ML20058N018
| ML20058N018 | |
| Person / Time | |
|---|---|
| Site: | Catawba, McGuire, Mcguire  |
| Issue date: | 08/08/1990 |
| From: | Tucker H DUKE POWER CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| TAC-75138, TAC-75139, TAC-75140, TAC-75141, NUDOCS 9008130171 | |
| Download: ML20058N018 (41) | |
Text
{{#Wiki_filter:= ifT iC > x,eyff ll: ,. l , l ' ' bl' Duke . IWer Compaiy > ti a lin R Twur. '1 L' x > PO Bat 33198 , , . bice President "F; Y Charlotte, NC 28242 Nuclear Production 1 (704)373 4L11 m p.; , DUKE POWER:. ,. : : August 8, 1990- ., 1 o :U. S. Nuclear Regulatory Commissic.n
> ' ATTN: Document Control' Desk
.' Washington, ' D.C.- 20555
Subject:
McGuire Nuclear Station Docket Numbers 50a369 and -370: '
. Catawba Nuclear Station Docket Numbers 50-413 and -414
;r Response to Request for Additional Information Regarding BAW-10174-(TACS;75138/75139/75140/75141)
'By!1etterLdated March 27, 1990, the NRC staff requested information.on
opical' Report BAW-10174, " Mark-BW Reload LOCA Analysis for Catawba and !
, -McGuire." ' Attached are responses to the last 9 of of the 29 Round 1-Questions. <1 ease note'that responses to the other 20 Questions were a transmit / by letters' dated June 7, 1990 and July 25, 1990. -
$1f1there are any,_ questions, call Scott Gewehr at (704) 373-7581.
;Very t.ruly yours, t
~
s
\ g
-Hal'B. Tucker i- LSAG/224/lcs Attachments t
t t t ) 00
; Sbbk000000s, " N PDC
jc .1, 0
' I ';
J -
,U.'. S. Nucire.r. R:gulctory Comicssoni Augurt 8,-1990
[ 3 Page'2' xc t - Mr.-S.1D. Ebneter, Regional Administrator
' U.. S.: Nuclear _ Regulatory Comission Region II, 101'Marietta Street, NW, Suite 2900-i . Atlanta,, Georgia 130323
- Mr. Darl S. Ilood, Project Manager
-Office of Nuclear Reactor Regulation i .U.- S. Nuclear Regulatory Comission i- Washington, D.C.- 20555 Dr. Kahtan 'Jabbour, Projects Manager Officetof Nuclear. Reactor Regulation U. S. Nuclear Regulatory Comission -
Washington, D.C. 20555= Mr. L.-L. Losh ! 3315-Old Forest Road I P.O. Box 10935 Lynchburg,-_ Virginia 24506-0935 Mr;-W. T. Orders . j NRC. Resident Inspector- l
~
Catawba Nuclear Station i Mr.'P. K. VanDoorn NRC Resident Inspector
!i -McGuire Nuclear Station _j l
h 1 l
+
p
~ '
- = ~ - , - , ,
g; , ,~ , . . f lif , . 7.
- f. .t:
.:... . r. :3 ,
- 7. ; ;, _
[' i'k. .',. } l f
'.1 e t te .
ATTACHMENT = i' RESPONSES TO QUESTIONS. n -
(\p 4, 8,.12, 14, 15, 18,'24, 25, and 28:
c, [ N t . }l l t
. .l l
l 4 I I I 1 i r
, o f l -. ' l s. J .
!s
. (-
p tc j! [Q)
- c
- 4. Thecreport stated that rod' insertion was not' credited until after- blowdown. The. Standard: Review Plan requires: rod insertion to be-justified by an appropriate' analysis before =it can be credited. Provide -a - blowdown load and structural
-analysis 1to verify that the c core - and upper. _ vessel geometry n will..' permit rod insertion at the time credited- in the analyses.
ea ' :k ~
~'
Response: As discussed in the response to Question 4_of the-
' first _ aund of questions : on. the RSG LOCA evaluation model, calculations of large_ breaks take~ credit for - reactivity
' insertion following blowdown in order to ensure a subcritical-margin during reflooding.
By . maintaining a core k,,, . o f approximately :- O . 95 ' the' post, blowdown fission power can be conservatively modelled as a constant fraction of the decay-heat power. This eliminates the=need for_ extended reactor kinetics calculations. The negative reactivities credited for 1this purpose have included void. formation, boron injection, and control rod ~ insertion. A recent examination of the-Catawba-FSAR'shows'.that, although the fuel assembly geometry
.following LOCA remains' amenable to control rod insertion, the upper plenum control rod guide structures :may, in some locations, be displaced to the extent that rod insertion cannot be assurert. Therefore, it cannot be concluded- that all of the control rods will enter the core by the end of blowdown. Because of this, a . revised accounting of ' post blowdown' reactivities has been done to verify that the reflood modelling of fissio'n power remains conservative.
Post-blowdown reactivity concerns stem from an assumption of the BWFC RSG LOCA model that post blowdown fission power can. be conservatively evaluated as a constant fraction of the decay heat power. As explained in the response to-Question 4 of the first set of questions, the ratio of fission power-to-decay heat is set at the end of blowdown and kept constant thereafter. Because fission power decreases more quickly than decay heat, this is a conservative treatment so long as a
'< negative reactivity of at least 5 percent is maintained. The
~ !
5h. yx 3 l' . 1; < >4L r , I contributors to core reactivity following the end of blowdown, q' without consideration'of the control rods, are' voiding'above. l
; - the ; quench front (negative reactivity), liquid , below thel ,R . quench- front. (positive reactivity), decreased fuel j
- p. temperatures (positive reactivity), and. increased. boron 2 concentrations (negative reactivity). Using information from-
, Table 4. 3. 2-2 of the Catawba FSAR,1987 update, . the reactivity ;
4 during reflood for the region below the quench front can be . lE shown to be somewhat more negative than -0. 05 without' .; j consideration of the core void content. (For zero - power,
' cold, with no control rods k,,, = 1 at 1650 ppm - Boron --
equilibrium Xenon is worth 300 ppm Boron -- giving kg,, = 1 at.
^
[ , reflood conditions with 1350 ppm Boron -- because 1 ppm Boron is worth 9 pcm,.k,,,.is about 0.9C for the reflooding LBoron concentration of 2000 ppm or p =
-0.05.) Considering .the effects of partial voiding below the quench front '(boiling) and near complete - voiding above -the quench front, the net l reactivity is expected to be between -0.35 and -0.05'Ak/k.
Therefore, the core. reactivity during reflooding lies.within-the range specified'in the response to Question 4 of the' first I set of questions and the conclusions of that response remain 4 valid without taking credit for control rod insertion. In summary, mechanical evaluation of . the control rod guide- j housings in the upper plenum of. the reactor vessel cannot I
, 1 demonstrate that all of the control rods are able to enter the reactor ' core following a large break LOCA. Reconsideration of l the reactivities during refill and reflood, however, show that the conclusions in the response to Question 4 of the first ;
round of questions on the RSG LOCA evaluation model remain 0
. valid- even when the control rods .are no longer included.
