ML20140G556
| ML20140G556 | |
| Person / Time | |
|---|---|
| Issue date: | 01/29/1977 |
| From: | Lainas G Office of Nuclear Reactor Regulation |
| To: | Miner S Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20140F372 | List:
|
| References | |
| FOIA-85-665 TAC-2285, NUDOCS 8604020470 | |
| Download: ML20140G556 (5) | |
Text
. _ _
s 1 -)
j o
~
/
umTro starts e,
NUCt. EAR REGULATORY COMMISSION 8
i i
wAsmNGTON, D. C. 20656 jp. 2 e 10 ' -
MEMORANDUM FOR:
S. Miner, Project Mar.ager, Light Water Reactors Branch No. 3 DPM FROM:
G. Lainas, Chief, Containment Systems Branch, DSS THRU:
R. Tedesco, Assistant Director for Plant Systems, DSS
SUBJECT:
REVIEW OF GE TOPICAL REPORT, NEDM-20988, "CAORSO RELIEF VALVE LOADS TESTS - TEST PLAN" (TAC-2285)
As requested, the Containment Systems Branch has reviewed the GE topical report, NEDM-20988 entitled "Caorso Relief Valve Loads Tests - Test Plan,"
and has prepared the enclosed coments including the following significant areas of concern.
1.
Effects of leaking SRV on quencher loads has been identified as an area of concern. The Caorso test plan, however, has not addressed this concern.
It is our position that tests on leaking SRV should be conducted and that a test matrix addressing this concern should be provided for our review.
2.
The test matrix shows that the parameters of interest will be repeated only once. We believe that this is not* sufficient to uemonstrate the repeatability of test data. We, therefore, suggest that additional tests be conducted to demonstrate repeatability.
It should be noted that our review has excluded consideration of fluid /
structure interaction effects. Coments concerning this area should be obtained from SEB.
- f. f
~
Gus C. Lainas, Chief Containment Systems Branch Division of Systems Safety
Enclosure:
As Stated cc: See Page 2
Contact:
T. Su, CSB 492-7711 koj4020470060114 C
Fingg&gQ-665 pJ pop
b S. Miner cc:
S. Hanauer F. Schroeder R. Tedesco J. Glynn C. Long O. Parr G. Lainas J. Knight I. Sihweil D. Vassallo R. Tedesco J. Kudrick J. Shapaker R. Boyd (w/o encl.)
W. Mcdonald (w/oencl.)
T. Su 1
i 4
1 i
I 4
i 1
3' 1
Y3 l
Request for Additional Information General Electric Company Topical Report: NEDM-20988, Revision 2 Test Plan - December,1976 Reviewed by:
T. M. Su 1.
Effects of a leaking SRV on quencher loads has been identified as an area of concern. The Caorso test plan, however, does not address this concern.
It is our position that tests on a leaking SRV should be conducted. Therefore, a test plan considering a leaking SRV should be provided for our review.
2.
The Caorso test matrix indicates that most of the parameters of interest udll be repeated only once. Tests performcd either for ramshead (Quad City ar.d Monticello) or for quencher (NEDE-ll314-08) exhibited a great degree of data scatter. Therefore, we believe that the current Caorso test matrix is insufficient to determine the repeatability of the test data. We recommend additional tests be conducted for first actuation of a SRV and subsequent actuations of the same SRV to demonstrate repeatability. The number of tests should result in test data with statistical significance.
3.
Based on our evaluation of previous SRV test data, we find that the SRV discharge time and duration between first actuation and subsequent actuation influence the quencher load due to subs >
it actuation.
Therefore, provide the value and the basis for the selection of the times for the Caorso tests.
. 4.
Page 4-6 states that a complete understanding of the subsequent actuation effect requires data on pool temperature in the vicinity of the quencher, pipe temperature and pressure following valve closure, flow rate of air through the vacuum breaker and dynamics of back flow of water. We agree that the air temperature history inside the pipe ~could be important. However, insufficient information has been given in the test plan regarding the measurement of air temperature in the pipe. Clariff what measurements or calculations will be made to monitor this temperature.
5.
The sensor failure rate was found to be quite high in the Monticello Plant test program. Sensors of the same manufacturer model used in the Monticello will also be used in the Caorso. In light of this experience, we believe that redundant instrumentation is needed in critical areas. For instance, redundant sensors should be provided i
in the following locations:
a.
The vicinity of quencher A and the place by which combined loads from multiple SRV's actuation will be determined, b.
SRV line between elevation 51.612 and 45.770.
In addition, level probes should be added between L1 and L12.
e
i 3-6.
Submerged structure loads have been identified as a primary design load for the Mark II containment. We believe that the analytical program indicated in the Mark II owner group meeting which was held on February 16 and 17,1977, is insufficient to support the l
design loads for submerged structure without experimental data.
Therefore, we recomend that additional pressure sensors should be installed on support columns and downcomers to measure the drag load during SRV operation.
7.
Provide the locations for pressure sensors Nos. 19, 23, 35, 36 and 37.
I
O T H. Q4 96Y 21 GENER AL h ELECTRIC suC' era samaav PROJECTS DIVISION
[, -
GENERAL ELECTRIC COMPANY,175 CURTNER AVE., SAN JOSE, CALIFORNIA 95125 MC 682, (408) 925-55040
- p. I
[N f few '#
4b December 18, 1978
~,,,gif 5 K' Dr. Harold R. Denton U
.fy /
Director, Nuclear Regulations N
U. S. Nuclear Regulatory Comission
/_ g 4 7920 Norfolk Avenue L7 Bethesda, MD
Dear Dr. Denton:
This is to provide General Electric's coments on the status of the five BWR-related items in the NRC's list of " unresolved safety issues." I believe these coments show that progress is being made on these issues, and they are not of safety significance to BWR plants.
Based on discussions between Walt D'Ardenne and Del Bunch, only five of the issues involve the BWR.
We are pleased to provide you these status reports.
Yours truly, i
A 4.
G. G. Sherwood, i anager Safety & Licensing Operatior.
GGS:bp/1097 Attachment cc:
L. S. Gifford W. H. D'Ardenne D. F. Bunch-NRC O
y-3_
/
/
1
~
PROPOSED " UNRESOLVED SAFETY ISSUES" ISSUES FOR 1.
Waterhammer (A-1)
PWR 2.
Asymmetric Blowdown Loads on the Reactor PWR Coolant System (A-2) 3.
Pressurized Water Reactor Steam Generator Tube PWR Integrity (A-3, A-4, A-5) 4.
8WR Mark I and Mark II Pressure Suppression BWR Containments (A-6, A-7, A-8, A-39) 5.
Anticipated Transients Without Scram (A-9)
BWR Nozzle Cracking (A-10)
BWR 7.
Reactor Vessel Fracture Toughness (A-11)
PWR 8.
Qualification of Class IE Safety-Related PWR, BWR Electrical Equipment (A-24) 9.
Reactor Vessel Pressure Transient Protection PWR (A-26) 10.
Residual Heat Removal Requirements (A-31)
- 11. Seismic Desiga Criteria (A-4'))
NRC CRITERIA 12.
Pipe Cracks in Boiling Water Reactors (A-42)
BWR 13.
Emergency Sump Reliability (A-43)
- 14. Station Blackout (A-44)
NRC CRITERIA ISSUES PWR 9
BWR 5
NRC 2
~
PBS:at/23M1
_,-..-.e,
,--..-.-,,.,..--_,-n-.-
.w--
~
4.
Title:
BWR Mark I and II Pressure Suppression Cont'ainments (A-6,A-7,A-8,A-39)
Issue:
As a result of an ongoing GE testing program, new hydrodynamic containment loads associated with a postulated loss-of-coolant accident and the anticipated discharge of safety reliaf valves were identified which had not been explicitly included in the original design of Mark I and II containments.
BWR Status:
GE and the Mark I and Mark II owners groups are in the final process of defining dynamic forcing functions for Mark I and II containments.
Each utility is in the process of perforair.g appropriate evaluations to determine the response of their containments to these loads.
