Text

Intervenor Exhibit I-SC-2,consisting of Requesting Addl Info Re Plant Spent Fuel Pool Reracking & Use of Boraflex Neutron Absorbing Matl.Comm Ed & NET-042-01 Ltr Re Quad Cities Plant Anomalies Encl
	ML20237K236
Person / Time
Site:	Diablo Canyon
Issue date:	06/17/1987
From:	Trammell C; Office of Nuclear Reactor Regulation
To:	Shiffer J; PACIFIC GAS & ELECTRIC CO.
References
	OLA-I-SC-002, OLA-I-SC-2, TAC-64470, TAC-64471, NUDOCS 8709040264
	Download: ML20237K236 (101)
	v • d • e

~

50- Z 75,f3 2 3 -Cdf $U.

\\

'f 6fS/r;7 b 8C-7

' c..

2 gy 150 W,'

Docket Nos. 50-275 and 50-323

~ ~ ~

5

'87 gr 26 P4 :02 0:

Mr. J. D. Shiffer, Vice President D0D l !"- - -

Nuclear Power Generation D*

c/o Nuclear. Power Generation, Licensing Pacific Gas and Electric Company 77 Beale Street, Room 1451 San Francisco, California 94106 i

Dear Mr. Shiffer:

l

SUBJECT:

REQUEST FOR MDDITIONAL INFORMATION - DIABLO CANYON SPENT FUEL P0OL RERACKING (TAC NO. 64470 AND 64471)

We have recently learned that some anomalies have been found-in the Boraflex neutron absorbing material used in the spent fuel pools at the Point Beach Nuclear Plant and the Quad Cities Station. See the enclosed letter and report from Commonwealth Edison dated May 5,1987 (Enclosure (1)). Since Boraflex is planned to be used in the' proposed modified racks at Diablo Canyon, we request that you provide the additional information identified in Enclosure (2) as soon as possible.

The reporting and/or recordkeeping requirements of this letter affect fewer than ten respondents; therefore, OMB clearance is not required under PL 96-511.

Sincerely, original signed by Charles M. Tramell, Project Manager Project Directorate V Division of Reactor Projects III, IV, V

& Special Projects

Enclosures:

1.

Commonwealth Edison Letter dated May 5, 1987 2.

Request for Additional Information cc: See next page Office:

P PDi V

Surnana:

ell /tg G 4 hton

/ (/87 06/2/87 Date:

fl& F SC 2

j

{R09040264 870616 b^C l

0 ADocg 05000g75 cly; (O

{

g y

PDR

NUCLEAR REGULATORY COMutSS104 3C E1 50-?_16-0 4 otnci,i g,3, a hrbet Co.

in the matter of 2 $' b O * (O A

/

5 tiff ICENfiflID

/

Applicant R[ttlVED Intervenor _ _

- __ RIJECTED Cont's Off'r

~]

c Centractor DATE O'

Other Yatnen bI N

Reporter

\\

8 l

l 9

Mr. J. D. Shiffer t

Pacific Gas and Electric Company Diablo Canyon cc:

Richard F. Locke, Esq.

NRC Re'sident Inspector Pacific Gas & Electric Coipany Diablo Canyon Nuclear Power Plant Post Office Box 7442 c/o U.S. Nuclear Regulatory Commission San Francisco, California 94120 P. O. Box 369 Avila Beach, California 93424 Janice E. Kerr, Esq.

i California Public Utili' ties Commission Mr. Dick Blakenburg 350 McAllister Street Editor & Co-Publisher San Francisco, California 94102 South County Publishing Company P. O. Box 460 Arroyo Grande, California 93420 I

Ms. Sandra A. Silver 660 Granite Creek Road Bruce Norton, Esq.

Santa Cruz, California 95065 c/o Richard F. Locke, Esq.

Pacific Gas 'and Electric Company Post Office Box 7442 Mr. W. C. Gangloff San Francisco, California 94120 Westinghouse Electric Corporation P. O. Box 355 Pittsburgh, Pennsylvania 15220 Dr. R. B. Ferguson Sierra Club - Santa Lucia Chapter Rocky Canyon Star Route Managing Editor Creston, California 93432 San Luis Obispo County Telegram Tribune 1321 Jchnson Avenue Chairman P. O. Box 112 San Luispo Obispo County Board of San Luis Obispo California 93406 Supervisors Room 220 County Courthouse Annex San Lois Obispo, California 93401 Mr. Leland M. Gustafson, Manager Federal Relations Pacific Gas and Electric Company Director 3

1726 M Street, N. W.

Energy Facilities Siting Division Washington, DC 20036-4502 Energy Resources Conservation and Development Comission 1516 9th Street f

Dian M. Grueneich, Esq.

Edwin F. Lowry, Esq.

Ms. Jacquelyn Wheeler Grueneich & Lowry 2455 Leona Street 345 Franklin Street San Luis Obispo, California 93400 San Francisco, California 94102 s

q l

)

_-____-_a

Pacific. Gas & Electric Company Diablo Canyon

'r-cc:

Ms. Laurie McDermott, Coordinator Ms. Nancy Culver Consumers Organized for Defense 192 Luneta Street of Environmental Safety San Luis Obispo, California 93401 731 Pacific Street, Suite 42 San Luis Obispo, California 93401 President California Public Utilities Mr. Joseph 0. Ward, Chi'ef Commission Radiological Health Branch California State Building State Department of Health 350 McAllister Street Services San Francisco, California 94102 714 P Street, Office Building #E Sacramento, California 95814 Michael M. Strumwasser, Esq.

Regional Administrator, Region V Special Assistant Attorney General U.S. Nuclear Regulatory Commission State of California 1450 Maria Lane Department of Justice Suite 210 5580 Wilshire Boulevard, Room 800 Walnut Creek, California 94596 Los Angeles, California 90010 1

i Pacific Gas and Electric Company Diablo Canyon

-~2r-I cc:

Glenn 0. Bright Administrative Judge Atomic Safety and Licensing Board U.S. Nuclear Regulatory Commission Washington, DC 20555 Dr. Jerry Harbour Administrative Judge Atomic Safety and Licensing Board U.S, Nuclear Regulatory Commission Washington, DC 20555 B. Paul Cotter, Jr., Chairman Administrative Judge Atomic Safety and Licensing Board U.S. Nuclear Reculatory Commission Washington, DC 20555 4

1

s commona. ann u.ca hduaR 1

@. One First Naborist Ptaza. ChW Address Repty to. Post,0ffc? Som 757 Cheago, lancis 60690 0767

_w May 5,.1987 Mr. A. Bert Davis Regional Administrator U.S. Nuclear Regulatory Cormaission Region III 799 Roosevelt Road Glen Ellyn, IL 60137

Subject:

Quad Cities Station Units 1 and 2 Spent Puel Storage Racks NRC Docket Nos. 50-254 and 50-265

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 appar6nt anomaly in the neutron poison material, Botaflex, performance in the fuel storage racks.

Enclosed is a report frcan our contractor, Northeast Technology Corp. (NETCo), that describes the preliminary assessment of the anomaly.

The report is entitled " preliminary Assessment of Boraflex performance in the Quad Cities Spent Fuel 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 Quad 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 no criticality concern with the spent fuel at this time.

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 is hypothesized that Boraflex shrinkage caused by irradiation results in sufficient tensile stress to cause breakage when it is restrained as in the Quad cities rack design.

l l

A.B. Davis May 5, 1987 "n

l Additional testing and evaluations are being scheduled. We will keep you informed of future developments.

Please direct any questions regarding this issue to this office.

o Very truly yours, t.S.0 W '

M. S. Turbak Operating Plant Licensing Director im

Enclosure:

As Stated cc: Resident Inspector - 7Jad Cities T. Ross - NRR 3032X 4

6

1 y

ETCO NORTHEA$!TECHNOLOGYCORP

~

\\

Report No. NET-042-01 Revision No. A o

PRELIMINARY ASSESSMENT OF BORAFLEX PERFORMANCE IN THE QUAD CITIES SPENT FCEL STORAGE RACKS Prepared for Commonwealth Edison Company by Northeast Technology Corp 4/10/87

/

e

s

c TABLE DE CONTENTS 219A 1.0 Introduction...................................... 1-1 2.0 Spent Fuel Rack Description....................... 2-1 3.0 Description of Quad Cities Unit 1 Spen?. Fuel Pool and Refueling Practices........... 3-1 4.0 Evaluation of National Nuclear Corporation Measurements......................................'.4-1 4.1 Desc31ption 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 Radiation Testing of Boraflex.'...... 6-1 6.1 Summary of Irradiation Testing of Boraflex...... 6-1 6.2 Ev al u a tion o f Data.............................. 6-3 7.0 Postulated Radiation Damage Mechanisms in Boraf1ex.......................................... 7-1 7.1 Physical Properties of Filled and Unfilled Methlyla ted Polysilox ane........................ 7-1 7.2 Radiation Damage Mechanisms in Polymers......... 7-3 7.3 Effsets of Irradiation on the Physical and 9

Mechanical Properties of Methylated Po1ysiloxane.................................... 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..................................... 8-8 I

(CONTINUED) 1 1

1 J

l l

l I.

1 i

I TABLE OF CONTENTS (continued) 1 i

i 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 Ef f ects............. 9-7 9.4 Model/ Method' Conservatism......................

9-9 10.0 Conclusions....................................... 10-1 References

-l S

g O

q

4 s

2

=-.

A PRELIMINARY ASSESSMENT OF BORAFLEX PERFORMANCE IN THE QUAD CITIES-SPENT NUCLEAR FUEL STORAGE RACKS

1.0 INTRODUCTION

The spent nuclear fuel. storage racks at Commonwealth Edison's Quad Cities Nuclear Power Station consist-of a stainless steel structure and utilized Boraflex*

as a

neutron poison for criticality control.

Recent inspections of the Boraflex absorber by National Nuclear Corporation'has revealed that

' gaps

in the Borafles absorber have 1

252 developed.

The measurement technique, utilises a Cf neutron source and either BF-3 or Ee-3 proportional detectors.

The detectors do not record fast neutrons 252 (source) produced by the Cf source but' rather measure thermal 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' 0; :.e measurement in which the presence or absence of a gap in the Boraflex was determined.

