ML20246N004

Applicant Exhibit A-1,consisting of EPRI Final Rept NP-6159, Assessment of Boraflex Performance in Spent Fuel Storage Racks, Dtd Dec 1988

ML20246N004
Person / Time
Site:	Saint Lucie
Issue date:	01/24/1989
From:	Haley T, Kine D, Lindquist K NORTHEAST TECHNOLOGY CORP.
To:
References
OLA-A-001, OLA-A-1, NUDOCS 8903270283
Download: ML20246N004 (159)

Text

- - - - - ,

- 4 EPRI NP6159 l Topics.

Project 2813-4 l Spont-fus) storage Final Report j

' Neutron absorbers . . . .

December 1988 Material testing ..' i ' E." '

' arch Institute 8tric Power

~89 WR 22 P6 :55 grc e. .

vvCr L : . ,

'an, l

. . 2. w .a An Assessment of Boraflex icn G. .. Performance in Spent-Nuclear-Fuel Storage Racks l

l O'

NUCLEAR REGULAT~aY COMMISSION i I

Docket in the matter tio.of53-33FMd.[

f/mA - b V u 4 r& Official

- 5b b Exh. No.

Stdt IDENTIFIED W Applicant RECEIVED  ;

intervenor REJECTED Co,1t'; Ofl'r Ca.nracior _.DATE bMi[

Other Witness Reporter MNT% /J o fr %

Prepared ty Northeast Technology Corp.

Kingston, New York O .

9903270283 890124 PDR ADOCK 05000335 o PDR

REPORT

SUMMARY

SUBJECTS High-level radioactive waste management / Ught water reactor fuel TOPICS Spent-fuel storage Material testireg Neutron absorbers AUDIENCE Fuels engineers / R&D scientists An Assessmant of Boraflex Performance in Spent Nuclear-Fuel Storage Racks Experimental data and theoretical models Indicate that Boraflex, a neutron-absorbing material, will shrink during irradiation. Further-more, the physical properties of Boraflex can change with time and radiation exposure. Thuo, EPRI recommends that utilities conduct active surveillance programs to monitor the long-term performance of this material and thereby determine maximum service life.

BACKGROUND High-density racks for storing spent nuclear fuel in water filled pools are designed for a 30 to 40-year service life. The use of neutron absorber materials to control reactivity allows fuel storage at maximum density. One such material, Boraflex (manufactured by BISCO), contains boron carbide in a matrix of polydimethyl siloxane or silicone rubber. At one plant, small gaps formed in the Boraflex, and at another plant, Boraflex test coupons deteriorated unexpectedly. EPRI initiated this project to explore the factors determining long-term Boraflex performance.

OBJECTIVES

• To collect and evaluate data from utility coupon surveillance programs and from in-pool neutron radioassay measurements of spent-fuel racks.

.To evaluate data from test irradiations of Boraflex.

. To develop guidelines for Boraflex coupon surveillance programs.

APPROACH The project team collected and evaluated coupon and neutron radioassay data from nine utii; ties. They also evaluated data from earfy Boraflex qualifi-cation tests and recent irradiation tests. The team reviewed the literature to identify the mechanisms of radiation damage for materials similar to Boraflex and formulated models to prodlet the maximum shrinkage. They then compared model predictions with the compilation of experimental data. Using the results of th!s wrsk, the team dowloped guidelines for con-ducting coupon surwillance programs.

RESULTS Experimental data and theoredcal predictions of Boraflex savW were in general agreement. Cross linking in the polymer matrix of Boraflex causes the material to shrink when exposed to gamma radiation. The shrinkage EPRI NP4159e

~

_ _____---__----- - - - - - - - - - - - - - - - - - - - ' - - - - - - ^ ^ - ^ ' ^ ' ^ ~ ^ - - ^ ^ ~ - ~ ~

stops 'when crose linking saturates at a gamma exposure estimated c9 i about 1 x IOS rad. The projected maximum shrinkage is 3-4%.

The study also identified mechanisms for gap brmation that may result from shrinkage, as well as design and fabrication features of fuel racks l O that may contribute to gap formation. The maximum cumulative gap ^

size in a storage rack panel is estimated to be about 4-6 in.

Data were not available to estimate the combined offects of gamma radiation and long-term exposure to tho' aqueous pool erwironment.

Although no study findings suggest a rapid or dramatic change in the -

physical properties of Boraflex that would affect its neutron-absorbing function, EPRI recommends coupon surveillance programs to monitor its long term performance. An appendix to the t6,' ort provides guide- j lines for such programs.

EPRI This pro, lect makes an important contribution to understanding the fac- ]

PERSPECTIVE tors that affect the performance of Boraflex in spent-fuel storage racks.

The study has addressed one of the primary concoms with Boraflex use by presenting data showing that shrinkage of the material will be limited and that employing good design and fabrication techniques can eliminate gap formation.

Howoor, some uncertainty remains about the combined effect of radia-tion and extended exposure to an aqueous pool erwironment. There-fore, EFRI caations that continued surveillance of Boraflex is prudent to I assure adequate long term perforniance.

PROJECT RP2813 4 EPRI Project Manager: R. W. Lambert Nuclear Power Division O' Contractor: Northeast Technology Corp.

For further information on EPRI research programs, cd EPRI Technical information Specialists (415) 855-2411.

O V

An Assessment of Boraflex Performance in Spent-Nuclear-Fuel Storage Racks NP-6159 i

Research Project 2813-4 Final Report, December 1988 Prepared by NORTHEAST TECHNOLOGY CORR 26 Pearl Street Kingston, New York 12401 Principal Irwestigators K. Lindquist D. E. Kline Contributor T. C. Haley Prepared for Electric Power Research Institute 3412 Hillview Avenue Palo Alto, California 94304 EPRI Project Manager R. W. Lambert High Level Waste Program Nuclear Power Division

v O

v ORDERING INFORMATION Requests for copies of this report should be directed to Research Reports Center (RRC), Box 50490. Palo Arto CA 94303, (415) 965 4081. There is no charge for reports requested by EPRI member utdities and affiliates, U.S. utility associations, U.S. government agencies (federal, state, and local), media, and foreign organizations with which EPRI has an information exchange agreement On request, RRC will send a catalog of EPRI reports.

0 f

)

}

l O

i l

Elearc Powee Research inanuie and EPRI are ropeered serven rnarks of Electre Poaer Research inettuie. tre Copynght C 1968 Electre Power Research instaute. 6rc A3 nghts reserved NOTICE The report .es prepared ty tne organemon(s) named bes o . as en account of .crk sponsored ty the Electre Power Research Ingdule, irc ([PRI). Nether EPRI trw..nDers of EPRI. the orgaruation(s) named beoo., nor any person peng on behalf of any or thorn (a) makes any marranty. express or anphed. wrth respect to N use of any Wormetori, apperatus, method, or process drecioned m INe report or that such use r'isy not elffage pmetely owned nghts, or (b) assumes any hebdites won respect to the use of, or for damages resulting frorn the use of, ary Wormeson, apperous, method, or procese eedosed m she report Prepared by Nor;hees Technology Corp

,. Kegeon. New wk

O I

ABSTRACT i l

Data from utility surveillance prograss, test reactor Irradiations and the open literature have been collecttd and evaluated to eusens the effect of service enviror.mont in spent nuclear storage eccks on the neutron absorber material, Boraflex.

Radiation Induced changes in the properties of Boraflex have been Identified. The observed formation of gaps in the full length panels j of Boraflex in some spent fuel racks has been attributed to one such change, shrinkage, in combination with mechanical rettraint.

Mechanisms of gap formation and growth are also discussed. I 4

Factors which may influence the ultimate service life of Boraflex in spent fuel storage racks have been ident! fled. Continuation of coupon surveillance programs to verify the serviceability of Boraflex in the spent fuel pool environment is recommended. Guidelines for coupon surveillance programs have been developed.

e 4

111 f-. .

l Ol l l'

ACKNOWLEDGMENTS 1

In the course of this project Northeast Technology Corp (NETCO) has l had discussions with nine utilities. The data and comments pr,ovided were Invaluable to this p r e,t ect and NETCO hereby acknowledges and expresses its sincere appreciation for their help. Specifically, these utilities were:

e Alabama Power Company e Carolina Power and Light e Commonwealth Edison Company ,

o Duke Power Company e Florida Power and Light Company e Northeast Utilities )

e Northern States Power Company  ! I o Sacramento Municipal Power District I o Wisconsin Electric Power Company l

In particular, the contributions of Commonwealth Edison Company and Wisconsin Electric Power Company are gratefully acknowledged. The 4 valuable comments of the Electrt: Power Research Institute staff and, in particular, R. Lambert, on the final report as well as assistance provided throughout the project is sincerely appreciated. The test irredletion data provided by BISCO, the manufacturer of Boreflex, are l also acknowledged.

ll l

O v

_ _ _- _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _________m. . _ . _ _ _ _ _ . _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __

O CONTENTS l Sectton h 1-1 1 INTRODUCTION 2 TYPICAL NEUTRON ABSORBER CONFIGURATIONS IN 2-1 SPENT FUEL STORAGE RACKS Spent Fual Storage Rack Types 2-1 Boraflex Encapsulation and Retention 2-3 l In Spent Fuel Storage Rocks 2-3 Wrapper Design Pleture Frame Design 2-6 Absorber Insert Design 2-6 REVIEW AND EVALUATION OF DATA FROM UTILl1Y 3-1 3

SURVEILLANCE PROGRAM Scope of Utility Coupon Programs 3-1 l

Results of Utility Coupon Programs 3-3 l

I Quad Cities Neutron Radioassay Measurements 3-22 Turkey Point Neutron Radioassay Measurements 3-26 4-1 4 REVIEW AND EVALUATION OF BISCO TEST' DATA Early BISCO Quellfication Testing 4-1 Dimensional Changes 4-2 Gas Evolution Data 4-5 1 Mechanical Properties 4-5 l 4-7 j Noutron Attenuation Long Term Exposure to Hot Water 4-7 Recent BlSCO Radiation Testing 4-9 Dimensional Changes 4-10 {

