ML20079B847
| ML20079B847 | |
| Person / Time | |
|---|---|
| Issue date: | 05/31/1991 |
| From: | Page J NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| To: | |
| References | |
| REF-GTECI-079, REF-GTECI-NI, TASK-079, TASK-79, TASK-OR NUREG-1374, NUDOCS 9106180011 | |
| Download: ML20079B847 (149) | |
Text
_ _ - - _ _ _---
N U R EG-1374 1
Tecanica Fincings Re:a~:ec :o Generic Issue 79 An Evaluation of PWR Reactor Vessel Tliermal Stress During Natural Convection Cooldown U.S. Nuclear Regulatory Commission Office of Nuclear Regulatory Research J. D. Page jo tovy a
b
/.. $
O 1374 H PDR
-- J
~
. _. _ _ _. _.. _. _.. -. _ ~
I i
A:
a.h, ti c f f u 'e'r r e Maten '* Cqed in IN Pub: cat,ons M; st dc.uments c,'ec m N' p $c m., s vs h bn avai:atne f torn one cf the fotorg Jcer 1
The hW Pubh; Document Room. 2120 L Street NW Lower Level, Washington, DC 20 W 2.
The Supennter dent o' Docu'nents U.S. Gose nmont Pnntng O!!ico, P.O. Box 3700?
Wm.h:ngui, DC 20013-70E2 3.
The Nationc! lechreal Inf ormat cn Seruce, Spongheld. VA 22161 k:ho4h 10 0 M ting that to' aos reprosents the rnapoty of documents cated in NRC pubhca-t or,s. it is not mtended to be rohaustwo.
Referenced documents avaraW for im pection and copying for a fee frorT' the NRC Pubi c Doc ument Rcom nclude NRC correspondenco ena internal NRC m&noranda; NRC Ofhco o' inspect on and Enforcement bu4tns, cecutars informatran notices, inspection and investi-gation notices; Licenseo Event Reports; vendar reporte, end correspondonce; Conmssion pa;sers; and appeant ar.d hcensee documents and correspondence, The foNomng documents in thu N'JREG senes n'a asai:ab!c for purchase from the GPO SMes Pio/aw formM NRC staff and contractor reports, NRC-sponsored conference proceed-ny and NRC boWes and brochures. Also evadable are Regulatory Guidos, NRC regu!a-tons n the Code of f eceret RegMons, and Nucteer Regu!atory Commtssion issuances.
Documents ovalable f rom Ino National Technical Iniormatinn Service include NUREG ser.es reports and technicai reports prepared by other federal agencies end reports prepared by the Atomic Energy Commismo,. forerunner agency to the Nuclear Regulatory Commtss.on, DocumenN avanab'o from pubbc cod special technical 1branes includo all open htcrature items. Such as books, jou'nal and penodical articles, and transactions, Fodoral Reg ster note es, federal and state leg station, and congressional reports can usuM!y be obtained from these Lbranes Documents such as theses, d.ssertations, foreign reports and translations, and non NAC conference proceedings are evadabio for Durchaso from the organization sponsoring the pubbcation cited.
Sing'e copios of ARC draf t reports ero available free, to the extent of supply, upon wntten request to the Office of information Resources Management, Distribution Section U.S.
Nuclear Regulatory Commission, Washtngton, DC 20555.
Copies of industry codes and standards used in e substantivo monnor in the NRC regulatory process are maintained at the NRC Library, 7920 Norfolk Avenue, Bethesda, Maryland, and are 6v634b s thua for roferenco usn by the pubbc. Codes and standards are usually copy-nghted and may be purchased from the originating organdation or, if they ero Amencen National Standards, from the Amencan Nationa! Stancards Institute, 1430 Broadway, New York, NY 10018.
NUREG-1374 Technical Findings Related to Generic Issue 79 An Evaluation of PWR Reactor Vessel Thermal Stress During Natural Convection Cooldown Manuscript Completed: September 1990 Date Published: May 1991 J. D. Page Division of Safety Issue Resolution Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission Washington, DC 20555
,p" s.
@..i6),,
ABSTRACT His report presents the technical basis for the U.S. Nu-natural convection cooldown conditions. This report was clear Regulatory Commission's (NRC's) resolution of reviewed by the NRC staff and its contractor, and conser-Generic Safety Issue 79, "Unanalyzed Reactor Vessel vative independent confirmatory stress analyses were (PWR)nermal Stress During Natural Convection Cool.
performed by the NRC contractor in selected areas. To down." included are discussions of pertinent technical complete the review an inde,<ndent fracture mechanics background mformation, the historical development of evaluation was perfor.ued by the NRC staff.
the issue, the approach of the NRC staff and its contrac-tot to the resolution, and the NRC staff technical conclu-This report presents the NRC's review and evaluation of sions with their supporting bases.
the BWOG report and the NRC's conclusion that the BWOG document, supported by the additional consena-De B&W Owners Group (BWOG) prepared a detailed tive independent analyses discussed above, provides an analysis of its 177-fuel assembly reactor vessel undcr adequate basis for the resolution of the issue.
iii NUREG-1374
CONTENTS i
t a
d Page Abstract.......................................................................................
iii Acronyms and initialisms........................................................................
vil Ackn owl ed gm e n t s...............................................................................
ix 1
In t r od u ct io n................................................................................
I 2
Background................................................................................
4 d
2.1 St. Lucie 1 Natural Circulation Cooldown Event and NRC Generic Letter 81-21..................
4 2.2 Ide ntification of G eneric Issu e............................................................
4 2.3 B & W Own ers G rou p R e po rt..............................................................
4 3
Cod es and S tandard s.........................................................................
5 3.1 Reactor Vessel Design Requirem ents......................................................
5 3.2 Fract u re Toughn ess R equirements.........................................................
5 3.3 Inservice Inspection Requirements.........................................................
5 4
Contractor Confirmatory Stress Analysis........................................................
6 4.1 S t ress Analysis R esul t s...................................................................
6 4.2 Fa t igu e Analysis........................................................................
6 4.3 Con t ractor R eport.......................................................................
6 5
NRC Staff Fracture M echanics Evaluation......................................................
8 6
Staff Concl u sions............................................................................
9 4
7 References.................................................................................
11 FIGURES 1 Typical PWR reactor vessel showing internals....................................................
2 2 Typical PWR reactor vessel shell noting closure flange i tgion......................................
3 3 NCC tmnsient profile used for BNL confirmatory analysli.........................................
7 APPENDICES A Contractor Report. " Review of Unanalyzed Reactor Vessel Thermal Stress for the B&W FA-177 Reactor Vessel Under Natural Circulation Cooldown"............................................ A-1 B Fracture Mechanics Evaluation of the B&W 177. Fuel-Assembly Reactor Vessel During a Natural Circulation Cooldown Eve n t.................................................................. B.1 C Fracture Mechanics Analysis of the B&W 177. Fuel. Assembly Reactor Vessel Head Closure Studs During a Nat ural Circulation Cooldown Event.................................................. C.1 y
ACRONYMS AND INITIALISMS ASME American Society of Mechanical Engineers j
ASME Code ASME Boiler and Pressure Vessel Code
)
ASME Sec-ASME Boiler and Pressure Vessel Code,Section III, Nuclear Power Plant Components j
tion III ASME Sec-ASME Boiler and Pressure Vessel Code,Section XI, Rules for Inservice Inspection of Nuclear tion XI Power Plant Components HNI.
Brookhaven National laboratory 3
D&W Ilabcock and Wilcox Company H& W 177 Habcock and Wilcox 177-fuel assembly reactor vessel BWOG Babcock and Wilcox Owners Group CUF Cumulative usage factor DHR Decay heat removal system GI-79 Ocneric Issue 79 GL 81-21 NRC Generic Ixtter Number 81-21 ISI Inservice Inspection LOOP loss of offsite power NCC Natural convection (or circulation) cooldown PWR Pressurized water reactor i
RCP R: actor coolant pump RCS Reactor coolant system RV Reactor vessel SRP NUREG-0800, Standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants i
10 CFR Title 10, Code of Federal Regulations, $50.55a, Codes and Standards 50.55a 10 CFR 50, Title 10, Code of Federal ReEulations, Part 50, Appendix G-Fracture Toughness Requirements Appendix G vii NUREG-1374
i ACKNOWLEDGMENTS Several NRC staff members contributed to this repon. In R. E. Johnson, Engineering bsues Branch, Division of support of the staff's overall evaluation, the Brookhaven Safety issue Resolution, RES National Laboratory reviewed the repon provided by the B&W Owners Group and performed an independent J. D. Page, Engineering Issues Branch, Division of Safety confirmatory stress analysis of the B&W 177 fuel assem-Issue Resolution, RES bly reactor vessel. The authors or contributors to this repon are:
J. D. Page, as the Task Manager, coordmated the overall review of the GI-79 issue and preparation of this report under the direction of F. C, Cherny, Section Leader, R. L Baer, Engineering issues Branch, Division of Safety Engineering issues Branch (EIB), and R. L Paer, Chief, Issue Resolution, RES EIB. R. E. Johnson of the EIB performed independent fracture mechanics analyses. J. Fletcher of the EIB pro-F. C, Cherny, Engineering Issues Branch, DMsion of vided much appreciated word processing support.
Safety Issue Resolution, RES Thanks are due to the American Society of Mechanical Engineers for its permission to reproduce certain figures E. L Hill, Program Management, Policy Development, from Section XI, Appendix A, of the 1989 Edition of the and Analysis Staff, RES ASME Boiler and Pressure Vessel Code.
I 1
l l
l l
ix NUREG-1374
l 1 INTRODUCTION Generic Issue 79 (G1-79), "Unanalyzed Reactor Vessel Since the RCPs are only prmided with offsite power,
'Ihermal Stress During Natural Convection Cooldown forced circulation ceases during a loss of offsite power (NCC)," involved the evaluation of previously unanalyzed (LOOP) event. 'Ihis resuhs in a substantial decrea ;e in thermal stiesses in the closure flange region of the liab.
the mixing between the fluid in the head and the lower cock & Wilcox 177 fuel. assembly reactor vessel (ll&W part of the RV An extended loss of RCP availabili*y can 177). A typical PWR reactor vessel, including its inter-result in a substantial difference in fluid temperatures in nals, is shown in Figure 1.
thcsc two relatively isolated parts of the RV. In some cases. a steam bubble can form in the RV head as the RCS Although in compliance with ASMll Code requirements, is depressurized.
the closure flange region of the ll&W 177 and other reactor vessels (RVs) is relatively highly stressed even under normal heatup and cooldown conditions, primarily in most cases the loss of RCP availability, usually due to a because of the high preload requirements in the closure LOOP, will be limited to a relatively short time and will studs and the discontinuities associated with any closure not require bringing the plant to a cold shutdown condi-flange design. As discussed below, additional thermal tion. In these cases the plant will be placed in the hot stresses will occur in this region if there is a significant standby mode of operation with the primary system (RCS) temperature gradient during an NCC event.
being maintained near its ne-mal power operating tem-perature and the secondary system either 6'eaming to the PWR RVs by design have a physical structure, the upper condenser or vented through the atmospheric dump support plate (see Figure 1), that impedes mixing be-valves.The return of offsite power allows th e plant to be t ween the fluid in the upper head area and the fluid in the returned to the normal power mode with no significant part of the RV below the support plate. During normal thermal transient effects on the RV.
power operations, including heatups and cooldowns, some mixing occurs between the fluids of these two areas of the RV because of the forced circulation provided by llowever, if an extended LOOP occurs, limited availabil-the reactor coolant pumps (RCPs). This forced circula-ity of secondary-side water may require the plant to cool tion and the technical specification limits on heatup and down to a cold shutdown condition. It is under these or cooldown rates ensures that RV temperature gradients, other similar circumstances that the previously unana-and thus thermal stresses, are moderate for most tran-lyzed thermal stresses may occur in the RV closure flange sients.
region (see Figure 2).
1 NURiiG-1374
1 Introduction b
b ld Eg i
'" E
'm"=i sEsa Ib.k!'} lII ' hN,,,,
M 1 on L
hs:
il c b in%c f,M,'
tries us lV
/? l
_ko i
i 9VistRW p<fl
.g yg, u dtsj. 5 5 g =- -..-
=-
3 g m _,
woi trw f
f g
con sneen t
t
(
i 3
)
Q F _cwem ese news saan N
wea.no.s civine 7
i n,en..
s
- l T
cusn morar
.n ua, e
Q i
i R R,,& =,h_ i 5
2 s
c k
-Y n
g
'Nd Y 0 wne coar tus -
N-f y
y W'L'
~~ -a. x
%g; m
\\
)'
,I1,gi l
d
/ affni W l#'041 r
aucios vissat 9
p 4
p
{
actris posi N
j
'N
).1 > # # # f 8,4
%beg,
,_z t,y <
y-. <w..
li '/
8
, N i
s p
O h
}
&&$46 $wf0lti
==
g : :
t r
qg,. >
G
'te= =ss eau
(,
g o
l, RJ (esis e onteaw=5 norm semi cau=c d
"i JI YI g... ~,
% _ yr Figure 1. Mical PWR reactor vessel showing internals 1
4 1 Introductioa
- i i
i
)
Closure Flange i
m",:,,i1}L3.
Region
"_'x I
f :
^
~~g l
s f
/
\\
a
/
6 j
]
u I d i
a si r ";
/
j E1 g
1 s
/
MA
{
l' I
A
. // 4 r
i I
1 i
i i
}
i 4
)
t 1
I i
I i
I Figure 2. Typical PWR reactor vessel shell noting closure flange region 1
i 1
i i
2 LIACKGROUND 2.1 St. Lucie 1 Natural Circulation the normal reactor cooldown mode (i.e., the transition to Cooklown Event and NRC Generic dec y heat removal (DilR) operation). Item 2 was later clarified as discussed in Section 4.
Ixtter 81-21 On June 11, 1980, an NCC event that resulted in the 2.2 Identification of Generic Issue formation of a void (steam bubble)in the reactor vessel heac' occurred at the St. Lucie Unit 1 plant. As a result of Following receipt of the B&W notification (Reference 1),
the St. Lucie 1 NCC event, the NRC issued Generic the NRC reviewed the concern as a potential generic
!Itter (G1 ) No. 81-21, " Natural Circulation Cooldown,-
safety issue.The Office of Nuclear Regulatory Rescas.'h on May 5,1981, to all operating PWR power reactor (Rl!S) evaluated its safety sigmficance and it was desig-licensees and applicants for PWR operating licenses (ex-nated as Generic Issue 79 (GI-79)in July 1983 with a cept for St.1.ucie 1). Gl. 81-21 requested addressecs to pri rity ranking of Medium, determine whether operator training and plant proced-urcs were adequate to effect a controlled NCC from 2.3 Ilp Owners Group Report operating conditions to cold shutdown.
In January 1984, the ll&W Owners Group (llWOG)initi-During its investigation into NCC conditions, the llab-ated a program to perform a detailed evaluation of the cetk & Wilcox Company (B& W) identified a concern and stresses induced in the B&W 177 closure flange region for notified the NRC (Reference 1) that thermal stresses, these transients, and in October 1984, a report (Refer-beyond those considered in the original design of PWR ence 2) documenting the evaluation results was submitted RVs, may develop in the RV flanges and studs due to to the NRC.'lhe NRC then contracted with the llrookha-targe axial temperature gradients across the RV closure ven National I;tboratory (llNI.) to assist in the review of region, i.e., a condition that was potentially outside the the llWOG report and to perform u independent con-design basis of PWR RVs. Initially, ll&W stated that firmatory computer stress analysis of the ll&W 177 clo-these thermal stresses could occur as a result of two sure flange region for the effects of an NCC transient, different transients. (1) non-uniform cooling (coolant
'lhe ll&W 177 is used in all operating Il&W reactors.
stagnation in the head) of the reactor coolant durins, an Resultsof thellNI.studyare presentedin Appendix Ato NCC or (2) after the reactor coolant pumps are secured in this report.
NURIiG-1374 4
l 3 CODES AND STANDARDS 3.1 Reactor Vessel Design of the ASME Boiler and Pressure Vessel Code as refer.
Requireme'Its enced in 10 CrR 50.55a, and SRP Sections 5.2.4, "Reac-tor Coolant Precure Boundary Insenice Inspection and in 10 CFR 50.55a, the NRC has endorsed the design Testing," and 5.3.2, " Pressure Temperature Limits."
criteria for reactor pressure vessels specified in Section l$ ('nn'aygMB " i','{o"n",d[,'*5' ',, bas don $esedies 3.3 Inservice Inspection Requirements h
V oi In 10 CFR 50.55a the NRC has endorsed the insenice 3.2 Fracture Toughness Requirements inspection 0sI) crit eria tor reactor pressure vessels speci.
fied in Section XI of the Code. De flaw size detection ne fracture toughness critcria used for the evaluation of capabilities associated with these requirements were con-the RV closure flange region, includmg the closure studa sidered in the NRC staff fracture toughness evaluations were from Appendix 0 to 10 CFR 50, Sections III and XI (see Appendices B and C),
c 5
4 CONTilACTOlt CONirlitMATOltY STitESS ANALYSIS Following receipt of the llWOO Stress analysis in Octo-calculated streves). 'the infmite friction flange interface ber 1984, the NRC Staff andits contractor (llNL) began a conditions used in Case 2 effectively coupled the closure review of this report and IINL began to perform a conser-flanges to each other, resulting in matrnum calculated vative confirmatory stress analysis of the ll&W 177 clo stress conditions for the RV i, hell.
sure flange region.
4.1 StrcSS Analysis It0SullS Ilased on the init;al review, numerous questions were transmitted to llWOG by Reference 3 and a inecting As a result of the conservatisms incorporated in the con-with IlWOO and IINL was requested. IlWOO provided firmatory analysis, some of the tesultant stresses were draft responses to the questions (Reference 4) prior to higher than those ralculated by 11& W; howes er. -H calcu.
the meetmg, which was held on April 25,1988. *lhe re-lated stresses were less than the applicab!c Co,e allow-sults of the meeting were documented in Reference 5 and able values. 'lherefore, the IlliS staff concludes that ade-IlWOO provided draft and fmal responses by Referer4ces quate design margins do exist in the closure flange region 6 and 7 respectively, in these responses, llWOG in-of the ll&W 177 for the NCC conditien analyicd. Details formed the staff that the concern originally stated with of the confirmatory stressanalysisare providedin Appen-respect to the normal reactor cooldown mo.le had been dix A.
incorrectly stated.~lhe corrected response stated that the thermal stresses of concern could occur only during.in NCC, which includes the subsequent transition to DilR 4.2 lullgtle AllalySIS system operation,llased on this information, the staff and llNL evaluation of the ll&W i17 was hmited to the ef.
- lhe llN L confirmatory analysis also evaluated the eff ects of NCC cycics on the ll&W 177 closure region with ru fccts of the NCC condition, spect to fatigue. Section 111 of the ASMll Code requites
'Ihe numerous conservatisms incorporated in the llNI.
that the total cumulative usage factor (CUF) for the RV, analysis (Appendix A) included using a maximum cool-including the closure flangen and studs, not exceed a value down rate of R)0'F/hr. Il&W advised the Rl!S staff that of LO.1he analysts determmed that 40 NCC cycles will this cooling rate is achievable in some ll&W operating contribute approximately 10% of the total CUF allowed plants under NCC conditions. 'lhe NCC transient profde by the ASMii Code for the RV studs and less than 2% of used by llNL is shown in Figure 3. Additional conserva-the total CUF allowed for the RV closure flanges. For the tisms included maintaining the fluid temperature in the ll&W 177 closure region, even with the addition of these RV head at 60tPF for the entire NCC transient, and NCC fatigue effects, the Code specified CUF is not ex.
cceded.
either a frictionless flange interface (Case 1)or an infinite friction Range interface (Case 2).
- lhe frictionless flange interface conditions used in Case I allowed the upper and lower closure flanges to reach the Following the completion of its stress and fatigue iru ly-maxirr.um possible horizontal deflections with respect to ses, llNL provided the results to the NRC by a report each other. 'lhis, combintd with a minimum allowable dated May 1989 and transtnitted by a letter dated June 19 stud effective length, was considered to proside a bound-1989. '!he llNL report is inc'uded as Appendix A to this l
ing case for the closure studs (i.e., it r:sulted in maximum report.
NURl!G-1374 6
4 Confirmatory Stress Analysis bOC STAo i
500 2500 i
i D
\\
l k
l Q400.\\
2000 q
\\
E
\\
j Tcitcf k
l 300 If00$
i
\\
hio tec 1
\\
n kJ i
\\
- - /./ktssutE M
o oo looo %
7 l
\\
i
\\
loo (00
~
l j
f l
i l
o 2.
4 6
7 F
9
/o
//-
12.
/a I4
/c Time (llours) l Figure 3. NCC transient profile used for !!NL confirmatory analysis 1
7 NUREG-1374 i
5 NitC STAFF FitACTURE MECilANICS EVAL.UATION In evaluating the fracture toughness adequacy of the 5.2.4. *'Heactor Coolant Pressure Iloundary inservice in-Il&W 177, the RiiS staff concluded that the RV closure spection andTesting." and $.3.2," Pressure Temperature region, with the ciception of the closure studs and the Ilmits."
nonle shell coun.c remained at a sufficiently high tem-perature throughout the NCC transient to avoid brittle fracture.
Compliance with the criteria utilised in the fracture me.
chames arialysis was demonstrated; therefore, applicable l'or the closure studs and the noule shell course, which regulatory RV fracture toughness requirements are Latis-can be exposed to somewhat lower temperatures, the fied for the ll&W 177 closure region for the NCC condi-staff performed individual evaluations. 'lhe evaluation tion analyzed. Appendix A and Reference 7 contain data criteria used were from Appendix 0 to 10 CFR 50, Sec-that w ere used in the fracture mechanics analyses. Details tions!!!and XIof the ASMI!Iloiterand PressureVessel of the fracture mechanics analyses are provided in Ap-CWe as referenced in 10 CFR 50.$5a and SRP Sections pendices 11 and C.
1 NUREG-1374 g
1
6 STAFF CONCLUSIONS As a result of the staffs analysis of GI-79, five conclu.
Although the llNL analysis results indicated that mem-sions were reached, some of which were amved at by brane plus bending stresses in the RV studs were 98% of qualitative extrapolation of the analysis results.1hese the ASMl? Code allowable values, the Code contains conclusions are presented below along with the staff's safety margins and certainly provides assurance of RV justification for each.
integrity for a single NCC event.
Conclusion No A NCC events of the type analyzed (i.e.,
From a review of stresses calculated by llNI. (Appendix NCC events that result in the plant bemg brought to a A) and 11& W (Reference 2), the staff dern*cd the follow-cold shutdown condition) have a low frequency of occur.
ing understanding: Under the worst condition, the princi.
rence.
pal stresses in the noizle shell course ugion remained relatively constant throughout the NCC transient. 'Ihc hasir Although numerous PWRshave experienced NCC only change was the origin of Ihose stresses.1 hat is, at the conditions, the staff is aware of only one NCC that re, initiation of the NCC, the principal stress contributors sulted in the plant bemg placed in a cold shutdown condi.
wett high pressure and telt.up loads, but at the end of tion.'Ihat event took place at St. l.ucie Unit 1 on June 11,
'hv N "C the pressure contribution had dropped mnsid.
1980.
crably, and had effectively been replaced by a thermally-induced mechanical stress.1he thermally induced rne-Conclusion No 2: 1he ll&W 177 is consi!cred analyicd chanical stress resulted from thermal contraction of the for NCC events that are bounded by the transient profile lower part of the RV w hile the RV head remained in a hot shown in Figure 3.
" expanded" state. From this the staff concluded that the highest potential for RV failure for a single NCC event Basir A masirnum RCS cooldown rate of 100'F/hr was would be when the RV metal was coldest, creating the used by llN!. in its conservative confirmatory stress analy.
limiting rtiess and temperature combination,1his oc.
sis of the ll&W 177. As shown in Section 4 and Appendix cursed nt the end of the NCC transient. For these rea.
A, applicable ASMii Section 111 allowables were met.
sons 'he staff performed fracturc loughness evaluations Additionally, the NRC staff performed a conservative for the conditions at the end of the NCC(see Appendices fracture toughness evaluation of the ll&W 177 using 11 and C).
stresses calculated by llNI As shown in Section 5 and Appendices 11 and C, the applicable ASMl! Section X1
'lhe high NCC cooldown rates evaluated by llNL and acceptance entena were satisf ed.
Il&W occurred in the early stages of the NCC event, Ho M previously stated, the worst stress and RV Conclusion No. 3: It is extremely unlikely that a single metal tunpnatwe mmWadon in We clome Dange re-NCC event will cause the failure of any U.S. PWR reactor p n was found to occur toward the end of the transient vessel. even if a cooldown rate of 100*F/hr is exceeded.
when the temperature differential between the fluid in the RV head and the fluid in the lower parts of the RV Besir lhe staff based this finding on its evaluation of was greatest. Additionally, the staff's evaluation found stresses and!racture toughness adequacy in the 11&W 177 that the metal temperatures in the RV closure flange closure flange region at the end of the NCC transient regi n were sufficiently high during the high-cooling rate when the worst case stress versus RV metal temperature p rtion of the NCC so that rapid flaw propagation, and was obtained, ASMii Code safety margins, and the simi-the potential for vessel failure from related brittle frac-luit@ctween significant 11&W 177 dimensions and those ture, would not be a concern. For these reasons, the staff of other PWR RVs (i.e., Westinghouse, Combustion lin-concluded that, even for moling rates in excess of 100'F/
gineering, and liabcock & Wilcox RVs other than the hr, a 11&W 177 would not experience a brittle fracture for gg 37g a single NCC event,
'the staff restricted its evaluation of the il&W 177 to the in order to ascertain the applicability of the results of the analyses performed on the ll&W 177 to other U.S. PWR closure flange region of the RV since it was polnled out by RVs, the staff reviawed available pertinent dimensional ll&W (Reference 1) that this was the only part of the RV which could pot entially be affected by thermal st ress from information (e.g., closure stud and closure flange dimen.
an NCC event. Temperatures and pressures experienced sions) of RVs other than the 11&W 177 and compared by other parts of the RV will not exceed those associated them to those of the ll& W 177. From this review, the staff with other previously evaluated normal operating condi-concluded that adequate geometric similarity exists to 1
tkms. An evaluation of these stresses for the ll&W 177 support extending this finding to all U.S. PWR RVs.
showed that all were within ASMil Code allowable val.
