ML20198P243
| ML20198P243 | |
| Person / Time | |
|---|---|
| Site: | 05000000, Oyster Creek |
| Issue date: | 03/19/1986 |
| From: | NRC |
| To: | |
| Shared Package | |
| ML20151H203 | List:
|
| References | |
| FOIA-86-26 NUDOCS 8606060269 | |
| Download: ML20198P243 (13) | |
Text
._
/
n SAFETY EVALUATION REPORT l
, f ' gla OYSTER CREEK PLANT EFFECTIVENESS OF EXISTING CORE
~
SPRAY SPARGER IN A' STEAM ENVIRONMENT 1.0 Introduction and Backaround l
In 1976 the staff requested (Ref. 11 that the licensee.iustify spray co ling (heat transfer) coefficients assumed in the Oyster Creek ECCS analysis in light of test data which suggested core spray flow to some bundles in a steam environment may be reduced below that measured in design tests in air.
In their response (Ref. 2), they indicated that core spray flow to the hottest (central) regions of the core would not be significantly.
i affected because the spray nozzles designed to cover this region had shoan little sensitivity to steam environments in single nozzle tests, and that; elthough the nozzles designed to cover the periphery of the core indicated some spray cone contraction in steam, peripheral bundle flow rates were expected (based on GE tests) to remain above the design minimum flow rate.
Based on the above, they concluded that spray heat transfer coefficients used in ECCS analysis (i.e. Appendix K values) remained applic'able, safety 9
analyses for the Oyster Creek facility were unaffected and no further testing was needed. GE supported this evaluation in their. licensing topical report (Ref. 3).
But staff approval of the report was. contingent upon completion of tests and analysis which verified that sufficient flow existed to support Appendix K heat transfer coefficients (Ref. 4).
The staff requirement that tests and analyses be completed was reiterated in 1982 in (Ref. 5). The licensees response was recently submitted in topical report form and is the subject of this report (Reference 14).
2.0 Discussion l
To address the problem of spray cooling effectivenes's the licensee has pursued a program with General Electric Company which includes both tests and analysis. The objectives of this prgoram were to identify the minimum bundle spray flow rate which supports the 10 CFR 50 Appendix X heat i
transfer coefficents, and to demonstrate that flow to all bundles during i
spray cooling would be greater than the minimum required flowrate.
An analytical model, qualified with experimental data was used to determine the required minimum bundle flow rate.
Results obtained with this model show that pred.icted heat transfer is equivalent to the Appendix K assump-e tions at a bundle flowrate of 1.8 GPM.
General Electrics semi-emperical C@
design methodology was used to determine the spray flow distribtuion in a Q
steam environment.
This method combines plant specific single nozzle and gg fall scale sparger test data with analytical results to predict the minimum i
bundle flowrate in the core. The methodology has been successfully h$
benchmarked against 30" sector steam tests of a BWR/6 Core Spray Sparger gg
(. Reference 6).
The minimum bundle flow rate predicted with the methodology car includina uncertainties, is 2.0 GPM.
The calculation of this result SN assumes single sparger (lower) operation in a 30 psia steam envirenment.
While this result is less than the original Oyster Creek desian minimum an o.a.
value of 2.45 GPM, it is above the r.eouired flowra~te of 1.8 GPM.
On this basis the licensee concludes that the assumption of Appendix K heat transfer coefficents in the Oyster Cieek ECCS evaluation model remains valid.
~
i
-7 i
3.0 Evaluat % F The pur' pose of the evaluation is to determine whether the effects of steam environment on core spray distribution degrade the minimum bundle spray flowrate to the extent that the assumption of 10 CFR 50 Appendix K heat transfer coefficients in LOCA analysis becomes invalid. The bases for the evaluation are analyses submitted by the licensee in response to staff, requests for additional information (Reference 5), and review of other tests and analyses germane to the issue.
j The evaluation is divided into two parts. The first section addresses l
the minimum buriale spray flowrate expected in a steam environment. The second section addresses the question of how large a bundle flowrate is needed to achieve App'endix K heat transfer coefficients.