A substantial amount of interim / preliminary analyses has already been completed and are being reviewed by the NRC. The Mark I Load Definition Report will be submitted to the NRC starting in December 1978 with the final sections provided in March 1979..The Mark II Dynamic Forcing Functions Information Report was issued September 1975 and has been periodically updated with Revision 3 having been issued in June 1978. The NRC's final Zimmer Safety Evaluation Report which includes typical lead plant Mark II assessments is expected to be completed January 1979.
Basis for Continued Operation and Licensing The acceptability of Mark I continued operation has been evaluated based on the Mark I Short Term Program and reported in the NRC's Safety Evaluation Report issued December 1977.
The acceptability of Mark II operation has been demonstrated through licensing evaluations leading to operating licenses on Zimmer.
t PBS:sj/23H4
_ _ _. _... _.. _ _.. _ _ _.. _, _ ~ _.. _ _.. _.,. _
5.
Title:
Anticipated Transients Without Scram, ATWS (A-9)
Issue:
The NRC has modeled LWRs using extremely conservative probability input data. The results suggest that design changes ranging from minor to significant are needed to meet NRC objectives.
BWR Status:
i GE has modeled the BWR using representative probability input data. The results conclude that the failure to scram is extremely unlikely because of the redundancy in the BWR control rod system.
In summary, there is no need to install additional shutdown systems, and above all, no justiff-cation for making ATWS a design basis event.
At the request of the NRC, GE has provided descriptions and costs for Reactor Pump Trip, Alternate Rod Insertion, and modified Standby Liquid Control Systems which can be utilized to mitigate the postulated ATWS event even though GE believes that the event is so unlikely that no mitigation is necessary.
Basis for Continued Operation and Licensing The probability of the event is so small that there is no safety issue and ATWS should be treated as a class 9 event. Even based on the conserva-tive NRC modelling, continued operation is an acceptable risk because of the low probability of an anticipated transient without scram between.
f
.l 1
1 l
l l
l l
PBS:sj/23HS
l 6.
Title:
BWR Nozzle Cracking (A-10)
Issue:
Cracks have been reported in feedwater nozzles and in control rod The cracks have been hydraulic return line nozzles in operating BWRs.
initiated by high cycle thermal fatigue.
BWR Status:
Improved nozzle designs, inspection techniques, 'and operating procedures a
have been developed to avoid the nozzle cracking problem or to detect i
All BWR plants are committed cracks before they become a safety concern.to incorporate design modific problem, to perform verification tests to assure the adequa design changes to confirm the adequacy of the modifications.
t Basis for Continued Operation and Licensing Operating condition envelopes imposed on operating plants and improved nozzle designs for plants under construction provides adequate safety margins for the plant life based on evaluations performed by the NRC and GE.
PBS:sj/23H6
8.
Title:
Qualification of Class IE Safety-Related Equipment (A-24)
Issue:
It is the NRC's position that construction permit applicants for which a Safety Evaluation Report was issued after July 1,1974 should qualify all Class 1 electrical equipment to the requirements established in IEEE 323-1974, "IEEE Standard for Qualifying Class 1E Electric Equipment for Nuclear Power Generating Stations".
BWR Status:
GE is developing a Licensing Topical Report (LTR) which sets forth criteria by which Class 1E Equipment is qualified per the guidance of IEEE 323-1974 and the GE interpretation of Regulatory Guide 1.89.
The Licensing Topical Report is being developed to satisfy an open issue on the GESSAR docket and will be directly applicable to GESSAR Reference plants. Other projects committed to IEEE 323-1974 will reference the Licensing Topical Report.
For Projects which did not contractually commit NSSS Class IE equipment to IEEE 323-1974 but are or will be required to comply with that standard for licensing, a plant specific retrofit program is planned.
Basis for Continued Operation and Licensing Present equipment qualification programs are adequate; and there is sufficient basis to continue with plant licensing and to allow plant operation pending ultimate evaluation of programs and results.
I PBS:sj/23H8
h
,(
~
i kN v,1, 12.
Title:
Pipe Cracks in BWRs (A-42)
Issue:
Iritergranular stress corrosion cracking has occurred in BWR Type 304 stainless steel recirculation system piping weld heat affected zones.
BWR Status:
Both GE and NRC evaluations have concluded that Intergranular Stress Corrosion Cracks (IGSCC) at weld heat affected zones of the piping are not a safety concern since (1) IGSCC can be detected with current in-service inspection procedures which are highly effective; and (2) detectable leaks always precede major pipe cracks. All plants perform scheduled inspecticns and are equipped with continuous leak detection systems to assure plant safety.
Numerous measures to ameliorate IGSCC, including the use of alternate r.11oys, are currently being implemented on both operating plants and plants under construction to improve availability.
Basis for Continued Operation and Licensina Intergranular Stress Corrosion Cracking is a BWR plant availability concern but is not a safety issue since inservice inspection programs and leak detection systems preclude major cracks.
PBS:sj/23H12
q
\\-
i s
UNITED STATES j
NUCLEAR REGULATORY COMMISSloN j [,
g..
wasuiucTom. o. c. 2 cess
- %(m.//
r.:: M gl f (
g f
.iA.~. 1 1573
[
i
~
Generic Task No. A-39 50-220, 50-237, 50-245, 50-249, 50-254, 50-259, 50-260, Docket Nos.:
50-263, 50-265, 50-271, 50-277, 50-278, 50-293, 50-296, 50-298, 50-321, 50-324, 50-325, 50-331, 50-333, 50-341, 50-354, 50-355, and 50-366 i
Members of Mark I Owners Group (excepting Jersey Central LICENSEES:
Power & Light Company - Oyster Creek)
SUBJECT:
SIM1ARY OF MEETING HELD ON FEBRUARY 13, 1979 WITH REPRESENTATIVES OF THE MARK I OWNERS GROUP PROGRAM COPNITTEE 13, 1979, a meeting was held in Bethesda, Maryland, with On February representatives of the Program Connittee of the Mark I Owners Group.
The purpose of the meeting was to discuss the generic Mark I program relating to the safety / relief valve loads and its application to individual plants. It is noted, however, that the discussion will not be applied for Oyster Creek since they are not using the quencher device developed under this particular progrei.
An attendance list and a copy of the meeting handouts are provided as enclosures 1 and 2, respectively.
The following summarizes the significant points of the meeting:
4 It 1.
Representatives of General Electric described the SRV program.
consists of several subprograms, including a 1/4 scale test program, scaling analysis, full scale testing and an analytical model Test results from the 1/4 scale testing and
-l development effort.
full scale testing (Monticello) were used to verify the analytical 1
L models. These include three key analytical models; namely, SRV discharge line reflood model, SRV discharge line clearing model and j
torus shell load model. The three models are closely related; the calculational results from one model serves as input information for another model. GE is responsible for the model development and provides guidance for AE's and utilities in their appl'eation of these models. *Je expressed our concern about the level of guidance provided. We believe that the guidance as outlined in the Mark I Containment Program Load Definition Report is too general. We will i
require that more specific guidance be developed and provided for l
our review and evaluation.
C
,n
-p jvj j)l
~2-G73 Multipliers and other factors are used in various'models. We questioned 2.
the justification regarding these constants. GE indicated that the justification for some of the factors and multipliers cannot be clearly However, they will attenpt to provide the justification in identified.
the next meeting with the staff.
on the assumption for delay time, which There was also some discussion 3.
is defined as the time required for the SRV discharge line to depressurize following SRV closure and as a result allow the reflood transient to The reflood model assumes that the delay time is proportional to begin.
the air volume stored in the SRV line. We requested that GE provide its o
justification for this assumption.
These We expressed our concerns related to the torus shell load model.
4.
are the substantial discrepancies between the data reported in the GE Monticello test report and the data used to verify the model.
indicated that clarification of this matter will be provided at the next meeting.