Conclusion No. 4: An NCC event that does not exceed a
- ues, total cooldown of 100*F, independent of rate, would not 9
NURIIG-1374
)
l
ll ti Conclusions L
be expected to compromise the safety of any U.S. PWit Conclusion No. A lirposure of U.S. l' Wit 1(Vs to certain reactor vessel; however, it may result in the itV being NCC transients may result in a condition that is outside l
outside its design t) asis, the f(V design basis.
Basist "Ihe detailed analyses by 11& W an1 the NI(C and its Basis: "Ihe significant thermal stresses found during the c<mtractor (learly pointed out the estremely complex 1
staffs evaluation of NCC conditions for the ll&W 177 nature of this type of analysis, it induded numerous resulted from the large temperature differential between thermal. hydraulic and mechanical modeling assumptions the hot water (or steam)in the itV head and the cooler that, although considered to be conservative, were not i
fluid in the lower portion of the 1(V. As previously dis-confirmed by specifically measured test data. Calculated cussed, this occurred late in the transient. A total bulk stress results for the ll&W 177 were as high as 98% of i
4 J
11CS temperature eduction of only 100'F would not ASMI! Code allowable values in the 11V studs. While it is result in a sigmficant head to vessel axial temperature recognlied that this Code allowabic value includes mar.
gradient. For this reaan, and since adequate geometric gins differences between stresses calculated by ll&W similarity exists between the 11&W 177 and other PWil and those calculated by IINI. pointed out the possibility of j
l(Vs(discussed in mort detail under Conclusion No. 3),it an 1(V beitig in an unanalyn d condition for certain NCC is the stafPs judgment that this finding can be extended to events, particularly for events complicated by other fac.
j all U.S. PWit itVs.
tors (e.g., stuck open atmospheric dump valve).
i 1
1 e
1 1
4 1
I 1
l i
l 1
)
t P
NUltl10-1374 10
7 RCFERENCES t
1.
I <tter for it. C. DeYoung (NRC) from J.11. Taylor 4.
Ietter for J. D. Pare (NRC) from J. it. Paljug
('11&W\\ 'Unanalyred Reactor Vessel ' thermal (ll&W/IlWOO). Apnl 8,1988.
Stress During Cooldown," htarch 18,1983.
5.
hiernorandum for R. L Ilaer (NitC) from J. D. Page 2.
- l. citer for N. P. Kadambl (NRC) from F. R. hiiller (NitC). *htinutes of hiceting With Il&W Owners (llWOG), "I'ransmittal of RV llead Stress !!valu-Group Regarding Generie issue 79," hiny 6,1988.
ation Program Results Ol&W Document Numbers 32-1151155410 and 77-1152S46-00)," October 15, 6.
letter for J. D. Page (NRC) from J. R. Paljug 1984.
(Il&W/IlWOG), June 23,19S8.
3.
letter for W. S. Wilgus OlWOO) from D. bl.
7.
14tter for J. D. Page (NRC) from J. R. Paljug Crutchficid(NRC)
- Request for hiceting With the Ol&W/IlWOO), October 26,1988.
Il&W Owners Group Regarding Reactor Vessel
' Thermal Stresses Durmg Natural Convection Cool.
8.
letter for R. L Daer (NRC) from bl. Reich (llNI.),
down-Genene issue No. 79," October 2,1987.
August 3,1990.
11 NURiiG-1374 l
API'ENDIX A IINL Report," Review of Unanalyzed Reactor Vessel Thermal Stress for the H&W FA-177 Reactor Vessel Under Natural Circulatioit Cooldown,"
dated Alay 1989 transmitted by letter dated June 19,1989
Heactor Vessel'!hennat Strew hh BROOKHAVEN NATIONAL LABORATORY QQl ASSOCIATED UNIVERSITIES, INC.
Upton, Long 100nd. New York i1973 (516) 282s Doportmont of Nucleor Eregy FIS w 2448 June 19, 1989 Mr. Joel Page U.S. Nuclear Regulatory commission Mail Stop NS217A 5650 Nicholson Lane South Rockville, Maryland, 20852 Dear Joelt We are enclosing the final copy of the report " Review of Unanalyzed Reactor Vessel Thermal Stress for the B&W FA-177 Reactor Vessel Under Natural circulation cooldown," incorporating your latest changes.
Very t
'y you s, Morri ch, Division Head Structur Analysis Division MRigts Enclosure (Original)
A-1 NUltl!G-1374
Reactor Veuel Thermal Streu REVIEW OF UNANALYZED REACTOR VESSEL THERMAL STRESS TOR THE BW FA-177 REACTOR VESSEL UNDER NATURAL CIRCULATION COOLDOWN Y.S. CHUNG, J.A. PIRES and H. REICH Hay,1989 l
1 NUREO-1374 A-2 l
Reactor Veuel Thennal Stren g
.P.
- ff.
TABLES.........................................................
vii FIGURES........................................................
viii
1.0 INTRODUCTION
1 2.0 REVIEW AND EVALUATION OF BWOG REPORT No. 77-1152846-00.........
1 21 Transition to DHR........................................
2 3.0 HEAT TRANSFER FINITE ELEKENT H0 DEL.............................
3 4.0 THERKAL STRESS FINITE ELEKENT MODEL....................
4 5.0 ANALYSIS.......................................................
4 5.1 Cenera1...................................................
4 5.1.1 Transient condition 1..............................
4 5.1.2 Transient Co n d i t i o n 2..............................
5 5.1.3 Transient Condition 3..............................
5 5.2 Description...............................................
6 6.0 RESULTS........................................................
7 6.1 Reactor Vessel She11......................................
7 6.1.1 Special Stress Limits..............................
7 6 1.2 Primary Plus Secondary Stress Intensity............
7 6.1 3 Fatigue............................................
8 6.2 Reactor Vessel Studs......................................
8 6.2.1 Average Stress.....................................
8 6.2.2 Ma x i m um S t r e s s.....................................
8 6.2.3 Fatigue............................................
8 6.3 Compliance of the kt Factor with the Requirements of 10 CFR 50-Appendix G...................................
9
' 3.1 RV Closure Head Region.............................
9 o.3.2 RV Studs...........................................
9
7.0 CONCLUSION
S REG ARDING CHAPTERS 1 THROUGH 6..................... 9 A-3 N.UREG-1374
Reactor Veuel'Ihennal Streu 00NTENTS (Continued) l E.* Ed.
l 80 RE-EVALUATION OT VESSEL STRESSES USING NEW FILH COEFTICIENT RICOMMENDED BY B&W TOGETHER WIT}l SOKE CHANCES IN 10 EFFECTIVE $TUD LENGTH..........................................
1 10 8.1 Ge n e r a 1...................................................
10 8.2 Heat Transfer Analysis....................................
11 8.3 Stress Analysis...........................................
11 8.4 Results...................................................
- 8. 4.1 Reactor Vessel She11...............................
11 8.4.1 1 Special Stress Limits.....................
12 l
8.4.1.2 Primary Plus Secondary Stress Intensity...
12 12 8.4.1.3 Tat 1gue...................................
8.4.2 Reactor Vessel Steds...............................
12 8.4 2.1 Average Stre.s............................
13 8.4.2.2 Haximum Stress............................
13 8.4.2.3 Fatigue...................................
13 8.5 Compliance of the ki Tector With the Requirements of 10 CPR 50-Appendix C...................................
13 8.5.1 RV Closure Head Region.............................
13 8.5.2 RV Studs...........................................
14 8.6 Conclusions...............................................
14 90 REFERENCES.....................................................
15 APPENDIX 1 APPENDIX 11 APPr.NDIX III APIENDIX IV APPENDIX Y
APPENDIX VI APPENDIX V11 NURl!G-1374 A-4
.ii i
Reoctor Vessel Thermal Stress TABLES
)
j i
]
Table f
f_ ale a
1 Material Properties.......................................
16 2
Cooldown Transient Analysed...............................
17 3
Tilm Coetfic1ents.........................................
17 4
Maximus Triaxial Stresses in the Head to Head-Flange l
J 1
Juncture................................................
18 l
5 Stress Intensity Range at the Head to Head-Flange
}
Juncture..................................................
18 l
6 Maximus Membrane and Membrane Plus Sending Stresses f
I in the Studs............................................
19 i
7 Maximus Triaxial Stresses in the Head to Head-Flange j
J unc t ur e - R e-e v al ua t e d Ca s e...........................
19 i
8 Stress Intensity Range at the Head to Head-Flange I
Juncture -- Re-evaluated Case.............................
20 1
9 Maximun Membrane and Membrane Plus Sending Stresses j
in the Stude -- Ke-evaluated Case.......................
20 10 Maximum Membrane and Membrane Plus Bending Stresses I
in the Studs -- Re-evaluated Case With Effective Stud l
Length = Free Length of the Stud........................
21 11 Maximure Primary and Maximum Secondary Stresses in the Stude (Membrane + Bending Stresses)...................
21 1
1 h
i, 3
i j
I 4
I i
l i
f 4
4 4
g i
il 1
.i d
4 A-5 NUREG-1374
Reactor VenelThernd Streu FIGUKES Figure 1
Finite Element Model-Node Numbers (RV Head and Head-Flange Juncture) 2 Tinite element Model-Element Numbers (RV Head and Head-Flange Juncture)
Cooldown Transient Analysed 4
Film Coefficients 5
Finite Element Model of stud Frestreesing 6
Void Level Locations (Reference 1) 7A Temperature Contours-time = 0.02 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> 78 Temperature Contours-time = 0.30 hours3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br /> 7C Temperature Contours-time = 0.78 hours9.027778e-4 days <br />0.0217 hours <br />1.289683e-4 weeks <br />2.9679e-5 months <br /> 7D Temperature Contours-time = 1.26 hours3.009259e-4 days <br />0.00722 hours <br />4.298942e-5 weeks <br />9.893e-6 months <br /> 7E Temperature Contours-time = 1.74 hours8.564815e-4 days <br />0.0206 hours <br />1.223545e-4 weeks <br />2.8157e-5 months <br /> 7F Temperature Contours-time = 2.22 hours2.546296e-4 days <br />0.00611 hours <br />3.637566e-5 weeks <br />8.371e-6 months <br /> 70 Temperature Contours-time = 2.70 hours8.101852e-4 days <br />0.0194 hours <br />1.157407e-4 weeks <br />2.6635e-5 months <br /> 7H Temperature Contours-time = 3 52 hours6.018519e-4 days <br />0.0144 hours <br />8.597884e-5 weeks <br />1.9786e-5 months <br /> 71 Temperature Contours-time = 8.58 hours6.712963e-4 days <br />0.0161 hours <br />9.589947e-5 weeks <br />2.2069e-5 months <br /> 7J Temperature Contours-time = 11.50 hours5.787037e-4 days <br />0.0139 hours <br />8.267196e-5 weeks <br />1.9025e-5 months <br /> 7K Temperature Contours-time = 12 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> 7L Temperature Contours-time = 13.0 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> 7M Temperature Contours-time = 13.7 hours8.101852e-5 days <br />0.00194 hours <br />1.157407e-5 weeks <br />2.6635e-6 months <br /> 7N Temperature Contours-time = 14.7 hours8.101852e-5 days <br />0.00194 hours <br />1.157407e-5 weeks <br />2.6635e-6 months <br /> NUREG-1374 A-6
__. _ _. _. _ ~
Reactor Vesse1 %ennal Stress 10 INTRODUCTION On June 11, 1980, an event at the St. Lucie plant resulted in the formation of a void in the upper head. This event prompted the nuclear industry to, invest-igate the problems associated with voiding in the upper head region of a reac-tor vessel.
Preliminary investigations performed by Babcock & Wilcox (B&W) on the Reactor Yessel (RV) identified that thermal stresses, beyond those consid-ered in the original design, may develop in the PV flanges atd stude due to large axial gradients across the RV.
These gradients could occur as a result of non-uniform cooling in the vessel during the transition to decay heat removel system operation or during a natural circulation cooldown. During i
either mode, a relatively stagnant area may exist in the RV upper head result-ing in poor thernal mixing between the fluid in the head and the fluid in the plenum and nostle regions of the RV.
In January 1984, the B&W Owners Group (BWOG) intiated a program to investigste the stresses produced in the RV flanges and stude due to large tesperature gradients produced by a 50'F/h natural circulation cooldown from 580'F to 180'F.
A report submitted in October 1984 suasarises the results of the pro-gram. The report is entitled, " Stress Analysis of the Reactor Vessel Closure Region for a Natural Circulation Cool Down Transient," and is identified as report No. 77-1152846-00 (Reference 1).
Some details of the analyses perform-ed in the investigation are given in BWOG Document ID 32-1151155-00 (Reference
- 2) which was submitted as an attachment to Report No. 77-1152846-00.
2.0 REVIEW AND EVALUATION OF BWOG REPORT No. 77-1152846-00 A thorough review and evaluation of the thermal hydraulics, heat transfer and stress analysis aspects of the BWOG report " Stress Analysis of the Reactor Vessel Closure Region for a Natuial Circulation Cool Down Transient" (Reference 1) was perforued by BNL.
In performing the review use was also made of References 2-10 and 14-15 for pertinent information. Appendices I through III present questions and comments that were raised during the re-view.
The major issues resulting from the review and the approach followed by BNL to perform the evaluation are discussed below:
l Twn major concerns with the BWOG evaluations ares i
1.
The background section of Reference 2 states that thermal gradients can he created "during the transition to decay heat rewoval system operation or during a natural circulation cooldown." However, the E&W analysis treats only the natural circulation transient without justifying it to be a limiting case.
1.
$sstion 4.1 of Reference 1 gives the coolant rete as 20' to 100*F/h but only a 50*F/h cooldown rate is used in the calculations.
The iustification of a cooldown rate of 50'F/h provided in Reference 1 is not conclusive. While it night be argued that for higher cooldown rates some fluid would circulate through the upper head, it is conceivable that a transient with upper head circulation and a 100*F/h cooldown rate would still l
l l
l A-7 NUREG-1374 l
Reactor Venel Dennal Stren b3 equally or more severe than a case without upper head circulation but with only 50'F/h cooldown rate. Moreover, regarding the cooldown rate Figure 1B-9 of Reference 5 shows a neteral circulation cooldown rate of 100'F/h.
The saee cooldown rate (i.e. 100'F/h) is shown in Section 6.2.2 of Reference 5 (these oro transients for cooldown f rom 8% to Of power).
As stated in Section 4.1 of R3ference 5, these transients are to be used for design purposes only and are not intended for actual transients.
Since the purpose of the report being roviewed is to show adequate design margins, it appears to be appropriate to apply a 100'F/h cooldown in the analysis of the natural circulation transient.
In addition to the above, some of the heat transfer coefficients used for the l
cnolyses, as well as some details pertaining to the stud model, are also open to question (see Appendices I through III).
However, At is not clear which of ths coefficients are important and which are inconsequential for the purposes of the evaluation analysis (see Appendix III). Since, there apparently is a significant safety margin in the final thermal stresses (Reference 1) it is the reviewers opinion that it might be simpler and less cumbersome to use a s(splified, liattinA temperature transient with 100'F/h coo,1ing, together with I
the following assumptions:
1.
Consider the inner surface of the head above the void level to be at a constant coolant temperature of 600'F (i.e. very large inside heat j
transfer coefficient).
2.
Neglect all heat lose through the outside thermal insulation (i.e. Very small outside heat transfer coefficient) in the reactor vessel head.
3.
Use one time and space variable heat transfer coef ficient for the RV inside and below the void level depending on main flow and internal vent valve position.
These guidelines were followed by BKL to independently perform a heat transfer and thermal stress analysis of the reactor vessel closure region for a natural circulation cooldown transient with a 100'F/h cooldown.
The analysis was ccrried out using the computer program ABAQUS-EFGEN (see Reference 11).
As stated in Appendix II of this report BKL was informally advised 'sy B&W that the natural convection transient was considered to be more severe than a forced flow transtant (transition to decay heat removal) system operation.
2.1 Transition to DHR Subsequently, by Reference 16 B&W stated that this concern had not been cor-roctly stated in Reference 1 and that the statement should have read:
"These gradients could develop as a result of non-uniform cooling of the reactor l
coolant within the reactor vessel that are created during a natural circula-l tion cooldown and the subsequent transient to decay heat removal operation following the natural circulation cooldown." Therefore, this concern was not ovaluated by LNL.
)
4 l
NUREG-1374 A-8 4
. ~ -, - -.. - - - - -
.I
'l Reactor Vessel'Ihcrmal Stress Since the normal transition to decay heat removal operation could be discounted, BNL performed an evaluation using the natural convection condition 1
as the basis for temperature distributicus. Moreover, the model used in the BNL evaluation differs from the structural model used by 86W in their stress 1
analysis in such a way (i.e. shorter effective stud lengths were assumed)
{
vhich lead to higher stresses in the studs.
Essed on the B&W judgement and the conservatisms introduced in the SNL model, the BNL stud stresses are i
i considered to provide conservative estimates of the stud stresses and forms a bounding condition for this transient.
l 3.0 HEAT TRANSTER TINITE ELEMENT N0 DEL
}
1 An axisyssetric finite element model of the entire RV was made using the i
ABAQUS-EPGEN computer code.
The model is comprised of 618 elements which con-l tain 712 nodes.
Tive hundred and eighty six of ths elements are four-noded (DCAX4e heat t ransf er elements. Three are four-noded axisymmetric (DINTER2AO) j interface heat transfer elements and twenty-eight are (DCID2) link transfer elements.
Tive elements are used for modeling the thickness of the RV,shell, e
i The RV head and shell flange region is modeled using six to seven elements across the thickness for the lover flange region, while ten elements are used i
to model the thickness of the upper flange.
At the intersection of the RV head shc11 and the RV upper flange the finite element mesh was refined to account for possible thermal stress concentrations. All of the modeling details are depicted in Figures 1 and 2.
J T fo-dimensional axisymmet ric interf ace heat transf er elements are used for modeling the upper flange and lower flange contact region (located on the j
inside portion of the RV). The contact between the interf aces was modeled i
takinf into account the tapering of the cladding material at the interface region. Eight link elements (DClD2) were used to model the RV stude, using i
the c:terial properties, the conductivity and specific heat for the stud mate-rial.
Twenty-one connective link elements were used in the air annulus region for modeling the studs in the RV head flange. Table 1 summarises the material 4
l properties used for the analysis.
?
i i
The natural circulation cooldown transient used in the analysis is shown in Tigure 3 and is tabulated in Table 2.
As can be seen from both Tigure 3 and Table 2, a cooldown rate of 100'T/h was considered. All the heat losses l
through the outside thermal insulation in the RV head were very small and thus l
can ne neglected.
The heat transfer film coefficient for the remaining areas of the RV were assumed invariant with time and are tabulated in Table 3.
i i
1 4
i 1
A.9 NUREG-1374
- ~.. - - _. -. - - -. -
}
Reactor Venellhermol Streu 4.0 TitEPJtAL STRESS y1N1'fE ELEKERT MODEL 4
The finite element model for the stress analysis ir almost identical to the heat transfer model. Thus, nodal temperatures computed during the heat,trans-for analysis are be used directly as input for the stress analysis. The loca-tions of all modes are the same in both the stress and heat transfer finite element models. The stress model is made up f rom 586 four-mode CAx4M amisye-metric stress-analysis finite elements. These elemente coincide with the 586 four-node DCAX4 elements used in the heat transf ar analysis. In addition four-roded axisymmetric interf ace elements (IKTEk2A) are used to model the flange contact region. The tapering of the cladding material to taken into account in determining the initial conditions for these interface closents.
Eight bona elements (521) are used to model the R7 studs. At the top of the RV studs a gap element is used to model the prestressing load acting on the stude. Two very stiff horisontal boas elements are used to represent the ef fect of the nuts, which restrain the rotation at the top of the stude and,-
apply the prastress load on each stud over a finite area.
Details of the finite element model at the top of the RV stude, including,7the initial gap for the prestressing load are shown in Figure 5.
Note that the free length of the studs used in this analysis is 33.5 inches as opposed to the 40.25 inches effective stud length used in Reference 1.
Several constraint equations between the stud nodea and the nudes on the RV flange are used to properly model the interaction between the RV flange and stude.
Specifically the following constraints are imposedt (1) identical horisontal and vertical displacements for the stud nodes and corresponding flange nodes in the lower flange (namely stud nodes 705 and 706, and flange nodes 410 and 676 in Figure 1), (2) the three flenge nodes at the bottom of the flange must define a st.raight line before and af ter the deformation (modes 409, 410 and 411 in yigure 1), (3) the three flange modes at the top of the RV t
stud suet define a straight line before and af ter the deformation. The int e r-action between the stude and the RV shell is modeled in such a manner that it results in a worst case stress condition for the cooldown transient analysed.
50 ANALYS15 j
5.1 ceneral since the exact boundary of the void level is not known, three possible loca-tions for the void level were considered in Reference 1 (Section 4) for the purpose of thermal stress determinations. These three locations give rise to three transient conditions which are defined as follows (see Figure 6).
5.1 1 yrenaient condition 1 For this case the void level in the vessel flange region is assumed to be at the base of the cover.
Since the stagnant region between the cover and the RV is only in thermal communication with the RV head fluid, this redion to considered as part of the void (see Figure 6).
NUREG-1374 A.10
Reamt VG65el Thermal Stress 512 Transient condition 2 For this case the void level is assumed to be at the head and vessel flange boundary.
This condition is considered because it is assumed to generate the usximum temperature differentials between the head and vessel flanges'and thus induce maximum membrane plus bending stress in the RV stude (see Figure 6).
5.1.3 Transient Condition 3 The assumption here is that the void level in the RV head tu at the top of the head flange elevation.
It is anticipated that this condition will generate maximum linear stress distributions in the head to head flange juncture (see Figure 6).
As mentioned, the heat transfer and thermal stress analysis described in Reference 1 were performed for the above three transient conditions. To address the impact of the QA review comments on Page 94 of Reference 2, it was deemed necessary by the authors of Reference 2 to redo the analyses previously perferued incorporating the QA review comments. However, the ncy analysse were only carried out for one of the transient conditions. The transient con-dition considered was selected based upon a review of the stresses in the RV shell and the stresses in the studs obtained in the original analyses for each of the transient conditions. Transient Condition 2 was selected.
It is t9 be noted that this location may not be the critical one for the reactor vessel shell. Ilowever, since the stresses calculated by BWOG for the reactor vessel shell (i.e., the areas not shown in Figure 6) are well below the allowables with ample margins of safety it was felt that stresses in areas away from the closure region not shown in Figures 1 and 2 need not be pursued. Therefore, in the analyses performed by BKL and reported herein, Transient Condition 2 was also selected.
Under Transient Condition 2, the void boundary location is at the head and vessel flange interface.
This position is also the location where maximum stud stresses occur (including the alternate stresses applied for the fatigue evaluation).
If the thermal gradients in the interface region were to de-crease, then, the stresses would also decrease.
The thermal gradients can decreene ift (1) the temperature in the head flange decreases, or (2) the lower flange temperatures increase.
Both of the above are related to the location of the void boundary.
If the void boundary were to move upward from the head and vessel flange interface, temperatures in the interface region of the head flange are expected to decrease, thus reducing the thermal gradients in the head flange. On the other hand, if the void boundary moves downwatd i
from the interface region, the lower flange temperatures would increase, also reducing the therati gradients. Thus, the assumed location for the void is chosen beenote it is believed to be the most critical one for the studs.
Additionally, the BNL analysis also assumed a shorter ef fective stod length and frictionlesa flange-to-flange interface.
According to Reference 2 (Page 69), the friction coef ficient for the cladding j
surfaces at the head and vessel flange interf ace is not known.with any degree
(
of certainty. This friction coefficient is necessary for the interface ele-nents used to model the head and vessel flange interface.
It is stated in Reference 2 that the gasket itself has probably a low resistance to shearing A-11 NUREG-1374 I
l
l 1 tactor Vessel'1hermal Stress i
forces and could be modeled with a friction coefficient of aero. However, the l
metal to metal contact surfaces could have a friction coefficient ranging from 0.3 to 0.74 (Reference 2).
i in the BWOG analyses (Reference 2) it is assumed that the actual stresses could be enveloped by considering two separate friction conditions for the l
selected transient condition:
CASE 1 - a friction coefficient of sero is used at the head and vessel flange interface; and CASE 2 - the radial displacements on either side of the interface elements are coupled. This approach is also l
f ollowed in the SE analyses reported herein.
For the selected transient condition then the two cases considered were CASE 1 - a friction coefficient j
of aero, and CASE 2 - the head shd vessel flange interface assumed to remain j
locked in the radial direction.
l Bh'L did review all three transient conditions referred to above as well as the two separate friction conditions for the flange interface, as submitted by l
BWOG in References 1 and 2.
From the results of that review, DE performed an independent conservative confirmatory analysis of what was considered to be i
the most limiting case.
1 The BE analyses also differs from the BWOG analyses (Reference 2) in the choice of the effective length of the studs.
In the SWOG analysis an l
effective length of 40.25 inches is used while in the SE analysis an l
effective length of 33.5 inches is used. However, the deformed length of each i
j stud was the same and equal to 0.047 inch in both the SE and SWO(i studies.
{
Therefore. the initial prestress in each stud is 34,914 psi in the SWOG study (References 1 and 2) and 43,170 psi in the BE analysis.
In the SE analysis i
the prestressing force is applied by solving an initial interference problem.
g The deformation length of the studs at prostressing is a sensured constant 1
which is set at 0.047 inches. As the other loads (internal pressure and
]
thermal gradients) act on the vessel, the studs will further be strained in a
]
nonur.iform manner along their length.
It is this additional straining of the i
studs which will be responsible for the increase in stresses in the studs.
5.2 Description 1
The cooldown transient used in the analysis is shown in Figure 3 and tabulated i
in Table 2.
A steady-state heat transfer analysis was performed for the con-l ditions at the beginning of the transient.' The temperatures calculated in the steady-state analysis are then used as initial conditions for the transient i
heat-transfer analysis under the cooldown transient shown in Figure 3.
Plots i
of the temperature contours at certain transient times are shown in Figures
}
7A-7N for upper portions of the reactor vessel. As can be seen f rom these -
figures the maximum temperature gradients occur at the end of the transient.