3.1 Minimum Available Bundle Spray Flowrate in Steam u
The General Electric design methodology has been used to detemine the.
spray distribution in a steam environment.
The method treats condensatio'n and drag effects as well as effects of interacting sprays.
The methodology l
was developed under the key assumption that the condensation (thermodynamic) effects in a steam environment can be handled independently from the hydrodynamic effects of multiple nozzle i'nteractions. Under this assumption the thermodynamic effects are evlauated from single nozzle spray distribution tests in a simulated steam environnent. The hydrodynamic effects of multiple nozzle interactions are determined from single nozzle and full scale sparger spray distribution tests in an air environment using simulator nozzles.
Simulator nozzles are specifically developed spray nozzles which produce spray patterns.in an air environment 'similar to corresponding reactor nozzle spray patterns in a steam environment.
Single nozzle tests in steam are perfomed at 30 psia where the steam density is approximately eoual to 4
the density of atmospheric air.
This maintains dynamic similarity between I
performance in air and steam. The methodology has been successfully benchmarked against 30* sector steam tests of a BWR/6 core spray sparger (Ret ence 6); and approved by the staff (RSB) for use in BWR/6 design (Reference 7).
In addition model predictions have been made for the BWR/4&5 -
218 sparger design and compared with full scale 30* sector steam tests of that design.
(Reference 8).
The Comparison shows that as with BWR/6, predictions and test results agree very well within a certain error bond. The BWR/4 comparison is also significant in that it confims the methods ability to treat differences in nozzle and sparger design with variation in input.
In implementing the design methodolopy for the Oyster Creek evaluation the licensee has made one modification.
In the standard desion method the contribution to bundle flow c'ue to spray interaction is detemined based en a full scale sparger flow test in air with " simulators" nozzles that have been designed to reproduce single nozzle flow patterns in I
steam for the same nozzle flowrate.
In the Oyster Creek evaluation existing spray test data for a. reduced nozzle flowrates was utilized to l
l
~..m.,,.
,- -, m m. n.. n... m
.. ~, ~... -..,,,,
-.. - ~
i l
simulate the major effect of a steam environment.
Single nozzle. tests in steant indicate that the primary effect of a steam environinent is a contraction in the spray pattern which for horizontally aimed sprays results in less flow reaching the central core regions and more flow
. reaching intermediate and peripheral bundles. A similar pattern in air, which matches central region flow yet underpredicts intermediate and peripheral region flow, can be achieved by reducing nozzle flowrate.
The licensee has used experimental and calculated distributions taken at a sparger flowrate of 3400 GPM to characterize the distribution for a sparger flowrate in steam of 4690 GPM.
In response to a staff. request for supporting data the licensee has provided data comparing distributions taken at sparger flowrates of 3400 GPM and 3100 GPM to support a claim of conservatism in the approach. The data shows little sensetivity over a range of 300 GPM. The small difference that does exist indicates that the 3100 GPM case (i.e. the lower of the two) yields a less limiting spray interaction effect.
Thus, we believe that the data presented by the licensee is inconclusive with respect to the conservatism of the licensee's deviation from the standard design methodology. To demonstrate conservatism the licensee should present data which shows that multiple interaction effect (MIE) term detemined based on actual simuTator nozzle '
~
_is bounded by MIE term determined with the reduced flowrate Oyster Creek approach.
Core Spray Sparcer Flowrates Sparger and individual nozzle flow rates assumed in the analyses were based on the following:
i
. 1.
Core Spray system perfomance data (pump-head /#1ow characteristics) obtained during pre-operational tests in 1969 and confirmed by surveillance tests in 1978.
l 2.
Spray distribution tests perfonned in 1968 with a full scale mock-up of the Oyster Creek sparger (nozzle flowrates).
3.
Calculated system pressure drop characteristics of the delivery pipino including the ring spargers.
4.