Sample Weight Changes 4-15 l l

Shore A and D Hardness 4-15 f l 4-18 .i Nsutron Attenuation 4-16 f Visual Appearance Specific Gravity Measurements 4-18 j Discussion of Interim Report Conclusions 4-20

' S vii l

.- 1

-f CONTENTS (continued)

O Mac+1en M DISCUSSION 5-1 5

Radiation Damage Mechanisms and G-Values 5-1 Change in Specific Gravity and Specific Volume 5-5 Versus Dose Changes in Boreflex Dimensions Versus Dose 5-7 Changes in Specific Volume Versus Geometric Volume 5-12 Gap Formation and Growth 5-14 l Water Permeetion and Coupon Wolght Changes 5-18 l Test Reactor Versus Spent Fuel Pool Radiation Conditions 5-26 Boraflex Service Life 5-27 6 CONCLUSIONS AND RECOMMENDATIONS 6-1 7 REFERENCES 7-1 APPENDIX At PROPERY:ES AND CHARACTERISTICS OF POLYSILOXANE POLYMERS A-1 APPENDlX B: RADIATION EFFECTS IN POLYSILOXANE POLYMERS B-1

\O))T l

APPENDlX C: GUIDELINES FOR A STANDARD 12ED BORAFLEX COUPON SURVEILLANCE PROGRAM C-1 l

l 0

O viii

___. "9

G' ILLUSTRATIONS t

F f gor a EAg3 2-1 Flux Trap Fuel Storage Rack for Unirradlated PWR Fuel 2-2 l 2-2 Eggerate Fuel Storage Rack for PWR and BWR Fuel 2-4 2-3 Wrapper Design Fuel Storage Cell 2-5 l

2-4 Alternative Type of Wrapper Design Fuel Storage Cell 2-7 s

2-5 Picture Frame Design Fuel Storsgo Cell 2-8 2-6 Absorber Insert Design Fuel Storage Cell 2-9 3-1 Utility Coupon Measurements: Post Irradiation Shore A and Shore D Hardness 3-15 3-2 Utility Coupon Measurements: Percent Change in Coupon Length or Width versus Pool Residence Time 3-17 4 3-3 Utility Coupon Measurements: Percent Change in Coupon Weight versus Pool Residence Time 3-19 3-4 Utility Coupon Measurements: Percent Change in Coupon Neutron Attenuation versus Pool Residence Time 3-20 3-5 Boraflex Gap Size Distribution, NNC Special Test Measurements at Quad Cities 3-25 3-6 Axlal Distribution of Gaps, NNC Special Test Measurements at Quad Cities 3-25 4-1 Not Change in Specific Volume versus Gamma Exposure 4-4 4-2 Gas Evolution from Boraflex During Radiation Exposure 4-4 ,

4-3 Percent Change in Sample Length and Width versus l Gamma Exposure 4-11 4-4 Linear Fit of Percent Change in Sample Length versus Gamme Exposure (Co-60 Irradiations Only) 4-13 4-5 Percent Change in Sample Thickness versus Gamme Exposure 4-14 4-6 Percent Change in Sample Wolght versus Gamma Exposure 4-16 ix 9

-w

_ _ _ _ _ - - - - - + - - - - - - - - - - - - - - - ' ' - ' ' ' - ' ' '

--,m__

ILLUSTRATIONS (continued) l 4-17 Shore A and Shore 0 Hardness versus Gamma Exposure h 4-7 4-19 4-8 Specific Gravity and Specific Volume of Boraflex versus Gamma Exposure 5-1 Specific Volume Change of Boraflex versus GammaEx 5-6 Tests 5-8 5-2 Change in Boraflex Length and/or Width versus l

Gamma Exposure '

5-13 5-5 Change i n Boraflex Geometric and Specific Volume versus Gamma Exposure 5-15 5-4 Estimated Range of Boraflex Open Porosity versus Gamma Exposure induced Strain in a Restralned Pgnolto of 5-5 Shrinkage 5-19 Boreft x Over the Exposure Range of 1 x 10 i

1 x 10 Reds 5-20 l 5-6 ElasticModulusoftheNg-1toPolymer1 x 10 9 0verthe Rads  !

Exposure Range of 1 x 10 EstimatedStressinaRestrainedganelofBogaflex 5-21 5-7 to 1 x 10 Rads

() 5-8 Over the Exposure Range of I x 10 Maximum Cumulative Gap Size as a Percentage of 5-22 initial Boraflex Panel Length versus Exposure Scanning Electron Microscopy of the Flat Surface of 5-24 5-9 Boreflex et 270x and 2800x Magnification Scanning Electron Microscopy of the Cut Edge of 5-25 5-10 Boraflex et 280x and 2800x Magnification l O .

X e

m-um um

I i

O TABl.ES Tahlet g

Sources of Surveillance Program Data 3-2 3-1 Scope of Coupon Data Collected and Evaluated 3-4

. 3-2 3-3 Extent of Coupon Surveillance Program Measurements 3-6 Pro irradiation Characterization 3-4 Extent of Coupon Surveillance Program Measurements Post irradiation Characterization 3-7 3-5 Shore A Hardness as a Function of Sample Thickness (Unteradiated Boraflex) 3-14 4-1 Fractional Change In Coupon Weight, Specific Gravity, True Volume and Specific Volume versus 4-3 Gamme Exposure 4-2 Gas Evolution from Boraflex During Reactor -

4-6 irradiation Effect of Gamma Radiation on BISCO NS-1 Polymer 4-8 A-3 l

l q

1 o

i O

xi

O Section 1 INTRODUCTION When the nuclear power stations now in operation or under construction were designed it was envisioned that spent nuclear fuel would be stored at the reactor site for a period of 6 months to one year.

After this Interim cooling period, the fuel was to be shipped to a j reprocessing plant for separation of the fissile isotopes and recovery from the residual fission products. In the mid 1970's, delays in commercial reprocessing and ultimately a moratorium eliminated fuel reprocessing as a near-term option. Reactors at the time had storage capacity for one full core plus one or two additional discharge batches of fuel. Accordingly, the abandonment of commercial reprocessing and subsequent delays in the Federal Repository have caused a severe shortage of at-reactor spent fuel storage capacity.

One alternative to expand fuel storage capacity was to remove the original storage racks and replace them with racks which offered higher fuel packing density. The original racks supplied with these plants rolled on large center to center spacing- between fuel assemblies to control reactivity. Typically, PWR fuel assemblies were ,

arranged in the racks in a square array with a specing of as much as 21 inches between centers. Higher packing density, i.e., closer center to conter assembly spacing, was achieved by taking credit for the stainless steel structure of the storage racks and/or by Incorporating plates or sheets containing a neutron absorber material for reactivity control.

One such neutron ebsorbing material used quite extensively since the late 1970's for reactivity control utilizes a polymer, dimethyl  !

polysiloxane or silicone rubber, as a matrix to contain the neutron l absorber material, baron carbide. This material is Boratlex (Trademark of BISCO). The sole purpose of Boraflex is to provido

.t - 1

- ---_--.-_m____________ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ . . , , _ _ _

i

reactivity control in the spent fuel storage racks.

• I.

i l Recent examinations of surveillance coupons and an Irradiated Boreflex sheet at Point Beach 111 as well as neutron radioassay measurements of the Boraflex In the spent fuel storage racks at Quad Cities 121 have behavior of 'the Boraflex absorber. For Indicated unanticipated example, the Point Beach coupons showed a discoloration 'of the Initially shiny black Boraflex to a steel gray color and some tendency Also, some of the of the material to sof ten and rub of f when handled.

and coupons exhibited mechanical dameDe such as broken- corners chipping attributed to handling when the coupon cladding was removed.

While the overall condition of the full length panel was good, some discoloration and softening of the Boraflex along the edges, similar to the color of the coupons, were observed.

Subsequently, a neutron radioassay technique was used at the Quad Cities Pool to verify the presence of Boraflex in the spent fuel racks. This investigation showed the presence of small gaps (or the l ~

absence of Boraflex) In some of the fuel storage cells. The gaps 1/2 In. (approximately the limit of ranged in size from about detectability) to 4 in.. In a separate study, funded by Commonwealth Edison, the neutron radionssay measurements were evaluated ill. The formation of gaps and their subsequent growth was postulated to result from shrinkage of a restralnad sheet of Boraflex due to radiation Induced crosslinking between adjacent chelns in the polymer ill. l The present evaluation is a follow-on to the work performed for Commonwealth Edison and represents a broadening of the scope to include data collection and evaluation from a number of surveillance and test programs currently underway in the Industry. Specifically ,

the scope of this project includes j

l .

e Utility Poll and Data Collection: A poll of utilities with spent fuel racks utilizing Boraflex<was conducted. l The status of each coupon surveillance program was established and where available, data were requested.

() e Data Evaluation: The data received from the utility 1-2 i

__ -Q

coupon prc;---? has been reviewed and evaluated. In addition, recent test Irradiation of Boraflex samples  ;

in a Co-60 facility as well as a test reactor at the University of Michigan were also evaluated. These later Irradiations were performed by BISCO, the manufacturer of Boraflex, after the Quad Cities gap phenomenon was observed.

e Literature Search: A computerized search of the liter-oture was conducted to identify relevant studies of re- '

dietton effects in materials similar to Boraflex. This study was initiated to quantify the radiation damage

' mechanisms in Boraflex and to provide a basis for pro-Jocting the long term effects of irradiation on Bora-flex.

i e Gap Formation and Gap Grow +ht Based on the data col-lected and evaluated, postulated mechanisms for gap for-mation and growth have been developed. A gap growth model provides a method for projecting the maximum gap size and a basis for evaluating the worst case

• reactivity effects of gaps.

e Recommendations: Based on this comprehensive evalua-tion of the available data, recommendations as to areas where further research, testing and surveillance may be desirable have been developed. These recommenda-tions are intended to provide data to determine the ade-quacy of Boraflex in the spent fuel pool environment for i extended service life.