The end of the transient is at 15 hours1.736111e-4 days <br />0.00417 hours <br />2.480159e-5 weeks <br />5.7075e-6 months <br /> and not at 14.7 hours8.101852e-5 days <br />0.00194 hours <br />1.157407e-5 weeks <br />2.6635e-6 months <br /> (which is the j
transient time in Figure 7N). However, a comparison of the temperature dis-l tributione obtained at 14.7 hours8.101852e-5 days <br />0.00194 hours <br />1.157407e-5 weeks <br />2.6635e-6 months <br /> with those obtained at 13.7 hours8.101852e-5 days <br />0.00194 hours <br />1.157407e-5 weeks <br />2.6635e-6 months <br /> (Figures 7N and 7M. respectively) shows that the two results are very close. This anu m tem sua6 steady-state conditions have been achieved and thus it was considered that there would be no benefit 1.1 extending the analysis.
i i
i i
)
1 A-12
)
Reactor Venel Thermal Streu Once the thermal analysis was completed, the node temperatures were used as input for the thermal stress analysis.
In the stress analysis besides thermal l
loads, the prestressing forces in the stude and internal pressure acting on the vessel walls are also considered.
The prestressing forces and how they are applied in the analysis have already been discussed in Sections 4.2 and 5.1.
The pressure load is considered to be unif orm inside the RV and changes with time as shown in Figure 3 and Table 2.
The analysis for these loading conditions (i.e. thernal load, prestress load, pressure) is per(ormed f or each of the friction conditions derctibed above (re CASE 1 and CASE 2).
Section 6 discusses the stresses in the reactor vessel she?.1 and in the studs and compares them with ASHI Section 111 Code allowables.
The critical locations for strasses in the RV shell, were in the head dome to head flange area, and for the stude at the top and in the head and shell flange interface area.
6.0 RESULTS 6.1 teactor Vessel Shell For non-bolting materials, the stresses for emergency conditions must satisfy the primary stress limits given in Figure NB-3224-1 of Reference 12 and the special stress limits of NB-3227 of Reference 12, both of which are modified to meet " Level C Service Limits" of NB-3224 of Reference 12.
As stated in Reference I the only significant primary loads applied to the reactor vessel during the natural circulation cooldown transient ares intern-al pressure and stud preload.
During the natural circulation the internal pressure does not exceed the design pressure (2,500 psi). The primary stress limits for emergency condition are greater than the limits for design condi-tion.
Since for the present emergency condition the loads do not exceed the design loads and the closure analysis (Reference 14) has shown the stresses in the reactor vessel to be acceptable for design condition. The emergency condl. ion limits are also met.
6.1.1 Special Stress Limits Since special stress limits were addressed in Reference 1, they are also ad-dressed in this review. According to NB-3224.3 of Reference 12, the allowable special stress limit is 1.2 (4.0 S.), which should be greater than the tri-l axial stress (the algebraic sum of the three primary principal 6 tresses, 01, o2 and o3). For the RV shell steel the value of S, at 600'T is 26.7 kei, which corresponds to a special stress limit of 128,200 psi. The maximum tri-axial stresses at the RV head and head-flange junction are shown in Table 4.
It can be seen from the Table the special stress limit is not exceeded.
l 6.1.2 Prlaary Plus Secondary Stress Intensity l
According to Reference 2 the only transient that could cause a stress reversal in the RV head and head-flange junction region is the heatup transient. The stress intensities for the inside and outside surfaces of the RV shell for A-13 NUREO-1374
Reactor Vessellherrnal $ tress that region are given in Reference 2.
It is not known if the BNL calculated stresses coincide exactly in location with those given in Reference 2.
Also, the directions of the principal stresses for the heatup transient are not known. To estimate the maximum stress intensity range, however a combina-tion of the Reference 2 and BNL calculatod stresses must be made.
For these evaluations it is assumed that the principal stress directions are constant and that their locations in the BKL calculation are the same as those of Reference 2.
Having accepted the above, the methods of NB-3216.1 (Reference
- 12) are used to calculate the alternating stress range, rather thar ae methods of NB-3216.2 (Ref erence 12). The stress intensity ranges calculated in this manner are shown in Table 5.
As it can be seen from Table 5, the max-imum stress intensity range does not exceed the stress limit of 3 Sm (60,100 psi) spect(ied in NB-3222.2 (Reference 12).
6.1.3 Fatigue The f atigue usage f actor due to 40 cycles of the cooldown transient analyzed is 0.0166. This results in a cumulative usage factor for the RV shell of 0.0500 (using the usage factor of 0.0342, for other design considerations, f rom Ref erence 14 as quoted in Ref erence 2).
6.2 Reactor Vessel Studs The stress limits considered for this evalustion are those given in NB-3232.1 and NB-3232.2 of Ref erence 12.
6.2.1 Average Stress The maximum average membrane stresses f or the stud are shown in Table 6.
It can be seen from the Table that the maximum everage membrane stress is 49,013 pai. This stress does not exceed the 71,800 poi stress limit specified in NB-3232.1 (Reference 12) for this case.
6.2.2 Maximum Stress According to NB-3232.2 (Reference 12) the maximuu service stress at the periphery of the stud cross-section must not exceed 107,700 psi. The calcu-lated maximum stress at the periphery of the stud cross-section for the cool-down transient analyzed, is shown in Table 6 to be 104,027 psi.
It does not exceed the service stress limit of 107,700.
6.2.3 Fatigue The maximum primary plus secondary stress intensity range calculated for the emeTyred trarsient occurs at the location of maximum stud stress and equals 43,434 psi, the maximum dif f erence between membrane plus bending (ref er to Case 1, Table 6).
For 40 cycles of the cooldown transient under consideration the fatigue usage factor is 0.0548, resulting in a cumulative usage factor for the kV studs equal to 0.753 (using the utage factor in the RV studs for other desi n considerations of 0.698 f rom Ref erence 14 as noted in Ref erence 2).
F Reactor Yessel' :rmal Stress l
I l
NUREG-1374 A-14 l
l
__mh
j Reonor Vessel'Ihernet Stress I
6.3 Compliance of the ki Tactor with the Requirements of 10 CFR I
_50-Appendix G i
Since the additional thermal stresses due to this emergency event are located I
in the RV head to head-flange juncture (i.e., the areas shown in Figure 6, i
including the studs), the stress-intensity factor only needs to be considered in those locations and be compared with requirements given in Appendix G - 10
)
CFR 50 to ensure adequate margins.
J 6.3.1 RV Closure Head Region 1
See Section 8.5.1 below.
6.3.2 RV Studs The BNL computation has shown that the maximum membrane and bending stress I
during the worst case for the event is 104,027 psi (Table 6).
B&W calcula-tionn gave a corresponding stress value of 69,486 psi (Page 5-3 of Refe'rence 1).
i In Re-ference 15, B&W has shown that the service stress.'Imit at the periphery of the stud cross-section is 107,700 psi.
Thus the stream value has not been exceeied even with the more conservative BNL assumptions. Accordingly, the requirements of 10 CFR 50 - Appendix G will not be compromised during this emergency event.
1
7.0 CONCLUSION
S REGARDING CHAPTERS 1 THROUGH 6 l
An independent evaluation based on a conservative assumption of the reactor
{
vessci thermal stress due only to natural convection cooldewn for the Babcock and Wilcox FA-177 reactor vessel was performed by BNL.
Based on the review of the 86V Owners Group (BWOG) evaluation (References 1-3, 6, 8 and 10) and related material, the BNL evaluation is considered to represent a worst case situation. As indicated above, the BNL analysis results satisfy the specified ASHE ode criteria.
The closest approach to the allowables occurred on the st.., which exhibited a maximum service stress of 104,027 psi vs. the service Ifmit of 107,700 psi.
Since this assessment was based on the very conservative assumptions of zero friction, at the head and vessel flange interface this close result to the allowable is deemed acceptable.
It should be noted that according to the BWOG reports, the friction at this interface can vary from 0.3 to 0.7 thus lowering the stud stresses. Further, since these conservative results based on linear methods met all criteria, it was considered unnecessary to pursue non-linear evaluations. Thus, although the natural circulation cooldown transient produced high stress levels herein, they were nevertheless found to be within acceptable limits.
It recains to be noted that in the BNL analysis a f atigue usage f actor due to 40 cycles of cooldown transients was assumed. This is double the number of cycles used by BWOG in their analysis and thus the BNL evaluation incorporates a factor of safety of two on the cycles. As shown in Section 6.2.3 the current assucptions result in a total cumulative usage factor for the studs l
that is equal to 0.753.
A-15 NUREG-1374
Reactor Vessel 1hennel Stress Finally, it should be pointed out that the analysis performed herein is based on axisymmetric model assumptions. As such, localized three-dimensional effects are ignored. These effects could impact the fatigue evaluations locally increasing the usage factor in the reactor vessel shell but probably not the usage factor for the studs. However, since the cumulative usage fac-tor for the reactor vessel shell and the fatigue usage factor for the analysed thermal transient are ema11, the effecto probably will not impact the overall safety of the vessel assembly.
8.
RE-EVALUATION OF VESSEL STRESSES USING NEW TILM COETTICIENT RECONKENDED l
BY B&W TOGETHER VITH SOME CHANGES IN ETTECTIVE STUD LENGTH As a consequence of the BNL review described in Chapters 1 through 7 above, several questions pertaining to the BWOG reports were sent via the NRC to BWOG (see Appendices 111 and V).
BWOG responded to the NRC questions in a letter f rom Mr. J.R. Paljug (BWOC) to Mr. Joel Page (NRC) dated October 26, 1988 (see Appendix V).
BWOG's response to Question 11.5 of Appendices III and V was,1 that the outside film coefficient for the res: tor vessel surface below the-upper head region given in References 1 and 2 is too large by an order of mag-nitudv.
BWOG's response to Question 111.7 of Appendix V states that the prestress in the studs is calculated based on test measurements and that the calculated value is 34,914 psi, not 43,170 poi as used in the BNL analysis (see Assumption 8.3(1) in the pages that follow). To assess the impact of the error for the overall outside film coefficient (below the upper head region),
BNL performed new heat transfer and theraa1 stress analysis using the following assumptions:
8.1 General A simplified limiting temperature pressure transient with 100'F/h cooling instead of 50'F/h cooling as used by BWOG (see Figure 3) was used in the analysis.
8.2 Heat Transfer Analysis (See Figure 4 and Table 3) 1.
Considered the inner surface of the head above the void level to be at a constant coolant temperature of 600'F.
Assumed a very large inside film coefficient (both assumptions the same as BWOG).
2.
Assumed a very small outside film coefficient for the RV upper head.
This implies very ses11 heat losses through the thermal insulation for the RV uppi head (this assumption is the same as BWOG).
3.
Used a constant film coefficient for the RV inside and below the void level (differs from BWOG model which assumed that the film coefficient changes with time and space).
4.
Used a constant film coefficient for the outside surface of the reactoryessel. Thevalueforthiscoefficienywas0.0016 miU/h-in
'T.
It replaces the 0.0197 BTU /h-in
- F value used in the original analysis, which was found to be in error by an order of g
magnitude. The value of 0.0016 BTU /h-in
'T was chosen for the teactor head in order to have a uniform value for the entire vessel.
(See Appendix VI and response 11-5, Appendix VII, NOTE: Units in Appendix VI are B11J/h-f t
- F.)
NUREG-1374 A-16
_-. = - - - _ _ - _ - _ _
t Heanor Vessel Thermal Stress 83 Stress Analysis 1.
Used an effective stud length of 40.25 inches (the free length of the stud plus one-half of the engaged length) in order to calculate the prestress lotd.in the stud.
In this manner the preload value matcheb the test value given by BWOG.
2.
Interaction between the stud and flanges i
s.
Identical horizontal and vertical displacements for the stud l
nodes and corresponding flange nodes on the lower flange (i.e.,
i flange node 410 has the same displacements as stud node 705; sim-ilarly, flange node 676 and stud node 706 have the same displacement; see Figure 1).
j b.
The three flange nodes at the lower end of the stud model define a straight line before and af ter the deformation (i.e. nodes 409, 410 and 411 in Figure 1).
c.
The three flange nodes at the top of the stud define a straight line before and af ter deformation (nodes $32, 520 an1526 in Figure 5).
3.
Transient condition - void levels i
l The void level is assumed to be at the head and vessel flange inter-face. This condition is assumed to generate maximum teuperature dif-ferentials between the head and vessel flanges and thus induce maxi-l mum membrane plus bending stresses in the studs (it is not known with 1
certainty where the actual location is).
4.
In the BNL analysis approach, it was considered that the actual stresses could be enveloped considering two separate friction conditions for the head and vessel flange interface: CASE 1 - a friction coefficient of zero, CASE 2 - the head and vessel flange j,
interface are assumed to remain locked in the radial direction (same es BWOG). According to the BWOG report being revie,ed, the friction coefficient for the cladding surfaces at the head and vessel flange interfaca is not known with any degree of certainty. It is stated in the BWOG report that the gasket itself has probably a low resistance to shearing forces and could p~robably tc eedeled with a friction coef ficient zero, while metal-to-metal contact surf aces quald have a friction coefficient ranging from 0.3 to 0.74.
8.4 Resuate 8.4.1 Reactor Vessel Shell For non-bolting materials, the stresses for emergency conditions must satisfy the primary stress limits given in Figure NB-3224-1 of Reference
- 12 and the speef al stress limits of NB-3227 of Reference 12, both of which are modified to meet " Level C Service Limits" of NB-3224 of Reference 12.
A-17 NUREG-1374
Heactor Veuel *lhermal Streu As stated in Reference 1 the only significant primary loads applied to the reactor vessel during the natural circulation cooldown transient are internal pressure and stud preload. During the natural circulation the internal pressure does not exceed the design pressure (2,500 psi). The pri-mary stress limits for emergency condition are greater than the Ifnits for design condition.
Since for the present emergency condition the loads do not exceed the design loads and the closure onalysis (Reference 14) has shown the stresses in the reactor vessel to be acceptable for design condition. The emergency condition limits are also met.
8.4.1.1 Special Stress Limite According to NB-3224.3 of Reference 12, the allowable special stress limit is l.2 (4.0 f ), which should be greater than the triaxial stress (the alge-braic sum of the three primary principal stresses, 01, 02 and 03).
For the RV i
shell steel the value of Sm at 600'T is 26.7 kei, which corresponds to a special stress limit of 128,200 psi. The maximum triaxial stresses at the RV head and head-flange junction are shown in Table 7.
It cap be seen from the Table that the special stress limit is not exceeded.
8.4.1.2 Primary Plus Secondary Stress Intensity According to Reference 2, the only transient that could cause a stress reversal in the RV head and head-flange junction region is the heatup transi-ent. The stress intensities for the inside and outside surfaces of the RV shell for that region are given in Reference 2.
It is not known if the BNL calculated stresses coincide exactly in location with those given in the Ref-Also, the directions of the principal stresses for the heatup transi-erence.
ent are not known.
To estimate the maximum stress intensity range, however, a combination of the Reference 2 and BNL calculated stresses must be made.
For these evaluations it is assumed that the principal stress directions are i
constant and that their locations in the BNL calculation are the same as those of Reference 2.
Having accepted the above, the methods of NB-3216.1 (Reference 12) are used to calculate the alternating stress range, rather than the methods of NB-3216.2 (Reference 12). The stress intensity ranges I
calculated in this manner are shown in Table 8.
As it can be seen from Table 8, the maximum stress intensity range dees not exceed the stress limit of 3 Sm (80,100 psi) specified in NB-3222.2 (Reference 12).
8.4.1.3 Fatigue The fatigue usage factor due to 40 cycles of the cooldown tiansient analyzed is 0.0166. This results in a cumulative usage factor for the RV shell of 0.0508 (using the maximum usage factor of 0.0342 from Reference 14 as quoted in Reference 2).
8.4.2 Reactor Vessel Stude The stress limits considered for this evaluation are those given in NB-3232.1 and NB-3232.2 of Reference 12.
NURI'.O-1374 A-18
____._____m._
Reactor Vessel 1hermal Stress I
4 i
i j
8.4.2.1 Average Stress l
J l
The maximum averaged membrane stresses for the stud are shown in Table 9.
It can be seen f rom the Table that the maximum everage membrane stress is 39,510 l
psi. This stress does not exceed the 71,800 psi stress limit specified in j
NB-3232.1 (Reference 12) for this case.
I i
8.4.2.2 Maximum Stress 2
l According to NB3232.2 (Reference 12) the maximum service stress at the periph-i ery of the stud cross-section must not exceed 107,700 pai. The calculated maximum stress at the periphery of the stud cross-section for the cooldown transient analyzed, is shown in Table 9 to be 92,650 psi.
It does not exceed the service stress limit of 107,700.
)
In addition to the above calculations which are based on the B&W stud lengths, i
another calculation using the original BNL effective stud lengths was made,.
J Tor this esse the maximum membrane plus bending stress at the periphery of the stud cross-section in 105,100 which is less than the limiting value of 107,700 l
psi.
8.4.2.3 Fatigue
)
The maximum primary plus secondary stress intensity range calculated for the analyzed transient occurs at.the location of maximum stud stress and equals 45,050 psi, the maximum dif ference between membrane plus bending (refer to
]
l Case 1. Table 9).
For 40 cycles of the cooldown transient under consideration j
the fatigue usage factor is 0.0583, resulting in a cumulative usage factor for i
the RV studs equal to 0.756 (using the maximum usage factor in the RV stude of 0.698 of Reference 14 as noted in Refsrence 2).
e l
l It can be seen from Table 10 that the maximum primary ano secondary stress intensity range remains the same regardless of effective stud length i
l assumptions.
i l
8.5 Compliance of the kg Factor With the Requirements of 10 Crt 50 -
Appendix G i
Since the additional thermal stresses due to this emergency event are located j
in the RV flanges and the stude, the stress-intensity factor only needs to be constoered in those locations and be compared with requirements given in Appendix G - 10 CFR 50 to ensure adequate margins.
l 8.5.1 RV Closure Head Region X3 values are a function of the prima'y plus secondary loading. Expressed r
i in equation form 1
- 2(K ) pressure + (K ) thermal + (K ) residual K
1 1
1 where (K)) pressure corresponds to the primary load and (Kg)therani +
(Kg) residual correspond to the secondary loads. As can be seen the K3 i
+
j A-19 NUREG-1374 i
J i
Reactor vessel'thennal Stress t
j due to the primary loads has a weight factor of two in the above equations.
Reference 15 has shown the adequacy of this region for the design and normal operating conditions.
For the void event pressure is reduced, the ther' mal stress conditions increase, and residual stresses rossin essentially unchanged. When changes in pressure and thermal stress are equivalent the not effect is a reduction of the K value as can be seen by the equation.
Even for a condition where the increase in thermal stress is twice the decrease of the pressure dependent stress the K3 value remains unchanged.
Der condition lies between these two conditions and thus Kg is somewhat less than the one i
given in Reference 15 and the requirements of Appendix C of 10 CFR 50 are not l
compromised.
I 8.5.2 RV Studs The BNL computation has shown that the maximus membrane and bending stress during the worst case for the event is 92,650 psi (Table 9) or 105,100 psi (as shown in Table 10).
B&W calculations gave a corresponding stress value of 69,486 poi (page 5-3 of Reference 1).
In Reference 15 B&W has shown that the limiting stress value at the periphery of the stud cross-section is 107,700 psi. Thus, the stress value has not been exceeded even with the more conservative BNL assumptions. Accordingly, the requiroecnts of 10 CTR 50 - Appendix C will not be compromised during this emergency event.
8.6 Conclusions The conclusions of Section 7.0 of the BNL report are not altered by the new analysis except for (1) the cumulative fatigue usage factor for the reactor vessel studs increases very slightly from 0.753 to 0.756, (2) the closest i
approach to the allowable stresses occurs in the stude which exhibited a maximum service stress of 105,100 psi (see Table 10) vs. the service limit of i
107,700 psi. Thus, the re-evaluation of the thermal stresses using the more accurate film coefficient for the outside surface of the reactor vessel below the upper heat region results in adequate margine of safety.
t l
t A-20 NUREO-1374 y
...g.-.c-,,__.,.n..-..,~,=-,em.
,.,.vv._,
-.,_.-.---.m,-_--.,,-
w-,
,,w_,--.
J Ileactor Venel Thermal Stress j
I
]
9.0 REFERENCES
1 J
1.
" Stress Analysis of the Reactor Vessel Closure F:gion For a Natural j
Circulation Coo? Down Trai.a;.'nt
- Labcock & Wilcox Ovvete Group Analysis Committee, Report No.
77-l'*,'A'-
00, Jul;, 1944, l
2.
" Stress Analysis of the Aer' twt " 3sr1 Closurs Ragior L. i a Natural Circulation Cool Down T ins!r*,t." 3**.vock & 4 ;;ccx, f. Avul $2f >n No.
j 32-1151155-00, June 21, 1984.
3.
B.L. Boulman
- Reactor \\s Jsel Head 56 seM a Anal!.1. Inputs," babcock &
Wilcox, Calculation No. aJ-1150499
- f, March 10 1964.
I j
4.
Crystal River Unit 3 FSAR.
j j
5.
General Functional Specification ist R.C.S. Components for CPCO, t
18-1092000012-05.
1 l
6.
R.W. Winks and R.C. Tuilley, Jr., " Natural
'culation Occurrences at Operating B&W Plants,' labcock & Wilcox Co., day 7,1979.
~
7.
N.Y. Simms. *RETRAN-02 Comparison of Natural Circulation Flow Rates at i
B&W 177-TA Plants," Proceedings of the Third International RETRAN Conference, EPRI NP-3603-SR, 1985.
8.
" Reactor Vessel Head Stress Analysis Inputs," for all 177-FA Owners Group contracts, Babcock & Wilcox, N6W Celculation No. 32-1160499-00.
9.
J.P. Holman, Heat Transf er, McGraw-Hill Book Company,1963.
10.
"CPCO R.V. CROM Hotions," H5512 to 13 contracts. R&W Calculation No.
32-1140915-00.
I i
11.
43AQUS-EPCEN - User's Manual. Electric Power Research Institute, October l
1982.
12.
ASHE Boiler and Pressure Vessel Code,Section III, Subsection NB,1983 Edition.
l
- 13. ASME Boiler and Pressure Vessel Code, Appendix 1,1983 Edition.
1 14.
Stress Report for Reactor Vessel, Consumera Power Company, Midland Unite I and II (Record Center Microfile Rolls 80-1-80-2).
i 15.
Methods of Compliance with Fracture Toughness and Operational Requirements of 10 CTR 50, Appendia 0 - B&W report 10046, Rev. 2, December 1984 by H.W. Behnke et al.
16.
Letter for J.D. Page (NRC) from J.R. Paljus (B&W), October 26, 1988.
17.
Ou111eret, J.-C., "Re-examining Reactor Vessel Rebrittlement at ChoosA ",
Nuclear Engineering International, pp. 44-46, November 1988.
1 l
l A-21 NUREG-1374 i
Reacto? Vessel nermal Stress TABLE 1 - MATERIAL PROPERTIES COEFFICIENT YOUNG'S OF THERMAL SPEC?FIC E
FOISSON'S CONDUCTIVITY HEAT M&TERIAL TEMP MODpUS,E 10{FANSION
(*F)
(10 psi)
(in/in/ 'F)
RATIO BTU /h-in *F BTU /lb *F Reactor 100 29.8 6.50 0.3 1.975
.107 Vossel 200 29.5 6.67 0.3 2.0
.115 A-508 CL.2 300 29.0 6.87 0.3 1.99
.120 3/4Ni-1/2Ho 400 28.6 7.07 0.3 1.97
.125
-1/3 Cr-V 500 28 0 7.25 0.3 1.93
.128 600 27.4 7.42 0.3 1.87
.133 Stude 100 29.8 6 27 0.3 1 64
.107 A-540 Gr.23 200 29.5 6.54 0.3 1.72
.115 2Ni-3 /4 Cr-300 29.0 6.78 0.3 1.77
.320 1/4MO 400 28.3 6.98 0.3 1.78
.125 500 27.4 7.16 0.3 1.78
.128 600 26.7 7.32 0.3 1.77
.133
' skirt 100 29 8 5.53 0.3 1.99
.107
- & 516 200 29 5 5.89 0.3 2.03
.115 Cr. 70 300 29.0 6.26 0.3 2.03
.120 C-Mn-Si 400 28.6 6.61 0.3 2.02
.125 500 28.0 6.91 0.3 1.975
.128 600 27.4 7.17 0.3 1.925
.133 0.964 0.052 Equivalent 100 0.976 0.056 Material for 200 0.972 0.059 Thermal 300 0.960 0.061 Modeling of 400 0.939 0.0o2 Closure Head 500 0.911 0.065 Flange and 600 Air Annulus i
(1) From Reference 13.
(2) From Reference 2.
i NUREG-1374 A-22 4
4 1
l Reador Venellhennal Stren i
J f
TABLE 2 - COOLDOWN TRANSIENT ANALYZED TEMPERATURE
- l T' iJ (F')
TIME PRESSURI j
("..u : s )
(Hours)
(Psig)
VOa0 HOT LEG COLD LEG O
600 580 530 0
2,500 1
2.6 600 320 270 2.6 1,000 11.4 600 320 270 7.5 1,000 14.0 600 200 140 11.0 350 j
15.0 600 200 140 15.0 350 l
- The temperature outside the RV was assumed constant and equal to 120*F.
4
)
4 TAB?.E 3 - FILM COEFFICI*NTE FILM COEFFICIENT FILMCOEFFjCIENT ID NO.*
IDENTIFICATION *
(BTU /h-in
'F) f VOID 0.799 1
INSIDE RV 0.521 2
j OUTSIDE (RV UPPER HEAD) 0.0016 3
OUTSIDE (OTHER THAN 0.019:
4 RV UPPER HEAD)
- See also Figure 4.
4 L
A-23 NUREG-1374
Rextor Vessel Thermal Stress TABLE 4 - MAXIMUM TRIAX1AL STRESSES IN THE HEAD TO HEAD-FLANGE JUNCTURE TRANSIENT MAXIMUM TRIAX1AL STRESS
- ELEMEl(T CASE TIME (hours)
(psi)
NUMBER 1
0 73,340 504 15 55,853 504 2
0 59,914 504 15 80,820 504
)
The triaxial stress is defined as the sum of the three principal stresses TABLE 5 - STRESS INTENSITY RANGE AT THE HEAD TO HEAD-FLANCE JUNCTURE OUTSIDE SURFACE STRESS-INTENSITY * (ps1)
TRANSIENT 0 -73 0 -02 CONDITION CASE 03-01 2
1 Cooldown 1
-48,676 28,290 22.466 2
-59,680 39,340 18.305 NO STRESS REVERSAL Heatup 57,680 39,340 22,466 Range c01>02>03 INSIDE SURFACE STRESS-INTENSITY * (psi)
TRANS1Ehi
- '41 0 -02 CONDITION CASE 01-03 2
3 Cooldown 1
33,411
-17,440
-21,348 2
49,743
-30,721
-25,904
-12,200 8,800 19,500 Hut up 61,943 39,520 45,404 Range Col 1 021O!