Predicted leakage flow through sparger crack's.
We find the licensees approach in determing system and nozzle flowrates acceptable since it is generally based on test data. The affect of a single through wall crack in the Oyster Creek No. 2 sparger has been.
evaluated previously by the staff (Reference 8).
I,n their evaluation the staff concludes that Crack leakage would have a negligible effect on distribution of flow among nozzles.
Since the discovery of de o.riginal crack in 1978, three additional crack indications have been identified.
In determining crack leakage flow the licensee has treated the'se indications also as through wall cracks. All crack leakaoe flow is treated in the analysis as a ret loss in available core spray #10w.
He believe this to be a conservative approach and ' find it acceptable.
1
. i
~
The tottrl7 ingle core spray sparger flowrate in 30 psia steain, including an allowance for crack leakage, has been determined to be 4690 GPM based on tests conducted at the plant. This flow rate is considerably larger than the original single sparger design flowrate of 3400 GPM. The flow difference appears to be the result of different assumed reactor system pressures (i.e. 30 psia backpressure vs.125 psia for rated flow).
We note that actual sparger flowrate determined at the plant r:ust be shown to bound that assumed in plant safety analyses as per plant Technical Specification 3.4.
Thus, an amendment to the plant Technical Specifications is reouired in order to reflect the increased sparger flow recuiremen,t of 4690 GPM.
Uncertainty Analysis In deriving th'e minimum bundle flow rate the licensee has accounted for the uncertainties in the calculational and experimental steps of the design methodology. The net affect of applying the uncertainties in the analysis is to reduce the predicted core radius average flow rate of 3.5 GPM/ bundle by a factor 1.75 to a core wide minimum bundle l flow of d
2.0 GPM/ bundle. To evaluate the uncertainties we have compared the overall uncertainty factor of 1.75 with the differen~ce between predictions and tests documented for BWR/4, BUR /5 and BWR/6 systems.
(Reference 7 and 9).The comparison shows that the analysis methodology generally l
overpedicts bundle flowrates, and that the. differences between predicted -
and measured bundle flowrate generally falls in the range between 25%
and 50% of measured flow. The largest deviation observed'was slightly less than 70% of measured flow. Based on this comparison we conclude that the licensee's uncertainty factor is accbptable for application to results obtained with the design methodology discussed above.
Effects of Hioher System Pressure The P'rediction of' core spray distribution discussed perviously was determined for steady state conditions based on a reactor system pressure of 30 psia. The predicted core spray distribution for this pressure is representative of large break loss of coolant accident conditions ~ in which the vessel depressurizes rapidly to pressures below 30 psia prior to initiation of core spray. Core spray distribution predictions at higher -pressures are needed in order to support Appendix K e
heat transfer assumptions in small break loss of coolant accident analyses.
This is because core spray is credited in small break analyses during the period in which vessel pressure is below 125 psia (i.e. pressure for rated core spray flow) yet significantly above 30 psia.
For the limiting small break case this period is about one minute, i
I e
,a aaa s-d+mmeam e s is, e a weeeee w we w e e a we mee, e e.w e e -e es e e e w r e ere-we e w r e
+e--
we e s a.
eve-r,
, - - W -e~.u++e.-me w-e,,--
.+.-e..-,,-e.m, rw*
-w--<-a-es-e--.==~
i
. The licensee has made predictions of core spray distribution for pressures
~
up to 55 psia.
Predictions at pressure above 55 psia have not been made.
The rea.on is that spray interaction effects detennined from full scale tests in air at atmospheric pressure become highly unrealistic.
In addition confirmation tests of spray distribution in steam (Reference 6) were performed only for pressure up to 73.5 psia (5 atm).
Base.d on the uncertainty in spray flow at high system pressure the licensee has re-analyzed the affected small break LOCA cases taking no credit for j
core spray heat transfer prior to reaching a vessel pressure of 55 psia (i.e. assuming a minimum bundle spray flowrate of zero).
