ML20246M508
| ML20246M508 | |
| Person / Time | |
|---|---|
| Site: | Saint Lucie  |
| Issue date: | 01/25/1989 |
| From: | Turbak M COMMONWEALTH EDISON CO. |
| To: | Davis A NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III) |
| References | |
| 3032K, OLA-A-007, OLA-A-7, NUDOCS 8903270175 | |
| Download: ML20246M508 (95) | |
Text
. - -- , , - - -
bbbt }
I - 4 Commonvicolth Edison one Fire Neors Pata. CNcago. knos_
- - Gran Mgfy to: Pos Omco Box 757 _ _ _
CNcago,lanos 80600 0767 M___x NT- F M .MMF 1
. m.~
O. , *, * .~ . . . . . s. ..;. . . .s s. ,.t- .:,, .. ->:r -s-\ - . *. . . ., v ., w's -w *-
' .t
- v= :..1- -
. ~~.r.es~..n,, x g ,
May 5. 1987 '89 M 22 P7 :02
~
NUCLEAR REGULATORY COMMISSION
' Docket No. N T5a l[ (fhiki Eih. No. 9 in the mztter of f lcw; L h., J W h AJ $), Q ,
v 4taff k Mr. A. Bert Davis IDENTIFIED Regional Administrator Apphant "
RECEIVED U.S. Nuclear Regulatory Commission intervenor REJECTED Region III Cont'g Off'r 799 Roosevelt Road Contractor DATE / "M-[f Glen Ellyn, IL 60137 Other V/itness -
Reporter dl,' Oiu u ,,g %
~
subject: spent Dwcities Puel' station storage R- anitrkanep T- m NRC Docket'Nos.' 50-254'and50- G c,
. _ _ . _ _ ___._,,m
Dear Mr. Davis:
A meeting was held in Region III offices with Messrs. R. Warnick, M. Ring, and others of your staff on May 1, 1987 regarding the Quad Cities station spent fuel storage racks. We requested the meeting to inform the staff of our preliminary assessment of an apparent anomaly in the neutron poison material, Boraflex, performance in the fuel storage racks.
Enclosed is a report from our contractor, Northeast Technology Corp. (NETCO), that describes the preliminary assessment of the anomaly.
The report is entitled " Preliminary Assessment of Botaflex Performance in the Quad Cities Spent Puel storage Racks", Report No. NET-042-01, Revision 0 dated 4/10/87.
During the past several months efforts have been made to understand the anomalies discovered in the fuel storage racks at guad Cities station. '
Nondestructive testing of the racks have shown gaps in the Boraflex extending the full width of the cell wall up to 4 inches in length. There is s .. h a . u q m .... '/2. r : ,i.rfurt BY"thirtist.
NETCO was retained to evaluate test data and fuel rack design to explain the phenomenon and predict the extent of gap growth with further irradiation. It 8- !Ji; :-' - ' r r ' " -' T-- - " -- -'- ' "- -- -- :r ' by.
irradiation-rervits-irl suffici~errtensite""strss. w .. . L.&:;: "-- it-is.restrainee as irt tWp(Tat ctYrefr~idE"Tesi~gfr.'
e 0175 890125 7 h32[ DUCK PDR o 05000335
s
~
M.x 5. 1ee, .
O o vi-h l
Additional testings ant. evaluations arrbeing' scheduled.,f We will 8 keep you informed ~of^fU~tste dev,elopments.
Please direct any questions regarding this issue to this office'. >
Very truly yours, f
i l
Y
- q..m l M. 5. Turbak operating Plant Licensing Director 1
1m l
Enclosure:
As stated
( l j
cc: Resident Inspector - Quad Cities T. Ross - NRR 3032K O .
~~ ~
=, ,
. \
i O[27C0f NORTHEASTTECHNOLOGYCORP
> 0 I
f '
Report No. NET-042-01 Revision No. A l PR5H k.f5 arf 555555ME1@ @F ~
- BORAFLEX' PERFORMANCE ~IN' QUAD' CITIES: SPENT FU
-.a uumumg STORAGERACESWF{
O Prepared for ,
Commonwealth Edison Company by l
Northeast Technology Corp 4/10/87
/
r O
I
v J
I
(
, n+
C:) l I i 1"
~
ZARLE QE CONTENTS I.ASA 1.0 Introduction...................................... 1-1 2.0 spent Fuel Rack Description....................... 2-1 3.0 Description of Quad Cities Unit 1 Spent Fuel Pool and Refueling Practicas........... 3-1 4.0 Evaluation of National Nuclear Corporation Measurements...................................... 4-1 4.1 Description of Test Methods..................... 4-1 4.2 Results of the Special Test Measurements........ 4-2 4.3 Results of the Standard Test Measurements...,.... 4-5 5.0 Audit of Fuel Rack Manufacturing Process........'..-5-1
)
6.0 Assessment of Madiation Testing of Boraflex.s..... 6-1 6.1 Summary of Irradiation Testing of Boraflex...... 6-1 6.2 Evaluation of Data.............................. 6-3 7.0 Postulated Radiation Damage Mechanisms in Boraf1ex.......................................... 7-1 7.1 Physical Properties of Filled and Unfilled Methlylated Po1ysiloxane........................ 7-1 7.2 Radiation Damage Mechanisms in Polymers. . . . . . . . . 7-3 7.3 Effects of Irradiation on the Physical and Mechanical Properties of Methylated Polysiloxane.................................... 7-7 8.0 Gap Formation, Gap Growth and Long Term Integrity of Boraflex in the Spent Fuel Pool Environment.................................. 8-1 ,
8.1 Potential Mechanisms of Gap Formation........... 8-1 8.2 Estimate of Maximum Gap Size................. . .. 8-5 8.3 Long Term Integrity in the Spent Fuel Environment..................................... 0-8 (CONTINUED)
O ..
I e
4
___ ____.-___.______m.__ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _
., s O
5 f 1
~ I;l TABLE OF CONTENTS (continued) 9.0 Reactivity Effects of Gaps in the Fuel / Rack Neutron Absorber.................................. 9-1 9.1 Reactivity calculations......................... 9-2 9.2 Probability of Gap Occurrence...................
9-6 9.3 Local and Global Reactivity Effects............. 9-7 l w 9.4 Model/ Method Conservatism...................... 9-9 10.0 Conclusions ...................................... 10-1
-1 I
Referencent l
1 O
O ..
's 9 8 O. l I
A PRELIMINARY ASSESSMENT OF BORAFLEX PERFORMANCE IN THE QUAD 'ITIES SPENT NOCLEAR FUEL STORAGE RACES
1.0 INTRODUCTION
The spent nuclear fuel storage racks at Commonwealth Edison's Quad Cities Nuclear Power Station consist of a stainless steet structure and utilized Boraflex* as a
. neutron poison for criticality control. Recent inspections of the Boraflex absorber by Nationat Nuclear Corporation has ,
revealed that ' gaps" in the ' Boraflex absorber have developed .
1 The measurement technique utilizes a Cf 252 neutron source and either BF-3 or Be-3 proportional detectors. The detectors do not record fast neutrons (source) produced by the Cf 252 source but rather measure thermai neutrons which have been transmitted through gaps in the Boraflex and have been thermalized and reflected back to the detector.
The measurements conducted by National Nuclear Corporation consisted of two types. The first, designated as standard measurements,. consisted by a 'go-no-go' type measurement in which the presence or absence of a gap in the Boraflex was determined. The second tests, designated as O - Trademnr= of > rand 2nda,tria1 Set ices Cor oration 1-1 S
+
A.
O I
~
- special tests', provided a meksure of the gap size and the axial elevation of the g ag. . This study focused on an l evaluation of the later measurements.
The report herein documents the results of an evaluation of the National Nuclear measurements, test data developed by BISCO at the University of Michigan at Ann
- The objectives of this evaluation are Arbor and other data.
as follows:
o To understand the mechanism of radiation damage in Boraflex o To estimate future performance of Boraflex in the -
Q environment of the spent fuel pool at Quad cities o To quantify the implications of Botaflex gap formation on the criticality state of the Quad cities spent fuel storage racks The following sections o'f this report document the, results of this evaluation.
O k 342..
. 1-2
+ 0
e O
I
.h a 2.0 SPINT Z,Q74 .BACK DESCRIP11DR .
The spent fuel storage racks for the Quad cities Station were fabricated by the Joseph Oat Corporation of Camden, New Jersey. The rack modules and pool layout have been described in Reference 2. A brief description is included here to aid the reader in interpreting the following Sections of this report.
The' fabrication process starts with the manufacture of a series of elements (hereafter designated by their form and noted as ' Tee', ' Ell", or ' cruciform" shaped) which are O subsequently welded together to form an ' egg crate" l structure which ultimately provides storage locations or cells for spent fuel assemblies. The basic Tee, Ell and cruciform elements are manufactured by starting with Ell shaped subelements of stainless steel 6' on each wing, 165" long and .0754' thick as shown in Figure 2-1. On the E11 and Tee elements the outer stainless steel is .120" thick. A cavity for the Boraflex is created by using end strips of stainless steel to form a " picture frame" between adjacent Ell's as shown in Figure 2-2.
In the process used for manufacture, the strips forming the ' picture frame" are tack welded to the stainless steel Ell's. To retain the Boroflex in place during manufacture, O aow E111 con adhesi.e coow o .. 33 is u,ma. A head of 4
2-1
~
O I
adhesive was applied to roughly the center of the prity and subsequently was distributed into a 2-1/2 *-3' wide strip along the entire axial length of the cavity with a stainlass-steel scraper. The Borafles was then rolled into the cavity pressed in place. The nominal dimensions of the and Bormflex are .070' thick, 5.9' wide and 152" in length with a Boron-10 loading of .01728 gs/cm2 (areal density).
With the basic E11 sections fabricated, the Tee, Ell and cruciform elements were assembled as shown in Figure 2-1,. During the welding process, copper chills were used to minimiite the area of the heat affected zone and to preclude O overheating the Boraflez.
The basic elements are joined by welding with welds applied at 4" centers and 2" diameters (see Figure 2-2).
The Ell, Tee and cruciform elements are then assembled, tack and MIG welded to form the egg crate storage rack structure as shown in Figure 2-3. The egg crate structure is then welded to a base plate (Figure 2-4) with feet (Figure 2-5) to form an individual rack module.
O 2-2
O l i
0g y
T
,5 7
6( E 0 ..
~'
'11 E .
6 l l
i ,
. S
~
.) S : , '
. / "
5 ~ S
~
" 7 ,
60 , K ~
.. H T
- (6 o E ~ 6 .
p _
z E
g L S %
l
- i. .
- . . 0 T A L
=
A - _[
) J E
f
) , 3 0 - (
c .
C E
(J )
Tp T b e
6 0 7 ;
\ E E
l
(
_ ' i
_ H l T
6 - S i
S N
( .
S s E
_ / / M S
- ". E
. L N S KE O H T
T E N T. H T R E .
. ~5 HE 0 TE 7 .S M 2 1
l l
0 K E O.O S H XL P T E E x}1
_ I P. ~ L R ). 1 0 F Y
=0 6 0. O
_ D (T 2 ' - B
_ N o. O E -.
_ I l .
- > n M
=
% R
- +l6 O
, . .I ;s F
I C
U
_ a "5 R 6 "5 7 C 7 .
0
_ n 0,- o 5* o
)
o Q
e 6 .
1
(
y a
i
+ ,
~
O ,
- i. .
i h
l
%N _
. ? .
STAI N LE S S STEEL ANGULAR, N)
S U B E L E M E N T A, . TOP END ,
( FI G. 3.3 ) .
(STAINL SS STEEL)
O.
LX i ONG EDGE
[T L STRIP
( T Y P, )
{ STAINLESS STEEL)
\ ,
,/
O FIGURE 2-2: ISOMETRIC VIEW OF A CRUCIFORM ELEMENT As (From Reference 2) na.
O E
N X S SL LG O I E
L EA A AT F T LI C RA A E R w I
RE N E P OC OH I
A T Y TO B S TA T SL SM A s l 1
a,< '
{ ' n ', l ,
l ,'
. - . l s-A 5 0 2 0 _ 1 0 /c 1 .
- + - C _
. _ c _
r
- u 7 _
2P '
- 2. Y T 2
O E.
- 6 _
- . . l l ,
l'
'r , -
w .
[
. j, >
J- I jl i ,[
A
=
- z . . _
o _ _. .
P - _
$*_Q S Y .
- l . ,l ! . I
,6T
- _ c .
< . r g- . . I
(
r
' J O .
ua,
m O
Top W ;
f M
- b e,. : ::::: , 889 ie i, n i l
, j M
b l g E i i .
. 5
- i' s
a Jf ,
o :
Swenes I ! '
I.
., s w .
h- .
.j* l
/ '*
} , sf ,, f
}
l
{
~
A ,
,, , E '
/ ,
k / '
8 i
3
/ :
O 1 m i.es .
I i A . ,,
/ -w
- I Seeser sir, i ,
t
" (4) %@ Fiew L p,m, o !. . ,
4 i E
':x
.J8
- , fr.AI. _
~
I Sans Piste
%*. # 45*
,, i , /
y=?r
- 2 m-i g
5%"
g I O
FIGURE 2-4: CE8L ELEVATION WITH BASEPLATE ,
(From Reference 2) l
y.
o ,
1 i
"i
~
O .
i rl CELLSh i
^
MODULE B ASE PL ATE ,'[ ~ ~ ~ ~,
~
~y
. 1, J . .I' ll
.[
s m u s,-
ii ,y J . .
