%. power 2 Reactor critical at less than 5% power 3 Reactor coolant loop at 547 F (Hot Standby) 4 Reactor coolant loop temperature is greater than 200 F but it is less than-350.F (Hot Shutdown) 5: Reactor coolant loop temperature is less than 200 F (Cold Shutdown) 6 Refueling t l. lll 1 s e l 80 4
4 i l actual plant operation could not be derived from existing plant operating logs and other recorded-plant data (e.g., steam generator level was observed to rise, fall, or remain constant during the most significant local thermal stratification cycles). Plug
- flow, trickle flow, AFW injection and heatup have been identified by references 9.16, 9.17, 9.18 and 9.19 as potential sources for the significant cycling phenomenon.
6.5.2 THERMAL STRATIFICATION EVENTS During the evaluation of local stratification it became apparent that the design number of thermal stratification events were much greater than the actual number of events experienced by the plant. Table 6.5.1.2-1 identifies the two plant events (heatup and reactor trip) that were associated with the local thermal stratification phenomenon. The count of plant design events is presented in Section 6.5.2.1 and the number of actval plant events is presented in Section 6.5.2.2. 6.5.2.1 REVIEW OF SYSTEM STANDARD DESIGN CRITERIA 1.3 REVISION 2 Westinghouse Corporation, the nuclear power steam system supplier, has issued several documents concerning plant design transients. The most appropriate design transient document for Beaver Valley Power Station-Unit 1 was determined to be the System Standard Desi.gn Criteria (SSDC) 1.3 Revision 2. This SSDC is the closest to the vintage of Beaver Valley Power Station-Unit 1 and is the first design transient document issued with secondary side information. Section 6.5.2 identified the two plant events associated with significant local stratification (heatup and reactor-trip). SSDC 1.3 Revision 2, was reviewed to determine the number of design transients associated with those identified events (i.e., the. number of heatups and reactor trips). There are 200 postulated heatups and 760 postulated reactor trips that are applicable to the station. See Table 6.5.2.1-1 for a listing of these events based on a 40 year plant life. e 81 l
e - ] TABLE 6.5.2.1 1
SUMMARY
OF SSDC 1.3, REVISION 2 TRANSIENTS (Reactor Trip And Heatup Only) l REACTOR TRIP C1.QLEE HEATUP CYLCES II.1-Loss of load 80 I.1 Heatup 200 II.2-Loss of power 40 II.3 Partial loss of ~ flow. 80 II.4. Reactor trip 400 ' II.5 RCS depressuri-zation 20-1 ' II.7 Control rod drop 80 II.8 Safety injection _.fQ j '760 (Total) All-case.II, III, and-IV events result in a reactor trip. Case ' II.6 (loop out-of sevice) was excluded because it is not applicable to the Beaver Valley Unit i design basis. Case.I - Normal- - Case II Upset Case.III Emergency Case IV - Faulted-s e 82
- 4 6.5.2.'2 ACTUAL NUMBER-OF EVENTS SINCE 1979 The actual number of plant-heatups that have-taken place between 1979 and 1988 is 20.
The actual nunber. of reactor trips that have occurred between 1979 and 1988 is 30. Prorating these -values for the remaining 37 years of life since 1979 results in'37/9 x:30 123 reactor trips. Additionally the = prorated number of events for a-forty year ' plant life would be 102 heatups ( 37/9 x 20 + 20 heatups _ before 1979) and 156 (123 +-33 3 reactor trips before 1979) reactor trips. A { heatup-event is considered to occur if significant time was spent in Mode 3. It is 'noted that the actual number of expected cycles is much less than the' postulated design 1 events (see Table 6.5.2.1-1). 6.5.3 ESTIMATE OF_ CYCLES (AND CALENDAR TIME) TO INDUCE CRACKING The estimate of cycles to inaece cracking was_ developed by the following analytical m eps: The one' dimensional heat transfer program' HTLOAD c '{Ref
- 9. 60) ' was utilized to generate a family of
-outside Wall _ temperature versus time profiles for 9 an up transient-and.a down transient (see Figures . and. 6.5.3-2). These profiles are '6.5.3-1 conservative because the act'ual profile.is probably a ramp. The up transieat was_ input.as a1 step from 80' F.to 557'1F. The down transient was also input as'n step with a-temperature range from-557' F-to 80' F. Velocity and the' corresponding'co'efficienti -of heat, transfer were' varied based.on a constant-but arbitrary flow rate of 1000'gpm and-changing-the flowing portion of the pipe (hydraulic radius). The! flow. rate of 1000 gpm was selected because it .is _ approximately ' 10% of-the design flow rate for L thec line. and would therefore allow. thermal L stratification to occur. 'The full intensity (477' B F)' up transients and down transients were reviewed L to determine-the representative up transient an_d down transient. A best fit-approach'was then used-. l I to select the most. appropriate up and down 'i transient -profile from ' the HTLOAD results (i.e.,. HTLOAD outside ' wall temperatures). Lesser transients considered _only a fraction of the full i* range profile 'results (i.e., 400' F intensity ranges consider 84% (400/477) of - - the full range case). Westinghouse analysis results correlate-83
UP TRANSIENT - STEP FUNCTION 600 557 T E-M P E A- 'T U R E (F) 80 O TIME (sec) 5 Figure-6.5.3-1 DOWN TRANSIENT - STEP FUNCTION-t -600 557 i .T, E. M. P E' R A L .T U' -R E' (F) 80 I 0-TIME-(sec) 5 Figure 6.5.3-2 84
a 7. .well with the best fit curves selected. The maximum coef ficient of heat transfer selected by Westinghouse.at the fluid interface was 2000 -Btu /hr-ft* 'F [Ref. 9.17). This analysis selected 2028 Btu /hr-fta ep, An analysis was then performed to determine the calculated cumulative : usage factor (CUF) for the two plant events identified. Subparagraph NB-3650 1989 was used to provide i from ASME Section III guidance for the analytical methods utilized. The up transient peak stress was determined to be i 101,000 psi, while the down transient peak stress was determined to be 74,000 - psi for the full-potential transients (i.e., 80' F to 557' F and 557' F to 80' F respectively). Stratification of the steam generator nozzle also generates a global i bending moment-range. This range cycles along with the local thermal effects above and this peak stress range was calculated to 5000 psi. The total peak stress-range is approximately 180,000 psi. The CUF was determined to be 0.0596 for one. plant heatup and 0.0623 for one reactor trip. These values are based solely on the 1989 post 7R restart-data. Whea.these values are. extrapolated they indicate that approximately 16 plant events (i.e. plant events-are defined here as heatups or reactor trips) can occur before cracking is predicted .9968). When the cumulative usage (. 0623 x '16 = ~ factor exceeds 1.0, cracking is considered to have been initiated in accordance with design basis criteria. Fourteen plant events have occurred 4 since-the taper. transitions (elbow)-on loops B and C' adjacent - to the steam generator.were replaced. The loop B and C elbowsLwere last replaced in 1988-while the loop A elbow-was replaced in 1989. These 14 events' occurred over 2 1/2 years. The plant has also experienced ~an average of-over 5 events per year since 1979 (30 reactor trips + 20 heatups / 9 years). The operating period between 1976 and 1979 ~ has been excluded from the average events-per year calculation because this was the unit's first fuel cycle. The 1976 to 1979 time period is. not indicative of how - the plant is operated today. -Since the average number of events is 5.5 16/5.5 is approximately 3. years. Therefore the taper u, l transition is considered to be crack-free until the summer of 1991. This roughly corresponds to operation between 8R and 9R. Note that cracking L was discovered at the taper transition in 1979 and in 1988 after about 50 events. This would equate 85 l-
~ .~ r .y - toia CUF-of about 3.- This may_be-a realistic-value o foricrack initiation.=~The postulation of cracking t atia component'with a CUF of 1~is consistent with-E- 'ASME Section III. 6.5.4-SIGNIFICANCE OF LOCAL THERMAL STRATIFICATION VERSUS PIPE GLOBAL STRATIFICATION TO INDUCE m CRACKING-The ' global bending : moments _ generated by global thermal) stratification ' are considered to be self - e .Llimiting-for-the, pipe geometry and support I locations at-Beaver -Valley Power : Station-Unit -1 cased on. _ extensive analysis. The global-s stratification bending : moments: are a' function of pipe geometry, support location,: and.. top;to bottom differential thermal: growth due to: topL to bottomt ? 'differantialitemperature. The stress generated by' lgicbal-thermal stratification ~ wasfound to be S .relatively insensitive to. support function changest for _ thesel ' specific -piping configurations.. The4 ~ thermal-stratification' Class-2< stress'was basically; -unaffected andialways less-than-_15 kai intensified'. for'the current pipe-support scheme considering the. 1
- deletion of-one or more supports.
Since-this value= d ~ - is welly below the' yield stres's, limit. (35 ksi) ~ and; c the'significant global. stratification cycles-were-determined to be infrequent' and - time E independent w thermal stratification, when Ecompared' to-local globalithermal stratification was determined not to.
- \\
1'.
- significantly contribute} to the cracking-at
-feedwater nozzleitaper~ transition. E6.6LINSTALLATIONI to the irregularc_pipa JAnothero possible;: contributor 'No
- deformation : pattern that was considered Lwas la lack of -
- -installation control procedures.
LCracking was first identified-on' the. taper transition to the steam: generator nozzle in E1979.at/ Beaver. Valley ' Power Station-Unit: 1. t e, 4 Alli three - loop elbows adjacent : to the' steam generator.: u 4 -were replaced - twice : prior; to : the ' discovery of. ' the y firregul_ar. pipe.deformati'on pattern,, once 'in 19791andionce : in 1988.
- Since'no.~ installation controlling procedures-
,y were. provided to, construction to recheck: rupture l t Qf t restraint. gaps 'and support locations, fit is possible that some or~all of-the! misalignment'of:theifeedwater' lines ? occurred during the elbow replacement-tasks. A NUPIPE-SWn ~ t So' computer- - analysis was performed to quantify the pipe t a.. movement during Linstallatica and' is - discussed : in the ~ following"section. , ; FJ 86 li.: ..,z,.. ~.t
,m 4-e e
0 x O C + 7 q O m C Z [ O Z d 3, E e u e c : oeo" O C O Z ) i a a a O' a rO-C3 0 2 O u = ,2 x e a O ~ n n_ i > o ~ m h b mee g_ [x w N Y N O I h, 7h( < y i N 1 mb O ex; 'h 2 0 5 E . o = M M b s th M O N Q Z Em c S g.!... ...!,,,... =. L s W e
- ,. N.9 z :2 a
. ' =.. = s 't', ~ L C F g r\\ to e,' ' e* x- /1* 't { ~ y 1 /9 p U ? s / r a e c = I W w a o e cc m W u')- 7 a LC C M (M C v_ __ ~3 3 T g' W es-c E >== t C L s 87
.~ i 5 4 I 3 e O 2 O X O J$~h O, v
- c. 'o
-j - g = en Q E cc g , c = = *., wL c = w o o a 2,c o o - oZ
- d e E' J LM e.