Therefore, the BWFC technique for specifying the fission power lg during refill and reflood remains valid and there is no longer ! any requirement that the control rods enter the core during a large break LOCA. de io _ _ . - _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
- i.% l?' ' '
N 4
- 8. iThA 'following questions are related'.to the containment pressure:
a .- A partial-justification of'the containment pressuretwas n discussed on page 4.9. _ Provide a comparison:of the mass-and~ energy: releases to the containment for the Catawba-t FSAR and B&W LOCA EM calculation cases analyzed.'~This-comparison-should-include-break-flow and spillage from
;both the maximum and minimum ECCS cases. Justify.that' the dif ferences between.the calculations are small enough to allow-= the use of. the. Westinghouse calculatedt containment pressure.in FSAR.with the B&W LOCA EM.
Response::: A comparison of the integrated mass and' energy
. releases is provided in Figures 8-1 and 8-2, respectively.- These plots.show that differences:inithe releases (particularly-the energy release) as predicted ~
by the Catawba FSAR and Mark BW _ Reload analyses' are - generally _small. The comparisons in Figures-B-1 and'8-2'are-based on the DECLG, . C= d 0.6, minimum safeguards ' case. Mass and=
.p energy release data :are provided in the Catawba FSAR
.(Table 6.2.1-56) in reference to the minimum containment pressure evaluation for this case only. The information-T
-needed for a comparison of maximum safeguards-cases is, therefore, not available. Nonetheless, it is not-
- expected, that the results of such - a comparison would dif fer significantly from those of the minimum safeguards case comparison.
Three sources of effluent are considered 'in the M/E curves: the break flow, the bypassed accumulator fluid and the overflow of downcomer fluid that occurs during reflood. Note that overflow results when the downcomer is filled to the bottom of the cold leg and is
! transferred to the sump. Spillage, the flow of pumped injection and accumulator fluid delivered to the broken loop and assumed to spill directly to the containment, is not it.aluded in either the FSAR or the Mark-BW curves. l
- e. .
E [,R L ,; / ?- 3L ] . J i ,
'The rate of spillage is dependent on the containment' pressure and would be'.similar for the FSAR calculations-and the Mark-BW reload analysis.
The Mark-BW Reload analysis and the FSAR mass and energy _ releases follow the same general profile .with the release computed for the Mark-BW Reload being.somewhatthigher. The ' containment pressure that would result from the Reload M/E release would exceed ~ the FSAR' curve. The .
" -higher pressure would (1) reduce resistance to the relief of; steam during reflood via a reduction in steam specific
, volume; (2) result in a better reflood rate, ~ and (3) effect a lower peak cladding temperature. Therefore, it is appropriate to use the FSAR containment pressure curves as boundary conditions for the reload analyses.
- b. Justify how the use of the Catawba FSAR containment pressure adequately accounts for changes to containment-heat transfer surfaces- and structures for both the Catawba and McGuire containments since'the Catawba FSAR containment pressure calculation was completed.
Response: The containment heat transfer surfaces and structures for both the McGuire and Catawba containments were recalculated in 1987. The results of- that recalculation were used in the most recent minimum containment pressure analyses given in Section 6.2.1.5 of each plant's FSAR. .As can be seen.by comparing McGuire FSAR Table 6.2.1-45 (1987' Update) and Catawba FSAR Table 6.2.1-59A (1988 Update), the heat transfer surfaces and structures for the two stations are very similar, to the point that the heat structure modeling for minimum containment pressure analysis is identical. The heat transfer surface and structure calculation are periodically reviewed by Duke Power Company to verify that they remain conservative with respect to subsequent plant modifications.
.c
t ,
'Q ! r y
L f, ' w... ,
, . j
-c. Compare the Catawba:. and. McGuire FSAR containment: ;
pressures to verify the. lower pressure was used for the - combined Catawba /McGuire analyses r.ubmitted in BAW-10174. ;
' Response: Figures 15. 6. 5-41' and '15. 6. 5-4 2 of the Catawba- ,
5 FSAR illustrate the minimum containment pressure i responses for minimum and maximum safety injection cases, respectively. FiguresR 15. 6. 4-37.' and 15.6.4-39 of.the. i McGuire FSAR illustrate-the corresponding-responses for- .! E .McGuire (these figures are attached for convenience) . By. - inspection, the figures show that the minimum' containment
. pressures occurring during - ' reflood are . the pressures specific to . Catawba. Using the lower pressure as n' !
L boundary condition causes-lower reflood. rates and higher .
~ peak cladding temperatures. ;
1 i." d. Clarify if the containment pressures in Figures 7-44 and. 7-47 were ~ taken ' from the Catawba FSAR as discussed.on ' Page 4-9. . If not,- dis .ss how .the pressures were calculated. l
. 'l
; Response: Figures ~7-44 and 7-47 are from the minimum
- containment pressure curves of the Catawba FSAR.
Comparison to the Duke applications report plots to
- Figures 15.6.5-41 and 15.6.5-42 of the- Catawba FSAR i L
(included as an attachment to the response t3 question-8.c.)- substantiates this. The FSAR information was-L linearly extrapolated from the endpoint at approximately j L 234' seconds out to 600 seconds to provide needed boundary ; L condition data for the reflood analysis in this' period.
- 4 l.-
t I
.._ 2- - -
.n _
p; .:gr;$ _. -9
- .a-FIGURE 8-1 INTEGRATED MASS RELEASE COMPARISON -
m ..
~
. ; ...... ~~~
810 - ...... .. 500 - .'.* m 2 fB 7 400 -- J v "g (O to ( 300 - 2 CATAWBA FSAR 30 . ................ MARK BW 100 - 0- i i
+ '
N~ 40 80 1N- N O .. TIME (SEC) ; MCOMP12 -
,: g 7.. _
.- . -> 2.-e
, g : .,
s
*; 4
.c - 3 :.,
- FIGURE 8 INTEGRATED ENERGY RELEASE COMPARISON mo aso . .......
3o0 - t
^
a 250 - e-. m
> i a
E a 5 200 - m - Z CATAWBA FSAR m iso - MARK BW - 100 - So - f , . . , . , o , , , , , , 2e i 12e im. 200 I o e 80
. TIME (SEC)
ECOMP12
T I .
,,(
I i I i i i
! i I
l . r R i . 1 1
)
/
i IwM w l I j
~, !
l I s Y I > l 1 N i (
% m I 75 l
," ii ~%. .
. .ime==28
=,,a...---.wom=== a........-.
----------------------------------- l
-- . ...---.mmm -=------------------------------
J f i i ,. l 2 . , . a e *a - I amwmmmuw -4m l l t I COMPARTMENT PRESSURE l MINIMUM SAFETY INJECTION (onuni CATAWBA NUCLE AR STATION j FIGURE 15.6.5-41 1987 UPDATE
. ._.. ., . , _ . - - . ~ _ _ . _ . . . . . . .._.--...._ _ .._.. ._.~... _ _ _ . _ , _ _ - . _ _ _ _ _ _ , _ . _ _ _ . _ . _ _ _ . . _ . _ _ _ _ , - _ _
t l e., c< -- l fl, . A r . e- ,. j
, ;e : ;* j i
. -j 1
I t- , l I l l .- t a ! ' I ! [] ;g
! n 1 I r
A 4 E l _. K J R l , 1 - J I r g I
! J l /
I < a E
}
I r 1 i I J 4 l 1 I i !