NUREG-1374 A-24
- =.
Reactor Venel Thermal Stress 1
l j
TABLE 6 - MAXIMUM MEMBRANE AND MEMBRANE PLUS BENDING STRESSES IN THE STUDS l
TRANSIENT EMBRANE STRESS MEMBRANE PLUS BENDING STRESS CASE TIME (hours)
(psi)
(psi) a b
j 1
0 49,013 60,593 1
15 42,656 104,027 1
l 2
0 48,902 77,918 15 42,656 77.934 4
i PREOTTESS 43,170 74,350 a
l I
i J
TABLE 7 - MAXIMUM TRIAXIAL STRESSES IN THE EAD TO HEAD-FLANGE JUNCTURE -
RE-EVALUATED CASE STRESSES (See Section 8.0) 1 TRANSIENT MAXIMUM TRIAKIAL STRESS
- ELEMENT j
CASE TIME (hours)
(psi)
NUMBER 1
0 72,640 504 15 55.540 504 l
2 0
58,020 504 15 83,250 504
- The triaxial stress is defined se the staa of the three principel stresses a
l 4
{
A-25 NUREG-1374
Reactor Vessel Thermal Stress TABLE 8 - STRESS INTENSITY RANCE AT THE HEAD TO HEAD-FLANCE JUNCTURE RE-EVALUATED CASE STRESSES _ (See Section 8.0)
OUTSIDE SURFACE - RE-EVALUATED CASE STRESSES STRESS-INTENSITY * (psi)
TRANSIENT 0 -02 02 -03 CONDITION CASE 03-01 1
Cooldown 1
-45,060 21,870 23,810 2
-58,010 37,960 20,015 NO STRESS REVERSAL Heatup 58,010 37,960 23,810 Range 001102183 INSIDE SURFACE -- RE-EVALUATED CASE STRESSES STRESS-INTENSITY * (psi)
TRANSIENT 0 -02 02-03 CONDITION CASE 01-03 3
Cooldown 1
31,700
-16,750
-14,496 2
50,030
-30,230
-20.070
-12,200 8,800 19,500 Heatup 67,230 39,030 39,570 Range 001102103 TABLE 9 - MAXIMUM MEMBRANE AND MEMBRANE PLUS BENDING STRESSES IN THE STUDS -
RE-EVALUATED CASE STRESSES (See Section 8.0)
TRANSIENT MEMBRANE STRESS MEMBRANE PLUS BENDING STRESS CASE TIME (hours)
(psi)
(psi) 1 0
39.510 47,600 15 34,810 92,650 2
0 39,420 62,210 15 34,830 63,810 PRESTRES5 34,914 61,880 NUREG-1374 A-26
N Reactor VesselDermal Stress TABLE 10 - MAXIMUM MEMBRANE AND MEMBRANE PLUS BENDING STRESSES IN THE SfUDS -
RE-EVALUATED CASE STRESSES WITH EFFECTIVE STUD LENGTH = FREE LENGTH OF THE STUD (See Section 8.0) i TRANSIENT MEMSRANE STRESS MEMBRANE PLUS BENDING STRESS CASE TIME (hours)
(psi)
(psi) 1 0
46,750 60,070 15 42,050 105.100 2
0 46,670 74,590 15 42,070 76,190 PRESTRESS 43,170 74,350 TABLE 11 - NAXIMUM PRIMARY AND MAXIMUM SECONDARY STRESSES IN THE STUDS (MEMBRANE PLUS BENDING STRESSES TRANSIENT PRIMARY STRESS SECONDARY TOTAL STRESS CASE TIME (hours)
(ksi)
STRESS (kei)
(ksi) 1 0
55.02
-2.19
.52.83 15 40.05 6.23 46.28 2
0 43.90
-1.42 41.48 15 38.75 27.30 66.05 A-27 NUREG-1374
)
l Reactor Vesselnerrno! Stress
?
6:
i I
N se y
6d4 WtMI ta b ED Sid
]
58 I
5d1 E IN:
NL 5tB sql 4
9 gg jy 4,8 6dt sp02 Na B&B 51t 1
llj) lin BJ6MB 633 na s;I sgl a
j
%@je BdD SJ 4WR Nt fil Elp y% o<s sie sau; Ni si n ses I
%' ELMbN TN3 Ell 13XL XI E l M4 413 511 elsMt al l 415 elf ell M 7 11 E3 El c
di STB sah 81) all all all adp d1R ein elf N
l
.z j
.,..n i.ii.n
....n
.o
.n 45NL eli $1L 814 all 4' O 411 491 415 45d g
j I
e4bMaMR55 410 433 diI alt 4E6 4E5 415 3
ha allEl edD B12 44 4 Mt ed3Me 44) edB ed'I sitat 41 1ma a)( j4s
- 41. 41 B ell 4J7 ei n a
. \\/
sn at i ss D ata att are j
/24 54 4&2 58 4E3 40 g.8 6al - 4 37 gg aggg 4 pg 4
i ia al l als elleil 4(7 ell all i
o S;:
l 4,e tes ecsni sage di t 41:
Il!
b 4$$
MkM!
b NI l
11: 31 4 Alt Ely El E L E7 1
4
}
, a,, <
\\
l A-28 NUREG-1374
!lI1 l
0
[a !SWi 5 i.
h W
l E l i
k.n.
o i
i l 5i E i
i i i 5
i l
l l
)
i c -
l I l E
i i
p' -
F l
l i
i i
i 4
i i
l l
l E
h p
hFE E
PE E
0 E
h P
R hEPD F
h E
h P
P F
h PD E
P h
E P
n a.
EE P
E E
N R
E a.
PE E
N E
N F
a.
h PR W
N E
E E
P E
m h
N PE F
N E
E F
P h
5 EE E
E E
E E
E n
ir(
E a
h p
E E
N E
P E
F
.n p
E E
h E
E
~
E p
N F
N E
E n
.n p
N E
N R
P 2
P p N
h P
N n
N p
E N
P E
Mp E
g%
E p
N g
p F
,C OL.5 yu l
t
Reactor Vessel Thermal Stress 717eso
..A y_
l l
i l
90
\\
\\
mk Q400.\\
T ar u g R
K acc
\\
Teen ter t
\\'
)
200
\\
S&cssus
\\
\\
/00
\\
~
i o
z 4
6 7
8 9
/r>
//
Iz
/s 14 IE Time (Hours)
Figure 3 Cooldown Transient Analyzed NUREG-1374 A-30
Reactor Vessel 1hermal Stress I
3 k....
5-1
't fl Cl l u
r
$y
'4 is 0I,'
@Q'4,ib h @
g
- g..-
@t M@
@ F...
g 7
'L 'T..
T III Tr,-
o g-y I:::
x
.2
-it..-
w I:::
L s l I kI Jf ie z
4 4
4 4
Figure 4. Film Coefficient (Refer to Table 3)
{
i j
A-31 NUREG-1374 l
Reactor Vesseinermal Stress 680 a\\ (E s{2 d@
)
nii) s20
~
~
S26 l
597 o 7/2 E6m d Nsk l
Nods 6%1 adm y
n,a 44 39.f 6,et-s97 7/2.
Cao s20 212.ao o 322-Eo o g2 4-wj-s95 sa2 524 y26 1$h B2 +-Mb 519 5*'
32
.qog SAS NO j
7/2.
3 / g.g3 0
-( k ree) l fe44;aeMuc-o. fos I
j a
1 j
Figure 5. Finite Element Model of Stud Prestressing i
j 1
I NUREG-1374 A-32
Reactor Vtssel Hermal Stress 1
l t
1 3
M+m' Transient Condi ion 1 3
$2 ; 1 I
Ii a.
i
~I e
i 2
t/1 i Transient Condition 2 ll l l Il l
Transient Condition 3 i
l\\1 /
i l
l l
l l
l l
l Tigure 6 Void Level Locations (Reference 1) l l
t I
A-33 NUREG-1374
W D
2 O.,
i h,
S
?
E
~
w Y
"a E
B i
i L
l TEMP.
p I.D. YR. UE I"
T 1 +1 00 E-02 i
2 +1 50 E+02 1
i j
3 +2 00 E+02 l
f 4 +2 50 E402 5 +3.00 E+02
\\
i i
6 +3.50 E+02 l'
7 +4 00 E+02 g
8 +4.50 E+02 9 +5 00 E+02 10 +5.50 E402 I
/
1I +6 00 E+02 e
i 1
Figure 7A. Temperature Contours-time = 0.02 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> i
. =. -.
. _. -. ~. -.. _. -
i i
l Reccter Vessel Dermal Stress a
1 i
i i
P 4
)
4 s'
i a
l i
i h
4 1
w -
i w
4 j
'~~Myg y
= _..g..gm e
- = - -
.g 4
C f*1 e
A e
i k
i.
U l
4 44 1
W D0 4
M C
0 w
D u
80W G
i m
5 4
Gw D
I to wb i
i i
i k
f f
A-35 NUREG-1374 l
4
' Reactor Vesse!7hermal Stress
\\
N W
D O
.C:
to tw e
O E
U 1
to4 DCu C
8
.4 D
U 4k&
i i
tJ N
e I
k3 l
80 w
IEe J
I f
i i
a a
NUREG-1374 A-36
Reactor Vessel hermal Stress
\\
.w
.E noeemm a
I w
U ln4eo ueou e
kawab
-4 s
I I
i ci 6
1.
1 w
E i
w h
i i
1 4
j i
l i
r i
1 l
A-37 NUREG-1374 l
l t
i f.
=. _ _ _ _ _.
Reactor Vessej hermal Stress
-- --_ y w
ho w
W I
'9 mage,
N.
=
4 s
3 8u
?
5 5
N Co E
f NUREG-1374 A-38
1 1
Rextor Vessel'lhernd Stress k
i i
)
4 i
1 1
1 i
i il
,- =
b g
e k
i h....... e a F
~
+
a i
1 M
i M
W D0 u
C 8
t 4
.W Q
u N
WW H
4 w'
l o
W9 i
i i
b i
1 j
i-i 4
f 1
4 A-39 NUREG-1374 4
Reactor Vessel Thermal Stress 4
I a
1 l
j I
l f
i l
i A
o O
.C
- i
,A j
a = = = e e %'
s I,
.I I
U I
g' to W
j a
i o
"c I
"o e
w Da
.s i
Mb 3
e
'I
- ~
5
)
e t
d i
6
)
8
=
t w
n 1
ee i
M>
i l
I i
1 i
I
)
i 1
)
i l
5 i
I i
I NUREG-1374 A-40 i
l
=
..n.
i i
]
Reactor Vessel Thermal Stress
.i 2
'I a
9 i
(
1 i
i 4
m 4
W 1
f
.c j
\\
l n
4 m
e s
-~
O
~
l N
u., e.._..- -.
i a
I 4
40 W
1 a
i' O
u c:
6
.S
.W 1
1 E
i u
(6p y
H.
~
4 i
1 e
I 3*=
5
.w3
]
00 w
1 EAe l
I 1
i l
l l
l 1
1 1
j i
i o
3 l
1 j
A-41 NUREG-1374
~~..-
Reactor Vestellhermal Stress v
\\
o f
~ ^ ~
hf od T.
v b oer = M"~
a w
Mi Mu i
ac o
e i
1 w
D u
0 e
W r
e 5
H t
J
^
i l
,w aM w
$4 i
i I
i 1
3 i
I i
l 1
1 NUREG-1374 A-42 I
I
i
.I t
Reactor VesselThermal Stress l
1 4
1 1
9 i
s i
1 s
I i
1 li OW D
l
\\
3 g
]
M-i,
_~ _.
(
a 1
___,,,,,,,p
,3 a
1 1i O
u i
C 4
oo GW3 u
4 k
43 Q.
5
\\
e 4
D.
4 6
G i
4 r
i 1
1 l
i i
1 1
+
i n
I 1
A-43 NUREG-1374
Re:ctor Vessel Therm:1 Stress e
40
\\
j i
i 9
_- x s,
=
_g b eongnoon#
g D
b 8
e U
G 2
A.
.F.
H 5
m J
NUREG-1374 A-44
Reactor Vessel Thermal Stress
@W D
a a S o
M 1
MW D
Q u
C O
O
/
m D
rs W.
Q.
EG H
.3 e-WH D
00 w
$4e I-4 i
1 i
l i
i l
l l
A-45 NUREG-1374 l
r L
_. ~. -.. - ~.. -.. =. _... ~ _ - - - -. _.
Reactor Vessel Dennal Stress E
E.c 5
5
\\
e t
5
,s Y
a y
b c-e n = = ; Ml
,E c
0 t
B B
es H
ai 5
0h u.
I l
l l
l NUREG-1374 A-46
.~
i
)
}
}
iMactor Vessel *Ihermal Stress i
t j
i 1
a 1
i 1
3 a
g i
W 8
\\
4
- n. -
a t
.a en j
% w go a
g l
ec o a' e n. e..
w y
I 10 1
W l
D0 T
uc0 U
G k
'Du EU l
b e
.b N
5 1
Uw
{
D 4
no 5
w i
i 1
l t
i i
l s
4
(
l 1
J l
)
A-47 NUREG-1374 1
m
F l
Recctor Vessel nermal Stress i
4 i
i i
k 9
}
l a
l I
l l
i i
l l
APPENDIX I
~1
'l E
i e
4 4
i I
NUREG-1374 A-48
Reactor Venel nermal Stress BROOKHAVEN NATIONAL LABORATORY MEMORANDUM DATE:
February 5,1986 Te:
Files FROM:
W. Shier i
suesccT:
Ruview of Thormal-Hydraulic Aspects of B&W Stress of Natural Circulation cooldown Transient This mamo provides comments and observations that were generated during a review of the thermal hydraulic aspects of a Babcock and Wilcox (B&W) stress analysis of the reactor vessel closure head during a natural circu-lation cooldown transient (Reference 1).
(This work was regbested by Dr.
J. Pires). The information available for this review consisted of two B&W reports (No. 77-1152846-00, Reference 2 and No. 32-1151155-001, Reference 3); however, several other reports, that were referenced extensively, were unavailable.
Thus, the results of the initial review could be subject to some modification if the additional documents are obtained and reviewed.
The following paragraphs summarize the comments that have been obtained thus far.
l 1.
The Background section of Reference 2 states that large thermal grad-1ents and stresses could be generated in the reactor upper head region dur-ing a natural circulation (NC) cooldown transient or during the transient preceding operation of the decay heat removal system.
However, only the natural circulation transient has been considered in the stress analysis; no justification is given for the use of this transient as more limiting than the decay heat removal system transient or cther possible system tran-sients.
2.
Reference 2 states that the reactor coolant system can cooldown at a rate up to 100*F/hr.; in addition, the " General Functional Specification for R.C.S. Components" (4) is referenced as including a cooldown rate 100*F/hr for an NC transTent.
However, the stress analysis is based on a 50*F/h'r. cooldown. Justification should be provided for the reduced cooldown rate.
t 3.
Figure 4 of Reference 2 provides the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> pressure and temperature transients that have been used in the stress calculations.
If the tempera-ture transients are assumed to be correct (see comment above), then the pressure transient associated with the hot leg temperature transient ap-pears to be conservatively high.
The degree of hot leg subcooling has been calculated at various times in the transient and is of the order of 100*F to 200*F indicating that the saturation temperature associated with the as-sumed pressure transient is high. However, more information is required to i
obtain a more definitive evaluation of the transient.
In addition, actual plant data on natural circulation transients hac been obtained for several l
B&W plants ( Arkansas Nuclear One, Crystal River, Davis-Besse and Oconee, l
References 5 and 6) and could be useful if further evaluation of the as-l sumed pressure and temperature transien s is desired.
Recdor Veml'1heract Streu 4 Page 11 of Reference 3 states that the inner surf ace film heat transf e.-
2 coefficient for the reactor vessel heaJ is 115 BTV/hr ft.*F for the upper 2
region and 2.83 BTV/hr-f t ef for the lower region.
These values were ob-tained from a reference that was not available during this review.
How-ever, if significant voiding were to occur in the upper head reg 10n during tnis transient, the film coefficient would be expected to be smaller in the voided (upper) region than in the solid or two phase (lower) region.
The determination of these film coefficients should be reviewed in more detail and additional information would be required.
4 5.
Page 17 Reference 3 computes an equivalent heat storage capacity of the reactor head flinge (CEQ) where CEQ = 0.44 Cmetal.
The coefficd et (0.44) represents the ratio of the area occupied by the 9
flange metal to the total flange area and
, meant to exclude the closure stud holes.
However, Cmet3) or CEO can be cilculated directly from geomet-ica? considerations and does not need to be reduced for the effect of the stod Poles.
In addition, it is not clear if, or how, CEQ is used in the analyses.
6.
On pages 19 and 20, an effective film coefficient is computed to repre-sent radiation heat transfer (with fourth power temperature dependence) by an equation that is linear in temperature, in addition, in the computation of this effective film coefficient, temperatures and temperature differ-ences are assumed without justification.
The use of this constant, effec-tive film coefficient ignores the fourth power temperature dependence on the heat transf er rate and will only be correct when the structure tempera-tures are at the values assuned in the approximation.
in addition, it is not clear why T,= 450*F (or 910*R) and AT = 50'R or 40'R were used. These assumptions could be evaluated through a sensitivity study or at a minimum, checked for their validity after the stress calculations were performed.
)
7.
A conve.tive heat-transfer coefficient for the closure head (identified as Fluid Block 1) is computed nn pages 39 through 44.
Several comments are applicable to this calculation.
- The miss flow rate in this region is an "med tc be 8% of the natural circulation 1000 flow.
However, no just. 'ica*. ion is provided. When g
the 8% flow rate is converted to an equivalent 100% flow rate in Ibm /
nr, tne refult is sonewhat high when compared with the data used by BR in other R5W analyses. This would tend to make the heat transfer coefficient high.
- The velocity in this region is co.1puted using a flow area that was computed for a dif ferent region (Fluid Block 4),
a t
NUREG-1374 A-50
4 Veurt hermal Streu
- The characteristic length 1. is defined as De L*T in the heat transfer correlation; howe'ver, in the Reynolds Number cal-culation, it is used as
- 1. = De.
- The report states that the characteristic length is 2,5 feet but a value of 0.125 is used to compute a " conservatively high" heat transfer coefficient.
However, the reduced length results in the conclusion that the correlation is not valid and, therefore, riot used.
(The use of a length of 2.5 feet would reverse this conclusion). The analyses should be performed with consistent values of length and flow area.
- A natural convection correlation is evaluated using a 15'r difference (assumed to tie between the metal and the fluid) without justification or verificatien after the stress calculations were completsd.
- An independent calculation of the natural convection coefficient, us-ing Reference 7, gave a heat transfer coefficient of approximately 300 2
BTU /hr-ft.or which is the same order of magnitude as calculated by B&W.
8.
The evaluation of film coef ficient for Fluid Block 4 assumes the riow rate to be 87, of the natural circulation loop flow rate without j tifica-tion. However, in this case; the correlation used is'for forced convection as opposed to the natural convection correlation used for Fluid Block 1 wi5h the same ficw rate.
9.
Tnere is a general co ment on the film coefficient calculations for each of the fluid blocks: the geometry bein cult to charactcrize by lengths, flow areas,g analyzed appears to be diffi-etc. and it would be desirable to know the sensitivity of the most limiting stress calculations to the heat transfer parameters and to the assumed thermal-bydraulic transient used as a boundary condition.
This sensitivity would provide information on the accuracy required in the preparation of the thermal-hydraulic input parameters.
These commer.ts have resulted from the review of the information provided in Refarence 2 and 3.
It is recormended that a more detailed review be per-formed ' hat includes an independent verification of the natural circulation pressure and temperature transient (possibly tnrough a comparison with plant data), a review of the supporting B5U documents and an independent ev luation of the heat transfer coefficients.
WS/lr cc:
J. Guppy J. Pires M. Reich U RoSatgi A-51 NUREO-1374
~4*
Reactor Vet.sellhermol Stress References 1.
F.R. Miller, "Transmitts) of RV Head Stress Evaluation Program Re-suits," Letter to N. Prosed Kadambi, B&W Owner's Group October 15, 1984 2.
B&W Owner's Group Analysis Conmittee, " Stress Analysis of the Reactor Vessel Closure Region During a National Circulation Cooldown Tran-sient " 77-1152846-00, Babcock & Wilcox, (1984).
3.
A.D. Nara, " Stress Analysis of the Reactor Vessel Closure Region for a j
flatur:1 Circulation Cooldown Transient " Babcock and Wilcox General Calculation 32-1151155-00, June 1984 4
" General functional Specification for R.C.S. Components for CRPCo,"
IS-1092000012-05.
5.
R.W. Winks, R.C. Twilley, Jr.. " Natural Circulation Occurances at Operating B&W Plants," Babcock and Wilcox Co., May 7,1979.
6.
N.T. Simms, "RETRAN-02 Comparison of Natural Circulation Flow Rates at B&W 177-FA Plants," Proceedir.js : Third International RETRAN Conference. EPRI NP-3803-SP, (1985).
7.
J.P. Holman, " Meat Transfer," McGraw-Hill Book Company, (1963).
1 J
f l
NUREG-1374 A-52 1
I s
i J
Recetor Vessel Thermal Suess l
4 1
i J
t f
i i
1
.i 1
j 1
4 i
l 1
i l
4 i
i 1
i APPENDIX II 1
(
I
?,
l 1
i i
j I
I 4
a a
s f
4 i
i i
P i
I 4
I i
l A-53 NUREG-1374 I
Reactor Venet nermal Streu DROOKHAVEN NATIONAL LADORATORY MEMORANDUM DATE:
October 6, 1936 To File rnoM:
P. G. Kroeger
/
sunsccT:
Review of Ther 1-Hydraulic Aspects of B&W Stress Calculations During Upper Head and Reactor Vessel Cooldown in a previous review of tnis subject by W. Shter (Ref.1), it was con-ciud:d thct several crucial references were not available and that a full evaluation was, therefore, not possible.
Some of these references have now been provided, and a somewhat more coTplete judgement is now possible.
The two major points raised in Ref. I were as follows:
1.
The background section of Ref. 2 states that thermal gradient 3 can be created during the transition to decay heat removal system operation or during a natural circulation cooldown." However, the B&W analysis treacs only the natural ccnvection transient without justifying it to be a limiting case.
2.
Secti on 4.1 of Ref. 2 gives the coolant cooloown rete as 20' to 100'F/hr.
But only a 50'F/hr cooldown rate is used in the calcula-ti on s.
In brief and informal telephone conversations with B&W engineers (Garry L. Weatherly, Charles Tally, and Athok D. Nana), the writer was advised that the natural convection transient was considered to be the more severe one since during it the upper head fluid would remain more stagnant, while under forced flow cooldown conditions about 7 to 97. of the flow would reach the upper head.
Regarding the cooldown rate, it was felt that for natural circulation cooldown even 50'F/hr would be high.
While these arguments may well be correct, they are not necessarily conclusive.
Regarding Point 1, from the given in f ormati on, including a review of several BAW SARs we were not able to establish when and why any fluid would circulate through the upper head.
And even if that is so, a transient with upper head circulation and a 100*F/hr cooldown rate might still be equally or more severe than a case without up;.er head circulation but with only 50*F/hr cooldown rate.
Thus, if the natural convection transiert is indeed the more severe one, this should be demonstrated more conclusively.
NUREG-1374 A-54
l Reactor Vessel 1herinal Stras Memo Kroeger to file October 6,1986 Page 2 Regarding Point 2 Ref. 3, Figure 18-9 shows a natural circulation cool-down rate of 100*F/hr.
The same is stated in Section 6.2.2.
These transients are for cooldown from 8% to 0% power.
As stated in Section 4.1 of Ref. 3, these are transients to be used for design purposes and do not intend to be actual transients. However, as the thermal transients of the current work are used to show adequate design margins, it would seem appropriate to also apply 100'F/hr cooldown in the current natural convection transient.
Once we know why t'here is more upper head circulation in the forced convection cooldown, it may indeed be acceptable to consider a 100'F/hr natural circulation cooldown as the limiting case.
However, this would also depend on the action of the iaternal vent valves during these transients, which is also not clear at this time (see also below).
Many of the heat transfer coef ficients being used in the analysis, as well as some details of the stud model, could be questioned.
But it is not clear which of these are important, and many may be inconsequential- (see addi-tional detailed comments below).
In particular, since there apparently is a significant saf ety margin in the final thermal stresses, it might be simpler and more convincing to use a simplified, limiting temperature transient, for instance, as follows:
1.
Consider the inner surf ace of the head above the void level to be at constant cool ant temperature of 600'T (i.e., infinite inside heat transfer coefficient).
2.
Neglect all heat loss through the outside thermal insulation (i.e.,
zero outside heat trensfer coefficient).
3.
Use one time and space variable heat transfer coefficient for the RV inside below the void level depending mainly on main flow and inter-nal vent valve position.
Several further points appeared questionable during this review, and since it is not practical to check all computations, these are not necessarily c ompl et e.
1.
The above void level heat transfer coef ficient of Ref. 6, Fluid Block 1, Transient 3, Pgs. 40-43 is based on elaborate computations, but the ultimate coefficient used (natural convection) is controlled by assumed temperature dif ference of 15'F and a vertical height of an plate of 50 in., while the correlation is then apparently used for the complete upper void surf ace (or is ttris value only used in that part of Block I which is below the void level?) (see sketch on Pg.
37).
As this number may be significant since it ilmits coolant to metal heat transfer, more carefully obtained numbers (or an upper limit, see above) should be used.
Is it indeed sufficient to use a single constant heat transfer coefficient for this region?
A-55 NUREO-1374
Reactor Vessel nermal Stress Memo Kroeger to File October 6,1986 Page 3 2.
Since most of the heat transfer coefficient computations use correla-tions out of B&W design handbooks, it is not practical to check all of these in detail.
3.
Pr.ge 5 of Ref 4 deduces a top region (Fluid Block 1) heat transfer coef ficient of 115 BTV/f t: hr/F for a "non-venting case", to be used by Re f. 6.
But Ref. 6 considers cpen vent valves (see Pg. 55, with Fluid Block 5).
15 there a conflis ?
'hiat is the ef fect of vent valves on Fluid Block 1 heat transfe. Mrf ficients, if anyl 4
Ref. 2. Section 2 refers to a " void" f srhing in the upper head, i.e.,
saturated vapor?
Ref.
4, Figure 3 gives the film coef ficient as being f or subcooled water, S.