O D O O c c z = g co o m A s' o W C W I - r p =0 W ts y c 3 g c e p w =J O C E Z - 2 L a c 2 n 3 JO O 3 % O o w E c y*
- ~
ag E5' he: 5 5r 8 3 w; E y-c.< -l -Q y 3 o y s xx e. x o g. O g $44 eu e-c w h s o 4 k &Nb c o g. a e Oo g a lg 5 G u ( h 'e ~ W .\\ 7 p to .e,,. a co. Z / e - M 4 _O b m 0 Q a '4m e e $ *0 / W o -O w H E s M ~* C w O' 3 g h a o O ga) 2 Q Q W e M L A N C h N N .s W 88
1 k r O e .i~ L7 x X J O 2 o 4 = n w w = 2: e C o C y.O O 0., 2OOO 0 C O 2 + i o o L7 f.h L7 C -O O N i g V O. O CC C L7 o C::: 2 n 1 y c H i 11 O C Z LO 47 U E h 4 L x >- N y w =ne* o xlgs e = o m 5 2 w N 0. s o y g c: 4' sg -._. 3. <i. Pr.-+ l b / U r-5 r Q a is: o 5 w ~ g N i N g O to 2 l 's c +,. M C +==* .C s.
- r.
tr.' ~4 =c .ba - g' W 6 [ Q .e = w w e, N- .k d z h 's- $ 4 = . h} .},C in. m ~ G'. a: 9. \\'. ' r. Q e - c. O \\,- l LC a W \\ 'n, s e C D x <, *. -o.- = c m. ' r., / b ..,e, ' ' e. *,,,,,. 'sl 7 /
- *-m l
. =" 8.am i
- v#
I l 3 ~ 89 l [ L.
h ?g# 6.6.1 DISPLACEMENT MODEL ,[ Loop A and C models were developed in order to f,/ analyze the effects of reducing the steam generator nozzle 1e'ngth by one inch. The one inch dimension was arbitrary, but was also considered to be within the reasonable bounds of nozzle length lost due to wald preparation / shrinkage. A displacement of one inch in the direction of tne steam generator nozzle was input in both loopa A and C. See Figure 6.6.1-1 for loop A monoballs free. Figure 6.6.1-2 for loop A monoballs locked up OH Figure 6.6.1-3 for loop C (monoballs unlocked) for the horizontal displacement results which were plotted by NUDRAM-SW (Ref.-9.51). A ucmmary of the cor~'-tion with the as-found pipe position is provid." 6.6.2. n Section p 6.6.2-CORRELATION WITH AS-FOUND DISPLACEMENT POSITIO The as-found position of bon loops A and C and in towards the reactor on the annulus portion was up of piping in the vicinity of the steam generator. t The figures in Section 6.6.1 show that both loops A and C are drawn towards the reactor when the steam generator nozzle length is reduced. No change in elevation is identified by the figures for the input horizontal displacement. However, if the weld preparation on the elbow.under the-replaced elbow was cut long, the pipe would shift upward. p T I 90
O
7.0 CONCLUSION
S + The root cause for the main feodwater
- lint, irregular deformation pattern and taper transition cracking have been identified and are discussed in this section of the rrr.rt.
In addition to ron can.se determination for the two ing problems identified above. the potential for gloca;
- mal stratification and for significant future water hah e
identified and are also presented here. Design modifications are required to prevent reoccurrences of the irregular deformation pattern and the steam generator nozzle taper t ansition cracking. 7.1 PIPT. IRREGULAR ALIGNMENT All of the potential root causes were evaluated and are categorized as factors and nonfactors. They are both presanted below: 7.1.1 ROOT CAUSE FACTORS ,re root cause of the irregular deformation pattern has bosn determined to be most likely the result of pipe misalignment due to the 1979 and 1988 elbow replacements. 7n addition to the effects from the elbow replacements some of the irregular pipe deformation pattern on loop A could be attributed to the 1981 water hammer event. The effects of the two types of events are cumulative. Each type of event is discussed below: 7.1.1.1 UNCORRECTED ALIGNMENT FROM THE 1981 WATER HAMME't EVENT In 1981 the linkage on the loop A main fcedwater regulating valve (FRV) failed. This failure allowed the plug to close rapidly and caused a water hammer event that failed two pipe supports inside containment and small branch lines of the main feedwater system. Subsequent to this event on 05/06/81, EM No. 60322 was issued identifying rupture restraint shim stacks on loop A that were dislodged and which could not be returned to their design locations because of a shift in the clearance around the pipe. This relocation may have been caused by the 1979 elbow replacement or this water hammer event. 91
e The piping was not relocated and a portion of the shim stacks were installed creating a small gap on the upper inboard to the reactor side between the feedvater line and the rupture restraint. Most transient events experienced by Beaver Valley Power Station-Unit 1 per the Licensee Event Report (LER) listing are not of a significant magnitude such that an irregular pipe deformation pattern would be expected to occur. 7.1.1.2 PIPING MISALIGNMENT DUE TO THE 1979 AND 1988 ELBOW REPLACEMENT The piping which may have been relocated by the water hammer loads was not zero gapped against the rupture restraint on loop A after the last significant water hammer event had been recorded-in 1981 in the LERs. Only the installation could have relocated the piping from 3/64 inch gap (EM tio. 60322) to zero gap after the 1981 FRV linkage failure induced water hammer on loop A. Only .olping misalignment due to the 1979 and 1980 lbow replacement could be responsible for the irregular deformation pattern on loop C. The other possible root causes were determined to be nonfactors and are summarized below. 7.1.? NONFACTORS Local stratification as the name implies is only a local phenomenon and could not permanently relocate the piping. Global stratification class 2 stress levels with or without monoballs locked were determined to be well below the yield stress limit. Therefore global stratification could not have contributed to the irregular deformation pattern. Since no pipe support damage was reported, steam generator water hammer and feedwater regulating valve resonance / vibration were also eliminated as possible sources of piping irregular deformation j pattern (pipe support damage would be expected to occur before pipe yielding could begin). 7.2 STEAM GENERATOR NOZZLE TAPER TRANSITION CRACKING All of the potential root causes were evaluated and are categorized as either factors cr nonfactors. Factors and nonfactors are presented below: 92
j' 7.2.1 ROOT CAUSE FACTORS The root cause of the steam generator nozzle taper transition cracking has been determined to be local thermal stratification. Rapid fluctuations of the through wall pipe temperature during Modes 1, 2 and 3 (See Appendix C) are associated with significant fatigue of the feedwater nozzle taper transition. The effects of the rapid through wall temperature changes were quantified and determined to be consistent with the issues identified in NRC IE Bulletin 79-13 (Ref 9.12). The analytical results indicate that cracking on the steam generator nozzle ti,>er transition could begin to occur after about 1( events. One event is defined here as one heatup from the refueling mode or one reactor trip (See Appendix C). This corresponds to a time period of concern beginning during the rummer of 1991. (see Section 6.5 for the detailed oiscussion of analysis results). 7.2.2 NONTACTORS Several different feedwater system transients were identified as possible root causes er contributors to the taper transition cracking. The following list of transients were determined not to significantly contribute to the cyclic fatigue cracking identified by metal 3urgicsl examination: a. Global stratification was eliminated because of the small number of events recorded during the post 7R heatup, the apparent independent nature of the global and local stratification cycling and the relatively low magnitude of stress levels generated at the taper transition (see Section 6.5.4). b. Main feedwater regulating valve (FRV) instability alone was eliminated because the magnitude of 'oads generated due to this event is below the elastic limit of the pipe. Although resonance has been identified on ?he A loop at 66% power (Ref. 9.35), instrumentation has confirmed that large w 'ssure oscillations and the resulting large forcing functicas are not being generated into the piping system. For the water 93
hammer event to occur, the resonance must f cause a mechanical f ailure (i.e., linkage ) at the valve as was the case in the 1981 event. Water hammer due to the failure of the FRV is a separate issue j and has only occurred once on loop A at Beaver Valley Power Station-Unit 1 since the last replacement Of the elbow and was, therefore, also eliminate". c. Steam generator water hammer was eliminated because no evidence of a steam generator water hammer was identified after the replacement of the elbows in 1979. d. Installation, which was identified as a contributor to the irregular pipe deformation pattern, is a one-time event and, therefore, could not contribute to cyclic fatigue. 7.3 GLOBAL STRATIFICATION Global thermal stratification of the loop A feedwater line at Beaver Valley Power Station-Unit I was identified during the reduction of post 7R heatup data. Based on the collected data, global thernal stratification causes the feeduater line to deflect with a larger magnitude and in the opposite direction of thc. design basis e talysis. This phenomenon occurs on all three feedwater lines at the station. Analysis of all three lines for the global stratification phenomenon indicated that the piping met the ANSI B31.1. 19 67 design basis equation allowables. Certain pipe supports when considering global thermal stratification and locked monoballs, could not meet their original design besis equation requirements but have been shown to meet alternate criteria (see Table 6.3-4). Global stratification could also account for the dislodged shim stacks at restraint FWR-1. 7.4 POTENTIAL FOR FUTURE WATER HAMMER During the investigation of the irregular pipe deformation pattern it became apparent based on review of LER's and the lack of support damage that although Beaver. Valley Power Station-Unit I had not recently experienced l a water hammer initiated at the steam generator other l PWRs of a similar vintage and design had. Summarized below are four items of concern that are directly related to the potential for future steam generator water hammer at Beaver Valley Power Station-Unit 1. 94 i l
i-i 7.4.1 SIMILARITY TO TROJAN NUCLEAR POWER STATION Trojan Nuclear Power Station is a four loop Westinghouse PWR designed and constructed during the sanie time period that Beaver Valley Power Station-Unit 1 was designed and constructed. 3 Specific feedwater system similarities between Beaver Valley Power Station-Unit 1 and the Trojan Nuclear Power Station includet a. Both have J-tubes installed on the feedring. b. Both initiate auxiliary feedwater (AFW) immediately after a reactor trip. c. Both have short horizontal lengths between the loop seal and the steam generator nozzle. Prior to the Trojan steam generator water hammer the two units also provided maximum ATW flow to the steam generators and had slip fit thermal sleeves. For Beaver Valley Power Station Unit-1 the maximum auxiliary feedwater flow to each steam generator is 460 gpm. Trojan has since limited the maximum AFW flow rate, has repaired the thermal sleeve olip fit and has replaced its eroded J-tubes. Trojan experienced a severe steam generator water hammer event prior to its seventh refueling outage (Ref. 9.36). During the refueling outage several pipe supports were found to be damaged and the piping was f ound to be relocated. Subsequent te the discovery an analysis was performed. Tnis analysis estimated that the maximum forces l generated by this event was 45 kips. The 45 kip j load results from a 22 cut-ic inch steam void tnat developed in the f eedring. The void in the f eedring formed because of crosion in the feedring thermal sleeve slip fit connection in the vicinity of the steam generator nozzle. Beaver Valley Power Station-Unit 1 is susceptible to the same thermal sleeve erosion problems that caused this significant but by no means bounding event. As observed at BV-1 following the 29% reactor trip on December 26, 1989, the steam generator level dropped to as-low as 11% narrow range band. Auxiliary feedwater initiated approximately 1 to 2 minutes following the feedwater isolation (FW::) signal. 95