..;._ ,~.A w v 'mp';
a ,
d i a i,,
,, 4 s/ i ss 2
1s j' u 7
m7 $- ,
o 2 x g g gxNg g g gN a
" ' et *- ~
O ,
ik J\\ \\ \\l l \\ \\\\k Y
'/,/ * /
/ /
13 '/4
.. I -
I 7 e/
50 / /
/ /
/ ' /
-j _ (T Y P.)
//L Al ~
' fNN\\s W 7 ,
_V\\\\\Ni a i
l 1
FIGURE 2-5: FUEL RACK MODULE SUPPORT FEET l
(From Reference 2) l * -
l
.~.-
O
- b i!
I
(
. p
3.0 DESCRIPTION
DI, Igg Q M CITIES D,EII & 1 ZRIL i 22QL A52 REFDELING PEACTICES .
The Quad Cities' spent fuel pool contains 15 spent fuel storage modules as described in section 2.
The pool layout is shown in Figure 3-1 and provides storage capacity for 2907 fuel assemblies. The fuel transfer canal providing access to the reactor is located.along the south wall of the pool approximately adjacent to module Kl.
During a normal refueling, the complete complement of fuel assemblies from the reactor core is initially moved into modules E2, K1 and C4 with some assemblies being stored in cells located in adjacent modules (A4, C3, B3). These modules are designated as the ' refueling storage racks".
Subsequently, those assemblies scheduled . for re-insertion into the core are removed from the racks and transferred to the reactor. The assemblies scheduled for discharge are then transferred to the modules in the pool designated as
" discharge" rack modules. During previous refueling operations freshly discharged fuel may reside in the reload rack modules for periods of 1.5 to 3.0 months. As a consequence of this refueling mode, the refueling storage racks tend to accumulate gamma exposure faster than the discharge rack modules and may be subjected to additional thermal cycling.
3-1
O
~ .
I t
The procedure of designating refueling and $1scharge e I racks is described since it may be one factor in interpreting the behavior of Boraflex anomalios described subsequently.
O l
9 O +-
3-2
'h , ==in..
e I
O FIGURE 3-1 QUA.D CITIES UNIT 1 SPENT FUEL STORAGE POOL '
t O O O O C C C C i
i M T f '
.c . , .
, , , i .. . i r- e o a u .
1 l , , . . , .
i i 8 I i
'
- i t i . !
t ' '
! ' i I ' i i
e s i ! t i 4 , a r i i ; e i t i i l i i t t ' i I t r i i i O 6 l 3 4 j i
I l I 3
j e , T i
p 4 3 1 i i I
- i t
i e 3 ! I h t I r 1 I I 4 ! i o ; p g p i j 3 I I ) r si g i ' m e m a m e I
.r s .s a r- 6 r. 6 I I e 4 i i 8 I 6 i t i i t i i t ! P f I 4
i . i 6 i i 8 8 ' i e i 1 ' ' I I I '
r i 6 e i I i ' '
l 6 4 '
LA E i I I l'
]' ll i i i a e i e e
i s 1 l l 8 1 5 I t 3 ' i I I i I a s e i 8 ' I i i l
, , 1 3 I t i i i I ' 6 i f
, i a i '
e I e i e i i
- I I I l 5
, i j i
' 3 I
, i a i i I l 9 l E 1 l t 3 1 i 1 e 1, li i i !
- P 1 i F T i l ! i I e i i i i !
I I i l ! f O
, M* W W~a W 3, al a mi da, m u e e
4 O
I 4.0 IVALDATTON Qf, Igg RATTONAL HUCf PAR.
h '
MrAsontwents CORPOPATTOJ s' 4.1 Dernietion g the 2331 Methods An initiab. testing -campe&ge-eendweted Nuclear" Corporation --by- Mationat hadWY""centteobt:
toor pince -usinewhah h*= Ceamenwealtb Edison ,
Met!!et?""The Standard Test =*"4'a=
Method 5- - A--*ca=t*J!.Ah3Aan
, E
_Cf 25 s ec r d&~E66"*four-- BWp ropett fEHT"""68 s ensiliirTT6*th etsrtr"n e ttTon1f. Hettrr5"""Whfetr" arw ==
so that each detector is adjacent toThe equipment is so designe '
when it is placed in a fuel storage e .
c llone panel of Borafles Cf252 The BF-3 detectors do not record f source but do detect thermal transmitted through the cell wall neutronsast neutrons which have been back into the cell containing th, thermalized and scattered wall contains Boraflex, e detectors.
If the cell significantly attenuated; the back scattered neutrons are backscattered whereas where gaps exist, neutrons the '
undergo attenuation. significantly less During a measurement, two passes are made in each cell--first from the top to the botto bottom to the top. m, and then from the and a peak in The count rate is continually recorded the count rate is discontinuity in the Borafleic abso b indicative of a r er.
The Standard Tests
e O .
~!
2 3
4.0 EVALDATTON QI, If1 NATTOMAL Eg CORPORATION i
)
MEASUREMENTS 4.1 Description stf. tha Ital Methods ,
An initiab testinecempe4gn==eenesoto6=- br" Nat iona) tindWF" coffer... t: C1---":-itb-Ediac %
Nucleer* Corp 6Yatish too W m s % w h '-- 5-- '--4ca=**d the stg;1 d q M gg L Metuwe. The Standard Test Method ? 'i===
d -M U D '-- -
=>
sourde"~E65""*four== BF=^ y6 wpoil.fEfMT""E8Yteterr""hu sedillve te thVtst!"hHtf6Ef. The equipment is so designed so that each detector is adjacent to one panel of Boraflex when it is placed in a fuel storage cell.
The BF-3 detectors do not record fast neutrons from the 252 source but do detect thermal neutrons which have been Cf transmitted through the cell wall, thermalized and scattered back into the cell containing the detectors. If the cell wall contains Boraflex, the back scattered neutrons are '
significantly attenuated; whereas where gaps exist, the undergo significantly less bachscattered neutrons attenuation.
During a measurement, two passes are made in each cell--first from the top to the bottom, and then from the ,i bottom to the top. The count rate is continually recorded indicative of a' and a peak in the count rate is l discontinuity in the - orafie assorser. The se ndard Tests O l 4-1 )
l
O -
I are considered a "go-no-go' Sype of measurement pt do not' provide an accurate indication of the size or axial location of anomalies.
After anomalies in the neutron absorber were' indicated in the initial testing campaign, a special test method was developed. This method utilizes a Cf 252 source and two He-3 proportional detectors and is intended to provide a measure of neutron attenuation in a single panel of Boraflex in a ,
k storage cell. The detectors are wrapped.in lead (to reduce-the potential of gamma interaction) and cadmium to form a one half inch high window in the front of the detector Q sensitive to thermal neutrons. The source and the detector housing are suspended from the refueling mast bridge.
During a measurement, the detector and source housing are moved in one half inch increments through a storage cell.
The mast position and detector count rate are continuously recorded. By comparing the shape of recorded peaks in the i
count rate with peaks from measurements on a calibration j i
standard containing gaps of known size, the approximate size .j and axial elevation of the gaps can be determined. j l
4.2 Results 21 tht 1111 Measurements l In the initial testing campaign using the standard Test Method, a total of 203 panels containing the Boraflex j absorbers in the refueling rack,ydules were tested.
Of 4-2 1 i
1 6 . -
l 1 -
O -
I these, 77 panels showed indication of anomaliesd In the l discharge rack modules, a total of 103 panels were tested of which 18 showed indication of anomalies. Although the dose to the rack modules has not been rigorously calculated, the dose in refueling racks has been estimated to be 10 8 rads.
Review of the magnitude of count rate peaks recorded in w
these measurements conducted by CECO personnel indicate the.t the average gap size in the refueling rack modules is larger than those in the discharge rack modules. Based on those panels determined to have gaps from the standard tests, 28 panels were selected for testing using the special test O method. x11 28 ,. nets were in the region of the ,o.1 containing racks designated as reload racks.
For the current evaluation of the NNC Special Test Data, the following procedure was used. If the NNC data showed indication of an anomaly in the range of 0.0" 'to ,
1.0", the occurrence is defined as falling into Gap Size Interval 1; anomalies in the range of 1.0* to 2.O' as Gap Size Interva1 2, and so on. In a similar manner, each Boraflex panel was divided into 15 axial intervals or bins approximately 10 inches long.
Table 4-1 summarizes the gap size, gap size interval, axial elevation, and axial interval for each of the 31 gaps
- detected. The average gap size is 1.35" t/ .87 (1 sigma) .
O te * =1d be ao= d ** * *a a i= d *= vra med mee==re r I
4-3 .
i O. -
I gap elevation as measured from the bottom of each cell. The Boraflex does not continue to the bottom of each all and appropriate corrections based on manufacturing drawings and detector design details were made. Accordingly, gap axial distribution data has been normalized to the bottom of the i Boraflex sheet. l A review of the data in Table 4-1 shows that of the 28 panels tested, three (NO7-W52, NO2-W54 and NO1-W54) contained two gaps each. Accordingly, the data was reanalyzed assuming a cumulative gap size (sum of two gaps) for those three occurrences. The results of this analysis ;
are shown in Table 4-2. and Figure 4-1. The averagt i Q cumulative gap size on this basis is 1.5" +/ .85'.
The axial distribution of gaps is shown in Figure 4-2 in which the number of gaps versus axial interve.1 is 1
This distribution exhibits several characteristics plotted.
which should be noted:
o There are no gaps in the first four intervals.
o There is a distinct peak in occurrence around the mid-plane of the cell.
o There appears to be a second peak near the top of the cell.
Possible mechanisms responsible for these observations will be described subsequently.
O 4-'
0 0
O .
'l 1 l
4.3 Results g th,3 Jtandard M Measurements Thestandardtestmethodwasappliedto118cakisaf'ter [
tihe second refueling outage primarily in those modules j l
designated as the ' refueling racks'. Of those 118 cells tested, 45 cells were found to have all four panels intact without detectable gaps and the balance (73 cells) were l l
found to have at least one panel with one gap. Table 4-3 s
contains a summary of the standard Test results and shows the cell location, total number of gaps detected, panel location and number of gaps in each panel as well as the j number of panels in the cell containing gaps. It should be noted that for some individual panels. indication of more .I O than one on, per ,anet wa. observed. ror the eur,ose of obtaining the number of panels in a cell which contained one or more gaps (last column in Table 4-3), those panels with l
more than one gap were counted once. The data in Table 4-3 j
is used subsequently to develop the fraction of panels per cell having 0, 1, 2, 3 and 4 panels with one or more gaps.
4 0
4-5
O lv l
a -
ar ie xt 7344346565926830888486345918120 11111 1 1 11 1 111 1 11 An I -
e l
l t 5555550055505505005005050000050 e e h C) g 8351059680149012912110332683491 si 6234335455915830788486345808010 11111 1 1 11 1 111 1 11 mfeR ooh r c fmnt 5500555555505550005505550050005 s oif p nt( e 8151058680149092912110232673490 a ot L 6334335455915830788486345808010 1 1 11 1 111- 1 11 Gs io 11111 t tB ls a ae v uT e -
' d l 9999994499949949449449494444494 il Ea t va r h 4917615246705678578776998249057 g 6133235454815729777375234807019 ii lo 1 1 111 1 1 dc aB)i 11111 1 1 ne I p l sR .
xf e S Aoh 1 c 3 : mnt 9944999999949994449949994494449 n oif . . .
. ro t( e 4717614246705668578776898239056 1oi t L 6233235454815729777375234807019 111 1 1 ft o 11111 1 1 1 1 4 a B nr eoo O lip btr auo TbC el za iv sr i e 1224221114211112111212132233113 rr pt ta an se GI il Dc u
dN n e) x 0000000000000000550000000000500 al zsa a i eM 1224221114211112001212132233013 en Sh zo ii pnn c
5000005550055550005050500000050 St aai a C( M 0113110003100001000101021122002 pN a
G l
e n
a WWNWNESWWWWWNNN8NNNNENSNNESSNE P
D IW 2788664222154444771214666655443 45555222222555555525 111444555 l
l 6223799776671225124565334443166 eN 1111 1
- C r
ae tb .
am 1234567890123456789012345678901 O Du N
1111111111222222222233 eE-
i -
ll1 iljll.
lj s
A O l n
il -
a-a rv- ., -
v er- -
r be- 9342 -
e mt - 1 t un- ,-
n NI =
hI c
ne -
Iz l i ar- w 1S v e 12 3 4 _
rb- _
p a
em-t u-G nN-I a
l e
ns at Ps e
8T el
' 2 za l iv ra sr oi pt e 1224221242122211121213223313 fc np e an GI oS i
t us .
bn 2io ri 4tt sa ei r 0000000000050055000000000000 O .lDo b
adr Tno p e) a zsN ie s
1224221242112200121213223313 aC sh c
er pn za aI n 500000500.0550000505050000050 n W ie G(i 0113110131001100010102112202 ti --
Sl N a puc hs tp aN a -
G sg e l e -
ea t2 ,
vn a l io ti cee i rn at daa la l n p uN e 2 22 Ie m n WwNWNES/WWW//SNNNNENSNNESSEN rt u a N NM 2ea -
C P /hh r tt m 2788664221544477121466665543 DN I
11144455 455522222255555525 l
l 6223799766732512456533444366 e 1111 1 CN .