As data were collected it became apparent that the scope of the l Individual surveillance programs varied from' utility to utility.

This, is some cases, made the data evaluation and Interpretation l difficult. In order to assure uniform data of high quality in the future, guidelines were developed for a " Standard Boraflex Coupon Surveillance Program". These Guidelines, included .as Appendix C to this report, are Intended to supplement, not supplant, any test procedures recommended by BISCO or the fuel rack manufacturers. The Guldelines are generally intended for new rack Installations where sufficient quantitles of Boreflex can be specified for coupons. Some elements of the Guidelines may be applicable to existJng programs.

1-3 O

O Section 2 TYPICAL NEUTRON ABSORBER CONFIGURATIONS'lN SPENT FUEL STORAGE RACKS SPENT FUEL STORAGE RACK TYPES High density spent fuel storage racks utilize nsutron absorber materials in different ways depending on the type of fuel the racks are Intended to receive, fuel rack design and other factors. An overview of the various types of high density fuel racks currently In )

use has been given previously ill. Accordingly, only those features of fuel rock design important to Boraflex behavior are addressed here. )

l This Includes the manner In which the Boraflex is encapsulated and j retained within the rack structure.

High density spent fuel storage racks can generally be classified as being of either two types: a " flux trap" design or an "eggcrate" design. For PWR applications, both types are currently used and in some cases both may be used, in the same pool. For BWRs, on the other hand, the designs are exclusively of the "aggerate" type.

The " flux trap" type of rack is designed to store unirradiated PWR fuel with enrichments generally up to 4.5 w/o U-235. A fuel storage  !

location is formed by using a square tube of stainless steel typically 0.060 to 0.100 In. In thickness with inside dimensions of 8.5 to 9.0 In as shown in Figure 2-1. The storage cells are typically arranged in a square pitch with center to conter spacing of 10.5 to 10.8 in..

Four neutron absorber sheets are located on each side wall of the stainless steel tube and closure plates of stainless steel are provided to capture the sheet. The water gap between storage cells (flux t r a.p ) provides an effective means to thermalize neutrons maximizing the effectiveness of the absorber. A lattice of fuel storate cells is created by either connecting the In6ividual cells at the corners or by placing the cells in a rigid framework structure.

l Ol, 2-1

1

_. \

_ i i

Fuel Assembly Fuel Assembly Flux Trap Water Gap l

Fuel Assembly Fuel Assembly l

10.510.8 In. -

Fuel StorageTube Neutron Absorber Plate (4 per storage cell) Inner and Outer Shell of Sta!ntess Steel Figure 2-1. Flux Trap Fuel Storage Rock for Untrradiated PWR fuel.

O 2-2 I

l l

i The "eggerate" type of rock differs from the Sflux trap" rack in that the water gap is eliminated and each sheet of neutron absorber l

material is common to the adjacent cell. Figure 2-2 Ilivstrates a typical "eggerate" storage array for PWR fuel. Cells are generally arranged on a square pitch with conter to center spacing of 9.0 to 9.2 l

In. for PWR fuel. In determining the rasctivity state of the j fuel / rack system, credit is taken for fuel burnup. Accordingly, the racks are designed to store Irradiated PWR fuel which has accumulated a specified exposure. For BWR fuel types, racks sistler to those in Figure 2-2 are used for both unfrradiated and Irradlated fuel.

Typical conter to center cell spacing for BWR fuels is 6.0 to 6.5 In..

l in both "eggerate" and " flux trap" designs, the cavity created to encapsulate the Boreflex is vented to the pool water. This allows for an escape path for gases produced when Boraflex is subjected to gamma radiation. This Is designed to preclude bulging of the stelnless steel closure plates due to pressure bulldup of the offges if the cavity was sealed. Venting therefore allows entry of the pool water and contact with the Boraflex.

BORAFLEX ENCAPSULATION AND RETENTION IN SPENT FUEL STORAGE RACKS The rack manufacturer receives Boraflex from the supplier pre-cut to

• the size appropriate for a given design. Typically the material is used in thin (0.040 to 0.11 In. thick) sheets 5.0 to 8.5 in. wide and j approximately 12 ft. In length, in the as-produced condition, 1

i Boraflex exhibits many properties of an elastomer and as such is not free standing. Means must therefore be provided in the design of the rack to support the material and prevent it from slumping. This Is accomplished in a number of ways depending on the rack designer. The methods of Boraflex encapsulation and retention used in the majority of racks currently Installed are discussed subsequently.

"Wrnener" Danten A typical wrapper design storage cell is shown in Figure 2-3. The structure of the cell is provided by a thick wall (0.060 to 0.100ln.)

Inner tube of stainless steel with square cross section. The 2-3 e A _ _ _ _ _ _ _ _ . - . _ . _ _ _ . _ _ _ _ . - - ------

D -

MEME MEME l

- EMEE E E El R +--8 5-9.2 in.-*

Fuel Assembly / PWR BWR *-6.0-6.5 in. --

Stainless Steel Rack Structure Neutron Absorber Plates ob R a b B ue i i

2-4 m

O O

, n y

f f M f i 1 1 - c,

,r 1 a s Stainless Steel Cell Wall ,

Boraflex

/

l

/ l

)

N #

i

'q .

r r

1 1

Stainless Steel [

Wrapper Plate -

Figure 2-3. Wrapper Design Fuel Storage Coll.

2-5

Boraflex sheets are located on each face length.

of the structural tube to Thin (typically correspond generally to the active fuel 0.020 in, thick) stainless steel sheet (s) are then place over the and along the length of the Boraflex and spot welded at the corner (s) tube. In some cases manufacturers use an adhesive to hold the Boraflex in place during tube assembly and wrapping.

In Another verlation of the wrapper design is shown in Figure 2-4 this case, two L-shaped sheets are pressed firmly over the Boraflex and spot welded to the structural tube at diagonally opposite corners.

1. Pfetura FremaN nectan in the "pleture frame" design, an Inner, heavy well stainless steel tube is used to provide structure for the storage cell as in the l wrapper design. Instead of using a wrapper, a cavity is formed on the exterior of each side of the cell using stainless steel closure strips tack welded in place. The closure strips are attached on all four sides of the cavity in a " picture frame" fashion as shown in Figure j 2-5. The Boreflex is rolled in place and a thinner gage stainless steel closure plate is wel'ded in place.

l While the basic element of the Quad Cities racks is a cruciform (not a square tube as shown in Figure 2-5), the picture frame method is used to enclose and encapsulate the Boraflex. During the manufacturing process for these racks, an adhesive (Dow Sillcone Sealant No. 999) {

l vas used to hold the Boreflex in place during rack assembly.

1. Abencher ineme+# Dacfon l

The " absorber Insert" design utilizes removable cartridges to retain the Boraflex sheet in place. The cartridges are designed to silp in f between stainless steel storage cells. The cartridges are assembled I

from a series of U-channels as shown in Figure 2-6. Two V-shaped f f

grooves are rolled into the channels along their entire length to exert a positive force against the Boraflex sheet to hold it firmly in -

place and prevent slumping. .

2-6 1

- m 1

O

)

3

[ Stainless Steel Cell Wall i

Boraflex l

N I)

% j L Shaped Wrapper Plate Figure 2-4 Alternative Type of Wrapper Design Fuel Storego Coll.

l l

l e;

2-7

- - - - l

O Inner Stainless Steel Cell Wall Picture Frame -

Stalniess Steel 5

Strip l

l Boraflex g x

fStainless Steel g 7 Closure Plates s

1 Figure 2-5. Picture Frame Design Fuel Storage Coll, 1

lO 2-8 I

l .

G Fuel \

Fuel Storage Storage

_[L Cell Cell Absorber Insert inner Stainless teel Channel Outer Stainless Steel s Channel Boraflex

\ s,

/

V Grooves ><~.%. =

Figure 2-6.

N/

Absorber Insert Design Fuel Storage Cell.

O 2-9 L

The obsorber Insert design is used in the Point Beach spent fuel storece rocks. No edhesives were used to b0nd the Boraflex to the channel well 1.11.

i O

O 2-10

~

O i Section 3 REVIEW AND EVALUATION OF DATA FROM UTILITY SURVEILLANCE PROGRAMS Those U.S. utilities with spent fuel storage racks utilizing Boraflex had been previously identifled 1.d,1. As of this writing, nine utilities have responded with data contributions solicited under EPRI project RP-2813-4 The data received consist of pre Irradiation characterization of small coupons and post Irradiation examinations and measurements of the same samples. In one case, two full length absorber Insert panels were removed from the racks and examined.

These data have also been included in the current evaluation.

In addition, measurements were conducted with neutron radioassay equipment at the Quad Cities and Turkey Point Stations. The former data were previously evaluated under separate contract. A summary of the results of neutron radioassay data are also provided. Table 3-1 contains a compilation of utilities which contributed data to the program.

SCOPE OF UTILITY COUPON PROGRAMS Based on observations made of utility coupon surveillance data received to date, most utilltles with spent fuel racks uttilzing i Boraflex do have a coupon surveillance program in place, but the specific elements of the programs very from one utility to another.

The programs generally consist of a series of samples prepared from production batches of Boraflex used for the specific racks. The samples are clad or enclosed in metal (generally stainless steel) and range In dimenelons from 2 in. x 2 In. to 8 in. x 12 in. A series of coupons are placed in the pool at the time of rack Installation, l adjacent to Individual storage cells designated for discharged fuel.