The second tests, designated as 1

Trademark of Brand IndustMal Services Corporation-1-1

m

'special tests", provided a measure of the gap size and the axial elevation of the gap.

This study focused on an 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 Arbor and other data.

The objectives of this evaluation are as follows:

o To understand the mechanism of radiation damage in Boraflex o To estimate future performance of Boraflex in the environment of the spent fuel pool at Quad Cities o To quantify the implications of Boraflex 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.

e 1-2

4

~

2.0 REh"I ZEL EJLCE DESCRIPTION The spent fuel sto, rage racks for the Quad cities Station were ' fabricated by tha 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 subsequently welded together to form an

' egg crate" structure which ultimately provides storage locations or cells for spent fuel assemblies.

The basic Tee, E11 and cruciform elements are manufactured by starting with E11 shaped subelements of stainlass 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 Boreflex is created by using and strips of stainless steel to form a ' picture frame" between adjacent Ell's as shown in Figure 2-2 Yn the process used for manufacture, the strips forming I

the " picture frame" are tack welded to the stainless steel Ell's.

To retain the Boroflex in place during manufacture, 1

a Dow Silicon Adhesive (Dow No.999) is used.

A bead of a o

e

' k' t !

- 1 2

' t, l

j :..

v a

s

~

adhesive was applied to roughly the center of the. cavity and subsequently was distributed into - a 2-1/2*-3' wide strip.

1 along the entire axial len,gth of the cavity with a stainless l

. sy e

(

steel scraper.

The Boraflex was then rolled into,'the cavity

[

and pressed in place.

The' nominal dimensions of the

.I Botaflex are 070" thick, 5.9" wide and 152" in" length with I

a Boron-10 loading of.01728 gm/cm2 (areal density)..

1 With the basic E11 sections fabricated, the Tee, E11 and cruciform elements were assembled as shown in Figure 2-1.

During the welding process, copper chills were used to 1

minimize the area of the heat affected zone and to preclude I

overheating the Boraflex.

1 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 e

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) s to form an individual rack module.

I en e

2-2 l

5 7.i f

,5 06 7

6' (

T E

0 E

6 l

q l

i S

S

)

i

5

/

S 7

60 K

6 t

~(

i T

0 E

6 2

O, L

7 5

jt

~

1 E

8 %.

L

/

m h<

0 T

A

)

E

~

J c

E

)

(

)

,5 Tp T

b 7

E

(

E 0

1 l

H T

l 6

i S

S N

(

S

E M

/

S

". E L

N S

KE H

T T l

i T

E N T R E

~5 H E 0 E T

S M 2

7 l

1 0

x E

.l uXL OO S

}

T EE P

e.

+

o L I

R ).

v F L TP T

T AL SY "0

'6 p

o RA D (T 2

B

.O N

O O

E E

g y,

M R

f.'

I

.I 5L O

F I

C U

g R

"6 a

C 5

o 7

g*

0

- a

)

a o

6 a

e

(

7 i

r u

(

- i l

-'- - l w-a l

l I

^

Ms,

b STAINLESS D

STEEL i

ANGULAR l

S U B E L E M E N T 'A TOP END e

l S TRIP

( FI G. 3.3 )

(STAINLESS STEEL) l

\\

,, r=~

/

BORAFLEX Q

NG EDGE

/

LO l

SHEET SIDE E N D STRIP l

( T Y P, )

(STAINLESS STEEL) l i

6

,/

FIGURE 2-2:

ISOMETRIC VIEW OF A CRUCIFORM ELEMENT (From Reference 2)

N X

S

,i E O E

SL LGI L

EA T

A AT F

LI E

R C RA A

N R E E

IP OC I

O H A T Y TO B S T A T SL SM A

y - '

4 e

i L

)

=

I I

f

~

w 50 2

w 20 k

1 0 /c

+C n

i 2P 4

n 4

Y J

2. T 1

6

=

l*

. l l,

~

f 1

3 :;

6 u_

f 2

=

o P

l.

1 Y

T

=

F'-

.p L

L Y.

0

~

l

/

TeeEnd j

5 pacer g,:::::,

8W I,

n

-~

[

a a

t y

y 6

I P'

v

'i' n

g twafie o

i I

.l.

.,e f

1

)

, f

,/

Fuellede j

f I

g 3

j

/>-.. /

t

/q/

h i

.E 2

j stem (see kimin I

i a

/.

, spee.

w sine i

i 4

4.

y (4) %@ Flew t

Pres o

L t

1 Ql>..

1, t

a ',

&o v

o t

x Base Ptses 45' o

.w ts t

i a-Dia.

(%"

~

0A Sleeve Dent

~

FIGURE 2-4:

CE8L ELEVATION WITH BASEPLATE (From Reference 2)

-e CELLS I

MODULE

,/ ~

~f.

~ ~ ~ n

\\

B ASE PLATE e

l.

1 I

i i

i g

u

_ h__ _.~ay__!_._ cog-

'n :

