ML20236F793

Structural Adequacy of Waterford 3 Basemat, Technical Evaluation Rept

ML20236F793
Person / Time
Site:	Waterford
Issue date:	09/30/1987
From:	Costantino C, Chris Miller, Reich M BROOKHAVEN NATIONAL LABORATORY
To:	Office of Nuclear Reactor Regulation
Shared Package
ML20236F740	List: ML20236F736 ML20236F765 ML20236F793
References
CON-FIN-A-3837 A-3837-3-9-87, NUDOCS 8711020296
Download: ML20236F793 (21)

Text

,..

Q

.t~

.4 l % N!'

sh y

e 3

,+

v:.

s%%

1 l +

,e a.

.w tq,.

~;

A-3837-3-9/87,,

• r +

'2

.+

am

, %m..

uy

~

TECHNICAL. REPORT ~

PS -

y f

g.

J

.w cg%g ;'

.....J s7 w $,

+

l~

0,;,

M g.;,

.. 'u.

w g

n.. g?s. m. p@..c

e q %~

4

.u n

p%p v

e

,n.

,.4 h.i

.N.

[.M g t

TECHNICADEVALUAIIONREPORTE.gg[AvI

.g.. m. e

.cs STRddORAEADEQUACYOFWATERFORD3'BASEMATc.A W

, m;.m..+:w p s w. 94mMikW.,

M D

C.

..a 97, 4

. ; msma w. s, n

MM@,,,:,v y g. y /j 4,,;.j,M$$. N.h5,$U ke dM}$' : p M

[

3 n ~n~b;qgg;~g,, gp.

g.

-y K.

r C. A. # MILLER,(m.x., M OSTANTINo, A.J. PHILIJPACOSg p.g h

e

.ggN; L.

=.&

~ $p,

W w AND-M. REICH

1. ' %qqw 4g,pk$g

.:% p

..a+ sgR y%.

^

327 n$$y#;f'Qpg g

' x w m w/

..dL

,g a

<y <,..

o c

.n gp j

3-p,.,

[.'h.Y hMy Y' QMk d.

. 'rch

}

(f y:s' g i &', Q g Q Q y g. ; 7 % g Q.

,,7 g_.W

'^

'p m pg~d(H y ni.pn(g r % w @4 g

9W 4

pp q

-ewghge rsaradut ;

nA M

d&

.-. q q-gsdm.TaticTjiF AL7 ANALYSIS 1DIVISIONa 5-c j m.e:genw ww yq:C y1 j

9,-

DEPARTMENT QFc UCLl!ww NJ.ERGYg=ra~n 7.g.ggy g

~

~

ARjW mw:,w g. w.1 a

M VEN,pw.61yLAbORATORYh;g.~ ~

J NAT10 A y'

a '

J 1 Nm 4p.,;,.

00K

.x y

.o.mn ~gggcq

'y g.

r PTO,IEW ;ORK g11973 a ; w x

f y

e

e. m

?

q.;g aggy yy p:y. %.

p e.

q e

m

r,

A y, ?y'a m.l.o;w s.9 w

a

+

W.S@p% %y9.p ] Wh.c_:f 3 e

p V.' A,w%. 339,.;

f

.- +

.y

.m A

l:

"%)$A1..;

~

~Q~

M y.n L..,x.

jn

~.

L

... 3 _.+

.w a-

~

q.

.m_.

e,p',,w Y g# u^ l T nn M;.l:N.:%h;%;L e

>w uw. y

' n' "

X S$n

. 9 'Mf e'..

> F "m Hi 4

g;..g

v.. s -

.g.g

EPTEMBER 987. g.g. gig,. WA

. ~

f.

..ge iMg.@

. 4 y

w.

wy p

mmq t

v

- 3

-g 4

i.

up my.

J gn. d ag N qw.,

u x

w 3

pe.d;y.7(:.:

WM A;. 2, 4

"W

'c

$fwr

.g^~

/G m

a, 3

f4

@C PAR LEAR.

ERG EN NATIONAL.LABORATORF y i:,J

jb -

47

q

'qp e;y<

U[i d

q. t 'M g%a e

.j

+

,;i i

d fy-< ' i M.

E

'h

~:q &(%

g.t 3<

s1 a

1,.

i r'

. s, Q

, ;?

};

3..

4 6y "l. l

-.g s

.l[

n ru

? q Q~.

Ed

.opated for the 4.N fj'$ h,j, j s:

LJ j}h' & p f Office o,fMdeles(,,,.,,., '

y y g,

~l E,;7 u,.

f S Contract No

>bACO2-t V:

,1 N).w v y@,,ppwA.

s

.1 3:,

)

J g

x y

s..

h!N[ kMINME'

[

i 87110L0296L071027>

A

';g M cCK;O p_

fR p

gg gj wm m;r memyanng gn av.

m.x

..~

c.m+. e..

, m ~m n.'r 7,n,,,.%., u _ m, eno m,,I,.s 4.s ' m

.c.-

%.:-. \\.(p$.rg 1

m'[ m 3 ' k-y I

' e i

m.g'i " '"i i j f 6., m,' a n,

.<,.hp: d:, f

.sme c m',, 4

[, -[{ N. 4, ' P ;

3 ' A' O f..-

Y ' * '  ;

j l 4 =,g 3

b

' I

'f_} '.i 4, 1i y]

'I' 1

b y

/

5[n.,,

t

+

i,

{ -

5,1, M

,r

m. 'd 4. L.j +,f.c4 i -

m,7,*

t, m. x,,.

r i

2

+- w e 3, 4,

e.. '

i y

a v.,,

, ; 3:

, 3 ;;

f.

s,

~t 3 ;il 1

1 4

2 e

.' i s

s s

o

,.+e r y. g

-+..:

.M< 0 y

',;s.. nr 4,g, 1, e: i

-.,m.>" u 't. u.,;.. A, f w',

S '

y s

. !l' 4'i

,e z ( -,, -'

E

' t,,e t-. g(l.

-} ')',

.g u 2,

't-

..., a r--

y r-2 I.

s

{pek,i d

f' I

. $ /,

=

u-Jt'-

g, pj

",'f

,. c y, j u ;.

D','

,1, f._;

c.,

p,,.,.,,

i t

7 t

.t:-# y -} 7,;

'p..

- -(}, 3 /

• 1. (,.,

y,,I

.p"

"-(l.,,

. n 'n

+

t '.l4

.r f

p. [q(-

lj><

2 i

.t.

t, c..] a og

,p5...,,

r.

i cp' qq ;.

=. a %.

3 f

..t:

a ~

p

,i s

i i w'

'.y ;; f_ b y

,p n.

s p.

. jj

• ng

, c--

, e

?

x w.

r s.

+

.~-

ag+u

.r 1;r; j 3

s 9

f

,ve%f t) ; f i

i O'

j-

.ms

Q:

,, q%

s t

s..

.f v

f f

a -..

3

,L=

%_4 s,-

,t, i'.,'

g J

N

'T 1 Technic a l1 Ev. a lu'a t. ion.s Re po r t.a.,

4W

m..

m.

d,.'

rJ.

l' Oi i........

.t c.

~L 3

>t

,e Struct!ura11 Adequacy.of Watkrfordc31Basemat)

' I=k f..

, ~.

J g 'b

'l) - 1

-(

Y, l [}

' f 3;

3i s

-i,s=.

,i.

,s s

ow

..o o

D g ^:

.i.f

, 3 rv

' l.