~
+
r ;
ae e O tb am Du N
123456749a 123456789012345678 111111111222222222 m
78'
Table 4-3 O
Evaluation of NNC Standard Test Aesults j Cell Total Total Numbe: ,
Location Number f of Panels i i
, Data N 8 E W with Gaps Number N W ,
of Gaps 0 0 0 0 0 1 1 26 0 2 1 0 1 0 2 2 1 27 0 0 0 0 0 3 1 28 0 0
26 0 0 0 0 0 4 2 2 2 27 2 1 1 0 0 5 1 2 28 1 0 0 0 1 6 0 26 0 0 0 0 0 7 3 2 1 1 0 0 2 8 3 27 0 0 0 0 0
- 9 3 28 0 2
4 20 2 0 1 0 1 10 0 2 21 2 1 0 1 11 4 2
22 2 0 1 0 1 12 4 0 2 5 20 2 1 0 1 13 2 5 21 2 0 1 0 1 14 3 5 22 3 1 1 0 1 15 2 5 23 2 1 0 1 0 16 2 5 24 2 1 0 0 1 17 3 18 6 20 3 1 1 0 1 2 2 a >
- 2 <
O 2*
20 6 22 22 2
1 0 0
1 1
0 0
0 2
1 2
21 6 23 3 1 1 1 1 4 22 6 24 4 22 2 1 0 1 0 2 23 7 7 23 2 1 0 0 1 2 24 25 2 24 4 1 1 1 1 4 0 0 0 0 0 0 26 1 17 0 0 0 0 0 0 27 2 17 0 0 0 0 0 0 28 3 17 !
4 17 1 0 0 0 1 1 29 1 5 17 1 1 0 0 0 30 0 6 17 0 0 0 0 0 31 0 7 17 0 0 0 0 0 32 0 33 8 17 0 0 0 0 0 18 0 0 0 0 0 0 34 1 1 0 0 0 1 1 35 2 18 0 0 0 0 0 36 3 18 0 37 4 18 0 0 .0 0 0 0 0 0 0 0 0 38 5 18 0 0 0 0 0 0 0 39 6 18 0 0 0 0 0 0 40 7 18 18 1 0 0 0 1 1 41 8 1 0 1 0 0 1 42 9 18 s
O 4-8
Table 4-3 '
O Evaluation of NNC Standard Test Results (codtinued)
Cell Total [ Total Numbe-Data Location Number r of Panels .
Number N W of Gaps N 8 E W with Gaps 43 10 18 0 0 0 0 0 0 44 11 18 1 0 0 0 1 1 45 12 18 3 2 0 1 0 2, 46 13 18 3 0 2 0 1 2 47 9 17 0 0 0 0 0 0 48 10 17 0 0 0 0 0 0 49 11 17 0 0 0 0 0 0 50 12 17 2 0 1 0 1 2
' 0 51 13 17 0 0 0 0 0 52 19 3 3 1 1 0 1 3 53 3 29 0 0 0 0 0 0 54 16 1 1 0 0 0 1 1 55 16 2 2 0 0 1 1 2 56 16 3 1 0 0 1 0 1
- 57 16 4 0 0 0 0 0 0 58 16 5 1 0 1 0 0 1 59 16 6 0 0 0 0 0 0 60 16 7 0 0 0 0 0 0 O 51 62 15 16
=
9 2
1 1
1 a
0 1
0 a
0 2
1 l 63 16 10 1 1 0 0 0 1 64 16 11 0 0 0 0 0 0 65 16 12 0 0 0 0 0 0 66 16 13 0 0 0 0 0 0 67 16 14 0 0 0 0 0 0 68 16 15 0 0 0 0 0 0 69 16 16 0 0 0 0 0 0 70 17 1 0 0 0 0 0 0 !
71 17 2 0 0 0 0 0 0 !
72 17 3 0 0 0 0 0 0 73 17 4 0 0 0 0 0 0 l 74 17 5 1 0 0 0 1 1 75 17 6 3 2 0 1 0 2 76 17 7 0 0 0 0 0 0 77 17 8 3 1 1 0 1 3 78 17 9 4 1 1 1 1 4 79 17 10 3 0 1 1 1 3 80 17 11 0 0 0 0 0 0 81 17 12 0 0 0 0 0 0 82 17 13 0 0 0 0 0 0 83 17 14 0 0 0 0 0 0 84 17 15 0 0 0 0 0 0 85 17 16 0 0 0 0, 0 0 O
t..
/
/
e Table 4-3 Evalua' tion of NNC Standard Test Results (continued) !
Cell Total Total Numbc:
Data Location Number W
f of Panels with Gaps Number N W of Gaps N 8 E 1 0 0 0 1 1 86 6 51 ;
6 52 2 0 0 1 1 2 j 87 53 2 1 0 1 0 2 88 6 51 1 0 0 1 0 1 89 7 7 52 2 0 0 0 2 1 90 -
53 4 1 1 2 0 3 91 7 8 51 1 0 0 0 1 1 92 52 0 0 0 0 0 0 93 8 94 8 53 2 0 0 0 2 ' 1 95 7 44 3 0 0 2 1 2 7 45 3 1 0 1 1 3 j 96 7 46 5 1 2 1 1 4 97 98 8 44 1 1 0 0 0 1 99 8 45 3 0 1 1 1 3 100 8 46 2 0 1 1 0 2 101 9 44 2 0 1 1 0 2 102 9 45 2 1 0 0 1 2 103 9 46 1 0 0 1 0 1 104 1 54 1 1 0 0 0 1 0
- 105 106 1
1 55 56 0
2 0
1 0
1 0
0 0
0 0
2 107 2 54 6 3 2 1 0 3 106 2 55 1 1 0 0 0 1 109 2 56 2 1 1 0 0 2 110 3 54 3 2 1 0 2 111 3 55 2 1 1 0 0 2 112 3 56 2 1 1 0 0 2 >
113 4 54 2 2 0 0 0 1 114 4 55 3 1 1 0 1 3 1 115 4 56 3 1 1 1 0 3 l 116 5 54 2 1 1 0 0 2 117 5 55 2 1 1 0 0 2 i 118 5 56 1 0 1 0 0 1 Total Number of Gaps = 155 Number cf Panels = 141 with Gaps . ;
i j
Number of Panels / Cell ,
with Gaps Number of Calls '
0 45 1 25 2 32 22 O
4-10
9 ._
KGE ...
_ CL ~
N A .
R-
- O I
S TT E
- UNG BEA-I RM R-TE O r
1
- SRU TS I
D SL
^ 4EAE ZEU 7 5
o EIS MF
=_R LT UP GAI AN FG C E I
EP X PS E S S
_ L
_ F CEI -
. A NT R
. I O NC B D - -
A - - - - -
8 6 4 U02 8 1
6 1
4 1
2 1
0 1
_ Q O _
m k o u O $i @ v Z
,A"
O _
K D " .
CNE 3
__ 2 1 o A GS g Hl t
_ REL MmEPJ E SS
_ PT E~
NG- -
_ A )
$c GM EA- R- - _
FEO- -
_ :OR T- S-F< r'i.ii' g$a h <
2 U -
- 25 NSL- -
7V/'d 4 OAE T EU I
_ 7//
E U MF - .
1 O R B -
_UIRAN-LT- -
7f ;
- i G TIC E-
_ I I
FS EP
_ I ). ;::j ,
i/
DPS-S -
- $p L S -
ACE I I
._ XNT -
ANC I
_ D- -
A ^ - - -
_ U-7 ^
6 5 4 3 2
. Q' _ $5 O
26 ioa :al2Dz 1
}
. , ! t ,i!
rC i
I l
O I
5.0 AD.D.II E I2IL RACE MANhPACTURING PROCESS An ' audit of the fuel rack manufacturing process used for the Quad Cities racks was conducted at the Joseph Oat l
Corporation. The purpose of this audit was to determine whether any fabrication methods or processes may have contributed to the formation of gaps in the Boraflex.
During the audit specific attention was paid to:
o Method used to afix the Boraflex to the various Tee, E11 and cruciform' elements o Welding process and potential for overheating the Boraflex
({} o Any other factors such as the use of clamps, etc.
which could contribute to the obssrvad Botaflex beha-vier The major conclusions of that audit are summarized belows o The adhesive used to hold the Boraflex in place during manufacture was Dow Silicone 999 whereas BISCO had tested Dow Silicone 732 (Ref. 3). The difference be-tween these two adhesives has not been evaluated although Joseph Oat stated that they are similar.
o The adhesive was applied to approximately the center of the cavity in a discontinuous bead along the entire length. The bead was spread out to a width of approx-imately 2-1/2*--3' with a stainless steel scraper. The O - - - ~
5-1 " - - -
}
1
() F I l 1
i, Boraflex was then rolled.into place and pressed cgainst I
thestainlesssteelE11sub-elementandallowehto i.
cure. There were no specific procedures for this process since the only intended function of the adhe-sive was to hold the Boraflex in place during assembly of the Tee, E11 and cruciform elements.
o While the use of discontinuous strips of Boraflex cannot be ruled out, it is unlikely since the Boraflex is received in full lengths for the various matching stainless steel components. j O
,I O
5-2
O
[
.?
6.0 EVALUATION g RADTATTON TESTING E BORAFLEX 6.1 fuum&I.Z 21 irradiation Iggg ggpm As part of a larger program to-qualify Bora*1-- fos-use.
inespenbfue&-estereT*"raekry* BISCO sponsored a series of i r r adiatian-teste-et-the- Feve- ReesteNbe= ca&ees sity. oQ w Michigaa-,(Ann Arbor)4 The purpose of these tests was to d==n_mer=*= *k= radiattee-stability-e6-Besaflaz The tests were ca dua*-' i- -
roastee-fee 6&ity-te~esee6eer "" __
accumul.atio_n __ of- umaxp,gAggs Accordingly, it must be noted that diffar==a== i= di seeket6ee-enhesist %
between-the-twet-experiinvass anu unir vv.. shi.. .,.... fue A paa' There are probably differences in the gamma spectrum in the test reactor and in the Quad Cities pool.
Additionally, the a"=a* a' - '---
'-- p-to--the= test irradi=*ian i= aa* k a a=_a .
In these tests, samTT"samptes=oP9ere6&es (of both 25 and 40 w/o 54C) approximately 6' in length, .25' in width and .100* in thickness wooe-isse46etee-in '
di=*111gd
- v. u ..a h.,a*.a.ua*.c tonnn ra-s *a ----- --- ia *ha
_ range ,
og A s r in10 ,, n a, , sall_ rada, The samples were characterized both pre and post irradiation by physical dimensions, sample weight, specific gravity, hardness and tensile strength (not all samples were tested for tensile strength after irradiation). During irradiation, Boraflez
~6-1
O samples in each of the three environments were morIitored for volume, rate and gas evolution in terms " of total l
composition. In addition, ona sample was irrad ated to a 8 rads in air) in the spent fuel storage low dose (2.81 X 10 Additional data on area of the For'd Nuclear Reactor.
have been reported in the irradia.tion of Boraflex litsrature ,6 5 and have been included in this review.
The- me asu r ements- of- phy s i- =' di--- = i a = = sho t in most _
w =*iaa- The cassa_A naLabidage-of-the-samples-aftee-iss _
about 2-34 data is variable but the general trend is The accuracy shrinkage in width and up to 84 in thickness.
of these measurements is not known but it is suspected that accurate dimensional measurements on small samples would be difficult.
This is particularly true in the pre-irradiation state when the material still has the properties of an elastomer.
Some general trends in the data includes o Increase in the specific gravity after irradiation o Increase in shore A hardness o Tensile strength variable due to sample configuration (some indication it increases with exposure) o Sample weight variable (some increased, some decreased after irradiation)
During the irradiation of some of the samples, offgas
/ produced when Botaflex is irradiated was collected and O
6-2
i O- .. ;
analyzed. The offgas consisted primarily of E2 vith some and hydrocarbons. The' N,0 2 and lesser amounts of'CO, CO2 2 f source of N is not clear, however, potentiak sources 2
include air entrained in the samples during manuf acture or leakage in the sampling lines. The other off gas products would be expected based on the radiation are what i' radiation damage mechanisms discussed environment and 7
- subsequently. For the samples irradiated in air at 7 X 10 rads / hour, the gas evolution rate diminished to zero after 10 approximately 150 hours0.00174 days <br />0.0417 hours <br />2.480159e-4 weeks <br />5.7075e-5 months <br /> (accumulated dose of 1.05 X 10 rads). The samples irradiated in distilled and borated water showed continued ges evolution which is believed to be due to radiolysis of the water in the reactor environment.
Q 6.2 Zy.gloatten af the DAAA Since the physical dimension data may not provide a indicator of the total extent of Boraflex reliable l j
shrinkage, the weight and specific gravity data (pre and post irradiation) from References 4, 5 and 6 have been j
evaluated. Table 6-1 contains a summary of all the data reinted to weight and specific gravity changes. The sample volune change based on both weight and specific gravity changes as well as changes in the specific gravity only have changes in volume have been also been computed. The computed as:
1 O
/
6-3 f
~
O .
I AV/V = (V, f - Vg )/Vg where: V, = final volume, post irradiation y = initial volume, pre irradiation r[
The data contained in Table 6-1 has been plotted in Figure 6-1. Review of Figure 6-1 indicates the following:
- 1. There are no data between 2.8 X 10 8
and 1.5 X 10 10 1
I rada.
s 2. It appears that initially all the samples underwent a reduction in voluze. From the data it would appear 10 that at an exposure in the range of 1 to 2110 rads (gamma), the volume reduction ceases and the samples begin to swell.