/
E I [
' ' ]i
' I I I ] ! I t i l
1 I I I I I 1 1 I i1 ! i f ; i 1 I I i i I I I 1 I I ! i I I i i i i / ! d6 t I l !
! I i I I 1 ! / 3 1I i ! ! ! ! !! ! 1 I I f ! i 1
! i l J !
! ' i
! I i i ! I i ! ! ( l I .
I i I ! I I I t i ! i i ! ! i i ! I ! l1 / I i am
! ! i i ! 1 !
! ! 'i I J [ J . b) ')
i 1 2 I I ! l ! i i l 1 i IiT If I I E , I I 1I I i i ! I![ I [ ] l 1I I I ! I A I ' ! I I r r 1 ' i r ] I r i i l 1 I i i I !1 11 1 i ./ i y~ '
! ! I I ! 1 1! I I I ' ' ' : / i I l i i 2 g
'. I ! ! I i I I I J! ! T I I # =
, . i i i > I i l l ' "
i >
; [f I ' ! l
]
! ' i 1 I I I 2
! ! I ! I
! i i I I ! ! I i ! i m..
! ' ' I I i I I i ! I NI
] ! [ l .
i 1 1 I I 1 I i ! I 1I 1 I i :7m 1 i ! I I i i !
' T
! 1 'm
' i ' ' =- >
i a i i i I I i 1 i i i ! ! ! 1 I I 4 H-w, , , w,," M M """""" "" , 4 Mg*@
=- h*-
=
r
+w,. -{-, W',i: j 7p d
-------m-m----=
*,MMME mmyyym m -m===m mmmmmmmme
,,s pg g un gmM M -- m
/ ! ! ! I i 1 I I i ! i I i i I t t
, 1 ! '~' i i1 1 I I i i i 1 . 1 1 t 1 1 ! i I 1 I il i i i i ! i I i i1 I I I i I l l l I i I
- S: $ $ $ $ $ $
e a d S m .d -
. e e 6
&WWWW3WW =&WmW=
COMPARTMENT PRESSURE MAXIMUM SAFETY INJECTION 4 O CATAWBA FIGURE 15.6.5-42 1987-UPDATE NUCLEAR STA '
;a7 '-4 3 .
a .. o
' u O
. m
.,- - . o;
. . w.
n n > 1
. . 1 e
, o-
.- N 1 .,
, .. o
, . > e I
- m I I .
H-o m. 1 :: N
=
. . . o
. -1 o
. , m f 8 M L O' A
[ EI
^
I I eI **'8 i
,o. -
i to
."- fJ a.- o y
R g.
~
- o
, N
. o i
s: o { h -
.N.
>* MI g v'
_ , ebe r i N I ' c. me l
~
- 9. o 4
m I E 1 _ m L .:
~~
. o
# * * % & m (OTat0 a m usaag COMPARTMENT PRESSURE, MIN SI O McGulRE NUCLEAR STATION Figure 15.6.4-37
; 1987 Update
- n
. , . . + - - - . - . - - n -r-~~ ~ ~- * - ~ .
WL* y- '
-l- $
U' .: as,e y- .
..e e, j sT. j .
s 1i I
- f. j ;- , k
,1
-l i i I
.J; ;
,C , s j
s.>< ,
~
k t F-
.,..........,..no..,....e.-.
i...
...e...e.a.. .. .. . 4. 4
,. e 6 .- :---
o m
........e.,..6.......,.
e..... n
. . ....6.......-
,=-
. .. .. . .. .b.6.....- .
*. - O
..e . ...6........ . .. : . . .-
. o
...............e.,...-
e . . . - m-6 e e +. : - e ... , 4 . b t....., . , . . . - .. O 00 '
. . . . . .. . . ,e,. .,.:
.- N
. . .....e.......- e..... .
.....4....,...
. * . ...e.......-
, O
. . . ......4... ..4.- ,-
- e. - N-
......6..e
. . . 4...
m
. . . . . .e. . . . . . . . .
i.e.....,..g.,..,.-
, .4
, o,
. . . .e.....
. , ..+....o. . . . ., .6-
.. e ...
g
....,.....,.6..:
. .. . .. . . . e . e . -
g" ;o
..e.4e..........
. . N
..e 7- .. N
.. ....,,.4
- e. . . e .e,.e.4.
.e
. .. 8 -
. .... e . .. "'"'
. . . . . . . e.,...... 'I C
.e,..... ...,.. , .. . .. " .4 C
I
........6...
. . . . , . 'i
_ . - 8
- e. . . . e ..... .. - 6. .. .-
" . . . . < N
- 6. . . . w . l' t***................,. _
........I.......+-
**** ~ *
.m o -I t
. ....e.-..".."......"."..
.. ..e..
. e , = ===f ~~.
cc ,!
.I ..
. . ... )
. . . . i,...e.........,,....
f- . o 'E"
. e . a.. . .. . . . .6.. . . .,1..... , ..-
.c
.9 .
. ..e.. . . , .. .-. . .4..e...,.,....
. . e-
. . . . . ..e.........
. . . . .... ... .s . . . .
. ..- .w. f .~
o r
.. .[
. . . . . .. a ""* 'T
.e
... . . . [ .
. . .... . . . . .......6......-
. .. .e C
. . , -- N
. . . .....,.....e....
. .... . . . . . , .. . . .4 4 .s
...e..., 4. ,4i . . . .:.... ..
.* **.**... M
. , . 6.e...
,, . . h. o. ,
. . . . . . . . . .. .e. . . . . ..........e,.,....,.p.e. . g.
. .. . ~
..L. . , .p
. . . ...n........
. . . . . . . . . . , . ... 4.. o....
. .+.L, . . . . . .
.: . p.., e ... .. . . .
==
. _ . - . _ e
... e.
.... : .~ .
co
. ...ea...e........h,.., e.e.. . , . . 46..
. ....e .. .. ., e. e.... ....... 4 .+.44
. + :
.. - m........ . . ..
. 4. 4
.. .e. 96.
e
. .6
. . b .>.s.o4.
.4 ..9. . . =.. ...-.-.-.----
.. O
. ....,.,..e....
.e . .e .e...r.,........,.*I.......d...
.. .... 6. .... . . * . .o a t...: :
. . e.. C. e. e. 3 6.e .4
. , , 6 4
t e .e e- , F . 9.I .tt. ,, c ; _,9. . F 6 9 .e. e ,. . +. 4.. . .t. ..
. . . . ~
o
. . . .. .. .~. .. .. .. .. ...... .. .. ...... -. .. .. .- .-
.. e.. .!. . .. . .. .4. ..........: !.. . .. ,.. ..
...... . .....e ...... ..