Cl M f,,-

p

.. 37 g t;

e

- L.

t ll1l Jf;:

3 1

so w y,.

v p

.e-v, y 4... r3

1..

s c.

.b~J.

~".

+,,

.:n

.v.~...

m.,.

V.,.

s

a. os.

e.

g 2

+

11.

_v,

m 1

,r/

' ik,{.,1

^k-'

(

i

{w 1 4 r

3 1.,

..,,m

.cr

. C.A.-Miller l C.J. Costantino / A'.Jr Philippacopoulos,,

r.

T r

t y

n "J

m l

n.

->f',s~

O l ' ~. e

+g.1 g'-

?

V 3.l a

w1 g,..

fIf I,]

• E'

. : L U, v} -(

A f.

,[-l'~

j :i i. -

t

.:w.

4r c

< t: (,,

,a r

,,g t

y x;;,

i i,,

O

?

l

, w,t, _

n T.

t 1 &",, A' ' f b

"J

Struc tura1($naly,, sis: Divi, ion y

s e

, g BrookhavenLNat1onaliLab'oratoryi..

f M "J me n'

4

.:Dep'artmentisfzNuclear Energy; M

We.

~

i s.

W ; 1,% d, f7 n.,

LUpton, New T,ork h11973L

!.':) Y H,. lg y

s

- 0

.s

._ y }

['y

'^i i v

.,.c 3.,

, s

,a

f i u

v. x..:< w k?.,

'r i

n

.,t-y

.c

(

j'^-1f

'[,

i t

g i

4,,,,

r y

f l

r - [' q.

-J.

J.,

is i

-September 71987:

y

..t -

o

)

.t 4

'j=.

j,'y a

f

'{

n g

I L

I u.

s

.. q n, q

'.8 v.

't

.L

.Ig I

I-r i

+

f

_ j jg m

t

d. g, i 9

b

._t

m

.q d

._' f q 1

s 1

[,,

i A

-. ; f.

g V.

m, m

'Ap l;',

t

, y 1

1 n

3 y

4 a

+

-i

,, +

jw W 's.

?

n h4b

=_

, 3 o.s f !'

,'f

- i FIN A-3837 4

g.'

j y

I

' d5

. i p

i g4, h..

1 l:,I%..;

j

,.o 4 T' :

,' 7

^

'I

,n

(

3

'l' (

., 7:g.gd[

.iJ

}

('

'-' 1 t

4

, d r.5 p s,

s..

Ig. k ':. j '. s'./, #. 'e

(:f.,

".f

). '

b

-f CJ a..,~,

= -

A

r..rir l k,, a, 3

g

.a 1

wn., :

-<,<~4*

.= -,',

' ?yi :._.j _.

_g

]

n

_(r" r.~

d

^

i:.c g,,. a c 5....,

s 4

i t

s s"

,a k r

1 >., a.

m i.

,.CA._ _ mE_1.1_ L __. w,., _ _.._ ; J.,

3

y _ _ _..

." ji

[tl,

{ P.'

91~

c

.:R' 0

LA;

'y, 3

4

<T i

Q

"( ;

g~y

• u, f--

4 y;

7:1

h. 4 t

t

}[?

) f.

c,<

e 1

4-LQ s

g TABLE OF CONTENTSi g

.ic g

Pag'e

<b S UMMARY ;..... ~...................................... '. i...... ;........ Il.

l s

h

'1.

INTRODUCTION ? AND B ACKGROUND,........ i.............. i....... ;. c../'? 1

.M

.........i..........v..........<.....-

(10 L1.1-Basernat Cracking......

1.1.1' Pre-licensing ' Crackj Hi s to ry ' '.....'....'..... [...... i..... '... - si" H

~

iL 121 1.1.2.l Pre-licensing [ Crack.' Studies -.............................

. g 135 1.1.3.

Post-licensing Crack History.........~.............;......:

4.....t

- 3' 4

1.2-Confirmatory Analyse's; e

I Ba semat Surveillance Program o..... (......... c..........'...L } 5j 1.3:

........'s.....4.........i.........:

'5

' l.4 - Scope'of BNL; Review; M

.'.....i.......- L6i

2., DYNAMIC EFFECTS l 0F.. SOIL / WATER. LOADINGS (.

-2.1 Introduction :.......................'.....i.

........v.....

16c 2.2.Results of' Analysis ~

-61 2.3.' Findings and' Conclusions,...............................'.;

7J E

l0 m

3.

DYNAMIC COUPLING ~0F THE REACTOR. BUILDING AND BASEMATJ:

17 3.1 Introduction

..........<.........i........................; ; 7-3.2 Results of Analysis'.................;.....................-

8 i8' 3.3 Findings and Conclusions'-

.......1 8?

4.

ARTIFICAL BOUNDARY CONSTRAINTS ' IN FINITE - ELEMENT MODEL-D.... i.:

8-4.1 Introduction '.....................................

4.2 Results of~ Analysis.

....................................4.

9<

~

4.3 Findings and Conclusions

........<;................'.......; ;9 5.

FINENESS OF FIN 11?); ELEMENT MODEL:.......................i...... -10 5.1 Introduction.

................................'.......'....'c10s 10--

5.2 Results of f Analysis 5.3 Findings and-Conclusions

...........4........i........~...

jl0

.......v....-111J 6.

BASEMAT ~ CONSTRUCTION SEQUENCE ANALYSIS 6.1 Introduction

........................'...........'......... i llu a

6.2 Results of.. Analysis.

.........'.....4....................

lit 6.3 Findings and Conclusions

.............................'...it12-

.- a n

t-r b

d I

I

.?1 s

C)

u y

s n "l

(

---__.-_.a.

__'\\.

____}]__

y..

i 1

F k

.t..

3

(

l g

a

~

, 1

.,.1 t.m i

i

'1

.t..

t

.}

c.

.'.i

+

TALBEJ0F CONTENTS

I

. l

.-COMBINATION OFICONSTRUCTION' SEQUENCE AND/.0THER LOAD

<j

~

7.

Jio

.a 7.1 Introduction 1................~..'.....'.........'.......'.......-

121

' 7. 2 ' Re s ult s - o f ' Ana ly s i s

.... '........ ~.......................... : 13 7 '.3.: c Findings. and. Conclusions

...........;......................: 131 a

wm!

B ASEMAT ' SU RVEILLANCE ' PROGRAM........... '............... u........ ' 13 '

8.

t, 8.1 ' I n t r o d u c t i o n :...'...... '............. '........................ 2 13.

1 49 1

9 8.2( Scope-of. Review

..........................................1 8.3.-

Finding. and : Conclusions :.'.................................. :: f 14'

. t 4

9.

SUMMARY

..........................'.'.....'..'................e...,11~-

161-

/)

RE F E RE N CE S -........... '........... '......,..................... ~...... :

,, s

.1' 3

.{ -

t I 1

i nP yi s

/

i

'l

• g w

i

.1 3

1

-i

1..g

' ?

i

?

+

[.i {

'D Ii s

-r

a n

'i"

j, 7,-

g.-

..)

t d

n

SUMMARY

s

'10 Thiefreport addres'sesi th'e' confirmatory ' analysis !performeid Iby'. Louisiana-Power and Light.(LP&L) 'in ~ compliance, with condition 2.C.17.,,of lthe operating r

-license'for the plant. LInladditio'ni to the confirmatory analysis the utiilityi wa's; required to ins'titutefa basemat; surveillance ~! program to assureLthat mat.-

b performance is as. expected., This item-is'also. addressed.inithis TER.:

~

. As stated.in::the. report, BNL concludes:that.the'.resultslof-the

. confirmatory analysis carried out by'LP&L generally providel. quantitative data i

in support of.the adequacyfof the cracked mat. KMoreover, the;basemat surveillance program which:is: planned'fortthe lifefof"the' plant.waslfoundjtol beacceptable."In'viewof'the.shorttimespanfthatLthefmonitoringEprogramihas been in place (1985), it is recommended'thatJdata obtained'during? routine.-

surveillance be reviewed.by NRC until'lsome;overall trendsfcan,be established.