7.4.2 THERMAL SLEEVE EROSION i Erosion of the Westinghouse low toleranca slip fit thermal sleeves have been examined. Localized gaps of up to 1/4 inch have been found on operating steam generator feedring slip fit connections. The void of 22 cubic inches postulated at Trojan could easily develop in itss than 30 seconds with this magnitude of gap. The void would be generated before the auxiliary feedwater system could supply water to the partially uncovered feedring. Since the integrity of the thermal sleeve slip fit is unknown at Beaver Valley Power Station-Unit -1 (Westinghouse has velded this connection in more recent installations), the potential for a water hammer equally or more severe than the event at Trojan is significant and.should be quantified. 7.4.3 J-TUBE EROSION Another potential concern which is sinilar to the erosion of the thermal sleeve is erosion of the J-tubes on the top of the feedring. The J-tubes were installed as a design modification to prevent steam from entering the feedring. These tubes have been found to erode at a faster rate than the corrosion allowance originally anticipated and specified by Wessinghouse. Several eroded J-tubes (the latest - inspection performed during the 7R outage indicates that the worst erosion loss is as great as 50% locally) could result in a signif!; ant water hammer. The verification of.the erosion to the J-tube lends additional credence to the concern that the thermal sleeve is also eroding. The sleeve and J-tubes are composed _of similar material. The ~ integrity of the J-tubes should be verified. The l teedring was originally provided with holes in the bottom.. These holes were plugged when the J-tubes were installed. The integrity of these plugs should be verified to prevent a severe water hammer. s 7.4.4 AUXILIARY FEEDWATER IMPLICATION The auxiliary feedwater supply is provided _ by a safety related source (demineralizer_ water storage tank) which is located in the yard and'is protected from frezzing by space heaters. The rapid injection of cold auxiliary feedwater will collapse a steam void more rapidly than the injection of auxiliary feedwater at a slower rate. Westinghouse Technical 96
c i i Bulletin [Ref. 9.58) and NUREG-0291 (Ref. 9.31) recommended. this. to. minimize the effect of collapsing. any volds that occur in the feedring. l The adoption of a maximum auxiliary feedwater flow rate limit 'is advised in order-to minimize the auxiliary feedwater flow. Reduction of the flow from the existing maximum available flow rate of approximately 460 gpm per steam. generator is considered to be a prudent. water hammer prevention precaution. Any consideration of imposing a maximum upper limit i on auxiliary feedwater injection to minimize water hammer potential. has to consider that minimum auxiliary feedwater flow rates must be provided to l satisfy-safety / accident analysis. i I I t L f B i b-i I-, t e' 97 ...-..:-...-..,,.-.-...-__.__-.-.-....-..-~.-.;....,._
8.0 RECOMMENDED CORRECTIVE ACTIONS Recommended corrective actions to prevent reoccurrence of the irregular pipe deformation pattern and the steam generator nozzle taper transition cracking are presented in this section. In addition corrective actions are also provided to prevent future steam generator water hammer to which Beaver Valley Power Station-Unit 1 is susceptible and to address and qualify p)pe supports under certain conditions (i.e., global stratification and monoball locked conditions). The correctivo actions are categorized as mandatory, strongly recommendtd and recommended. Mandatory corrective actions are the minim un actions-required to meet design basis requirements. Strongly recommended corrective actions are those actions that significantly enhance the reliability of the unit. Recommended actions are those actions which establish baseline information in support of future NRC inquiries and or enhance the reliability of the unit. -Mandatory, strongly recommended and recommsnded corrective actions are presented in Sections 8.1, 8.2 and 8.3 respectively. 8.1 MANDATORY CORRECTIVE ACTIONS l-The mandatory corrective actions involve two of the four concerns discussed above. The irregular deformation pattern and the potentially locked monoball supports require correction. There are no mandatory corrective l actions required for the future possible water hammer or steam generator nozzle cracking concerns. l 8.1.1 IRREGULAR-DEFORMATION PATTERN l The mandatory corrective action for the irregular j pipe deformation pattern is to realign the piping l to its as-designed position. This was accomplished during the unit's 7th refueling outage. l l 8.1.2 GLOBAL STRATIFICATION AND POTENTIALLY LOCKED MONOBALL SUPPORTS 1 The mandatory correctivu actions for global stratification and potentially locked monoball supports are tot a. Confirm analysis assumptions through an enhanced instrumentation program (see Section 8.1.2.1). 1 98
A -b. Revise pipe rupture criteria to be consistent with Mechanical Engineering Branch Technical Position 3-1 (MEB 3-1) (see Section 8.1.2.2). c. Measure the existing elbow wall thickness (see Section 8.1.2.3), d. Modify or replace the monoball supports R-3, R-4 and R-11 which do not meet plant design basis criteria (see Section 8.1.2.4). e. Inspect-supports to verify . input assumptions (see Section 8.1.2.5). f. Perform an interim evaluation, implement OBE seismic li'gits and issue a justification for continued operation (JCO) (see Section 8.1.2.6). g. Revise design basis documentation to include ASME Boiler and Pressure Vessel Code Case H-318 _(see Section 8.1.2.7). These issues are discussed. in detail in the following paragraphs. S.1.2.1 CONFIRM ANALYSIS ASSUMPTION THROUGH INSTRUMENTATION l The loop A and C current design basis analyses which qualify the piping and pipe supports are i based on a scale factor (62%) of the potential l thermal stratification. during auxiliary foedwater injection, post-reactor trip at' 100% . power.- The loop B. piping is based on 100% of the potential. _ 'This is a. conservative approach for this loop and provides long term i qualification _ for the piping components independent of results from future data collection activities. The scale factor of 62% is based.on-reduction of-plant data j recorded from temporaryi instrumentation on i loop A-during one plant' restart. and one reactor trip. The AFW event occurred at 29% l power and these results 'were then extrapolated to the-100%-' power case. -Since these results are. based on only one reactor trip at less than' full power on only one of the two loops -of concern, the recommendation is made to gather additional baseline data. Instruments l~ 99 l \\;
should be placed on loop C. In addition to the loop C instrumentation, loop A instrumentation must also be upgraded to obtain the most accurate thermal stratification profile. The additional instrumentation will provide a more detailed thermal stratification profile for analysis if the current 245' F of thermal stratification is exceeded. See Figures 8.1.2.1-1 (loop A) and 8.1.2.1-2 (loop C) for instrumentation locations, typeu and quantities recommended. 8.1.2.2 REVISE PIPE RUPTURE CRITERIA (MEB 3-1) Pipe stress analysis equations for all design basis criteria were satisfied except for the pipe rupture criteria. The current Beaver Valley Power Station-Unit i design basis pipe rupture criteria is more conservative than the current NRC criteria. Review of the NRC Mechanical Engineering Branch Technical Position 3-1 (MED 3-1) (Ref. 9.6) of Standard Review Plan 3.6.2, the most current NRC
- position, indicates that relief from the conservative Beaver Valley Power Station-Unit i
pipe rupture criteria is available and applicable to Deaver Valley Power Station-Unit 1. Based on the newly identified global stratification condition, it is necessary to adopt the current NRC pipe rupture stress limits in order to show compliance to a long term plant design basis for pipe rupture. 1 Therefore, the latest NRC criteria, MED 3-1, was adopted for the feedwater lines. It is also recommended to adopt the latest criteria for other systems if reanalysis is performed on them. 8.1.2.3 MEASURE EXISTING ELBOW WALL THICKNESS In 1981 vall thinning of the teedwater lines was identified. The manufactured wall thickness per catalog data is.843 inches. However,-the wall thickness may be as low as .750 inches. The current analyses for the unanalyzed condition consider a wall thickness of.843 inches (for the piping and elbows) for the structural analysis and considers local thinning down to.75 inches. Since the loop seal elbows which are not replaced during the addition of the thermal sleeve are at 100
~" .g +. PROPOSED INSTRUMENT PIAN IDOP. A TT7 - EXTING THERMALCOUPLE g P1-P2 L.= EXISTING IJLtIARD i y - PROPOSED THERMOCOUPLE LOCATION-T '/ .PE N T M y - PROPOSED IANTARD LOCATION L f 76 p1 - IDCATE'MID WAT BETVEEN EKISTING L,,y 7 THERMOCUPIE T AND 45* SIDF DOWN T3 P Tpg F2.= IDCATE MID POINT OF PIPE RUN ON T ELEVATION 776'-11"' &Ly2 = IDCATE MID WAT BETWEEN Ty3 PENETRATION AND RISER L. - LOCATE ADJACENT TO FWR-19 .p y THE EXISTING THERMOCOUPLES AND IANYARD - VILL REMAIN IN PIACE FOR FUTURE DATA RECORDING SEE FIGURE 4.5-1 FOR DETAILED m LOCATIONS. EACH THERMOVOUPLE LOCATION UILL HAVE A l-MINIMUM OF (3) THERMOCOUPLES.IDCATED AT. . L ^2 T, 3, 6 AND 12 O' CLOCK AROUND THE PIPE. s STEAM CENERATOR EACH IANYARD LOCATION WILL HAVE (3). RC-E-1A LAYNARDS TO MEASURE THE VERTICAL. IATERAL p l l AND AXIAL DEFELECTIONS.. l FIGURE 8.1.2.1-1 LOOP A _____'m-___ ..E mm .m.,._1,, c
i PROPOSED INSTRUNENT FIAN LOOP C j Ty.- PROPOSED THERMOCOUPLE LOCATION 4 l~ Ly - PROPOSED IAYNARD LOCATION , Tyt - IDCATE MID WAT BETWEEN IDOP SEAL AND CRANE WALL P3 - -k - P2 / \\ 4 TF2 - LOCATE MID WAY BETWEEN Lyt AND LF2 Ty3 IDCATE 6 FEET FRON LF2 - T \\ P2 / Ty4 8. Ly3 . IDCATE MID WAY BEIVEEN / p - PENETRATION AND RISER p Lyt IDCATE ADJACENT TO FUR-1 ~ IDCATE ADJACENT TJ FUR-13 PENETRATION I 78 LF2 PI P3 EACH THERMOCOUPLE IDCATION T 4 VILL HAVE A MINIMUN OF (3) P4 THERMOCUPLES IDCATED AT T i 3, 6 AND 12 O' CLOCK AROUND P1 THE PIPE. EACH IANYARD LOCATION VILL HAVE (3) IANYARDS .h _. ' 7 TO MEASURE THE VERTICAL, gg l IATERAL AND AXIAL RC-E-IC DEFELEt TIONS.- W i LOOP C FIcURE 8.1.2.1-2 5 O O e e t..~ --+,.- ~ < d ~ m ,--E -e m-~-
critical stress locations and will be easily accessible during the thermal sleeve installation, the wall thickness of these elbows should be measured. This will confirm that use of a.843 inch wall thickness for the elbows (a critical stress area) is appropriate. This will not be required if these elbows are replaced per Section 8.3.3.5. 8.1.2.4 PIPE SUPPORT MODIFICATIONS Inspection of the monor til support R5 during 7R indicated that it may not have been functioning properly. Pipe stress analysis of f the feedwater lines indicated that the remaining monoballs were potentially not functioning properly. The best correlation of loop A recorded data occurs when all loop A monoballs are considered as 3-way restraints (i.e.,.monoballs are designed as vertical _only 1 supports but are also restrr.ining the axial and lateral directions).: Analysis of the monoball supports indicate that they do not meet the plant design basis -criteria. If these supports (H-3_and H-4 on loop A) are not functioning properly and they do not meet the . plant design basis criteria, they must be replaced or redesigned to function as vertical only supports and strengthened to carry the additional global stratification loads. The monoball supports meet-the one time loading requirements per ASME 'III Appendix F _ which - allows continued operation of-the unit (see Section 8.1.2.6). 8.1.2.5 PIPE SUPPORT INSPECTIONS Several inspections of snubbers and ' hanger frames are. also required to confirm assumptions. Snubbers to be-as-built are HSS-201, HSS-202,'HSS-203, HSS-204, HSS-20d HSS- - t 206, HSS-212 AND HSS-212A. Frames to be-as-built are H-3, H-4, H-5 and H-11.. 8.1.2.6 INTERIM EVALUATION All piping systes components did not meet the design basis acceptance criteria _ when considering effects of global stratification and potentially locked monoballs. In order to justify the cor.tinued operation of the unit, 103
appropriate interim acceptance criteria were developed and utilized in the evaluation of piping and supports. These alternate criteria were applied to qualify monoballs support frames and to evaluate pipe break stress levels. Appendix D discusses specific criteria utilized and affected piping systam components. An acceptable limit for an OBE event, equal to.25 of the existing OBE limit, was determined for interim plant operation. 8.1.2.7 REVISE PIPE SUFPORT CRITERIA Pipo support SH-6 was analyzed using ASME Boiler and Pressure Vessel Code case N-318 (Ref. 9.64). This code case provides alternate criteria for the evaluation of piping local stresses due to integral welded attachnents. Code Case N-318 identifies current criteria available for the analysis of rectangular integral welded attachments. It provides relief from overly conservative stress indices and allowables provided certain geometric limitations are satisfied. Therefore, Code Case N-318 was adopted for the analysis of integral welded attachments. 8.2 STRONGLY RECOMMENDED CORRECTIVE AfTIONS The strongly recommended corrective actions are primarily concerned with steam generator nozzle cracking and future possible water hammer events that were identified during the review for the root cause of the irregular pipe deformation' pattern. 8.2.1 STEAM GENERATOR N0ZZLE CRACKING The recommended corrective actions for the steam generator nozzle taper transition cracking are a modification to install thermal slooves to prevent reoccurrence of the
- cracking, and pre / post installation incpection of the feedwater line after the modification.