In these analyses the 1.icensee has modeled the effect of steam cooling and shown that the Appendix K clad temperature limits can continue to be met when steam cooling heat transfer is credited in the critical period.
In order for the staff to concur with the licensees position it must first review the 1
revised ECCS evaluation nodel on which it is based. The licensee's presently approved ECCS evaluation model does not include the effects of steam cooling heat transfer.
3.2 Minimum Reouired Bundle Spray Flowrate I
The licensee has determined that rod bundle heat transfer under spray cooling is eouivalent to 10 CFR 50 Appendix K Assumptions at a minimum bundle flow rate of 1.8 GPM. This detennination was made through calculations with an analytical model called.COREC00L.
Spray flowrate was varied in the COREC00L calculation until the. predicted clad surface temperature equaled the clad surface temperature calculated with the approved ECCS evaluation model incorporating Appendix K heat transfer coefficient.
COREC00L i's a model for evaluation of core heat-up transients for a fuel element and for evaluation of the performance of core spray systems.
It consists of two basic models, a. fuel rod model and a;model for the two-phase flow in the system.
The fuel rod model is a heat conduction model, which is also applied to the channel. The two-phase flow model is based on solution of th~e conservation equations for mass, momentun and energy, and the equation of state. The two phases are treated separately, and physical models and correlations are developed for the interchange of mass, momentum, and energy In addition, for the liouid phase, films and droplets are treated separately.
Thermodynamic equilibrium is. not assumed, and the steam is allowed to be superheated and the watdr subcooled.
The coupling between the fuel rod model and the two-phase flow model is taken into account through a numb' r of physical models and correlations for the heat transfer, e
which includes conduction, convection, and thermal radiation.
COREC00L has been qualified through comparisons with spray cooling experiments in full scale electrically heated BWR fuel burdles.
(Reference 10 and 11)
(23) experiments carried out by General Electric and 13 experiments carried out by A. B. Atomenergi in Sweden have been used for the qualification.
As shown in table 1, the experiments cover a wide range of conditions, such as bundle power, pressure, initial temperature, and spray flow rate.
In general the agreement between COREC00L calculations and experiment is very good.
In. the General Electric (GE) experiments the average erro~r in calculated peak clad
. TABLE 1
~~~~
RANGE OF PARAMETERS IN GE AND SWEDISH SPRAY C00 LING' TESTS PARAMETER RANGE GE Tests Swedish Tests Ihitial Rod Temperatur*
1220*F - 1616*F 1274*F - 1364*F Bundle Power 250 kw - 300 kw 151 kw - 341 kw Spray Flow Rate 2.5 GPM - 8 GPM
.64 GPM - 2.55 GPM Pressure 1 Bar 1 Bar - 20 Bar temperatures is.7% and the R.M.S. error is 3.1%.
In the Comparison with the swedish experiments the average erior is 1.0% and the P.MS error is 3.1%, which is in good agreement with the results obtained for the comparison to the GE experiments.
In addition to the tests performed by General Electric and Riso National Laboratory (Sw'edish) other independent tests in this country and in Japan support the COREC00L result.. Tests by Exxon Nuclear Company (Reference 12) on a 268 kw 8 x 8 BWR fuel tundle show a 1.8 GPM/ bundle flowrate as being 'nese tests (Reference 13) on an 8 X 8 full scale BWR sufficient to provide Appendix K heat transfer coefficents.
Japa bundle show that Appendix K heat transfer coefficents are obtainable at spray flowrates between 1.0 and 1.8 GPM/ bundle depending on rod location.
We conclude that the licensee's determintion of required minimum bundle flowrate is acceptable because it is well supported by experimental da.ta.
~ 4.0 Conclusion and Findinos Based on the analyses presented by the licensee and our review of those analyses we conclude that core spray distribution in the Oyster Creek plant is significantly affected by the presence of a steam environment.
The analyses presented by the licensee indicate the mininum bundle flowrate including uncertainties is 2.0 GPM based on a single core. spray sparger flowrate of 4690 GPM.