~

-1 1

O

'l T

i i

__ - -- - =.

---_2 2=. _ =L 0

.. a i a

R 4

o-r s g7 7 E

zy

' i, P N

, p; j' i "[

g.,

g, m

d i,-

l

.Y-

, TECHNICAL' EVALUATION. REPORT'

? !!

'~

[

STRUCTURAL ADEQUACY OF WATERFORD 3 BASEMAT-

,O n

' 1'. 0 INTRODUCTION AND BACKGROUNDS Fu

,..[ :

g Condition.2.C.17 Ofithe operatingilicense;for the-Waterford Steam Electric-Station,; Unit.3doperated by, Louisiana:PowerLand(LightjCompany.;(LP&L,.~

3 the licensee) (contains;requirementsLthat:the licenseelperformMconfirmatoryJ analyses' associate',with:the' structural adequacy of theiplant!sibas.ematfand?

d monitor its1 performance.: Thesefrequirements were; imposed on;the'licenseeLa_sLs

result of cracking of the basemat. j Althoughi several". questions. regardingi the J

.cra'cks remained atjth'eLtimelof licensing,aitiwas" judged?by the'NRC;staffgandL Lits. consultants, based.:on;analys,es,lreviewp andinondestructive 't'esO res'ults",

thatthesecracks;wouldhaveminimal(significanceon;thestructuralja'dequacy}

'"J of the.basemat.'jThe;objectivefof thefconfirmatory;analypes program 1was(to @f

. 'j, '

provide quantitative: data - substantiating : that' judgementi;The1 purpose. of) th'e surveillance program was to:det'ect changes!thatjcould" affect thialbasemats-Rl,; j

'q performance. The results 'of (the confirmatory analysesf are documented in

._

^

Reference 1, and the. surveillance prograafinLReference12L This IER containst the'Brookhaven National Laboratory.BNL re' view and' evaluation'of "hese-reports._:Before discussing each of'theiconfirmatory analysesfand.th'e surveillance program;(in Sections 2:through.8),'some background of the is.uesi j

s a

are prer,ented, a) 0 1.1 -

Basemat cracking 1

1.1.1 Pre-licensing Crack History All structures of the Waterford 3; nuclear plant.islan'd'are founded'onia; common basemat. 'The basemat isL380 feet;(in the N-S! direction) by'267; feet

,3 (in the E-W direction) in plan and is -12 feetJthick. : : Flexural: reinforcing steel was placed near - both the top and bottom:of; the mat,-lwith the - topfonlyc.

1 lightly reinforced (to satisfy temperaturedand; shrinkage: requirements as well' d

as some localized bending moments).

Somesvertical; shear reinforcement was!

provided in high shear areas. The. bottom of the mat? is.. located 'about -. 641.f ee t below grade, with the foundation soils Lbeing primarily. silts) and ' clays., The

. roundwater ' table is loca ted close to the. ground; surface..

g The mat was constructed in' blocks approximately 60. feet' square.

~

Construction proceeded.by first completing'several: blocks:in'an.E-W. strip.near

'q

~

the' center of the mat,and then proceeding:out'tojthe north and; south ends o'f=

t 1

the mat..This construction sequence,{ combined'with-the relatively 1arge(

i settlements, resulted in bending moments producing tensile 7 stresses acting on the top of.the mat-along E-W sections.-

Crackstwere first observed.in some'ofE he vertical? walls ~in: March 1976.

t These cracks lwere attributed to. shrinkage. effects'andiwereinot considered?

further. LThe first cracks in the basemat itself were observedlin July (1977 when the" area's was cleared.of' construction' debris.. These cracks'are; oriented 2 '-

1 u,

[] h.

j p

8.

9

i

.f 1

j in the E-W direction and are located inside the shield building area. The l

licensee determined that these cracks have no-structural significance and repaired them with an epoxy gel prior to placement of fill concrete inside the ring wall. A crack map was prepared documenting the extent of the cracks l

inside the ring wall structure.

l Additional basemat cracks were observe'd in May 1983 when the area outside' l

the shield building was cleared of construction debris. These newly observed j

~

cracks, together with those found in July 1977, formed a system of seven major

{

cracks that appeared to span across the entire mat.

New crack maps were pre-l pared documenting visible cracks in all accessible areas of the basemat in 1983.

At that time, questions were raised as to the -impact 'such cracks might

]

have on the performance of the mat.

l 1.1.2 Pre-licensing Crack Studies l

i Several actions were taken to resolve the questions concerning the '

significance of the cracks. The following studies were undertaken in the. time

)

l period from mid 1983 to early 1985:

l l

a.

A non-destructive test program of portions -of the basemat and walls was conducted by Muenow & Associates (Reference 11) for the licensee in 3

1984. A pulse-echo sonic technique was used to search selected areas of.

the mat and walls with the objective of defining the geometry of the

.l cracks. The results of this investigation demonstrated that the major j

cracks in the mat were vertically. oriented and' deep. That is, in many 3

locations, cracks extended from the top surface of the nat to the bottom j

reinforcement. The cracks widths were found to be small (less than 7 the cracks were I robably induced by i

mils). This data indicates that bending moments which put the top of the mat in tension.

As indicated above, such bending moments existed during construction.- The test results also eliminated the concern that the cracks might be shear-induced diagonal tension cracks. Such cracks would be orientedL45 degrees to the j

vertical rather than the vertical orient'ation which was ' observed.

The i

results of the investigation of the wall cracks indicated:that the'y were shallow and had no apparent-relationship to the cracks in basemat.

i b.

The licensee and their engineers undertook studies (References 3-6) to interpret the data collected by Muenow and to ' evaluate the impact the cracks might have on the performance.of the plant. It was concluded that the cracks likely occurred during the construction period and that the I

cracked mat could safely transmit the bending moments and shears which would be expected for operating and' accident load combinations.. It was.

also demonstrated that the shear stiffness of the mat would not I

be adversely affected by the presence of the cracks.-

c.

A series of. studies (References 7-9) were undertaken by BNL, acting as consultants to the NRC staff, as part of the overall technical review of the licensee work. These studies included analytical and experimental

' t

<c l

work having the objective of providing independent verification of the studies being performed by the licensee as described in (b).above.

The I

-BNL studies generally supported the conclusions developed by the licensee f

but raised several concerns whose resolution required additional con-firmatory analyses. These confirmatory analyses became the basis for the licensee condition imposed on the licensee.

1.1.3 Post-licensing Crack History During a scheduled crack monitoring survey in June 1986, additional cracks were observed' which did not appear' on the 1977 'nor the 1983 crack j

The licensee informed NRC of these findings.and' initiated a study to l

maps.

assess the implications of the "new" cracks. The assessment concluded that l

these. cracks were present'during the perioud of_the' previous mapping efforts q

l but were missed either because they were located in portions of the mat which were inaccessible-at times of mapping, were. judged to be not structurally f

significant, or were so fine as to be. difficult to classify as a crack..

NRC/

j BNL visited the site and inspected these cracks in July.1986.