- 3. The extent of apparent swelling tends to depend on
() whether the samples were irradiated in air or in water 11 (either distilled or borated). At 1.03 X 10 rada, che samples which were irradiated in an aqueous envi-ronment show the greatest extent of swelling when the net volume change is computed using the weight and specific gravity data.
Further to the first point, References 4, 5 and 6 refer to other data and continuing testing although they have not been identified as having. been reported in the literature or, for that matter, in reports issued by BISCO. It should be noted, however, that a detailed search of the literature has not been a -t;.ted at this point.
O s O
O
~
Fromthedataitappearsthatintherangeofflto2X 10 10 rads, a maximum volume change occurs. This might be consistent with the gas evolution measurements which indicate that gas evolution ceases at approximately 1 X 10 10 l
rads. It is postulated that crosslinking of the polymer is
)
complete at this point (all available sites for crosslinking w
expended) as discussed in section 7.0. Crosslinking is reported to occur at sites where radiation has caused I release of B or CEj.
The samples ' irradiated in distilled water or borated water show the greatest apparent swelling at 1.03 I 10 11 O rads when ,o 1 . e chan,es ca1cu1sted using the s ,ecific gravity and weight data are considered. One possible explanation for this is that the samples are developing some l porosity and absorbing water. A Boraflex panel at the Point Beach Station has recently been removed from the spent fuel storage rack and examined after a gamma exposure of 1.0 x 10 10 rads.7 In addition, surveillance coupons have also been removed from the pool and inspected. The coupons were removed from the pool after 1.6 X 10 10 rads exposure, dried and radioassayed. These measurements indicate residual beta, alpha and gamma activity, presumably arising from containments in the pool water. Since the sample weight measurements in Table 6-1 may noc represent the true weight O < 9 tv > ==a ri11 r,is i 6 11 a sh * **
- ti st-6-5 i
O .
indicator of volume changes versus exposure can be derived from the specific gravity measurements.
k 1 The results of , these data are used subsequently to estimate the maximum overall magnitude of vo5uma reduction of the Boraflex in the Quad Cities spent fuel racks.
h O .
l
& 4 O
6-6
O i2 X-
/- e m __
u 9
0 0
7 6
0 59 40 01 501 387 110 9
5 1
9 0
1 55 03 01 694 971 110 9
0 11 11 781 111
) - 000 i- l 00 000 0 0 00 000 0 00 e X- o 0 0 r - V - - - -
u - -
s - '
13 o f -
1 -
613 5 958 1 p X e 7 0 1 163 3 72 x ( m 0 3 568 110 0 110 e0 10 E u 0 1 e- l 0 000 0 00 n r- o 0 0 000 -
o u- V - -
i s-t o-p-
a 598 i x- c Eiy_ 9 2 82 597 9 2 57 984 6 52 33 122 d 517 8 2 05 211 0 a - ft_ 0 7 42 220 1 11 221 120 1 1 01 r s ii_ 0 0 '01 r u 00 000 I s r-cv_
prea_ 0 0 00 000 0 0 00 000 0 s e SG_
u v- -
9 s -
4 532 4 274 7 r s- t_ 2 2 221 7 812 6 1 e e- h_
g_
0 000 1 000 1 0
' v g-n- _
0 0 0 0 0 000 l
x e
a-h-
C-i_
W e 0 0 000 f
a r
o 8 0 36 011 2 0 48 667 0 07 484 B t_ 3 2 85 514 9 21 335 c s_ 416 8 3 79 008 9 99 000
~
iy o_ 6 9 78 909 0 0 69 1f o ft)P_ 11 121 2 2 11 221 1 11 212 6 iic 1 1 s cvc ee ea/ 351 0 33 481 O lg bn aa prg SG( o_r P_
3 7
6 0
0 8
30 99 66 901 752 668 9
0 7
0 0
8 33 99 66 752 668 0
8 99 66 752 668 Th 11 111 1 1 11 111 1 11 111 C 1 1 e 9 976 7 978 4 4 m t_ 7 7 181 1 6 u s_ 0 8 0
725 819 1 006 7 1 l to_ 0 o hP_
g 5 3 533 4 933 3 3 V i d e 6 075 7 358 2 4 n W e_ 8 6 350 5 330 8 0 a 9 5 307 1 1 r_ 9 1 000
, P_ 634 3 833 3 3 y 4 3 t
i v
a 001 0 0 01 001 0 01 00 r e - 8 0 01 11 111 1 1 11 111 1 11 111 G r) us 0
+
1
+ ++ +++ + + ++ +++ + ++ ++
EE EEE E E EE EEB E EE EE c sd E E 093 0 0 03 093 0 03 09 i oa pr 1
8 0
5 03 60 640 0 5 60 640 5 60 64 f
i x( _ 5 1 11 121 1 11 12 c E _ 2 1 11 121 .
e p e_
S ac_
tr 5 6 44 444 6 44 44
, au 4 6 44 444 t
h g
Do_ S i tC _ 0 55 000 + 55 000 55 00 44 e 4 22 444 22 444 22
/ 4 _ d W WB _ e a_ r tr
.i e ae rt O rd re IM i
A r t W
a oa BW
y f
e e ,
O h
s . *- (
3 s * . . . . a s, $ . * , e .a.
. , , .J, :j41
. t as
- * *. e .. ; irJ s(-Lv.7,T:
p b
I; ,d ,,
.. A M
~
s b4 1 y.u._-- ,.
g D r.
M e .
6 M- i. - ca C
M
, ca a t W
I w
M E
M m M C
> s
- k g
4 O '
R w
-ce C
N O ,a M
U f
5 5
= o M be i a e em M
N
- o e E
ed D s
@ ,b<
Q em U C b ,y
- G tw t -e e U {
- b. E
.c i E< C
& C A q 80 e. (
g .4 cd g {(
O c e e' 1 E. o
(
i 4 9
O menvas sursea aan 2
E
, C f-I S
O.
I 7.0 POSTULATED RADTATTON DAMAGE MEcRANTEws 3 BORAFLEX M ETMTLAR ELASTOMERS 7.1 Phvnical Prenerties g Filled u d Unffiled Methvinted Polvsilevano an elastomer (methylated polysiloxane)
Boraflex is filled with finely divided boron carbide 4 (B C) powder.I Although the repeat unit is proprietary to BISCO, it is
' assumed to be very similar to polydimethyl siloxane as reported in the literature. The elastomer is intended to serve as a matrix to contain the B 4C powder, which by virtue of the large thermal neutron absorption cross section of B-10,'provides reactivity control when this material.is used O in spent fuel storage racks. Boraflex can be manuf actured over a range of composition (B 4C content) and thickness.
8 Typical composition for spent fuel storage applications is :
Boron 31.5 w/o Carbon 19.0 w/o l
' Silicon 24.5 w/o Crygen 22.0 w/o Eydrogen 3.0 w/o In its as-produced condition and unfilled, an ideal elastomer posesses the following properties's
- 1. It must be rapid stretchable to very high extensions (on the order of 5004).
- 2. It must possess high tensile strength when fully stretched.
1 O
7-1
+
O I
3.
It must snap back (quickly) whenthestressy re1ea,ed.
4.. It must retract completely (no permanent set).
When the above behavior is observed, certain molecular -
and environmental conditions usually existI :
l
- 1. The material is a polymer.
- 2. The material is amorphous, i
- 3. The temperature is above gT , the glass transition tem-perature. 1
- 4. The material is lightly crosslinked.
Fillers in methylate'd polysiloxane polymers such . as l O *araa = r$ta = a4 *= i=>ra
- a 12 =*r aven aa taar -
~
the elastic modulus but would be expected to reduce extension and elongation to break. For a B-10 loading of 0.020 gm B-10/cm 2 at 2.5 mm thickness as-produced Boraflex exhibits the following propertiesI: _
Modulus of elasticity 1000 psi Tensile strength 200 psi l specific gravity 1.7 gm/cc Bardness 75 shore A Temperature stability 200*C minimum without i
variable distortion The above properties are according to the manufacturer's specifications for typical as-produced Boraflex.8 7-2 O
v
. l 6
O Radiation DAEASA neehanisma g polymers 7.2 ,I i
Prior to discussing postulated radiation damage ;
mechanisms in polymers, it ks useful to define nits and terms used to describe the effects of radiation on changes l in the physical properties.
Roentgen is defined in terms of the number of ion pairs As such it depends produced in air by lonising radiation.
on the ionization enerijy, and it requires conversion ' and
' adaptation to apply it to other materials rather than air.
Typically it can be conveniently used to describe a gamma ray field. Depending on the agreed value of the ionization energy for the ion pairs, the energy value is 80-90 ergs /gm (energy depositedT.
energy O Rad is a term that signifies 100 ergs of deposited in a given material by a given radiation environment. It simplifies some of the technicalities of describing radiation effects and is rather widely used. For irradiation of materials, many types of observable damage tend to occur at the negarad (10 0 rads) region and beyond.
When a polymer such as Boraflex is subject to a radiation field, changes in the atomic / molecular structure occur. Radiation results in the breaking of atomic bonds crosslinking between atoms in adjacent and subsequent polymeric chains can occur. Figure 7-1 shows the molecular structure of as-produced methylate polysilozano. The spine l'
of the polymer is comprised of the repeating chain of 81-0 0
7-3
O
~!
atoms. Attached to each Si' atom are two CH 3 yadicals.
Figure 7-2 illustrates possible crosslinking mechMisms in methylated polysiloxane. In the first ' case (left hand side), radiation results first in the release of a hydrogen radial and subsequently the formation of atom from the CH3 radicals in adjadent chains. In crosslink bonds between CH2
- the second case, CH 3 radicals are released from the main chain and crosslinking between 81 atoms in adjacent chains occurs. Crosslinking is characterize 3 by the release of E2 '
- CH and possibly C 26 5 . Typically, as crosslinking occurs, 4
the chains in the polymer are physically pulled closer
=dergoes shrin= age -d -
O together -d the ateriat increase in density.
The other radiation damage mechanism in polymers is described as chain scissioning in which bondsi between atoms I
in the polymers are broken and crosslinking does not occur.
An example of scf.ssioning in methylated polysiloxane would be the breaking of the 51-0 bonds (main chain scissioning) .
l In general, for crosslinking polymers' when there are many sites available for crosslinking (at low doses),
i crosslinking is the predominant effect. As potential sites for crosslinking are expended (or saturated), crosslinking decreases and scissioning and other types of degradation become the predominate damage mechanism.
A 'G value' is a description of the number of events of Q
i 7-4 l _
~ .
L i O -l I I a given type that occur when 100 eV of energy is deposited.
For instance, G(scission) =2impliesthat2scissihnevents a of molecules occur for each 100 eV absorbed.- For crosslinking of polymer molecule repeat ur.its , which are already incorporated in molecular chains, G (ILU) can be defined as the number of polymer repeat units which have crosslinked to others for each 100eV of energy been deposited by a radiation field. G(XL) refers to the number of crosslinks formed and is one half of G(XLD) since two units are involved in one crosslink.
Several investigators have studied the behavior of methylated polysiloxane both filled and unfilled (see e.g.
O References 10, 11, 12) at low doses (megarad range).
-l conclusions which can be deduced from this work includes l o Crosslinking is proportional to radiation dose.
o Crosslinking is independent of the type of radiation f (electron, reactor, deuteron and gamma) as well as the intensity of dose (from negarads in a few seconds to a few hundred negarads in days).
Furthermore, at low doses it appears that G(EL) for methylated polisilozane is in the range of 2.5 to 3.0. At higher doses and as the available sites for crosslinking are expended, G(KL) approaches sero. If for the purposes of itiustration we assume the average value of G(EL) over a
~
large range of exposure is 1.4, then the 'irradiaton dose O
7-5
...d
O .
involve potential units in in rads required to all crosslinks can be calculated:
1 rad = 100 ergs /gm 1 megarad = 10 8 ergs /gm = 1020 ,yjg, 1.6 .
\
For G = 1.4, then for every eV absorbed we haves f 1 megarad -> LA 1 10 18 crosslinks 1.6 Polysiloxane (repeat unit) has a molecular weight of about 74 gra and therefore:
l 1 megarad - > 1.417411018 = .65x1020 crosslinks/ mole 1.6
~ 23 units, we can calculate the Since a mole contains 6.023:10 dose to crosslink all units:
6.023r10 23 creamlinks/ mole = 9.3 x 10 3 magarads
.65x10" crosslinks/ mole negarad So that approximately 10 10 rads is required to crosslink all units in methylated polysilozana. :
10 rads),
Long before the accumulation of this dose (10 the polymer will probably be severely changed and the G(1L) value will have changed. One notes that if the G(IL) decreases measurably because favorable sites for crosslinking have decreased, and, at the same time G(scission) remains about the same, then scissioning might predominate as the accumulated dose increases. Polymer degradation would then be expected.
O 7-6
O- .
I anfeal 7.3 Ef fects d Radiation h h Physical gy1 Mee Preeerties g Methvinted Polysilovana The data which are currently available relative to .