Original full scale tests in air showed a minimum bundle flowrate of 2.45. GPM based on a singel sparger 'lowrate 3400 GPM.
Specific conclusions and major findings of the staff review are given below.
1.
The minTEum bundle flowrate of 2.0 GPM predicted in the licensee'.s analysi.s was arrived at using a slightly modified version of the General Electric design methodology for determining core spray distribution.
We required that this plant specific (0yster Creek) method be shown to be conservative with respect to the original methodology confirmed by tests in steam.
We have reviewed the data presented in the licensee's submittal'for the purpose of demonstrating conservatism and find it inconclusive.
In order to resolve this concern appropriate data generated with the Oyster Creek method must be shown to bound that generated with the confirmed method, as discussed in Section 3.1 of our evaluation.
2.
In the analysis used to predict minimum bundle spray flowrate the licensee has assumed a single sparger flowrate of 4690 GPM based on plant tests.
The Technical Specification on Core Spray System presently reflects minimum allowed sparger flowrates of 3400 GPM for System I and 3640 GPM for System II.
Plant Technical Specification 3.4 must be amended to reflect the revised safety analysis assumption.
3.
In order to demonstrate acceptable consequences for certain small break loss of coolant accident scenarios, the licensee has re-analyzed the small break cases assu. ming no credit for spray cooling and taking credit of cooling of the rods by heat transfer to steam rising through the fuel bundle.
This was done because of uncertainties in core spray distribution at elevated reactor system pressure.
The present Oyster Creek approved ECCS evaluation model does n6t treat the affects of steam cooling. ' Review of the licensees revised model must be performed by the staff prior to evaluating the safety significance of Core Spray distribution uncertainty in small break LOCA Scenarios.
4 The licensee has determired that the minimum recuired bundle spray flowrate needed to achieve Appendix K heat transfer coefficients is 1.8 GPM.
We accept this determination because it is well supported by test data.
5.
We have reviewed the uncert'ainties applied in the determination of i
minimum bundle spray flowrate and find them acceptable based on comparison of test results with analytical results.
5.0 References l
1.
Letter to Mr. I. R. Finfrock, Jr., JC P&L Co. from Mr. George Lear, Chief, Op'erating Reactors Branch #3, NRC; December 10, 1976.
2.
Letter to Mr. George Lear, Chief, Operating Reactors Branch #3, NRC
{
from Mr. I. R. Finfrock, Jr.; January 17, 1977 3.-
NEDO 20566-3. GE licensing topical report, " General Electric Comoany Analytical Model for Loss of Coolant Analysis in Acenrdance with 10 CFR 50 Appendix K Amendment 3 - Effect of Steam Environment on BWR Core Spray Distribution"; April 1977.
w
=
i 8-4.
Letter and enclosure to Dr. D. G. Sherwood (GE) frorr. Mr. Olan Parr (NBC), June 13, 1978; Interim Safety Evaluation of NEDO-20566.
5.
NRC memorandum, Thomas Ippolito to Dennis Crutchfield, October 7,1982 6.
NEDO-24712, " Core Spray Design Methodology Confirmation Tests,"
August 1979, General Electric Topical Report 7.
NRC memorandum, P. S. check to R. L. 'edesco September 8, 1980 T
8.
Letter and enclosed Safety Evaluation for amendment #34 Dennis Ziemann (NRC) to I. R. Finfrock, Jr. (JCP), November 24, 1978 9.
NUREG/CR-1707, "BWR Refill.- Reflood Program Task 4.2 -
Core Spray Distribution Final Report," March 1981.
10.
Letter and Enclosure from R. L. Gridley (General Electric) to M. Caruso (NRC); August 16, 1983.
11.
L. Nilsson, L. Gustafson, R. Hauy, " Experimental Investicjation of I
Cooling by Top Spray and Bottom Flooding of a Simulated 64 Rod Bundle for a BWR,." STUDSVIK/RL - 78/59 June 1978.