- _ _ %. . N m N 4 w + m u o
s (670 musud COMPARTMENT PRESSURE, MeGUIRE NUCLEAR STATION 9MAXSt Figure 15.6.4 39 1987 Update t
V ,
~
_ 5 4. s q.7 ' q q c *-- 12 . - Section 4'.7 compared only the guillotine and. split breaks with
~
a discharge coefficient.of 1.0.. However, Appendix K requires
- the discharge coefficient be varied from1 1.0 to 0.6 for both 1
types of breaks'. Therefore, provide - the results of split. break analyses with discharge coefficients of 0.8 andio.6 to verify the worst- case. break type / discharge coefficient combination for Catawba and McGuire was identified.- Response:- It is an established NRC practice to accept mini - spectrum _ analyses for licensing applications -when a full ~ spectrum has previously been performed. All PWR vendors have, as appropriate,- limited their IACA analyses forEevaluation model upgrades, problem evaluations, and reloads to the mini ^ - spectrum approach.. This includes applications for core power
- w. - upgrade, ECCS system change, change of steam generator tube:
. plugging limits,. alteration of reactor vessel internals.
configuration, and. fuel assembly design change. A'recent example of the application 7f a mini-spectrum approach is the justification for the removal of the' Upper Head Injection ECCS system at McGuire.. This change involved the application of a substantially different evaluation model to' the McGuire plant and-was performed with a three-break, mini-spectrum that did not need to reconfirm break type, break location, RC pump status, or fuel.burnup.- The approach to licensing . of the McGuire/ Catawba units for' Mark-BW reloads selected by BWFC comprises somewhat more than a mini-spectrum. In addition to the mini-spectrum, break type, break location, and worst case fuel burnup have been . reestablished. Additionally, the-
> , - response to ' Question 16 of this set confirms the RC pump status. Because a full spectrum analysis has been performed for McGuire/ Catawba, the mini-spectrum approach will adhere to
\.. . <
NRC practice for demonstrating compliance with 10'CFR 50.46 since there is no reason to believe that an alternate break type will not trend with the mini-spectrum. The identification of the double-ended guillotine as the worst break type is not unique to the BWFC evaluations for McGuire/ Catawba model. Calculations performed in support of l f 0 i
l /.L J
. i
. il: s the- BWFC -RSG LOCA evaluation model' - on the standard
,Westinghouse'3411 Mwt design also identified the guillotine asi i
- a. worst-break type. Both B&W, for B&W-designed plants, and Westinghouse, for the 4-loop plants, have shown that the split is not the limiting break type through _ full spectrum evaluations,-references 12.1 Table 6-1, 12.2 Table-6-1,.and ;
- 12. 3 Table 2. 3. Thus, a consistent trend showing that the ,
guillotine is the- most limiting break- type has been established. The reason that the guillotine is consistently the worst break type is related to the break modelling dif ferences between the split and the guillotine. For the guillotine, fluid from one ll side of the break discharges _ freely through one-half of the total break area without interacting with fluid from the other side of the break. Split modelling mixes the fluids coming to the break fromxboth sides and then discharges through the total break area.. Generally, and specifically in the McGuire/ Catawba analysis, there are no substantial differences in the' blowdown results between-the two break simulations. During- core reflooding, however, the guillotine break j relegates one half of the break area to the discharge of flow l from the reactor vessel, whereas for afsplit simulation, the , t effective break area for reactor vessel #1ow is less than half L the total break area. This allows more steam to vent from the downcomer for the guillotine t break than for the split, :! reducing the downcomer pressure and thereby lowering the core l reflood rate. The difference is small, resulting in only a 60
- F change in cladding temperature, but observable (see Figures i
7-4 and 7-38 of BAW-10174 - for the results of double area l breaks with discharge coefficients of 1.0). The same effect , will occur at smaller break areas or reduced discharge ! coefficients. i The acceptability, then, of the mini-spectrum approach to ! licensing analyses is well established. Consistent trends, l: t
i e , l ,. ? .g
- s
[ 'L t
,w g ,
.i L. observed < in diverse - calculational models,: ' show - that- the - .i
. guillotine is the ' more severe break - type.. This trend is ')
I h evident.in the' calculational results for.McGuire/ Catawba and
> - a' causative ' relationship between the etfact and the break' h',, ,
. modelling 'is p';esent. Therefore, the peak cladding- ;
i temperatures ~ for split: type' breaks of areas less than the full two-areaLbreaks or of= reduced discharge coefficients willibel 1 L ' less.than the-temperatures of the. corresponding guillotine ; y -breaks. For the i guillotine' break type, .the peak cladding - temperature decreases with decreasing break area or discharge' j coefficient. Therefore, a spectrum of split breaksLwill-not! q
" produce a - cladding temperature higher than the worst - case l
. guillotine identified. A mini-spectrum of' splits need not be l
~
specifically - evaluated to assure that the LOCA limits Lare - based onE the : worst case break. Thus, the . mini-spectrum approach as conducted for McGuire and Catawba is valid and , x appropriate for plant licensing. l
^1
,i i
1
.h C 1 .m l i 1 .f2 ,
. y l l
T,%9 : D';df/A.,. 3 4
+4
! l i References 12.1' ECCS - Evaluation 'of- B&W's 205-FA NSS - Rev. 2, BAW-10102',LBabcock & Wilcox,~ December 1975.
;g 12. 2 ECCS Analysis ~ of B&W's 177-FA Lowered-Loon NSS - ' R e v . 3',,
BAW-10103, Babcock:&=.Wilcox, July 1977. 12.3 Westinahouse ECCS: Four Loon Plant (17 x 17) Sensitivity
. Studies, WCAP-8566, Westinghouse- Electric ' Corp. , . April .
1976. s,
- s A , '
1 . 7 , e , 14.- The B&W LOCA EM provides models to enhance-cooling due to grid ieffects and rupture. This implies the PCT calculated depends
. on the positioning of the axial peak'and the axial power-shape ~
- relative;to the grids and the location of rupture.- If the axial peak and power shape analyzed result in~the PCT beiag calcuhted in a node eith grid and/or rupture cooling effects, the Pr?r calculated would be lower than if, the positioning of the axial peak' and the power shape were such that the PCT occurred in' a node without grid and/or rupture cooling.
effects. Response' Nodalization of the BEACH hot channel model, as described in Appendix C of the BFACH topical report, BAW-10166, is based' on the axial distribution ' of the
; important cladding cooling _ mechanisms across ~ the grid span. j I
A pattern of three nodes per grid span is repeated up the fuel pin. Nearly all of the direct grid cooling effects .! I are accounted ~'for in the lowest or first node cf the s pattern. Physically this node models the fuel from the bottom of the grid to about 5 inches above' the . top of the ; grid and is referred to as the grid node. Droplet grid ) interaction and the majority of the grid-induced ) convective enhancement effects are modelled within this ] node. The second node of the three node pattern models f a small amount of residual convective enhancement but i primarily represents a region of the cladding.near the center of the grid span that is essentially free of grid 4 effects. The final node of the pattern models:the fuel-from about 7 inches below to the bottom edge of the upper bounding grid. There are no direct grid ef fects on heat' ! transfer within this node. Heat removal is enhanced j indirectly as a result of the grid . form loss pressure drop coefficient, applied at the junction between this-and the adjacent grid control volume. The pressure drop causes the node to retain entrained liquid causing an increase in interphase heat transfer and effectively reducing the vapor temperature. The effect may or may fu
A
- W.M@
y 9 not be substantial because it depends on the availability of entr..ined droplets in the flow field. During the early portion of reflooding or for upper core elevations, this cooling mechanism is minimized. For the McGuire/ Catawba cal ulations presented in BAW-10174, none of the peak cladding temperatures occur in nodes directly effected by grid-induced cooling mechanisms. The hotteat temperatures early in the reflood period may be in either the second or the third node of a span. One of these nodes will rupture and cool while the_other continues to increase in temperature and may become the location of peak cladding temperature. It is not unusual for the location of peak power to obtain the highest temperatures early in reflooding, become the ruptured location, and not be tha location of the peak cladding temperature.