In some programs the coupons receive accelerated gamma exposure by placing freshly discharged fuel next to the coupons after each 3-1 e

J

Table 3-1 SOURCES OF SURVEILLANCE PROGRAM DATA

_ ,,f1=+m_ Tuna _

,, , , ,P 1 m rt+ . _ . _ _ ,

_,, p + t I t +x _ _ __

coupon Farley Alabama Power coupon, neutron Commonwealth Edison Quad Cities 1 radioassay coupon Quad Cities 2 coupon Carolina Power & Light H.B. Robinson 2 Oconee 1, 2 coupon Duke Power Company ,

Turkey Point neutron radio-Florida Power A Light assay coupon Northeast Utilltles Millstone 2 coupon Northern States Power Pralrle Island

[

G,f coupon Sacramento Municipal Rancho Seco Utility District Point Beach 1, 2 coupon, full pan-Wisconsin Electric als Power O 3-2

)

refueling. The surveillance programs call for the periodic removal of coupons for examination. While the frequency of coupon removal verles l

from program to program, a typical program would call for removal after 1, 2, 4, 7, 11, 15, 20, 25, and 35 years of service.

During ccupon preparation, the Boreflex is generally, but not always, procheracterized with respect to the following attributes:

e Sample weight l

• Sample length, width and thickness l e Shore A and/or D hardness e Neutron attenuation

{

I When the coupons ere removed from the pool they are visuelly examined and generally subjected to measurements similar to those conducted during the procharacterization. In some cases, only nominal as-built dimensions are avsilable and this makes evaluation of dimension changes difficult. With few exceptions, the Integrated gamma dose to which the coupon was subjected is not accurately known. In those f I cases where the dose is known it has been either rigorously calculated or estimated based on the operating history and discharge schedule of s fuel placed adjacent to the coupon sample train. Table 3-2 summarizes i the scope of each of the utility surveillance programs. Tables 3-3 l and 3-4 contain a summary of the pre and post Irradiation coupon characterization for each program. The results of coupon inspection are described in the following sections.

RESULTS OF THE UTILITY COUPON PROGRAMS vicuni Framinn+1nn Fnrlav. Coupons were removed from the Ferley pool after I and 2 years of residency. In addition to the 1 year Irradiated coupon (01282A),

an unirradiated control coupon (03152C) was examined. The irradiated coupon experienced a major increase In hardness and had lost most of 3-3 9

m

v Table 3-2 SCOPE OF COUPON DATA COLLECTED AND EVALUATED Pool Residence Nominal Cou- Time (mos.) or pon Size Gamma Dose Coupon ID (Inches) (rads)

Plant -

4x6x.060 12 FarIey 01282A

• 12 05152C*

• 24 01142E 2.75x3x.11 11 Mllistone 2 la

• 11 2

1(.25)x4.5x.075 68 Oconee 1/2 AD16 AD25 A027 A045 AD49 AD51 N1 2 x2r:.11 12(4x109 )

Point Beach "

N2 "

N3

" 24(1.0x1010)

S4 "

O S5

$6 N7 1.44x10 10 NB " " i N9 10 )

" 1.55x10 N10 "

N11 *

• N12 10

• 1.60x10 N13 84 Panel # 8x144x.11

• 84(1.0x1010)

Panel 8x12X1.25 6 Pralrlo Island 05220C/01 05200E/04

" 12(5X109 )

4x4x.070 3 Quad Cities 1 05102A " 6 043028 " 12 04272D 60 04502D

" l 1

I

" 3 Quad Cities 2 05112C " 6 ,

05192A

• 12 i

05192C l

---

(CONTINUED)-------

i O 3-4 ,

9 l

Table 3-2 SCOPE OF COUPON DATA COLLECTED AND EVALUATED (Continued)

Pool Residence Nominal Cou- Time (mos.) or 'l pon Size Gamma Dose Plant Coupon ID (Inches) (rads) l Rancho Seco 1 2x4x0.085 3 l 2 *

• 3

• 6 1 4 a a )

$ " 12 6 " # l H.B. Robinson MM48A N/A 6 l MM488 " a i MM48C 8 a j

i

'Untrradiated Controls

\

O 35

___ ^'

Table 3-3 EXTENT OF COUPON SURVEILLANCE PROGRAM MEASUREMENTS Pro Irradiation Characterizations Neutrcn Shore Dimen- Attenua- A Hard-We!ght slons tion ness Other Plant x x x Farley Millstone 2. x x x Oconee 1/2 Point Beach > x N/A N/A x x x Pralrlo Island Quad Cities 1,2 x x x x Boron content Rancho Seco x xI x x H.B. Robinson x 0 , Nom nal dimensions only.

O 36

m Table 3-4 I

EXTENT OF COUPON SURVEILLANCE PROGRAM MEASUREMENTS l

Post Irradiation Characterization:

Neutron Shore A /

Dimon- Attenus- D Hard-Plant Weight sions tion ness Other Farley x x x Millstone 2 x x x x Oconee x x x x Modulus of Rup-ture Point Beach x x x x Full panels in-spected. Dose i calculated.

Pralrle is. x x x B-10 content Dose estimated.

Quad Cities x x x x Boron content I and 2 Rancho Seco x x x x  ;

H.B. Robinson x - x 37

__ _ _ _ _ _ . _ . _ _ _ _ _ _ _ _ _ _ - _ . _ _ _ _ . _ . . _ _ _ . _ _ _ . _ _ . _ _ _ 8 _

Its flexibility G which is expectedCoupon based03152C on tho' was BISCO reactor essentially Irradletions discussed subsequently. noted on coupon 01282A; were unchanged. Slight discolorations described in the however, the nature of the discolorations were not of the 2 year coupon (01142E) was report. The general appearance in that the only significant similar to that of the 1 year coupon loss of flexibility.

change was an overall increase in hardness and I after Two coupons were removed from the Millstone 2 pool was e Wille+nna 2 coupon from capsulo #1 residency. One an eleven' month The other coupon from capsule unirradlated and was used as a control. f rom discharged fuel although l

1. 2 had been exposed to gamma radiationIt has been reported M that the is not known.

the integrated dose has becomo significantly harder, more fragile and  ;

Irradiated materiel changes and neutron absorptivity changes stiffer. Minor dimensional f Changes in Boreflex color are noi reported.

were also observed.

each were inspected after of three coupons Ornnan. Two sets The samples were approximately 5 1/2 years residence in the pool. d prior to l configured as tensile specimens and were not procharacterize samples had It has been reported m that all l placement in the pool. in about one millimeter a grey coloration along the edges extending i from the edge.

Radioassay measurements showed radiation levels of 1 NR/hr gamma and 3-4 MR/hr bata at a distance of about I centimeter.

brittle and cracked easily The coupons were reported to be very Although the gamma dose to which the coupons were during handling.

subjected was not reported, Shore A hardness 9 rads. measurements Indica least 1 x 10 they probably received a dose of at The surveillance program at Point Beach has included the '

En.l.n1 Baach.

of coupons starting one year after the fuel racks periodic removal The coupon sample train was subjected to accelerated l were Installed.

gamma radiation exposure by placing a freshly discharged fuel The latest coupons examined assem adjacent to It following each refueling. 10 reds.

The had received a gamma dose in the range of 1.0 to 1.6 x 10 Co.

Wisconsin Electric Power examinations reported were conducted by Lo 3-8 1

k

(WEPCO) personnel and an Independent laboratory. The results of these examinations have been documented elsewhere ill. The summary reported here is based on Reference 1 and discussion with WEPC0 personnel. The major findings includes e The samples showed degradation in the form of cracks and chipping thought to have resulted from packaging /

shipping or unpacking / mechanical damage.

e The samples had changed in color from a shiny black to a greylsh color.

e immersion of samples into clean water darkened the water with a graphite-like particulate material shed from the Boraflex. Successive immersions into clear water darkened the water, but each time to a lessor ex-tent. This in turn lightened the Boraflex to a whit-Ish grey color.

e Several samples showed evidence of thinning in the spent fuel pool environment.

e Radioassay of the coupons Indicate the permeation of pool water and containments (primarily Cobalt-60, Coslum-134 and Silver-110.)

e The coupons yloided dust or powder when rubbed.

Since there were significant differences in geometry and encapsulation method of the Boraflex in the coupons and in the fuel racks, two full length panels were removed for inspection. These panels had exportanced 6 to 7 years of residence in the pool. The Point Beach racks are of the removable absorber Insert design as described in Section 2. One panel had received a high gamma exposure (1 x 10 10 rads) and the other no significant exposure. The inspection results have been summarized by WEPC0 personnel ill as follows:

e The unfrradiated panel was " brand now" In appearance with the exception of a whitish powder which covered the panel where It was in contact with the stainless steel clad.

• Radioassay of the unfrradlated panel showed very lit-tie beta / gamma activity Indicating little permeation of pool water.

e The Irradiated panel showed good overall Integrity with no missing pieces, no cracking or other degradation.

Some discoloration (greying) was evident along the edges of the panel ano anpaared as randomly occur-39

Ing scallops. The largest such discoloration was about 5 inches long and penetrated 0.5 to .75 inches into the panel. The scalloped areas covered approxi-O mately 1-25 of the total surfece area of the panel.

The gray areas yloided a dust or powder much like the coupons when rubbed. ,

I e Radioassay of the Irradlated panel Indicated water less permeation in the nondiscolored areas although than the coupons. Radiation levels on contact with the grey areas were aboutThisseven timesthat suggested those of the water may  ;

nondiscolored areas.

be preferentially permeating the panels at the edges.

e The length of the panel was not measured during the examination so that no definitive conclusions The panelwith widthre-gard to length changes can be made.

was measured, however, and the shrinkage based on the

• nominal width including manufacturing tolerances have been computed as described subsequently.