12.
XN-NF-78-12 " Spray Cooling Heat Transfer Test Results phase II l
Facility and Test Data ENC 8 X 8 BWR Fuel," Exxon Nuclear Company Proprietary Report; May 1979.
13.
M. Naitoh, " Heat R moval by Too Spray Emergency Core Cooling,"
presented at Second two-phase Heat and Mass Transfer Symposium Miami, FL; April 1979.
14 Letter and Attachment from Mr. Peter.B. Fiedler' (GPU Nuclear) to Mr. Darrell G. Eisenhut (NRC) June 30, 1983.
g r
ay 99 e
y
,--r-----
,-+
w 2
)
ATTACHMENT l
JERSEY CENTRAL POWER & LIGHT OYSTER CREEK NUCLEAR GENERATING STATION SAFETY EVALUATION REPORT MATERIALS ENGINEERING BRANCH DIVISION OF ENGINEERING
Background:
On the basis of cracking in core spray systemsf discovered at Oyster Creek in October 1978 and Pilgrim in January 1980, IE Bulletin No. 80-13 TIM was issued.
IE 80-13 mandated that at next scheduled and following 3
refueling outages until further notice, a visual inspection was to be rade of the core spray spargers and the segment of piping between the inlet nozzle and the vessel shroud.
]WTCD MM 3 1 GPUN Topical Report No. 013, Rev. 0 summarized the. video visual 3
inspection of the Oyster Creek Spargers:
1978 - 1 through wall crack; fix involved installation of a clamp assembly over crack.
1980 - 19 indications observed, which were identified as cracks.
1982 - enhanced video reassessment of 1980 indications reidentified 3 indications as cracks, 2 indications as possible cracks.
Reassess-ment concurred in by 3 NDE qualified inspectors.
1983 - Reinspection disclosed no ir:dications. (Ajo74; Aff44 of /A/SA/CF#45 t,,,m epam
=
BWAcKSTS)
A. st.
4 s
\\
~
// > 8 g >
wA<~,
n.t a ua
.es-[f w LLv & l A cu L4L aJisy n
Av ot "
,cw" k u,i8k n
dan i
n teaa n
est L d, do, eme u &~ Jy cine 6, eta a da-f sbla 7
.%As
'214. t abyi 2 /ns ~ s,.
L a s k,l, y a A <
.m y
a z
k H sensy L cp y / A y tJa e ce /w we saf,you M
/e c
m en v~
de c a<--~e ef ' &
J{ $wn uen~p a
a n >&
w i
- s
4
- Discussion:
g inconsistency between the evaluation of the 3 NDE qualified
/
7-The M
G/
inspectors who called 3 relevant indications on reassessment of the 4
1980 video tapes and the 1983 reinspection, which did not reveal these I
indications, indicates that the examination procedure lacks a con-firmed reliability.
Factors which could contribute to the lack of reliability include the inability to view the bare base metal surface, because of its reflectivity, and the inability to date to focus on g
,c P-f
/ ';
an in situ artificial flaw, such as a vibrotooled (or engraved),o 'g & f o4 t, s
i component or part identification marking, part assembly match A, e
marking, or induced surface scratches at mapped locations.
~. _ _ _. _ _
/ While the capability of the examination procedure to resolve a one mil )
.g' wire _has_been _ demons _tratgif, i
/
reliability
- espect to-determining the relevancy of indication ying beneath a no redictable, probably
- non-uniform thickness of crud remains to be determined..
bu..g o
n Since_the deferment-of _ the.replacementmf-thetster-tYeek core
^
,*A
,n -yy Spray Spargers is--to be based on ev.idenca_that -lut-lezor-no pr_ogrgs-t slan_of=-cracking has-occurred, an inspection method that has the sensitivity to allow crack dimension measurement is necessary to l
evaluate crack progression.