- a. Verify such conditions are net possible within the ..
technical specification limi',s for the Catawba and McGuire plants, are bounded by the analyses already presented, or analyze the LBLOCA with this/these power shapes and peaking factors. It should be noted that, of N the analyses presented in Section 8, only the 8.0 ft l axial peak case did not have the PCT occur in a node ) influenced by grid or rupture cooling effects. , Response: As discussed in the response to Question 15, the technical specifications do not limit or control the location of peak power. The peak power will, however, be located near the center of a grid span because of grid ef fects on the neutron flux. Tha thermal neutron flux is depressed in the vicinity of the grids because of i increased neutron absorption and a lowering of the local moderator content (the grid displaces water) . For the grid span within which the peak power is located, the power will be lower in the grid node and pre-grid node: bottom and top nodes, than it is in the center node. The
- , 3
V.,
,O' ,
v e nominal peaking in percent of total peak from the lower o node of the span to the top is 0.97, 1.00, and 0.98. 3 (These relative values apply to the peak power grid span. For grid spans that are off-peak the overall power distribution een overshadow the grid flux depressions.) The reference power distributions used for the LOCA limits calculation in the RSG LOCA evaluation model are simple, relatively smooth, functions intended to represent a generalized distribution and not refined-to account for effects of the grids on local power. Their curvature near the location of peak power, however, is quite reasonable when compared to reality. Table 14.1 presents the nodal power distribution within the peak power grid span for the 4.6-ft, 6.3-ft and 8.0-ft 14cA limits cases in comparison to the realistic values. The power shape used in the LOCA calculations is
'presentative of the real shape, or may be slightly
- otter than can actually be expected. For the highly skewed shapes, 2.9-ft and 9.7-ft peaks, the severity of the skew of the overall shape combines with the grid neutronics effects to cause a larger depression for the node nearest the end of the core. A reference realistic distribution for these highly skewed cases hhs not been developed but it can be expected to verify the shapes used in the.LOCA analyses.
As shown, the power distributions used for the LOCA calculations are not significant*.y dif ferent from the profiles that result naturally fron grid flux depression. Therefore, ' although the power profile within a grid span is not dictated by technical specifications,. the. distribution used in the LOCA calculation for the high power grid span is assured. Further flattening of the distribution near the peak power location would be a small, but unwarranted, increase in conservatism.
Table 14.1 Flux Patios Node Typich 4.6' LLIM 6.3' LLIM 8.0' LLIM Grid 0.97 0.98 0.99 1.00 Middle 1.00 1.00 1.00 1.00 Upper 0.98 0.99 0.99 0.97
~
l
- b. In Section 8.2, *:he highest PCT was calculated.for the case with the asial peak at 4.6 ft. The- PCT was j calculatud to occur in node 11, a node with some grid cooling effects from the grid in node 10. Could a higher PCT'be calculated if the axial power shape was adjuste, slightly (for example, move the axial peak from node 8 to .;
node 9) so that the PCT was calculated to occur in nodo i 12, a node without grid cooling effects? Other cases where this combination of PCT and grid coaling effect occurred were the 2.9 and 9.7 ft axial per casos. Response: The central distribution of power within the peak power grid span is discussed in the response co part (a) of this question. A shif t of power peak upward from 4.6 ft, within the framework of the EM nodalization l scheme, would result in shiftimi the peak power to ) another grid span, a case that is already analyzed. The 1 power in Nodes 11 and 12, however, could be increased by ! an alteration of the power shape from the reference used in the LOCA calculations without repositioning the peak ! powor. 1 The response to Question 15 discusses alternate power shapes and the need to evaluate them with the conclusion that the present set of power distributions used in the BWFC RSG LOCA evaluation model provide a realistic bound to the shapes that can occur at the limits of plant 1 (m
~:...---------..__.........--....
c.., i it' . O .
. operation. In reaching that conclusion, the sensitivity of peak cladding temperature to an unrealistically flat axial power profile was determined to be less than 60 F.
Applying the same technique to the evaluation of a power shape that increases the power in Nodes 11 and 12 by an amount suggested in Part (b) of this question would gives an increase in peak cladding temperature of about 30 F. 4'; In summary, then, moving tne peak power location f rom the middle node of a grid span is not supportable physically;-
- 1 Adjustment of the axial power distribution is not
- !' warranted because the sensitivity of cladding temperature to the shift is small and the distributions presently
- employed form a realistic bounding set. The current approach set out in the EM is, therefore, sufficient fo't the evaluation of peak cladding temperature and the setting of local power operation limits in accordance with 10CFR50.46.
p- s . p , 4
- 15. Due to the slope of the axial shape on either side of the peak power location, the power at the PCT location was less than at the peak power location (see Figure 8-3). Clarify if a worst case axial shape could be defined within the Catawba and L McGuire technical specification limits so that the axial shape- ,
was, flatter in the vicinity of the axial peak and the power was higher at the PCT location. For. example, in the 8.0 ft axial peak case, the axial peak was in node 14 and the PCT occurred in node 12, at 6.9 ft. Would a higher PCT have been calculated if the axial shape was flatter around the peak power location so that additional power was applied at the 6.91 ft elevation? A similar question applies to all other cases. Response The limit on power shapes imposed within plant j , technical specifications because of LOCA is restricted to ( specification of the peak power as a function of elevation in the core. No provision is made to specify a shape or a distribution of the total peak between the axial and radial factors. To determine power shapes with which the allowable l peak power versus . core position val'ues will be set, l conservative, pragmatic, and phenomenological factors are balanced. In so doing, the RSG LOCA evaluation model requires that five separate power shapes, each peaked at a different elevation in the core, be evaluated. The basic philosophy for determining the power shapes is to use realistic power shapes with normal but high peaking and-1 increase both the axial and radial peaking in a reasonable fashion until the absolute or the desired power limit is reached. Although operation is allowed up to and beyond the ! limits (the limiting is by administrative control such that being above a limit only requires a plant to take control measures to return below the limit), it is seldom if ever . achieved, and there is no most probable way by which peaking would increase to the limits. Therefore, by starting with real shapes and pushing them to the limit in reasonable ways, the power shapes that result can be described as possible shapes if operation at the limiting local power levels were to occur. P P
, - , - - - ~ . , - . - - , -,_n-- - , <- ,-, _ -_ ~-,-,,m-- e
f
* ~
- I
. i While the power shapes used are representativa of shapes ;
possible for operation at the limits of local power, they are bounding for the operations of the plant because they are at the limits of allowable operation. This, in combination with . ! the use of five separate axial distributions, provides broad l coverage of the possible power shapes that can be achieved and , makes the LOCA limits power distributions, as a set, essentially bounding for operation of the plant. !' Because the IDCA limits shapes are derived from abstractions 1 of actual shapes, they are curved over the entire length of the core. Within the vicinity of the peak, - however, the curvature is not so severe that complete flattening would q seriously alter the results achieved. To provide a measure of the effect of total local flattening of the power shapes a ! study was made that, using the LOCA limits cases as a base, , 1 adjusted the cladding-to-vapor temperature dif ference at each l I axial position by the ratio of the peak local power to the local power at the axial position. The results are considered credible for the grid spans above and below the span that contains the peak power in the base. Within this range the peak cladding temperatures were increased by 20 to 60 F. The highest cladding temperature within the LOCA limits set would increase by 57 F. Therefore, further flattening of the power shapes will not substantially alter the results obtained by current practice. l l In summary, a set of five power shapes are utilized for the LOCA limits studies under the BWFC LOCA evaluation model. , These shapes provide, as a set, a realistic bound to the possible power chapes that can occur at the limits of plant operation. Moreover, the effects of an increase in the flatness of the power shapes has been shown to be limited to a 60 F possible increase in peak cladding temperature. ; Therefore, the methodology of the BWFC RSG LOCA evaluation 1
I i l
-model offers a practical yet realistically bounding approach l
'+ to the determination of LOCA-induced operational limits. ! i l 1 1 k l l l l i 1
?