The overall U valves prepared in Ref. 4 for use in Ref 6 were inad-vertently misquoted in Ref. 2 (Pg. 4-3) and Ref. 6 (Pg.11), which shou'.d read:
t RV head inner surface 115 BUT/ft2 hr F;'
RV head outer surface 0.23 BTU /ft2 hr F; (upper region) '
RV head outer surface 2.83 BUT/fta hr F; (lower region) i 6.
The outer heat flows ssed in Ref. 4 to deduce overall effective U's are varying spatially, and it appears doubtful that this averaging done here is meaningful.
However, it 'may have little effect on the results.
One would expect that the thermal insulation in the main thermal resistance, and since that is apparently 3 in, thick every-wnere (Table 4-3 and Fig. 4 4 of Ref. 7) one would expect it to not vary by more than one order between top and sides.
What is the l
material, what are its k and c7 l
l 1
l i
NUREG-i374 A-56
= -. -
Reoctor Vessei nermal Stress Memo rioeger to file October 6,1986 Page 4 References 1.
W. Shier, Memo to Files,
Subject:
" Review of Thermal Hydraulics Aspects of B&W Stress of Natural Circulation Cooldown Transient," February 5, 1986.
2.
B&W Own er 's Gr oup Analysis Committee, " Stress Analysis of the Reactor Vessel Closure Region During a National Cli*culation Cooldown Transient,"
77 1152846-00, Babcock A Wilcox, (1984).
3.
"Tunctional Specification 18-1092000012-05 f or Reactor Coolant System".
I 4
B.
L.
- Bowman,
" Reactor Vessel Head Stress An aly sis Inputs," Calc.
1 32-1150499-00, B&W, March 18, 1994 J
5.
V.
A.
Hand i ek ar, "Str es s Analysis Report No.
1, Closure Analysis for Consumer Power Company Midlant Pl ant Unit B&W Contract No.
l 620-0012-51, October 1973.
6.
A. D. Nana, " Stress Analysis of the Reactor Vess.el Closure Region For a i
Natursl Circulation Cooldown Transient," Babcock and Wilcox General Calcu-Iation 32-1151155-00, June 1984 7.
Crystal River Unit 3FSAR.
PGX/lh cc:
J. G. Guppy i
M. Reich J. Pires I
i l
A-57 NUREG-1374
Reactor Vessel Thermal Stress APPENDIX III NUREG-1374 A-58
1 Reactor Vest.el'Ihermal Stress b
BROOKHAVEN NATIONAL LABORATORY Q
ASSOCIATED UNIVERSITIES, INC.
Dep r a nt
'ucles n rgy Upton. Long Island, New York 11973 Building 129 7933 (Si6) 282 s F T S 666' Hay 26,1987 Mr. J. Page Division of Engineering Of fice o! Nuclear Regulatory Research Hail Stop 212 U.S Nuclear Regulatory Commission Washington, D.C. 20555
Dear Hr. Page:
Enclosed please find the set of questions for the Babcock & Wilcox Owners Group (PWOG) pertaining to the review of the documents entitled:
" Stress Analysis of the Reactor Vessel Closure Region for a Natural Circulation Cooldown Transient," BWOG, Report No. 77-1152846-00, dated July 1984; and
" Stress Analysis of the Reactor Vessel Closure Region for a Natural Circulation Cooldove Transient." BWOG, Calculation H2. 32-1151155-00 dated June 21, 1984.
Please contact me if there are any questions.
Sincerely yours, s4 k Jose Pires bee enclosure A-59 NUR11G-1374 l
___ _J
ltractor Vcuel'lhennat Stren UNANALYZED REACTOR VESSEL Ti!ERMAL STRESSES (FIN A-3822)
INTRODUCTION On June 11, 1980, an event at the St. Lucie plant resulted in the forma-tion of a void in the upper head. This event prompted the nuclear industry to investigate the problems sesociated with voiding the upper head region of a reactor vessel. Preliminary investigations performed by Babcock and Wilcox.
(B&W) on the reactor vessel (RV) identified that thermal stresses, beyond those considered in the original design, may develop in the RV flanges and studs due tu large axial temperature gradients across the RV.
These gradients could develop as a result of non-uniform cooling of the reactor coolant within the vessel that is created during the transition to decay heat removal system operation or during a natural circula; San cooldovu.
During *eithet mode, a relatively stagnant area may exist in the RV upper head region resulting in poor thermal mixiag between the fluid in the head and the fluid in the plenes and nossle regions of the RV.
In January 1984, the B&W Owners Group (BWOG) initiated a program to investigate the stresses produced in the RV flanges and studs due to large temperature gradienta produced by a 50'F/hr natural eftculation cooldown from 580'F to 180'F.
A report submitted in October 1984 summarizes the results of the prcgram. The report is entitled, " Stress Analysis cf the Reactor Vessel Closure Region for a Natural Circulation Cooldoen Transient", and is identi-fied as report No. 77-1152846-00 (Reference 1).
Some details of the analyses performed for report No. 77-1152846-00 are given in the BWOG Doc. ID 32-1151155-00 (Referencs 2), which has been submitted as an attachm:nt to report No. 77-1152846-00.
Reference 1 through 10 were the documents used for this review.
QUESTIONS AND COMME!US 1.
General
- 1. The background section of (BWoG) report being reviewed states that thermal gradients can be created during the transition to decay heat removal system operation or during a natural ciculation cooldown.
However, the (BWOG) analysis only addresses the natural circulation cooldown transient.
Please provide justification for not considering the transition to decay heat removal system operation in the analysis.
- 2. Section 4 of the report being reviewed gives the coolant cooldown rate as 20' to 100' F/ hour. Only a 50' F/ hour cooldown rate is used in the analysis. Please provide justification for selecting only the cooldown rate of 50' F/ hour for the analyais.
NUltliG-1374
Reactor Vessel %ermal Stre:2 i
II. Heat Transfer Analysis I
J
- 1. Provide justifiestion for selecting a temperature difference of 15' F and a vertical plate height of 50 inches to compute the void level host transfer coefficient for Fluid Block 1, transient 3 (pages 40 to 43 Reference 2).
)
Can the film coefficients cr.1culated for the void level for all the transient conditions be considered as upper limits?
1
)
cos a top region (Fluid Block 1) heat transfer 2.Page5ofReference3dedy/h*/'Tfora"non-ventingcass,"tobeusedin coefficient of 115 BTU /ft j
Re,ference 2, but Reference 2 considers open vent valves (ree Page 55, with Fluid Block 5).
Is there a conflict? What is the effect of vent valves on the Fluid Block I heat transfer coefficient.
- 3. Section 2 of Reference 1 refers to a "coid" forming in the upper head, is i
th1. superheated stone or saturated vaport Figure 3 of Reference 3 gives i
the flim coefficient as being for subcooled water. Why was this value j
chosen?
I j
- 4. The overall U values prepared in Reference 3 for use in Reference 2 were j
inadvertently misquoted in Reference 1 (Page 4-3) and Refstence 2 (Page 11), which should reads a
2 RV head inner surface 115 BUT/ft hr FI 2
l RV head outer surface 0.23 BTU /ft, hr FI (upper region) j RV head outer surface 2.83 BUT/ft hr Fg (lower region)
- 5. The outer heat flows used in Reference 3 to deduct overall effective U's I
are varying spatially. What is the effect of the averaging done heraf One l
would expect that the thermal insulation is the main the:3a1 resistance, and since that is apparently 3 lu.
thick everywhere (Table 4-3 and Figure 3
4-4 of Reference 4) one would expect it not to vary by more than one order of magnitude between top and sides. What is the insulation asterialf What l
are its therasi conductivity and emissivity?
- 6. On pages 19 and 20 of Reference 2 an effective film coefficient le oosputed to represent radiation heat transfer. Provide justification for selecting i
T2 = 450'F (or 910'R) and AT = SO'A or 40'R for computing that affective l
film coefficient.
4 l
i A 61 NUREG-1374
1(cactor Venel 7 amal Streu II.
Heat Transfer Analysis (tont.)
- 7. A convective heat transfer coef ficient for the closure head, identified as Fluid Block 1 is computed on pages 39 through 44 of Reference 2.
Several questions are applicable to this calculation.
A.
Provide justification for assuming the mass flow rate in this region to be 81 of the natural circulation loop flow.
B.
Justify computing the velocity using the flow area that was used for Fluid Block 4.
C.
The characteristic length L is defined as L = De/2 in the heat transfer correlation. Why is it used as L = D in the Reynolds e
Number calculation?
D.
Are the values of length and flow area used consistent with each other?
E.
Provide justification for using a 15'T difference between the metal and the fluid.
- 8. Provide justification for assuming the flow rate to be 8% of the natural circulation loop flow rate in t:
evaluation of the film coefficient for Fluid Block 4.
Provide Justific, tion for using a correlation for forced convection as opposed to the natural convection correlation used for Fluid Block 1 with the same flow rate.
III. Thermal Stress Analysis
- 1. Since the inlet and outlet openings in the reactor vessel are localized, provide justification for the choice of a finite element model for the stress analysis which is axisymmetric about the reactor vessel vertical axis. Also, provide details regarding the degree of accuracy in the modeling of the temperature and film coefficients within the vessel as axisymmetric.
- 2. Provide justification for not including in the enalysis the reaction forces on the reactor nozzles.
- 3. In Section 9.0 of Rafarmace 2 it is sentioned that the thermal stress analyses were performed for three structural models, identified as caso 1, case 2 and case 3.
For all the stress analyses the shear stresses in the vessel closure bolts have not been reported. What is the magnitude of the shear stresses in the vessel closure bolts for case 1, which uses a f riction coef ficie nt of zero et the flange interf aces? Included in the above question are the thermal stress runs made to comply with review corment 2 and 6 (page 94 of Reference 2).
1 NUl(EG-1374
^42
l Reactor VesselDermal Stress III. Thermal _ Stress Analysis (cont.)
1
{
- 4. The radial displacements of the stud node 921, (yigure 4 of Reference 2, page 26) and the flange node 259 at the interface, were coupled to allow i
for stud shest force and bending soment to be transmitted to the vessel flange at two locations (page 96 of Reference 2).
Subsequently, the i
thermal stress analysis for transient condition 2 and the structural casse j
1 and 2 was repeated.
Why is a linear constraint equation such as the one at the bottom of page 34 of Reference 2, not provided between nodes 259 and I
its adjacent shell nodest Is addition, why are nodes 259 and 921 not j
coupled in the vertical direction?
)
i
- 5. For the reanalysis done to comply with review comments 2 and 6 the reactor i
vessel was assumed to be free to grow radially at the bottom of the nostle 3
belt region. Does this this boundary condition have any effect or limiting i
the strains in the vessel and on the magnitude of the shenr forces in the vessel closure bolts?
In addition, it is not totally cient from the answer to review comment 3, page 96 of Reference 2, if the reaction of the plenus cover on the uppsr head was considered in the initial thermal / stress analysis, or only in the analyses performed af ter the review comments.
)
- 6. In Table 1 (page 22) of Reference 2 the coefficients of thermal expansion, j
a, are average coefficients of thermal expansion in going from 70'F to the indicated temperature.
Provide justification for using the average values in lieu of the instantaneous values.
{
- 7. On page 27 of Reference 2 the effective length of the bolt was assumed to 1
be 40.25 in.
Provide justification for choosing this bolt length to j
calculate the bolt prestress?
i l
i t
j A-63
Recetor VesselTherinal Stress REFERENCES
[1]
" Stress Analysis of the Reactor Vessel Closure Region for a Natural Circulation Cooldown Transient," Babcock & Wilcox Owners Group Analysis Committee, Report.4o. 77-1152846-00, July 1984
[2]
" Stress Analysis of the Reactor Vessel Closure Region for a Natural Circulation Cooldown Transient," labcock & Wilcox, Calculation No.
32-1151155-00, June 2, 1984 I
3.L. Boulsan, " Reactor Vessel Head Stress Analysis Inputs," cal:.32-115
[3)* 0499-00, R&W, March 18, 1984.
[4] Crystal River Unit 3 TSAR
[5] Ceneral Tunctional Specification for R.C.S. Components for CPCO.
18-1092000012-05.
[6)
R.W. Winks, R.C. Twilley, Jr., " Natural Circulation Occurrences at Operating B&W Plants", Pabcock & Wilcox Co., May 7,1979.
1 (7)
N.T. Simms "RETRAN-02 Comparison of Natural Circulation Flow Rates at R&W 177-FA Plants", Proceedings:
Third International RETRAN Conference, EPRI NP-3803-SR, 1983.
[8] R&W Calc. 32-1150499-00, " Reactor Vessel Head Stress Analysis inputs",
For all 177-TA Owners Croup Contracts.
[9]
J.P. Holman, " Heat Transfer", HeCraw-Hill Book Company,1963.
l
[10) Babcock & Wilcox Calculation 32-1140915-00, "CPC. R.V. CRDH Hotions" i
H55 12 & 13 Contracts.
r 4
i i
?
i l
t i
NUREG-1374 A-64
_ _ _ _.. _ _ - ~.., _
I Reactor Venel Thermal Stress APPENDIX IV i
5 l
i 1
I I
t l
l l
l A-65 NUREO-1374
~.
Reactor Veucilhermal Streu 8abeock & Wileox avo..ere-.,o.a.i.a e McDettnott company 3315 Old rotest Road April 8, 1988 P.O. Dos 10935 ESC-308 tynchbuts. VA 24506 0935 (604) 385 2000 i
l Hr. Joel D. Page Task Manager, Section B, Eng'ineering Issues Branch Division of Licensing office of Nuclear Regulatory Research U.Se Nuclear Regulatory Commission Washington, D.C.
20555 Dear Mr. Paget Please find attached draft responsen to the NRC questions regarding Generic Issue No.
79, RV Thermal Stresses.
These responses were prepared by B&W for the B&W owners Group and are proliminary to the requested meeting of April 25.
Please review these responses in preparation for that meeting.
r very truly yours,
[
J R. Paljug Project Manager Owners Group Engineering Services ORP/leh l
Attachment cc:
B&WOG Analysis Conitt.g.g C. H. Turk
- AP&L P. F. Guill
- DPCo J. E.
Burchfield, Jr. - DPCo E. H. Davidson
- FPC A. Irani'
- GPUN R. Little
- SMUD R. H.
Bryan
- TVA J. F.
Dunne
- TED NUREG-1374 A-66
b a
Reactor Vessel 7hermol Stress Draft Hesconse to NRC Ouestions V
I.
Egneral ouestion 1 The background section of Reference 1 states that thermal gradients can be created during the transition to decay heat removal system or during a natural circulation cooldown.
- However, the analysis only addresses the natural circulation cooldown transient.
Please provide justification for not consid-l l
ering the transition to decay heat removal system operation in j
this analysis.
Additionally, please provide justification for j
considering only 20 natural circulation cooldown cycles.
1 i
Besconse 1 Transition tb the decay heat removal system was I
considered.
The pressure vs temperature curve used for the J
stress analysis input, Reference 2 Page 97, includes transition to the decay heat removal system (DHRS) beginning at approximate-i ly 11 hours1.273148e-4 days <br />0.00306 hours <br />1.818783e-5 weeks <br />4.1855e-6 months <br />.
Total primary system flow before and after DHRS J
actuation are estimated to be quite similar.
Natural circulation flow is estimated at approximately 8500 GPM, and DHRS flow is approxirmately 7500 GPM.
Consequently, the change in flowrate into the upper head region following actuation of the DHRS would not be significant.
l A-47 NUREG-1374
Rextor Vessel Derraal Stress The design basis of the NSSS is 20 natural circulation cooldown cycles.
To this date no operating Dr.W plant has performed a natural circulation cooldown.
BriW does not foresee the 20 cycle limit to be exceeded by any plant during its lifetime.
Ouestion 2 Section 4 of Reference 1 gives the coolant cooldown rate as 20 F to 100 F/h; however, only a 50 F/h cooldown rate is used in the analysis.
Please provide justification for selecting l
only the cooldown rate of 50 F/h.
A natural circulation cooldown of the NSSS would Resronse 2 only occur when offeite power is lost for an extended period of time.
The loss of offsite power precludes the use of the turbine bypass valves for cooldown of the primary thus requiring a cooldovn with the atmospheric dump valves (ADV's).
The total capacity of the ADv's in B&W plants in general does not exceed 10% of the total rated steam flow, thus limiting the long-term primary natural circulation cooldown rate to 50 F/h or less.
II.
ligat Transfer Analysis Question _1 Provide justification for selecting a temperature difference of 15 F and a vertical plate height of 50 inches to compute the void level heat transfer coefficient for Fluid Block 1,
transient 3 (pages 40 to 43 of Reference 2).
Can the film NUREG-1374 A-68
3 l
1 Reactor Vessel Hermal Stress I
I i
coefficients calculated for the void level for all the transient conditions be considered as upper limits.
k 4
Response
1 A vertical plate height of 50 inches roughly 1
l corresponds to the elevation of the liquid level above the plenum j
cover.
For transient condition 3,
the fluid block element j
temperatures below the steam void level are assumed to equal the i
core outlet temperature.
l The temperature differential ~between the liquid and the wall was i
l chosen to be 15 F, which is a reasonable approximation based upon
'i l
the detailed finite difference fluid model which was used to l
generate the inner surface film heat transfer coefficient l
l (Reference 3).
I I
The filn coefficient calculated on pages 39 - 43 of Reference 2 l
is only applicable to fluid block i for transient condition 3.
i The film coefficient for the voided region is discussed in l
question 3.
l f
i Ouestion 2 Page 5 of Reference 3 deduces a top region (Fluid l
Block 1) heat transfer coefficient of 115 BTU /ft2-h-F for a "non-venting case,"
to be used in Reference 2,
but Reference 2
l l
considers open vent valves (see Page 55, with Fluid Block 5).-
Is there a conflict?
What is the effect of vent valves on the Fluid Block 1 heat transfer coefficient?
A-69 NUREG-1374
l(coctor Vessel'Ihcrmel Stress i
Venting is in reference to the vent line which Rennonso 2 connects the reactor vessel upper head to the hot leg.
Conse-quently, the "non-venting case" in reference 3 refers to an upper head cooldown analysis which was performed without the vent line attached.
The vent valves are expected to be active during the natural circulation cooldown.
The fluid flow into the opper head region (fluid block 1-transient 3), with reactor vessel vent valves operating, is estimated to be approximately 10% higher than with no vent valves.
Vent valve operation would result in a slightly higher heat transfer coefficient for fluid block 1.
Question 3 Section 2 of Reference 1 refers to a " void" forming in the upper head.
Is this superheated steam or saturated vapor?
Figure 3 of Reference 3 gives the film coefficient as being for subcooled water.
Why was this value chosen?
Responp_p_1 - The void in the upper head is assumed to be satura-ted steam at 600 F.
The film coefficient for subcooled liquid was mistakenly applied to the steam region.
However, there is high resistance to heat flow between the head and the vessel.
The only paths aret from the head, through an air gap to the studs, and then back to the vessel; and from the head to the vessel through 2 small "0"
rings.
Therefore, the artificially NUlt!!G-1374 A-70
Reactor VeselThermal Stress high film coefficient on the inside of the head will not change the stresses, up or down, significantly.
Question 4 The overall U values prepared in Reference 3 for use in Reference 2 were inadvertently misquoted in Reference 1 (page 4-3) and Reference 2 (Fage 11), which should read RV head inner surface 115 BTU /ft2-h-Fi RV head outer surface.23 BTU /ft2-h-Fi (upper region)
RV head outer surface 2.43 BTU /ft2-h-F; (lower region)
Resoonse 4 - That is correct.
The heat transfer coefficients are RV head inner surface film heat transfer.115 BTU /ft2-h-F (sub-cooled liquid)
RV head outer surface
.23 BTU /ft2-h-F (upper region, which includes an air gap between the metal and the insulation)
RV head outer surface 2.83 BTU /ft2-h-F (lower region, which does not inciude an air gap between the metal and the insulation)
The RV head inner surface heat transfer coefficient is applied with respect to the temperature difference between the upper head fluid and the metal wall inside surface temperature.
The RV head outer surface coefficients are overall heat transfer coefficients witn respect to the difference between the average metal tempera-ture and containment temperature.
A-71 NUREG-1374
Reactor Vessel hermal Stress ouestion 5 The outer heat flows used in Reference 3 to deduco overall effective U's are varying spatially.
What is the effect of the averaging done here?
One would expect that the thermal insulation is the main thermal resistance. and since that is apparently 3 inches thicX everywbara (Table 4-3 snd Figure 4-4 of Reference 4) one would expect it not to vary by more than one order of magnitude between top and sides.
What is the insulation material?
What are its thernal conductivity and emissivity?
i The representative U's for the lower region and Resoonse 5 upper rat on ware obtained by adding the representative heat i
fluxes and dividing by the temperature differential between the average metal tamparature and the containment temperature for the respective regions.
The values represent an average U for each of the regions.
The spatial variation in the heat transfer coefficients is insignificant (i.e standard deviation is small).
The thermal insulation provides the main thermal resistance for the ' lownr portion of the RV upper head, as the insulation is af fixed to the metal surface.
The upper portion of the RV head contains an air gap between the RV head metal and the insulation, thus oroviding additional resistance and a lower overall U (.23 Btu /h-ft2-F for the upper region compared to 2.83 Btu /h-ft2-F for the lower region).
l MUREG-1374 A-72
Reactor Venel Thermal Stress The insulation material is mirror insulation with the following heat transfer proportless I
o Thornal conductivity =.079 BTU /h-ft-F o Density
= 14.4 lbm/ft3 o specific heat
=.134 Btu /lbm-F o Surface Emissivity
=.160 Question 6 On pages 19 and 20 of Reference 2 an effective film coefficient is computed to represent radiation heat transfer.
Provide justification for selecting T2= 450'F (or 910 R) and dt=
50 R or 40 R for computing the film coefficient.
Response 6 The choice of T2 450 F and a DT
=
50 F were
=
assumptions based upon the engineers knowledgu of the steady state temperature distribution from previous analyses.
The corresponding heat flux for this region agrees well with the heat flux calculated in reference 3.
Questian__2, A convective heat transfer coefficient for the closure head, identified as Fluid Block 1, is computed on pages 39 through 44 of Reference 2.
Snveral questions are applicable to this cair n1ation.
QueJtion_2A Provide justification for the mass flow rate in this region to be 8% of the natural circulation loop flow.
A-73 NUREG-1374
Reactor Venellhermol Stress Resnonse 7A - Detailed upper head thermal-hydraulic calculations have been prepared by Bt.W vhich predict flowrates into the upper head region, above the planum cover, to equal 84 to 12% of the total system flow.
The flow, however, penetrates approximately 6 to 12 inches above the plenum cylinder cover.
Consequently, the heat transfer correlation for fluid block 1 (11guld port.ior.)
should not be calculated via forced convection and in fact was calculated based upon natural or free convection (pg 42 ref. 2).
Dyestion 7B Justify computing the velocity using the flow area that was used for Fluid Block 4.
Resoonso 7B - The final film coefficient was correctly based upon natural convection as opposed to forced convection (pg. 42 of Ref. 2).
Consequently the velocity calculation with respect to forced convection is not applicable.
Question 7c The characteristic length L is defined as L = De/2 in the heat transfer correlation.
Why is it used as L = De in the Reynolds Humber calculation?
Response 7C - Please see response 7B.
ouestion 7D Are the values of length and flow area used consis-tent with each other?
NUREG-1374 A-74
Reactor Vessel 1hermal Stress Eissonne 70 - Please see response 7B.
Question 7E Provide justification for using a 15 F difference between the fluid and the metal.
Resoonse 7E - Please see response to question 1.
Question 8 Provide justification for assuming the flow rate to be 8% of the natural circulation loop flow rate in the evaluation of the film coefficient for fluid block 4.
Provide justification for using a correlation for forced convection as opposed to the natural convection correlation used for Fluid Block I with the same flow rate.
Resoonse 8 - The 84 flow, which penetrates the uppar head above the planum cover, returns to the. outlet annulus as shown on pg 51 l
of Reference 2.
A forced convection correlation is therefore applicable for fluid block 4.
III.
Thermal stress Analysis ouestierl_1 since the inlet and outlet openings in the reactor vessel are localized, provide justification for the choice of a finite element model for the stress analysis which is axisyn-metric about the reactor vessel vertical axis.
Also, provide details regarding the degree of accuracy in the modeling of the A-75 NUREG-1374
Reactor Vessel Thermal Stress temperature and film coef ficients within the vessel as axisym-metric.
Essoonse 1 - The vessel nozzle openings werc designed in accor-dance with the area replacement rule (ASME Boiler and Pressure vessel codo,Section III, Subsection NB, Pr ragraph NB 3332.2).
Th'us, on any plane passed through the oper 65e area of metal
. ment in a local-removed by the opening is replaced by ra ized area around the opening.
Therefore the stiffness of the vessel is essentially axisymmetric.
During the natural circulation cooldewn transient
- analyzed, reactor coolant flow in the system is very low.
Therefore, below the assumed void'line, flow in the reactor vessel in the axial and tangential directions are both low and there is no inaccuracy in the axisymmetric modeling of film coef ficients.
Above the assumed void line, flows in all directions are essentially zero.
Question 2
Provide justification for not including in the analysis the reaction forces on the reactor nozzles.
Response 2 - The reaction forces due to system heat-up are small.
Recent calculations indicate that the stresses in the 3.3 inch thicx hot leg (largest) pipe at the reactor vessel are on the.
order of 500 psi.
The resultant stress in the 8.4 inch thick l
l l
NUREO-1374 A-76
Reactor Vessei nermal Stress l
reactor vessel shell are even smaller and therefore not signifi-cant to the analysis.
Question 3 In Section 9.0 of Referance 2 it is mentioned that the thermal stress analyses were performed for three rtructura) nodels.
For all the stress analyses the shear etress in the ve'asel closure bolts have not been reported.
What in the magnitude of the shear stresses in the vessel closure bolts for case 1, which uses a friction coefficient of z,sro at the flange interfaces?
Included. in the above question are the thermal stress runs made to comply with review comments 2 and 6 (page 94 of Reference 2).
Resoonse 3 Three structural models were run for Transient condition 1 and two structural models each for Transient condi-tions 2 and 3.
The in plane results for the studs are shown on l
pages 79, 80 and 81 of Reference 2.
The transverse shear stresses in the studs are small and do not affect the analysis result.
This is shown below where the shear stress, extracted frem the computer runs, are combined with the nombrane stresses to obtain the resultant stress intensities.
The an.dysna performed in response to review comments 2 and 6 give similar results.
A-77 NUREO-1374
Reactor Vessel 7hermal Stress AVG.