pm;V 7

n s

1,,

d i

'-d =

4

- ~ ~

~~~

5V i

v 2

' Ij' v 4

2 NNNNNNNNNNNNNNNNNN e

a I/ ei 15 4 S Q,

=

o 2xx xx xxi iss ssg

'/ r p/

f/

9

/

I d

8 I..

~-

13 /4 7

m fl 50,

/

] (T YP,)

/

Al

/L

~

_VNN N s\\ sJ tsNANsx %

h 1

1 FIGURE 2-5:

FUEL RACK MODULE SUPPORT FEET (FIL3 Reference 2) 1 j

___._._____._.____._______________j

---=

3.0 DESCRIPTION

DE Lil DHaQ CITIES REII & 1 SPJJ[I fjZEL 20QL AHD PEFUELING PRACTICES 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 completh complement of fuel assemblies from the ' reactor core is initially moved into modulec B2, 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 f9 ster than the discharge rack modules and may be subjected to additional thermal cycling.

3-1

I l

The procedure of designating refueling and discharge racks is described since it may be one factor in interpreting the behavior of Boraflex anomalies. described

)

subsequently.

e p

'M ee<

se

.u O

use 3-2

4 4

e anwe.

FIGURE 3-1 QUAD CITIES UNIT 1 SPENT FUEL STORAGE POOL "

O O

O O'O O

O C

8 8

. A 9.

. M m-O 9

O d e 9 es b,

M 2,

I '

f I

t l j

[ {

6 1

f l

I h

I 4

6 4 6 3

l 9

g,

,, g g

g p M

r i i e a

i e 6

+

t i t

t l I I 1 I h f

I I

i e

a i

I I '

I i t I t t

I i

3 1

MF9

- M 4

m at

' se e !

+

4e 5

'b J

6

%s n.

I i

6 6

4 t

e 5

8 4 3 5 i l

(

E 5

3

& I e

I g i i !

i i i

i i !

,I i

5 I

I I

l 3

1 1

l s

e a

i g

1 i

1 i,

,~ ~ ~

ll T

.Y lA Z

1 i I

L s

i 3 l a

~

T I

8 g i

i s i

i s s i s I,

I 1

6

, 1 i 6 l 6 8

f I i i I I

I 4

i, e i s i 4

e 4 a i

i 1

f g j

g I

\\

i

. 4 1

I l

5 I

I f

a i

i a

e i

' l i

i 1 4 i

I i t I

i i Ie I;

I f

P' I i t

yr I.

, i

! i e i i

e i

i i

a i

I i

J i

., i f

1

! t i

6 6 i i i

w Ne os IS M

a 26 ai 40 de a

u

4.0 EVALDATTON QE,THE M E NUCLEAR CORPORATION MEASUREMENTS 4.1 D e s c r i et i o n 21.th.t.T. tit Me t h o d s An initial testing campaign conducted by National Nuclear Corporation under contract to Commonwealth Edison took place using what has been designated the Standard Test 252 Method.

The Standard Test Method utilizes a Cf neutron source and four BF-3 proportional detectors which are sensitive so thermal neutrons.

The equipment is so designed so that each detectior is adjacent to one panel of' Boraflex when it is placed in a fuel storage cell.

The BF-3 detectors do not record fast acutrons from the 252 Cf scurce but do detect thermal neutrons which have been I

transmitted through the cell wall, thermalized and scattered back into the cell containing the detectors.

If the cell j

wall contains Boraflex, the back scattered neutrons are significantly attenuated; whereas where gaps exist, the backscattered neutrons undergo significantly less 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 i

l and a

peak in the count rate is indicative of a

A l

discontinuity in the Boraflex absorber.

The Standard Tests i

l 41

___- __ _ J

t are considered a "go-no-go' type of measurement but do not provide an accurate indication of the size or axial location of anomalies.

Af ter anomalies in the neutron absorber were indicatied in the initial testing campaign, a special test method was M2 developed.

This method. utilizes a Cf 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 storage cell.

The detectors are wrapped in lead (to reduce the potential of gamma interaction) and cadmium to form a 7 56h high window in the front of the detector one half 1

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 count I:te with peaks from measurements on a calibration standard containing gaps of known size, the approximate aize 3

and axial elevation'of the gaps can be determined.

4.2' Results g,{ 1h; 2g31 measurements In the initial testing campaign using the standac3 Test

Method, a total of 203 panels containing the. Boraflex fj absorbers in the-refueling rack. modules we e tested.

Of 4-2

1 j

j l

I l

these, 77. panels showed indication of anomalies.

In the discharge rack modules, a total of 103 panels were tested of which 18.showed indicati'on of. anomalies.

Although the dose to the rack modules has not been rigorously calculated, the 0

dose in refueling racks has been estimated to be 10 rads.

Review of the magnitude of count rate peaks recorded in these measurements conducted by CECO personnel in'dicate that 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

~

i panels were selected for testing,using the special test method.

All 28 panels were in the region of the pool containing racks designated as reload racks.

For the current evaluation of the NNC Special Test i

l l

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 1r anomalies in the range of 1.0" to 2.0" as Gap Size Interval 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 s,ummarizes the gap size, gap size interval, axial elevation, and axial interval for each of the 31 gaps J

A detected.

The average gap size is 1.35 * +/

.67 (1 sigma).

j i

It should be noted that the basic data provided a measure of i

l l

l

~

i v.----

1 l

i gap elevation as measured from the bottom of each cell. The I

Boraflex does not continue to thk bottom of each cell and appropriate corrections based on manufacturing drawings and detector design details yere made. Accordingly, gap axial distribution data' has been no.rmalized to the bottom of the Draflex sheet.

i A review of the data in Table 4-1 shows that of the 28 panels

tested, three (N07-W52, NO2-W54 and N01-W54) 1 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 average cumulative gap size on this basis is 1.5' +/

.85'.

The axial distribution of gaps is shown in Figure 4-2 l

in which the number of gaps versus axial interval is plotted.

This distribution exhibits several characteristics

{

which should be noted:

J 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.

I o There appears to be a second peak naar the top of the

cell, i

l Possibit mechanisms responsible for. these observations 1

will be described subsequently.

_q 1

4-4 i

i

4.3 Results g thg Standard M lleasurementa The standard test method was applied to 118 cells after the second refueling outage primarily in those modules 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 found to have at least one panel with one gap.

Table 4-3 contains a summary of the standard Test results and shows I

the cell location, total number of gaps detected, panel location and number of gaps in each panel as well as the number of panels in the cell containing gaps.

It should be noted that for some individual paaels indication of more than one gap per panel was observed.

For the purpose of obtainit; the number of panels in a cell which contained one or more gaps (last column in Table 4-3), those panels with more than one gap were counted once.

The data.in Table 4-3 is used subsequently to develop the fraction of pan 61s per cell having 0, 1, 2, 3 and 4 panels with one or more gaps.

l

)

4-5 O

la lv ar ie 7344346565926a30888486345918120 xt 11111 1

1 11 1

111 1

111 An I

l l

t 555555' 0055505505005005'050000050 e

h C) g 8351059680149012912110332683491 si 6234335455915830788486345808010 mfeR 11111 1

1 11 1

111 1

111 ooh

~

r c

fmnt 5500555555505550005505550050005 s

oif

~

p nt( e 8151058680149002912110232673490 a

ot L

6334335455915830788486345808010 Ga io 11111 1

1 11 1

111 1

111 t

tB ls a

ae v

uT e

d l

999999449994' 9949449449494444494 il Ea t

va r h 4917615246705678578776998249057 ii lo g 6133235454815729777375234807019 dc aB)i 11111 1

1 1

1 111 1

11 ne l

sR I p nfe S

Aoh 1

c 3:

mnt 9944999999949994449949994494449 n

oif ro t( e 4717614246705668578776898239056 1oi t

L 6233235454815729777375234807019 ft o

11111 1

1 1

1 111 1

11 4

a B

nr eoo lip el btr za auo iv TbC Sr i

e 122422111421111'

'.1232132233113 rr pt ta an se GI il Dcu dN n

e) s 0000000000000000550000000000500 al zsa a

ieN 1224221114211112001212132233013 en Sh zo c

ii pnn 5000005550055550005050500000050 St aii a

C(N 0113110003100001000101021122002 pN aG len WWNWNE8WWWWWNNNsNNNNENSNN8s8NEN a

P D -

TN 2788664222154444771214666655443 111444555 45555222222555555525 l

l 6223799776671225124565334443166 eN 1111 1

C'

~

r ae tb am 123456789012345678901234567890'1 Du 1111111111222222222233

' N

- E

n il -a rv-lav er i

r be3342 e

mt -

1 t

un-n MI -

hIcne iz l

i ar-1S v e 12 3 4 rb-p em tu-a G

nN-I s

lens at Ps e 8T el 2

za l

iv ra Sr oi e

1224221242122211121213223313 fc pt e

an np GI oS i

tu:

bn 2io

- ri 4ttsa eir 9000000000050055000000000000 lDc x

e b

p e) a 1224221242112200121213223313 adr zsM Tno i e-aC Shc er pn 50000050.4.0550000505050000 za ain ie G(i 0113110131001100010102112202 Sl N

c pu aN G

l ea vn io ti at la l

uN e

2 22 m

n WWNNhES/WWW//SNNNNENSNNESSEN u

a W

NN C

P 2' 78866422154447*i214666655 DN 11144455 455522222255555525 I

l l

6223799766712512456533444366 e

1111 CN 1

r ae tb am 123456789012345678901L 4345678 Du 1111111111222222222 N

7+

Table 4-3 Evaluation of NNC Standard Test Restits Cell Total Total Number Data Location Number of Panels Number N

W of Gaps N

S E

W with Gaps 1

1 26 0

0 0

0 2

1 27 2

1 0

2 3

1 28 0

0 0

0 4

2 26 0

0 0

0 t

5 2

27 2

1 1

0 0

2 6

2 28 1

0 0

0 1

1 7

3 26 0

0 0

0 8

3 27 2

1 1

0 0

2 9

3 28 0

0 0

0 10 4

20 2

0 1

2 11 4

21 2

1 0

2 12 4

22 2

0 1

2 13 5

20 2

1 0

2 14 5

21 2

0 1

2 15 5

22 3

1 1

0 1

3 16 5

23 2

1 0

2 17 5

24 2

1 0

0 1

2 18 6

20 3

1 1

0 1

3 19 6

21 3

1 1

1 0

3 20 f

22 1

0 1

0 0

1 21 6

23 3

0 1

0 2

2 22 6

24 4

1 1

4 23 7

22 2

1 0

2 24 7

23 2

1 0

0 1

2 25 2

24 4

1 1

4 26 1

17 0

0 0

0 27 2

17 0

0 0

0 28 3

17 0

0 0

0 29 4

17 1

0 0

0 1

1 30 5

17 1

1 0

0 0

1 31 6

17 0

0 0

0 32 7

17 0

0 0

0 33 8

17 0

0 0

0 34 1

18 0

0 0

0 35 2

18 1

0 0

0 1

1 36 3

18 0

0 0

0 37 4

18 0

0 0

0 38 5

18 0

0 0

0 39 6

18 0

0 0

0 40 7

18 0

0 0

0 41 8

18 1

0 0

0 1

1 42 9

1B 1

0 1

0 0

1 4-8

i Table 4-3 Evaluation of NNC Standard Test Results (continued)

Cell Total Total Number Data Location Number of Panels Number N

W of Gaps N

8 E

W with Gaps 43 10 18 0,

C 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 48 10 17 0

0 0-0 0

0 49 11 17 0

0 0

0 50 12 17 2

0 1

2 51 13 17 3

0 0

0 5'i 19 3

3 1

1 0

1 3

i 53 3

29 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

57 16 4

0 0

58 16 5

1 0

0 1

59 16 6

0 0

60 16 7

0 0

61 16 8

2 1

0 1

0 2

62 16 9

1 l'

O 0

0 1

63 16 10 1

1 6

0 0

1 64 16 11 0

0 0

0 65 16 12 0

0 0

0 66 16 13 0

0 0

0 67 16 14 0

0 0

0 68 16 15 0

0 0

0 69 16 16 0

0 0

0 70 17 1

0 0

71 17 2

0 0

72 17 3

0 0

73 17 4

0 0

74 17 5

1 0

0 0

1 1

75 17 6

3 2

0 1

0 2

76 17 7

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 81 17 12 0

0 0

0 82 17 13 0

0 0

0 83 17 14 0

0 0

0 84 17 15 0

0 0

0

?

85 17 16 -

0 0

0 0,

0 0

4-9

m Table 4-3 Evaluation of NNC Standard Test Results (continued)

Cell Total Total Number Data Location Number of Panels Number N

W of Gaps N

8 E

W with Gaps 86 6

51 1,

0 0

0 1

1 87 6

52,

2 0

0 1

1 2

88 6

53 2

1 0

2 89 7

51 1

0 0

1 0

1 90 7

52 2

0 0

0 2

1 91 7

53 4

1 1

2 0

3 92 8

51 1

0 0

0 1

2 93 8

52 0

0 0

