ML20052D684
| ML20052D684 | |
| Person / Time | |
|---|---|
| Site: | Zion File:ZionSolutions icon.png |
| Issue date: | 02/22/1982 |
| From: | Zudans Z FRANKLIN INSTITUTE |
| To: | Okrent D Advisory Committee on Reactor Safeguards |
| References | |
| ACRS-CT-1422, NUDOCS 8205070098 | |
| Download: ML20052D684 (4) | |
Text
--
C / -hD nklin Research Center E
~
A DMalon of The Frankhn insexute I
February 22, 1922 Dr. David Okrent Professor, Sebool of Enginsering and N
E l V E ti Applied science 80Wrt 5531 Seelter usil ggM%rIIm untu Salwersity of California g3 See Angeles, CA 90024
'Re:
Bion Probabilistic Safety Study
_ h gfgIg{
Bear 3r. Skrenti k
As requested by Dr. Griesmeyer, I undertook something like a mial review of the subject study.
)
h approach I used was as follows:
i 1.
Ouick look at the structure and contents of i
the entire PRA study.
2.
Closer look at the derivation of containment ultimate capability.
3.* Ouick look at arv failure modes.
i 4.
Quick look at structure and component fragility.
1.
h overall PRA rs to he very comprehensive and structured la a aanmer al to different classes of readers.
Stocutive summary clearly defines the results of the study, sumanary report gives to diestseien.for the reader to find his way through the a bety of the report.
h differences between WhBa-1400 and Sion PRA are clearly stated, one of the most important ease being the assignment of levels of eenfidmace associated with verises fregeencies of the exceedamos.
Al not-reviewed at th3s time la detail, it appears that the physion1 na associated with the progression of acelear peuer plant accidents are addressed la e semprehensive meaner (rege 32.5-8).*
1 "For esemple, the conclusion that ladependent of degree of initial fragmentatica, mechanisme exist that drive sore debri systen toward ult'. mate coolable goemetry, is similar to that given bs; T.S. theotamous and M. Saito, in put-91-140, June, 1991, Purdue Salversity.
Dr. ::.w r. c.t : w pgin* *
=
c.n.,.,,Gy
\\
CT-1422 POR The Sev$nmin Feen Robey.PhAnd@Ns.k.1912 Q1514et-130 TWx-nr m 't. '
r r o
3
!l Dr. David Okrpnt February 22 IR; 2.
Containment response is analyzed for 21 entry states the containment event tree leading to 11 exist states, each associate-tc with specific release estegory.
e Exit to entry states are coupled by the containment matrix (Table 12.5-1, Page 21. 5-10, vol. 1).
Bach of the entries in this matrix represents a conditional probabi-i 11ty that a given entry state will result in an exit state correspond-ing to a release category.
I' the contaiammat integrity are effected byIt a.s shown that the event behavior and a
1.
BCs pressure at the time of RPV failure.
2.
5 venting from BCS prior to RPV failure.
2 3.
timing of fual melting.
e 4.
Operation of contaiamsnt sprays and fans.
t Failure of Item 4 or ecstainment bypass (failure of two RER isolation valves) lead to a sirect release.
(2 )
and containment analysis is performed by use of MARCH code. Integrated core I recall from another class 9 Accident discussion at an Acns meeting, As time scale MARCN analysis may have eiseable variability.
of.that one know how signi Because be in'aitigating the response. ficant the structural heat sinkt will Primar 149' psia. y containsaant ultimate capacity was determined to be It was onloulated by sar Se analysis was supposed to oover: gent & Lundy (Appendix 4.4.1).
1).oontainment structure, 2) trations. 3) ' rate of loads
- 4) ancertainty bounds, 5) miens.
failure daly the first two items are dealt with,in some detail.
The heaeter Cent 4& ament to $a the shape of 3'6" wall cylinder with (ese F foendaties s. 2), 3'8" thiek aballow domed roof based on 9' flat She erlindriest portion is prestressed by a
-tenoien The
,0ystem semeisting of berimental (boop) and vertical o25gm elab is soeventlenelly reintereed.has a three-way post-tansioning system.