T. M. Su, Task Manager Task Action Plan No. A-39 Containment Systems Branch Division of Systems Safety Enclosures-i As Stated Distribution w/ enclosures w/o enclosure 2 Docket File H. Denton i
)
NRR Reading R. Mattsen CSB Reading R. Boyd R. Fraley, ACRS (16)
D. Rcss S. Hanauer R. DeYoung R. Tedesco D. Vassallo I&E (3)
D. Skovholt NRC pDR Local POR W. Butler T.,Ippolito(4)
G. Lainas V. Noonan D. Ziemann J. Kudrick C. Grimes L. Ruth T. Su
^ '
~
ATTENDANCE LIST
SUBJECT:
NRC T-QUENCHER ROADMAP DATE:
February 13, 1979 LOCATION: Bethesda, Md.
NAME REPRESENTATING/ ORGANIZATION T. Martin NUTECH C. Tung NRC/BNL C. Economos BNL S. Hucik GE C. I. Grimes NRC/ DOR T. M. Su NRC/CSC/ DSS L. C. Ruth NRC/CSB/ DSS L. Steinert GE T. J. Mulford GE a
e r o e
e
,-n--
- - -, - - - - - --~
-m-- - -, - -, - -
,-----,,---+-*-~-------n------
w
--y-w
--,-e-----~~l~
r
l M AftX I con TAIN ME NT' Paccagm
's
~
NRC LDP.
REVt E w/
S/RV Ro AD M AP 5 FEr$RU.ARY l'5, 1979 j
1 l
l SAR-i 2.-13-11
~
LOAD DEFINITION REPORT SCOPE I
INTEGRATES RESULTS FROM THE TESTING AND ANALYTICAL e
TASKS COMBINES METHODOLOGIES AND PLANT UNIQUE LOADS e
e PROVIDES LOAD COMBINATION TIMING DOCUMENTS FINAL DESIGN BASIS LOADS USED BY OWNERS /AE'S e
WITH STRUCTURAL ACCEPTANCE CRITERIA, PROCEDURES FROM OTHER TASKS AND AE SUPPLIED LOADS FOR MARK I PLANT EVALUATION AND DESIGN OF MODIFICATIONS.
REFERENCES MARK I CONTAINMENT.LTP TASK REPORTS FOR COMPL e
JUSTIFICATION OF LOAD DEFINITIONS.
e g u.a
'1-0-79 f
e O
LOAD DEFINITION REPORT CONTENT AND ORGANIZATION e
LOADS CONSIDERED IN LDR LOCA RELATED LOADS SAFETY RELIEF. VALVE DISCHARGE LOADS
~
OTHER CONSIDERATIONS (SECONDARY LOADS) e LDR SATISFIES LONG TERM PROGRAM REQUIREMENTS FOR EVALUATION OF ALL L0 ADS e m en S A H -3
~
- 2. 7 9 i
O e
l I. ---
$NtVDL CLE ARING MODEL MARK I E OAfD DE F INITION 84L PollT l
ENIVOL PitES$URE y
PetOVIDES CA Cut ATIONAL PHOCEDUIILS F Oll:
OSTAIN SNIVL't AND TORUS i
GEOMEllIV AND OPER A18NG f,AtVOL TIsauST LOADS 5/RVDL TEMPEllAflillE CONulIlONS FOst 11sE CASE
'1o ut EVALUATED Ti1UENCllER PRESSURE WATIft JLT DilAG LOADS ON l
WATER CLE ARING VELOCITV yWATEis CLEARING ACCELLitATION WATER CL E AltING TIIRUST O
LOADS ON T-QULNCIIEtt AltMS
/
E E
WAiCR cLWW4 TsME O
s El to j
u.
p I*
3 y La co 8
%^
o3 l
StRVOL RE FLOOO MODEL CALCULATES-5/HVDL REF LOOO HEIGHT TORUS SHELL LOAO MODEL AND TIMING If Ost SuuSEQUENT S/RV CALCULATES:
ActuAIION ANALY$f 58 l
Asa mue0LE PRESSURE A
AIR Sue 0LE R AOluS
, N A:R ouusi E un Ac. uOoE L I
CALCULATES:
Am suseLE VERTICAL POSITION &
AIR 80ullt E INDUCED DRAG TORUS self LL PftESSURE LOADS On suuu[HGED STituCTultts g
DisTRisuTION l
e-3.
na N
Figure 5.l-1.
T-thiencher Land I)efinitioR Scheme 3)
'YI.
a
~a +
- - -. ~ ~ ~ - - ~
~-
- - ~ ~ -
5/HVDL C1 EAHING MODEL MARKILOAD DEFINITION REPollT
+
CALCULATES:
SAtVDL PHESSURE 00TASN SNIVDL AND TOHUS SNtVDL TINtusT LOAos GEOutiTilY AND OPERATSNG PROCEDURE FOR.
CONDITIONS 7064 YllE CASE PROVlOES CALCULATIONAL 10 DE EVALUATED WATEH C4 EARING ACCELERATION AND VELOCITY I
i WATER JET LOADS ON SUOMEttGED SiHUCTullLS DIScilARGE PetES$URE
- ~
llOflIZONTAL POSITION OF SuellLLS e--
l 1
1 r 14 r
b ta L
SAlv AIR suest E PRESSURE MODEI.-
[
ShlVDs flEFL OOO MODEL CALCULATES:
ts
.oo CALCULATES.
AIR SUSSLE PitESSURE S/Blv0L llEF LOOO ItEIGili
&~
ANo YlueNO lFOH SueSE-AM StMSLE RADIUS CUENT ACTUAYlONS8 IIElGif f ANO TIMING AIR Sue #LE VERTICAL POSITION t
1 f
1 P i
SJRV AIR BUSSLE PRESSURE AIR BuesLE Oft AC MODEL j
ATTENII4 TION MODEL CALCULATES:
CALCtf',4TES:
E
'D AIR BUSSLE INDUCED DilAG g r. m$
T011U8 Silf Lt. PRESSURE LOADS ON SUBMElIGED DISTRieuTielle STRUCTURES
,44
o J lf) o Figure 5.1-2.
llamshead lead Definition Scheme o
i I
~ ~
S[RV DL CLEA Rt NG MODE C$7 LDR SECTioM 5 2.l s/R\\/
DtscHARGE LtME CLEA RIMG TRAtJSlEMT LOADS McoEL (M Pu.T PtAMT u.MlQu G INFbRMATIOM e
Ko DE L ouTPu.T LOAD DEFIN ITtoM
- S/RV OL PR.CS50.R6 5,11 Pi PE SEQ M.EM T PRESSURE-
.52 7 SIRV DISCHARGE LIME me, TEMP, S/RVOL THRu.5T loa D 5
- 5. 2., i THRasT LOADS og s/gypt pipe Sc6Ncq 5 T-QU.E M CH E R. Pg.E s su.a.E 5 2., \\
T-ctpEMCHGR.
PgESSuR.E loa.DS,
\\AlATER. C.LEARt RG VELoc IT Y S.2.4 v)ATER., SET DRACs LOADS o AJ SasMERGED g7gQcTMRES,
- 5. 2-G THRu.sT LOADS OM T-Qu.EMcMGR ARMS
~
I SAH-6 1-t 5-71
Str& D L Cl.EA RlWG 40 D &'L.
WATER CLEA RING AccELE/ EAT /ON
.G~ 2.. (a 77/ RUST 40 ADS ON 7: QuencNEA ARMS DISCH ARGtE P RES S u RE-5.2.5 AtR. BUBBLE ~.D246 t a4D.S o AJ Suemers eD Snwcru/eEs.
Sa 2, ~2 7*0 A M..S GNELL. AlGESUALG~
l warea.
cLene.ws rime-S. 2. G~
/112 BuBBLG~
DRAG L.oAD.S CA/
Su.MERG EO STAU.cruRES.
~
'T~oaa.s SHELL PRESSUR.G' f
L MoDEL-DATA comeAatsoNS i
N EoE-13749-oI - P F
l e
G SAH-7 1-13 -7 9
ToRu.5 SW. ELL LoAo MooEL
.t l
LDR.