STRESS TRANSIENT CASE IDAD STEP MEMBRANE MAX. SHEAR INTENSITY CONDITION TIME-HRS STRESS-PSI STRESS-Pfl PSI 1
1 0
35,609 56 35,609 3
1 6
35,130 165 35,132 1
1 24 37,998 59 37,998 1
2 0
35,050 300 35,055 1
2 6
35,412 284 35,417 1
2 24 38,797 304 38,802 1
3 0
35,348 419 35,358 1
3 6
35,411 284 35,416 1
3 24 38,796 303 38,801 2
1 0
35,554 521 35,569 2
1 6
34,37*
148 34,375 2
1 24 36,925 86 36,925 2
2 0
35,000 300 35,005 2
2 6
34,700 287 34,705 2
2 24 37,794 311 37,799 3
1 0
34,850 547 34,867 3
1 6
22,169 435 22,186
(
3 2
24 21,517 457 21,$36 J
3 2
0 34,230 301 34,235 3
2 6
21,554 193 21,557 3
2 24 21,038 254 21,024 t.
Qugstien 4 The radial displacement of stud node 921 (Figure 4 of.
R9ference 2,
page 26) and flange node 259 at the interface were coupled to allow for stad shear force and bending moment to be transmitted to the vessel flange at two locations (page 96 of Reference 2).
Subsequently, thc. thormal stress analysis for transient condition 2 and structural canes 1 and 2 was repeated.
Why is a linear constraint equation, such as the one at the bottom of page 34 of reference 2, not provided between nodes 259 and its adjacent shell node?
In additisn, why are nodes 259 and 921 not coupled in the vertical direction?
o NUREG-1374 A-78 i
Reactor Vessel 1hermal Stress The stud model waa coupled to the vessel at one Resoonse 4 loca tion each at the top and bottom of the stud.
The coupling 4
allows transfer of vertical and radial (shear) forces and bending moments.
The top connection represents the bearing of the r.tud nut against the top of the closure head flange and is representa-tive of the actual configuration.
The single connection at the j
bottom of the stud is conservative since all loads are trans-i i
ferred to a point instead of through the entire length of the j
engaged threads as in the actual configuration.,
1 i
I Question 5 For the analysis done to comply with review comments 2 end 6 the reactor vessel was assumed to be free to grow radially at the bott.om of the nozzle belt region.
Does this l
boundary conditio'n have any effect on limiting the strains in the i
vessel and on the magnitude of the shear forces in the. vessel i
closure bolts?
In addition, it is not totally clear from the answer to roview comment 3,
page 96 of Reference 2,
if the l
reaction of the plenum cover on the upper head was considered in f
the initial thermal /stra,as analysis, or only in tha analysis performed after the review comments.
Eggp_qnge_E - The stresses in the reactor vessel in the area of j
interest are unaffected by the radial boundary conditions at the i
1 and of the model since the model cylindrical length is greater
]
than the characteristic length of the vessel.
The cylindrical 3
j model length is 130 inches while the characteristic leng'.h of a 1
i A-79 NUREG-1374 3
~
i Reactor Vessel'Ihennal Stress I
cylinder 180 inches in diameter end 8.4 inches in thickness is 64 i
inches.
l.
j The original analysis did not include the plenum cover reaction 1
on the upper head.
This omission was corrected in the analysis I
performed after the review comments, t
Ouestion 6 In T.1ble 1 (page 22) of Reference 2 the coefficients of thermal expansion, oc, are average coefficients of thermal i
i expansion in going from 70F to the indicated temperature.
Provide justification for using the average values in lieu of the instantaneous values.
I Pesconse J - The'ANSYS computer code uses average coefficients of thermal expansion in tabular form.
For intermediate tempera-tures, the code interpolates.
i Question 7 On page 27 of Reference 2,
the effective length of the bolt was assumed to bo 40.25 inches.
Provide justification j
for choosing this bolt length to calculate the bolt prestress..
i i
Be5Pon50 7 - The bolt (stud) length was modeled as the total free i
length plus one-half of the length of engaged threads.
This yleAus an axial stiffness which is in close agreement with data 4
measured during stud tensioning.
The stud preload is given, not calculated.
The computer analysis is uced to iterate on the stud 1
MUREG-1374 A-80
4 i
Reactor Vessel 7hennal Stress i
i 1
preload until the given value is achieved; therefore the preload 4
is not determined by the soud length.
i REFERENCES i
a 1.
" Stress Analysis of the Reactor Vessel Closure Region for a Natural Circulation Cooldown Transient," Babcock & Wilcox 3
Owners Group Analysis Committee, Report No. 77-1152846-00, July 1984.
I 2.
" Stress Analysis of the Reactor Vessel closure Region for a l
Natural Circulation Cooldown Transient," Babcock & Wilcox, i
Calculation No. 32-1151155-00, June 21, 1984.
3.
B.
L. Bowlman, " Reactor Vessel Head Stress Analysis Inputs,"
Calculation No. 3.1-1150499-00, B&W, March 18, 1984.
I l
}
4.
Crystal River Unit 3 FSAR.
l 5.
General Functional Specification for R.C.S.
Components for CPCO, No. 18-10?2000012-05.
E.
R.
W.
- Winks, R.
C.
- Twilley, Jr.,
" Natural Circulation Occurrences at Operating B&W Plants," Babcock & Wilcox Co.,
May 7, 1979.
A-81 NUREG 1374
Reactor Vessel Thermal Stress Flow Rates at B&W 177-FA Plant,"
Proceedings:
Third International RETRAN Conference, EPRI NP-3803-SR, 1985.
8.
B&W Calculation No.
32-1150499-00,
" Reactor Vest Head Stress Analysis Inputs,"
for all 177-FA Owners Group Contracts.
(Same as Reference 3) 9.
J.
P.
Holman, " Heat Transfer," McGraw-Hill Book comoany, 1963.
10.
Babcock & Wilcox Calculation No. 32-1140915-00, "CPC.
R.V.
CRDM Motions," *l55 12 & 13 Contracts.
3 A-82 NUREG-1374
Reactor Vessel Thermal Stress APPENDIX V I
4 i
l 1
l I,
d L
i f
A-83 NUREG-1374 1
bactor Vessel Thermal Stress l)
BROOKHAVEN NATIONAL LABORATORY (l(ll ASSOCIATED UNIVERSITIES. INC, Upton. Long Isfand. New York 11973
($16) 282s FTS 666/
2448 Depcrtment of Nuciso' Energy g
n'. 2 :.3 April 20, 1988 Mr.'Jc=1 Page MS 217A U.S. Nuclear Regulatory Consission 5650 Niche' ion Lane South Lockv(He,t40 20351
Subject:
Comments pertaining to BWOG Draft responses to NRC cocnents dated 4/8/88 Dear Joel 1.
General Response 1 (page 2) a.
This response seems to r.cstradict what was stated to BNL during an informal telephone conversation with B&W engineers (i.e. under natural convection cooldown conditions the upper head fluid would ree:ain stagnant, whereas under forced flow (DHRS) between 7 to 9 percent of the fluid would reach the upper head resulting in more uniform temperature conditions) see our letter to you d: tad 3/31/88.
b.
BWOG should clarify the un her of cycles expected for DERS.
c.
It would also be advantageous to obtain the DHR3 transient temperatura time history, d.
Having aviewed several TSARS, it is still not clear to why there would h 7 to 9% flow thrcugh the upper head. We would appreciate a er;;snation, supperted by reference to drawings, figures or sketches.
R(.spons,e 2 Since this is a design problem, it seems that conditions similar to snose of the original design assumptien shuld be used. This is the reason that a cooldown rata of 100 r/hr was used in the evaluations that were performed at BNL.
MUREG-1374 A-84
.. =
_ - ~. - - -
. _ _ _ - _. _ _ = -
i Reactor Vessel'Thennal Stress 1
i DRAPT s
Part II Heat Transfer Analysis Response 5 l
2 This response does not clarify the question. The insulation data given corresponds to an equivalent h = k/d = 0 079/0.25 = 0 316 pTV l
f brF Any overall resistanea can only be lower than that number. Thus, 2
1 where does the overall U value of 2 8!
TU/yt hrF for the lower j
region come from?
Part III Therma" Stress Analysis 1
j Response 3 d
1 The transverse shear stresses reported by BWOG for the studs are low.
Could these be a result of the modeling assumptions made by BWOG1 (i.e. coupling assumptions regarding the reactor studs and the reactor vessel shell).
)
Response 4 We have questions concerning the coupling of the R.V.. stud with the reactor vessel lower flange at the boundary interface region.
In reference 2 pages 94 and 96, it's shown that the radial displacements of the stud (node 921) and of the lower flange shall at the same 1
location (node 259) are coupled. Whay aren't these nodes also coupisd in the vertical direction? Moreover why didn't BWOG apply a similar constraint at this position as that applied at the bottom of the 1
threaded portion of the stud at nodes 265, 266, and 2677 (In j
fact, such constraints at the interface regions could lead to highet i
stresses in the stude because of the reduction in effective stud l
1ength).
\\
l Response 7 l
The calculations on page 27 of refera tes 2 seem to indicate that the prestressing stress in tha bolts is calculated from the defermed length of the bolt using an effective length of 40.25 inchs.: for the bolt. On page 25 of reference 2 it is stated "for proloading, the bolt the parameter INTF listed in tables was determined by trial and error so as to cause the preload stress also shown in table 3...".
i Ane preload stress shown in Table 3 is the one calculated from the j
deformed Irangth of the bolt and the bolt effective leng'th of 40.25 i
l i
4 l
1 i
A-85 NUREG-1374 P/E'd S2:Pt S8, T2 &f3 j
l
,,1 bd[*h['
=
Reactor VesselDermal Stress 4
in.
If the preload is given, what is the purpose of the aalculations shown on page 27 of reference 21 (Obviously one can readily vary the stresses by assuming different effective lengths.
4 I
Yary truly yours, pl Morr s $ch,' Esad Structural Analysis DivisionH1:db i
i i
i 4
i i
4 J
i l
l i
i l
l i
A-86 I
Reactor Venel Thermal Stress 4
i I
,1 9
?
4 1
1 i
i 1
s 1
i 1
4 4
i a
1 4
4 APPENDlX VI 1
l, 1
i 4
l s
l e
r i
I.
l d
4
.f 1
J
\\
e A-87 N!! REG-1374
Reactor Vessel *!hermal Stress Babcock & Wilcox we., e...
ni,;.i.,
a McDertrett company 331S oto ro,est Road October 26, 1988 P.O.soito935 ESC-9.'9 trac *'s. Y A 24506 0935 (804) 385 2000 Mr.'Joel D. Page Task Har.ager, Section B, Engineering Issues Branch Office of Nuclear Regulatory Research Hail Stop NL/S-302 U.S. Nuclear Regulatory commiscion Washington, D.C.
20555
Dear Mr. Page:
Please find attached final responses to the NRC questions regarding Generic Issue No.
79, RV Thermal Stresses. '
These responses were originally submitted via letter of June 23, 1988 and were inadvertently labeled " draft" responses on the title page.
The title page has been corrected wit.h no other changes made to those original responses, The response document is attached hereto, and replaces the original issue in total.
Very truly yours, J.,
. Pal ug' Project Har..ser owners Group Engineering Services JRP/leh Attachment cc:
B&i4oG Analysis committee C. H. Turk
- AP&L P.
F. Guill
- DPCo J. E. Burchfield, Jr. - DPCo E. H. Davidson
- FPC A.
Irani
- GPUN J. X. Atwell
- SMUD J.
F.
Dunne
- TED NUREG-1374 A-88 l
Reactor Vessel 1hermol Stress Final Response to NRC Ouestions I.
General ouestion 1 The background section of Reference 1 states that thersc1 gradients can be created during the transition to decay heat removal systen or during a natural circulation cooldown.
- However, the analysis only addresses the natural circulation cooldown transient.
Please provide justification for not consid-ering the transition to decay heat removal system operation in this analysis.
Additionally, please provide justification for considering only 20 natural circulation cooldown cycles, a
Transition to the decay heat removal system was Response 1 considered.
The pressure vs temperature curve used for the stress analysis input, Reference,2 Tage 97, includes transition to the decay heat removal system (DHRS) beginning at approximate-l i
ly 11 hours1.273148e-4 days <br />0.00306 hours <br />1.818783e-5 weeks <br />4.1855e-6 months <br />.
Total primary system flow before and after DHRS I
actuation are estimated to be quite similar.
Natural circulation flow is estimated at approximately 8500 GPM, and D: IRS flow is approximately 7500 GPM.
Consequently, the change in flowrate into tne upper head region following actuation of the DHRS would not-be.significant.
l l
4
Reactor Vessel' Thermal Stress The B&W plants' original design basis was to remain at hot shutdown following an unplanned NSS transient until a normal plant cooldown could be initiated.
Natural circulation capabil-ity was designed into the plant as a means of removing decay heat following loss of forced flow, but natural circulation couldowns were not included in the original design basis of the p3 ants.
In recent years natural circulation capability has been viewed as a significant contributor to achieving col 6 safe shutdown following severe plant upsets and has besn evaluated 'at various times to address operational or technical
- concerns, including RV head stress issues.
For these types of evaluations B&W has recom-mended a relatively small number of cycles (20) for natural circulation cooldowns.
This value is consistent with other i
emergency category cycles specified in the RCS Functional Specifications and the component stress reports in the original design basis.
Operating plant experience supports the use of 20 cycles as a reaso able upper limit for the number of natural circulation cooldown events.
To date, no natural circulation cooldown has occurred at a B&W plant.
The design basis of 240 cycles (Table 5.2-1 of Oconee FSAR) is only applicable to a normal forced flow cooldown.
A natural circulation cooldown is considered an emergency event and is limited to 20 cycles.
The total number of, cycles for both forced flow cooldown and natural circulation cooldown is 260 cycles.
NUREG-1374 A-90
Reactor Vessel Thennal Stress i
1 Sentence number 4
of the
Background
section of the stress l!
analysis report Reference 1 should be revised as follows:
These f
gredients could develop as a result of non-uniform cooling of the 4
]
reactor coolant within the reactor vessel that are created during t
a natural circulation cooldown and the subsequent **ansient to I
j decay heat removal system operation following the natural i
circulation cooldown.
)
I Ouestion 2 Section 4 of Reference 1 gives the coolant cooldown rate as 20 F to 100 F/h; howeve r, only a 50 F/h cooldown rate is
{
used in the analysis.
Please provide justification for selecting only the cooldown rate of 50 F/h.
l Response 2 - A natural circulation cooldown of the NSSS, for all B&W plants excluding the Oconee unita, would occur only when offsite power is lost for an extended period of time.
The loss 1
f of offsite power precludes the use of the turbine bypass valves 4
j for cocidown of the primary thute requiring a cooldown with the atmospheric dump valves (ADV's).
The total capacity of the ADV's i
in B&W plants in general does not exceed 10% of the total rated steam flow, thus limiting the long-term primary natural circula-
{
tion cooldown rate to 50 F/h or less.
Oconee Units 1, 2 and 3 have implemented procedural limits of i
50 "/h for natural circulation cooldowns.
i 4
A-91 NUREO-1374 4
4
Reactor Vessel'lhermal Stress II.
}icat Tra ns f er SIlaJ vsis Ouestion 1 Provide justification for selecting a temperature difference of 15 F and a vertical plate height of 50 inches to compute the void level heat transfer coefficient for Fluid Block 1,
transient 3 (pages 40 to 43 of Reference 2).
Can the film coefficients calculated for the void level for all the transient conditions be considered as upper limits.
A vertical plate height of 50 inches roughly
Response
1 corresponds to the elevation of the liquid level above the plenum covel.
For transient condition 3,
the fluid block element temperatures below the steam void level are assumed to equal the core outlet temperature.
The temperature differential between the liquid and the wall was chosen to be 15 F, which is a reasonable approximation based upon the detailed finite difference fluid model which was used to generate the inner surface film heat transfer coefficient (Reference 3).
The film coefficic.nt calculated en pages 39 - 43 of Reference 2
.is only applivable to fluid block 1 for transient condition 3.
The film coefficient for the voided region is discussed in question 3.
NUREG-1374 A-92
1 Reactor Vessel Thermal Stress i
I Duestion 2 Page 5 of Reference 3 deduces a top region '(Fluid i
f Block 1) heat transfer coeffislent of 115 BTU /ft2-h-F for a "non-j venting case,"
to be used in Reference 2,
but Reference 2
l considers open *.ent valves (see Page 55, with Fluid Block 5).
Is there a conflict?
What is the effect of vent valves on the Fluid Block 1 heat transfer coefficient?
1 Response 2 - Venting is in reference to the continuous vent line which connects the reactor vessel upper head to the hot leg.
Consequently, the "non-venting case" in reference 3 refers to an upper head cooldown analysis which was performed without the vent t
line attached.
The reactor vessel vent valves are expected to be active during i
the natural circulation cooldown.
The fluid flow into the upper head region (fluid block 1-transient 3), with reactor vessel i
vent valves operating, is estimated to be approximately 10%
l l
higher than with no vent valves.
Vent valve operation would result in a slightly higher heat transfer coefficient for fluid block 1.
l l
l l
Ouestion 3 Section 2 of Reference 1 refers to a "vcid" forming 1
in the. upper. head..Is this superhetated steam or saturated vapor?
{
Tigure 3 of Reference 3 gives the film coefficient as being for 1
subcooled water.
Why was this value chosen?
I I
I
^43 NUREG-1374
Reactor Vessel'Ihermal Stress Reseense 3 - The void in the upper head is assumed to be satura-ted stean at 600 P.
The film coefficient for subcooled liquid was mistakenly applied to the steam region.
However, there is high resistance to heat flow between the head and the yessel.
The only paths are: from the head, through an air gap to the studs, and then back to the vessel; and from the head to the vessel through 2 small "O"
rings.
Therefore, the artificially high film coefficient on the inside of the head will not change the stresses, up or down, significantly.
Question 4 The overall U values prepared in Reference 3 for use in Reference 2 were inadvertently misquoted in Reference 1 (page 4-3) and Reference 2 (Page 11), which should read:
RV head inner surface 115 BTU /ft2-h-F; RV head outer surface.23 BTU /ft2-h-F; (upper region)
RV head outer surface 2.83 BTU /ft2-h-F; (lower region)
That is correct.
The heat transfer coefficients Resconce 4 are:
RV head inner surface film heat transfer 115 BTU /ft2-h-F (sub-cooled liquid)
RV head outer surface
.23 B"U/ f t2-h-F (upper
- region, which includes an air gap between the metal,and the insulation).
RV head outer surface 2.83 BTU /ft2-h-F (lower region, which does not include an air gap between the metal and the insulation)
NUREG-1374 A_94 I
)
i Reactor Vessel Thermal Stress The RV head inner surface heat transfer coefficient is applied with respect to the temperature difference between the upper head f
fluid and the metal wall inside surface temperature.
The RV head outer surface coefficients are overall heat transfer coefficients with respect to the difference between the average antal tempera-l ture and containment temperature.
j ouestion 5 The outer heat flows used in Reference 3 to deduce overall effective O's are varying spatially.
What is the effect of the averaging done here?
One would expect that the thermal insulation is the main thermal resistance, and since that is apparently 3 inches thick everywhere (Table.4-3 and Figure 4-4 of Reference 4) one would expect it not to vary by more than one order of magnitude between top and sides.
What is the insulation material?
What are its thermal conductivity and emissivity?
Ressonse 5 The representative U's for the lower region and upper region were obtained by adding the rcpresentative heat fluxes and dividing by the temperature differential between the average metal temperature and the containment temperature for the respective regions.
The values represent an average U for each of the regions.
The spatial variation in the heat transfor l
coefficients is insignificant (i.e standard deviation is small).
A-95 NUREG-1374
Reactor Vessel nermal Stress The overall heat transfer coefficient for the lower region (2.83 Btu /h-ft2-f) was found to be high by an order; of magnitude.
B&W has determined the impact of the error to be conservative.
The combination of a higher heat transfer coefficient in the lower upper head region, which would yield lower metal tempera-tures, coupled with the higher temperatures predicted for the upper head region, a result of the 115 Btu /h-ft2-f heat transfer coefficient applied to the inner surface of the steam region, yields conservatively high RV closure head and upper shell region stress predictions.
The insulation material is mirror insulation with the following heat transfer properties:
o Thermal conductivity =.079 BTU /h-ft-F 1
o Lensity
= 14.4 lbm/ft3 o Soecific heat
.134 Btu /lbm-F
=
o Surface Emissivity
.160
=
Cuestion 6 On pages 19 and 20 of Reference 2 an effective film coefficient is computed to represent radiation heat transfer.
Provide justification for selecting T2= 450 F (cr 910 R) and dt=
50 R or 40 R for computing the film coefficient.
The choice of T2 450 F and a DT 50 F were Response 6
=
=
assumptions. based upon the engineers knowl' edge of the steady state tenperature distribution from previous analyses.
The TiUREG-1374 -
A-%
(
r
s---
_a 4
i Reactor Vessel Thermal Stress corresponding heat flux for this region agrees well with the heat flux calculated in reference 3.
Qpestion 7 A convective heat transfer coefficient for the closure head, identified as Fluid Block 1, is computed on pages l
39 through 44 of Reference 2.
Several questions are applicable to this calculation.
}
l 4
{
ouestion 7A Provide justification for the mass flow rate in this region to be 8% of the natrial-circulation loop flow.
1 Detailed upper head thermal-hydraulic calculations Response 7A have beer prepared by E&W which predict flowrates into the upper head regiun, above the planum cover, to equal 8% to 12% of the total system flow.
The flev paths leading from the core exit to i
the hotlegs are presented in Figure 1.
A description of each flow path is provided below:
i PATH PATH DESCRIPTION 1
From fuel assembly to open plenum area 2
From fuel assembly to column'weldments (control rod guide tubes) 3 Through lower exit ports in column weldmr,ts l
.4
.Through.3 inch diam. holes in plenum cyl. to outlet nc?tle 5
Through 22 in, and 34 in, diam. heles in plenum cyl.
Over top of plenum cyl. into outer annuius l
A-97 NUREG-1374
Reactor Vessel Thermal Stress 7
From column weldments (69 of them) into upper head and to outint plenum 8
Through outlet nozzle The percentage of total flow through path 7 has been calculated by B&W to be 10.5% of the total flow through the hot legs.
The ficw, however, penetrates approximately 6 to 12 inches above the plenum cylinde:c cover.
Consequently, the heat transfer correla-tion ior fluid block 1 (liquid portion) should not be calculated via forced convection and in fact was calculated based upon natural or free convection (pg 42 ref. 2).
Question 7B Justify computing the velocity using the flow area that was used for Fluid Block 4.
1 Response 7B - The final film coefficient aas correctly based upon natural convection as opposed to_ forced convection (pg. 42 of Ref. 2).
Consequently the velocity calculation with respect to forced convection is not applicable.
Ouestion 7C The characteristic length L is defined as L = De/2 in the heal transfer correl ; ion.
Why is it used as L= Da in the Reynolds Number calculation?
Response 7c - Please see response 7B.
i NUREG-1374 A-98
t Reactor Vessel Hermal Stress Question 7D Are the values of length and flow area used consis-tent with each other?
Rgsp_pnse 7D - Please see response 7B.
Question 7E Provide justification for using a 15 F difference betwe3n the fluid and the metal.
Response 7E - Please see response to question 1.
Question 8 Provide justification for assusing the flow rate to be 8% of the natural circulation loop flow rute in the evaluation of the film coctficient for fluid block 4.
Provide justification for using a correlation for forced convection as opposed to the natural convection correlation used for Fluid Block 1 with the same flow rate.
Response 8 - The B'4 flow, which penetrates the upper head above the plcnum cover, returns to the outlet annulus as shown on pg 51 of Reference 2.
A forced convection correlation is therefore app 2icable for fluid block 4.
III.
Thermal Stress Analysis Question 1 Since the inlet and outlet openings in the reactor vessel are localized, provide justification for the choice of a
+99 NURIiG-1374
Reactor Vessel nermal Stress finite element model for the stress analysis which is axisym-metric about the reactor vessel vertical axis.
Also, provide details regarding the degree of accuracy in the modeling of the temperature and fi' m coefficients within the vessel as axisym-metric.
Response 1 - The vessel nozzle openings were designed in accor-dance with the hrca replacement rule (ASME Boiler and Pressure Vessel Code,Section III, Subsection NB, Paragraph NB 3332.2).
Thus, on any plane passed through the opening, the area of metal removed by the opening is replaced by reinforcemene. in a local-ized area around the opening.
Therefore the stiffness of the vessel is essentially axisymmetric.
During the natural circulation cooldown transient
- analyzed, reactor coolant flow in the system is very low.
Therefore, below the assumed void line, flow in the reactor vessel in the axial and tangential directions eie both low and there is no inaccuracy in the axisymmetric modeling of film coefficients.
Above the assumed void line, flows in all directions are essentially zero.
Question 2
Provide justification for not including in the analysis the reaction forces on the reactor nozzles.
Egmoonse 2 - The reaction forces due to system heat-up are small.
Recent calculations indicate that the stresses in the 3.3 inch NUREG-1374 A-100
Reactor Vessel Hermal Stress thick hot leg (largest) pipe at the reactor vessel are on the order of 500 psi.
The resultant stress in the 8.4 inch thick reactor vessel shall are even smaller and therefore not signifi-cant to the analysis.
i Question 3 In Section 9.0 of Reference 2 it is mentioned that the thermal stress analyses were performed for three structurkl models.
For all the stress analyses the shear stress in the vessel closure bolts have not been reported.'
What is the magnitude of the shear stresses in the vessel closure bolts for, case 1, which uses a friction coefficient of zero at the flange interfaces?
Included in the above question are the thermal stress runs made to comply with review comments 2 and 6 (pag. 94 of Reference 2).
Rosconse 3 Three structural models were run for Transient Condition 1 and two structural undels each for Transient Condi-tions 2 and 3.
The in-plane results for the studs are shown on pages 79, 80 and 81 of Reference 2.
The transverse she s t-stresses in.the studs are small and do not affect the analys;.a result.
This is shown below where the shear stress, extracted from the computer runs, are combined with the membrane stresses to obtain the resaltant stress intensities.
The analyses performed in response to review comments 2 and 6 give similar results.
/r 101 NUREG-1374 i
Reactor Vessel Thermal Stress AVG.
STRESS TRANSIENT CASE LOAD STEP MEMBRWE MAX. SHEAR INTENSITY CONDITION NO.