These two actions are discussed below in Sections 8.2.1.1 and 8.2.1.2 respectively. 8.2.1.1 MODIFICATION Analysis of the specific steam generator loop A nozzle taper transition and associated thermal transients indicates that the cracking 104
could initiate after approximately 16 plant events. Each of the first two times (1979 and 1988) when cracking was discovered, the elbows were simply replaced. A better solution than just replacing the elbows. When they crack is to protect the taper transition from cracking by adding a thermal sleeve. The thermal sleeve will reduce the coefficient of heat transfer which is estimated to be as high as 2000 Btu /f t8-hr *F withcut the thermal sleeve to as low as 20 Btu /ft8-hr
- F with the thermal sleeve (Refs.
9.16 and 9.17). This will eliminate the thermal f atigue concern analyzed in Reference 9.20. Thermal sleeves should be installed and should consider thermal cycling data recorded during "9. The thermal sleeve will be provided witt, two 90 deg*ee short radius elbows which will replace ono half of the existing loop seal on each loop and will aid in the installation. These elbows are also upgraded from grade B material (yield stress = 35 ksi) to grade C (yield stress = 40 ksi) to provide additional strength in a high stress location during global thermal stratification transients. 8.2.1.2 PRE / POST INSTALLATION INSPECTIONS Preinstallation and post instaAiation inspections are recommended for the implementation of the thermal sleeve modification. These inspections should identify and confirm all rupture restraint gaps.. This information along with construction procedures will prevent misalignment of feedwater lines due to installation. 8.2.2 FUTURE POSSIBLE WATER HAMMER EVENTS Steam generator water hammer has been identified at other operating plants of similar vintage. Inspections for two potential causes of steam ganerator water hammer to which Beaver Valley Power Station-Unit 1 is susceptible to are discussed below. 8.2.2.1 FEEDRIN3 THERMAL SLEEVE INSPECTION The feedring thermal sleeve of the steam 105 i i
f generator must be inspected to assure that significant erosion of the slip fit connection has not occurred. This corrosion has been documented at Trojan Power Plant and was determined to be the cause of a severe water hammer event. The inspection can be easily performed during the installation of the stcam generator nozzle taper transition thermal sleeve, when the two closest elbows of the loop seal to the steam generator are removed. Erosion of the slip fit must be corrected to prevent a eignificant water hammer event similar to Trojan's from occurring at Beaver Valley Power Station-Unit 1. 8.2.2.2 J-TUBE EROSION INSPECTION Another possible cause of steam generator water hammer is the erosion of J-tubes or the loss of bottom plugs on the feedring. Only a 22 cubic inch steam void is considered to have caused the 40 kip loads postulated at Trojan.' Several missing plugs and/or leaking J-tubes could' allow enough steam to enter the feedring to cause a severe water hammer at Beaver Valley Power Station-Unit 1. The J-tubes and feedring plugs should be inspected. Missing plugs or severely eroded J-tubes should be replaced to prevent a significant water hammer event from oc.rring at Beaver Valley Power Station-Unit 1. 0.2.3 TACK WELD SHIM STACKS The rupture rastraint shims stacks on the Beaver Valley Power Station-Unit 1 feedwater lines have been dislodged several times due to water hammer events and thermal movements in combination with the irregular pipe deformation pattern. Tack welding the shin stacks would better maintain the appropriate rupture restraint gaps. This would eliminate periodic replacement efforts by construction and maintain the appropriate rupture restraint gaps. Existing design basis water hammer and seismic displacements would not contact the rupture restraints however global stratification l displacements can close the gaps. 1 l 8.3 RECOMMENDED CORRECTIVE ACTIONS 1 l-The recommended corrective actions are primarily l concerned with future possible water hammer events and 106 i w
other concerns. These action items provide additional information in the event of further inquiry and minimize the impact of various transients by replacing materials or by limiting conditions. The future possible water hammer is discussed in Section 8.3.1 and installation of MIN-K insulation is discussed in Section 8.3.2 and is applicable to both future possible water hammer and the other F *cerns as well as routine inspections. The other conce i: >Te discussed in Section 8.3.3. 8.3.1 FUTURE POSSIBLE WATER HAMMER Two additional actions are identified below whose implementation would further reduce the probability of a feodwater system water hammer. 8.3.1.1 PERIODIC REVIEW OF FRV AND TURBINE BUILDING PIPING Failure of the main feedwater regulating valve (FRV) linkage cause a significant water hammer event at Beaver Valley Power Station-Unit 1 in 1981. The linkage failure caused primarily by resonance / vibration of the valves and piping has only been partially corrected. A review of the resonant period should be undertaken to avoid or eliminate the resonant peaks f or this piping system. This will prevent a future water hammer due to FRV f ailure. For example, steps should be taken to decouple /detune system piping from the valve stem / plug resident frequency at the 17 Hz range. 8.3.1.2 LIMIT MAXIMUM AUXIL1ARY FEEDWATER FLOW A limit on the maximum AFW flow rate should be established and implemented in order to reduce the magnitude of a steam generator bubble collapse water hammer. Reducing AFW flow will slow down the rate of tre steam bubble collapse (if a steam bubble exists in the steam generator feedring) which will reduce the magnitude of water hammer loads. This was recommended in Reference 9.58. Beaver Valley Power Station-Unit 1 currantly uses an AFW flow rate of approximately 460 gpm per steam generator. This should be reduced to 150-200 gpm. Although 150 gpm is recommended by References 9.58 and 9.37, the current minimum design flow is approximately 175 gpm. The BV-1 plant specific steam generator flow requirements may require a flow rate greater 107
than 150, gpm, but one less than 175 gpm can be achieved. 8.3.2 INSTALL MIN-K INSULATION AT ALL RUPTURE RESTRAINTS Installation of MIN-K insulation at all feedwater rupture restraints inside containment will facilitate inspection of the rupture restraint gaps. Future required inspections of these gaps are considered to be likely based on the feedwater lines operating history and concerns about water hammer and global stratification. The installation of MIN-K insulation would require a review of affected design calculations to ensure design basis requirements are maintained. 8.3.3 OTHER CONCERNS Recommended corrective actions for other concerns include the following items tot a. Review global stratification effects in the MSVH and Turbine Building (see Section 8.3.3.1). b. Perform sample NDE's inside containment to verify wall thickness (see section 8.3.3.2). c. NDE the auxiliary feedwater nozzle in the MSVH (see Section 8.3.3.3). d. Replace the remaining loop seal grade B elbows with grade C material (see Section B.3.3.4). e. Inspect other monoball supports (see Section 8.3.3.5) These issues are discussed in detail in the following paragraphs. 8.3.3.1 REVIEW GLOBAL STRATIFICATION EFFECTS IN THE MSVH AND TURBINE BUILDING The feedwater lines inside the MSVH and Turbine Building should be reviewed for global thermal stratification effects. Instrumentation and selective nondestructive examf. nation of these lines should be undertaken to ensure that no detrimental unanalyzed condition is occurring in the MSVH and Turbine Building. 108
o 8.3.3.2 PREFORM SAMPLE NDE'S TO VERIFY WALL THICKNESS Concerns exist from previous evaluations that portions of the feedwater lines inside containment have wall thickness of.750 inches instead of .843 inches. Selective nondestructive examination (NDE) of each feedwater line would determine the magnitude of the feedwater line wall thinning problem i and help substantiate the existing analysis. The loop seal elbows are exempted from this recommendation because their inspection is required per Section 8.1.3.3. 8.3.3.3 NDE AFW N0ZZLE IN MSVH The auxiliary feedwater nozzle in the MSVH is subjected to relatively high thermal stresses after a_ reactor trip. After a reactor trip and water from tank WT-TK-10 at a temperature ^ the auxiliary feedwater system is activated as low - as 45' F is injected into 441' F feedwater line. NDE of this connection would assure integrity of the nozzle and of the remaining piping system up to the steam ~ ' generator nozzle. The AFW connection is-the most thermally stress fatigued region of the feedwater system with the exception of the steam generator nozzle because of the 45' F to 441' F step change. Fortunately the' number of events is relatively small and no indications would be expected upon NDE. 8.3.3.4 REPLACE REMAINING LOOP SEAL GRADE B ELBOWS WITH GRADE C MATERIAL The four existing steam generator loop seal short radius elbows on each feedwater line are currently stamped as grade B material. These elbows are the most highly stressed components of the system due to global thermal stratification. The two elbows closest to the steam generator are being replaced with grade C material during the thermal _ sleeve i* installation. Grade B and c materials have l yield stresses of 35 and 40 ksi respectively. Replacement of the remaining two elbows in each loop seal could also be made thereby increasing the margin of safety. These elbows 4 109