SECT (o M 5' 2. 2 ToRu.s sastt Pressure MODEL
>l N plt.T s/RVOL CLEA Rt M G moDEL M A%t Mu M-DtSCHARGE PREssu.ns WATER CLE A Rt M S TIME SIRUDL.
hND T-Q aGNCHER GEQM6*7RY PLAMT u.MlQLtf GEOM ETRY PLAM T u u t Qu.G IMtTI AL ccM DITioM S E
Fog 5/RVOL AMD WETWELL 1
MCOEL OMT Pu.T L
TORu.5 SHELL Pressure TRAMSIENT i
TORu.S SHELL Low ra t Tu.Di M AL PRESSURE DisTRt Bu.T!OM f
- TORu.S 5l-lELL RADiat PRESSURE DlsTRisaT7aA1 A\\R Bu.BBLE PRE 5suRE, RAD 16s, VERTi CAL Pos tTi o tJ 5' z. 5' AIR BUBBLE DRAG LOADS eM i
SuSMERGED STRu c. Tut 2 Es MODEL DESCRI PTlow) - DATA c.oMPAR\\ sons e
NEDE-11 e7 8 - F san-e 1-t 3 "14
';fx.
+
- ,' 7
$/RVDL PsE FLOO D MODE L.
LD R.
SECTt Old
- 5. 2. 3 S/RVDL R.EFLooD TRA M St GM T MOO 6 L iMPuT
- 5/RVOL AND T-QuEMCHER 6EOMETRy
- PLAMT u.Mt QLLE-IMtTtAL couDITroM S MoDEL OMTPu T S/RVDL REFLOOD tM ATER.
TRANSlE M T-
- D Et AY TIME Fo t2.
REFLoo D IM PuT~
TO $h\\/DL C.LEARL M G MODEL (5.z. t )
Fog.
SuasEqugMT s/av AcTuA.rtoN AMALyses l
MoDEL DESCRtPTloM - DATA C.oMPARtsoN 5 l
NE D E.- 7.3 87 8 - P I
SAH-9 2.-13 -7 1 f
.. ~ i..
t
METHo0 OLOGY LDR s'R\\/
NRC, APPROVAL-
~
CLEARt MG Mo0EL I
- s RVDL P
- NEDE - 2.3749 -o t -
MOO 6L S RVOL REFLooD NEOG -23890-P PRESSURE ~
M0DE L ToRu.5 SHELL e
N EDE - 2.t 87 6 - P Mo0GLS STRivTu.RE S/RV Su.8m6RGED
- t4EDO-2.\\47(
ucc - et+7z -P e
~
- NEDS-24 589 -P l
i S AM to
- 2. 7 9
5/RV LOAD DEFtMITION SCH EM E
~
PLANT u.M tou.E AW ALYSt S Appuc.ATIOM GtAt OE
(,Pu A A4)
LOAD DEFlu t Ti o N REPOR.T
(, LOR)
Naco-2.1888 y
S/R.V AN ALY Tic.AL soosts/c.caga.rea.
PR.o4R AMS PLAN T u.MtQu.E/'
Lt N E u.M t q u.E S/ P.V LOADS e
eTJK 1-13-79
~
,0() AAG
.48^#
3/df A0A0 E ///14 5 A 7 / 6 g C,9SES
.f S
sff y
M9 ssqe3
- N s.s
- r c3 42 c A35*
4A 4 42.
MV
$RV SSA SEA + EQ SSA + RV SEA + R V + EQ 00A + RV DBA + EQ + RV 18A ISA + EQ ISA + RV ISA + R V
- EQ d
tytmT m tuftm3 CD.
CD.CN CD.
CD.CN PS CO.
P5 CD.OI CM CM CM TYPE OF EARTHQUAKE O
5 0l 5 0
5 0
5 0
5 0
5 0
COP'8tMATION NUPegER 1
2 3
4 5
6 7
8 9
10 11 12 13 14 15 16 17 18 19 20 21 LOA 05 norest(2) n I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
tarthquake (Q
I I
I I
I I
I I
I I
I I
I I
CV Olscharge
$RV I
I I
I I
I I
I I
I I
I I
I I
LOCA Thermal T
I I
I I
I I
I I
I I
I I
I I
I I
I I
g LOCA Reactions R
I I
I I
I I
I I
I I
I I
I I
I I
I I
3 LOCA Quest-Static P
I I
I I
I I
I I
I I
I I
I I
I I
I I
4 Pressure LDCA Pool $well P,$
I I
I LOCA Condensation P
I I
I I
I I
I I
I gg Oscillation LOCA Chugging P
I I
I I
I I
I I
I g
0nt.
5/AY t's}lll 8 d.
G C/1 3/ W C /=e d /MWy
/Y p /a n f is eg ojpped', as/M fnm s NOW '
4 T7M 1- 0-71
Su.en t rT ED To MRC S'RV REPORTS k
- S/RVDL CLEARtMG MODEL
- Neos-13749 -o t - P 3
e S RV OL REFLooD M,006L NEDG - 2.3Bef6-P PRESSu.RE~
MODE L e
ToRu.s SHELL Nepe - 2.i s7 8-P s/R.V su.smeRGEo Srauc.TuRE MoOGL5 N ED O - 2.\\4'li
/
E~ 1.I +7 z -P NED
~
~ HEDf-2.4 5 89.- P sjnv resr ae.pos:.rs T-hu1NCHER TEST RGPQRT
.- MourlCEll 0
-NEDE-118(a4-P
'/4. GcALE T-QU GNCHGA TEST RCPQtt.T
-prog-1.9 5+9-P
~,~
~
^
-m sw
[g&D MGg:
UNITED STATES
[, ],,,,j NUCLEAR REGULATORY COMMISSION WASM NG TON, D. C. 20555
%f,. s,f JUL 161973 Generic Task No. A-39 MEMORANDUM FOR:
S. H. Hanauer, Director, Unresolved Safety Issues Program, NRR R. P. Denise, Acting Assistant Director for Reactor Safety, DSS FRDM:
T. M. Su, A-39 Task Manager, Containment Systems Branch, DSS
SUBJECT:
SUMMARY
OF MEETING HELD ON APRIL 4,1979 WITH REPRESENTATIVES OF THE GENERAL ELECTRIC COMPANY TO DISCUSS SRV METHODOLOGY On April 4,1979, a meeting was held in Bethesda, Maryland with representatives of the General Electric Company. The purpose of the meeting was to discuss the methodology for predicting bubble phasing during multiple valve discharges for all Mark III containments where the GE designed cross quencher device is used in the safety / relief valve discharge line.
An attendance list and a copy of the meeting handouts are enclosed.
Backaround In April 1978, the General Electric Company submitted an Interim Containment Loads Report, Mark III Containment (22A4365). Attachment M to the report provides an outline of the methodology fo= detennining multiple safety / relief valves bubble-phasing. Since tnen a series of discussions had been held between GE, the staff and their consultants. Following.these discussions, GE had gathered all staff concerns and provided justifications for each concern.
GE, therefore, requested the meeting to discuss these justifications.
Summary 1.
The bubble frequency distribution curve was generated on the basis of 132 data points obtained from tests at reactor pressures ranging from 150 to 1000 psia. Since the wide range of initial testing condition will affect the bubble frequency distribution, we requested that GE generate a bubble frequency curve based on initial testing pressure close to rated reactor pressure.
In response to this request, GE presented the results of their study, which was based on selected reactor pressure and initial pool l
temperature. This selection criterion reduced the number of data points from the original 132 to 38. Analyses based on these selected data points resulted in a standard deviation of 1.7 Hz instead of 2.3 Hz as the origi.ial curve indicated. The mean also changes from 8.1 Hz to 8.9 Hz. Based on the J
e
'1(11
'l
[g -
^/ f I
/ ',
- 3og, m.
s
. JJL 1 6 1373 results of this study, GE proposed a standard deviation of 1.7 Hz and a mean of 8.1 Hz as the design values. Note that the selected mean is based on all data because it results in a higher confidence level.
2.
The results of the study also confirmed that ifne air volume is the most dominant parameter for determining bubble frequency; the other parameters such as SRV opening time, line air temperature and submergence have no statistically significant effect on bubble frequency.
i 3.