J changes in physical and mechanical properties of Boraflex with increasing irradiation dose are somewhat limited. It reported data for j is therefore useful to examine .some
- methylated polysiloxane found in the literature. I Hardness, Tennile streneth, gyi Elonention 3 M Warwick (see e.g. References 10 and 11) studied the i
(megarad range) ganaa and electron effect of low dose
==ai ei== = iti== ritt a ==a r$ = ai22 4 **rt== a O It should be noted that in these experiments l polysilozano.
the polymer was prepared with fillers in the range of parts per 100 parts of polymer which is significantly less than The results of hardness, Boraflex (40 w/o 3C4 filler).
tensile strength and elongation to break are shown in Table increases with accumulated dose 7-1. The hardness 4
consistent with the data presented for Boraflex . Tensile.
strength increases with dose initially and then decreases as the dose increases. Elor.gation to . break initially increases at low dose and then decreases to about a fif th of its initial value at 40 megarads. While there are differences in the filler composition used in these tests and that of Boraflex, similar trends would be expected in Boraflex. It 7-7
=\
O .
i is also difficult to ascertain the true dose in! rads for rays are l
every case, especially if both neutrons and gammt '
al present plus other effects.
Elastic Modulus:
12 studied 'the effect of low.
In another report Bueche dese (up to 60 megarads) electron radiation on polysiloxane He including a series of unfilled methylated polysiloxane.
found that, except for very low doses where presumedly the change in l polymer chain end effects are important,
elastic modulus was well represented as a linear fit of the' data with positive slope.
O The theory of tem,erature -d crossunung on the I tensile properties of polymers evolved from the analysis of the entropy change. For an initial unit length, ais the e,
length in the stressed condition and the extension, f, for a simplified equals a-1. Then the applied stress, 3-D crosslinked network is:
(7~1) f = dF/de = -T(ds/de) = jo(RT/M,) { a- 1/s2) where:
F = free energy a = entropy T = temperature R = gas constant M = molecular weight between crosslinks M= density For small extensions we can write:
7-8
O I
f = 3joRTe/M g (7-2)
Therefore E (modulus) = 3 jORT /M, (7-3) or:
E = 3 joRTq/w (7-4) where: q = crosslinking density (fraction units crosslinked) w = repeat unit molecular weight and: E = 3,A5R?e r = 6.24 X 10-6 joRTrG(XL) (7-5) w where: r = absorbed dose (megarads)
R = gas constant G(XL) = G-value h g ggg,ggg h feradiation:
The test irradiations conducted by BISCO were carried out in essentially a stress free condition. The use of an adhesive in the Quad Cities fuel racks suggests that as the Boraflex shrinks tensile stresses accumulate assuming there is still some type of bond between the Boraflex and the stainless steel structurae while the effect of irradiation on the condition of the adhesive is not known, it is possible that the Boraflex at Quad cities is accumulating dose under tensile stresses.
13 under Bopp and 81sman irradiated polysilozanes compressive stress in a jig. After irradiation the jig was removed and the percent of recovery to initial dimensions was measured. A control sample was also examined which was not irradiated. For the control sample (0 megarads) the f
, 7-9
/
e t F e
O -
I I
l percent recovery to the initia1 configuration was 984. At a dose of 40 megarads recovery _was limited to 2ft of the ,
l initial configuration. For a dose o,f 230 megarads the recovery was approximately 04. While these data are for compressive stresses, .it is possible that overall volume changes experienced by Borafles under tensile stress may not be the same as when irradiation occurs in the stress free w
condition. .
Temperature Effects 14 has studied the 6tfect of one investigator temperatures on G(IL) in methylated polysiloxane. Using 2 mov electrons at one negarad/sec., G(XL) varied with temperature shown in Table 2. The data suggests that G(XL)
< increases monotonically over the range of temperatures studied.
The data presented are for materials similar to Boraflex (i.e., similar polymer) but with different fillers (and unfilled as well) and are at low doses relative to the Borafles test irradiations. The data have been included to provide some basis for interpreting and evaluating the behavior of Boraflex in the Quad cities spent fuel racks.
It should be noted that the literature cited were only those
~
at hand at the time of this evaluation. A more detailed and extensive search of the literature is planned as part of j '
O 7-10 t
i l
l 1
.o
?
another project in the near future.
f ,
4 O l
\
i 1
I O
7-11 '
8
- Me
O l O- .! ,
T5B'LE 7-1 h n Effect of Low Dose Gamma and Electron Radiation ggnyg Silica and Carbon Filled Methylated Polysiloxane
. Megarads Hardness Tensile Elongation Gamma or Strength to Break Electron psi w
Silicon filled: 1 18 135 550 2 15 153 750 5 27 1180 750 6 26 742 605 10 29 876 580 20 43 679 250 25 53 916 158 40 52 561 117 Carbon filled: 6 26 709 805 )
10 35 7 87 435 Q' 20 47 581 200 i
TABLE 7-2 Effect of Temperature on GiXL)
Temperature, 'C G ('XL) 150 4,7 20 3,1 0 2.8
-78 2.6
~*
O 7-12
?
.y j
i a
'0 5 O S m
' m ,
B
.i 5 m 5 J
U 0
=
E M g m '
I O 8 g U- U1 - O E
$ O r-fa3 g O -m-O E
6 7-u 1 l
O .
O q I
- e 5
o e D.,-m6 ;
U l 5 O O .
0
,~
k m y i
5 f e 5 enj
=
U
=
I o .
8 5-m 5A d
= o o <
D b b
~
0 l O , -.
7-14
1
.\
l O
I" f
8.0 Q&2 FORMATION,-G&R GRONTEM'LQ]E*(2 Egg INTEGRITY'gg*
AgAA. M IM &-128M2'ED.EL'20QErENYTRONMEffbe j From the outset it should be noted that the mechanisms for gap formation and gap growth described are preliminary as the extent of data currently available is limited. As j such, any conclusions drawn from this material are I
preliminary and may change as more data relative to Boraflex behavior under irradiation is documented. Further. j experiments will' probably be required to ~ determine the l causes for all effects noted.
O 8.1 Potentini Mechanimma af Gag Formation A major uncertainty governing the mechanism of gap 4
formation is introduced by a lack of knowledge as to the condition of Dow silicone 4999 adhesive used to afix the Boraflex to the stainless steel cell during manufacture of I
the racks. Nevertheless, it is useful to postulate three different bounding scenarios including:
o The adhesive bond completely breaks down at low doses of gamma exposure.
o The bond between the Boraflex and stainless steel is !
,e- " perfect's uniform and the mechanical properties of the Boraflex are uniform along its entire length.
Q o The bond is intact at the' ends of the Boraflex sheet
/
8-1
m O
(i.e., region expected to receive the lowest dose) and has failed or partially failed in the central region. F R In the first scenario, since the bond is postulated to ]
have f ailed along the entire length of Boraflex, the sheet I
is unrestrained and would be expected to shrink in a stress free condition as irradiation proceeds. A not shrinkage in all three directions would occur but tearing of the sheet !
' and subsequent gap formation would not be expected. ,
If the bond were ' perfect", as is the second scenario, high local stresses would develop in the sheet as the material tries to shrink and one might expect the material .
to tear at many locations forming many small gaps along the O 1enseh of the sheet.
In the third scenario, with the Boraflex sheet restrained at the ends, and as the material shrinks, the greatest accumulation of local stresses would be expected at ,
the midplane of the sheet. Therefore, the sheet may preferentially tear at or near its center. This scenario seems to be supported at least partially by the National Nuclear Measurements of gap occurence varus axial elevation (see Figure 4-2).
Evaluation of the distribution of local stresses is further complicated by the manner in which the adhesive was applied. In section 5.0 it was noted that the adhesive was applied in a strip 2-1/2' to 3' wide in the center of, the O
8-2
- ges> ammm
,, i 1
l O
F Boraflex sheet more or less continuously along the length, i The center of the approximately 6" wide panel over 2-1/2" to' l 3' is therefore initially bonded whereas the outer 1.5' on j each side of the width is'not. The outer 1.5" on each side may be under a lower tensile stress condition than ' the -
central portion. Non uniform axial shrinkage across the width of the panel leading to relatively high shear stresses I
could then be a factor leading to incipient tearing followed by gap formation and growth. These materials may be weaker under shear stress than under tensile stress.
For the purpose of illustration, consider the Borafles ;
sheet as essentially a one dimensional stress problem and we Q
write for the stresses in the axial directions a=Ec where E is the elastic modulus and c is the amount of strain that would have developed if the material would have been .
allowed to shrink stress free. Further for the purpose of
- . illustration, it is assumed that after a tear occurs, the l material."anaps back" to the length it would have assumed in the stress free condition. Then for a 1.5" gap the strain is .01 and based on the as-produced value of E = 1000 psi, the stress 12 10 psi. This value is well below the i as-produced tensile strength of 200 psi as well as the tensible strength for ir radt .x ced Boraflex cited in
~
' References 4, 5 and 6.
i q
8-3
- .I
I 1
1 1
l fi If the effect of irradiati'on on the elastic modulus is the m; considered, a crude order of magnitude estimath of change in elastic modulus can be made using Equation 7-5.
If we assume that the G value in Equation 7-5 does not 6
to l chsnge, then for a two decade increase in dose (say, 10 10 8
rads) the elastic modulus increases by a factor of The observed increase in Shore A approximately 100 N
hardness suggests a large increase in modulus. Under these conditions, 0.01 strain would result in a stress of 1000 psi which may be in excess of the tensile strength of irradiated ~
Boraflex. Even if the G value decreased by a factor of 2 in going f rom 10 6
to 10 8 rads, an increase in the elastic O modulus of 50 would be indicated. In this case .the estimated stress is 500 psi.
Equation 7-5 also indicates that the elastic modalus is a function of temperature. When a freshly discharged fuel assembly is placed into a storage cell, the temperature at the top of the cell is higher than the bottom of the cell due to residual decay heat of the fuel. This may be a f actor influencing the axial distribution of gaps as shown in Figure 4-2 (i.e., no gaps in the first four axial intervals and a peak at the top) . Other factors which may axial distribution of gaps include the influence the temperature dependence of the G-value and a non uniform gamma dose distribution along the fuel assembly. It was O
8-4 e
O ni noted in section 7.3 that the G-value increases with !
temperature which would suggest more cros linking, additional shrinkage and greater increase in th elastic n modulus at higher, temperatures.
Owing to the lack of specific data for Boraflex at low dose as well as a lack of knowledge of the conditon of the silicone adhesive, the discussion presented is only qualitative. It does however suggest a mechanism (s) by t which local stress could exceed the yield stress of Boraflez leading to the inception of tears. It also provides one potential explanation of the observed axial distribution of gaps observed in the special Tests conducted by National Nuclear Corp.
O 8.2 Ert f ante g Marimum fa&D h 1
Although the irradiation data for Boraflex is limited and certain material properties must be assumed, a rough estimate of the maximum gap size can be made. Data from gas evolution measurements as well as calculations of the dose required to crosslink all available sites suggests the crosslinking (and hence shrinkage) is probably complete at an exposure of 1 to 2 x 10 10 rads.
Figure 6-1 shows that at this exposure the maximum volume reduction is approaching 206. If for the moment it is assumed that volume changes are isotropic, this would correspond to a change in any O ai ea toa or th nor e2 = aeet or 5 55'- ror - aeet or .
8-5
\
l
)
O
'I Boraflex 152 inches long, th[s would correspond to a maximu=
gap of approximately 10 inches. Bowever,thiseshinatemust be qualified with the following caveats:
o The assumption that volume changes are isotropic may or may not be true. In one report it has been noted that changes in sample thickness are greater than changes in sample length and width6, o The estimate of maximum gap size is based on data ob-tained from small sa=ples irradiated under essentially t
stress free conditions. While the condition of the silicone adhesive is not known, it is possible that the Boraflex is either under stress or was previously under stress resulting in a tear. The estimate of a 10' gap assumes the Boraflex snaps back to its length had con-ditions been stress free. Furthermore, uncertainties may be present owing to the extrapolation of test data from small test samples to a 152' length of Boraflex.
o it is not known what is happening at the upper and lower ends of the Borafler. If the adhesive 'gives',
then the ends may tend to pull toward the center limit-ing the maximum gap size to somewhat less than has been calculated here. '
o The test data on which the estimate is based were ob-tained in a test reactor where neutron radiation damage i
may have been a factor in addition to that caused by O
8-6
O 1 k
the' gamma dose. Furthermore, the gam =a source spectrum in the test reactor is likely to be different than in the Quad Cities Fool.
o There are no test data in the range of 2.8110 8 to 10 10 rads which bounds the range of the gamma dose to which the Quad Cities racks have been exposed.
Another uncertainty introduced by conducting the irradiation tests in reactor is the effect of neutron-alpha reactions in the entrained boron carbide. Boron-10 under i neutron irradiation undergoes transmutation:
510 + n1 -> Li7 + E6 + 2.8 May A not result of this reaction is that boron carbide powder i and pellets tend to swell and release helium as neutron irradiation progresses (see, e.g., Reference 15). It is not
]
known whether this effect contributes to the swelling observed in the test Irradiations conducted at 'the ,
University of Michigan after 2.5 1 10 10 gamma exposure. In any case, it would not be a f actor in the spent fuel pool due to the low neutron flux.
To place the estimate of maximum gap size in perspective, the effect of uncertainties must be considered.
For example, if volume changes are not isotropic, then'the estimated maximum gap size would change. Data presented in References 4, 5 and 6 suggest that the greatest shrinkage O "
8-7
I i
d v
may take place in the materials thickness. A revi of the a pre and post irradiation dimensions- in Reference 4 at an 10 rads shows typical shrinkage of 3-64 exposure of 1.6 x 10 in the material thickness but only 2 to 34 in the width.