must be removed to f acilitate the installation of the new thermal sleeves. 8.3.3.5 INSPECT OTHER MONOBALL SUPPORTS All other monoball supports on Beaver Valley Power Station-Unit 1 that are not addressed in this report should be inspected to ensure that they_are constraining the pipe as designed. - J l .I + i + ^ .r t ~ . r e b h e i I i 1 110 -.[4._--.,,
i
9.0 REFERENCES
l 9.1 DBD-24A, Design Basis Document for BV-1 Steam Generator Feedwater System, Rev. O. 9.2 DBD-24B, Design Basis Document for the BV-1 Auxiliary i Feedwater System, Rev. O. 9.3 Operating Manual Chapter 24, BV-1 Steam Generator Feedwater System, Rev. 11. 9.4 UFSAR Section 5.2.6, Missiles and Pipe Rupture, Rev. 1. 9.5 NUREG 0800, Section 3.6.2, Determination of Rupture Locations and Dynamic Effects Associated with the Postulated Rupture of Piping, RLv. 1, dated July 1981. 9.6 Mechanical Engineering Branch Technical Position, MEB 3-1, Postulated Rupture Locations in Fluid System Piping Inside and Outside Containment, Rev. 2, dated June 1987. 9.7 Calculation 11700.34-NP(B)-783X, Problem 783-Pipe Stress Calculation for Steam Generator Feedwater Fiping Inside Containment (RC-E-1A), Rev. 2, dated November 19, 1990. 9.8 Calculation 11700.34-NP(B)-784X, Problem 784-Pipe Stress Calculation for Steam Generator Feedwater Piping Inside Containment (RC-E-1C), Rev. 2, dated November 19, 1990. 9.9 Calculation 11700.34-NP(B)-785X, Probi tm 785-Pipe Stress Calculation for Steam Generator Feedwater Piping Inside Containment (RC-E-1B), Rev. 2, dated November 19, 1990. 9.10 Calculation 8700-DMC-2358, Feedwater Rupture Restraint Cap Cold Set Evaluation, Rev.0, dated December 20, 1989. 9.11 Calculation 8700-DMC-2355, Analysis of Main Feedwater Lines, As Found During 7R, Rev. O, dated February 5, 1990. 9.12 NRC IE Bulletin No. 79-13, Cracking in Feedwater Piping, dated June 25,1979. 9.13 UCRL-52265 "An Investigation of Pressure Transient Propagation in Pressurized Water Reactor Feedwater Lines", Lawrence Livermore Laboratory dated July 22, 1977. t 111
9.14 BV-1 Response to NRC 1.E. Bulletin 79-13, Stress Analysis of the DLW Feedwater Piping, dated June 25, 1977. 9.15 Not Used 9.16 W.H. Bamford, A. Thurman and M. Mahlab, Fatigue Crack Growth in Pressurized Water Reactor Feedwater Lines, ASME Paper 81-PVP-2 (1981). 9 17 A. L.
- Thurman, M.S.
Mahlab and R.E. Boylstein, 2-D Finite Element Analysis for the Investigation of Feedwater Line Cracking in PWR Steam Generators, ASME Paper 81-PVP-3 (1981). 9.18 M.H. Hu, J.L. Houtman and D.H. White, Flow Model Test for the Investigation of Feedwater Line Cracking for PWR Steam Generators, ASME Paper 81-PVP-4 (1981). 9.19 R.
- Branchel, M.
Miksch and G. Schucktanz, Thermal Stratification in Steam Generator Feedwater Lines, Trans, ASME Vol. 106, February 1984, pp. 78-85. 9.20 Calculation 01296-NP(B)-4001, Pipe St ,os Anclysis of Local Thermal Stratification Fatigue Analysis of Beaver Valley Power Station-Unit 1 Loop A FeedWnter Nozzle Taper Transition and Root Cause Evaluation for the observed 7R Irregular Deformation Pattern, Rev. O, dated November 19, 1990. 9.21 NRC-AEOD
- STUDY, Review of Thermal Stratification operating Experience, by office for Analysis and Evaluation of Operational
- Data, dated March
- 1990, Prepared by Nelson T.
Su. 9.22 SWEC
- Report, Request for Additional Information, Evaluation of Main Feedwater Piping Vibrations, submitted under Docket No. 50-334, dated January 21, 1977.
9.23 Letter ND1NSM 2490, Feedwater System Perturbations dated q 7anuary 8, 1987. h 9.24 NUREG-0582, Water Hammer in Nuclear Power Plants, da' ed July 1979. 9.25 NUREG/CR-2059, Compilation of Data Concerning Known and Suspected Water Hammer Events in Nuclear Power Plants, dated May 1981. 9.26 NUREG/CR-2781, Evaluation of Water Hammer Events in Light Water Reactor Plants, dated July 1982. 9.27 NUREG-0918, Preventit n and Mitigation of Steam Generator 112
Water Hammer Events in PWR Plants, dated November 1982. 9.28 NUREG-0927, Evaluation of Water Hammer Occurrence in Nuclear Power Plants, Rev. 1, dated Merch 1984. 9.29 NUREG-0993, Regulatory Analysis for OSI A-1, " Water .( Hammer", Rev. 1, dated March 1984. 9.30 NUREG/CR-3939, Water Hammer F.vw Induced Vibration and Safety / Relief Valve Loads, dated September 1984. 9.31 NUREG-0291, An Evaluation of PWR Steam Generator Water Hammer, dated December 31, 1976. 9.32 Not used. 9.33 D. A. Van Duyne and W. Yow, Water Hammer Events Under Two-Phase Flow Conditions, TO 89-56, dated December 10, 1989. 9.34 EPRI Research Project 2856-3, Task 5 Water Hammer Prevention, Diagnostic and Assessment Guidelines, dated March 23, 1990. 9.35 J.
- Lynch, Impedance-Coupled Valve and Fluid System Instability.
9.36 Trojan Report, Main Feedwater Piping Restraint Failure, submitted under Docket No. 50-334, dated June 15,1987. 9.37 NUREG-0800, SRP Section 10.4.7, Condensate and Feedwater System, Rev. 3, dated April 1984. 9.38 Branch Technical Position ASB 10-2 Design Guidelines for Avoiding Water Hammers in Steam Generators, Rev. 3, dated April 1984. 9.39 Engineering Memorandum 60322, Feedwater Restraint No. 38, dated May 6, 1981. 9.40 Engineering Memorandum 20724, Acceptance of Shimming of Feedwater Restraint No. 38, dated May 7, 1981. 9.41 Report No. 01036-535-AV1, Revision 0, dated DRAFT " Loop A Feedwater Line Recorded Data-Beaver Valley Unit 1 Post 7R Heatup". 9.42 NRC Bulletin No. 88-11, Pressurizer Surge Line Thermal Stratification, dated December 20, 1988. 9.43 Stone and Webster Calculation No. 17801-NP(B)-X005-FA-
- 50K, Water Hammer Analysis of tne Main Feedwater System 113
_ _. _ _ _.. _ _ _ ~ 4 Due to check Valve closure Following Pump Trip and Flow Control Valve Closure, Rev. O, dated January 18, 1989. 9.44 STRUDL-SW, ST-346, Computer Code for the Analysis of Structural Elements. 9.45 NUPIPE-SW, ME-110, Computer Code for the Analysis of Piping Elements. 9.46 ANSYS-4.4 PC based Engineering Analytical Computer Code. 9.47-Correlative Analysis of ANSYS 4.4 pc And STRUDL-SW Thermal Stratification Profiles for Main Feedwater Piping Beaver Va.My Power Station-Unit 1, by Carol Allen, DRAFT. 9.48 Stone and Webster Pipe Support Analysis i
- a. 11700.34-NP(B)-783-Z 1 through 11 r
- b. 11700.34-NP(B)-784-Z 1 through 3 c.
11700.34-NP(B)-785-Z 1 through 6
- d. 8700-DMC-2335 e.
8700-DMC-2329 l 1 9.49 Metallurgical Examination Report Letter: NDIMNE 4278 I dated September 2, 1988 DLC Nuclear Group to: T.P.Noonan-from: N.R.Tonet. 9.50 STDRAW, GR-147, Computer Code for Plotting STRUDL-SW Models. 9.5L NUDRAW, GR-159, Computer.. Code for ' Plotting - NUPIPE-SW Models. i 9.52 1981 Water. Hammer Report Response Filed Onder LER 81-032 (Westinghouse). 9.53 Calculation _ 01296-NP(B)-4002 Feedwater Nozzle ; Load Evaluation for Steam Generators RC-E-1A, 1B & 1C, Rev 0, dated November 19, 1990. H . 9. 54 Calculation- 01296-NP(B)-4003 Containment' Penetration Review (Penetrations No. 76, 77 & 78 Only) -For Loads Resulting From-Thermal Stratification Analysis, Rev. O, l-dated November 19,1990. l. 9.55 Not-used 9.56 Calculation 8700-DSC-156-T, Embedment Plate As.41ysis for Feedwater' Loop A,. dated November 14,-1990. 9.57 IR 89-113,-Damaged Shim Packs on Pipe Restraint. 114
4 9.58 Westinghouse Technical Bulletin NDS-TB-75-7, Rev. 1, dated' March 9, 1977. e 9.59 NRC Information Notice No. 84-87, Piping Thermal Deflection Induced by Stratified Flow, dated December 3, 1984. 9.60 HTLOAD, ME-142, Computer Code for Heat Transfer and Thermal Loading. 9.61 LER 81-032 0J62 ANSI'B31.1, Power Piping Code, 1967. 9.63 ASME ~ Boiler and Pressure Vessel
- Code, Section III Appendix.F, 1989.
9.64 ASME Boiler and Pressure Vessel Code case N-318-2, dated . July 12, 1984, ' Procedure for Evaluation of the Design of Rectangular Cross Section Attachments on Class 2 or 3 Piping,'Section III,-Division 1. 9.65 Calculation 8700-DSC-156-U, Embedment Plate Analysis for Feedwater' Loop B, dated November 14, 1990. [ 9.66. Calculation 8700-DSC-156-V, Embedment Plate Analysis f or Feedwater Loop C,-dated November 14, 1990. 9.67)DCP-248, Feedwater. Control Valve Modification, Implemented During Second Refueling Outage, March 1982. 9.687 Letter from A. Schwencer, NRC to C. Goodman,Jr., Portland General Electric Company,
Subject:
. Safety Evaluation-Report for Steam Generator _ Water Hammer At Trojan Nuclear-Plant, Docket.No. 50-344, dated' October 18, 1979. -. 9.69 Letter from J. L. Williams,. Portland General Electric Light Company,'to N. R. Butler, NRC,.