GE will include the Caorso test results in their final analysis for '
predictions of bubble phasing during multiple valve actuations.
The preliminary analysis indicates that the current methodology predicts conservative results when compared with the Caorso data.
4.
The staff and their consultants concluded that the general approach for predicting multiple valve bubble phasing is valid. We will require.
however, that GE include the following in the final analysis:
a.
Effect of pool temperature on bubble frequ'ency; b.
Sensitivity study of standard deviation of bubble frequency distribution and its effect on SRV loads; c.
Effect of pool boundaries on bubble frequency; and d.
Structural and equipment re'sponse for determining the design case.
T. M. Su, A-39 Task Manager Containment Systems Branch Division of Systems Safety
Enclosures:
As Stated Distribution:
Central File R. Martin NRR Reading
- 0. Ross CSS Reading I&E (3)
H. Denton NRC p0R R. Mattson Local p0R F. Schroeder S. Varca R.Fraley,ACRS(16)
W. Butler R. DeYoung J. Kudrick D. Vassallo T. Su D. Skovholt R. Denise t
l ad
~ _. _ _ _ _ _ _
e Mark III SRV Meeting April 4,1979 Nam Organization T. J. Su NRC/CS8/ DSS P. Huber MIT/BNL Forrest Hatch GE F. Reuter GE P. Stancavage GE I. Uppal GE L. Sobon GE R. Patel Bechtel R. A. Hill GE t
P. Moskal Bechtel R. Beck Bechtel C. Tung BNL e
J. A. Kudrick NRC/CS8/ DSS I
a D
e e
e e
e
l 1
e e
I o
e MLTIMI QUENDER ETTES SIM%RY f
t e
i P. P. ST#rAyrf GEEAL BICTRIC QPP#ff l
QNAINNT ENGIERING t
k l
i l
t e
i l
i i
PPS-1 1
W/9 4
j
--M Jmw s
e e+
l VALIDDATABASE i
O FREQLUCY DISTRIBlFION FRai IN - PIM DATA
- FUl.L SCALE IN M BWRS
- 410LE P#E OF EXPECTED CDNDITIONS i
9 DISTRIBUTION C0 WIRED AT TYPICAL COEITIONS i
2
- RJLL REACTOR PRESSURE
- tGERATE POOL TEPERATURE l
- FIRST ACTUATION
~
i O
C40RSOTESTVALIDATESDISTRIBUTION i
- FPEr1 G CY IS RANDCN l
j
- STRCE SIGNN.S 5 TO 10 He j
i i
I t
e i
E-2 W4m 4
. _ _..._._. ~.
-,.__,_..,_.__.~_.,_r.-
_____7_s
FUJID - SiRUCTURE ItBKTION 8
GLOBAL tMtus SMALL
- TORUS K0EL SIGS to FSI IF MINOR DIVmICMESS LESS THAN 600
- LICENSEE PLMS MVE D/T < 600
- MARK II/III PLMS HAW D/T < 300 i
0 LOCAL EFFECTS S*ALL
- TPMSDUCERS ON CONCRETE At0 SHEL SIG SAT FREUCIES
- NATURAL FREDUENCIES'0F LOCAL STRUCTURES KlCH HIGHER THAN BUBBlf i
FREDUE?CIES i
i t
e i
O 4
.PPS-3 wyn e
e 6
l 4
f.
p.y m
CA0RSO DATA 0
PRESSURE #R.ITUDES BOUNDED BY PREDICTION i
- EAN VA.UES PREDICTED TASURED FIRST ACTUATION
+9.1/-6.5 4.3/-2.8 SUBSFAENT ACTUATION
+15.9/-9.2
+7.2/-4,5
- DESIGN VALVES PREDICTED (90/90)
NASURED(MAX)
FIRST ACTUATION
+12.8/-8.1
+5.0/-4.3 NUENT ACTUATION
+29.6/-11.6 4.0/-5.7 9
TIE #0 DIST#E ATT91]ATION MORE RAPID THAN PREDICTED s
BUBBLE FRE0LENCY IS P#0m
- CYCLE TO CYCE
- TEST TO TEST INTEGRAL SPECIALDENSITY
- VALVE TO VALVE I
I i
PPS - 4 4/VI9
_._.___.__,___.__.__.___..._.,_._.______________._,______..__.m_
BELEPHASIE e
SMALL INTEPETION Dutut.u
- BIELES SEPARATED BY 5 DIA
- IR! BEE ON.Y 10%
- NO EFFECT ON INITIAL CYCLES S
S%LL INTEPETION C0tFIRED BY TEST
- 01)LTIPLE VALVE PRESSlJRE B00EED BY SIELE VALVE PRESSURE
- P1NICELLO
- CA0RSO
- ttJLTIPLE VALVE WAVEF0fMS MIXED 9
PPS-5 W /9 A
we s..
n-+
m y
(DNCl.USIONS l
8 DATA PASE IS VALID RJLL R#E OF (DDITIONS l
t C0 WIRED AT TYPICAL OPERATING STATE VERIFIED BY CA0RSO TESTS I
4 FUJID STRUCTURE INTERACTION IS SM.L C4.CBAL EFFECT UNim)RTANT D/T < 600 i
I.CCAL tttti,IS E N i
8 b N NIM
- l PRESSWE ATUTl0ES l
TIE #0 DISTANE ATTBlJATION R#DE FREDUENCIES 8
0 i
i I
e 4
4 r
0 PPS-6 4/4/79 j
\\
l L_._
l 4
3 4
i I
[
i i
l MULTIPLE 1
QUENCHER t
1 METHODS i
i NRC QUESTIONS
.I O
I. S. UPPAL CONTAINMENT ENGINEERING l
APRIL 4, 1979 1
h
?
I
[
i i
I e
4 i
/
QUESTION 1:
DISCUSS THE STOCHASTIC NATURE OF BUBBLE FREQUENCY
RESPONSE
BUBBLE FREQUENCY IS RANDOM DUE TO V.ARIATIONS IN o
INITIAL CONDITIONS LINE TEMPERATURE WATER LEVEL STEAM CONTENT o
DYNAMIC PROCESS TAYLOR INSTABILITY BUBBLE FORMATION EXPERIMENTAL EVIDENCE SHOWS FREQUENCY VARIATION EVEN WITH SIMILAR INITIAL CONDITIONS 15U-1 4/4/79 u
m
~
SELECT DATA BASE L
QUESTION 2:
THE DATA WERE OBTAINED AT REACTOR PRESSURES RANGING FROM 150 TO 1000 PSIA.
HOW DID THIS AFFECT THE QBF OBTAINED? A SEPARATE QBF INCLUDING THOSE DATA OBTAINED AT FULL REACTOR PRESSURE SHOULD BE GENERATED.
RESPONSE
A SUBSET OF THE DATA BASE WAS SELECTED FROM THE IN-PLANT-TESTS WHICH MOST CLOSELY REPRESENTS EXPECTED CONDITIONS FOR AN ALL VALVE ACTUATION EVENT.
THE SELECTION CRITERIA ARE:
FIRST ACTUATION, SINGLE VALVE P0OL TEMPERATURE BELOW 110*F REACTOR PRESSURE AB0VE 950 PSIG STANDARD NUMBER OF MEAN DEVIATION TESTS (Hz)
(Hz)
ALL DATA MEETING CRITERIA 38 8.9,
1.7 USE TWO ALL DATA 132 8.1 2.3 SIGNIFICANT FIGURES DESIGN USE N/A 8.1 1.7 THEREISREASONABLEAGREEMENTBETWEENSELECTEDDAIAAND ALL DATA.
0%&
~
\\
i o MEAN FREQUENCY OF DATA MEETING CRITERIA IS WITHIN l
TEN PERCENT OF MEAN FREQUENCY OF ALL DATA.