An in situ artificial flaw of known dimension, as discussed above, could be used to scale crack progression.
g
a J
Conclusions:
The " state of the art" of the video visual examination procedure used in the inspection of the Oyster Creek Core Spray Spargers, at this time, precludes the assignment of the reliability on the crack length measurement upon which the deferment of replacement of the spargers could be based.
An ultrasonic inspection in evaluating the Oyster Creek Core Spray Spargers a455-does not appear feasible because of the limited access.
We further concluded that a UT inspection is not practical due to the high radiation.
The visual air test disclosing the presence only of through wall cracks is tilgs limited as an inspection method.
The method is further limited by constraints of piping configuration which precluded air test of System I, as noted in GPUN Topical Report No. 013, Rev. O.
E 6
'vrb N4 4 M +,~ a,, I "' cf-P' W S a
it na tut
? + :( "g 7
,~
Obm-UPod
-O r L.. :tti;..., 4the examinat,ippff the Oyster Creek Core Spray j h h h, f System 1$.gthe tenth refueling outaget,g.. -_.
..__., conclud e r ') (;
D u st M a-( k iEhe!d(77/l 3Mff
?r :
1 c
I.. [.
r
-e 4masseet that major progression;"tT, -..t.ly-uQue,,t IU M has not e T r.
F
)A t
occurred and operation with the present soargers for another fuel cycle q ;- E :
tiuucr maw asn)imten: Change Rcquest,"o-dM,
[. fi is acceptable.
. s..... i co @ d.
pv 'I b ]
wt
~
Ress4on4 =is ecceptreble. The staff recommends that core spray spargers 3' D q
3 a
h 6
at Oyster Creek be replaced at the end of the next fuel cycle, unless E
- f' A@ NKC AffMot/CD 1
subsequent inspection using improved visual examination techniques p
3 s
?
v disclosee no new cracks and no further progression of existing cracks.
.3
- f x
v
)
h
/
l
h ysh & Lk f9
/ / '7 fe[
~
'{
q{
ENCLOSURE 1 Evaluation Findinas go 1.
The minimum bundle flowrate of 2.0 GPM predicted in the licensee's analysis was arrived at using a slightly modified version of the General lip0 Electric design methodology for determining core spray distribution.
We D
required that this plant specific (0yster Creek) method be shown to be conservative with respect to the original methodology confirmed by tests in steam.
We have reviewed the data presented in the licensee's <)
submittal for the purpose of demonstrating. conservatism and find it
\\
inconclus'ive.
In order to resolve this concern appropriate data generated i
with the Oyster Creek method must be shown to bound that generated with
\\
the confirmed method, as discussed in Section 3.1 of our evaluation.
\\
2 In the analysis used to predict minimum bundle spray flowrate the licensee b l,, M.
has assumed a single sparger flowrate of 4690- GPM based on plant tests.
I gth\\
The Technical. Specification on Core Spray System presently reflects i
minimum allowed soarger flowrates of 3400 GPH for Systen I and 3640 GPM for System II.
Plant Technical Specification 3.4 must be amended to I
reflect the revised safety analysis assumption.
4""
3 In order to demonstrate acceptable consequences for certain sniall break g
loss of coolant accident scenerios the licensee has re-analyzed the small u
T,
? v break cases assuming no credit for spray cooling and taking credit of i
cooling of the rods by heat transfer to steam rising throught the fuel bundle. This was done because of uncertainties in core spr.ay distribution at elevated reactor systen pressure. The present Oyster Creek approved j
ECCS evaluation model does not treat the affects of steam cooling.
Review of the licensees revised model must be performed by the staff prior to j
evaluating the safety significance of Core Spray distribution uncertainty in small break LOCA Scenarios.
4.
\\gy~
The licensee has determined that the minimum required bundle spray flowrate g
needed to achieve appendix k heat transfer coefficen'ts is 1.8 GPM. We
'j j accept this determination because it is well supported by test data.
t 5.
We have reviewed the uncertainty factor applied in the determination of minimum bundle spray flowrate and find it acceptable based on comparisons y
of test results with calculated results.
A-s3