Subject:
Report on Testing of-Auxiliary Feedwater Addition, Following J-Pube 'y Modifications to -the S/Gs oof the Trojan Nuclear Plant, Docket:No. 50-334, dater s0tober 21,1975. s i r b I 115 1 i e, y a e,--...,.9y,,--,,r-, y ..v~ t +, -.w c-.e-
r-ir vy e
n'ew*e-w,-ww*-w-*-*w+--
- +*=****-++-e-*-
a APPENDIX A 10.1 CHRONOLOGY OF EVENTS POST 7R HEATUP t 9 ? 0 0 116
1 APPENDIX A DATE TIME LVENT 12-14-89 0615 START TO FILL STEAM GENERATOR THERE WAS NO ACTIVITY BEFORE THIS DATE 12-14a89 1752 COMPLETED FILL OF AL1, STEAM GENERATORS USED AUXILIARY FEEDWATER PUMP 3B PLANT STATUS RCS PRESSURE 300 PSI RCS TEMPERATURE 256' F 12-15-89 1900 MOV-FW-151D XF WILL NOT CLOSE ELECTRICALLY 12-15-89 1945 B STEAM GENERATOR 1.EVEL INSTnUMENTS DO NO'1 CORRELATE (485, 486) A AND C LOOPS CONSIDERED OPERABLE 12-15-89 1956 2/3 B STEAM GENERATOR LEVELS ABOVE 75% PIPE STRATIFICATION OCCURRED FEEDWATER ISOLATION SIGNAL TO MOV-FW-150A,XB WHICH CLOSED AUX FEED INJECTION RESULTED PLANT STATUS MODE 5 RCS PRESSURE 300 PSI RCS AVERAGE TEMPERATURE 265* F 12-15-89 2029 .iiHUT OFF AUX FEED PUMP 12-16-89 0300 STROKING MOV'S FW-151A AND C NOT B WITH AND WITH OUT FLOW i 12-16-89 0530 STROKED MOV,S-FW-151-B,D,AND F 12-16 89 1148 WORKING ON B STEAM GENERATOR LEVEL TRANSMITTER 485,486 FEEDWATER ISOLATION SIGNAL MOV'S-156-A, B AND C CLOSE 12-17-89 1915 INSTALLED DUMMY SIGNALS ON $T-FW-4R5 AND 486 TO WORK ON THEM 12-18-89 0708 ENTERED MODE 3 RCS PRESSURE APPROX 300 LBS RCS TEMPERATURE APPROX 256' F 117
12-18-89 1323 OPENED TV-MS-101; STEAM GENERATOR SWELL [ OCCURRED FEEDWATER ISOLATION SIGNAL 12-19-89 0500 FT-FW-476 EXHIBITING SPIKIrtG \\ DECLARED OUT OF SERVICE (005) 12-19-89 1641 SHUT DOWN FW-P-1A 12-19-89 1853 ATTEMPTED TO START FW-P-1B 19-19-89 1905 REATTEMPTED TO START PUMP 1B 12-19-89 2104 REATTEMPTED TO START PUMP 1B 12-19-89 2341 -WHILE RUNNING FW-P-2 FOR OST FOUND VALVE FW-36 LEAKING 12-20-89 0212 STARTED PUMP FW-P-1B 12-20-89 0217 SHUT DOWN FW-P-1A 1?-20-89 0320 START UP PUMPS FW-P-1A AND FW-P-1B
- .2-20-89 0441 START UP FW-P-2 FOR OST 12-20-89 0630 DECLARE FT-FW-477
'A' STEAM GENERATOR FEED FLOW 005 ALARM AND CONTINUOUSLY SPIKING 12-20-90 09/0 FT-FW-476 DECLARED OPERABLE BASED ON SUITABLE OUTPUT AS SEEN ON RECORDER 12-20-89 1355 DURING MSP 104 A STEAM GENERATOR 3ROPPED TO 25% WITH F-477 TRIPPED FOR FSP _ 22-89 1004 FT-FW-477 DECLARED OPERABLE FOLLOW 7 NG STABLE BEHAVIOR ON RECORDER TRACE 12-24-89 1902 COMMENCED REACTOR STARTUP 12-25-90 0000 ENTERED MODE 2 12-26-90 0800 ENTERED MODE 1 12-26-90 0852 LATCHED TURBINE I g 12-26-89 1043 TURBINE TRIP 12-26-89 1325 GENERATOR TRIPPED ON OVER EXCITATION 12-26-89 1413 TURBINE REATTACHED 118 y
12-26-89 -1457 TURBINE TRIP 12-27-90 0120 APPROX 29% POWER 12-27-89 0121 REACTOR TRIP TURBINE TRIP MAIN FEEDWATER ISOLATION AUX FEEDWATER INJECTION RESULTS WITH..JX FEED PUMPS 3A AND 3B MODE 3 RESULT OF REACTOR TRIP 12-27-89 0124 STEAM GENERATOR LEVEL ( APPROX 11.8 INCHES FEEDRING NO LONGER COMPLETELY SUBMER'iEC 12-27-d9 0128 STOP BOTH MAIN FEED PUMPS (1A AND 1B) 12-27-89 0144 STOP AUX FEED PUMP 3B 12-2*J-89 0152 START MAIN FEED PUMPS 12-27-89 0243 STOP AUX FEED PUMP 3A 12-27 0250 ATTEMPTED TO-OPEN MOV,FW-150A MAIN FEED REALIGNED 12-27-89 0912_ STOP BOTH MAIN FEED PUMP 1A AND 1B 12-27-89. 1921 REACTOR CRITICAL (I.E., MODE 2) .12-27-89 2300 PLANT AT 10% POWER (BEGIN 8 HOUR SOAK AT 10% POWER) 12-28-89 0700 COMPLETE 8 HOUR SOAK AT 10% POWER 28-89 ???? INCREASE TO.24% POWER 12-28-89 1500 RAISE POWER TO 30% POWER FOR 30 HOUR SOAK 12-28-89 1700 BEGIN 30 HOUR STA3ILIZATION AT 30%' POWER '12-29-89 ALL 30% POWER __ 30-89 ALL 30% POWER gg 32-01-b9 ALL 30% POWER 01-01-90 ??? ELEVATE TO APPROX 50% POWER r 01-01-90 ??? ELEVATE TO APPROX 58% POWER 01-01-90 1500 ELEVATE TO 70% POWER / 119
., n ~ . - ~.- 'I g' k.4 t - 3 s 1 04'-90 :??? ELEVATE-TO:88% POWER--- ~ E 01-05-9 0 ; i ?? ? _. ELEVATE TO 96% POWER-01-19-90 :.1632: COMMENCE PLANT SHUT DOWN 4 L f 01-19 -9 0 - -- 17 0 2 2 - SHUT DOWN FW-P-1A SOLATE-MAIN-FEED REGULATING VALVES . 01-19-90--1803 TURBINE' TRIPPED-N t ) '?i lt' ) ll}l. 5 , V b t s fi L(. j 3 . s. l .1 t 6 l -_g t. i; W p k < : } t! _- =, ,.p h - i " I p, . (
- 1..
i p 5 L -h q ). ( d n , n. I
- ip
in. _, I '; g t' L %g -t j' - F -y pogg i 1 e i y E- -120 4 i l'.4; ?5 p-1 2 s'. d r-v ,-5, r.c- ,.'4 + - -v E,, r,-+ eh-m o e4 eJ J
7 f t 4 APPENDIX B 10.2 GRAPHS OF OCTOBER 6, 1990 MANUAL COOLDOWN J G : 't l s l l I e 'I i 121 l. (- I !1' >
4 BV1 FEEDWATER LINE-A 05:20, October 6,1990 600 4 SGO E 400' Oa Uw w tt 300 - I ~ g -TB n. 2 [ a T9 0 i i i i i 1 3 i t O 2 4 6 8 10 12 14 16 18 20 (Thousands) TIME [SEC]
t BV1 FEEDWATER LINE-A 05:20. October 6,1990 600 ') A soo 7 hr MWM6 l 1 h-400 __ b A I _7 s m,L l i t = r i 9A og/ 4 a ,c, f 300 - T4 T5 ,i,iiiiiiiiiiiii o = O 2 4 6 8 10 12 14 16 18 20 (Thousands) TIME [SEC]
e e 4 i e i I i l L F^ 1 I } L I CC 5 F i i em- - i l Lc . j, w f I ~ i i .i i e y-7 i n 1 - h I C l I' I O-1 j Ch - l I. ~ O l 2 t (y ~ l-4
- aj o
!w r 8-1 -.-1;.--
- 1
'n- .t ao 4 i w J t i O % := O mE: - } r 2 I ( wm -y I g e x N l
- LLj;j s
1 4.C it ^j y 4 l I 4 y g -~ l.- i W w ~ 1 I,, H ,j; 'I ' I g j 'N - I) N () f) Q g) O O N i i. Y h k 124-e ' [b i.
9' 1 - +=, 'e; l l 4 t 'Mlh-s 3 'ec 4 ~ is i F-' I. i .iw 1 i l. F-1 b ,.O k.* c, ~~' ~ ^ di ~ i k 7 ,1 ...,)., # h, .L. l t 7, 4- .w -. s Ji
- U
\\ = ') '~ p w -Q. Id y + x e -.O. c Q. a --6 L -i
- j ~
t [E I-01 .. j y. _. ~ ' t. I ~ i 1 w s
- m. -
} i i ~ i j '*. j ^* .I I i
- w
'g Q .C. V q f} I O C' }- \\ g_ N l. I i _[N1).LN3W3DY^ d$lC
- n.,
125 .a .s; [ s'.
s i i 0' 2 r h T 8 i 1 i i 6 T 1 i tr T A l r4 1 E N i i IL0 2 9 i 1 ) 9 s 1 d 3 R i_ n a}L E6 sC 0u! r T e Ao 71 oS 'v b i[ "- f i E t WO l ( c r M i D 0 E2 i 8 5 E 0. r F i ( 1V r ~ B [> / 4 i f-p -r J T / i j 1 f il 5 0 5 5 2 m 2 5 g 1 0 o_. m 1 1 s C1. s ao o t ~ m
-4 4 a .-s.o ..e6 I .6 - ,s; 4 I t j l 1 i I ~ 4 5 Lr Bi -e W i b l i "n ,Ll;. se 7-r o ~ c4 .%b .m P g "g i v i w m \\ ^
- t 9E~
- - t
~ % a f.