MEAN FREQUENCY (8.1 Hz) USED IS BASED ON ALL DATA BECAUSE 132 TESTS GIVE HIGHER CONFIDENCE.
d a
STANDARD DEVIATION OF 1.7 Hz IS SELECTED FOR DESIGN USE 1.7 Hz IS LESS THAN THE STANDARD DEVIATION BASED ON ALL DATA AND HENCE CONSERVATIVE 1.7 Hz IS BASED ON TEST DATA WITH FULL REACTOR PRESSURE ND OTHER CONDITIONS THAT ARE TYPICAL OF BWR PLANTS i
e I
e ISU-3 4/4/79 i
e
~
FREQUENCY DEPENDENCE QUESTION 3:
WAS THE QBF OBTAINED UNDER CONDITIONS OF CONSTANT AIR LINE VOLUME, VALVE OPENING TIMES, SRV LINE LENGTH AND HYDRAULIC RESISTANCE, PIPE TEMPERATURE ETC.?
SEPARATE CREDIT IS TAKEN FOR THE POSSIBLE MITIGATING EFFECTS OF SOME OF THESE VARIABLES:
IT IS, THEREFORE, IMPORTANT T0 l
ESTABLISH THAT THEIR INFLUENCE NOT ALREADY IMPLICIT IN THE QBF.
RESPONSE
THE FREQUENCY DISTRIBUTION WAS OBTAINED FROM TWO IN-PLANT TESTS WITH COMPARABLE CONDITIONS DATA MEETING s
PARAMETER PLANT A PLANT B CRITERIA LINE AIR VOLUME (PT3) 50 47 47-50 SUBMERGENCE (FT) 15 13 13-15 POOL TEMPERATURE (*F)85-104 92-169 85-110 EC) 150-1000 220-1475 275-1475 REACTOR PRESSURE (PSI) 13-1066 120-1030 950-1066 VARIOUS PARAMETERS WERE INVESTIGATED VIA REGRESSION ANALYSIS FOR THEIR EFFECT ON FREQUENCY.
PRESSURE RISE RATE, VALVE SETPOINT, VALVE OPENING TIME, AND BUBBLE FREQUENCY ARE INDEPENDENT VARIABLES.
1 i
,RESULTS OF REGRESSION ANALYSIS BUBBLE FREQUENCY IS NOT INFLUENCED BY VALVE OPENING o
TIME OR REACTOR PRESSURE o
RESULTS SHOW NO TREND R'EGARDING EFFECT OF OTHER PARAMETERS ON BUBBLE FREQUENCY o
LINE AIR TEMPERATUR,E WAS FAIRLY CONSTANT 3
I o
LINE AIR VOLUME WAS (47-50 FT ) CONSTANT EFFECT OF VOLUME NOT IN DATA BASE
'i EFFECT OF VOLUME IS SIGNIFICANT
- VOLUME l'S INCLUDED SEPARATELY i
4,
e 6
l i
1 ISU-5 4/4/79
+
m em CA0RSO DATA QUESTION 4:
THE QBF SHOULD BE UPDATED ON THE BASIS OF THE CA0RSO TEST AS S00N AS THESE ARE AVAILABLE.
RESPONSE
o CAORS0 SVA DATA IS IN PRELIMINARY FORM, ANALYSIS SHOWS CAORSO MEASURED FREQUENCY IS RANDOM.
CA0RSO MEASURED FREQUENCY IS WITHIN RANGE OF PREDICTED FREQUENCIES.
o TWO WAYS TO OBTAIN BUBBLE FREQUENCY p, TOTAL OF BUBBLE OSCILLATION CYCLES TOTAL TIME OF CYCLES POWER SPECTRAL DENSITY PLOT (SEE FIGURES 6 & 7) o PSD USED TO DEVELOP CURRENT QBF o PSDPLOTSHOWSTHATMORETHANONEFREQUENCYHhS SIGNIFICANT ENERGY o
PSD FREQUENCY SPREAD IS MUCH LARGER THAN TIME AVERAGED FREQUENCY (TABLE 4) o OVERALL CA0RSO PRESSURE IS A FACTOR OF 2 BELOW MEAN PREDICTED ISU-6 4/4/79 a
e
,3
p h}K i
T1 o
IN TABLE 4, PREDICTED MEAN FREQUENCY.
=8.1xlhi=7.4Hz o MEASURED FREQUENCY IS ONE PREDOMINANT FREQUENCY PER TEST TABLE 4 MEASURED PREDICTED MEAN (Hz) 6.05 7.4 S.D. (Hz) t.41 1.6 LOWER BOUND (Hz) 5.3 4.6 UPPER B0UND (Hz) 6.8 10.9 o
TABLE 3 SHOWS THAT EACH CYCLE HAS ITS OWN FREQUENCY o CA0RSO MVA MODEL/ DATA COMPARISON UNDERWAY.
THIS COMPARISON WILL SHOW THAT SRVA IS CONSERVATIVE BY A LARGE MARGIN ISU-7 4/4/79 e
FREQUENCY DATA BASE QUESTION 5:
HOW WELL IS THE PROBABILITY DISTRIBUTION KNOWN?
WHAT IS THE DATA BASE?
RESPONSE
THE FREQUENCY PROBABILITY IS BASED ON 132 IN-PLANT QUENCHER TESTS AT 1TK) LICENSEE FACILITIES.
THESE TESTS PROVIDE THE DATA FOR:
- MEAN 8.1 HERTZ STANDARD DEVIATION 1.7 HERTZ UPPER BOUND 12 HERTZ LOWER BOUND 5 HERTZ A CHI - SQUARE TEST SHOWS THAT THE NORMAL DISTRIBUTION IS APPROPRIATE AT 5% LEVEL OF SIGNIFICANCE.
t ISU-8 4/4/79 I
m.
QUESTION 6:
THE POSSIBILITY THAT THE TEST DATA (ESPECIALLY THOSE RELATING TO BUBBLE FREQUENCY) WERE AFFECTED BY FLUID STRUCTURE INTERACTION SHOULD BE ADDRESSED.
RESPONSE
GLOBAL FSI EFFECTS o APPLIED SRV FORCING FUNCTION TO A COUPLED FLUID STRUCTURE MODEL OF MARK I TORUS o MODEL SHOWED FOR MINOR TORUS DIAMETER TO SHELL THICKNESS (D/T) RATIO UP TO 600, FSI IS NEGLIGIBLE o
PLANTS A & B E$600.
CONCLUSION OF REFERENCED STUDY IS APPLICABLE o
HENCE GLOBAL FSI EFFECTS IN PLANTS A & B ARE NEGLIGIBLE LOCAL FSI EFFECTS o
LOCAL STRUCTURAL INFLUENCE IS SHOWN BY TRANSDUCERS AT DIFFERENT LOCATIONS c
o FIGURE 10 SHOWS THE LOCATIONS OF PRESSURE TRANSDUCERS DA 13, 14, & 16 o
FIGURE 11 SHOWS THAT EACH HA'S A BUBBLE FREQUENCY s8 Hz o THEREFORE, THIS IS A GENUINE BUBBLE FREQUENCY, NOT AFFECTED BY LOCAL FSI I
FREQUENCY DATA FROM TYPICAL PLANTS o ADDITIONALLY PLANTS A & B ARE TYPICAL OF MARK II AND III PLANTS.
FSI EFFECTS (NEGLIGIBLE) PRESENT IN A & B WILL ALSO BE PRESENT IN MARK II & III.
(SEE FIGURE 12)
OD!9
.,__,,,y
._--..-w
-- - -- - - -~'
---w- - - - - - - - - - " " * - * - - - ~ - - ' ' ~ - - "
6 t\\
l y,q FREQUENCY DEPENDENCE QUESTION 7:
TO WHAT EXTENT IS THE DISTRIBUTION PLANT -
DOES IT DEPEND ON LINE LENGTH?
LINE TEMPERATURE?
FIRST OR SUBSEQUENT ACTUATION?
RESPONSE
o DESIGN DISTRIBUTION VARIES WITH LINE VOLUME ONLY o DISTRIBUTION IS BROAD EN0 UGH TO ACCOUNT FOR ALL OTHER VARIABLES o MARK II AND III PLANTS HAVE SAME BASIC GE0 METRY AS FAR AS SRV LOADS ARE CONCERNED.