0 94 8

53 2

0 0

0 2

1 95 7

44 3

0 0

2 1

2 96 7

45 3

1 0

1 1

3 97 7

46 5

1 2

1 1

4 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

1 105 1

55 0

0 0

0 106 1

56 2

1 1

0 0

2 107 2

54 6

3 2

1 0

3 108 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

3 115 4

56 3

1 1

1 0

3 116 5

54 2

1 1

0 0

2 117 5

55 2

1 1

0 0

2 118 5

56 1

0 1

0 0

1 Total Number of Gaps =

155 Number of Panels =

141 with Gaps Number of Panels / Cell with Gaps Number of Cells 0

45 1

25 2

32 3

12 4

4

D SN E

KG E

CL A

N R

O S I

TT E UNG-BEA

=

RM R I

TEO SR TS I

D U 1

SL p

+ EAE yr Z E U EI MF S

R L T

- UP A N-

- GAI FG C E I

E P X P S ES L

S FC E I

AN T R N C I

O B

D A-U0 8

6 4

2 0

8 6

4 2

1 1

1 Q

n%e u.

rh*EDZ 1

a e

S 4

a2 1

K D

CN 3

2 0

1 E

A 3 5 R E $ME3 G$

N L !

E S S P T E' NG-y: @ j + 5 A

g EA.

1

~

G

~

M R-4 FEO-

- 1 3

3j 1

4 T

!lu[l

- :OR S-1 2

_ 2 U

-5 1

NSL-

~

1 7

yI $@ 1 4 OAE-f l E

0L IT U

7 A

1 E U MF

~~

V 9 R

- R B E

- UI LT-T RAN-

e$~

I!

i l

8N 7//Ii:

I

_ G TIC E-y@fM 7 i

L I

~

A

. FS EP-I u*6 X

1: :j

1 I

l J;J, DPS-A I3j2 5

S 4

L S-

~

AC E-d 4 I

I XN T

' 3 i

I ANC-F 2 D-1 A-

~

U-

~

7 5

4 8

2 1

0 Q

$ E )

E;a u.

n'd I.2 l

l t:

9 5.0 AUDIT DE E. EEL M MANUFACTURING PROCESS An audit of the fuel rack manufacturing process uced for the Quad Cities racks was conducced at the Joseph oat 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 Borsflex to the various Tee, E11 and cruciform' elements Welding process and potential for overheating the o

Boraflex Any other factors such as the use of clamps, etc.

o which could contribute to the observed Boraflex beha-vier The major conclusions of that audit are summarized belows The adhesive used to hold the Boraflex in place during o

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 Joraph cat stated that they ate similar.

o The adhesive was applied to approximately the center of the cavity in a discontinuous head along the entire i

length.

The beed was spread out to a width of approx-l imately 2-1/2*--3" with 5 stainicss steel scraper.

The 5-1

- - - ~ -

Boraflex was then rolled into place and pressed against the stainless steel E11 sub-element and a.11 owed to cure.

There were no specific procedures for this process since the only intended function of te adhe-e sive was to fiold the Boraflex in place during assembly of the Tee, E11 and cruciform elements, o While the use of discontinuous strips of Bornflex cannot be ruled out, it in unlikely since the Boraflex is received in full lengths for the various matching stainless stt11 components.

i i

5-2 1

4 l

6.0 EVALUATTON E E TATTON TESTING 2 BORAFLEX 6.1 Summary sti.Tr.Ijolation M Preeram As part of a larger' program to qualify Boraflex for use in spent. fuel storage racks, BISCO sponsored a series of irradiation tests at the Ford Reacton at the University of Michigan (Ann Arbor)4 The purpose of these tests was to demonstrate the radiation stability of Boraflex.

The tests were conducted in a reactor facility to accelerate the accumulation of gamma exposures.

Accordingly, it ' mu.:t be noted that differences in irradiation environment exist between the test experiments and the Quad Cities spent fuel pool.

There are probably differences in the gamma spectrum in the test reactor and in the Quad Cities pool.

Additionally, the effect of neutron damage in the test irradiation is not known.

In these tests, small samples of Boraflex (of both 25 and 40 w/o B c) approximately 6" in len'gth,

.25" in width 4

and.100" in thickness were irradiated in air, distilled water and borated. vater (2000 ppm) to exposures in the range 10 11 of 1.6 X 10 to 1.03 X 10 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 irradiatics)..

During irradiation, Boraflex 6-1

_ _ _ = _ _ _ _ _ _ _. _ _

a samples in each of the three environments were monitored for gas evolution in terms of total

volume, rate and composition.

In addition, one sample was irradiated to a 8

low dose (2.81 X 10 rads in air) in the spent fuel storage area of the Fo r'd Nucidar Reactor.

Additional data on irradia. tion of Borafier have been reported in the literature,6 and have been included in this review.

5 The measurements of physical dimensions show in most cases a net shrinkage of the samples after irradiation.

The data is variable but the general trend is about 2-3%

shrinkage in width and up to 8% in thickness.

The accuracy 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 j

pr$duced when Beraflex is irradiated was collected and 6-2

e 4

with some analyzed.

The offgas consisted primarily of 52 and hydr carbons.

The and lesser amounts of CO, CO2 N,02 2

source of N is not clear,

however, potential sources 2

include air entrained in the samples during manuf acture or leakage in the sampling > 1ines.

The other off gas products are what would be expected based on the radiation environment and radiation damage mechanisms discussed 7

subsequently.

For the samples irradiated in air at 7 x 10 rads /bour, the gas evolution rate diminished to zero after 10 approximately 150 hours0.00174 days 0.0417 hours 2.480159e-4 weeks 5.7075e-5 months (accumulated dose of 1.05 x 10 rads)

The samples irradiated in distilled and borated water showed continued gas evolution which is believed to be due to radiolysis of the water in the reactor environment.

6.2 Evaluation g & n Since the physical dimension data may not provide a reliable indicator of the total extent of Boraflex shrinkage, the weight and specific gravity data (pre and post irradiation) from References 4,

5 and 6 have been evaluated.

Table 6-1 contains a summary of all the data related to weight and specific gravity changes.

The sample volume change based on both weight and specific gravity changes as well as changes in the specific gravity only have also been carguted.

The changes in volume have been 1

computed as:

J 63

~

AV/V (Vg - V )/Vg

=

g where:

V, = final volume, post irradiation y = initial volume, pre irradiation The data contained in Table 6-1 has been plotted in Figure 6-1.

Review of Figure 6-1 indicates the following:

8 10 1.

There are no data between 2.8 x 10 and 1.5 X 10 rada.

2.

It appears that initially all the samples underwent a reduction in volume.

From the data it would appear 10 that at an exposure in the range of 1 to 2 X 10 rads s

(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 1 10

rads, the 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 net been conducted at this point.

3 i

6-4

l 1

l-~

1 From the data it appears that in the range of 1 to 2 X l

10 10 rads, a maximum volume change occurs.

This might be l

consistent with the gas evolution measurements which j

10 indicate that gas evolution ceases at approximately 1 X 10 rads.

It is postulated that crosslinking of the polymer is complete at this point (all available sites for crosslinking i

expended) as discussed in section 7.0.

Crosslinking is reported to occur at sites where radiation has caused release of H+or CHj.

The samples ' irradiated in distilled water or borated 11 water show the greatest apparent swelling at 1.03 X 10 l

rads when volume changes calculated using the specific l

gravity and weight data are considered.

One possible 1

l explanation for this is that the samples are developing seme

\\

porosity and absorbing water.

A Boraflex panel at the Point I

l Beach Station has recently been removed from the spent fuel l

l storage rack and examined after a gamma exposure of 1.0 X 10 10 rads.7 In addition, surveillance coupons have also i

i been removed from the pool and inspected.

The ;;i?ons were j

10 removed from the pool after 1.6 X 10 rads exposure, dri=*

and radioassayed.

These measurements ir:dicate residual I

beta, alpha and gamma ac.drity, presumably arising from containments in the pool water.

Since the seule weight

_l measurements in Table 6-1 may not represent the true weight of polymer and filler, it is believed that the most reliable l

l O

l indicator of volume changes versus exposure can be derived.

l from the specific gravity measurements.

The results of, these data are used subsequently to estimate the ' maximum overall magnitude of volume reduction of the Boraflex in,the Qua'd Cities spent fuel racks.

i.

l i

i i

I i

l l

I I

I i

l u

_i l

l I

6-6 4

lI l

j

!); f j jj i

2

~

X e

9 7 59 501 9

9 55 694 5

97 763

/

m 0

6 40 387 5

0 03 871 9

11 781

)

u 0

0 01 110 01 110 0

1 1

11 111

?

I l

e X o

0 0 00 000 0

0 00 000 0

00 000 r

V

~. - -

u s

t o

f 1

p X-e 7

0 613 5

958 1

13 i

xc x

( -

m 0

3 568 1

163 3

72 E

u 0

1 110 0

110 0

10 e-l n

r-o 0

0 000 0 000 0

00 o

u-V i

s-t o-a p-d Eiy 9

2 82 597 9

2 57 984 6

52 598 a

ft 0

7 42 517 8

2 05 211 0

33 122 r

s ii 0

0

'01 120 I

s ea 0

0 00 000 0

0 00 000 0 00 000

~

1 1

01 220 11 221 1

g r

u cv r-pr s

e SG u

v-sr s-t 2

4 532 4

274 7

94 e

e h

0 2

221 7

812 6

15 v

g g

0 0

000 1

000 01 1

n i

x a

e 0

0 000 0

000 0

00 e

h-W l

C-faro B

t 8

0 36 011 2

0 48 667 0

07 484 c

s 8

3 79 416 3

2 85 514 9

21 335 1f iy o

6 9

78 909 o

ft)P 0

0 69 008 9

99 000 6

iic 1

1 11 121 2

2 11 221 1

11 222 s

cvc ee pr/

ea lg g

3 0

30 901 9

0 33 351 0

33 481 bn sG( e 7

0 99 752 0

99 752 aa r

6 8

66 668 7

8 66 668 8

66 668 Th P

C 1

1 11 11?

1 1

11 111 1 11 111 a

e t

m t

7 9

976 7

978 4

49 a

u s

0 8

725 7

181 1

68 d

l to 0

0 819 o

hP 1

006 7

13 y

v g

5 3

533 4

933 3

34 t

y i

i l

d e

v n

n W

8 6

075 7 358 2

44 a

o a

e 9

6 350 5

330 8

00 r

9 1

000 r

5 307 1

18 g

a P

c y

4 3

634 3 833 3

33 c

a t

i d

i f

v i

y m a

c t

c r

e 8

0 01 001 0

01 001 e

i /

G r) 0 1

11 111 1

1 11 111 1

11 111 p v g

us

+

++

+++

+

++

+++

+

++

+ ++

s a

c sd E

E EE EEE E

EE EEE r

2 i

oa 1

0 03 093 0

03 093 d

g 0

f pr 8

5 60 640 0

5 60 640 5

60 640 n

i x(

a c 0 c

E 2

1 11 121 5

1 31 121 1

11 121 i

e t

f f

p e

h i o

s ac g

c tr i

e g

au 4 6 44 444 5

6 44 444 6

44 444 e

p n t

Do w

n i h

S d

g n

n a

i tC 0

55 000

+

55 000 55 000 o

o o

e

/

4 4

22 444 22 444 22 444 l

W WB d

d d e

e e 0 a

r tr s

s 1

.i e

ae a

a rd r

t rt B

B +B re i

a oa I

2 IM A

W BW l

R

' h?hWe i f>

I.

c l'

; j'

} pl l

3 r

[

- ; j c i-pf *.,

$ }5$r

t

?

ji

,Ti

f j' lf

_ j j.

V

! j

1 m

I r

. D b i t

E I}

1 jIl 6 l5!!

U S

O P

X

/

I E

A l

h ;,i

, {

{

I 0 1

M M

i A

m G

l i lj Il l-,

J SU S

R ll I!

,i E

g l'

l!

,J l

i S

V D

E AR GN l

1I A

E H

0 R

C 1

U E

9 1

M O

P UL I

O R

V A

T M

E M

N A

G 1

i y

6 i

t E

v R

9 U

0 a

1 r

G

. m g

x I

- RE

!'jI i

F c

i ll f

f ce e

s i

l f

ma thg 8

i 0

1 e

0 0

w 0

1 3,

n o

de

,g!i s ens I

m._

7.0 POSTULATED RADTATTON DAMAGE MECHANTSMS H BORAFLEX M ETMTLAR ELASTOMERS 7.1 Phvnf emi Prenereien g riiled d anfliied Methviated pelvmilevano i

Boraflex is an elastomer (methylated polysiloxane) filled with finely divided boron carbide (B c) powder.8 4

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 C powder, which by virtue 4

of the large thermal neutron absorption cross section of B-10, provides reactivity control when this material is used in spent fuel storage racks.