As a result of this inspection, the following observations were offered by WEPCO relative to the progression of changes observed in >

Boreflex 111:

e When Boraflex is immersed in the spent fuel pool with-out gamma radiation, it appears to retain its "Ilke ,

O- now" appearance.

e When gamma radiation is present, the material changes. i The changes include color,'and from a material of good ,

Integrity to one that is brittle and yields a powder when rubbed. Once the Boraflex reaches the crev condl-tion, weight loss, thinning and other change,s in physt- ,

cal properties occur. '

e The presence of gamma radiation also appears to allow water permention initially along the edges where the material was cut to size. The flat surface of the panel seems to be more resistant to permention of the pool water.

e WEFC0 believed that the onset of discoloration, permea-integrity tionandsubsequepgchangesinthematerial begin at about 10 rads gamma.

Besed on the good overall condition of the Irradiated panel, after 6-7 years in the Point Beach pool, WEPC0 concluded that the Boreflex was expected to retain its serviceability for another 10-20 years.

O 3-10 m'

Discussions with WEPCO personnel 1A1 have indicated that silica contamination in the spent fuel pool watar has been noticed with levels in the range of 3 to 5 ppm measured. Possible sources of sillcc have been Identified a6 Boraflex and/or a grease used on some underwater tools. An Irradiated sample of Boraflex was placed in a I beaker of water for an undisclosed period of time at the Point Beach chemistry lab. The water subsequently tested positively for silica.

As discussed in Appendix A, Boraflex is believed to contain a significant amount (20 to 25 w/o) of a silica compound.*

I Prmfrfa faland. The surveillance program at Pralrle Island utilizes To date two coupons large coupons (8" x 12") clad in stainless steel.

beve been removed and examined, one after six months and one after one f l

yerr 121. Freshly discharged fuel had been placed next to coupons to dose accelerate the accumulation of gamme exposure.

The g ar.ma received by the 1 year coupon has been estimated at 5 x 10' reds.

Visual inspections 11L.1.1.1 showed "waterspotted" areas on the one year coupon which covered more of the coupon than the earlier sample.

Roylew of the photographs of the one year coupon cannot conclusively identif y the " water spotting" as being the same type of discoloration ,

There are, however, as that observed as the Point Beach panel.

similarities such as scalloped discolorations at the coupon edge and corners.

Quad ct+ fan. Coupon data from the programs at both Quad Cities 1 and Four coupons from the 2 have been received and evaluated (12 13 141 Quad Cities 1 program have been removed to dato corresponding to 3, 6, 12 and 60 months pool residence time. Three coupons at 3, 6 and 12 months have been Inspected from Quad Cities 2. These coupons have received relatively low gamma exposure with the dose for the 12 month coupon estimated at less than 10 reds. According to Commonwealth 8

• Private conversations with BISCO Indicate the silica content may be less than this estimat6 although the exact composition of Boraflex has not been provided.

3-11

l Edison (CECO) personnel, the general appearance of four of the coupons some of their initial i

the coupons had lost although I was good Two samples from Quad Cities 2 (coupons 05192A and C) flexibility. (coupon 043020) showed small and one sample from Quad Cities 1 The ,

bilsters over approximately two thirds of their surface .125 In, in area.

dlumeter.

size up to about bilsters ranged in Embrittlement was observed in the blistered area as well and it reported that the material was very easily torn in those areas.f Radioassay of a corner of the blistered region Indicated permention o The color of these coupons was described as a greylsh the pool water. B, bilstering or bubbles in As discussed in Appendix shade. in the literature and is irradlated silicone rubber has been reported result of believed to be associated with trapped off-gas produced as a the radiation-induced crosslinking process.

l+ ls not kn.wn why only three of the coupons showed blistering , l although this ma' be related to variations in some process These are verlable the f during the manufacture of these batches of Boraflex.

only coupons for which data was collected during this project which showed the formation of blisters.

Two coupons each efter 3, 6 and 12 months residence time Rancho Saca. The exposure reconved by these l have been removed to date 11,11.

coupons is not known although freshly discharged fuel has been placed e

next to the coupon train to accelerate the accumulation of gamma dose. l it has been reported 11.1,1 that the condition of all coupons was The 3 month coupons were found to be only '

generally the same. The one year l slightly flexible compared to unfrradlated specimens.of the one year Photographs coupons were very hard and brittle.

coupons showed greying discolorations along one edge to a depth of ,l about 1.5 In.. I l R6hfnton.

Three coupons from the accelerated surveillance program at l H.B. Robinson have been removed after six months pool residence for These coupons were subjected to gamma radiation examination (16 17).

from freshly discharged fuel to accelerate the accumulation of gamma in f dose. The coupon configuration is that of a tensile speelmen clad O 3-12

stainless steel. Coupon dimensional data was not provided. Neutron I attenuation and hardness measurements were made. Visual examination reveoled no indication of coupon cracks or other signs of physical deterioration lill. The coupons were hardened but retained significant flexibility.

Shnra A/D Hardnaea All coupons for which data were received were subject to pre and post l l'rra d i a t i on Shore A hardness measurement. In'two programs (Robinson and Mllistone 2),

Shore D hardness was also measured. The pro Irradiation Shore A hardness measurements show considerable variability (from 66 to 85 on the scale). It is not known whether this variability is real or whether, in some cases, the measurements were made on coupons too thin to give an accurate result. ASTM D 2240-86 as well as supplier of Shore A durometers recommend that samples bs at least 0.25 in. thick and for thinner samples that several layers be I

stacked to this thickness 1181. To test this hypothesis, Shore A measurements were made on successive layers of nominally 0.040 In.

Boraflex. The results are shown in Table 3-5.

t i

i Since the cumulative gamma dose is not known for many of the coupons, I the post Irradiation Shore A and D hardness is plotted as a function l of pool residence time in Figure 3-1 Referring to the Shore A measurements in Figure 3-1, all coupons registered above 90 on the Shore A scale except those from the Quad Cities program which had

{ received a gamma exposure of less than 10' reds. It is therefore probable that these are two classes of coupons in the matrix of samples evaluated in this study, low dose (less than 10' rads) and high dose (greater than 10' rada and beyond). It should be noted that when the Shore A scale exceeds 90, the manufacturer recommends that a D scale durometer be used 1121. The hardness data in Figure 3-1 shows no obylous correlation with pool residence time since many of the samples with relatively short residence time received accelerated gamma radiation exposure.

O 3-13

. j Table 3-5 SHORE A HARDNESS AS A FUNCTION OF SAMPLE THlCKNESS O (UNIRRADIATED BORAFLEX)

Number of Layers Thickness (In.) Shore A Hardness l

1 .038 87 +/-25 2 .078 82 +/-1 3 .117 80 +/-1 4 .156 79 +/-1 5 .195 78 +/-1 6 .234 78 +/-1 7 .274 77 1/2 +/-1 i 1

• Estimated repeatabfifty.

O 2.i.

O 85 g l Key 3 75 - A Mllistone 2 M @ H d Robinson 2 O J M Farley j 65 - 3 Oconee  !

M @ f Point Beach g X Pralrle Island l

- l E 55 - " O Quad Cities 1 f

9 Quad Cities 2 l E  ;; E Rancho Seco I

8 J- -

5 100 -

X -Er s s E E D 48

EH -

. O

+

2 90 B 9 O l 5

e l

O O I I 80 O 25 50 75 Pool ResidenceTime (Months)

Figure 3-1. Utility Coupon Measurements: Post -

Irradiation Shore A and Shore D Hardness.

3-15 O

m

• l I

01 man =1nnat channas The percent change in coupon length3-2.

in Figure andThewidth dataare plotted as a included function of pool residence as shown Irradiation in Figure 3-2 are only for those coupons where pre available so that changes in dimensions can be measurements are in cases where coupon degradation due to Inferred. Furthermore, been omitted.

l handling and shipping was significant, the data have included

' reds).

Accordingly, (examined after 1 only coupons N1, N2 and N3 from Point B low shrinkage (less than I The data seems to again fall in two classess Of those coupons in the l 1.5%) and high shrinkage (2.5 to 3.5%). xposure latter9 class, the Point Beach coupons had the received hig (4 x 10 rads) reds. It is Interesting to note that been estimated at 5 x 10'  % whereas the length change for the 1 year Pralrlo Island coupon is 3.3 Indicative of no width change is only 1.05. This could be On the other hand, both dimensions in This the anisotropic shrinkage. same amount.

I Point Beach coupons changed by approximately the e Introduces the question of variations in the production process of

(~N Boraflex, perhap,s depending on filler compositlen, sheet thickness, etc..

' f Most of the Quad Cities coupons showed less than 1.5% shrinkage In length und width with no preferential shrinkage in either direction.

coupon, on the other hand, exhibited 8

The 5 year Quad Cities 1 l 4 in length and 1.25% in width. While the dose shrinkaps of about 2.4. Island coupon and the Farley coupons '

received by the 6 month Prairie Indicate that these coupons is not known, hardness measurements  % or less.

received a fairly high dose yet shrinkage is limited to 1.5 The 11 month Mllistone 2 coupons probably showed received a or very little fairly no high dose based on Shore D hardness measurements butMeasurements o shrinkage in either length or width. It has been showed a not increase increase in this dimension of 2 to 35 In coupon thickness may be postulated Mi,1 that this attributed to water absorption.

! O 3-16

n --

O

. i i 0 -

XU O Key A O J M Farley A Millstone 2 O O f Point Beach X Prairie Island O Quad Cities 1 e X e Quad cities 2 g - 1.0 o -

3?