THIER FREQUENCY IS NOT EXPECTED TO VARY DUE TO PLANT GE0 METRY o
FREQUENCY IS A FUNCTION OF LINE VOLUME o
LINE TEMPERATURE EFFECT IS IMPLICIT IN THE DATA BASE o
CA0RSO SECOND ACTUATION FREQUENCY HIGHER THAN FIRST ACTUATION.
EARLIER TEST SHOWED THAT FREQUENCY DOES NOT DEPEND ON FIRST OR SUBSEQUENT ACTUATION e
e p
=
- e_
,y
~
QUESTION 8:
WHAT ARE THE DATA BASES FOR THE PROBABILITY DISTRIBUTIONS OF VALVE SETPOINT AND VALVE OPENING TIME?
RESPONSE
VALVE SETPOINT o FOR TESTABLE INSTRUMENTATION, SD - 2 PSI APPLIED TO BOTH CROSBY & DIKKERS VALVES o
FOR NON-TESTABLE INSTRUMENTATION SD = 8 PSI IS USED BASED ON 24 SHOP TESTS o FOR TARGET ROCK VALVES SD - 5.9 PSI BASED ON 77 SHOP TESTS.
FIGURE 13 SHOWS THE DISTRIBUTION IS CLOSE TO NORMAL.
VALVE OPENING TIME o FOR CROSBY VALVES SD =.0092 Sec BASED ON 408 TESTS o
FOR DIKKERS VALVES D =.0097 BASED ON 50 TESTS.
THEREFORE, STANDARD DEVIATION FOR BOTH CROSBY AND DIKKERS IS SPECIFIED AS 0.009 SEC.
o FOR TARGET ROCK VALVES 187 DATA POINTS GAVE SD =.013 SECONDS
~
150-11 4/4/79
AIR VOLUME ON FREQUENCY QUESTION 9:
IT IS PROPOSED THAT THE QBF DISTRIBUTION BE SHIFTED TO ACCOUNT FOR SRV LINE VOLUMES THAT DIFFER FROM THE 50 FT' LINES USED TO OBTAIN THE DATA.
THE PROPOSED ADJUSTMENT IS BASED ON A SIMPLISTIC AND POSSIBLY NON-CONSERVATIVE ANALYSIS WHICH NEGLECTS THE KNOWN DEPENDENCE OF BUBBLE PRESSURE ON AIR LINE VOLUME.
A REASSESSMENT OF THIS ASSUMPTION IS REQUIRED.
~
RESPONSE
THE RELATIONSHIP GOVERNING FREQUENCY AND AIR VOLUME IS, FROM RAYLEIGH'S EQUATION 3
FREQUENCYo6/AIRVOLUME THE ENTIRE RELATIONSHIP HAS BEEN EXPERIMENTALLY CONFIRMED IN 1/4-SCALE T-QUENCHER TESTS FOR LINE VOLUMES RANGING FROM 24 FT' TO 99 FT' (SEE FIGURE 14)
G 9
e e
6
(
30
-*- PREDicTet Q-4 ME ASURED s
20
\\
N 1
3 o
~
f 7
%,.= Q 10 e
I I
0 0
4.5 t.0 t.5 u
PtPt As.4 vCLunst Tigure ld Effect of ?ipe Air.':lume on Frequency for Inicial Pipe and
'.'eeve'l Pressure ci 3.7 psia e
15 -
Isu-13 4/4/79 1
QUESTION 10:
ONE CENTRAL ASSUMPTION OF THE PROPOSED METHODOLGY IS THAT BUBBLES OSCILLATING SIMULTANE0USLY IN THE P0OL DO NOT INTERACT IN ANY WAY THAT WOULD TEND TO INCREASE THE LOAD AMPLITUDES TO MODIFY THE PHASE DIFFERENCE BETWEEN THE BUBBLE OSCILLATIONS OR TO HARMONIZE THE OSCILLATION FREQUENCIES.
THE BASIS FOR THIS ASSUMPTION IS ONE OF OUR MAIN PRESENT CONCERNS.
RESPONSE
o THEORY PREDICTS LITTLE INTERACTION QUENCHERS ARE 14 FEET APART BUBBLEDIAMETERS ARE 2.7 FEET MEAN SPACING IS 5 DIAMETERS EFFECT (1/R) IS 10%
o TESTS SHOW NO INTERACTION MONTICELLO TESTS SHOW SVA SHELL PRESSURES GENERALLY LESS THAN MVA SHELL PRESSURES CA0RSO 2 AND 3 VALVE MVA PRESSURES LESS THAN SVA PRESSURES CA0RSO 4 VALVE MVA PRESSURES $20% HIGHER THAN SVA BUT WAVEFORM INDICATES NON-PHASED BUBBLES CA0RSO 8 VALVE MVA PRESSURES $6% HIGHER THAN SVA AND WAVEFORM INDICATES NON-PHASED BUBBLES l
9
!!-);
,-r-n.m.,
n=,
o
_-,-,,_,,_-_.c
i 1
i o MULTIPLE QUENCHER METHOD ALLOWS APPR0XIMATE PHASING EXAMPLE IN ATTACHMENT M i
3 BUBBLES AT 0.123 SEC AND 9.3 Hz i
2 BUBBLES AT 0.128 SEC AND 8.6 Hz 1
4 4
i j
4 I
4 i
)
r i
I I
,i i
i I
i t
i j
l ISU-15 4/4/79
_., _ _ _ _ _ _.,. _ _. -.. ~. _., _ -, _.. _ _ _ _, _ _. _,
l 22A.~. 3 6 5 Rev. 2
..A-3 Valve :;o.
I'.*OT ( s e c )
Valve No.
IVOT (sec)
Valve No.
IV'OT (sec)
.067 7
0.067 13 0.056 1
0.J69 3
0.051 14 0.061 2
0.065 9
0.062 15 0.056 a
2.059 10 0.065 16 0.065 5
0.062 11 0.033 17 0.057 6
0.03S 12 0.057 18 0.071 19 0.069 I
- c:a :::: a =esa value of 0.057 see is included in the above nu=bers. Adding
- hese values to che group T calcula:ed in Step 3 and nor=alizing to have the g
firs: bubble arrive a: zero time resul:s in :he following bubble arrival ti=es:
Arrival Ti=e Arrival Time Arrival Tise Valve No.
(sec)
Valve No.
(sec)
Valve No.
(see) 1 0.125 7
LO.125 3 13 0.243 2
0.256 8
0.238
-w 14 0.127 3
Fu. w i 9
0.120
- 15 0.243 4
- 0. 2;F 10 0.0 16 0.124 5
LO.1221 11 0.246 17 0.245 6
0.225 12 0.116 4 18 0.129 19 0.256 M.3 1732*.I F?JEQtC*CIIS Subb.e f requencies for individual quenchers are randomly selected from a random i
nu=ter genera:or code using the, dis:ribucion shown in Figure M2-4 Typical rand:: bubble frequency values for the 19 quenchers are:
l i
Valve No.
Frequency (Hz)
Valve No.
Frecuency (Hz) 1 6.56 11 7.22 2
9.77 12 5.39 3
6.
4 5.01~
13 5.68 14 w8.60 5
G 15 9.86 6
6.38 16 7.04 7'
G 17 11.08 3
9.10 18
- 8.68 9
7.92 19 8.52 10 11.14 i
- CTI
For :his exa:ple, all lines are considered as uniform in length and fre-quencies are rando=1y selected fro = one Quencher Subble Frequency (Q3F) dis:ribu: ion curve (Figure M2-4). In this exa ple =ean = 8.23 H: and
= 1.30 H:..*i:n nonuniform line. lengths, Subsec:1on X3.2.1 is used to g
develo; unique Q3F distribucion curves from which a frecuency is rando-d y selec:ed,for each line.
101673 ISU-16 4/4/79 i
a
~ _.
c.
EFFECT OF LINE ON BUBBLE ARRIVAL TIME QUESTION 11:
AT PRESENT NO CREDIT IS TAKEN FOR POSSIBLE
]
CHANGES IN BUBBLE ARRIVAL TIME DUE TO DIFFERENCES IN SRV LINE LENGTH OR HYDRAULIC RESISTANCE.