NRC/BNL. staff agreed with the licensee assessment that these cracks were of the same, type as the old cracks and probably were previously present. This conclusion was-based on the similarity in appearance of the "new" cracks as. compared to the I

old (rounded edges, weathered), and the fact that most of the "new" cracks were located in areas which were likely inaccessible during the construction

)

I period when the prior crack mapping studies were performed.

1.2 Confirmatory Analyses As discussed above, the NRC staff concluded that the basemat could safely perform its function. There were five areas, however, where it was concluded that additional confirmation was required. These five areas formed the basis for the confirmatory analyses requirement contained in the license. : The following is a summary of these five tasks. Sections 2 through 6 will discuss l

BNL's review of each of ther,e tasks in detail.- The need for an additional task to combine construction sequence loads with the other. loads was identified during the BNL review. The objectives.and. details of this-task are-presented in Section 7.

9 Dynamic Effects of Lateral Soil / Water Loadings a.

The nuclear plant island structure (NPIS).is partially embedded in the soil and therefore is subject to a confining pressure. The combined effects of the soil and water pressure acting on the walls of;the.NPIS (water table is near the surface)' result in a compressive stress in the i

basemat. This compressive stress tends to close any cracks in the mat and

~

to enhance the shear transfer capability across the cracks. The objective of this confirmatory analysis is to evaluate the extent to which the confining pressure may be reduced during a seismic event. The review of this task is contained in Section'2.

3 __

m

.n.

,y,

v',

[ q + f, ml p

' f, M, ;

y~

Y

/

%.. }l? ' W qf.2,

i

.l.

\\

_..g A

y

.c

> : ~

?

lE y

,lt 7

e_

'Jr y:

s

. t

-a-

[7 t;

gg Dynamic Coupling ofl the Reactor' Building'and Basema't b.-

riginally!cAlculat$dibasEdfonftheassumptioni

[

^

The,seismicIresponse,was.

thatLthe:basematelwasJrigidC 'Thelobjectivefof:thisttask was:tolevaluateC

the. extent.to which the basemat' flexibility may[effectfaither;thei predicted iseismic l loads ' action-on theistructuresior onL thel floort re sponse: ; * '

t

spectra.~Thereviewlofithistaskris:containedinSection[37 g

J L

o ArtificiklBoundAryConstraint'sNniFiniteElementMod L

L

~c.-

w I

.U'

[)

The.basemateiseismic istress analysih was' performed on ?a) finiteielement" s

model off the NPIS.: :Vertidalisprings: we're; pisced[underfthe matisimulating 3 94 thel vertical l soil? pressure? acting?on;the bottosiof1the' mat.V The; that'were frequired]tok.. '

horizontal'soilipressures? acting'onJtheMalls;7 balance theihorizontal? inertia 1Llo'ading's, werelde'termined from?lateralt pressure computations'and inputidirectly tof thel finite ; element!L model'.i Artificialhorizontal" constraints:verethereforefrequired"sfo;thatithef'9 finite' element model stiffnessimatrix'wo'ld not be, singular. JTh'eU I

u

. objective of this task was:to:demonstratexthatithe'selestionofLthel location' of (these s artificial bestraints did ~ not ihave" a ise'rious ' hf fedtfin;.m.

the computed bh'semat' shears and'bendingfmoments. ;The reviewlofLthisitas @

a t'

is contained in-Section:4..

7 d '. Fineness = of Finite Element'MeshD 4

m.

The finite element model of the basematicontained many { areas where' large:

shear and/or moment. gradients Loccurred acrossua"few elements. :. TheD

~

i objective 'of this. task.was-to. evaluate.whether;affiner mesh in these areas would alter the results. This task is reviewed:in?section: 5.-

Basemat' Construction Sequence Analysis.

e.

G

.g

~... 2 >i.

n The cause_ of basemat_ cracking was: attributed to flexural; stresses "in.

,0 the mat that, occurred.during construction.: This! conclusion was based on!

j qualitative judgments combined ~with_ analytical'resultsJobtained from-q relatively' simple models. The_ objective (of?thisitaskjwas to perform a-

~

detailed analysis of= stresses in theibasematiduringithe}consteddtifon1

~

i

. periodLso thatLthe cause-of the' cracking could be based.on!more quantitative' data.- This1 task' is reviewed in 'Section' 6).

~

i Prior.to performing these. confirmatory analyses,Nthe'ilicensee; performed.a q"

• new baseline"!analysisifor the operating;andfaccidentfloudslacting onLthe'. '

basemat. This analysis;used a finite element-model of; the]NPISLwhichi y"

incorporated changes in modeling:of(the'basemat,[modeling'of"the structuresL

'a and! applied loading which havel occurredisince; the}: original analyses'.gThis wass

' done so L that; modeling inconsistencies - were J eliminated,in Lthe baseline (casei

against which the' confirmatoryjanal'ysis_ cases;couldjbescompsredt !ItDhaalbeen? l

,y s-s-

s b

j 4

i

'j t

4 Q

s 3s. g.x,.

3

,...o n

p:

2 i

L,__

W-'--

L

~

L._:._ _ _. _ _ _. _ _

_r

_;_y___

- l.

r

l l

.a l

demonstrated in Reference 1 that the new baseline shears and moments lie

.)

within the design allowables.

It has been further shown that the variations f-the new -baseline results f rom the original ~ solutions are localized and readily f

explained by consideration of the modeling. changes. Therefore, a comparison i

of the confirmatory analysis results with results from the new baseline are j

deemed to be acceptable.

l

'1.3 Basemat Surveillance Program In addition to performing the confirmatory analysis described above, j

license condition number.20.17 required the licensee to institute a basemat j

surveillance program. The objective of this program is to-ensure that the mat performance is as expected. The c:ats response. parameters being monitored include: settlement, groundwater chemistry, groundwater' level, and crack j

In each of the areas, except for groundwater level, upset values'are j

growth.

defined at which point the licenseo must inform NRC and perform.an. engineering j

evaluation of the data. The surveillance program was.first defined.by LP6L' and accepted by NRC early in 1985. The finding of "new" cracks in' June 1986-led to a modification of the crack monitoring phase'of the program so that the a

results would be less subjective. A completr ;escription and evaluation of

.l.!

the surveillance program is given in Section 8 of. this report.

l j

1.4 Scope of BNL Review i

The scope of the confirmatory analyses and the basemat surveillance f

j program grew largely out of concerns expressed by the NRC staff and. BNL during j

the pre-licensing audits of work performed by the licensee.. The specific approaches taken to perform the analyses were proposed by the licensee in-

.j January 1985. Several meetings were held between the licensee and NRC/BNL j

over the following several months until the plan was accepted in March 1985.

j About ten audits were held at the EBASCO offices from 1985 to present. The i

progress on each of the confirmatory analysis tasks was: reviewed 'and comments

.l' outlining potential problem areas were given to the. licensee. During the same time period, two site visits were made 'with the objective of inspecting the l

cracks and reviewing the implementation of the surveillance program.

l 4-The licensee issued a report of the confirmatory analyses in.0ctober 1986 i

i (Reference 10). This report was - reviewed by NRC/BNL and found 'to be deficient in several respects. As a result, the licensee issued the report (Reference i

1) which is being evaluated in this TER..It should be noted that the new.

report stands alone and does not require the older report (Reference 10) as a reference document.

One of the major concerns expressed about. Reference 10 was the moment contour plots which indicated that large areas.of the mat are in tension at the top under both operational and accident load combinations. A review of the testimony given in support of the licensee indicates that many judgments of the acceptability of the cracked basemat were partially based on the

. i

w w

'.=.