. z__ ___ __ _ _ _ _ . . . _ . . . _
- - ,e
j 1
- 18. Although the cladding '.emperature figures in Chapter 8 show the calculation of the PCT and the turnover of the cladding temperature excursion, the figures do not show a clear coeling trend after the PCT is calculated and they do not show when the rods quench. How do you ensure that reheat will not occur and therefore the peak cladding temperature has been ,
l - determined to stop further calculation? l 1 L t Response Reflood cladding temperatures are calculated by 1 BEACH, which contains suitable conservatism for use in I 1 l_ Appendix K calculations up to and after the time of peak cladding temperature. Following the peak, during cladding i cooldown and quench, however, BEACH becomes substantially over conservative. Although BEACH would predict a second or third j peak 'in cladding temperature, should cooling conditions degrade sufficiently, it is unable to obtain the heat trannfer .l improvements ' necessary to predict realistically more rapid cooldown and quench. To assure that the actual peak in ; cladding temperature.has been determined, the BEACH simulation l 1s extended past the time at which the core is quenched as measured by the REFLOD3B code. Once REFLOD3B shows quenching j the transient is considered over. A further-discussion of I this application of REFLOD3B is presented in the response to Question 2 of the second set'of questions on the RSG IDCA l evaluation model, BAW-10168. Although BEACH under predicts l- the improvements in cooling following the occurrence of peak l cladding temperatures, it does show a continuous increase in l cooling when conditions warrant. Therefore, BEACH can be applied to assure that cooling conditions between the time of peak temperature and core quenching continue to improve and no j secondary temperatura excursion occurs. l Benchmarks of BEACH against experimental data are described in l. Appendices C, D, and E of BAW-10166P. The benchmarks conclude E that, during reflood, BEACH overpredicts post-peak cladding temperatures by a significant margin. Delayed quenching and cladding temperature overprediction, in the post peak period, are prevalent in the FLECHT, FLECHT-SEASET, CCTF, and the L__-_________-_____-___-____--___-_____
L.
, . 1
., / 'l I
REBEKA-6 benchmarks. In many of the benchmark comparisons, neither quenching nor a substantial cooling trend is evident ! in the predictions for the duration of the analysis. These benchmarks provide a clear measure of the degree of overconservatism embodied in the present BEACH formulation. j The benchmarks also provide evidence that, provided conditions i warrant, clad reheat- will not occur after the reflood i temperature has reached its peak. While the BEACH results. j j indicate slow post-peak cooling of the clad, experiments show l that the clad both cools and is quenched following the peak i 1 temperature. Benchmarks show that BEACH will predict the peak clad temperature and will conservatively overpredict temperatures l thereafter. Even though BEACH is unable, for the most part, 1 to predict post peak cooldown and quench, it will predict the possibility of a secondary cladding temperature excursion ! o should cooling conditions degrade sufficiently. Furthermore, ,, experimental observation support the essentially monotonic I cooldown and qu'ench of the cladding following the occurrence
. of peak temperature. Therefore, using BEACH for the ,
determination of peak cladding temperature and to monitor l cooling conditions during the post peak period, while employing REFLOD3B to determine the core quench time, provides l l assurance that the transient has been tracked to completion and the peak in cladding temperature determined conservatively. l l l l m L-__--____._._._______-______________________. - . , _ _ . _ . . .
, ,4
- 24. Appendix A of BAW-10174 only discussed the effects of mixed core operation on a LBICCA. Provide similar information to justify operation of Catawba and McGuire with mixed cores as it relates to a $BLOCA. ;
Responset Chapter 12 of BAW-10174 addrtsses the differences between the Mark-BW and the OFA fuel assemblies ir regards to small break IACA calculations. Question 23, of this set, considers a design difference between the fuel assemblies that I was not included in Chapter 12. Both' discussions conclude j that the change from the OFA assembly to the Mark-BW wil3 not alter the SBLOCA calculational results and that the OFA based FSAR analyses may be applied for licensing of the Mark-BW I fuel. As will be presented here, that conclusion also applies to the mixed core conditions during the transition from the OFA assembly to Mark-BW. No interaction betueen the OFA and the Mark-BW designs has been identified that will alter- or l compromise the SBLOCA calculational results. Chapter 12 identified the following differences between the OFA and the Mark-BW fuel assembly designs as possibly l affecting SBLOCA calculations: unrecoverable pressure drops across the assemblies, initial fuel temperatures, initial pin internal gas tressure, and the axial power profile. Appendix ] A of BAW-101*i4 listed geometrical differences between the _ fuel assembly designs along with the unrecoverable pressure drops. Table 24-1 has been copied from Appendix A to show the , differences. Each of the differences between the OFA and the Mark-BW designs has been examined for possible adverse effects during mixed core operation, and none have been identified. Unrecoverable Pressure Droo o As discussed in the response to Question 22 of this set, cladding temperature excursions during small break LOCA calculations occur during the core boiloff period and are governed by hydrostatic considerations. The only resistance-
~
- l. .. ;
- l. l
. l g
related pressure drops of significance occur within the steam regions of the RCS to vent steam frem the top of the core to ; the break. With these, system flow is sufficiently low that I even large loop resistance variations are inconsequential. f 1 The pressure drop difference between the OFA and the Mark-BW occurs at the fuel assembly inlet with operational flows. l During the core boiloff period this loc.ation will be under water with flow rates so small that f.ow related pressure changes are inconsequential. The response to Question 22 , allowed that some benefit " ... mest . lihely too small to notice ... " might occur because of the decreased resistance { of the Mark-BW. However, because of the hydrostatic balancing , (cross the core, the benefit would apply to all fuel-assemblics, not just to the Mark-BW, Therefore, the decreased- , I inlet pressure drop of the Mark-BW fuel assembly does not interact with the OFA assembly to create any adverse consequence for SBLOCA during mixed core operation. Initial Fuel Temoeratures l-The itaue of initial fuel stored energy was addressed in ; Chapter 12 of BAW-10174 and has been further discussed in the i response to Question 20 of this set. The Mark-BW fuel pellet can be expected to contain approximately 7 percent more energy at operation than the OFA. .This extra energy is easily removed during pump coast down and removed via the steam < generators (see Question 20 response) and will not affect the t response of either fuel assembly during that period. During. l
- the vessel boiloff phase the Mark-BW assely will heat up slightly slower than will the OFA and :oci down slightly slower. The effect would be that of a buffer or a filter on the transient occurrences and, although most likely unnoticeable in a calculation, would tend to reduce the amount of core uncovering. Thus, if a coupling can be considered to exist between the OFA and the Mark-BW because of this, the effect would be to reduce the computed cladding temperature
+ r - -- re-w-. - - - -- . _ _ _ _ _e
i for the OFA assembly. Since the OFA temperature is the temperature of record for the Mark-BW, there is no adverse consequence on either assembly for mixed core operation. Initial Pin Internal Gas Pressure As discussed in Chapter 12 of BAW-10174, the fuel pin internal gas fill pressures are similar between the two fuel designs. The gas pressure differences that do exist could affect cladding rupture time slightly, but the impact of a rupture or rupture timing dif ference would be negligible. The occurrence or lack of rupture does not affect vessel inventories and cannot cause any interaction between the two fuel designs. Therefore, although small dif ferences could exist between 'the OFA and the Mark-BW, these differences will not produce an adverge SBLOCA result for mixed core operation. Axial Power Profile i As discussed in the response to Question 21 of this set, there is no difference between the power shape used for evaluation ,, of the SBLOCA between the OFA and the Hark-BW. Therefore, the potential difference in unembly power shapes for operation not approaching the a', vable limits will not affect the results of SBLOCA calculations nor cause a coupling between i the two fuel designs which would adversely affect mixed core J operation, l Geometrical Differences 1 The geometrical differences between the fuel assemblies and their impact on SBLOCA calculations are assessed in the response to' Question 23 of this set. The conclusion is that these differences will cause a negligible though positive impact by slightly increasing the minimum core mixture height. The trend in operation with both fuel designs will be I - - . . _ _ - ___ l
n_ -. _ _. . _ - _ - . _ .