If, for example, shrinkage in length were only 70% of that projected for the isotropic case, then a maximum estimated g gap size of approximately 7' is calculated. Whether the Boraflex is isotropic or not may depend on the process used to manufacture the sheet material. If, for example, rolling or extrusion is used, the chains in the polymer may have preferential alignment in one direction or another.
If, as has been discussed previously, crosslinking between adjacent chains in the polymer is responsible for the observed shrinkage, it might be reasonable to expect the of shrinkage to be greatest at low doses and rate dimininishing as the G-value for crosslinking diminishes.
Accordingly, it is possible that the rate ( d2 , change in dE gap size per rad absorbed) of gap growth may slow at higher doses. Unfortunately, there is not currently available low dose data to support this contention.
8.3 Is.2 ItIm Inteeritv Ag nerarier in iht St.n1 Enti i environment After crosslinking saturates, the predominant radiation damage mechanism is likely to be scissioning. At this point 8-8
~
[
O r I
from the data, i,t is not clear which bonds (Si-Cg or Si-0) i fracture most readily and have the highest frequency of fracture. For Si-C, Si-O and C-5, the bond strhgths are 58, 89, and 87 k-cal / mole. Once scissioning is the predominant mechanism, we can speculate that perhaps one of two mechanisms (or both) contribute to degradation of the polymer and the production of porosity which would allow water to enter the matriz. Scissioning (either main chain 1 or other) might cause specific atoms to be released from the polymer matriz resulting in highly localized voids, pitts-or micro-cracking. This in turn would introduce porosity allowing the. material to absorb water. Another mechanism ,
might be degradation of the surf ace of the polymer thereby O exposing the boron carbide particles to the aqueous environment. Subsequent erosion or corrosion of the boron carbide would leave a pit or void in the surf ace of the polymer. Further, degradation of the matrix would expose additional boron carbide particles increasing pitting and voids as well as increasing the total surface area of the material. At some point the pool chemistry (e.g., acidity or akalinity) any be an important f actor in influencing the rate of degradation with irradiation and exposure to the l l
aqueous pool environment. 3 L..., ;.... ... .: ,Hw=t..'.i.., ;; ..i.. 01...; ;...;1es a t- eteeeted=1emperatV f H"'Efif"Tf!""I(W5DF*WTWtone-bee-been- !
i O<- s-9 i 4
6 '
l l
f !
. '! l
.repostadMRef. 16). Tha,0ampleo-of='Sornfielmrer immersed- ,
irt, bo r a te d wateM appresinately= 3000ppeHat=249 9P" fo r* 6200* )
bourr. The,borat.ed.watas=wmaestralizetP"vittr-Ne0B=te= e pne-ofr 9.0-9.5. After testing the sample showed an average dimensional decrease along the side of the sample of 0.934 and an increase in weight of 0.244. Gas evolution from the
+
sample was measured with a total of 5.22 cubic inches of gas at STP being generated per square inch of sample surface over the entire test period. Gas evolution diminished with time with 41.3% of all gas evolved during the first quarter of testing and only 8.7% evolved in the last quarter of the O test period. The evolved gas was analyzed periodically for composition and showed approximately 20% I, 53% C=3, 54 CO 2 and the balance was comprised of various hydrocarbons.
Whether gas evolution implies thermal decomposition at these relatively low temperatures is not clear at this point. The observed gain in sample weight after testing suggests that the samples may be absorbing borated water.
hettfeeeeee-beteseen-ti. ;.. . etmdittens
- and-tL. ,..". vneseeneWikie-dif fice&+=> pee $eeblong -
tera *1hfegrity cayehose- the-teekd* * = - We have noted potential effects due to neutrons in the irradiation tests.
As noted above, chemical effects may be important as well.
The out-of-pile elevated temperature tests were conducted in O bar:.ted water with anon used to acid neutra11=e the seecimam 8-10
i l
el O ,.
8
'?
containers. Review of selected chemistry logs from the Quad cities pool indicates ph levels at the outlet o the prol 1 domineralizer to typically be in the range of 5.0 to 6.C.
The bulk pool temperature of the Quad Cities pool is l l
approximately 100'F. Furthermore, it is believed these
~
l tests were conducted with unitradiated Boraflex although.
this is not stated explicitly in Reference 16.
The inspection of Boraflex at Point Beach 7 indicates a steel gray substance on the surf ace of the Boraflex along some of the edges of the panel inspected as well as on sample coupons. The substance tended to rub off when wiped with the inspector's hand. On the inspected panel, the l Q formation of gray deposits appeared to start on the outer edge of the sheet where perhaps the surface to volume ratio is the greatest. Since chemical analysis of the gray material was not conducted, it is not possible to determine )
i its chemical composition.
Prodtq$i,qas_=of $A1,,qiggAll.,4tzshlife-06-50esilom-in%
a s m * *nal paal anrizonmant asa-aebposethle-abthis-tire __ _
The--rerrl" '
- l arge1P- pr* Tram-- ir"'whi etr"d stir-- f r om" surventran' ranpnam f r on- seves en - D . 5,- pl an t e-i s- g a the r e d and-evaluated sey-provide-some-answees-O t-11
,; .. . I
O r, ,
9.0 REAc?TVTTY Jt?rrcTS g GALS D[ TE FDrOPACK M"?PoN Ansopnrn The shrinkage of Boraflex and subsequent formation of gaps in the Botaflex absorber panels results in a c redistribution of the neutron pison material in the spent fuel storage racks.
tI%thAg3pg3319A JbA abs,en_ce o_f .
n e u tr obab sorbeh ih oah es.aere panei ra'r etttIt s**1rr*aw l o cab iner os se-in=veest ivi tr a s- Mas = an-f reereeso= 1>4he, reactivitprak the eettse storeyw ee n In the region of the fuel rack where gaps do not exist, the Boraflex is Q undergoing a volume reduction and hence the density of B-10 atoms is increasing.
This effect would tend to decrease the reactivity of the fuel / rack storage cell. The net overall effect ,however, is an increase in fuel / rack reactivity which is a function of gap size, number of panels per cell with gaps, and axial location of the gaps.
To obtain a quantitative determination of the effect of gaps on the fuel / rack reactivity, the . following calculational procedure has been developed. The' reactivity effect (delta k) of 1, 2, 3 or 4 panels in a cell with gaps in size interval 1, 2, 3, and 4, is calculated using a three dimensional PDQ17 model of the fuel / rack geometry. The gaps are conservatively assumed to occur at the fuel / rack midplane. Cross sections for the PDQ aodel are generated V.
y
t
~
O .
ti. i.
using CASMo-2E 18 ,
Given the reactivity effects, a statistic l' model is applied to compute the probability of gap occurence (frequency) versus gap size and number of panels / cell containing gaps. The NNC measurements described in Section 4.0 are used to provide discreet probability functions for
, gap size and number of panels per cell containing gaps. A computer program which utilizes random numbers is then applied to determine which events (characterized by a gap size interval and number of panels / cell with gaps) have occurred and the not reactivity effect of each. The program l
recomputes the not reactivity effect 10,000 times to obtain a good statistical sample. This calculation provides the mean increase in reactivity as well as the 95% probability, 954 confidence level. The calculational method utilizes several very conservative assumptions and demonstrates that the design limit of k,gg < .95 is still met with margin.
9.1 Reactivity cal cul at ien n The Quad cities fuel storage racks have been analyzed assuming an infinite array (in the. lateral extent) of the design basis fuel assemblies as defined in Reference 2. The design basis fuel assembly is an 8 X 8 BWR assembly containing a nazimum enrichment of 3.2 w/o 0-235 with 62
. k
~
O .2 .
8e
~ -
O fuel rods and two water rods. .The assembly is assumed to be fitted with a 11rcaloy flow channel which maxi ses the ,.
reactivity of the fuel / rack system2 . The design parameters for the reference fuel assembly are given in Table 9-1.
To assure that the actual reactivity will always be 1ess than the calculated reactivity, the following conservative assumptions were made:
- - o The water in the fuel assembly and rack is assumed to be demineralized, unborated and at full water density o In the x and y directions, the fuel / rack are infinite.
o No credit is taken for parasitic neutron absorption in the fuel assembly grid spacers or and fittings.
i===
O o : cd ruei e=$1r i ==ed to a t nrica-ment and no credit is taken for fuel depletion.
o No credit is taken for burnable poisons The CASM0/PDQ method applied in these calculations have u ,20,21 An explicit been described in detail previously CASMo-2E model of the fuel / rack geometry is used to calculate cross sections for the fuel, water gaps, Eircaloy j channel, stainless steel structure and Botaflex. The B-f actor option in CASMO-2E la used to provide transport !
i theory corrected absorption cross sections in the neutron !
I poison region. Using this option, the macroscopic l absorption cross sections in the poison are iteratively adjusted until the infinite. multiplication factor as O 9-3, O
l O
I}
calculated by PDQ matches the corresponding value rom the transport theory . . calculation. The cross sections in the region of the fuel / rack without gaps are developed asswning four panels of Boraflex in the cell. To obtain. cr'oss sections in the gap region, the Bo'raflex is replaced with water.
% The cross sections so developed are then used in an explicit three dimensional PDO-07 model of one half fuel assembly / rack geometry as show'n in Figure 9-1. The fuel region is conservatively assumed to contain 150' of 3.2 w/o 0-235 fuel, although the actual bundlen, at Quad Cities have 1
(
a 6' natural uranium reflector on each end. A 15 centimeter thick pure water reflector has been assumed at the top of the assembly and no credit is taken for the stainless steel end fitting in the actual fuel assembly. A sero flux 1 l
boundary condition is assumed at the top of this reflector.
Since the Boraflex is not symmetric on the side of the cell 1
vall, a rotational boundary condition is assumed about the center of the fuel assembly. At the cell boundary (x and y directions), reflected' boundary conditions are assumed. All gaps are conservatively assumed at the assembly / rack midplace and accordirigly a reflected boundary condition is applied at the center of the gap.
Figure 9-2 shows the mesh distribution in the x-y i directions of the PDQ aodel. Each fuel rod is represented O
. T, -
94
i
)
O o l i.
by one mesh interval as are t e water gaps, chanael, rack cell and Boraflex. In the axial direction, a f ne mesh I spacing (typically 0.500' for large gaps) is used in the in the gap and in the vicinity of the gap region. In the I central, portions of the fuel (away from the gap and reflector), a course spacing is applied. In the reflector i g and in the fuel / rack regions in the vicinity of the reflector region a finer axial mesh distribution is used.
l Using this model, a reference calculation (no gaps) was l
performed and resulted in an infinite multiplication factor (no axial leakage) of 0.91.3 and a k,gg of .9105. For the I infinite media case the 0.9129 can be compared with C.9155 as reported in Reference 2. The effect of gaps at the midplane was then computed as a function of gap size and number of gaps per cell. The differential reactivity (k,gg (w/ gap) .5105) is then computed and is shown in Table 9-2. The data points in Table 9-2 are based on calculations for a cell with pans 1s containing 4- 4' gaps, 2 - 4' l 1
gaps, 2 -2* gaps and 4 -2' gaps. The intermediated points have been developed using second order interpolation which is conservative. The differential reactivity for 2 - 3" !
gaps and 4 - , 3' gaps was also calculated and in less than the values shown in Table 2 as determined by. interpolation.
It should be noted that the differential reactivity I shown in Table 9-2 is a very conservative calculation and
., :.. *-5 I
O ,
?!
l~
can be viewed as an upper bound estimate of the eactivity '
increase due to gaps. This is the case since the use of a reflected boundary condition in the x and y directi.ons assumes that every storage cell has', for example, 4 - 4' gaps at the midplane. The data presented in Section 4.0 shows that there is an axial distribution of gaps (not all
% gaps line up at tha same axial elevations) and that not all cells have panels with the same number of gaps. In addition, no credit is taken for increases in the B-10 loading in the region of the fuel / rack cell without gaps due to shrinkage of the Boraflex.
The conservative differential reactivity values in Table 9-2 are used subsequently to compute the net reactivity increase based on the actual distributions of gaps.
9.2 Probability d W Occurrence The probability that a given event occurs is ca1 91sted using the following equation: I P
ijk * #1(s) x F (n) x P kI*) II"1) where P
- probability that a fuel storage cell has n ijk Boraflex panels with a gap in axial interval k with a size in gap size interval s
- ': 9-6 g ..:-. ..
m .
O Pg(s) = fraction of gaps in cumulative gap size inter-val s f
F)(n) = fraction of celle with n panels containing one or more gaps Fk (z) = fraction of gaps occurring in axial interval K The quantities Fg (s) and F(z) have been developed from the data obtained from the NNC special measure'ents m as described in Section 4.0. has been computed based on the F)(n) results of the 118 cells subjected to the standard test )
method. The quantities so calculated are shown in Table O 9-3.
9.3 L2 sal A:11 ciebal Reactivity Effects In considering the effect of gap formation on the reactivity state of the Quad cities fuel storage racks, both local and global effects have been addressed. The former relates to the probability of a cluster of adjacent storage cells having large gaps all at the same axial elevation.
The latter is the not increase in fuel / rack reactivity based on the actual distribution of gaps.