- <[- 1 t
p o. y-CI - m T l.- . LlJ - N-ie .; W. c3 H .[I I w x, - ~ w 1 g.:- _. ~ 1, T,- = C 0-o m o n g d 1 i N i q l I I I i (Nd.Lt42W3OyldsjQ -l 127
a 2%.- _ o m -;.[ ~ m ie- _ m - 0 ^ I 2 LU 7 C . _J 0 _ N ~g 1 E i . LL.J
- OUa
' F-g.. _ o 8h_ b- _u ~ E. C) g - e tg y ' LLj ' $ ' t 1 g .g. - c. 3 CD _ m g O i .i i c o n n un m c O O I 'N i [NI) AN3W3DvidSIC 128
e 0 4 I i ~ $r x b~.- e <d t L <C^ I uJ Z ~ W -O m Y-C . D_,s r s._. t _J * - }- - E l 09-
- OEZ,
- ( j-
- - l
.__:g ai ui s5. J g_ ~ uPy La - e uJ 8-M u.. _+ _ a -er 03 -$.g D f$ + C o O C eq o rN c O d I N I i l [NI).LN3W30V1dSIG l l 129
-7 i i 02 T i 8 1 i 6 i 1 i A ~ ly i I4 g E h' i N { n I L0 i ? 9 I )' 9 r. t 1 1 R u 1 n} 6 E t . 0 w 1, r T e fw" 1 Ab j o I t t l WO i t c i. 1 D 0 E2 i 8 E5 0 i 'F ~ Y r r-1V i B T 4 r 1 ? i o 5 2 p 5 5 5 1 1 0 0 1 1 ,?_ o4 -[1 n* 5 3 r
g e _oN l m -4 .v. ._ i _D F .4 -4 L:.J 7 w a, j I -o. .m O ..e m E ~, ~ j r w.= c H-t _ c y =:- ~1
- <(
- Y 5 y:
g' w o ~ C) g. -r y=-n 'Lij-@ j' l~ s. ? v L l l-l- e CD r 1 L.. h v L 4 u i L N i c m. m .c c o c o o 1 1 I i [NI)' 1N3 W 30Vid 510 131
en a 4 - 4t. 2 .4 0 0. 4 5 4 APPENDIX C 10.3 PLANT EVENT DESCRIPTIONS ',k ' ' i i e 4 2 1 ts -( ..O t 132 e c v v ,w, ,e g
..r i> APPENDIX C ' EVENT ~ ACTIVITY RESULTS-
- 1. Heatup.(1) ilot stand-by cycling Significant cycling of-from-Mode 3--to Mode:
local stratification at 1 1ow-poVer steam generator nozzle operation.- from-Mode 3 to. low power operation. s 2.-Heatup (2) From-Mode 5 to Mode Produced 1 o c a l' 3 and then returning stratifim! '. o n '. i . to Mode 5 No signa.lcant' global stratification.-
- 3..Manu sl 'cooldown From-: Mode 1 to Mode Intrasystem global-
-w (1)- 3 and then returning-stratification. - to Mode.1 Significant local A stratification. Not counted as a heatup
- cycle, y
.- 4i - Manual cooldown - Mode 1: to= Mode 5 -Produced-intrasystem' '(2): with subsequent global stratiffr'. tion. ~ -heatup Significant local stratification. Subsequent heatup. 3 counted as a. heatup-ti event. t ~5.. Reactor, Trip; Auxiliary. feedwater-- Produced =significant; (1) Linjection-resulting-cycling Jof - _. local-a in scooldown to-Mode - stratification c at the L.- - steam generator nozzle. EProduced-global" stratification. Not counted-as a heatup; event.
- 16. Reactor Trip
' Auxiliary. feedwater. Produced-significant--
- (2)
. injection-resulting cycling-o f. local in : cooldown - to Mode stratification -at the-3 and.heatup back to-steam generator nozzle'. ,v Mode;1'. Subsecuent heatup-countea-as a' heatup; j event. .h o lj 133
- m
^ u i S- & +
- ew e
<e- - - -., w ce m-. r +-c e v w
9 e i ,1;' 4 .6 APPENDIX D 10.4 JUST7FICATION FOR' CONTINUED OPERATION - t 4 1 4 e f 't e 'k e- '134
APPENDIX D BEAVER VALLEY POWER STATION-UNIT 1 FEEDWATER SYSTEM EVALUATION JUSTIFICATION FOR CONTINUED OPERATION (JCO) DECEMBER 13, 1991 Duquesne Light Company has been conducting an evaluation of the main feedwater piping inside containment at Beaver Valley Power Station Unit I due to the history of cracking that this piping has had in the vicinity of the steam generator nozzles. Init al concerns related to cracking in feedwater system piping were identified in NRC IE Bulletin b o. 7913. The short radius elbows on the Beaver Valley Power Station-Unit I feedwater piping immediately upstream of the steam generater nozzle have been replaced twice on loops B and C and three times on loop A due to inside wall cracks. The location and type of crack seemed to indicate that the steam generator nozzle was experiencing un:xpected thermal cycling causing fatigue. During the seventh refueling outage (7R) in November 1989 it was also noted that the loop A and C feedwater piping inside containment was misaligned in the vicinity of the system's rupture restraints. This misalignment was most pronounced at the first rupture restraint upstream of the steam generator nozzle (FWR-38 and FWR-1, in loops A and C respectively). The affected piping was cut and reinstalled, and the rupture restraint gaps v'ere reset to design nnditions. In order to help determine the probable cause(s) of the piping misalignment, tiie loop A feedwater piping inside containment was instrumented with thermocouples and lanyards prior to the plant restart after the 7R outage. Temperature and displacement data were recorded between 12-15-89 and 1-19 90, during which time the plant experienced several events including a reacter trip at twenty-nine percent power and a manual cooldown. Additional data were recorded between 10490 and 10-13-90 following a manual cooldown. A review of the recorded data indicated that the feedwater piping experienced signi0 cant global thermal stratification and that the nozzle was subjected to significant thermal cycling. Based on these data, thermal stratification was initially identified as.' possible cause of the local nozzle cracking and the global piping deformation. The thermal s ratification phenomenon results in the upper part of the feedwater pipe to be heated to a Egher temperature than the lower part. The difference in the temperature and resulting differential thermal expansion between the top and bottom of the piping induces bending moments and si3nificant deflections in the feedwater piping. This does not cause an overstress of the piping. Additional evaluations utilizing the recorded data indicated that the monoball type vertical restraints may be locking-up during certain system conditions. This locking-up of the monoballs results in unanticipated system restraint on the feedwater piping system inside containment. Recent thermal stratification analysis of the affected feedwater piping (loop A, B and C inside containment) was performed utilizing the recorded temperature and displacement data noted above. This analysis also addressed the impact resulting from the potential of locked-up feedwater system monoball supports. The loop B analysis considered the maximum theoretical potential thermal stratification (maximum main feedwater temperature - minimum auxiliary feedwater temperature). That analysis envelopes all probable thermal stratification profiles and no further action is required for loop B. The thermal stratification profiles analyzed for loops A and C considered a temperature 135
APPENDIX D differential of 245 degrees F which corresponds to about 60 percent of the maximum theoretical potential thermal stratification. This temperature differential of 245 degrees F is based on the limited data obtained between 12-15-89 and 1-19-90 and between 10-6-90 and 10-13 00. The acceptability of these temperature profiles will be confirmed pending the results of a more comprehensive instrumentation program to be performed on the main feedwater system at Beaver Valley Power Station-Unit 1. The above noted analysis concluded that all system pipe stress levels are within design basis allowable limits. Additionally, the revised pipe stress levels do not result in any new pipe break locations based on the criteria provided in NRC Mechanical Engineering Branch Technical Position 3-1 of Standard Review Plan 3.6.2. Beaver Valley Power Station-Unit 1 is not currently licensed tc MEB 3-1. However, MEB 31 is the most current NRC criteria and usc of it for interim plant operation is appropriate provided a future licensing change is performed to incorporate MEB 3-1. An evaluation of the steam generator nozzles and containment penetrations concluded that the aew loads are within design basis allowable limits. Analysis of the feedwater system pipe supports has resulted in component stresses that exceed design basis allowable values. A further analysis of the affected pipe supports concluded that there is an acceptable design basis which provides a justification for continued operation (JCO) for Beaver Valley Power Station-Unit 1. Specific details of this support review along with future system recommendations are contained in the following paragraphs. 1 Since thermal stratification of the feedwater piping can occur during normal plant operation, the resulting loads from this load case are considered to be normal thermal loadings for pipe support design conditions. That is, thermal stratification loads are enveloped with normal thermal loadings and combined with deadioad to determine normal support oesign loads. Specifically: DL + THERMAL = NORMAL SUPPORT DESIGN LOAD Where: DL = Deadioad t THERM AL - Thermal including thermal stratification effects This normal design load was compared to the existing design basis normal allowable to document component acceptability. Additional suppott design loads which include the effects of occasional loadings (seismic and time history effects from fluid transients. events) are also evaluated in determining support acceptability. The normal loading described above was combined with occasion'.1 loadings to determine the additional support design loads. Specifically: DL + THERMAL + SRSS(OBEl, TH) + OBEA = UPSET SUPPOR1 DESION LOAD DL + THERMAL + SRES(DBEI, TH) = FAULTED SUPPORT DESION LOAD Where: OBEi= Operational basis carthquake inertia OBEA = Operational basis earthquake anchor movements TH = Time history effects from f.uid transient events DDEI = Design bek urthauake inertia 136
APPENDIX D These support design loab were compared to existing design basis upset and faulted allowables to document support ar = '9y. In certain instances where these allowable limits were exceeded, the criteia b ao in ASME Section III, Appendix F, " Rules for Evaluation of Service Loading with Level D Service Limits", has been utilized to document support acceptability. This appendix provides alternate stress limits which can be considered when evaluating loads occurring during a plant faulted condition. Although the rules for Appendix F apply to Level D Service Limits, the basis for using this appendix for evaluating the upset OBE loads or the faulted DBE loads is that in the event of a seismic occurrence of OBE or DBE magnitude, Beaver Valley Power Station-Unit I will be shut down and a feedwater system inspection will be performed. That is, only one potential upset seismic OBE event can occur during plant operation for this interim period. Furthermore, prior to any plant restart following this postulated seismic event, long-term (life of the plant) system modifications described later in this JCO will be implemented which will result in component stress levels being within existing design basis allowable limits. Based on this fact, the utilization of ASME Section III, Appendix F methodology and allowable limits is considered to be a reasonable approach to evaluate the feedwater pipe supports for this potential one-time seismic event at Beaver Valley Power Station-Unit 1. The maximum support load from the upset and faulted load combinations previously defined was considered when utilizing the rules of ASME Section III, Appendix F. It should be noted that the upset load combination which includes occasional time history loads only still satisfies the existing design basis allowable limits. That is: DL + THERMAL + TH < UPSET DESIGN BASIS ALLOWABLE It is acknowledged that while Beaver Valley Power Station Unit 1 is not licensed to ASME Section III and all of the requirements of the code are not met, the application of the principles of Appendix F are considered reasonable based on the following facts. ASME III and associated Appendix F is valid code, utilized in the design of several nuclear facilities. The criteria contained in Appendix F are based upon sound engineering principles and material behavior. The materials used at Beaver Valley Power Station Unit 1, while not satisfying all of the requirements of ASME certification, are consistent with the materials specified by ASME III. Typical allowable stress values contained in ASME Section III Appendix F which can be utilized for component evaluation are as follows: Comrmnent Evaluation Allowable Stress