O 3 o O t; O j, Oe E

.s 8

a 8 - 2.0 - -

& o h

3 o

+

e Q- - 3.0 -

I l 0 25 50 75 Pool Residence Time (Months)

Figure 3-2. Utility Coupon Measurements: Percent Change in Coupon Length or Width versus Pool Residence Time.

O 3 17

M-  ::--- --.. --- - - - . , , . . -

For those coupons where the gar.ma exposure is known, the measured l shrinkage is included in a composite Figure in Section 5 along with-the BISCO test data and Quad Cities gap data. '

4 cnunen Wafcht

-The percent change in coupon weight is plotted es.a function of pool residence time in Figure 3-3. As in the previous case, samples with a exhibiting severe deterioration have not been piece missing or

• l Included. While the data shows considerable verlebility, the general trends are either very little sample weight loss or a weight gain. l l

This may be due to the onset of permeation of the samples by pool water as was observed in the high dose Point Beach coupons and Irradiated panel where this was Indicated by the absorption of redloactive contaminants. One exception is the Quad Cities 1 five year coupon which showed a 55 loss in coupon weight. Thia may be due to a missing piece of the coupon although other than the bilstering, no such coupon deteriorttlon has been reported (13 141 e

chanea in Neutran Attenuation The percent change in neutron attenuation pre and post Irradiation versus pool residence time is shown in Figure 3-4. The change in neutron attenuation for most coupons is generally within a band of

+/- 25. The two exceptions are two of the Point Beach coupons which had been in the pool 5-6 years and the six month Rancho Seco coupons. .

The Point Beach coupon reportedly 1.1.1. had significant damagie which mcy have been a contributing factor to the decrease in neutron attenuation. The pre-irradiation attenuation data for the Rancho Seco coupons were made on " representative" control samples of Boraflex (15) and not the actual semple removed from the pool. This may explain the relatively large decrease in neutron attenuating cherectoristics of these coupons.

To place the neutron attenuation measurements in perspective, the following example is cited. For a neutron absorber matorist with 3-18

O' l

5.0

• I I -

Key A Millstone 2 f Point Beach X Prairie Island Y '

O Quad Cities 1 9 Quad Cities 2 2.5 - E Rancho Seco _

E 3 O c

s d U X E O -

i =4 3 e

0 + '

e lii n.

l

- 2.5 -

- 5.0 I I 0 25 50 o

75 Pool Residence Time (Months)

Figure 3-3. Utility Coupon Vaasurements: Percent Change in Coupon Wolght versus Pool Re:Idence Time.

O 3-19 m

n

s O

l 2.0 l l 1

+

8 + + 1 l

08 5 O 0

E l e E O

o is O + j e

~~

e

- 2.0 -

+ .

O

+

l se ,

z

.c

! 1

~ '

2 - 4.0 -

o TE e '

$ E Key I

" l 5 O J M Farley 4 Point Beach ~~' '

- 6.0 - O Quad Cities 1 O Quad Cities 2 l i

E Rancho Seco i 1 50 75 0 25 Pool Residence Time (Months)

Figure 3-4 Utility Coupon Measurements: Percent Change in Coupon Neutron Attenuation versus Pool Residence Time.

O 3-20 t

n Initial Boron-10 areal density of .025 g B-10/cm2 a " rule of thumb" estimate is that e decrease in neutron attenuation of 1.05 corresponds to a 105 reduction in B-10 areal density. The neutron attenuation measurements are relatively insensitive to small changes In B-10 areal density since at 0.025g B-10/cm 2 the material I,s " black" to thermal neutrons.

The relatively large verlebility and changes in neutron attenuation measurements from the coupons programs are probably attributable to factors other than the loss of boron (except perhapc for the 5-6 year Point Beach coupons which had experienced thinning of 25 to 305).

These factors would include, as in the case of Rancho Seco, the use of different pro and post Irradiation samples and also differences in the neutron spectrum used in the pre and post Irradiation measurements.

Differences In the location of the measurements on the specimen could also introduce variability since there are verlations in coupon thickness from location to location.

In summary, the data from the utility coupon programs Indicate that as expected under gamma treadiation in the pool the Boraflex coupons undergo changes. These changes include shrinkage, a loss of flexibility, embrittlement and an increase in hardness, change in color from shiny black to a whitish greyish shade, absorption of pool water, and the formation of blisters in some coupons. The major conclusions from the coupon data relative to these changes can be summarized as follows:

e The shrinkage data are quite variable but Indicate a maximum shrinkage in any direction of 3 to 45. Shrink-age may not be uniform in all directions but may be anisotropic.

e Changes in welght exhibit verlability. This may be p tielly attributed to how the various coupons were prear-conditioned (dried) prior to weighing and also to miss-Ing chips in some cases.

Indicate that permeation of pool waterRadioassay measurements may into the Boraflex which might Indicate a not increase In coupon welght.

e At relatively low doses (109 rods or less) the material looses its initial rubber elasticity. The hardness In-O 3 21

~_ ,

I

'~'

and the material becomes creases to full scale (Shore A)At somewhat higher doses, The the material changes brittle.

color from the Initial shiny black to whitish grey. i discoloration begins initially on the edges of the cou--

pons and progresses inward with time.

e The verlsbility of some of the neutron attenuation measurements can probably be attributed to the use of different control samples and/or differences in the )

spectrum of neutrons used In the pre and post Irradla-tion measurements.

I i

QUAD CITIES NEUTRON RAD 10 ASSAY MEASUREMENTS Inspections were After the results of the Point Beach coupon and panel avellable, Commonwealth Edison Initiated a series of neutron {

radioassay measurements of the Quad Cities spent fuel storage racks.

These tests were conducted by National Nuclear Corporation and the The data data from these tests have been documented elsewhere 121.

from these tests have been evaluated previously 131 and the test methods and results are summarized below.

O Tatt Em+hnds The Two tyt's of neutron radioassay tests were used at Quad Cities.

252 neutron first, iermed the Standard Test Method, utilizes a Cf source and four BF-3 proportional detectors which are sensitive to thermal neutrons. The equipment is so designed so that each detector is adjacent to one panel of Boraflex when it is placed in a fuel storage cell.

252 source The BF-3 detectors do not record fast neutrons from the Cf but do detect thermal neutrons which have been transmitted through the cell wall, thermalized and scattered back into the cell containing the detectors.

If the cell wall contains Boraflex, the back scattered neutrons are significantly attenuated whereas where gaps exist, the backscattered neutrons undergo significantly less attenuation.

() During a measurement, two passes were made in each cell--first from 3-22 O

the top to the bottom and then from the bottom to the top. The count rate is continually recorded and a peak in the count rate is indicative of a discontinuity in the Boraflex absorber. The Standard Tests are considered a "go-no-go" type of measurement but do not provide an accurate Indication of the size or axial location of anomallos.

1 After anomalles in the neutron absorber were Indicated in the initial testing campaign, a special test method was developed. This method utilizes a Cf 252 source and two He-3 proportional detectors and is Intended to provide a measure of neutron attenuation in a single panel of Boraflex in a storage cell. The detectors are wrapped in lead (to reduce the potential of gamma interaction) and cadmium to form a one half Inch high window in the front of the detector sensitive to thermal neutrons. The source and the detector housing are suspended from the refueling bridge mast. During measurements, the detector and source housing are moved in one half Inch increments through a storage cell.

The mest position and detector count rate are continuously recorded. By comparing the shape of recorded peaks in the count rate ,

with peaks from measurements on a calibration standard containing gaps of known size, the approximate size and axial elevation of the gaps can be determined.

Results of Test Mancuraments in the initial testing campaign using the Standard Test Nothod, a total of 203 panels containing the Boraflex absorbers in the refueling rack modules were tested. The refueling rack modules are those adjacent to the refueling canal and which received freshly discharged fuel during each of the two refueling outages. These modules had therefore received the highest gamma radiation dose. Of the 203 panels tested, 77 panels showed Indication of anomalies or gaps in the Boreflex. In the rack modules which had received a lower dose, a total of 103 panels were tested of which 18 showed Indication of anomalles. Although the dose to the rack modules has not been rigorously calculated, the dose in refueling racks has been estimated to be 10 9 reds. Review of the magnitude of count rate peaks recorded (

3-23

In these measurements conducted by CECO personnel Indledte that the average gap size In the refueling rack modules is larger than those in other rock modules which were subjected to a lowur gamma radiation dose. Based on those panels determined to have gops from the standard tests, 28 panels were selected for testing using the special test method. All 28 cells were In the region of the pool containing racks designated as refueling racks.

l l For the purpose of evaluating the special test data, the following procedure was used. If the data showed Indication of a gap in the l range of 0.0 in. to 1.0 In., the occurrones is defined as falling into ]

I Gap Size Interval 13 gaps In the range of 1.0 in. to 2.0 in. as Gap Size interval 2; and so on. In e steller manner, each Boraflex panel was divided into 15 axial Intervals or bins approximately 10 In.long. l I

I l

For the 31 gaps detected In the special tests, the averegn gap size is l 1.35 in. +/-0.87 In. (1 sigma). Of the 28 panels tested, three 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 Figure 3-5.

The average cumulative gap size on this basis is 1.5 in. +/- .85 in, and the maximum gap detected was 4 in.. j The axial distribution of gaps is shown in Figure 3-6 in which the number of gaps versus axial Interval is plotted. This distribution exhibits several characteristics which should be noted:

o There are no gaps in the first four Intervals.

e There appears to be a peak in occurrence around the mid-plane of the cell.

I e There appears to be a second peak in occurrence near the top of the cell.

The number of data points In Figure 3-6 is limited; however, the data does suggest that the axial distribution of gaps may be bimodal. In order to test this hypothesis the Student t-test was applied to the 3-24 wh

O 20 . . . .

18 -

16 -

14 -

E g 12 -

3 10 -

$8 E

Q 6 -

4 _

2 -

0 01 12 23 34 Gap Size Interval, In.