. *i .'.
l
]
improvement of results.in direct proportion to the amount of Mark-BW fuel in the mixed core. Therefore, althottgh there would be so:se definite- interaction between the designs, - the
., effect will be relatively minor and somewhat beneficial to the j OFA results. Since the OFA' calculations are the calculations I of record for the Mark-BW, there is no adverse consequence on either assembly for mixed core operation.
Conclusion 1 The differences between the OFA and the Mark-BW assemblies are 1 such that small break LOCA cladding temperatures for mixed 1 core operation .will not vary substantially from those i calculated for OFA core operation. The quantifiable deviations would indicate a slight lowering of the peak temperatures as the core transitions to full Mark-BW. The calculations show cenLiderable margin to 10 CFR 50.46 9 criteria. Operational limits or technical specifications required by the OFA-based FSAR~ analyses will not be altered for application to the Mark-BW assembly. Therefore, the FSAR evaluations can be applied to the licensing of both the full Mark-BW core and the mixed or transition core. _ _ . . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _- - - -_ ._-- r - , - - r - - - -
Table 24-1 0FA/ Mark-BW Design--Difforences MK-BW OFA Guide Thimbles: Upper Section OD/f (in) 0.482/0.016 0.474/0.016 Lower Section OD/f (in) 0.429/0.016 0.429/0.016 Instrument Tube: OD/t (in) 0.482/0.016 0.474/0.016 Fuel Pin: Pin OD (in) 0.374 0.360 Clad Thickness (in) 0.024 0.0225 Pellet OD (in) 0.3195 0.3088 Pellet Length (in) 0.400 0.507.- Diametral Gap (in) 0.0065 0.0062 Pressure Drop Across Core (psi) 22.7 23.7 (at full flow) t
~. _ _ __ . - - -- -- __ - _ _ . _ ._ _ ~'
l
- 25. !Section A.2 contained an estimate of the differense in PCT expected in a LBLOCA for the blowdown and reflood phases due !
to nixed core operation .on Soth the Mark-BW Lnd OFA evaluations. Provide a quantiu tive justification for the differences in PCT discussed.in ; tat section. J ! Response: The temperature effects described in Section A.2 of I Appendix A of BAW-10174 are based on quantitative evaluations. ; The 30 F to 50 F impact of mixed core operation during i blowdown was obtained from an early REIAP5 evaluation of a L core configuration consisting of one Mark-BW and 192 OFA fuel I assemblies. Although the REIAP5 version, model, and input l were not current, their content was sufficient to extract the sensitivity attributed to the mixed core configurati'1 in j Appendix A. That case showed that little differen'ce exists between an all Mark-BW core and a core with one Mark-BW and 192 OFA assemblies during the positive core flow period at the - I beginning of the transient. Once the RC pumps cavitate, however, the core flow becomes negative and of much higher quality. During this period, the higher pressure drop of the l OFA diverts a small amount of flow to the Mark-BW, somewhat ) l reducing the cladding temperature for the Mark-BW. U 1 I As with blowdown, assessments of the flooding rate effects were based on actual REFLOD3B calculations. Because REFLOD3B is a single channel code, the evaluation compared a full OFA core to a full Mark-BW core and then applied the flooding rate deviation to the cooling of the other assembly. The full OFA , core was shown to experience a slower, by less than 2 percent, i reflooding. In crder to assess the single Mark-BW - 192-OFA assembly core, the effect of this lower OFA flooding rate on 1 the reflood cooling of the Mark-BW assembly was determined. i Because' flooding rate changes of this order of magnitude had been observed in sensitivity studies, no new cladding temperature calculation was necessary. The temperature ; differences reporte.J are those typical of 2 percent flooding wm _. -. _____s
t . e rate changes in the evaluation model and the McGuire/ Catawba sensitivity studies. Appendix A does not portend that the temperatures provided are exact measures of the effects of interactions between the OFA and Mark-BW assemblies during the transition cores. Rather Appendix A intends to describe the trend which would be observed in a fully implemented calculation, explain why this trend occurs, and provide a conservative estimate of any-potential change in peak cladding temperature. The position
-taken in Appendix A ist (1) Some degree of interaction can occur between the OFA and the Mark-BW fuel assemblies for the transition cores. (2) The variation of cladding temperature caused by the intsraction will trend in -opposite and offsetting directions between the blowdown and the reflood periods. (3) The variations will be limited to less than 50 F during either the blowdown or reflood periods with the not difference at peak temperature being :. ear zero. (4) Because of the limited potential interactions and the margins to limiting criteria available within the whole core calculations, no specific evaluation model computation of the mixed core configuration is warranted.
I 4
h.3 1 l
. l l
- 28. On page A.3 the statement is made that the only effect of the ,
mixed core that needed to be considered for the reflood phase i of the LBIOCA is the whole core pressure drop. The following
' questions are related to this statement.
- a. Justify why the pressure drop difference for the two fuel assemblies does not cause a flow diversion effect during !
the reflood phase similar to the blowdown phase. Provide appropriats data or analyses to support your conclusions. j Responset The large break LOCA reflood phase is modelled = with a one-dimensional simulation that conservatively ignores radial effects within the core. Some flow l diversion toward the Mark-BW is likely and could be j represented in mora detailed modelling in a larger ! calculation. Those techniques would also substantially reduce reflood cladding temperatures by incorporating two- or three- dimensional effects that cause preferential flow to the hot assemblies. Because it ; would be inconsistent tr evaluate a negative multi-dimensional effect without consideration of the positivt. effects, only the overall resistance impact of the OFA l ,= versus Mark-BW assemblies was included in the reflooding q comparison.