Founda-
,eestainment is li6ed with 1/4* teelded steel plate.The inside of the entire A cylindrical, of 31 feet is leonted at tho'senter of the base mat.staster pit with a Sag internal y. r m espebility and subcontreeted Cas! to derive the strectural empebility of the'egeipment batch.
The analysis used seterials proportses derived fras the mill testa and from the 90-day senereto ey11ador tests.
teeGen'as assened, se$ metAs the failure mode,1% strain in hoop
, Speelfied by the PS sentt open flow area of le to 15 in2, as b
I f,
k v.
1
Dr. David Okrent February 22, H6 Analysis consists of hand calculations and of an axisywwtrit finite element computer calculation.
It is strictly determinist 3 :
and the conclusion that confidence level of 95L is associated v'.ta the amiculated containment ' internal pressure capability is not sepperted (presumably based on knowledge of materials property statistics, ese Item 4 below for further comment).
D e hand calculations is one of the type p*
is the yield stress, A is the corz sapending material s wre oy f.andon, rebar) cross section area and R is tne radius of the cylinder.
His assupues that all concrete has cracked and that liner has not lost its leaktightness.
The calculate. result is the 134.4 poig, a reasonable member.
Finite element model is quite complete.
It allows for gradual concrete cracki through the well.
Included in the sedel are:
I sat, sylinder done.
Account is given for non-linear behavior et steel (tendens and rebar) and of concrete (by cracking).
We resulta confism internal pressure at 14 boop tendon strain obtained g
hy hand calculation.
i Mis calculated result and the observed mode i
of emetile failure is favorable compared to 1:14 scale model test t
{
et a thin galled post-tensiemed concrete containment.
In addition to this failtes mode, transwerse shear stresses are examined at sarises locations.
Of these I believe the location of cylinder detail for potential leak path formation.eteachamat to foundation met wo i
Sie is because at It etrain in the cylinder, the radial displacement in the i
syl r to of the order of 3 to 7 inches greater than.that at the mat.' Also however, pross, thermal espansion differencesthermal effects are written off a (between the cylindu well and the met) asy cause severe transverse aboar and potentially form leakage path.
Caatainment well around the penetrations (including the equip-meet batch) is significantly thicker than in the cylinder, hence strenger for the railure' mode postulated.
Cas! analysis of the egelpment hatch is well organised and provides credible number of 134 yeag for ultimate buckling capability of the hatch dosie, effeut of the projected ultimate internal pressure on the hatchne heumdery is not somsidered in the equipment hatch analyeis, tial for leak path at the hatch boundary has to be evaluated with poter-realistic boundary conditions.
o a
,u S
Br. David aburent February 22, 1962 Osetaisosat structural espebility analysis does not addrear-the erstems med staaetures atteshed to the eentainanent wa11.
Ytu ee seven M rodial 51sp3eesment of the enstainment wall will itely egese same 41stsees et the umil ettactunent of theae.
Se 149 w to be e reemenable mester for the cylinder, getential es.e sensegeance of some penetration to this pessoare has not been included in the study a
m 3.
Stetter peessure vessel dellure dynamics are addressed in am 3.>44.
De treateemt ypears to be reasonable and the SMe et am 1estriesent mese3e sa the betten head is the likely
.es.e.s erspiut, e, cessess sofety,e2at.d sw.ctu,.s end 23.7-2) use deem-imad by Structural mechanics les 7.g.2).
W W taken was to identify the 6 heets geselesetion and %facter of eefety for each itest guttested.c h as8ety fener une them eseamposed into various M&heting elemente, eash of teseh one emelysed for its variabi-38 ties est e eestataties.
M ese insiehilities and sneertainties I
- W them em to get te veeleh!!!ty and uncertainty of the Settee'end of the esselesetion sepability for the ites.
i e m to be Setiemel and resulting families of 1
sugges eheeld he esettble.
I believe the goestion of peessere altaaste esposity shoeld have been emb)ected te e sta13er teostaest, leading to feedly of eurves seek as II.b3. eether then en e$agle curve. seek as rig.11.5-5, has se jestified confidense level associated with it.
Very truly yours, g
WV WU Seemas ees Senter Vice Presider.t ee Dr. J. Sreiemeyer, 3CRS 6
_--n a
J
.