Figure 3-5.

Boraflex Gap Size Distribution, NNC Special Test Measurements at Quad Cities.

7 , , , , , , , , , , , , , , ,

Key 6 -

EliSS 3-4 in. ~

'//J 2 3 in.

g E 12 in. ,

y4 _ !aila 01 in.

o

$3 E

z 2 -

l 1 -

0 ' ' ' ' -

l 1

2 3 4 5 6 7 8 9 101112131415-Axlalinterval Figure 3-6. Axial Distribution of Gaps, NNC Special Test Measurements at Quad Cities.

3-25 l w s

P I data, first under the hypothesis that the distribution is represented by a single population. It is then hypothesized that the axial ,

distribution of gaps is represented by two populations and the Student t-test applied again. These tests Indicate that there is a higher l level of confidence that the data are represented by two populations than one.

The Chl-squared test was applied to the data to determine whether the data is represented by one population normally distributed. The results of this test Indicate there are Insufficient data to determine whether the distribution could be normal.

The Chl-squared test was again applied under the hypothesis that the data are randomly distributed in axial Intervals 5 through 15 with a mean number of gaps per Interval of 2.8. If the distribution is random, the deviation between the actual number of gaps per axial Interval and the mean should be represented by a normal distribution.

The Chi-squared test Indicates that this distribution could be normal and therefore the axial distribution of gaps random. Unfortunately, there are insufficient data to determine whether there Is a higher confidence in the data being random or bimodal.

TURKEY PolNT NEUTRON RAD 10 ASSAY MEASUREMENTS Using neutron radioassay equipment similar to that used at Qcad Cities for the standard tests, 18 storage cells in the Turkey Point Unit 3 pool were surveyed .L221. Of the 18 cells, 8 were in the Region I rack modules (see Figure 2-1) and 10 were in the Region 2 modules (see Figure 2-2). A. total of 32 full length panels of Boraflex In the Region I cells and 22 panels in the Region 2 cells were tested.

Both the Region 1 and 2 storage racks at Turkey Point are of the

" wrapper" design described in Section 2. A mastic was used to affix the Boraflex to the -rapper plates prior to storage cell assembly.

The details of ho. ., mastic was applied are not available.

t 3-26

During the measurements, gamma radiation background levels from pool contaminants was relatively high thereby llatting somewhat the sensitivity of the measurements. A benchmark cell with gaps of knore size was not used to callbrate the equipment as had been done durie,';

the Quad Cities special tests. It has been estimated 12.Q.1 that the minimum gap size detectable with the equipment used at Turkey point was about 1.0 to 1.5 in..

The storage cells selected for radloessay were those which had received the highest cumulative gamma exposure from the discharged fuel assemblies. The gamma dose has been estimated to be 7.8 x 10' r a d s 12.Q.l. No detectable gaps were observed in any of the 18 cells surveyed. At this exposure level a maximum cumulative gap size of approximately 6 in. would be expected based on data developed in Section 5 of this report had gap formation occurred. Although the details of how the adhesive was applied are not available, differences in manufacturing methods and potentially the manner in which the mastic was applied may explain why gaps occur in the Quad Cities racks and not in the Turkey Point racks as discussed subsequently.

l I

\

O 3-27

O l

Section 4 REVIEW AND EVALUATj0N OF BISCO TEST DATA EARLY BISCO QUALIFICATION TESTS As part of a larger program to qualify Boreflex for use in spent fuel storage racks, BISCO sponsored a series of irradiation tests (1977 --

1980 time frame) at the Ford Reactor at the University of Michigan at Ann Arbor 1.221. The purpose of these tests was to demonstrate the radiation stability of Boraflex and its suitability for long term service in spent fuel pools.

The tests were conducted in a reactor facility where both neutrons and gamma ra'lation d are present to accelerate the accumulation of effective gamme exposure. Accordingly, it must be noted that dif f erences. In Irradiation environment exist between the test experiments and spent fuel pool environment. In addition, there are probably differences In the 1emma spectrum in the test reactor and in the spent fuel pool environment. The results are reported in terms of rads gamma dose but the equivalent dose (when fast neutron and recoli particle damage from neutron alpha reactions in BC4 are considered) is probably much larger and the radiation effects may differ.

In terms of total energy deposited, the test Irradiation conditions are estimated to result in a dose approximately an order of magnitude greater than that projected for gamma exposure only over the service life of spent fuel racks.

In these tests, small approximately 6 in.

samples of Boraflex (of both 25 and 40 w/o B C)

In length, .25 4 In. In width and .100 In. In thickness were Irradlated in air, distilled water and borated water (2000 ppm) to exposures in the range of 1.6 x 10 10 gamma. to 1.03 x 10lI rads The samples were characterized both pre and post Irradiation by physical dimensions, sample weight, specific gravity, hardness and tensile strength (data not reported for all samples Irradletion). During Irradiation, Boraflex samples in sach ofafter the three environments were monitcred for gas evolution in terms of total l

4-1 Iw_ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

volume, rate and composition. In addition, one sample was irradiated 7~ to a low dose in the spent fuel storage area of the Ford Nuclear Q data on irradiation of Boraflex have been Reactor. Additional reported in the literature (23.241 and have been included in this review. In a separate test, samples of unirradiated Boraflex were l subjected to long term exposure in hot water. These test results have been summarized and evaluated.

Olmentional Chanpas The measurements of physical dimensions have demonstrated in most cases a not shrinkage of the samples after I r r a di at i on . The data are variable but the general trend is about 2-35 shrinkage in width and up to 65 shrinkage in thickness. The accuracy of these measurements is not known but it is suspected that accurate dimensional measurements on small samples would in general be difficult and especially so in the case of measuring the thickness of very thin samples. This is particularly true in the pre-irradiation state when the material still has the properties of an elastomer.

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 1221, 1211 and 1211 have been svaluated. Table 4-1 contains a summary of all published data related to welght and specific gravity changes. The specific gravity measurements were made by weighing the sample dry and immersed in water. The true sample volume is then the sample welght times the reciprocal of the specific gravity. This change in sample volume then reflects geometrical changes in length, width and thickness as well as any open porosity which may have developed. The data contained in Table 4-1 have been plotted in Figure 4-1. Review of Figure 4-1 Indicates the followings 10 e There are no data between 2.8 x 10 8 and 1.5 x 10 reds.

e it appears that initially all the samples underwent a reduction in true volume. From the data It would appear that at an exposure in the range of 1 to 2 4-2

l Table 4-1 FRACTIONAL CHANGE IN COUPON WElGHT, SPECIFIC GRAVITY, TRUE VOLUME AND SPECIFIC VOLUME VERSUS EXPOSURE i Fractional Change irr. w$ Data Exposure Specific True Specific Medla BC 4 Source (rads) Weight Gravity Volumo Volume Air 40 1221 2.81E+08 0.002 0.009 -0.007 -0.009 1211 1.50E+10 0.072 -0.067 25 1121 1.60E+10 0.048 25

-0.045 1221 1.03E+11 -0.024 0.122 -0.130 -0.109 40 1221 1.60E+10 -0.025 0.155 -0.156 40 -0.135 1221 2.49E+10 0.023 0.219 -0.161 -0.180 40 1221 1.03E+11 -0.012 0.077 -0.083 -0.071 j Water + 1211 5.00E+10 0.189 -0.159 1211 1.50E+10 0.122 -0.109 25  !

1221 1.60E+10 -0.005 0.005 1 25 1221 1.03E+11 0.174 0.157 40 0.015 -0.135 1221 1.60E+10 0 082

. 0.229 -0.119 -0.186 40 1221 2.49E+10 0.017 0.218 -0.165

-0.179 40 1121 1.03E+11 -0.024 0.014 l

-0.038 -0.014 )

Borated 1211 1.50E+10 0.106 -0.095 Water 25 1221 1.60E+10 0.135 25 -0.119 1221 1.03E+11 0.167 0.132 0.031 -0.117 40 1221 1.60E+10 0.215 40 -0.177 1221 2.49E+10 0.019 0.229 -0.171 -0.186 40 1221 1.03E+11 0.154 0.128 0.023 -0.113

+ Boron-10 loading of 0.02 g/cm2 l

9 4-3

I i

O 0 g

0.0

& S o E

• 6 o E

O o 2 0 O .

$ - 0.10 -

s  % 9

• O 8 O Specific Volume e o l e True Volume g

-0.20 e i i l

l 10s 108 1 010- 10" 1012 Gamma Exposure, Rads Figure 4-1. Het Change in Specific Volume versus Gamma Exposure. (Soures: I r r a d i at i on 12AILE ni en r a + 1 er ua u t ro n Shlaid1na MatarIafa, 748-10-1 Rev. 1, Brand Industrial Services, Inc., Park Ridge, Illinois, August 1981.)

l O 20 -\ l I I I I -

1 i

\

? _

l g 16 -

1

[ Irradiated in Borated Water

\

f 12 -

'q,,'-

3 s, j 8 -

Irradiated g is in D.I. Water .'.,N 4., , ,

y, a -

'**~ a .

irradiated in Air l l 8 I I 40 80 120 160 200 240 Irradiation Time (Hours)

Figure 4-2. Gas Evolution from Boreflex During Radiation Exposure. (Source: f r e a d f a + 1 o n 11xsi,x ni E n r a f f a r N a u t e n n Khfaldina Materfala, 748-10-1 Rev. 1, Brand Industrial Services, Inc., Park Ridge, Illinois, August 1981.)