- l The combined effects of added assembly resistance and
! two- or three- dimensional reflooding can be discussed considering two possible core arrangements: (1) a Mark- ; BW assembly as the hot assembly surrounded by OFA assemblies or (2) an OFA assembly as a hot assembly surrounded by Mark-BW assemblies. In either situation the flow diversion, if any, will be toward the Mark-BW. Therefore, with diversion allowed, the cooling of the Mark-BW assembly under the first arrangement would be j enhanced over present evaluations and that condition need i not be considered further. i l' ? For the second core arrangement, with an OFA assembly as l a hot assembly, any flow diversion toward the Mark-BW t
,i 1
. 1 I
.would ta'u to reduce the flow of water entering the I bottom of the OFA assembly. The pressure drop difference )
between'tha OFA and the Mark-BW is small (less than 5 j percent), and any flow diversion toward the Mark-BW would be limited to the square root of that difference (about l 2 percent). Such a diversion is not large and would be I compensated for by a buildup of elevation head in the Mark-BW assemblies. The pressure drop across the nominal core during reflooding is less than one psi. Converting l 5 percent of the core pressure drop into elevation head shows that the water level in the Mark-BW need be only , 1 0.12 _ f t higher than that in the OFA to eliminate ths diversion of flow. At 2 percent flow diversion, with a carryout rate fraction of 80 percent, and a nominal inlet L flow of 1 inch per second, only 30 seconds are required l to establish a 0.12 f t elevation head dif ference between the assemblies. Therefore, any not ficw diversion will l be short-lived. Actually, the required head difference L is most likely set up in the first 5 or 10 seconds of reflooding when the flooding rates are high and
. oscillatory.
l_ l As demonstrated by the SCTF' and CCTF experiments, the conservatism of not considering multi-dimensional ef fects on core reflooding far out. weighs the effects of any I possible flow diversion caused by the difference in the ) design of the two assemblies. Within the hotter fuel ! casamblies, substantially more water is boiled and l entrained by the reflooding process than within the j average and colder assemblies. As the core fills,
]
hydrostatic head imbalances are set up that lead to greater flow of watSr to the hottest assemblies and i correspondingly less to the coldest assemblies. As a reasonable estimate, a measure of the flow imbalance is I the radial power peak, whereby inlet velocities for the hot assembly 30 to 50 percent higher than those of the 1 I
1 L ,y. .' ( 1 average channel can be expected. The result is a flattening of the cladding. temperatures across the core' l a radially and a reduction of the hot spot cladding i temperatures of 300 to 500 F. The SCTF and CCTF j i experiments have demonstrated the improvements in hot ) channel reflood cooling for radially peaked conditions. I
-(The response to Question 2.a of the second round of l questions on the RSG IACA evaluation model, BAW-10168, q presents a discussion of the effects observed in the l experiments, or reference can be made to the appropriate l
experimental reports, references 28.1 and 28.2.) Thus, l: the consequence of simult. tion of all sources of flow' ! ! diversion is a substantial decrease in cladding temperatures from the present calculational results. , As discussed, there is no flow diversion from one fuel l l assembly to another during reflooding with the -{ calculational . approach taken by the BWFC RSG I4CA evaluation model. The RSG I4CA model, in not considering l the reflooding process at a level of modelling detail sufficient to evaluate reflood flow diversion, incorporates conservatisms of far larger effect than those possible because of. mixed core-induced flow diversjon. Therefore, the evaluation of Appendix A, without the consideration of flow diversion, remains appropriate for the licensing of the transition cores. l
- b. Clarify how the fuel loading pattern during mixed core operation affects the possibility of flow diversion from one type of bundle to the other.
l Response: The actual fuel loading pattern for a reload is determined relatively late in the reload design process and cannot be readily predicted. It is reasonable to assume that the transition from a moctly OFA core to a mostly Mark-BW core will take place over three cycles with one third of the fuel replaced in each
cycle. Even af ter the third cycle, it is likely that the core design will continue to utilize a few 0FA assemblies for several years. Therefore, the IcCA calculations are not performed to an accuracy level .that would require precise knowledge of the fuel loading pattern. As explained in the response to Part (a) of this question, the consideration of flow diversion during reflooding as a mixed core consequence is not appropriate for calculations with current evaluation models. Therefore, the only effect of the loading pattern on the evaluation presented in Appendix A is the degree to which the mixed cores resistance is changed. The considerations of Appendix A bound the extremes of most resistive, all 0FA, to least resistive, all Mark-BW. As the loading pattern and the number of Mark-BW assemblies in the core change, the reflooding effect will transition from a 2 percent decrease in flooding rate to no effect. During the first cycle, for example, the decrease should-only be about 1.3 percent. Repeatiet the conclusions from Part (a), the evaluation of ref W
- ding at a level of accuracy that could properly treat flow diversion would result in cladding temperatures several hundred degrees lower than those predicted by present techniques. At that level of detail, however, care would be required to assure that both the fuel loading pattern and the skew in steady-state core flow assumed did not limit the core design.
At present, the conservative one-dimensional reflood treatment is independent of individual fuel assembly placement and, within the constraints of the results presented in Appendix A, independent of cycle design.
- c. Clarify how the above two items effect cooling of the OFA and Mark-BW assemblies during re*Pvd.
.E ." o' 1
.- i Response As developed in the responses to the other parts of this question, the evaluation of the coolability of the mixed core configuration as presented in Appendix
]
A of BAW-10174 is conservative and appropriate for plant licensing. Realistically, ' flow diversion within the l reflooding core would occur because of power distribution effects as well as for fuel assembly resistance differences. The results of the SCTF and CCTF l experiments (See. the response to Question 2.a of the second set of questions on the RSG I4CA evaluation model, BAW-10168, and references 28.1 and 28.2) show a substantially larger benefit from power distribution- ) induced flow diversion than the expected deficit from the 1 fuel assembly resistance mismatch. Thus, although some 1 o 1 small amount of fuel assembly-induced flow diversion, ! resulting in a small temperature increase, is expected during reflood, that increase would be imposed on a j cladding temperature several hundred degrees cooler than the present !icensing calculations. Therefore, the cooling considerations-for the mixed core configuration as expressed in Appendix A of BAW-10174 remain applicable l and appropriated for licensing. I i i 1 l l i u I l I I i
, -__.._....._s_ , . _ . , . . . . _ _ _ _ _ _ _ _ _ , . _ . _ _ . _ _ _- _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ . _ _ _ . _ _ _ _ _ _ _ _
y . s ;? ',o - . h .g. .g References 28.1 T. Iwamura, M. Osakabe,.and Y. Sudo, " Effects of Radical Core Power Profile on Core Thermo-Hydraulic Behavior' during Reflood Phase in PWR-IDCAs," Journal of Nuclear Science and Technology, 20(9), pp 743 - 751, September 1983. 28.2 H. Akimoto, T. Iguchi, and Y. Murao, " Core Radial Profile Effect on System and Core Cooling Behavior during Refle>od Phase of PWR-IACA with CCTF Data," Journal L of Nuclear Science and Technology, 22(7), pp 538 - 550, July 1985. w}}