Boraflex can be manufactured over a range of composition (B C content) and thickness.

4 8

Typical composition for spent fuel storage applications is,

Boron 31.5 w/o Carbon 19.0 w/o Silicon 24.5 w/o Oxygen 22.0 w/o Bydrogen 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.

7-1

l 3.

It must snap back (quickly) when the stress is released.

4.

It must retract completely (no permanent set).

When the above behavior is observed, certain molecular and environmental conditions usually existI 1.

The material is a polymer.

2.

The material is amorphous.

3.

The temperature is above T, the glass transition tem-9 perature.

4.

The material is lightly crosslinked.

l Fillers in methylated polysiloxane polymers such as boron carbide tend to improve tensile strength and increase the elastic modulus but would 'be expected to reduce

[

extension and elongation to break.

For a B-10 loading of l

l 0.020 gm B-10/cm at 2.5 mm thickness as-produced 2

l 8

Boraflex exhibits the following properties :

Modulus of elasticity 1000 psi Tensile strength 200 psi specific gravity 1.7 ga/cc Bardness 75 Shore A l

Temperature stability 200'C minimum without variable distortion The above properties are according to the manufacturer's specifications for typical as-produced Boraflex.8

=

7-2

)

I i

i

~T 7.2 Madiation Damace mechanisms h Polve*ra Prior to discussing postulated radiation damage I

mechanisms in polymers, it is useful to define units and j

terms used to describe the effects of radiation on changes in the physical properties.

Roentgen,is defined'in terms of the number of ion pairs l

produced in air by ionizing radiation.

As such it depends on the ionization energy, 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, tha energy value is 80-90 ergs /gm

! energy deposited).

Rad is a term that signifies 100 ergs of energy deposited in a

given material by a

given radiation I

environment.

It simplifies some of the technicalities of describing radiation effects and is rather widely used.

For I

irradiation of materials, many types of observable damage 6

tend to occur at the negarad (10 rads) region and beyond.

When a polymer such as Bortflex is subject to a

l radiation field, changes in the atomic / molecular structure occur.

Radiation results in the breaking of atomic bonda and subsequent crosslinking between atoms in adjacent polymeric chains can occur.

Figure 7-1 shows the molecular i

structure of as-produced methylate polysiloxane.

The spine

]

of the polymer is comprised of the repeating chain of Si-O j

J l

7-3 I

e

~

atoms.

Attached to. each si atom are two CH radicals.

3 Figure 7-2 illustrated possible crosslinking mechanisms in l

methylated polysiloxane.

In the first case (left hand 1

side), radiation results first in the release of a hydrogen atom from the CH radial. and subsequently the formation of 3

crosslink bonds between CE radicals in adjadent chains.

In 2

the second case, CH radicals are released from the main 3

chain and crosslinking between 81 atoms. in adjacent chains l

t i

occurs.

Crosslinking is characterized by the release of E '

2 CH and possibly C 5 '

Typically, as crosslinking. occurs, 4

26 f

the chains in the polymer are physically pulled closer together ar.d the material undergoes shrinkage and an increase in density.

i The other radiation damage mechanism in polymers is i

l described as chain scissioning in which bonds between atoms in the polymers are broken and crosslinking does not occur.,

An example of scissioning in methylated polysiloxane would be the breaking of the si-o bonds (main chain scissioning).

l In general, for crosslinking polymers' when there are many sites available for crosslinking (at low doses),

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 l

l 7-4 i

i S

i a given type that occur when 100 eV of energy is deposited.

For instance, G(scission) = 2 implies that 2 scission events of molecules occur for each 100 eV absorbed.

For crosslinking of polymer polecule repeat units, which are already incorporated in molecular chains, G (XLU) can be defined as the number of polymer repeat units which have been crosslinked to others for each 100eV of energy deposited by a radiation field.

G (XL) refers to the number of crosslinks formed and is one half of G(XLU) since two units are involved in one crosslink.

Several investigators have studied the behavior of methylated polysiloxane both filled and unfilled (see e.g.

References 10, 11, 12) at low doses (megarad range).

Conclusions which can be deduced from this work includes o Crosslinking is proportional to radiation dose.

o crosslinking is independent of the type of radiation (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(IL) for methylated polysiloxane is in the range of 2.5 to 3.0.

At higher doses and as the available sites for crosslinking are expended, G (XL) approaches sero.

If for the purposes of illustration we assume the average value of G(IL) over a

]

large range of exposure is 1.4, then the irradiaton dose 7-5

l 1

- l

~- -

I in rads required to involve all potential units in crosslinks can be calculated:

100 ergs /gm 1 rad

=

1 megarad = 10 ergs /gm = 1020,yjg, 8

f n

1.6 For G = 1.4, then for evdry eV absorbed we have l

18 1 megarad -> L.i X 10 crosslinks 1.3 l

Polysiloxane (repeat unit) has a molecular weight of about--

74 grm and therefore:

1 megarad

> 1.dr74rio18 =

65x1020 crosslinks/ mole i

1.6 23 since a mole contains 6.023x10 units, we can calculate the dose to crosslink all units:

23 3

6.023rio erosalinks/ molt

= 9.3 x 10 megarads

.65x10" crosslinks/ mole 'megarad 10 So that approximately 10 rads is required to crosslink all units in methylated polysiloxane.

i Long before the accumulation of this dose (10 rada),

10 l

1 the polymer will probably be severely changed and the G(XL) 1 value will have changed.

One notes that if the G(XL) 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.

m 7-6

d 7.3 ' Ef fects g Radiation u & Phvsf emi g Mechanicai Prenerties g Methylaggd Pelvmilerane The data which are currently available relative to changes in physical and mechanical properties of Boraflex with increasing irradiation dose are somewhat limited. _It is therefore useful to examine some reported data for methylated polysiloxane found in the literature.

Hardnema. Tensile Streneth, d Elonention Tj;t Eg,3,gk Warwick * (see e.g. References 10 and 11) studied the effect of low dose (negarad range) gamma and, electron radiation on silica filled and carbon filled met'hylated polysiloxane.

It should be noted that in these experiments the polymer was prepared with fillers in.the range of parts per 100 parts of polymer which is significantly less than Boraflex (40 w/o BC filler).

The results of hardness, 4

tensile strength and elongation to break are shown in Table 7-1.

The hardness increases with accumulated dose 4

consistent with the data presented for Boraflex.

Tensile strength increases with dose initially and then decreases as the dose increases.

Elongation to break initially increases at low dose and then decreases to about a fifth of its initial value at 40 megarada.

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

a in also difficult to ascertain the true dose in rads _for every case, especially if both neutrons and gamma rays are present plus other effects.

Elastic Modulus:

12 In another report Bueche studied the effect of low dose (up to 60 megarads) electron radiation on polysiloxane including a series of unfilled methylated polysiloxane.

He found that, except for very low doses where presumedly polymer chain end effects are important, the change in elastic. modulus was well represented as a linear fit of the data with positive slope.

The theory of temperature and crosslinking on the tensile properties of polymers evolved from the analysis of the entropy change.

For an initial unit length, als the length in the stressed condition and the extension, e,

equals a-1.

Then the applied stress, f,

for a simplified 3-D crosslinked network is:

f = dr/de = -T(ds/de) = jo (RT/M ) ( a-1/a2)

(7-1) c where:

F = free energy a

= entropy T = temperature R = gas constant M8 = molecular weight between crosslinks

=

density j

For small extensions we can writa 7-8

em f = 3 oRTe/M (7-2) j e

Therefore:

E (modulus) = 3 ORT /M I7~3) j c

or:

E = 3pRTq/w (7-4) where:

q = crosslinking density (fraction units crosslinked) w = repeat unit molecular weight

-6 and:

E = 3;mP*

= 6.24 X 10 joRTrG(XL)

(7-5) where:

r

= absorbed dose (megarads)

R

= gas cohstant G (XL) = G-value Effect g Strens Durine Irradiation 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 structure.

While the effect of irradiation on the condition of the adhesive is not known, it is possible that the Boreflex at Quad Cities is accumulating dose under tensile stresses.

13 Bopp and Sisman irradiated polysilozanes under compressive stress in a jig.

After irradiation the jig was l

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 l

7-9

4

~

percent recovery to the initial configuration was 98t.

At a dose of 40 megarads recovery was limited to 27% of the initial configuration.

For a dose of 230 megarads the recovery was approximately 04.

While these data are for compressive strekses, it is possible that overall volume changes experienced by Borafier under tensile stress may not be the same as when irradiation occurs in the stress free condition.

Temperature Effects:

14 One investigator has studied the effect of temperatures on G(XL) kn methylated polysiloxane.

Using 2 mev electrons at one negarad/sec.,

G(IL) varied with temperature shown in Table 2.

The data suggests that G(IL) 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 Boraflex 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

]

7-10.

4

~~

another project in the near future.

o e

e se 6

7-11 4

' w:;

l TABLE 7-1 Effect of Low Dose Gamma and Electron Radiationggnyy Silica and Carbon Filled Methylated Polysiloxane i

Megarads Hardness Tensile Elongation Gamma or Strength to Break l

Electron psi l

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 53 916 158 561 117 3,

40 52 1

Carbon filled:

6 26 709 805 1

10 35 7 87 435 20 47 581 200 TABLE 7-2 l

Effect of Temperature on G(KL)

Temperature, *C G(KL) 150 4.7 20 3.1 0

2.8

-78 2.6 I

-i 7-12 l

I 4

De m

i e

O M

S E

m M

i 5

m 5

E!

O s

n m

g 5

m5 m

O y

m M

I I

0

-m-O D

w=

7-13

]

1 L _ _---

- - - - - - - - - - - - - - - - - - - - - - = - - - - - - - - - -

4 4

e e

S 94 5

5' y

a v

_2 5

O O

1 D

G m

\\

s i

g 5.-% -yj g

e U=

n 3

N N

l 2

n

- 8 5J 5A - d I

]

e.

l O

O

!~

M N-5 m, 5g q

+4 e

mm qu>

7-14

I1 l

l I

8.0 G&2 FORMATION, M GROWTH Ag g 2333 INTEGRITY QE t

BORAPLEX IE & 12))E ZggL 2QQL ENVIRONMENT From the' outset it should be noted that the mechanisms.

{

for gap formation and gap growth described are preliminary I

as the extent of data currently available is limited.

As

such, any conclusions drawn from this material are preliminary and may change as more data relative to Boraflex behavior under irradiation is documented.