To address local effects, Equation 9-1 can be used to compute the probability of gap occurrence versus gap size interval, number of panels / cell with gaps and axial 97 Ti. . < '-
O -
i;
- i Es au elevation of the gaps. Tables 9-da, b, c, and d Bontain the 1 probability of occurrence of gaps in size interval 1, 2, 3 and 4 , respectively, as a function of the number of panels per cell containing gaps and axial elevation.
Of the refueling racks in the Quad cities Unit 1 pool, 446 cells have received freshly discharged fuel assemblies during both refueling outages and therefore have received j 22 For this population of 446 the greatest gamma exposure cells Equation 9-1 can be used to detumine a distribution of the number of cells containing 4' gaps (worst case from a reactivity standpoint) can he calculated. Table 9-5 shows that at any axial elevation less than 2 cells would be i expected to have 2 - 4' gaps. In the unlikely even that these cells occurred adjacent to each other, an upper bound estimate of the reactivity effect can be made. Table 9-2 shows that for 2 - 4' gaps in every cell all aligned at the fuel / rack midplane, the reactivity effect is +0.010. This may be compared with the infinite multiplication factor of (954 probability /954 confidence level)
O.931 reported in Reference 2.
The global effect of gaps on the fuel / rack reactivity has been assessed using a computer program which uses random numbers to calculate which events have occurred and the a
reactivity effect of each event. Since the reactivity O effects (Table 9-2) are based bn the assump?. ion :bst all 4
4 g
- - 9-8 -
l
(:)
l I}
gaps occur at the same axial elevation, Fk(x) Equation 9-1 was assumed to be 1.0 at axial interval 8 and zero elsewhere. The program is used to compute the net reactivity effect 10,000 times to obtain a good statistical sample. The mean increase in reactivity so calculated is
+0.0023, +/ .0037 (1 sigma level) . At a 954 probability 4'
and 954 confidence level, the total reactivity effect is
+.0097. Adding .0097 to the design value of 0.931 (Reference 2) provides a value of .941 which is less than the design limit of 0.95.
() 9.4 Medei/ Method conservatinum Throughout the previous discussion and description of the models and methods applied to assess the reactivity effects of gaps, several very conservative assumptions have j been used. These assumptions are summarized below to illustrate the overall conservative nature of the approach:
o The water in the fuel / rack reference calculation (Ref. 2) is assumed to be at 58'F whereas the pool wa-ter at Quad cities is approximately 100*F which corresponds to a reactivity effect of +0.004.
o The fuel / rack geometry is assumed to be infinite in the x--y direction and all gaps occur at the same axial elevation (midplane).
Tpisimpliesthateverycellin O
90 . .-
the rack has, for exa=p1'e, 4--4 gaps at the midplane I which maximizes the reactivity effect.
h o In the present calculations as well as in the reference 1
calculations (Reference 2), no credit is taken for l parasitic neutron absorption in the fuel assembly grid. I 1
spacers or end fittings. j o Every fuel assembly is assumed to be unirradiated and at a maximum enrichment of 3.2 w/o U-235 and no credit is taken for fuel depletion. In addition, no credit is taken for burnable poisons. The unirradiated reload fuel at Quad cities typically has enrichments less than l l
3.2 w/o U-235 and contains Gadolina burnable poisons.
O In addition, dischar ed asse hites have accumulated at least one cycle of exposure prior to being.placed in the racks.
o No credit is taken for increases in Boraflex B-10 load-ing in the region of the f.,el rack where gaps do not exist. For a 4' gap, the corresponding increase in B-10 density due to shrinkage is approximate 1y 84.
o If the NNC Special Measurements showed a panel with, for example, 2 - l' gaps, the cumulative gap size (2') was used in deve1oping Fg (s). The reactivity effect of 2 - l' gaps in a panel would be expected to be less than 1 - 2' gap, o For the purpose of calculating reactivity effects, a 0 -
9-10 t
e "
m
() .
l gap falling within a gap' size interval is assumed to be b
at the extreme of that interval. Forexampishagap detected at'2.5' falls in gap size interval 3 and for the purpose of reactivity calculations is assumed to be I a 3' gap.
o Fk (x) is assumed to be 1.0 at axial interval 8 and zero elsewhwere. This is equivalent to assuming all gaps
+ '
occur at the fuel assembly midplane.
)
l O
l l
O < g.11
. . e O
k TABLE 9-1 FUEL ASSEMBLY DESIGN SPECIFICATIONS 8r RM (Refereace}
Fuel Rod Data outside diameter, in. 0.483 Cladding thickness, in. 0.032
+
Cladding material Er-2 Pellet density, gra 00 /cc 2 10.41 l
Pellet diameter, in. 0.410 Mar. nodal enrich., wtt U-235 3.2*
l Water Rod Data outside diameter, in. 0.591 i
Wall thickness 0.030 I Material tr-2 i i
Number per assembly 2 Fuel Assembly Data Number of fuel rods 62 '
Fuel rod pitch, in. 0.640 j Fuel channel outside dia., in. 5.438 Fuel channel wall thick., in. 0.080 Fuel channel material tr-4
)
O
- Actual at bossfuel assemblies ende or rue 1 rod. have 6 inches of natural uranium 1
- - - - _ - - - - - - - - - -- d
~
o O 1i i
b he TABLE 9-2 f 4' DELTA K VS. NUMBER OF PANELS / CELL WITS GAPS AND GAP SIZE Err NUMBER OF PANELS / CELL WITB GAPS GAP SIZE
, INTERVAL 0 1 2 3 4 0 0.0000 0.0000 0.0000 0.0000 0.0000 1 0.0000 0.0006 0.0013 0.0021 0.0031 2 0.0000 0.0013 0.0034 0.0062 0.0097 -
3 0.0000 0.0022 0.0063 0.0122 0.0198 4 0.0000 0.0033 l 0.0100 0.0201 0.0335 4
e
- t i
O -
9-13
0 t,.
TABLE 9-3
, F (s), F (n) and F (z) i j k FRACTION OF GAPS FRACTION OF CELLS EAVING l VS l
n NUMBER OF PANELS / CELL GAP SIZE INTERVAL WHICE CONTAIN GAPS INTERVAL F (s) n F (n) 1 -
j
+ ....... ..
1 0.3214 0 0.3314 2 0.4643 1 0.2119 l 3 0.1429 2 0.2712 4 0.0714 3 Da1017
() 4 0.0339 FRACTION OF GAPS VS AXIAL INTERVAL
........................... =--
AXIAL F (z) ;
INTERVAL k
1 0.0000 2 0.0000 3 0.0000 4 0.0000 5 0.0323 6 0.1290 7 0.0323 8 0.1935 9 0.0645 10 0.0645 11 0.0645 12 0.0645 13 0.1290 14 0.1613 15 0.0645' O
I 9-14
~
l
. I
~
l l
TABLE 9-4a
)
P POR GAP SIZE INTERVAL 1 (0*-1" GAPS) ;
ijk NUMBER OF PANELS /CE",L WITH GAPS AXIAL F (z) --------------------- 9 1 2 4------------
INTERVAL k 1 4 .
I 1 0.0000 0.0000 0.0000 0.0000 0.0000 2 0.0000 0.0000 0.0000 0.0000 0.0000 3 0.0000 0.0000 0.0000. 0.0000 0.0000 1 4 0.0000 0.0000 0.0000 0.0000 0.0000 '
5 0.0323 0.0022 0.0028 0.0011 0.0004 6 0.1290 0.0088 0.0112 0.0042 0.0014 7 0.0323 0.0022 0.0028 0.0011 0.0004 8 0.1935 0.0132 0.0169 0.0063- 0.0021 WF 9 0.0645 0.0044 0.0056 0.0021 0.0007 10 0.0645 0.0044 0.0056 0.0021 0.0007 11 0.0645 0.0044 0.0056 0.0021 0.0007 12 0.0645 0.0044 0.0056 0.0021 0.0007 13 0.1290 0.0088 0.0112 0.0042 0.0014 i 14 0.1613 0.0110 0.0141 0.0053 0.0018 l 15 0.0645 0.0044 0.0056 0.0021 0.0007
() '1 '
TABLE 9-4b P FOR GAP SIZE INTERVAL 2 (l'-2' GAPS) ijk NUMBER OF PANELS / CELL WITH GAPS AXIAL P (z) ------- - - - - - - - - - - - - - - - - - - - - -
INTERVAL k 1 2 3 4 1 0.0000 0.0000 0.0000 0.0000 0.0000 2 0.0000 0.0000 0.0000 0.0000 0.0000 3 0.0000 'O.0000 0.0000 0.0000 0.0000 4 0.0000 0.0000 0.0000 0.0000 0.0000 5 0.0323 0.0032 0.0041 0.0015 0.0005 6 0.1290 0.0127 0.0162 0.0061 0.0020 7 0.0323 0.0032 0.0041 0.0015 0.0005 8 0.1935 0.0190 0.0244 0.0091 0.0030 9 0.0645 0.0063 0.0081 0.0030 0.0010 10 0.0645 0.0063 0.0081 0.0030 0.0010 11 0.0645 0.0063 0.0081 0.0030 0.0010 12 0.0645 0.0063 0.0081 0.0030 0.0010 13 0.1290 0.0127 0.0162 0.0061 0.0020 14 0.1613 0.0159 0.0203 0.0076 0.0025 15 0.0645 0.0063 0.0081 0.0030 0.0010 0 -
9-15 L - .____-___ _-___ _ _ -
i l
r, TABLE 9-4c P FOR GAP SIZE INTERVAL 3 (2"-3' GAPS) ijk i NUMBER OF PANILS/ CELL WITE GAPS AXIAL F (2) ------------------------------------
INTERVAL k 1 2 3 4 1 0.0000 0.0000 0.0000 0.0000 0.0000 2 0.0000 0.0000 0.0000 0.0000 0.0000 3 0.0000 0.0000 0.0000 0.0000 0.0000 4 0.0000 0.0000 0.0000 0.0000 0.0000 5 0.0323 0.0010 0.0013 0.0005 0.0002
, 6 0.1290 0.0039 0.0050 0.0019 0.0006 7 0.0323 0.0010- 0.0013 0.0005 0.0002 8 0.1935 0.0059 0.0075 0.0028 0.0009 9 0.0645 0.0020 0.0025 0.0009 0.0003 ,
10 0.0645 0.0020 0.0025 b.0009 0.0003 l 11 0.0645 0.0020 0.0025 0.0009 0.0003 '
12 0.0645 0.0020 0.0025 0.0009 0.0003 13 0.1290 0.0039 0.0050- 0.0019 0.0006 14 0.1613 0.0049 0.0063 0.0023 0.0008 15 0.0645 0.0020 0.0025 0.0009 0.0003 O
i TABLE 9-4d j P POR GAP SIZE INTERVAL 4 (3'-4' GAPS) ijk NUMBER OF PANELS / CELL WITH GAPS AXIAL F (s) - - - - - = = = - - - - - - - - - - - - - - - - -
INTERVAL k 1 2 3 4 1 0.0000 0.0000 0.0000 0.0000 0.0000 2 0.0000 0.0000 0.0000 0.0000 0.0000 3 0.0000 0.0000 0.0000 0.0000 0.0000 4 0.0000 0.0000 0.0000 0.0000 0.0000 5 0.0323 0.0005 0.0006 0.0002 0.0001 6 0.1290 0.0020 0.0025 0.0009 0.0003 7 0.0323 0.0005 0.0006 0.0002 0.0001 8 0.1935 0.0029 0.0037 0.0014 0.0005 9 0.0645 0.0010 0.0012 0.0005 0.0002 10 0.0645 0.0010 0.0012 0.0005 0.0002 11 0.0645 0.0010 0.0012 0.0005 0.0002 12 0.0645 0.0010 0.0012 0.0005 0.0002 13 0.1290 0.0020 0.0025 0.0009 0.0003 0.0024 0.0031 0.0012 0.0004 0
, 14 15 0.1613 0.0645 0 0010 0.0012 0.0005 0.0002 a.1 F
~
~
.O I
I l
l
- TABLE 9-5 NUMBER OF CELLS WITH 3'-4* GAPS IN TBE REFUELING RACES i
NUMBER OF PANELS / CELL WITB GAPS AXIAL - - - - - - - - - - - - - - - - - - - - ~ ~ -
INTERVAL 1 ,
2 3 4 1 0.0000 0.0000 0.0000 0.0000 O 2 3
0.0000 0.0000 0.0000 0.0000 0.0000.
0.0000 0.0000 0.0000 4 0.0000 0.0000- 0.0000 0.0000 5 0.2180 0.2789 0.1046 0.0349.
6 0.8705 1.1141 0.4178 0.1393 7 0.2180 0.2789 0.1046 0.0349 8 1.3057 1.6711 0.6267 0.2089 9 0.4352 0.5570 0.2089 0.0696 10 0.4352 0.5570 0.2089 0.0696 11 0.4352 0.5570 0.2089 0.0696 ;
12 0.4351 0.5570 0.2089 0.0696 13 0.5704 1.1141 0.4178 0.1393 14 1.0884 1.3930 0.5224 0.1741 15 0.4352 0.5570 0.2089 0.0696 O
9-17
~
Figure 9-1 Three Dimensional PDQ07 Model ,
e,l O ,.
l i
f I
. l WATER REFLECTOR
,l FDEL ASSEMBL7 N
\ \ EEIO FLUX IIRCALOT CEANNEL WATER CAPS
\ OTATIONAL Syy,yE;
\
STAINLESS STEEL CELL WALL
\
j l
t l
I l
l
. 1 O - p sonArLEx PANUS /
/
? /
l
%l s/
t Y
,- 4 f g EALF GAP f REFLECTED BODNDARY CONDITIONS
/
O 9-18
O r Figure 9-2 PDQ07 Mesh Description in the X-Y Plane l
l t
v.