- local stress at integral welded attachments lesser of 3Sm or 2Sy l
support members in tension lesser of 1.2Sy of 0.7Su support members in shear lesser of 0.72Sy or 0.42Su e bolts in tension lesser of 1.0Sy or 0.7Su l bolts in shear lesser of 0.5Sy or 0.42Su l Where: Sy = Yield stress Su = Ultimate st.ess Sm 3 Design stress intensity l l 137 l l
APPENDIX 'D l LThe pipe support-integral welded. attachments were evaluated in accordance with Welding Research Council Bulletin 107 (WRC 107), except in certain instances where ASME Boiler and Pressure Vessel Code Case N-318 (CCN-318) was used = While CCN 318 is not currently-included as part of the Beaver Valley Power Station-Unit I design basis, the application of CCN-318 is reasonable based on the following facts. CCN 318 is a valid code and has been widely used= throughout the_ nuclear industry and it is applicable to all rectangular attachments conforming to certain geometric limitations. Utilizing the methodologies described above, all pipe supports were evaluated and shown to be within acceptable limits. Based on these evaluations, it is concluded that the main feedwater piping system, including pipe supports, equipment nozzles and containment penetrations will be within acceptable limits during all postulated design basis events including the additional effects i resulting from thermal stratification and additional loadings due to potentially locked-up monoball supports. Since the pipe stress, pipe supports, equipment nozzles and containment _ penetrations have been judged to be acceptable, it can be concluded that the pressure boundary of the main feedwater system inside containment will be maintained. Furthermore, the noted reviews provide the basis for the justification for continued operation (JCO) for Beaver Valley Power Station Unit 1. i . Although the justification for continued operation has been concluded based on the noted evaluations, it is desirable and appropriate to perform additional engineering evaluations on the main feedwater system. inside containment, considering potential system modifications, to provide long term (life of the plant) qualification utilizing the existing design basis allowable-limits. ~ These evaluations would provide an acceptable long-term design basis for all plant conditions including the effects due to thermal stratification noted herein. Engineering evaluations shall be - performed in a timely manner to support the potential system modifications that may be required as part of this qualification. -Allimodifications resulting from these evaluations will be implemented during the next planned outage at Beaver Valley Power Station Unit 1. 1The following actions are considered mandatory to provide long-term qualification and to confirm assun.ptions made in the interim plant analysis. 1) . Develop and implement a complete instrumentation and data collection program for the-loop A and C piping to verify maximum thermal stratification levels. y ' 2): - Perform a licensing change to incorporate MEB 3-1. j j 3) Perform a licensing change to incorporate ASME Boiler and Pressure Vessel Code Case g N-318. J 4) Perform field inspections to verify the minimum wall thickness of the loop _ A loop. mal elbows. Ea 5) Redesign pipe supports WFPD R-3/4/11 to function as vertical supports. These supports. Dy shauld also be inspected to verify the as-installed configurations. 6) Pipe supports WFPD-HSS-201/202/203/204/205/206/212/212A should be inspected to p verify snubber piston settings and available travel in both directions. l 7) Implement an interim OBE limit. 1 i c 138
m- .s _,_4.,. S L. .t.: 4 o a APPENDIX E '10.5 CHRONOLooy oP' EVENTS WITH TEMPERATURE TIME ZONES i t h.' i t 139 [.'
APPENDIX E i ] DATE TIME EVENT DESCRIPTION ZONE 12-15-89 1750 Feedwater isolation. J s Auxiliary feedwater 2 injection. 12-26-89 0055 Increasing reactor A power level to 2% power. 12-26-89 0300 Regulating chemistry. B 12-26-89 1325 Generator / turbine trip. K 12-26-89 1800 Having difficulty C R maintaining turbine speed and governor valve control. 12-26-89 1900 Start 8 hour soak. D* .12-27-89 0121 Reactor trip. I 12-27-89 2000 Having difficulty E controlling reactor temperature and steam generator levels with steam dumps. 12-27-89 2300 Begin f3 hour' soak. F* 12-28 1100 Turbine costdown G following overspeed testing. .12-28-69 1235 Turbine synchronized to H system. '01-19-90 1800 Manual cooldown. L NOTES:
- Denotes zone associated with heatup of the-plant from a refueling outage; associated with initial fuel conditioning.
140
9 .-g. 9 10.6 CHRONOLOGY OF EVENTS WITH STRATIFICATION RESULTS -) s O h 9 141
y APPENDIX F DATE-EVENT RESULT 12-14-89 2 out of 3 steam generators Steady. 1oca1 high level. stratification. Feedwater isolation. G 1 o b a 1 ' Auxiliary feedwater injection, stratifioation. 12-16 -Erroneous reading on steam N/A generator level transmitter. y Feedwater isolation '12-18-89 Steam- : generator swell on N/A opening of TV-MS-101. Feedwater isolation
- i 1
12-18-89.- Mode 3 operation. Cyc1ic 1oea1 to stratification. . 12-25 -N/A . 12-26-89-- Commence: reactor-startup; Cyc1ic 1oca1 Heatup through Mode.2 into Mode stratification. 1 (2%) power. Latch turbine. Turbine = trip. Generator-trip. Turbine.reatta'ched. Turbine: trip. 12-26-89 -81 hour' soak-at-.10% power. .C'y c 1 i c 1oca1 ~' stratification. 12-27-89.. Reactor trip: (29%)> power. Cyc1ic l'o c a 1 Feedwater isolation.~ stratification.- 4 Auxiliary feedwater -injection G 1 o b a 1 .of-both motor driven pumps.- stratification. ' 27-89l 8chour soak-at 10% power. Cyc1ic-1 o'.c a 1-3 to -Increaselto'24% power. stratification. 4 12-28-89 l30ihour soak-at~-30% power. 1; 01-19 - Manual controlled cooldown Mode Stead'y
- 1.o c a 1 l'to Mode'3 only.-
stratification. Returned'to. Mode.1. Intrasystem global stratification-(29%) 'g. power. l I, i 142 r i 3 4 + + i
4 0 l '10-06-90 Manual controlled cooldown Mode Steady 1oca1 C 1 to Mode 3 only. stratification. Returned-to Mode 1. Intrasystem global stratification (29%) power. 4 143
4 APPENDIX G 10.7 ASME SECTT0;; III APPENDIX F 4 4 e 8 144
- (*%9*h IMAGE EVALUATION 8#
3 0 M? ////p [h k , + g (gg pf gi f/ TEST TARGET (MT-3) '% Y I l.0
- u n L24 Y $$$ M I.I E2 EB.
m 2 11.25 l i.4 ii=i.6 i i_ = 4 150mm 6"
- /, %i,
/ !b ~ O 9 s* t g z m '%+k+' Q ) //#bE d =-- o 1 %rd.hidhdNdi.,i. ,,,,., f/ ? ' ',[,.l ' 0 W..,. .7
P. r y fo i APPENDIX 0 .s-It is acknowledged that Beaver Valley l'is not bcensed to ASME Ill and 11 of the ' criteria are not' met. However, the use of Appendix F allowables for pipe support qualification for interim plant operation is considered acceptable based on the following reasons: ' A5 ME IIIis a valid code currently in use at several nuclear power facilities. The load colnbinations used in the pipe support analyses are consistent or more conservative than those specified in ASME 111, The materials used at Beaver Valley l'were procurred in accordance with e-ASTM specifications. While not satisfying all the requirements of ASME Ill - certification, the materials used are similar with those specified in ASME Ill. Furthermore, the minimum yield stresses, ultimate stresses etc. used in the ' pipe support analyses tre consistent with the specifications of ASME 111. The criteria presented in Appendix F is based on sound engineering principles and material behavior, it is reasonable to expect that the materials used at Beaver Valley 1 behave -as anticipated by Appendix F. Beaver Valley 1 was constructed and is maintained using good construction r and inspection practices. t 4 8 3 .O i 145
9 4 8 APPENDIX H 10,8
SUMMARY
OF SUPPORT LOADINGS AND STRESSES EVALUATED BY ALTERNATE CRITERIA 4 e x 4 146 __.-______m._.---__._.-__-_.._ms_m__. _m.-______m --______.m__
IfM Condition NORMAL UPSET A UPSET B FAULTED INTERIM - leads considered DI, tiler DI, TIIER, OCC DI, tiler, OCC, OllET Dl, tiler, OCC, DBEI Max of Upset B or Faulted Basis for allowables AISC AISC with 1/3 increase AISC with 1/3 increase AISC with 1/3 incicase ASME III, Appendix F for occasional / seismic for occasional /scismic for occasional / seismic loads loads loads 11-3 Member Shear Stress 8166 PSI s.4S 18126 PSI s 4/3.4S 20127 PSI > 4/3.4S !M23 PSI s 4/3.4S 21027 PSI s lessar of y y y y 1441XI PSI 1920tl PSI 19200 PSI 192fK) PSI 42SU" Y NG 25200 PSI M
- ti 11-3 Holt Shear 1(MMU PSI s ItkNOPSI 10933 PSI s 13333 PSI 18772 PSI > 13333 PSI 16:71 PSI > 13333 PSI 18772 PSI s lessar of E
ts 42S or.5S H g y NG NG 19800 PSI M 18227 PSI s lessar 11-3 Bolt Tension 9199 PSI s 14020 PSI 16546 PSI s 21586 PSI 18227 PSI > **" 15059 PSI s f 28tMU - 1.6f 2(F - f )t/2 2(F 2,g 2)1/2 74p.,r j /2 ,73 y, 3 2t 2 y y y y y y y NG NG 330N) PSI 0.92 s 1.10 11-3 Holt Interaction (f /F [ + (f /F )2 y y i 7 ( t.
7 e H LOAD LOADS CONSIDERED ALLOWABLE Ath>WABLE 11-3 11 4 CONDITION LOAD LOAD DESIGN BASIS INTERIM I DL 1.0 S 14701 PSI s 15fXX) PSI 14819 PSI s ISON) PSI h 2A DL, OCC L2 S 15293 PSI s 18tNN) PSI 155545 PSI s 1MNN) PSI h e 2B Dl, OBEl. OCC 1.2 S 2.4 S 20184 PSI > 1800 PSI 18076 PSI > 18 tan PSI y 3 3 20lM PSI s NdNN) PSI 18076 PSI s NdNH PSI l M ! 2: H M 3 DL, tiler, OBEA L25 (S + S ) Nd)74 PSI s 3754X) PSI 35446 PSI s 375fx) PSI e n 4 DL,OCC 1.8 S 1529.4 PSI s 2700 PSI 15545 PSI s 270H PSI h 5 DL, DBEl, orc 2.4 S 22334 PSI s Ndat) PSI 18874 PSI 5 NdMH PSI 3 e m a _}}