Further experiments will' probably be required to determine the causes for all effects noted.

l l

8.1 Potentini Meehaninen d 152 formation A major uncertainty governing the mechanism of gap formation is introduced by a lack of kncwledge as to the condition of Dow silicone #999 adhesive used to afix the j

Boraflex to the stainless steel cell during manuf acture of 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

' perfect"; uniform and the mechanical properties of the Boraflex are uniform along its entire length.

j o

The bond is intact at the' ends of the Botaflex sheet f

62

.I 1

(i.e., region expected to receive the lowest dose) and-has failed or partially failed in the central region..

l In the first scenario, since the bond is postulated'to have f ailed along the entire length of Boraflex, the sheet is unrestrained pnd woul'd be' expected to shrink in a stress f

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.

j If the bond were ' perfect", as is the second scenario, high local stresses would develop in the~ sheet as the I

material tries to shrink and one might expect the material a

to tear at many locations forming many'anall gaps along the length of the sheet.

In the third

scenario, with the Botaflex 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 4

preferentially tear at or near its center.

This scenario seems to be supported at least partially by.the National Nuclear Measurements of gap occurence verus 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 j

l 8-2

9 Soraflex sheet more or less continuously along the longth.

The center of the approximately 6" wide panel over 2-1/2' to 3" is therefore initia11y' bonded whereas the outer 1.5' on 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 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 purpsse of illustration, consider the Boraflex sheet as essentially a one dimensional stress problem and we write for the stresses in the axial directions o =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 material ' snaps 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 is 10 psi.

This value is well below the as-produced tensile strength of 200 psi as well as the tensible strength for irradiated Soraflex cited in References 4, 5 and 6.

m e

1 If the effect of irradiation on the elastic modulus is considered, a crude order of magnitude estimate of the 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

change, then for e two decade increase in dose (say, 10 to 8

10 rads) the elastic modulus increases by a factor of approximately 100.

The observed increase in Shore A

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 6

8 going from 10 to 10 rada, an increase in the elastic modulus of 50 would be indicated.

In this case.the estimated stress is 500 psi.

Equation 7-5 also indicates that the elastic modulus 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 beat of the fuel.

This may be a factor 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 influence the axial distribution of gaps include the temperature dependence of the G-valun and a non uniform gamma dose distribution along the fuel assembly.

It was B-4

t a-=

noted in Section 7.3 that the G-value increases with temperature which would suggest more crosslinking, additional shrinkage and greater increase in the elastic modulus at higher temperatures.

Owing to the 1sek of specific data for Boraflex at low dose as well ps 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 which local stress could exceed the yield stress of Boraflex leading to the inception of tears.

It also provides one potential explanation of the observed axial distribution of gaps obrerved in the Special Tests conducted by National Nuclear Corp.

8.2 E_s t ima t e d Ma r t mum GAR h 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 10 probably complete at an exposure of 1 to 2 X 10 rads.

Figure 6-1 shows that at this exposure the maximum volume, reduction is approaching 20%.

If for the moment it is assumed that volume changes are isotropic, this would correspond to a

change in any

]

dimension of the Beraflex sheet of 6.664.

For a

sheet of 8-5

pI c

_l Boraflex152incheslong,thiswouldcotrespondtoamaxiEum gap of approximately 10 inches.

However, this estimate must be qualified with the following caveats:

The assumption.that, volume changes are isotropic may or o

may not be true.

In one. report it has been noted that

]

changes in sample thickness are greater than changes 6

in sample length and width,

o The estimate of aazimum gap size.is based on data ob-tained from small samples irradiated under essentially 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 l

assumes the Boraflex snaps back to its length had con-i l

ditions been stress free.

Furthermore, uncertainties

{

may be present owing to the extrapolation of test data 1

from small test samples to a 152' length of Boraflex.

l l

o It is not known what is happening at the upper and l

lower ends of the Boraflex.

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 may have been a factor in addition to that caused by 1

l l,

8

the gamma done.

Furthermore, the gamma source spectrum in the test reactor is likely to be different than in the Quad Cities Pool.

8 There are no test d ta in the range of 2.8 I 10 to o

9 10 10 rad's 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 neutron irradiation undergoes transmutation:

510 + n1 -> Li7 + He4 + 2.8 May A net result of this reaction is that boron carbide powder 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 10 University of Michigan after 2.5 2 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 1

8-7

m may take place in the materials thickness.

A review of the pre and post irradiation dimensions in Reference 4 at an 10 exposure of 1.6 X 10 rads shows typical shrinkage of 3-64 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 nazimum estimated 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 rate of shrinkage to be greatest at low doses and dimininishing as the G-value for crosslinking diminishes.

Accordingly, it is possible that the rate ( dG, 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 Isng Itra Inteerity of Bornfler in the Spent Zugl Environment After crosslinking saturates", the predominant radiation damage mechanism is likely to be scissioning.

At this point 8-8

i

)

from the data, it is not clear which bonds (Si-C or Si-0) fractdre most readily and have the highest frequency of fracture.

For Si-C, Si-o and C-B, the bond strengths 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 matrix.

Scissioning (either main chain or other) might cause specific atoms to be released from the polymer matrix 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 surface of the polymer thereby 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) may be an important f actor in influencing the rate of degradation with irradiation and exposure to the aqueous pool environment.

Long term out-of-pile testing of unirradiated Boraflex at elevated temperatures'.and in aqueous solutions has been 8-9

1

~

k

~-

l t

f reported (Ref. 16).

The samples of Boraflex were immersed in borated water (approximately 3000 ppm) at 240'F for 6200 hours0.0718 days 1.722 hours 0.0103 weeks 0.00236 months .

The borated water was neutralized with NaOH to a ph-i of 9.0-9.5.

Nter testing, the sample showed an average dimensional decrease along the side of the sample of 0.934 l

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 surf ace l

over the entire test period.

Gas evolution diminished with l

l 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 test period.

The evolved gas was analyzed periodically for f

composition and showed approximately 204 E, 534 CH, 54 CO 3

2 i

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.

Because of the difference between the test conditions and the pool environment, it is difficult to project long l

term integrity based on the test data.

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 borated water with NaOH used to acid neutralise the speciman 8-10

containers.

Review of selected chemistry logs from the' Quad Cities pool ir.dicates ph levels at the outlet of the pool domineralizer to typically be in the range of 5.0 to 6.0.

The bulk pool temperature of tis Quad Cities pool is approximately. 100'F.

Furthermore, it is believed these tests were conducted with unirradiated Moraflex although this is not stated explicitly in Reference 16.

7 The inspection of Boraflex at Point Beach indicates a steel gray substance on the surface of the Boraflex along some of the edges of the panel inspected as well as on sample coupots.

The substance tended to rub off when wiped with the inspector's hand.

On the inspected panel, the 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 its chemical composition.

Projections of the overall service life of Botaflex in a spent fuel pool environment are not possible at this time.

The results of a

larger program in which data from surveillance coupons from several U.S.

plants is gathered and evaluated may provide some answers.

i

=

8-11

I

___- _,, _ [

e q

e a

e

.1

l O

w

.g E

6-m

+w 5

6 4

d

.}.

9.0 REACTTVITY EFFECTS DE GAPJ H H g FUEfJRACK NEUTRON l

ABSOREER The shrinkage of Boraflex and subsequent formation of gaps in the Boraflex absorber panels results in a

redistribution of the neutron poison material in the spent fuel storage racks.

In the gap region, the absence of neutron absorber in one or more panels results in a net local increase in reactivity as well as an increase in the reactivity of the entire storage cell.

In the region of the fuel rack where gaps do not

exist, the Boraflex is undergoing a volume reduction and hence the density of B-1.0 j

etoms is increasing.

This effect would tend to decrease the reactivity of the fuel / rack storage cell.

The net overall l

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 17 dimensional PDQ model of the fuel / rack geometry. The gaps are conservatively assumed to occur at the fuel / rack a

midplane.

Cross sections for the PDQ model are generated

T i

using CASMO-2E Given the reactivity effects, a statistical model is i

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 net reactivity effect of each.

The program recomputes the not reactivity effect 10,000 times to obtain a good statistical sample.

This calculation pr'ovidea the mean increase in reactivity as well as the 954 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 calculations I

The Quad cities fuel storage racks have been analyzed i

assuming an infinite array (in the lateral extent) of the a

design basis fuel assemblies as defined in Reference 2.

The I'l design basis fuel assembly is an 8 X 8 BWR assembly

)

I containing a maximum enrichment of 3.2 w/o U-235 with 62 s

9-2 1

l

)

fuel rods and two water rods.

The assembly is assumed to be fitted with a Eircaloy flow channel which maxiLizes the 2

reactivity of the fuel / rack system.

The design parameters for the reference fuel assembly are given in Table 9-1.

To assure. that the actuni reactivity will always be less than the calculated reactivity, the following conservative assumptions were mades o The water in the fuel assembly and rack is assumed to be demineralized, unborated and at full water density In tbe x and y directions, the fuel / rack are infinite.