+

g, 4

p ' W*

j.

=.r f

9C g

9 i

{ff

+

i r

\\

t q

l

,a assumption that'the top,of the basemat would betin: compression under suchD1had' combinations, tThe new moment =contourfplots.seemed7to contradictlthis!

The.licenseetthen? undertook-anLadditional task lto studyjthisi k c

~

. assumption.

issue. 'The'results of this task are? documented inLthe:licenseeireport:

1:

(Refer'ence 1) and.are discuss'ed inESection17L,

~

2; l DYNAMIC EFFECTS OF -SOIL / WATER LOADINGS :

1 i,

2.1' introduction y

The.Waterford basemat is located 'aboutI64L eet{below ground surface.HThe ' + '

f L

soil / water pressures acting.on the" side wallsLproduce a'compressiveistressJ across.the'.basematlwhich tends-to: force the' cracks to'.beiclosed h While)thisi i

i compressive stress was'not essential,toftheLargument..thatfsupportedLthei

-adequacy of the. mat, it;was another conservatism that(was: considered linimaking' 1~

.the,' decision that the maticould! perform.itsi. intended funct ons LThe~effectiof) i i

seismic loadingsLonLthe sidewall;pressurefwasL.a? concern rasied,during thei

'licesning proceedings. Therefore, Tone of the confirmatoryftaskanwasitoJ, q

j consider:theeffect-ofseismic;1oadsfon7theTmagnitudeofthis} compressive w

u stress.:

2.2 LResults ofEAnalysis Atwodimensionalfinitefelement'modelofJthep1Antwasmadebytakingja; L

slice through the NPIS.. The FLUSH computer' codefwas' used. to generate ;

~

Severale solutions for:the horizontal: seismic (response:of the model'.

preliminary analyses were' performed toLatudylthe) required extentrof thel soil?

t boundaries..Since the' site soil;is rather s'ft, this? presented'a difficult:

j o

1 task. The. sof t material requires : smail elements to! transmit ' the higher.

1 frequencies offinterest.; This requirement coupled withithe. required? size ofL

^

')

the. mesh, taxes the. capability.oflthelavailabletcomputerisoftware.1 N

Average soil properties,11.5 times the averagejand? average: divided;byJ 1.5, were used to. bound the effects lof scatterxin the? measured /soillpropertiesi D

From;these numerical resultsU the1computedfstress(in the?

~

in the' computations.'.

f o r t:

9 matdueto:sesimic:loadingsweredeterming.:TheseLwereeitheraddedJ subtracted from the stresses-caused by:-the at-rest soilbloadings(to arrive:at.

U net stresses developed in'the:basemat..

a In the initial FLUSH runs forJthis;. calculation,frensileistressesuwere y

i computed in the upper seventeen feet ofl soil. adjacent tokth'e'NPIS. !Since; s' oils cannot ~ sustain these tensions, a lsecondicalculati~on was made' remo'vingT this upper seventeen' feet of. soil-from'theiproblem. >A: third calculation was:

~

also run in which this-soil'was (included. inL the models but Lthe finite Eelement:

mesh of this. soil was not connected to the NPISO 9

M g,

y f

4' I

e a

q t

t N :

cl

~.

s l

o.

' f;_

-s s

b 4

4 i

9 The maximum reduction in basemat compressive stresses in the area of the reactor building (area of significant E-W cracks) were found to be 24 pai, 18L psi, and 10 psi for the average,1.5 average, and average /1.5 soil property cases, respectively.

The static pressure-induced compressive stress in the mat is 50 psi so that the net remaining basemat compressive stresses are 26 l

psi, 32 psi, and 40 psi, respectively, for the three soil cases.

l 2.3 Findings and Conclusions There is a great deal of uncertainty in the results of this analysis.

Much of the uncertainty arises because of the soft site conditions and the.

j resulting difficulties in developing suitable FLUSH models.

During the audits of the work several questions regarding model characteristics (size, boundaries, cutoff, frequency, etc.) were raised which contribute to the uncertainty. The difficulty in obtaining a solution for the average /1.5 soil j

property case is discussed _below.

]

i There is somewhat of an anomoly in that the average soil-property data results do not lie within the two bounding cases. This may well be caused by

)

l the results for the average /1.5 case. The FLUSH Code operates by inputting a criteria motion at a control depth. The code than convolutes this motion to the bottom boundary. This " bedrock" motion is then used as input to the model including the structure. This process did not coverge for the average /1.5 case when the control depth was taken at the foundation level, as required..

l J

The control depth was takne at the ground surface, with the results corrected I

to account for the difference between the criteria and calculated spectra at the foundation level. A correction factor of 2 was used (i.e., the computed results were multiplied by 2).

The actual frequency dependent correction f actors varied between 0.7 (at 5 cps) to 3.69 at-(2.5 cps).

1 The computed results show that there is a net compressive st ress in the basemate due to both at-rest and seismic loadings.

While the-exact value is i

rather uncertain (for reasons discussed above), it is concluded that the mat likely. remains in compression under all load combinations. The data discussed 1

in Section 7 of this report indicate that the performance of the mat is l

adequate, even for this caseiyhere no net compressive stress exists. This j

issue is therefore closed.

q 3.

DYNAMIC COUPLING OF THE REACTOR BUILLING AND BASEMAT 1

3.1 Introduction i

The seismic model used for Waterford 3 was based on the assumption that the basemat was rigid. A question was raised during the licensing review as to whether there would be at,y dynamic coupling between the vertical seismic-response of the structures and the flexural response of the basemat.

One of the confirmatory analysis requirements was therefore to determine whether such coupling occurred. During the audits, this scope was expanded to include both horizontal and vertical motions.

)

g N

,a This analysis was performed,using the same finite element model as j

described in Section 2.

The modulus of the concrete basemat was varied from the actual concrete modulus 1000 times and 100 times the actual modulus for the horizontal and vertical cases,-_respectively,' simulating the rigid mat assumption. Solutions were then obtained using the horizontal and vertical-seismic inputs. ' Peak' accelerations and response spectra'were computed at the' q

structural. nodes.

3.2' Results of' Analysis

,y q

The finite element models used by the licensee were' reviewed and the-numerica1' output from the computer runs audited.

The peak structural. node..

i accelerations computed with the rigid and flexible mat moduli, were found to: be I

very close. This would indicat'e. that. the -stresses in.the structures 'would not L

be.significantly affected by mat flexibility. Horizontal and vertical..

q response _ spectra at the structural' nodes were computed for both.the rigid'and j

flexible mat assumptions.

No specific trends were noted;between response:

'j a

spectra computed with either assumption. The largest. increase;in response '

M spectral values : vere found to be _ nineteen percent when the flexible'. mat -

assumption is used. This increase, however, is less'than the amount by.which the design spectra exceeded the criteria spectra.

l 3.3 Findings and Conclusions Since the finite element models used for this. task are the same as those-used in Section 2 (dynamic soil / water loadings), the' questions noted therein apply to these calculations as well. However, since the same modellwas utilized for both the rigid.and flexible mat analysis and since the changes noted by incorporating mat flexibility into the analysis wereLfound to be

)

small, it is concluded that the influence of'the mat flexibility does not. lead

.j to responses that will exceed'the' criteria.-- This issue'is therefore considered to be closed.

j q

4.