,m ,
I d
BORAFLEX PANEL WATER ROD
- l WAT G W s l
WATER GAPS g ,
q( ROTATIONAL EIRCALOY CHANNEL STAINLESS STEEL -'
CELL WALL ;
REFLECTED BOUNDARY '
CONDITIONS --. . . . - -
i O
9-19
O .
l
- 10. coNetestens This repoht describes the results of a preliminary ;
assessment of Boraflez performance in the Quad cities spent fuel storage racks. The results are considered preliminary since there are areas where data are not available. This is I particularly true with ;espect to Soraflex shrinkage over '
the intermediate range or gamma exposures to which the Quad .I cities racks have been exposed as well as the long ter:n stability in the spent fuel pool environment. Accordingly, l as additional data becomes available, the conclusions j developed as a result of the preliminary assessment could l change. Nevertheless, the study conducted herein points to the following conclusions with respect to observed behavior of Boraflex, irradiation damage mechanisms in Boraflex, the l effect of manuf acturing process, and the effect of gaps on the reactivity state of the Quad Cities spent fuel storage racks. These conclusions include:
Nat f enal Nuclear 2131 Results:
The special tests conducted by National Nuclear Corporation provide a means to develop gap size distribution and axial distribution of gaps for the Quad Cities racks af ter 2 refueling outages. The mean gap size detected in '
. t these measurements is approximately 1.5', the maximum gap O
10-1
O i
size detected is 4.0".
Review of the axial distribution of gaps sho several characteristics (see Figure 4-2):
o There are no gaps in the lower section of the cells, o There is a distinct peak near the midplane of the fuel' storage cell.
o There appears to be a second peak near the top of the
- cell.
Radiation damage acchanisms which offer a potential explanation of these characteristics have been discussed in section 8.0.
The standard NNC measurements have been evaluated and provide a means to develop probability functions for the number of Boraflex panels / cell which contain gaps. These functions have been subsequently used to assess reactivity effects.
l M g Manufacturing Process The fuel rack manufacturing process was audited and the maje- conclusions of that audit includes l o An adhesive (Dow Silicone 999) was used to afix the Borafier sheet to the stainless steel sub elements whereas BISCO had tested Dow Silicone 732 (Ref. 3).
The manufacturer, (J. Cat) has stated that the two i adhesives are similar.
e O 10-2
1 O h o Little control was exercised during the applic'ation of-the adhesive since its only intended pur se was to ,
hold the Soraflex in place during assembly and welding of the various sub elements. The adhesive was applied in a discontinuous bead along the entire length of the cell. The bead was then spread out to a width of 2 1/2' to 3' with a stainless steel scraper and the
^]
- Boraflex rolled into place. It is therefore not known whether the adhesive is continuous or of uniform width along the length of each cell.
. o While the use of discontinuous strips of Boraflex cannot be ruled out, it appears unlikely since the material is received from the supplier in full lengths l for the various matching stainless steel components.
Evaluation g Radiation Testine 21 Borafier l
Available data from the test programs sponsored by i
BISCO were reviewed and evaluated. In this study changes in 4
the sample specific gravity (pre and post irradiation) were used to compute changes in sample volume. Based on this evaluation, some samples showed a maximum volumetric 10 shrinkage approaching 20% at an exposure of 1 - 2 X 10 rads. Beyond this exposure, the samples appear to undergo swelling. Measurements of pre and post irradiation dimensions may not be a reliable indicator of the extent of O 10-3
~
O .
L
? !,
~
volume changes although ther's is some indicatio that the shrinkage may not be isotropic. Some samples shoWd greater 'l shrinkage in. thickness than in width.
j Gas evolution measurements during irradiation (in air)
]
show that initially the samples produce off gasses consisting of primarily 5 2
with N, 2
0, 2
CO, CO 2
and p hydrocarbons. Af ter an exposure of 1 x 10 10 rads, off gas production ceased.
Data in the range of esposures to which the Quad Cities racks have been exposed is not available. The test i 1
irradiations were conducted in a test reactor environment. l The effect of neutron damage and differences in the gamma-photon spectrum (test versus pool environment) have not been determined.
I Radiation Da mee Mechanisms h Berafier The main radiation damage mechanisms in Boraflex are believed to be crosslinking and scissioning. Crosslinking causes the material to shrink and when there are many available sites for crosslinking (i.e., at low doses), it is the predominant mechanism. As crosslinking saturates (at a '
dose estimated be 10 to approximately 1 1 10 rads),
scissioning becomes the predominant mechanism. Scissioning is characterized by the severing of atomic bonds in the spine of the polymer and may result in eventual degradation O
10-4
l 1,
)
of the material. Scission'ing may explain the observed swelling of the Boraflex samples in the BISCO t t program
'and the tendency of the material to absorb water (i.e.,
Point Beach examinations).
Q For-atien, g Crowth M h Iggg Etability g Berafier b & M Zggi M Environment Mechanisms have been postulated which offer a potential explanation for inception of tears in the panels. of restrained Boraflex.
The effect of radiation damage on the !
adhesive is unclear and therefore three different bounding scenarios have been examined. Radiation damage inBoraflex may cause the elastic modulus to increase substantially.
Accordingly, as a restrained Boraflez panel tries to shrink, relatively large local stresses (estimated in excess of the tensile strength) develop, potentially causing a tear.
Af ter a tear occurs, the gap formed continues to grow as the Boraflex is subject to additional shrinkage until crosslinking saturates (estimated at 1X 10 10 rads gamma based on test reactor irradiation). The magnitude of the >
maximum gap aire is difficult to project owing to primarily two factors. First, there is a lack of volume change data in the exposure range of 10 8 to 10 10 rads. Second, it is i 1
not known whether the Botaflex manufacturing process causes ;
shrinkage .o be anisotropic. There is some indication that '
O 10-5 i
O i
Boraflex shrinks less in lengt.h and width than in thickness.
Aworstcaseestimateofmaximumgapsizeof10'ipobtained assuming shrinkage is isotropic. The degree of anisotropy could alter this substantially. A preliminary model to predict gap growth developed as part of this study will be completed under an ongoing EPRI program when low dose data is available.
A The rate of Borafier shrinkage is likely to be greatest at low doses when there are many sites available for crosslinking. As crosslinking saturates, the rate is likely to diminish. Accordingly, the Boraflex in the Quad Cities racks may have experienced the greatest rate of gap growth g during the first two refueling outages and rate of growth with increasing dose may diminish during subsequent outages.
In order to prove this hypothesis, data in the exposure 8 10 range of 10 to 10 rads is needed.
The long term stability of the dimethyl polysiloxane matrix which contains the 3 C 4 powder in Boraflex cannot be projected at this time. The qualification program conducted by BISCO examined radiation effects and long term exposare to an aqueous environment separately. The combined effects after crosslinking saturates and scissioning predominates may likely depend on such f actors as pool water chemistry, water temperature, and local flow conditions around the Botaflex panels. The coupons inspected at Point Reach show O 10-6 9
v 1
O h
that the material tends to soften and rub off after exposures of 1 to 1.6 x 10 10 rads. While the coupols showed accelerated sof tening and erosion of the Boraflex, similar local effects were noted on a full panel which was inspected.
Reactivity Effects
- Conservative methods and models have been developed to determine the effect of gap formation on the reactivity
- state .of the Quad Cities spent fuel storage racks. The method uses CASM02E and PDQ07.,to determine the upper bound ,
reactivity effect as a functio'n of gap size and number of panels / cell containing gaps. A computer program which utilizes a random number generator to determine which
' events' have occurred, based on probability functions developed from the NNC measurements, is then applied. A f
total of 10,000 occurrences are sampled and the mean .
increase in reactivity as well as the 954 probability - 95%
l confid2nce level is determined. The analysis demonstrates that at a 954 probability /954 confidence level, the design limit of k,gg < .95 is still met. !
In considering the effect of Boraflex gaps on the reactivity state of the fuel storage racks, it is recognized that the reference fuel assembly design used for these analyses is considerably note reactive than the most !
O 10-7 I
___-________-_____ _ _ _ _ _ ____ ___ - - _ a
m l
I i .
i Bi D i reactive assembly used in practice at Qaad Cities. This is the case since the effect of gadolina burnable po on rods ,
in the reference fuel has not been considered. To demonstrate that future gap growth can be accommodated within appropriate design limits, the following is.
r ecorraanded o Itentify the most reactive fuel acaembly, when t consideration is given for fuel burnup and gadolina burnable poison.
o Determine the burnup corresponding to peak reactivity (estimated to be about 8000. MWD /MTU) for this limiting fuel assembly. l o Assess the effect of gaps on the reactivity state of O the racks using this fuel assembly at the burnup of ;
peak reactivity which is in practice the limiting case.
It is estimated that this approach can provide another 0.06 1 to 0.08 margin in k,gg to the design limit of .95.
In addition, fuel rack surveillance is recommended after the next refueling outage. such surveillance would include an inspection program utilising a neutron radioassay system providing resolution equivalent to that employed for the special tests. The same cells inspected using the special test method after the last outage should again be assayed to provide additional data with respect to:
o Rate of gap growth e O 10-8 4
e o Formation of new gaps
o Axial distribution of gaps h
o Gap size distribution It is anticipated that this two prong approach will provide the additional data and analysis necessary to assure that future gap growth can be accommodated within the design i
g limit of .95.
,Se O
O 10-9
}
l 1
REPERENCES
- 1. Special Neutron Attenuation Test for High Density Spent Fuel Storage Racks (Wet), National Nuclear Corporation for CECO Quad cities, CECO F.C. No. 309763, December, ;
I 1986.
- 2. Letter T.J. Rausch to B.R. Denton, " Quad Cities Station Units 1 and 2 Transmittal of supplement 8 to Revision 1
% of the Licensing Report on Bigh Density Fuel Racks NRC Docket Nos. 50-254 and 50-265', March 12, 1982.
- 3. "The Effects of Irradiation on Adhesively Fastened Botaflex", BISCO Test Report No. N-39, August, 1981.
- 4. ' Irradiation Study of Boraflex Neutron Shielding Materials", BISCO Report 74 8-10-1, Rev.1, August 12,
. ,1981.
l
- 5. Burn, R.B., Blessing. G., ' Radiation Effects of Neutron Shielding Materials", Trans. Am. Nuc. Soc., 11, Suppl.
1, 48 (1979).
- 6. Burn, R.B., Blessing, G., ' Radiation Effects on Spent Fuel Storage Rack Neutron Shielding Materials," Trans.
Am. Nuc. Soc., 21, 429 (1981).
l
- 7. Letter C.W. Fay to G. Lear, Docket Nos. 50 -206 and 50 - 301, 'Results of Examination of Poison Insert Assemblies Removed from the Spent Fuel Storage Racks Point Beach Nuclear Plant, Units 1 and 2", February 11, 1987
- 8. "Boraflex Neutron Shielding Material, Product Perform-ance Data", BISCO Report No. N-38, 748-30-2, August 25, 1981.
- 9. Billmeyer, F.W., Textbook of Polyme'r Science, Inter-science, New York (1982).
- 10. Warwick, E.L., Industrial Eng. Chem., 11, 2388, 1955.
- 11. Warwick, E.L., Piccoli, W.A. and Start, F.C., J. Amer.
Chemical Soc., 22, 5017 (1955).
- 12. Bueche, F., J. Polymer Sci., 11, 297 (1956).
e G e
\
i
. i
).-
(
E,
- 13. Bopp, C.D. , 'and Sisman, O., Nucleonics 11 (7), 8 (1955). -
Bopp, C.D., an'd Sisman, O., ORNL 1373 (1953).
- 14. Charlesby, A., ' Atomic Radiation in Polymers", Pergamon Press, (1960).
- 15. Strasser, A., Yario,.W., Goldstein, L., Lindquist, E., ,
and Santucci, J., ' Control Rod Materials and Burnable J Poisons: An tvaluation of the State of the Art and Technology Development, July, 1980', Electric Power u' Research Institute Report NP-1974, 1981.
- 16. 'A Final Report on the Effects Of Righ Temperature Borated Water Exposure on BISCO Boraflex Neutron ,
Absorbing Material", BISCO Report No. N-2, Test 748-21, August 25, 1978.
- 17. Pf eif er, C.J., PDQ-7 Ref erence Manual II, WAPD-TM-947
- l. f (L), February, 1971'.
() 18. Edenius, M., Ahlin, A., Eaggblom, D., 'CASMD-2E, A Fuel Assembly Burnup Programs User's Manual', Studsvik/NR-8 1/3, November, 1981.
- 19. Letter, J.D. O'Toole to S.A. Varga, ' Application for Amendment to Operating License *, Docket No. 50-247, November, 19, 1985.
- 20. Letter, J.D. O'Toole to S.A. Varga, ' Application for Amendment to Operating License', Docket No. 50-247, March 26, 1986.
- 21. Letter, J.C. Brons to S.A. Varga, " Proposed Technical Specifications Regarding Fuel Enrichment *, Docket No. .
50-286, June 13, 1986.
- 22. Letter, J. Boeller to E. Lindquist, March 4, 1987.
O
- 9