o I

o No credit is taken for parasitic neutron absorption in the fuel assembly grid spacers or and fittings.

o Each fuel assembly is assumed to be at maximum enrich-

~

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 19,

,21 An explicit been described in detail previously CASMO-2E model of the fuel / rack geonetry is used to calculate cross sections for the fuel, wt.ter gaps, tircaloy l

channel, stainless steel structure and Boraflex.

The l

H-factor option in CASMO-2E is used to provide transport theory corrected absorption cross sections in the neutron poison region.

Using thic

option, the macroscopic I

absorptien cross sections in the poison are iteratively adjusted until the infinite 2 multiplication factor as W

9-3

calculated by PDQ matches the corresponding value from the transport theory calenlation.

The cross sections in the region of the fuel / rack without gaps are developed assuming four panels of Boraflex' in the cell.

To obtain cross sections in the gap region, the Boraflex is replaced with water.

The cross sections so developed are then used in an explicit three dimensional PDC-07 model of one half fuel assembly / rack geometry as shown in Figure 9-1.

The fuel region is conservatively assumed to contain 150' of 3.2 w/o U-235 fuel, although the actual bundles at Quad Cities have a 6' natural uranium reflector on each end.

A 15 centimeter thick pure water reflector has been assumed at ths top of the assembly and no credit is taken for the stainless steel end fitting in the actual fuel assembly.

A tero flux boundary condition is assumed at the top of this reflector.

Since the Boraflex is not symmetric on the side of the cell wall, 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 accordingly a reflected boundary condition is applied at the center of the gap.

Figure 9-2 shows the mesh distribution in the x-y directions of the PDQ model.

Each fuel rod is represented 9-4

by one mesh interval as are the water gaps, channel, rack cell and Boraflex.

In the axial direction, a fine mesh spacing (typically 0.500* for large gaps) is used in the in the gap and in the vicin'ity of the gap region.

In the central portions of the fuel (away from the gap and reflector), a course spacing is applied.

In the reflector and in the fuel / rack regions in the vicinity of the reflector region a finer.azial mesh distribution is used.

Using this model, a reference calculatic, (no gaps) was performed and resulted in nn infinite multiplication factor (no axial leakage) of 0.9129 and a k,gg of.9105.

For the infinite media case the 0.9129 can be compared with 0.9155 f

as reported in Refere~nce 2.

The effect of gaps at the midplane was then computed as a function of gap size and number of gaps per cell.

The differen_lal reactivity (k,gg(w/ gap)

.9105) is then computed and is shown in Table j

l 9-2.

The data points in Table 9-2 are based on calculations 1

for a cell with panels containing 4-4' gaps, 2 - 4' gaps, 2 -2' gaps and 4 - 2* gaps.

The intermediated points have been developed using second order interpolation wbich 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 reaci:ivity shown in Table 9-2 is a very conservative calculation and 9-5

A e

mw can be viewed as an upper bound estimate of the reactivity increase due to gaps.

This is the case since the use of a reflected boundary condition in the x and y directions assumes that evpry stor' age 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 the 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 af raag occurrence The probability that a given event occurs is calculated using the following equation:

F (s) x F (n) x P I*)

II"1I P

a ijk g

3, k

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

.F (s) = fraction of gaps in cumulative gap size inter-g val s F)(n) = fraction of cells with n panels containing one l

or more gaps F (s) = fraction of gaps occurring in axial interval K k

The quantities F (s) and F(z) have been developed from the g

data obtained from the NNC special measurements as described in Section 4.0.

F)(n) has been computed based on the results of the 118 cells subjected to the standa-d test method.

The quantities so calculated are shown in Table 9-3.

9.3 Local and Clebal Reactivitv 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 foimer relates to the probability of a cluster of adjacent storage cells having large gaps all at the same axial elevation.

The latter is the net increase in fuel / rack reactivity based on the actual distribution of gaps.

To address local effects, Equation 9-1 can be u, sed to compute the probability of gap occurrence versus gap size

interval, number of panels / cell with gaps and axial 3i 9-7

elevation of the gaps.

Tables 9-4a, b, c, and d contain the probability of occurrence of gaps in size interval 1, 2, 3 and 4, respectively, an 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 22 the greatest gamma exposure For this population of 446 cells Equation 9-1 can be used to determine a distribution of the number of cells containing 4' gaps (worst case from a reactivity standpoint) can be calculated.

Table 9-5 shows that at any axial elevation less than 2 cells would' be 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 0.931 (954 probability /95% confidence level) 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 reactivity effect of each event.

Since the reactivity effects (Table 9-2) are based hn the assumption that all 9-8

gaps occur at the same axial elevation, F (3) in Equation k

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 l

sample.

The mean increase in reactivity so calculated is 1

+0.0023, +/.0037 (1 sigma level).

At a 95% probability 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 Model/ Method conservatinms

(

Throughout the previous discussion and description of j

i the models and methods applied to assess the reactivity

{

effects of gaps, several very conservative assumptions have been used.

These assumptions are summarized below to 1

illustrate the overall conservative nature of the approach:

)

i I

o The water in the fuel / rack reference calculation (Ref. 2) is assumed to be at 68'F whereas the pool wa-j 0

ter at Quad Citics is approximately 100 F which corresponds to a reactivity effect of +0.004, o

The fuel / rack geometry is assumed to be infinite in 1

I the x--y direction and all gaps occur at the same axial elevation (midplane).

This implies that every cell in 1

e 1

9-9 l

4 the rack has, for example, 4--4 gaps at the midplane which maximizes the reactivity effect.

In the present calculat' ions as well as in the reference o

calculations (Referen,ce 2), no credit is taken for parasitic nedtron absorption in the fuel assembly grid spacers or end fittings.

Every fuel assembly is assumed to be unirradiated'and o

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 3.2 w/o U-235 and contains Gadolina burnable poisons.

In addition, discharged assemblies 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 fuel rack where gaps do not exist.

For a 4' gap, the corresponding increase in B-10 'ansity due to shrinkage is approximately 84 o

If the NNC Special Measurements showed a panel with, for example, 2 -- l' gaps, the cumulative gap size (2") was used in developing F (s).

The reactivity g

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 9-10

'.j a..

1 1

c, l

~

gap falling within a gap size interval is assumed to be at the extreme of that interval.

For example, a gap detected at 2.5" falls in gap size interval 3 and for the purpose of reactivity calculations is assumed to be a 3' gap.

o F (z) is assumed to be 1.0 at axial interval 8 and zero g

elsewhwere.

This is equivalent to assuming all gaps occur at the fuel assembly midplane.

9-11 i

4' a

TABLE 9-1 FUEL ASSEMBLY DESIGN SPECIFICATIONS sx RR (Reference) o Fuel Rod Data Outside diameter, in.

0.483 Cladding thickness, in.

0.032 Cladding material Er-2 Pellet density, gra DO /cc 10.41 2

Pellet diameter, in.

0.410 Max. nodal enrich., wtt U-235 3.2*

Water Rod Data outside diameter, in.

0.591 Wall thickness 0.030 Material ar-2 Ngaber per assembly 2

Fuel Assembly Data Number of fuel rods 62 Fuel rod pitch, in.

0.640 Fuel channel outside dia., in.

5.438 Fuel channel wall thick., in.

0.080 Fuel channel material tr-4 Actual fuel assemblies have 6 inches of natural uranium at both ends of fuel rod.

y.~

t.,

l l

TABLE 9-2 DELTA K VS. NUMBER OF PANELS / CELL WITH GAPS AND GAP SIZE EFF NUMBER OF PANELS / CELL WITH 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 0.0100 0.0201 0.0335 4

9-13

t TABLE 9-3 l

F (s), F (n) and F (z) l 1

j k

FRACTION OF GAPS FRACTION OF CELLS HAVING VS n NUMBER OF PANELS / CELL GAP SIZE INTERVAL WHICS CONTAIN GAPS INTERVAL F (s) n F (n) i j

1 0.3214 0

0.3814 l

2 O.4643 1-0.2119 3

0.1429 2

0.2712 4

0.0714 3

0.1017 4

0.0339 l

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 ' O.0645 11 0.0645 12 0.0645 13 0.1290 14 0.1613 15 0.0645 9-14

', 4

I

\\

9 TABLE 9-4a l'

P FOR GAP SIZE INTERVAL 1 (0*-l' GAPS) ijk NUMBER OF PANELS / CELL WITE GAPS AXIAL F (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 0.0022 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 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 14 0.1613 0.0110 0.0141 0.0053 0.0018 15 0.0645 0.0044 0.0056-0.0021 0.0007 TABLE 9-4b P

FOR GAP SIZE INTERVAL 2 (l'-2' GAPS) ijk WUMBER OF PANELS / CELL WITE GAPS AXIAL r (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 0.0032 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 4

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

.y n 9-15 a..

~m TABLE 9-4c P

FOR GAP SIZE INTERVAL 3 (2'-3' GAPS)

I ijk NUMBER OF PANELS / CELL WITH GAPS AXIAL F (z)

INTERVAL k

1 2

3 4

o 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 0.0009 0.0003 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 TABLE 9-4d P

FOR 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' O.0002 0.0001 1

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 14 0.1613 0.0024 0.0031 0.0012 0.0004 15 0.0645 0.0010 0.0012 0.0005 0.0002 g

A

"M TABLE 9-5 NUMBER OF CELLS WITE 3'-4" GAPS IN THE REFUELINd RACKS NUMBER OF PANELS / CELL WITE GAPS AXIAL INTERVAL 1

2 3

4 1

0.0000 0.0000 0.0000 0.0000 2

0.0000 0.0000 0.0000 0.0000 3

0.0000 0.0000 0.0000 0.0000 4

0.0000 0.0000 0.0000 0.0000 5

0.2180' O.2789 0.1046 0.0349 6

0.8705 1.1141 0.4178' O.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 0.4351 0.5570E 0.2089 0.0696 11 0.4352 0.5570 0.2089 0.0696 12 0.4352 0.5570 0.2089 0.0696 13 0.8705 1.1141 0.4178 0.1393 14 1.0884 1.3930 0.5224 0.1741 15 0.4352 0.5570 0.2089 0.0696 M

9-17

Figure 9-1 Three Dimensional PDQ07 Model

~.,

WATER REFLECTOR FDEL ASSEMBLY N

\\

N EERO FLUX zIRCALOT CHANNEL mER msg pO2ArIONALsrxxtzR:

4 3

,,,I. E.,,, m -.

1 i

BORAFLEX PANELS Y

/

~

W

~

f N

%l.

SP T.

=

~

/'

1 gp w _

REFLECTED BODNDARY CONDITIONS e

9-18 9

\\

~

Figure 9-2 PDC 07 Mesh Description in the X-Y Plane "

=

t I

/

l FUE ROD BORAFLEX PANEL

/

OR WA N ROD WATER GAP k

ROTATIONAL-l WATER GAPS s

l ETRY f

ZIRCALOY CHANNEL l