ARTIFICAL BOUNDARY CONSTRAINTS IN FINITE ELEMENT: MODEL -

1 1

4.1-Introduction 1 1

l-i The seismic stresses induced'in the 'NPIS 'were determined f rom a large finite element model of the basemat and structures., The peak accelerations-were imposed as inertial loadings.and a static stress analysis performed. -The finite element model of the NPIS_ was supported vertically 'on soil springs.

No.

horizontal soil restraining springs-vere imposed to:the NPIS. This, of:

course, leads to an unrestrained system _so that artifical-constraints'must be applied before a numerical solution may be obtained. These artificial-restraints were applied by fixing both horizontal displacement components 5

j l

E i

L

_8_

g

['

f m

,y x

n-

,8

/

a.

s.

c a

i, m

1 along the N-S' centerline and. along the.s'outh edge of;the. basemat., 'The /

^

objective of this confirmatory analysisEwas ito" demonstrate that: stresses arising,from these artificial constraints areEsmall'.{ Thus,.the specific, y

locations of these constraints are unimportant.:

]

The confirmatory l analysis"was' performed byjrem'oving Lthe $rtificialf L.

constraints'and1 replacing them w1th a' set of two. horizontal. springs 1 attached.

L to each node in? the. basemat... A rigid.1 ink was placed from;midplanelof Lthe" mat' ch

.to the soil / mat interface. Two horizontal:-s'prings1(in: the E-W and N-S,

O 9

.f directions) were ' thenf placed f rom. this L11nk to j fixed (abdes. Thei spring

.stif fnesses were : selected to -representEtheihorizontal soil deformation thatl might ber expected..under the ' seismic -loading.; (Upper (20.8 lb/ inch )f and 3

lower 1(3.5'1b/ inch )) bound values were used for the analys.es.;1These bounds-

'l 3

a were obtained byLdividing-thelbase frictioni(setcat the soilicohesive E

t'o t3 l

3

- strength)' by! the, anticipated: horizontal deformation L (ranging: f rom l 0;5:

-inches). The STARDYNE computer program was'used to compute-seismic'bssemati

']

moments and shears.which could then belcompared:with the.. baseline <valuesiusing' D

d the' artificial constraints.

4.2 ~ Results of Analysis 4

'A comparison of the upper and lower. bound l spring'so10tions. indicat$dinol difference in basemat shearJforces and less than onetpercent-difference:.in'

.J basemat bending moments. The. moments computed by using~the 1'oweryspringf

]

constant were found to.be slightly 1arger'than;the moments;calculatediusing'

~

the higher spring constant. The results from' the' lower spring constant -

analysis were used for comparison with:the baseline analysis.

The' basemat shear forces 'were found to increase bv.noimore thanitwo s percent from the baseline analysis to that obtained usingithe distributed

~

spring restraints. The maximum increase in basemat.. bending 2 moments'was found:

j to be less than two and one half. percent.

4.3 Findings and Conclusions W The spring restraints are. representative of the: actual' support conditions, and the spring'stiffnesses used in thel analysis;representiboundingl 1

values. The increase'in basemat forces (shear and.bendingimoments)~were,found*

to increase only slightly (less than. two.'and onel half percent) whencthe worst case tof distributed springs' was used. ; These1sma11L changeslin forces ' are well

~

within the accuracy limits of.the analysis?and will notthave anTimpact oncthe'-

safety of the plant.: This. issue'is therefore considered ltc'be closed.;

.}

A.

?

1

>?

l'

~1 4

ll

\\

(,

q(

s z i i- -

hid__z. _ iii

V q

M Ag

'+y g

m.

m'-

[u N'

~

.Q y }il J

6 y.,

.,}'

go

..)'

ic q

m,

y 3 ~

7 x

u

~

.. ~ -

5.

FINENESS OF FINITE ELEMENT MODEL':

i_

.3q L

5;l ' Introduction v,

a The" characteristics of the! finite. element' mesh'.of Lthe basemat iused' in ~ the ~

d

- origina11 analyses.(1973);were influenced to some extent byc the Lsi.ne -

limitations'of:the"available; computers.;(Some compromises;were ther,efore made' in:the fineness oft he mesh usedito model the basemat'so that-the: total number t

of nodes > inlthe' modell could 'be maintained lessE than '1000.

y A review offshears,and; bending moment's:obtained:from this modellindicated- "

N1 1

~

that there!were areas.where large'shearfand/orfmoment gradients 7 occurred?

i across an element. ' The. elements ' used Lin ' the analysis; assume ; constant L aoment and linearly' varying shear across; the element ;(i.el, : con'stant:: curvature).7; Af

~

task was therefore included in theiconfirmatory? analyses"to demonstrate-thatt

y. g refinement ofithe finite element mesh would;not--result intlarge increasespinL

' basemat shear. or bending moments.'

J r

- 5.2 - Results of Ahalysis'-

i

The finite' element model of -thelbasemat was c refine'di n five ' areas rei

's i'

the: larger shear ~ and ' moment; gradients occuredb : Some 'ofi theTwall finite;:,

element;models were also required-to be changsd :to match the' changestin the -

ba'semat model. The normal' operation, W-E yseismic, S-N( seismic,. and L N-S seismic. load combinations were rerun with the frafined mesh,l using ;thelSTARDYNE' computer program. :The results showediafgeneral' improvement inLshear'and?,

y moment distributions when compared; with' the ibaselir.e analyses. '

f

\\

The basemat shear.lvalue's changed in-areas:.near walls.with ;aigeneralg (l

decrease occurring especiallylin those; elements;which are located;a. distance' 4

"d".away from a ' support,.where'"d" is the: effective-depth offthe; mat.. Code' L

acceptable procedures allow' that shear ~ does not. have to bei checked Jwithin a' d

distance "d"'of a-support. ;While thel baseline; analysis! indicated ajfew

-ej elements' where shear capacities: were exceeded, (this was : Justified by"using j ant average: shear. across' a. section comprised of -se'veral; elements),j the refinedf Q

mesh sho'wed actual shear loads. less. than 'capacityi for all elementsiwhich.are located further than "d" from a' support. There wa's-.a?similarfbut smallerf..

effect.on_ bending moments in the basemat.J0nc'e againithe' values, determined:

from the finer mesh model'are smaller than those'obtained using.lthe coarse.

g mesh.. 'All. bending moments are within the. allowable. limits.f 5.3, Findings and Conclusions

.c nj a

~i I

The refined: mesh contains a-sufficient-number of:) elements to reasonablyj,

representTthe expected shearsand bending moment distributions.yThe, solutions "y

'using this refined model. indicate smaller shear: loads 5and bending;momentsfin:

the basemat than were found with the baseline' analysis'.L L Furthermore, the; d

changes from the' baseline ~ analysis are' relativelyg small and 'of;a type to' bel f

tj'.

g

(

E

-10 =

i:

y.

1) r ;ip r

Q t

expected based on the finite element mesh refinement. This confirmatory analysis is.therefore found to be acceptable'and it is concluded'that further-

1. refinements.to the mesh would not result in any increases in mat loads.- This.

issue is therefore closed.

6.

BASEMAT CONSTRUCTION SEQUENCE ANALYSIS 6.1 Introduction The concerns regarding the adequacy of the basemat arose because of the

~

cracks. It was the BNL judgment that'these cracks occurred during the con-

~

struction period and that they-were caused by settlements induced by. the sequence of placement of the. construction blocks of the.basemat.- The tendency for cracks to form was certainly increase by thermal and shrinkage effects which were at play during the early life of the mat.

There, was.however, a dissenting opinion expressed by a member of the NRC staff that the cracks :

night have resulted from poor local soil. conditions.