TIME-HRS STRESS-PEI D. TRESS-PSI PSI 1
1 0
35,609 56 35,609 1
1 6
35,130 165 35,132 1
1 24 37,998 59 37,998 1
2 0
35,050 300 35,055 1
2 6
35,412 284 35,417 1
2 24 38,797 304 38,802 1
3 0
35,348 419 35,358 1
3 6
35,411 284 35,416 1
3 24 38,796 303 38,801 2
1 0
35,554 521 35,569 l
6 34,374 148 34,375 2
1 24 36,925 86 36,925 2
2 0
35,000 300 25,005 2
2 6
34,700 287 34,705 2
2 24 37,794 311 37,799 3
1 0
34,850 547 34,867 3
1 6
22,169 435 22,186 3
1 24 21,517 457 21,536 3
2 0
34,230 301 34,235 3
2 6
21,554 193 21,557 3
2 24 21,018 254 21,024 Ouestion 4 The radial displacement of ctud node 921 (Figure 4 of Reference 2,
page 26) and flange node 259 at the interface were coupled to allow for stud shear' force and bending raonent to be i
transmitted to the vessel flange at two locations (page 96 of Reference 2).
Subsequently, the thermal stress analysis for transient condition 2 and structural cases 1 ana 2 was repeated.
Why is a linear constraint equation, such as the one at the bottom of page 34 of reference 2, not provided between nodes 259 Gnd its adjacent shell node?
In addition, why are nodes 259 and 921 not coupled in the vertical direction?.
I NURiiG-l?74 A-102 l
Reactor vessel'Ihermal Strec3 Resoonse 4 The stud model was coupled to the versel at one location each at the top and bottom.
The coupling allows transfer of vertical and radial (shear) forces and bending moments.
The top connection represents the bearing of the stud nut against the top of the closure head flange and is representa-tive of the actual configuration.
The single connection at the bottom of the stud is conservative since all loads are trans-ferred to a point instead of through the entire length of the engaged threads as in the actuel configuration.
The location of the bottom of the stud model is at an intermediate. point along the length of engaged threads.
Question 5 For the analysis done to comply with review comments 2 and 6 the reactor vessel was assumed to be free to grow radially at the bottom of the nozzle belt region.
Does this boundary condition have any effect on limiting the strains in the vessel and on the magnitude of the shear forces in the vessel closure bolts?
In addition, it is not totally clear from the answer to review comment 3,
page ^ #,
of Reference 2,
if the reaction of the plenum cover on the upper head was considered in the initial thermal / stress analysis, or only in the analysis performed after the review comments.
Resoonse 5 - The stresses in the reactor vessel.in the area of interest are unaffected by the radial boundary conditions at the et) of the model since the model cylindrical length is greater A-103 NUREG-1374 i
Ret.ctor Vessel Thermal Stress i
than the characteristic length of the vessel.
The cylindrical model length is 100 inches while the characteristic length of a cylinder 180 inches in diameter and 8.4 inches in thickness is 64 inches.
The original analysis did not include the plenum cover reaction on the upper head.
This omission was corrected in the analysis performed after the review comments.
ouestion 6 In Table 1 (page 22) of Reference 2 the coefficients of thermal expansion, o(.,
are average coefficients of thermal expansion in going from 70F to the indicated temperature.
Provide justification for using the average values in lieu of the instantaneous values.
Resoonse 6 - The ANSYS computer code uses average coefficients of thermal expansion in tabular form.
For ' intermediate tempera-tures, the code interpolates between points in the table.
Question 7 On page 27 of Reference 2,
the effective length of the bolt was assumed to be 40.25 inches.
Provide justification for choosing this bolt length to calculate the bolt prestress.
Resoonse 7 - The bolt (stud) length was modeled as the total free l
length plus one-half of the length of engaged threads.
This yields an axial stiffness which is in close agreement'with data NUREG-1374 A-104
.I
-w e
m.
w S
w+wu-v-mfr a-m-t-1r yi-w
'e-+
-y--
7
4 Reactor Vessel Thermal Stress j
j measured during stud tensioning.
The stud preload is given, not calcolated.
(A copy of a typical stud tensioner data sheet is i
l attached.
Tne Average Strain readings are in micro inches per a
d inch).
The conputer analysis is used to iterate on the stud I
prelcad until thai given value is achieved; therefore the preload 4
is not determined by the stud length.
l BEFERENCES i
t i
1.
" Stress Analysis of the Reactor Vessel Closure Region for a i
)
Natural Circulation Cooldown Transient," Babcock & Wilcox owners Group Analysis Committee, Report No. 77-1152846-00, July 1984.
2.
" Stress Analysis of the Reactor Vessel closure Region for a q
Natural Circulation Cooldown Transient," Babcock & Wilcox, Calculation No. 32-1151155-00, June 21, 1984.
3.
B.
L.
Bovinan, " Reactor Vessel Head Stress Analysis Inputs,"
l Calculation No. 32-1150499-00, B&W, March 18, 1984.
4.
Crystal River Unit 3 FSAR.
5.
General Functional Specification for R.C.S.
Components for CPCO, No. 18-1092000012-05.
F i
A-105 NUREG-1374
Reactor Vessel Thermal Stress 6.
R.
W.
- Winks, R.
C.
- Twilley, Jr.,
" Natural Circulation Occurrences at Operating B&W Plants," Babcock & Wilcox Co.,
May 7, 1979.
7.
N.
T.
Simms,' "RETRAN-02 Comparison of Natural Circulation Flow Rates at B&W 177-FA Plant,"
Proceedings:
Third l
International RETRAN Conference, EPRI NP-3803-SR, 1985.
1 8.
B&W Calculation No.
32-1150490-00, "Reaci.or Vessel Head i
Stress Analysis Inputs,"
for all 177-FA Owners Group Contracts.
(Same as Reference 3) l 9.
J.
P.
- Holman,
" Heat Transfer," McGraw-Hill Book Comparv, 1963.
a i
10.
Babcock & Wilcox Calculation No. 32-1140915-00, "CPC.
R.V.
CRDM Motions," N55 12 & 13 Contracts.
i 1
i j
NUREG-1374 A-106
I Reactor VenelThermal Streu
?
~~
CD4TnOL hDD ORIVE
,[
i CONTROL ROD miunv cp y era ep 9
=
= =-=m fi%,3kk, I%)'
'[
~
'f
[
' "'" '" ' " ' " \\^
j g, g ai8 <
Q,
,j f
INTERNALS VENT VALVE
,g : ;
r e m,_ p3 i 101-s i
i.!
l h
COME SUPT 0BT
.fjg'j p'
OUTLET N0Z2LE l
1 f
U.SAFFLE e
a
,l g
,my_,
4 FUEL A33EM9Lv CORE BARREL N
E g
q 7
\\
l
/
s$
\\
}
/
/
\\
l RE AL" TOR VEMEL suayt#LLANCE PEClutN
\\
HOLDER TU9E y
g
\\
\\
3 M-
/
g THERW AL SHIELD s
J; s
\\
/
I LMER GSID g
/
\\
GUIDE LUGS
\\
k /
\\
p -INCORE GUIDE
\\
h;T W[i ---. r T SUPPORT PLATE FLOW DISTRIDLTTOR IHHHdHH INCORE INSTRuutWT 3 5 h _h h l k!$hlh!$.
\\
77 f '** "'uMI"EY*'"'
i s_.
i s
i3 m
's A-107 NUREG-1374 l
Reactor Vessel Thermal Stress DI AMOND POWER SPECI ALTY COR P.
Ay8na63 STWD Avtpasg a y t sta g g a E L.A A A T IO.e E P P E C T G VS' AC71V8 e.v c m a W L l C L Ch* O se t L D
- E80V*I
- T.
gTeagm g L C>e G a V i c ae DTaget ggme F A CT om
( S mo v ae L a s.GT o.
. 8. '. 6 s
ni t.
l=
=
2.500 TTN SION E R 299
.0150 8925 266860 1.520 50.0 NUT 197
.0085 5870 175530 43.1
%.000 TTt!S IONE R 602
.0 F 0_ _17954 536839 1.283 49.7 NtTT 469
.0200
- t3991 41*333 42.5 7.500 TEN 510NE R 892
.0A&5 26589 795012 1.262 49.8 NUT 706
.0290 21053 629505 41.0 10.000 TE N510NE R 1164
.0580 34687 1037147,
1.236 49.8 NUT 941
.0385 28056 83BP95 40.8 L'M TE N S ION E R 1462
.0725 435B2 1303116 1.237 49.5 NtTT 11R?
.04R0' 35211 1051408 40.6
..15. 0 2Q,.
TENSIONER 17%9
_nRA%
52L11 1567749 1.71%
h NUT 1424
.0575 42442 1269035 40.3 17.500 TENSIONER 2019
.1015 60777 1817?1%
1.240 49 7 N UT 1644
.0670 49013 1465505 40.7 20.000 TENS ICN E R 7 11.4
_11A0 69551 70MBM0
- 1. 7 T7 49 4 NUT 1910
.0765 56925 1702070 40.0 a rte:
STUD LOADING
SUMMARY
SHEET el'oU'e,.
502701-N eveco...
B tw FOR cosc;ini_c_R oaTa J ANUARY 2, 19 7 8 ov G. E. ROTHE.S BERGE R
, r A,/
CLASS 1.000 TON TENSIONER m ea in t a o.
2-74 6-1/2" Dia. Stud DWG. NO.
S K-179 8 8-D A r-( 8 Srs 179 69 P NUREG-1374 A-108 13 WH'k e65:'
Reactor Vessel hermal Stress APPENDIX VII 4
l l
i 1
i i
l l
A-109 NUREG-1374
Reactor VesselTherrnal Stress AUG 31 ' O 10:19 P.2/10 bj b 3 BROOKHAVEN NATIONAL LABORATORY
- (l(ll ASSOCIATED UNIVERSITIES, INC.
Upton, Long Island, New York 11973
,[
)
2448 July 20,1988 I
Mr. Joel Page HS 217A 4
1 U.S.' Nuclear Regulatory Commission 1
5650 Nicholson Lane South l
Rockville, MD 20852 Subjects Comments pertaining to BWOG Draft responses to NRC questions regarding Generic Issue No.79 dated June 23, 1988 (ESC-544) 4
Dear Joel:
I.
General Response I (page 1 to 3) s.
Its stated on page 2 that "to date, no natural circulation cool down has occurred at a B&W plant" - could B&W clearly state what the avant at St. Lucie on June 11, 1980 was? Was it a natural circulation cool dowr. or was it a normal forced flow cool down? Moreover, what is the i
actual temperature pressure time history for the June 11, 1980 St.
i Lucie event?
b.
Also, it is stated (on page 3) "these gradients could develop as a result of non-uniform cooling of the reactor coolant within the reactor vessel that are created during a natural circulation cool down and the subsequent transient to decay heat removal system...."
Can such gradients occur during a forced flow or nott Moreover, are there any other cases where such gradients can occurf i
l II. Heat Transfer Analysis l
l hesponse 5 (pages 7 and 8) a.
B&W states that the impact of the error for the overall heat transfer coefficient for the lowsr region is determined to be conservative.
BWOG should demonstrate that this is the case, especiallly for transient condition No 2.
From our viewpoint, it needs to be demonstrated that this will result in a.nore conservative stress condition.
NUREG-1374 A-110
Reactor Vessel 1hennal Stress Response 4 (page 15) a.
BWOG states that "the location of the bottom of the stud model is at an intermediate point along the lent,th of the engaged threads". What is the precise location of the intermediate point? How was this point chosen?
Regarding BNL's previous cornents pertaining to Response 4 (see BNL letter to J. Page 4/20/88), it is to be noted that the question remains unanswered.
Response 7 a.
With respect to attached table shoving a typical stud tensioner data sheet, it is to be noted that the effective stud length is still a calculated value which assumes that each thread along the " effective engaged length" contributes equal aeounts of strain.
In actuality this is probably not true. Different threads contribute diffe c.t amounts of strains and thus the effective length for stresses other than uniforu axial (prestress) would be different than 40.25 in.
(See attached figure). Thus, for bending stresses a shorter effective length should be used.
Yours truly, f
orr: s' ivision Head MRadb A-111 NUREG-1374
_N
. _.. -. _.. - - - -...-.-.~ -
g.G 3a '68 30820 P.4/10 Reactor Vessel Thermal Stress I
\\
g1
\\
SRELL
\\
-m%L,,ge st "Mu l m
a
' pg, '*%
Y)
\\
p 1
i a
I S
+
4 i
i 1
f 4
i NUREG-1374 A-112
I I
APPENDIX B Fracture Mechanics Evaluation of the ll&W 177-Fuel-Assembly Reactor Vessel During a Natural Circulation Cooldown Event Richard E Johnson
Introduction and Ilackground Flaw parameters were selected in anordance with the following conditions: the crack plane was oriented norinal
'lhe ll&W 177 reactor vessel (RV) was analyzed to evalu-to the masmum principal stress (i.e., an axial crack dic-ate conditions during a natural circulation cooldown tated by the circumferential stress); the crack front was event. A linear clastic fracture mechanics approach was sharp (e.g., as the leading edge of a fatigue crack); the used following procedures recommended in the ASME location was at the surface carrying the higher tensile lloiler and Pressure Vessel Code (the Code).
stress; and the shape was semielliptical with a 6-to-1 (total length to-depth) aspect ratio. Postulated defects werc Temperatures and stresses were calculated as functions assumed to have a depth equal to one-fourth of the RV l
of time into the event by the llrookhaven Nationallabo.
shell thickness as recommended in Section Ill of the j
rutory (ilNI.) technical staff. 'Ihe relatively high RV
- Code, i
metal temperatures at the start of the transient placed the 1
RV steel outside of the range of applicability of the Code 4
procedures; therefore, the condition could be dismissed Critical Crack Depth Calculation j
by inspection, even though the stresses at sorne elements were higher than those used in the following calculations'
'the procedure used to calculate the critical crack depth, By inspection, the worst combination of parameters oc.
is given in Appendix A to Section XI of the Code.
1 4
curred in the RV shcIl just below the closure flange (hg-According to Article A-3300:
i ure 11-1, element numbers 307 through 311).*Ihe stresses
)
and temperatures listed in Table Il-I were used in the K * "mMm (na/Q)1,8 + o M. (na/Q)p i
n j
fracture mechanics analysis, where: Mm can be found from Figure 11-3, which is a
^~
Stresses at the finite element centroids for the kicked
{
flange interface (infinite friction) case were plotted against position across the 12 inch shell thickness in hg.
Me can be' found from Figure 11-4, which is a ute 11-2. Extrapolation to the mside and outside surfaces copy of Figure A-3300-5; resulted in inner surface and outer surface stresses of 17 and 6.6 kst, respectively. A technically correct method of Q can be found frorn Figure 11-5, which is a copy handling the nonlinear stress distribution would be by of Figure A-3300-1*
factoring it into a u niform stress, plus a linear stress gradi' ent tangent to the actual distribution at the crack tip (sec For the given flaw parameters: all = 0.1667 and alt = 0.25.
t the dashed line on Figure B-2), plus the residual curvilin-car stress. For this analysis, a less cornplicated and more Enter Figure 11-3 at a/t = 0.25, interpolate to all = 0.1667 conservative method was used. 'Ihe inner surface and and read M. = 1.186.
i outer surface stresses were connected by an imaginary Enter Fi ute B-4 at a/t = 0.25, inte'P0 late to all = 0.1667 straight line, then that distribution was treated as if it 8
were the result of a tensile (" membrane"in Code lan, and read M3 = 0.8.
guage) plus a linear (or " pure") bending stress. The com-l ponents of stress distribution were found to be:
For the given stresses: (o + o )/o. = (11.8 + 5.2)/55 m
e y
= 0.31.
membrane stress = u = 11.8 ksi; Enter Figure B-5 at all = 0.1667, interpolate to bending strerb = om =.1.5.2 ksi.
(o + o )/o. = 0.31 and read Q = 1.212.
m o
y Additional parameters used in the calculations were de-l termined as described below, Solve for a, substitute the fracture toughness for K and i
let a = a :
c i
Heing generic, the analyses do not pertain to a specific a = (Q/n)[Ku/(o Mm + v3Me)]2 H&W 177 RV; therefore, the reference temperature for c
m i
the nil-ductility transition (RTuor) of the material is un-known. Following the guidance in Section 5.3.2 of the where the arrest toughness, Ka, was selected rather than NRC Standard Review Plan, a value of 60'F was as-the higher (less conservative) initiation toughness, K.
i sumed.
From Table B-1, the RV ternperature near the crack tip position is about 150 *F; therefore, T-RTwot = 150 - 60 =
Although of relatively small importance to the results.
90*F. Enter Figure B-6, which is a copy of the Code when the tensileyield strength, o was needed,a valuc of Figure A-4200-1, at +90'F and read Ku = 72.5 ksi y
about 55 ksi was used.
(in.) u2, H-1 NUREG-1374
i Reactor Vessel Evaluation Substitute values and calculate a,:
Flaw aceptance criteria for steel components 4 in and greater in inickness are given in IWil-3610,Section XI, e = (1.212/n) (*/L5/[(11.8)(1.186) + (5.2)(0.8)))'
AShiE Code. Becaure the region of the RV under exa.
s
= 6.15 in.
mination is close to a change in shell thickness (see Fig.
B-1), the appropriate Code paragraph is IWB-3613. "Ac-Since the critical crack slie, based on conservative valut s ceptance Criteria for Flanges and Shell Regions Near of several parameters,is about one half of the RV shcIl Structural Discontinuities. Funhermore, the calculated thickness (12 in.) or about three times the size of a flaw pressure after 15 hours1.736111e-4 days <br />0.00417 hours <br />2.480159e-5 weeks <br />5.7075e-6 months <br /> into the transient is 350 psi; there-that might escape detection and about ten times the size fore, the applicable criterion is the one given in IWD-of the smallest flaw detectable by ultrasonic testin8 3613(a) for conditions where the pressurization is not (about 1/2 in.), it was concluded that the natural circula-more than 20% of the design pressure, i.e.:
tion cocidown event is not likely to challenge the integrity of a ll&W r vessel' K < Ki./(2)"8, ASME Cale Section XI Analysis From the preceding calculation,14. = 72.5 ksi (in.)"8, U the procedure given in Appendix A to Section XI of so Ki./(2)"8 = 51.3 ksi (in.)"8, th% ode f or a postulated flaw as prescribed by Appendix 0 to Section 111(Design Bases)in Article 0-2120.
Since the toughness reduced by "the safety factor was greater than K,i.e.,50.6 ksi (in.) 8 the Code criterion i
a = t/4; for t = 12 in., a = 3 in.
was met. Certainly, if the transient were treated as an emergency or faulted condition where the Code accep-From article A-3300:
tance criterion is:
K = o htm(na/Q)ut + o hin(na/Q)"8 g,<g,f(2)us, i
n Execpt for the crack depth, a, values of all variables were given in the preceding section. Substitute and solve for the significantly larger (than Ki.) value of K, the initi.
i K:
ation toughness, would be enough to pass the test with a i
wide margin.
Y. = (ll.8)(1.186)(n3/1.212)us 4 (5.2)(0.8) (:r3/1.212)"8 = 50.6 ksi (in.)"8, The evaluation, based on a conservatively large 1/4-t flaw, demonstrated that the subject transient would not induce which is significantly less than the previously determined a failure of the RV. Noting that a natural circulation toughness:
cooldown event is a rare transient, it can be concluded that the analysis showed that the event is acceptable by K.- 72.5 ksi (in.)"8 AShiH Code criteria.
i Table B-1 Stresses and temperatures reported by BNL Conditions:
1.
flange interface locked 2.
15 hours1.736111e-4 days <br />0.00417 hours <br />2.480159e-5 weeks <br />5.7075e-6 months <br /> into the trans ent i
Element No.:
307 308 309 310 311 I
11oop Stress, psi 15,170 12,760 10,860 9,145 7,455 Temp., 'F 149.7 150.5 150.7 146.2 152.6 l
l l
NUREG-1374 B-2
- =
1 Reactor Vessel Evaluation j
i
. +y
{
'
- A
.., l... IE m.
s.
4,$x 39.a
,., l.. ? _,.,.,
na,,,
,, t..,,,,
, gxh, L.,
,e.
,, c_
,,. q=,
.n 31 11L cL OL n2 Nf3L (31 j
]
- 1L tai uta ut cz. at at us.
al j
s2L 21. nu!L 11L (IL PJ1. GL fE
- 1 i
ut i
EZL 2L dL(CL 1U.
(2. E ffJL C2.
6CL 8
s2L 21 n:3L
- n.
Ist m uL nl 4m L5 L m a nn E
m u ul E
a QL T.L HL:21 22L 3m. EL 322.
2n.
EL
=.
EZL 2Lnm at Z2. m 32L
' 125.
22L c22. 2c. na ut ara. nt ut u2.
mm.
11 ML i
w4wm/
4 a nt si az. n1 3:in.
us.
- tin.
di u
2iL L*L n sin.
ut za.
EL l
15.
in. n 112.
10.
sit j
11 3.1a. ui uL 222.
su.
ct 3:2.
- 33. n ua.
ut m
l ui 2:2.
rt n 323.
21L 3:L
_itt 11L ILL 31L E2.
w m
1 222.
- 2L 332.
un.
2n.
---> 1# %
Figure B-1 Finite element model element numbers (RV head and head flange juncture)
B-3 NUREG-1374 1
Recctor Vessel Evaluation f..,#
, i '
t j ^, -.
_g_--
r t.
,t, z..
3
. ' - ~...,
. a. :
.y
)
~
m
. L.
H j
.l.
}
I I i
.f.
7.:..
p*
L.,.}'
t-6 i
I
- l[
ll l. i
....l.....;.-
+; 5_ A'l
... --7--
L-
+_ 1
--.=a.--p-g~
i
.p l
l i
i.
I I
i j
Jg.a.h 2 y%...-=;
i-.
2 __
.3
- 1.. _:.:.
.M i
~~.:$2RkM5 '.5
' i
...f.._
i
-- Ei i
1.
.N t
e
..g
.-,. A p y;_-.. _...
- 1..._ ".._.y._
+:
c:
M -
OI.t.=:t..
- vf
- A. E
,,,,. = _
.--E=.=...
_.. ua. t c.: b._
q
- q' ' *
,4 jg,
_r'
.:p.d 2-ty.
.a.
.J
- 41. ' ?
?;; ?*.fJ: (. '.
- 7
- l.
1-- -~ s J-
. V :.-
L - r y
--]-- -- r.- -
i b,;t.
- 1.. :
i.
w:
=:. =.-
.:p' :
'. m
.. _ _ _ -i-f 6
i.-
..* ~ '
..1. :. ry=
-s ::82.= Wi?.N 5 C.' b * '
i' i
n0u
- m. e r_.cq ;-
, \\3
,e-
.. _.. =. -.
.:..:..__ _ _;..1 u 4
.,,_,,.- [ :
s~ b 25 NN.[t - hb
'\\
1-
. I.* -. r.:-
- p.,
I
.m-l t
T
- f _i- _:i :n.' U !. _-
4..
- i. ' b. E :N T 5' b ^
N'
- ':. d ;__? '?'.[.5 E?.
~~
=.
i
_,_.,'(.,
_' _ -. l,.
..::..L..
.3,. Q...
_.._r..
_. =.
- t=:.,
- L." ;..
b
- : A'O.;;'
-*t.'*-
..- :4:.: s... j..:
M
- _.p :-... _
=, :..-.. -'
o
=.:
. 5-
%=......-.-
i
.h
- ~.=
= - -
- t..
'. cc; :-
......l..
~;
..3
_M i 2
1 -
In..::
a ::=: =: p;_.
.s.
- - 4.. -. _.p:-
.nU&akW.Mh,
m
....ib.
....' :C. ~*"1 '. JCID: E d.'.
-.. 3lM.599TY ~M- - -- ---\\.
8 Un
-- j m ---
.1--
. L.,. ;-J
- aL_===.;'- =s"~~? L=.*
- - ; E
.I:-l
- !.W i..-
'l-sr.
J.- ;
- .
- s.
...r- - y--
m-
-- M r.
_r-22 ;
N Q... _..t. - :.
. l... -
. g__... ;.n,.. -..
-- - i.
2'....
- c. c.
- r.
- .2 :.
l i: ;:.
.r r
-. L.,.. m.-l..e-
- .:t.:.
.r.,..
i l
z,.:...
- 4::::..:w..w..
.r. _
.~. =... _. =. _.
. t-.
. d.. J.,...i.::.:.t_ ::..::: l
... l..- E..
..;t..;
. h. _.
.-+_._
..:._-:...,. :. :n. =...
.1.*.
- *~ f
- i:
- (
.f
- ' h/
.l i
- e
.l-j.
.N...,.
- - 3, -
- -
..y
~ ~.
p.4 :..... ;
i
- i
..:.; y. - - -
1.
j' _._
.l i
4 i
a_ _..
. 1...
,.- : #..i.:. g l
y.
i
..J.-..
An
+..._{.. '.
_:i.
1-,.
~ " '
w.
.e
.l.
.q i
i i
I 4.
F,
- 0..... l. N.
. *D
- h.,
tQ.....;. 1 L. 1...
I i
m ckie r wre n -m. verwet wau.
Figure B-2 RV shell stress distribution NURliG-1374 B-4
Reactor Vessel Evaluation 4
i APPENDIX A - NONMAFC ATORY rig, A.3300 3 l
i l
i.
2.0 t=0 0DS j = 0,1 j = 0.15 i
i 12 -
I*
i i
I Y = 0.2 l
1.7
- e i
+ t-e-f 1.6 -
f E
p = 0.25 g3 i
+
I j
1.4 -
I 4=u u-e
{e l80 s,2 s.t u
os th 4
s i
e a
e i
i i
O O.1 0.2 03 OA 05 OA 0.7 02 reaw Dese so Thicknees Reti. a/s i
j t = wee 81 tNetness.
l t =, i.e.4 et es circ eua*.e e new t
OC. A-3300 ', MEMSRANE STRESS CORRECTION FACTOR FOR SURFACE Ft.AWS i
(Reprinted with permission of the ASME) i Figure B-3 ASME Code method for M. determination l
l l
l B-5 NUREG-1374 r
n.
y
.,y-r---
Reactor Vessel Evaluation 4
APPENDIX A - NONMANDATORY Fis. A 3300 5 f
ie ya e/t-o e
n i.4
- f o else.
tocep.
j i.2 -
+
+e+
NgN's i.o -
a
..... %N N
.g.. w.4........
..f.'~.~..e / t os N %
4 m
/
i
.............. s Q...
.W..n.r
.~.r.~ 5 ---
8 cm
_x
=
,fr.oa AN' %
t sN N
% % s ~~.