STAINLESS STEEL CELL WALL REFLECTED BOUNDARY CONDITIONS 4

1 9-19 l

}

o 10.

CONCLUSIONS This report describes the results of a preliminary assessment of Boraflex 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 particularly true with respect to Boraflex shrinkage over the intermediate range of gamma exposures to which the Quad Cities racks have been exposed as well as the long term stability in the spent fuel pool environment.

Accordingly, as additional data becomes available, the conclusions developed as a result of the preliminary assessment could change.

Nevertheless, the study conducted herein points to the following conclusions with respect to observed behavior of Boraflex, irradiation damage mechanisms in Boraflex, the j

effect of manufacturing process, and the effect of gaps on the reactivity state of the Quad Cities spent fuel storage racks.

These conclusions include:

National Nuclear M Results:

I The special tests conducted by National Nuclear i

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 these measurements is approxImately 1.5',

the maximum gap i

10-1

4 size detected is 4.0.

Review of the axial distribution of gaps shows several l

characteristics (see Figure 4-2):

l o

There are no gaps in the lower section of the cells.

~

o There is a d.istinct 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 mechanisms which offer a

potential explanation of these characteristics have been discussed in Section 8.0.

The standard h1C 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 rc::t.1vity effects.

Audit gi Manufacturing Process The fuel rack manufacturing process was audited and the major conclusions of that audit includes o An adhesive (Dow Silicone 999) was used to afix the Boraflex sheet to the stainless steel sub elements whereas BISCO had tested Dow Silicone 732 (Ref. 3).

The manufacturer, (J. Cat) has stated that the twc adhesives are similar.

10-2

n.

q o

Little control was exercised during the applicatiori of the adhesive since its only intended purpose was to-i hold the Boraflex in place during assembly and welding of the various sub elements.

The adhesive was

~

applied in a discontinuous head 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 for the various matching stainless steel components.

Evaluation d Radiation Testina d Borafier Available data from the test programs sponsored by l

BISCO were reviewed and evaluated.

In this study changes in l

the sample specific gravity (pre and post irradiation) were used to compute changes in sample volume.

Based on this l

l 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 j

L 10-3 i

,P e

4

4 volume changes although there is some indication that the shrinkage may not be isotropic.

Some samples showed greater shrinkage in thickness than in width.

Gas evolution measurements during irradiation (in air) show that initially the samples produce off gasses consisf.ing of primarily 8 with N,

0, CO, CO "Ud 2

2 2

2 10 hydrocarbons.

Af ter an exposure of 1 X 10 rads, off gas production ceased.

Data in the range of exposures to which the Quad Cities racks have been exposed is not available.

The test irradiations were conducted in a test reactor environment.

The effect of neutron damage and differences in the gamma photon spectrum.(test versus pool environment) have not.been det er.nined.

Radiation Damaee Mechanisms in Borafier 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 cross 111. king saturates (at a 10 dose estimated to be approximately 1

x 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

~

10-4

1 E

of.the material.

Scissioning may explain the oliserved swelling of the Boraflex samples in the BISCO test program and the tendency of the material to absorb water (i.e.,

Point Beach examinations).

ggg Formation, ggg M d Lggg Zggg Brability 31 Boraflar g h M M 2g,gl 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 in Boraflex 1

may cause the elastic modulus. to increase substantially.

Accordingly, as a restrained Boraflex panel tries to shrink, relatively large local. stresses (estimated in excess of'the tensile strength) develop, potentially causing a tear.

After a tear occurs, the gap formed ~ continues to grow as the Boraflex is subject to additional shrinkage until l

10 crosslinking saturates (estimated at 1-X 10 rads gamma based on test reactor irradiation).

The magnitude of the maximum gap size is difficult to project owing to primarily two factors.

First, there is a lack of volume change data 8

10 in the exposure range of 10 to 10 rads.

Second, it is not known whether_the Boraflex manufacturing process causes shrinkage to be anisotropic.

There is some indication that 10-5

4 Boraflex shrinks less in length and width than in thickness ~.

A worst case estimate of maximum gap size of 10' is obtained

}

l assuming shrinkage is isotropic.

The degree of anisotropy 1

l could alter this substantially.

A preliminary model to-I l

o predict gap growth developed as part of this study will be completed under an ongoing EPRI program when low dose data is available.

The rate of Boraflex 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 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 B C powder in Boraflex cannot be 4

projected at this time.

The qualification program conducted by BISCO examined radiation effects and long term exposure to an aqueous environment separately.

The combined effects after crosslinking saturates and scissioning predominates may likely depend on such factors as pool water chemistry, water temperature, and local flow conditions around the Boraflex panels.

The coupons inspected at Point Beach show 1

10-6

o.

that 'the material tends to soften and rub off after 10 exposures of 1 to 1.5 X 10 rads.

While the coupons showed accelerated sof tening and erosion of the Boraflex, similar local effects were notgd 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 CASMO2E and PDQ07 to determine the upper bound reactivity effect as a function 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 total of 10,000 occurrences are sampled and the mean increase in reactivity as well as the 95% probability -- 95%

confidence level is determined.

The analysis demonstrates that at a 95% 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 l

that the reference fuel assembly design used for these analyses is considerably mote reactive than the most 1

10-7 1

l r.

reactive assembly used in practice at Quad cities.

This is i

i the case since the effect of gadolina burnable poison rods in the reference fuel has not been considered.

To demonstrate that future gap growth can be accommodated vithin appropriate desi'gn

limits, the following is recommended:

o Identify the most reactive fuel

assembly, when 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.

o Assess the effect of gaps on the reactivity state of the racks using this fuel assembly at the bernup cf peak reactivity which is in practice the limiting case.

1 It is estimated that this approach can provide another 0.06 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 utilizing a neutron radioassay system providing ' resolution equivalent to that employed for the \\ special tests.

The same cells inspected using the

special seat method 4tf ter the last' outage should again be assayed to provide additional data with respect to:

s 3

o Rate of gap growth a

e l

10-8 l

t s

'i s

4 3,

'}

}

.)

5

o Formation of new gaps o

Axial distribution of gaps o

Gap size distribution It is anticipated tirat this two prong approach will provide the additional data and analysis necessary to assure that future gap growth can be accommodated within the design limit of.95.

l e

10-9

lo m

RIJ3EFJGT.2 1.

Special Neu' tron Attenuation Test for High Density Spent Fuel Storage Racks (Wet), National. Nuclear Corporation for CEC 0' Quad Cities, CECO P.O. No. 309763, December, 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 High Density Fuel Racts NRC Docket Nos. 50-254 and 50-265, March 12, 1982.

3.

"The Effects of Irradiation on Adhesively Fastened Boraflex', BISCO Test Report No. N-39, August,.iS81.

4.

' Irradiation Study of Borafles Neutron Shielding

)

Materials", BISCO Report 748-10-1, Rev. 1, August 12, i

1981.

j 5.

Burn, R.B., Blessing. G., ' Radiation Effects of Neutron Shielding Materials, Trans. Am. Nuc. Soc., 32, 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).

7.

Letters 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.

i i

8.

"Boraflex Neutron Shielding Material, Product Perform-ance Data", BISCO Report No. N-38, 748-30-2, August 25, l

1981.

9.

B111meyer, F.W., Textbook of Polymer Science, Inter-science, New York (1982).

l 10.

Warwick, E.L.,

Industrial Rng. Chem., 11, 2388; 1955.

11.

Warwick, E.L.,

Piccoli, W.A. and Start, F.O., J. Amer.

Chemica3 Soc., II, 5017 (1955).

j 12.

Bueche, F., J. Polymer Sci., 11, 297 (1956).

V 00' g

D i

l 13.

Bopp, C.D.,

and Sisman, O., Nucleonics 11 (7), 28 (1955).

Bopp, C.D.,

and Sisaan, 0., ORNL 1373 (1953).

14.

Charlesby, A., ' Atomic Radiat en in Polymers", Pergamon Press, (1960),

e 15.

Strasser, A., Tario, W., Goldstein, L., Lindquist, K.,

and Santucci, J., " Control Rod Materials and Burnable Poisons:

An tvaluation of the state of the Art and Technolo3y Development, July, 1980', Electric Power 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. H-2, Test 748-21, August 25, 1978.

17.

Pfeifer, C.J., PDQ-7 Reference Manual II, WAPD-TM-947 (L), February,1971.

l 18.

Edenius, M., Ahlin, A., Eaggblom, E., 'CASMO-22, A Fuel j

Assembly Burnup Programs User's Manual', Studsvik/NR-8 1/3, November, 1981.

I 29.

Letter, J.D. O'Toole to S.A. Varga, ' Application for Amendment to @erating License *, Docket No. 50-247, November, 19, 1935.

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, 'Prcposed Technical Specifications Regarding Fuel Enrichment *, Docket No.

50-286, June 13, 1986.

22.

Letter, J. Boeller to K. Lindquist, March 4, 1987.

l l

t I

4 8

ie F ENCLOSURE 2

,,=

REQUEST FOR ADDITIONAL INFORMATION FROM CHEMICAL ENGINEERING BRANCH ON DIABLO CANYON SPENT FUEL POOL RE-RACKING 1.

Based on the recent experience pertaining to degradation of Boraflex in spent fuel pools at Quad Cities and Point Beach nuclear power plants, provide justification to demonstrate the continued acceptability of Boraflex for application in the Diablo Canyon spent fuel pool.

2.

Based on the recent information, provide any changes to the in-service surveillance program for Beraflex neutron absorbing material and describe the frequency of examination and acceptance criteria for continued use.

Provide the procedures for testing the Boraflex material and interpretation of test data.

l 3.

Describe the corrective actions to be taken if degraded Boraflex specimens or absorber is found in the spent fuel pool.

9 4

4 1

l l

)

v t e Letter Sequence Other
TAC:64470, (Open) TAC:64471, (Open)
Initiation Acceptance... Administration Questions 2 June 1987 Answers Results Approval Other: ML20237K236
MONTHYEAR1987-06-17 [Table View]

ML20237K236

Text

Dear Mr. Shiffer:

SUBJECT:

Enclosures:

Subject:

Dear Mr. Davis:

Enclosure:

1.0 INTRODUCTION

3.0 DESCRIPTION

=

=

Navigation menu

Search