All of f these judgments.

were qualitative in nature with relatively little_ quantitative data to support

. A confirmatory -analysis task was therefore established.

the cause of cracking.

to develop such quantitative data supporting the ' conclusion that cracks' formed because of the construction sequence.

A clearer understanding of-the crack ~

formation mechanism is required if one is co ensure that the mechanism which originally cause the cracks is not iatill operative and tnerefore does not have.

the potential to cause further deterioration ofJthe mat.

6.2 Results of Analysis The mat was constructed.in approximately 60-foot square blocks. A finite element model was used where each block was broken into either 6 or 9 elements. Solutions were obtained at different stages of construction by adding blocks to the model following the actual construction sequence. The concrete properties were-varied with age so ' that the dif ferent, blocks in the model had difference stiffnesses depending on the age of-the concrete in the block at the time of the solution.

Actual test ^ data obtained at. concrete ages of 1, 3, 7, and 28 days were used to model the concrete compressive strength

~

age variation. Modulus, tensile strength, and Poissons ratio were then related to the compressive strength.

Standard creep, shrinkage, and thermal properties of the concrete were used in the analysis.

At each stage of the. solution, field measured edge settlements -were used as input, causing stresses to_ develop in the mat. Space and time inter-polation were used when required to either fill in data that was missing inL the field survey information or to resolve discrepancies between measured data.

points.

(

+

s m

';i' e

i;.,

V"i (p' '

c >

v; j

-ten l

.n m.

y vp w

y

1 t'

n..

1 H

The solution at each atage(was atudied to--determine;whether.the ~

H section' cracking moment was. exceeded.

If~itlwas,'the:effectiveTsection fstiffness was determined based on the concreteLtensile strength at the current age. ;This was done' outside thei computer program (NASTRAN);and l

~

1 additional. cycles of solutions:obtained,.using'the cracked stiffnesses,,

L untilithe, process / converged.

s Mat forces were determined incrementalls from the; previous.'soluti.ns'.

i o

.to facilitate the handlingfof creepLeffects. LThe'effect of: creep during, Lthe; current' solution step-was: handled;by using'a: reduced concreteimodulus.

Aireduction factor.was' applied.to the loads; accumulated during; previous',

1 steps'to; account:for the creep-inducedirelaxation"ofjtheseil'oads during'the current. solution step. Thermal 'and shrinkage ef fects were. found to be:

relatively small and,lhence,Lwere omitted from'the finite # element calculations. -

1 i

The. construction;sequenceanalysisresultedin. predicted / concrete?_

crack patterns that.reasonally' match the observed. crack patterns.?1The, saximum. predicted rebar' stress was4found!to;be about-7 ksip with;thissvalue 7

-decreasing as the mat aged.

It.was further found that thefshear stres'ses: "7'

in the mat were always?less than the American.ConcretetInstitute Code:

allowable: stress values,. indicating that:the cracks were" induced-byLmoments:

rather than shear.

6.3 Findings and Conclusions The codels used to represent the constructionisequenceianalysis' vere fou'd to be reasonable.and a good representation of the actualiprocess.-

n The predicted crack patterns match the. observed patterns and provide quantitative data supporting.the concept that the cracks!.in the.basemat are

~

bending cracks which occurred during construction.. Thejtop rebar stresses-were predicted to reach a ! peak tensile.' stress of about 7 kai.'. The.;

magnitude of-this value is,probably'nottoo reliable since it.mayL. depend:on; the size.of the finite elements used f or : the' model'.' lit"is significant,-

.however, that the rebar stressesLare found;to decreasefafter reaching (a peak.. Combination of the stress;withLother.stre's; components.are' discussed e

in the following section. In summary,'itsis< concluded--that ample evidence-is provided to determine:the cause of the cracks,.andLthat;the ecate'of stress after cracking is such'as to;not impair the. performance of-the basemat. Thus,.this issue is, closed.

7.

COMBINATION 0F ' CONSTRUCTION SEQUENCE AND OTHER LOADS.

7.1' Introduction-During the licensing review for.Waterfordi3, several. testimonies: vere 3

.. filed'in support of the-license application..-These testimonies depended;in part on the assumption'that the basemat was. subjected to positive bending 3 1

at,

!I

..,s

3 3.,,.s

)

{

moments,for operating and accident loadings which produced compressive stresses in the top of the mat.

Since these, moments were opposite in sign from those,which caused the cracking, the conclusion was reached that the construction sequence moments would not be additive to the operating load moments and could therefore safely be neglected., The previous confirmatory j

analysis report (Reference 10) indicated thatLthere were large areas of the l

aat where the operating bending moments induced by operating loads produced tensile stresses on the top.o* the mat', and were therefore additive to the i

construction sequence bending moments.. At the request of the staff, another task was added to the confirmatory analysis program with the. objective of combining the construction sequence results with the other loadings.

7.2 Results of Analysis

't A direct addition of the moments induced by the construction sequence j

l loads to those induced by either normal operation' or accident. loadings is not valid since the construction sequence moments are deformation-dependent. The magnitude of'the construction sequence. moments arises from imposed 1

deformations and therefore depends. on the mats stif fness. The superposition of operating load-induced moments on the construction sequence moments results in additional mat cracking, reduced mat's stif fness, and corresponding reduced

-j construction sequence moments for the same imposed deformation patterns.

The superposition was therefore accomplished ~ by superposing the construction sequence curvature with the bending moments induced by the applied loadings.

1 Several parameter variations were considered in combining thei construction sequence and operating / accident loadings.

In all cases, the total bending moments and shear were found to be within the design allowables.

7.3 Findings and Conclusions

't While the combination of deformation-induced construction sequence i

moments and shears with operating / accident moments and shears is subject to l

several uncertainties, it is concluded that the studies generally performed

]

the combinations in a conservative fashion. The resulting combined moments j

and shears were found to be within the allowable values.

It is therefore gr l

concluded that the addition of construction sequence loads to the operating and/or accident loadings would not adversely af fect the performance of the basemat. This issue is therefore considered to be closed.

8 BASEMAT SURVEILLANCE PROGRAM i

8.1 Introduction l

A basemat surveillance program was instituted to monitor the baseest cracks to ensure that any changes in crack patterns would be detected and then-evaluated. The program has been in place since 1985.

4 1

. Q y ' % $ % ', ' % '? T

$. hl '

, ? * %g ?? N :p un ?

y" M s

W hNj$$ BOP

'y])g/' {+ W' p'wp

~

Q hp 1*

]l 4,

T' R

,y

, fy;, '

mi,,

sa:

4>

+,

g3 m

y d

, i-T :;

1 ni i

M 8.2 Scope of Review

<m A

,3 e sa p,

'There are four components to:the monitoring (programig

_ yg

.. d

.g. :., M s M;*

w r-i 1

s w

t;'

w

1.-

' Settlement datalis : collected ionc, the basemat k KWhenever atone-inch t >

differential settlement /is!found between1:the? center 7andfedgesiofLthe!'

mat /LP&L mustitakeLaction:.and ! review ;the. addbacy lof /the mat;i SThis; g"'.

WQ actionilimitfcorrespohds7to a?reba'r? stress"or.about42 keid.% F K>

. %ng a

h tiasemaQ

'y,,

.2.