T NN N.'-
y NN NN o.s -
N
.N,-
e / t - c.2 N
g s
N N
N N
N e/t-oJ e
,woti.aw-om3 N. e / e - o.e e,ti
. u - odesJ N e / t o.s estim.w u toe.,a l
o.s -
N i
o o
l a
i i
i i
i o
o.:
os os oA os o.s i
Flow Depth to Thicknees Pletto sh e -on thacun w t-meio< e.
et eneese se<cumeerW te= ha FIC, A.3300-5 BENDING STRESS CORRECTION FACTOR FOR SURFACE Ft.AWS (Reprinted with permission of the ASME)
Figure B-4 ASME Code method for M. determination k
NUREG-1374 B-6 i
e m
a m
1 Reactor Vessel Evaluation j
2 f
APPENDIX A - NONMANDATORY Fig. A 33001 j
1
]
}
1 0.s 4
i j
(e e egl / eg m
0.4 -
,y u
0.8 O.3 -
8 E
d 0.3 i
i 0.2 O
o.1667 i
.I O
y 0.1 -
4
-0 i
e i
i e
a a
4 0.6 0.3 1.0 1.2 1.4 1.6 1,3 2.0 2.2 J
Flew Shape Paremewt 0
)
i W'4 V' J.
1 w
r
}
E22).x
{
i a
l
._ 2 e i
(s) Sortees Fleur M Submerfees Flow I
e, sesenw m.,omum vieid nren,in r
i t e o.a.e ene eirevm. era,an, ene f 1
FIG A-3M01 SHAPE FACTORS FOR Ft.AW MODEL I
(Reprinted with permission of the ASME) l Figure B-5 ASME Code method for Q determination i
l i
B-7 NUREG-1374
- _ _ _ _ _ ~
. _ _ __ _ _ _ _ _ _ ~. _. _. _ _ _ _ _ _ _ _ _
l I
I T
x' C
z a
i m
o.
j g
220 o
i 5
m m
2-i m,
e_
c l
E too e
3 i
=
t 1
g i!
iso
- i Cb Q.
Kge
, 140 N
E b
5 gla M
(:1 k
Z 120 o
O 1
9 I
L a
a w
l 2
5 100 I
a f
1 T
n h
d 3
~
7 2..r ksl&
5 l
Q i
x i
O a
i N
eo D0 x
o.
l 0
x
=
a g-t i
o Ei 20 4.
j s
=
aruar j
b iL o
I f
f e
.1Oo
-50 o
+50
+100
+150
+200 (T-RTycy. 4 3
[
n t'
FIG. A-4200-1 LOWER SOUND K,, AND Ne TEST DATA FOR SA-533 GRADE 8 CLASS 1, SA-508 CLASS 2, AND SA-508
.l CLASS 3 STEELS g
i (Reprinted with permission of the ASME)
I I
6
i i
1 i
l APPENDIX C 1
{
Fracture Mechanics Analysis of the B&W 177-Fuel-Assembly Reactor Vessel Head Closure Studs During a Natural Circulation Cooldown Event
}
Richard E. Johnson e
l 2
k i
i 4
0 i
i I
l t
i c
l i
Background and Summary ing," gives relationships for stress intensity factor calcula-tions. De relationships are listed in Table C-!!, which a
Reactor vessel (RV) head closure studs were analyzed provides the numerical dirnensionless proportionality using ASME Code procedures for the case of the ana-factor relating N to the product of the nominal stress on lyzed natural circulation cooldown event in a B&W 177 the minimum cross-section of the threaded region, o, and plant. The analysis involved the calculation of the mode-the square root of the cnick depth, a, for values of the one stress intensity factor, K, based on the stud geometry relative (dimensionless) crack depth. Solutions are given i
j provided by B&W, the stresses calculated by the Brook-for two geometries: a circumferentially notched cylinder haven National laboratory (BNL), and the reference flaw and a single edge notched (SEN) plate. For both, tension and equations prescribed by the ASME Code, ne calcu-and bendmg loading conditions are addressed, ne SEN lated value of Ki was used to enter the ASME Code curve values in Table C-Il were used to prepare the curves in of plane strain fracture toughness (crack initiation tough-Figure C-1. According to Section 7 of Reference C-1, ness), K,to obtain a value of relative temperature threaded fastener analyses must use the notched cylinder i
(I-RTsor). A conservative value of the reference tem-valuesfor K (tension)andtheSENvaluesfor K (bend.
i i
perature for the nil-ductility (fracture mode) transition ing). K = K (T) + K (B) by the principle of superposi-i i
i (RTsor) was established based on information provided
- tion, by B&W, Stated in a different way, the purpose of the 1
exercise was to determine that temperature where the
- 3. Flaw RPV stud material exhibited a fracture toughness equal to the calculated stress intensity factor. Of course, if the in the threaded region, the total flaw depth is the crack input parameters were accurately representative of the depth plus the thread depth; B&W flange stud bolts gen-1 actual stud material rather than conservatively deter-erally use 8N threads that have a thread depth of 0.08 in.
1 mined, as they wcre, the equality of K = K, would signal for all diameters. For nominal diameters greater than 3 i
i conditions for fracture instability, in., the reference flaw (crack plus thread) depth is taken to be 0.3 in., acrording to Reference C-1.
To assess the safety inherent in the RPV studs, the tem-perature at which the stress intensity factor and tough-
- 4. Stud Parameters ness are equal was compared to the lowest service tem-perature (LST) for the studs in the given transient event.
The stud geometry factors were taken from a drawing It was found that the temtcrature corresponding to the provided by the B&W Owners Group, a portion of which above equality was 108'F. According to calculations re.
ts shown as Figure C-3.
ported try BNI, the studs would not go below approxi-mately 200'F during the transient; more likely the IST
- 000 D (gross) = 7 in. (nom, al); = 6.687.010 (max. on dwg.);
would be 300'F or more.nerefore, it was concluded that m
l because the temperature of 108F was well below 200'F, D (shank) = 6.25 m.
I the fracture toughnes.of the stud material would be more than enough to survive the transient without any reason-I.D. = 1 in. (axial bore).
able likelihood of failure.
+.000 d (at root of thread) = 6.33 in. (dwg.: 6.336
.010 in Analysis dI"5 I j
- 1. Stresses The thread depth = 0.08 in., so the maxunum diameter (at the threads) = 6.336 + 2 (0.08) = 6.496, or approximately 6.5 in.
The transient being evaluated would occur when the plant goes into shutdown with the main reactor coolant Reference flaw (based on ASME Code requirements):
pumps tripped resulting in coolant stratification and asso-ciated thermal stresses. De transient was analyzed by for D > 3 in., a = 0.3 in. (a = thread depth + crack); thus BNI stresses in the RPV closure studs from the reported the crack at the root of a thread will be 0.22 in. deep.
BNL results are given in Table C-1.
The relative crack depth, a/d = (0.3)/(6.5) = 0.046.
- 2. Stress Intensity Factors To evaluate the adequacy of the stud material fracture resistance. the procedures given in the ASME Code were From Table C-II, for tensile stresses (using the notched followed. Appendix G to Section Ill of the Code refers cylinder):
the analyst to WRC Bulletin 175 (Reference C-1). Refer-ence C-1, Section 7 " Toughness Requirements for Bolt-K (f) = 1.98c(a)*.
i i
l C-1 NUREG -1374 l
RV Head Studs Analysis From Figure C-1, for bending stresses (using the SliN closure studs are manufactured from SA.540, OR.ll-23 plate):
(or 11-24). For the usual minimum specified yield strength of 130 ksi,the actual oys willbe about 160 ksi/lhe current K (11) = 1.90a(a)"2 Code requires 45 ft Ib Cys. Generally, the steels used as i
bolting mater'als will reach their Charpy V notch impact Thevaluesof stresstobeusedin rohingfor K,eccording test upper shilf energy at about + 40'F. In a tclephone i
to the Code and Reference C-1, are based on the mini-conversation with A. L 1. owe, Jr., at il&W, Lynchburg, mum cross-section in the thread region (i.e., where d =
Virginia (one of the Reference C-3 authors), it was 6.33 in.). !!clieving that llNL based the stress calculations learned that the proprietary version, ll AW-10046 F, has on the minimum (thread root) diameter, the values re-a data bank (on page 3-27) from which one can deduce ported in Table C-1 can be und directly. At the onset of that RTuor = 45'F or less as the temperature where the transient, not only are the combined stresses rela.
closure stud steels meet both the energy (45 ft lb) and tively low, but the stud temperature will be high, the lateral expansion Charpy V-notch criteria. Reference fracture toughness will be hig'i, and a fracture mechanics C-4, in part, states:
analysis would be of no value. Therefore, the stresses I
reported for fifteen hours inta the transient are germane.
"Iflimited Charpy V-notch tests were performed Thus:
at a single temperature to confirm that at least 30 ft lbs was obtained, that temperature may be K = K (T) + K (II) used as an estimate of RTuor provided that at i
i i
least 45 ft lbs was obtained if the specimenswere
= 1.98a(a)us t 1.900(a)ut longitudinally oriented. If the minimum value obtained was less than 45 ft lbs, the RTuor may
= (1.98) (42656) (0.3)u2 be estimated as 20'F above the test tempera-tum
+ (1.90)(61371) (0.3)"3 psi (in.)n:
Therefore, RTwot = 45'F, and for T-RTuor = 57'F
= 46.26 + 63.87 ksi (in.)"2 from the construction on Figure C-2; K = 110.13 or approximately 110 ksi (in.)"2 T = 102'F i
as the temperature where Kie = K.
- 6. Determine Temperature for K = K e i
t The next step is to determine that temperature where the
- 7. LST Comparison; Margin toughness equals K because,at all higher temperapres, i
the material will have adequate toughness. The final step Since the studs will be exposed to no temperature less will be to compare the lowest service temperature (IST) than 200*F (more likely, no less than 300'F), there is a to that where the toughness equals K and assess the temperature margin of 98'F. On the same relative tem-i margin for failure. According to the ASME Code and perature scale of Figure C-2, the vertical arrow identified Reference C-1, the applicable bolting toughness require-as "Tmin" shows where the IEF for the subject transient i
ment should be based on the minimum static plane-strain would occur. The corresponding plane-strain fracture i
fracture toughness [the ASME Code lower-bound curve toughness, K c, would be well abo e 200 ksi (in.)u2, g,,,,
i for Kic = f(F-RTwor)] rather than the atrest toughness, the material would be fully ductile. With a toughness Ka, because of the absence of (1) dynamic loads on bolts about twice the stress intensity factor and the direct pro-and (2) significant strain rate sensitivity in bolting alloys.
portionality between K and o, the margin in stress also is Therefore, the applicable curve is the one for K e in at least a factor of two. Noting that the usual practice is to i
Figure C-2, which is a copy of Figure A-4200-1 from perform some nondestructive inspection (at least visual)
Reference C-2.* A value of 110 ksi (in.)ut on the K,,
of the closure studs at every refueling outage, the possi-curve corresponds to r. temperature of 57'F on the bility of a crack as large as was postulated for the above T-RTwor scale. Since the reference temperature for analysis is unlikely, and it can be concluded that the sub-studs will not change with time in service, the initial ject transient will not induce closure stud failures.
RTuor should be used.
References From Section 3.2," Impact Properties of Ilotting Materi.
als," of Reference C-3, it was determined that il&W RV C-1 *PVRC Recommendations on Toughness Require-S
- Anide A-1100. A nda A of Section XI a plies to ferritic materials 4 in. or more in thi ness and with a specified minimum yield strength Council Ilulletin 175 by the PVRC Ad Hoc Task ot 5o.o tsi or less. Aho. it may be extended to other rerntie matenah."
Group onToughness Requirements," August 1972.
NUREG-1374 C-2
I RV llead Studs Analysis l
C 2 American Society of Mechanical Engineers,lloiler Revision 2, liabcock & Wilcox, Nuclear Power Divi-i and Pressure Vessel Code, Setion XI, Appendix sion,1.ynchburg, Virgmia, December 1984.
A.
C-4 Nuclear Regulatory Commission Standard Review l
C-3 11. W. llehnke, et al.. " Methods of Compliance with Plan, Section 5.3.2, " Branch Technical Position-17tacture Toughness and Operational Require-hffEB 5-2,17racture Toughness Requirements,"
ments of 10 CFR 50, Appendix 0," IIAW-10046, page 5.3.214.
Table C-1 Stresses in the RPV closure studs for the natural circulation cooldown event based on the BNL report.
Time into the Stresses,* psi transient, hr Tension T+B Bending" 0
49,013 60,593 11,580 15 42,656 104,027 61,371
'I' rom the IlNI analysis of Case 1, which is conwavative. the results being higher than for Case 2.
"llendmg (II) stresses by difference:
11 = (*1 + II) - T; where the tension (T) and combined (T + 11) values were reported by llNI.
Note: IINL induded a tensile stress of 34,914 ;si from stud prestress in the T value reported.
Table C-Il Stress intensity factor solutions as given in Reference C-1.
K /a(a) ira Notched C)linder SEN a!d
- Tens, llend.
Tens.
Bend.
0 1.98 1.98 1.99 1.99 0.05 1.98 1.98 2.00 1.89 0.10 2.05 2.11 2.10 1.85 0.15 2.27 2.64 2.25 1.85 0.20 2.64 3.47 2.44 1.87 0.25 3.10 5.03 2.67 1.92 Note: stress, c. is based on the minimum cross-section in the threaded region.
C-3 NUREG-1374
4 4
4 5
j RV llead buds Analysis
)
l 4
.s s.u sg. e...e..n....
...s,a,<..--
)
..rt::11 *In 2.n :-*".' *ITn'l.
.i-A((
!*T=:
- !T
' 3..
~ " * * *
=. -tz..
JC ~~g;
- Cj=
j*It
.. +. iU *** :M
_4.
t!!!OEdif
)
- -b iE ~r ntE Eli ikii NITC ent
. +. _... ;I 2F O
]
2.r'..~i CEE ii?
C 2
n 1
- [rj C.*.*.'pt;'===0
""**"**nr~.C;.e J: OnP ::=+31:
1t11
- J1==.;;;
.*L =r'r=.
'7 t*-'~*-
it * =*O'
= f-r :.* ' :
4_ J _
2 1
-*-~ ' ::nn.. ;p:= :- =
===-
4=
==
- itt t=$;;;22==+
i._6 2 4;g '1.u
-.4
-"~~
33: nt:
.e 2-
.t!:d :13l' ((:=
=
=.
'I: '":1 =1...
!'j'.
~n n- ::
~"
i*.
= jUC
- p[*
=== :-
'~
1
!*.i in:
En"""=im 5 1~ ::0
=n c
- . un
,1,
~=
=
u:
}
- =
.-4
- 1
- !~
'*.7 ~ * "
t--~ e4
.._=.._*.:_-.::.**.
- 1 D.a.; In:
=_:. :**
'_p**'
" =g"
" j :}!~:::
g
... _., _ _ _ -g?
g %.*_""-'1**
.a.
r
"_~I~~~
~ ~ ~ ' "~~ O ~~1 C =UU 28 CC IO C
=
C
~*'
I"
" * * +
=
- CO=*;
" " * * ' * ~ ~ -
f
- T!!.i;;L-4-..ee
_..~. _
- 2 a t==
0:.-*
- =
.2a n! = c-m-
=
- A a :'*!
=.
.n
- :t:.-J1.
"?!
==
1
= - n*: IE.. =_:rm.
- .2% e_z n.:.gE
?:~i_
I.'
4
$t r rn02:
- 42..t In rb II= =*'
. il li;l. C.:. fp.1
- , =
a
.*d'-" :*'
)
. _.{..:
~2:.r~f
?-
- t;
.t!.i:'== * " On
- ! Or J
.i.
=
IT!!
,i J.n*ilj Iz.II
.2 -
- C; **: =:**C=
=:~i
- tt
-.p-m M33E P14
-F4'~
ig ;5:[
E
.~.~ M;E n
. _..E-',;
=
..O ti
- t.nn
~~--
t;n **::*.:=='-*"* * * *
.u.===
!==
rn m:::::=ng:::
I
- "1
_.- -. " _.s_..
Cl;
^
?.
~~t-
+.!'
+';.j{f T *r w.-
... gg.:**
g'
- {ifI't.:7
}
~~*
n
- L::
.T: :t! =.;.;.=.....t.*.[2.1.f ::
. 1,:{'1,._"U._g}._,- ***
4.;
t{;
E
.g
_.. ~ ~ ~.**N.
'*?-1:n t_;..-}p..
.g, 1
,1 u
au== yt= 2:+T=-=-un:n==
_-.. _=.:
1
. _. a-
- u w_un
.: un a
r
- pc
fn ::= nn (=:rr ::p p"p:
==o 2==:
:
.- b.
pt;==.. _ =
- :=:$:%t.i#t _t;}i..F:- *Ei m-m.n :mtF# liit':lossk==4F,: Srf.i,. tt'tm" j3f*il1
- 1 5*
- n.tn tt u -
d
+ + ' '
i n
n Jr =; $
utd:
i
=
=n :m :n mtJ.IE-
~m* +
-u' 1-u E"5s
$:+ met!!i Eli!H ;rl*!!t4t a
- s -ni
--a-
=
n a
fE
..m.
4, y
~tt:ThrHE EEEii
+
=r EEt:
g11 E;.--2:
in iiii?!EiEira nt. gpaniW - =
st* t.trtwir ~
- +~+*
+
- t: int =
= _,.
t
~
-g-*k- -
'tt --+tt J:ty =
g4'll-g' 2:=
nor =! name;3n. * ' ':
ng=2:
- :{r:- IEI"I g
- at tr lI
...n =t=
II*- ::I:
E 2' *E-
. _.Eli 2
- 1~1-n
- .,d,.i: l.
=nur d
tp
- pt i EK m:.E li
- _ifE. $c
.i ilE_ESE E
1,Il.ili@!h,
- EE
- i. -en n.,
,,, e B.li 1!i i
- t=: =r atm=:un:tili: f!!
L
=
.,.a
,.s. g g_.y 33
- u.- _.....
I$itC?
-b
. #i*".a=5555555:15 E}:.E.EE-\\v
u:=
- n E
-=
___.-.a 4
=.=.._ h.ii g __
__ _ _8L.
...._=__=,____a==._-_
_..E...._u..;
1
_ = =,.
- . Da-
-.m j
-+ :in ;gn=.
- . p
. r;r;
.=, y=j
_.,_IM$.-
! M :: 03=.Mr:
=
==
_E.r'![1ii2EiilldlEN.IEEM_. t!.!!ilin~!Y.:".a*5I.j; g.:4- _,
a*f_j!rn.a!. y;}; 3]t t+.t (33 :_t [..tu,2
-+
_edp:
.9,, =.2_.1gn ++-t t
tt-2
,u.
n%__._
.g.4t: r.n:
e m
n: w:m
-h,ir,n
=" J
-o
- E4.G.,jet i.*n"u.=.=.. =. _ =
.+.+'.t.
- n. i;..iis.s..e.e:ti.-.ga,-. :*n f,e.O.y=
t
==
+
n
==
t 4
u...g__.
+p;rgg,. -
. hyr:n ifpt".1 ;
- +
jyiff+ m.:.4nn..u. :t=
.!t.
- x
=-
4
=
?--
~
e
.+_...
+c.
IpEtth_t~+
.it*
g
_._iiE4 ff;* n'
- :!is u _iE EF.tE 3Eii m.
u.
tt y
..m...g.y,.y.
-&..=_ ;m..,,.rs
.g
=..=.a.=._==..
L.u g...:y5=..;.tt,;
7:
. _ n t
ny:=
=..t..
= _ =..
1! hift5E 5li5?*i MY :
=n=- EE
"~
- 4:2#1un;E 1 @
_ ini:
i
-.- ~.
2rM g~
- m
- ni
_4.._natit E_r_Q,, _n=.=
4
..a m =
l
=r.._
- pwr=
- ((.=..
.
.u..ta:t uuna.
- .,=.;;cd g=_;=
- y..
aa
-r
- m
- =.
g._.
un
~
,.4
._,m, j.w.n:.
J'f.
. _.::n,.=.
_..;1Or...m. ** :::
I,p :n.@..p.=.n. "r3_jg p_=: :'. y_3,_4
- t...t
- n{.
... '..i h,,j}[ 7 ~. : i
- 3
= t. ++T++' . ~..** O_. O** =?%, . _ -.s-.Oi.1% =; _.".*.*_.:._=.C2nt:,pgg -f,t,_ i ._=.q..-.
- t
_ **3h- ..ib p~;.~C; k13.....!!n2d. l _ =kQQ_. E. .* _.7g; . :. l.::= ~__.! u., ITT; C. g *=.*.:*7_~. - -u.+ 3 an-
- ..=
.c=
==:2.: gu 1pn! - ;p.up:.p.m up-h,it F-t - e. _ :Inf+;.-:=r _. __w.- -e t .a =nm .n r.: m r-.ar i
- =
- u=
..n m.n... n ::- r-==t-mr + ,: unl- --=E=.= p. .. = g:r ni:t:4: 1-==tn
== r:t :=====- ._:==~ .= =
- .t
- ..==:n neste :-- - :n:"l=:n=:r==u.._A..
n r -- -+
- :=
=s=x.
t
n mh= = n-n....... -.... _ ..-= === = = l l l Figure C-1 Stress intensity factor relationships for SEN plates I I l I. I I NUREG-1374 C-4 t i -,-p
RV Head Studs Analysis APPENDIX A - NONMANDATORY Fig. A.42901 3 5 i I f "8'1 i g a i R d 3 .gv 4 8
- s. u i as g,t d g a 8 la 0
1 = $. nag y -8 w3* 1 = = I t = 5 I E 7. 8 g .x =I e ~ g t 3 A g 8 g a s a a a g,,, w e -en.-, x Figure C-2 Copy of ASME Code, Section XI, Toughness curves with the construction discussed in the text C-5 NUREG-1374
RV licad Studs Analysis fN h t d A W \\ ,. s.. ~%f777 3 /l,/ A/ A M, (po /4; a-is
- g 1,l ll ig P3 % /
/ r/ [' j/ ' // w p $5 . / /_.. .1l1 *! .'2 an n-u )....k, I 1 ]O. t tB plg l.6,,,,. o...m. - -. ~.. s }- set E e' ,= q' y 4 f il I ks et M e el g. O> l a, j '> co ,-- _~~. i;. ?!!~[( f" f .k k\\hh\\' _.g/ N' t 3 s l M +i $$2 N ~. - s O o, u l 's. 2 M 9 f~ ( 9.. ~d dj.l5 a m
- 9..;.
4 % t@ fi ?' ht AT W { e Y w
- g 7, 2....7L Q
ll C. s = &,.h.5 V ule u..n a. A \\1 .{ eEg
- t-
.N.Esuo " 5 P I g. oneasa g~. f i r#i8m ~ ~~ rm e =. a s s s Nl 'I ,. ) t. it wt a Figure C-3 Portion of B&W Dniwing No.142159 E Revision 9 NUREO-1374 C-6
FJ40 FoHM us U.8. tAICEr AR REGULATORY COMMSSION
- 1. REPOR.T NUMBER (r.sei
{ Assign d v., NRe, Add Vol., ty NRCM t 102, upp., R a9d A63.ndum Nun'. 3m. 22 PIBUOGRAPHIC DATA SHEET
- d r 1 (s instruccons on ih. r.v.rs.)
NUREG-1374
- 2. tif LE AND $Uh tsf LE
- 3. DATE $4EPOHT PUUUSHLD Technical Findings llelated to Generic issue 79 g
,30nys ygxn An Evaluation of PWR Reactor Vessel Thermal Stress During May 1991 Natural Convection Coold7:m 4, FN OR GRANT NUMBER
- 6. AUT HOR (b)
- 6. T YPE Or-HEPOHT J. D. Page Topical
- 7. ecR.Oo COveREo (irciusiv. cat.si NA
- 8. FtHF OHM >NG OHGANIZAllON - NAML AND ADOHL bS (if NHC, previo. Divisson, Offic. or Ft.gion, U. S. Nuci.at H.gulatory Commession, and mamno na$.as; it conveciar, prw o. n.m. and mamna aoe.ss i Division of Safety issue Resolution Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission Washington, DC 20$55 6
bHONSOHI U S. Nucl.NG OHGANil AllON - NAME AND ADOHE SS (if t#4C, tysm ' Sam. as atov."; it contractor, provid. NHC Division Office or Revon, ar Regulatory Commission, and malling adG.ss.) Same as above
- 10. SUPPLEMLNi AHY NOIES t t. ABSTRACr (200 worps or I.ssi
'lhis report summarizes work performed l'y the Nuclear Regulatory Commission staff to resolve Generic Issue 79 "Unanalyzed Reactor Vessel (PWR) Thermal Stress During Natural Convection Coofdown (NCC)" The report evaluates the effects of an NCC cient on PWR reactor vessels (RVs), with particular emphasis on the closure flange
- region, A conservative independent confirmatory stress analysis of a ll1W 177-fuel assembly RV (ll&W 177) was performed by the NRC contractor, and an independent fracture mechanics evaluation seas performed by the staff. Ilased on these and a comparison of geometric similarity between the Il&W 177 and other PWR RVs, the NRC staff developed findings that are applicable to all U.S. PWRs.
7 kev WORDS/DESCRPTORS (Ust words or phrases that will assist r.s. arch.rs in locatang tn. t. port.)
- 13. AVAILABluTY ST ATEMFNT Unntnited Generic Issue 79 PWR Reactor Vessel Thermal Stress 1C SE URW LASSFICATION Natural Convection Cooldown PWR Reactor Vessel Closure Flange Natural Circulation Cooldown PWR Reactor Vessel Closure Studs Unclassified Natural Circulation Fracture Toughness
(""' '" P'") PWR Reactor Vessel Ilrittle Fracture Unclassified Reactor Vessel less of Offsite Power 16, NUMULH OF PAGLS
- 16. PHICE l
NRC FORM a36 (2-60)
>_m ...s# w l i l l THIS DOCUMENT WAS PRINTED USING RECYCLED PAPER
1 UNITED STATIS ~ ..., n. v. u NUCLE AH REGULATORY COMMISSION
- 6 " ' (' * ' 'c' $ ' '
- WASHINGTON, D.C. 20LLS ei., i, o co
~~ of rivat eusi'.its F-t P. At T V F Ott f RtV All U!,1,13f) 120 5 %13 9 5 g g I I A '* l A I l l A 19 L US.kC-0AJe f[f.' pod.A qj,q[ f*IH 1: Ai!0NS syc3 v. ; ;, 3 i W 4 C H I '* G I O N OC 235 % L i .....,.,..,}}