THESE FACTORS COULD, HOWEVER, TEND TO NEGATE THE FAVORABLE EFFECT OF DIFFERENT VALUE SETPOINTS.
THEY SHOULD BE ADDRESSED.
RESPONSE
o LINE AIR VOLUME AFFECTS BUBBLE ARRIVAL TIME SOMEWHAT BY CHANGING THE AIR AND WATER CLEARING TIMES 3
o FOR EXAMPLE, AN INCREASE IN AIR VOLUME FROM 57 FT TO 88 FT' CAUSES A DELAY OF 56 MSEC IN AIR CLEARING TIME o
IN INDIVIDUAL PLANTS, AIR VOLUME USUALLY LIES WITHIN 25%
o RESULT OF INCLUDING LINE VOLUME FOR A TYPICAL PLANT 5
(57 FT 88 FT')
THE AVERAGE FOURIER SPECTRA REMAINED ESSENTIALLY UNCHANGED (FIGURES 15-20) 9 e
I O
~
c.
FORCING FUNCTION SELECTION QUESTION 12:
DISCUSS IN GREATER DETAIL HOW FREQUENCY DEPENDENT " BOUNDING" FORCING FUNCTION IS DEDUCED FROM THE COMPUTED 59 MONTE CARLO SIMULATIONS.
RESPONSE
o THE FORCING FUNCTION IS BOUNDING IN THE SENSE THAT 59 TRIALS GIVE 95% CONFIDENCE THAT THE PEAK BOUNDS 95%
OF ALL EXPECTED RESULTS.
FEWER TRIALS WILL GIVE LESS CONFIDENCE o
SIGNIFICANT FREQUENCY RANGE IS DIVIDED INTO 3 FREQUENCY AND LARGEST SPECTRAL VALUE WITHIN EACH FREQUEllCY INTERVAL IS SELECTED FOR DETERMINATION OF EQUIPMENT RESPONSE o ADDITIONAL CONFIDENCE IN THE BOUNDING CHARACTERISTIC 0F THE FORCING FUNCTION IS PROVIDED BY:
HIGHEST BUBBLE PRESSURE GIVEN BY ANY DISCHARGE LINE IS USED FOR ALL. DISCHARGE LINES.
THE MAXIMUM DESIGN BUBBLE IS EXTREMELY CONSERVATIVE.
THE PREDICTED BUBBLE PRESSURE FOR CA0RSO SVA IS 15.1/-8.9 PSI AS COMPARED TO MEASURED MAXIMUM PEAK BUBBLE PRESSURE OF 5.0/-4.5 Pst.
THE DISTANCE ATTENUATION OF GESSAR/DFFR BOUNDS CA0RSO DISTANCE ATTENUATION Unh 1
^
GESSAR/DFFR TIME ATTENUATION B0UNDS TIME ATTENUATION OBSERVED AT CA0RSO (SEE FIGURE 21) o FIGURE 21 ALSO SHOWS THAT CA0RSO DATA IS BOUNDED BY PREDICTIONS BY A LARGE MARGIN o
FOR LINEAR SYSTEMS, HIGHEST INPUT FOR EACH FREQUENCY GIVES HIGHEST 0UTPUT 4
J l
I
- i ISU-19' 4/4/79 d
o
s l
dh l
R'$
CONFIRMATORY WORK QUESTION 13:
IT IS OUR OPINION THAT COMPARISONS BETWEEN PREDICTIONS BASED ON THIS METHODOLGY AND MVA IN-PLANT LOAD DATA ALREADY AVAILABLE AND TO BE OBTAINED FROM THE CA0RSO TESTS IS AN ESSENTIAL PART OF THE REVIEW PROCESS.
BOUNDING FORCING FUNCTION PREDICTIONS FOR DISCHARGE CONDITIONS CORRESPONDING PRECISELY TO THOSE ACTUALLY TESTED (TWO, THREE, FOUR AND EIGHT VALVE DISCHARGES AT CA0RSO, ALL MULTIPLE VALVE TESTS AT BRUNSBUTTEL) SHOULD BE GENERATED AND COMPARED WITH THE IN-PLANT LOAD DATA.
IT IS RECOGNIZED THAT THE IN-PLANT DATA CONSISTS OF DISCRETE PRESSURE MEASUREMENTS A BEST ESTIMATE OF THE INTEGRATED LOAD MUST NEVERTHELESS BE OBTAINED.
WE EMPHASIZE THE NEED FOR COMPARISONS BETWEEN TEST DATA AND PREDICTIONS BASED ON THE-PROBALISTIC PROCEDURE AS APPLIED TO DISCHARGE CONDITIONS IDENTICAL TO THOSE TESTED.
RESPONSE
o A QUICK LOOK INDICATES THAT CA0RSO MVA DATA IS BOUNDED BY PREDICTIONS BY A LARGE MARGIN o MODEL/ DATA COMPARISON UNDERWAY u
.=
~%
UNITED STATES
~
8 NUCLEAR REGULATORY COMMISSION g
E WASMNGTON, D. C. 20655 o
/
AUG 8 1979 MEMORANDUM FOR:
A. Thadani, Task Manager, A-9, Reactor Systems Branch, r33 FROM:
T. M. Su, Task Manager, A-39 Containment Systems Branch, DSS
SUBJECT:
PRELIMINARY QUESTION LIST FOR GE REPORT ON STWS Per your request on August 1,1979, I have prepared the preliminary question list for the GE report titled " Assessment of BWR Mitigation of ATWS, May 1979." A copy of the list is enclosed. Please note that the enclosed question list is intended only for discussion purposes during the coming meeting with GE on August 10. It does not, however, represent the result of our final review on that report. Based on the current review schedule, we expect to complete our review by early September,1979. We will provide you the results of our evaluation by that time.
T. M. Su, A-39 Task Manager Containment Systems Branch Division of Systems Safety
Enclosure:
As Stated cc: S..Hanauer W. Butler J. Kudrick L. Ruth 4
i e
_9 9
/
"m g
e c99
- i. ~ ~
o,o rp L
l Preliminary Question List Relating to. Sections 5.2.2 and 5.2.3 of GE Report on ATWS 1.
Provide a detailed description of the methodology used to calculate.ATWS SRV loads.
Information should include the following:
a.
All assumptions used in the methodology; s
b.
The initial conditions such as SRV line temperature, SRV line volume, suppression pool temperature, and suppression chamber p.ressurel c.
Data base and the associated interpretation and justification for the use of the data for ATWS conditions.
2.
It is noted that the Caorso test results were used directly for the ATWS SRV loads for both Mark II and Mark III containments. It should be recognized, however, the Caorso plant has its own plant unique conditions related to SRV design such as SRV line volume, submergence, pool area per quencher etc.
In addition, the Caorso primary systems and containment conditions during the tests were not near the ATWS conditions. Therefore, we believe that the Caorio test results should not be used without justification. Furthennore, the complete information on the Caorso tests will not be made available for staff review until the end of this year. A generic approach for the use of the test results has not been proposed either by GE or the Mark II owners group. Based on this status, until the above information is provided, the current statisti.a1 approach described in NEDO-21061-P (Mark II Containment Dynamic Forcing Functions) or NEDO-11314-08 for Mark III' containment should be used. Provide the calculated results for Mark II and III ATWS SRV loads.
3 d{
- 3. ' Provide justification for establishing the temperature difference of 14*F between local and bulk pool t'e$perature.
It should be noted that the Monticello T-Quencher tests on February,1978 were performed under certain specific operational conditions and plant unique geometry. The test results should not be used in a generic sense. Justification is required for the application of this data base for Mark Iidesign.
4.
The service water temperature and initial pool temperature of 75'F were l
used in the calculation of pool temperature response to ATWS events. The i
rational used to justify these temperatures, i.e., the T-quencher has the capability to condense steam up to saturated local temperature, has not been substantiated. Therefore, additional infonnation is required to justify the use of temperature below design levels.
5.
Provide the Caorso data base which was used to establish the ll'F temperature difference for Mark II and III containments.
\\
=
.-