Ground'wa'ter.chemistryjis? analysed bnsconcernybo

~

cracks is the potential?for>rebarYeberosionLastwater) propagates /into

'n

.them..;LThewatercanonlyjbisbrrosivephoseverfiffthe;chloridei

( / d contects f;theJgroundwaterJexdsedsJcertainLlimits; cA chloride o

% f,$

conten':'of-250 ppm:is:the action 111mit.: rIf(th'is! exceededf thej

~

,T lic'ensoe'mustlinform NRC'and jrepare an engineering; eval.uationh;

_.J s

m 7

> t.

t' u:.

, ^ ~.,

l..,

J..

O

'n~;

G

(.

'3.

The groundwater.flevellis monitored.c<.There arefnotaction. limits set) f ~

y for5this11evel.3: ItLisimonitoredibecads'e any?adverselbehssior ofjths]

R mat lwouldLm6st1'ikely;be'correlatedtwithitheiroundwaterlievel(i f

• b m

W g n

.s

..sc

.. a 4.

The crack cha'racteristiestareimoiiitore'd$1 ) Crack widthsfarefmonitored::

X>

a t s elected ' loca tions " using1Wj t teino re s t riin t gages d [AnNcrack? width?

.h reaching L15L mils; requiresj an -ass'essment(bpiLP&L.] [Crackhvidthe iand ! < >

lengths were'origina11yfmonitoredSby.visualfinspections M Anyfnewi

[

cracks. longer.thansten feetiorfany' crack widthLexceeding>15'milal 9

4 required an.' assessment"by LP&L.~. Theivisbal;inspectioniwas;foundIto g'

4 give results which.were veryisdbjective.( sit twasJtherefore; decided l

to eliminate the'vislual; portion)ofjthe; programp Crack-widthl

(

measurements are now'requin dLon-fifteen' instrumented cracks.

. :)

~

"a4 i.

.The surveillance < program wasdreviewed!dufing siteLvisitslandir'esult'sJwere-M discussed during the on-site audits.

[

o 9'

L 8.3 Findings and Conclusions 1 The surveillance program willlprovideLadequateTwarnihg(shoha yfundsuals n

behavior oc' cur in the mat. While th'ere'are'no discernable-trendsTin the datat

~

5 obtained to.date, the measured data, are all.withinithe allowable limits.:. Iti r

7 is recommended that - data' obtained duringithe routine 'abrveillance' be reviewed y

p until'some overall trends can be established.

,g q.

a 4

9.0. :

SUMMARY

=

3:$,

........, ~ym

~., y, The:Waterford 3? operating licenseicontainedda condition requiring a; i.

series of confirmatory l analyses beiperformediby'~ hhe< licensee.?Alseries ;ofs d

'five. analyses were required relat.ingjtoftheieffectidffthe cracks h thef,, i s'

-basemat on'the adequacytof theLmat.; The licansaa y.portj(Refer %eil) o

?

9 EThe details Wff+

\\q C

describing the results of. these Lanlayses' has 'be'en Jreviewed.1 fU this review are contained inI his TERM t

?

j

(

E q

d

~

4 l:,

O Yj t v,

h

)

i we,

)

n lfh:$,

b c

9

lli?!

f

.p

~

y i

- k

. [fb

(!., -...

[

1

i,t "

3:

1 g,. 3 i

f b4 i _

f 4

a

\\

N

(

The results' of the - confirm 4 tory analy9es generally-provide quantitative.

data in support of che. adequacy of theEcracked mat.

The;following. specific conclusions are made for each ofvthe five confirmatory. tasks:

(a) The first task concerned!the.effect that' seismic loading'might have; on reducingLthe' compressive stresses in the mat-due.to.eidewalli coil / water. pressure. The confirmatory -analysis demonstrated ' that the net-compressive stress never was. completely eliminated.! While:

large uncertainties exist in this calculation [ it.isjconcluded thatl

~

i the compressive stress willaaeverlcompletely. disappear'-

Furthermore, it. has been demonstrated that the mat. could performLits intended function even if the compressive stress vanishes.

(b) The second task dealt with the coupling'during:a: seismic event

- between the mat flexibility and.the~ superstructure flexibility..

.The mat was considered to beLrigid in the original; seismic analyses. The results of the confirmatory analysis' demonstrate that' the structural seismic loads and the floor: response. spectra are within the' design values when the mat flexibilitymis considered..

1 (c) The third task considered,the boundary conditions used in~ evaluating seismic stresses in the sac.

4 1t.has been shown that. upper andLlower

~

bound values of boundary restraints lead to' acceptable stress l -

levels. Therefore, the location of boundaries constraints have been l

appropriately oodeled.

1 (d) The fourth task evaluated the'basemat mesh sir.e effect on calculated shears and moments.

It has:been demonstrated that a finer mesh j

results in:more favorable moment and shear distributions.

Therefore, the data frod the-coarseLmesh model are conserv'ative.

(e) The fif th task developed a model of the:basemat applicable 'at various stages of construction.

it has ' been demonstrated that -

bending moments likely' developed during.the construction period which casued the major east-west cracks.

Of..

In view of these findings, it is the conclusion of-this TER that the confirmatory. analyses. have demonstrated the adequacy ~ of !the mat.

Another portion of the license condition required that the; licensee develop a basemat surveillance program to ensure that conditions within the mat do not' change significantly.

This program,!which will continue for the life of the plant, has been. reviewed and has been found to be acceptable..

-l

\\-

h

~,

i__i.i____________i_

3

,.e*

i a

REFERENCES 1.

Waterford Steam Electric Station Unit No. 3 ' Basemat ' Summary Report,

j Docket No. 50-382, July 1987.

s 2.

Waterford SES Unit 3 Basemat Surveillance Program,1 Docket No. 50-382, June 26, 1987.

i 3.

Waterford Steam Electric Station Unit No. 3 Summary Evaluation-Structura11 1

Significance of-Basemat Nondestructive Testing Results, Rev. 2, November 1984.

4.

tJfidavit of J. L.-Ehasz, Docket No. 50-382 0L filed before the Atomic Safety and, Licensing Appeal Board, U.S. Nuclear. Regulatory Commission, j

January 7, 1987.

1j 3

L 5.

Waterford Steam Electric Station,-Unit No. 3, NP1S Wall Hairline' Cracks

~

Evaluation, prepared for the Louisiana Power & Light. Company by Ebasco j

-Services Incorporated, New York, April 1984.

]

6.

Mat, Report No. 8304-2, prepared for Louisiana Power & Ligh't Company by j

Harstead Engineering Associates, Inc., Parkridge, New Jersey, October 12, l

1983.

)

7.

Review of Waterford III Basemat Analysis, prepared for the U.S. Nuclear Regulatory Commission by the. Structural Analysis Division, Department of Nuclear Energy, Brookhaven National Laboratory, Upton, New York, July 18,.

1984.

8.

Addendum No. 2 to the Reivew of Waterford III Basemat Analysis with Appendices, prepared for the Nuclear Reglatory Commission, August 7, i

1984.

l 9.

AddendumtotheReviewofWaterfo(jgIIIBasematAnalysiswithAppendices, prepared for the Nuclear Regulatory Commission by Department of Nuclear Energy, scookhaven National laboratory, Upton, New York, August 3,1984.

10.

Waterford Steam Electric Station - Unit i s. 3 Program to Perform Confirmatory Analyses, Nuclear Plant Islanet Structure Basemat.

Results of the Analyses, October 1, 1986.

11.

Muenow and Associates Inc., Charlotte North Carolina, Report,-.

"Non-Destructive Test Evaluation of Basemat Concrete At Waterford No. 3, l

Louisiana Power and Light Company", submitted by LP&L,- Letter dated October 26, 1984.

. 4