4-4 a6 I

\

x 10 10 rads (gamma), the volume reduction ceases and the samples begin to swell. A maximum fractional l change In ggample volume change of .171 is observed at 2.49 x 10 rads. '.

e The extent of apparent swelling tends to depend on whether the samples were Irradiated in air o in water (either distilled or borated). At 1.03 x 10{g rads, l the samples which were Irradiated in an aqueous envi-ronmeht exhibit the greatest extent of swelling.-

Gnc Funfuttnn Da+a During the irreolation of some of the samples, offgas produced when Boraflex is irradiated was collected and analyzed. The offgas consisted primarily of H2 with some N,2 0 2 and lesser amounts of CO l CO2 and hydrocarbons as shown in Table 4-2. The source of N 2 is not clears however, potential sources include air entralned in the samples during manufacture or leakage in the sampling IInes. The other off-gas products are among those expected for this type of material although their relative compositions differ from measurements by other investigators as discussed subsequently. For the samples Irradiated 7

in air at 7 x 10 rads / hour, the gas evolution rate diminished to zero l

after approximately 150 hours0.00174 days <br />0.0417 hours <br />2.480159e-4 weeks <br />5.7075e-5 months <br /> (accumulated dose of 1.0 x 10 rads) as shown in Figure 4-2 The samples irradiated in distilled and borated water showed continued gas evolution which is believed to be largely due tc radiolysis of the water in the reactor environment. I Machanteal Penneettan Pre and post irradiation measurements of various mechanical properties of Boraflex have been reported in a number of References (22--251 These measurements include tensile strength, elastic modulus and Shore A hardness for both the unfilled polymers and Boraficx with various  ;

compositions of BC 4 filler. The results as Indicated by these test data can be summarized as follows.

e Tenslie strength: The data show considerable variabill-ty with some samples showing increases and others show-Ing decreases. This verlability may be a result of sem-pie configuration and set during Irradiation. In one i

4-5 Ol C

d

---

x -

Table 4-2

() GAS EVOLUTION FROM BORAFLEX DURING REACTOR 1RRADIATl0N Gas Evolution i Constituent Percentage (5) l irradia- Average Gas CO CH 4 CH tion En- Generation Rate H 2

02 N 2

CO 2 26 vironment (ml/g) (ml/g-h) -

55 13 25 0 't 5 Dry 74 0.4 1 57 10 29 3 0.3 0.2 0.1 Domin. H 2O 206 0.9 (water (64) (10) (26) (0) (0) (0.5) (0.1) control)+ l 7 3 1 4 Bor. H 2 O 236 1.1 79 5 1 (water (2) (0) (0) (0.4) (0.5) control)+ (57) (40) i I

+1rradiation capsule filled with only water and no Boraflex.

Burn and G. Blessing.

• Radiation Effects on Neutron

~

Sour es R. B.

Cp -

Shielding Meterials", in Transac+1ons ni ibA Amartean Hus)ang sectatv, vol. 33, suppl. 1, 1979, pp.48-49.

l l

)

O 4-6 i

i 1

~_

report 1211, the tensile strength increased up to 7 x 108 rads (pure gamma radiation).

polymer in this report is unfilled It is Boraflex.

believed the NS-1 e Elastic modulus: Only one report 1211 was identified which of gamma presented dose. As data for elastic modulus as a function shown in Table 4-3, the elastic mode-lus gamma dose.

of the NS-1 polymer increases dramatically with e Elongation-to-break Table 4-3 contains the only re-ported measurements of elongation to break measure-monts as a function of gamma dose. It is noted that elongation to break decreases rapidly with dose as the material hardens and embrittles.

e Shore A hardness: There are several reported measure-monts of Shore A hardness as a function of gamma dose.

In one report 1221 an increase in Shore A hardness frgm 78 to 89 was observed after Irradiation to 2.8 x 10 rads in a pure gamma field. l.n the reactor tests, the hardness data reported were mostly in excess of 90 BShore A with variations depending on the composition of 4C. Shore A measurements of greater than 90 are not particularly meaningful 1121 and a Shore D durometer should be used.

The use of a Shore D durometer on irra-dlated Boraflex may however result in fracture of the sample.

Neutron Attenuation Neutron attenuation measurements have been reported for pro and post Irradlated samples.

The purpose of these measurements is to verify the presence of B-10 after Irradiation (22-251. Most samples showed very small changes In neutron attenuating properties pre and post Irradiation relative to the inherent statistical uncertainties.

Lone Term Frnosura +n Hnt Water Long term out-of-pile testing of unfrradiated Boraflex at elevated temperatures and In aqueous solutions has been reported 1261. The '

samples of Boroflex 0 were immersed in borated water (approximately 3000 ppm) at 240 F for 6200 hours0.0718 days <br />1.722 hours <br />0.0103 weeks <br />0.00236 months <br />.

with NaOH to a ph of 9.0--9.5. After The borated water was neutralized I testing the sample showed an overage dimensional decrease along the side of the sample of 0.935 and on increase in welght of 0.245. Gas evolution from the sample was 4-7 k

4

Table 4-5 EFFECT OF GAMMA RADIATION ON BISCO NS-1 POLYMER Dose Tensile Strength Elongation Elastic Modulus i (PSI 1 Menareds ,fPSI) _,

0 510 68 750 16 516 55 938 60 550 40 1375 III 504 38 1326 164 553 23 2404 713 896 3.3 27151 Samples .1 in. x 1 In. Ten'slie Bar pulled 8 10 inches / minute.

l Yource: Boraflav SuftabfIttv Renort. Ravfafon 1 Park Ridge, Ill.:

Brand industrial Services, May, 1978, 1047-1.

O l

l l

l I

O 48 m .

measured with a total of 5.22 cubic Inches of gas at STP being generated per square Inch of sample surface over the entire test period.

Gas evolution diminished with time with 41.3% of all gas evolved during the first quarter of testing and only 8.7% evolved in the last quarter of the test period. The evolved gas was analyzed periodically for compositten and showed approximately 205 H, 535 CH3 ,

55 CO 2 with the balance comprised of various hydrocarbons.

Whether gas evolution impiles thermal decomposition at these relatively low temperatures is not clear. Typically, the polysiloxane are known for service at elevated temperatures 12,21 The observed gain in sample weight after testing suggests that the samples may be absorbing borated water. In any case, the test is not particularly meaningf ul since the material tested was unfrradiated. As discussed subsequently, the effect of hot water immersion on irradiated Boreflex is probably different than on the unirradlated material.

RECENT BlSCO RADIATION TESTING I

As a consequence of the gap phenomenon observed at Quad Cities, BISCO 1

initiated a new series of test Irradiations in the spring of 1987, the 1 objective of which was to quantify the maximum extent of Boraflex shrinkage 12Al.

Two series of tests were conducted including the Irradiation of coupons in a pure gamma spectrum (Co-60) as well as test reactor Irradiations in the University of Michigan Ford Nuclear Reactor. In the first test, a serlos of ten relatively large (2.5 In.

x 12 In.) coupons were placed in a Co-60 facility and periodically removed for Inspection and characterization. The samples were then returned to the Co-60 facility for subsequent Irradiation until they 1

had received a total Integrated gamma dose of 1 x 10 9 rads. The ten j

samples were procharacterfred with respect to dimensions and weight and changes in these attributes were determined after each irradiation interval. After the final Irradiation period, Shore A and D hardness were determined.

in the reactor tests, twelve groups of nine small (1.6 In. x 1.6 in.)

coupons each were Irradleted.

At scheduled irradiation Intervals, one j group of nine samples was removed and characterized. After i

{

49 l

the samples were not reinserted for further character 12ation Irradiation but rather another sample group was removed after the next Irradiation period.

Each sample was procharacterized with respect to dimensions, welght, Shore A and D hardness and neutron attenuation and these attributes were determined after Irradiation as well. The last 1.12 x group of nine samples received a cumulative gamma exposure of 10 II rads, plus energy deposited by neutrons and recoli particles.

The total dose has been estimated to be approximately an order of magnitude greater than the gamma dose alone.

The results of the reactor tests were not particularly conclusive with respect to dimension changes due in part to the small sample size and Accordingly, BISCO in part to erosion from the edges of some samples.

has initiated the Irradiation of larger samples in a test reactor at

)

the University of Missourt.

nimenstanal channes Only the changes in sample length and width are considered here.

Measurements of sample thickness are considered meaningless owing to the large verlebility of the data. This is attributed to the The percent change measurement of small changes in very thin samples.

In sample length and width from both series of tests are plotted in l

l Figure 4-3. The open symbols in this figure represent changes in The coupon length and the closed symbols changes in coupon width.

symbols with error bands are data points from the reactor tests and represent +/- one sigma. The data points without error bands are from the Co-60 Irradiations. Owing to the large size of these samples, the

.05% in length standard deviation is much less (typically less than +/

and less than +/- .25 in sample width). At each exposure level the Indicated change is the a'verage of all nine or ten samples in the 1

group.

i

}

Several features of the data presented in Figure 4-3 are worthy of  ;

note:

O 4 10 l

1

-- -- ___J

,. .= =-mm l

1 1

i l I i 1 5 e e 2 0.0$, u e

3 'j> 0 B

r -1.0 -

0 e W -

73 5 O O

a O e - 2.0 -

U 0

c.

e (s, 5 - 3.0 -

Co80 Data.

o h O Length E e Width l"

' ~ ~ ~ ~ 7)]

O

- 4.0 -

I!

i j' E Reactor Data:  ; O c) l Length O

$ - 5.0 -

a. l 0 Width l'

l l l L. .. . . . .

107 10s 108 1010 10n Gamma Exposure, Rads Figure 4-3. Percent Change in Sample Length and Width versus Gamma Exposure.

Boratiew Neutron Abancher (Sources leradIatinn Interim .T.ast M nf.

ania, NS-1-050 Interim,November Illinois, Brand Industrial Services, Inc. Park Ridge, 25, 1987.)

i O

4-11 i

a ,

X-~~ - + - - - + a __

~s e Those samples irradlated in the Co-60 facility exhibit t about 505 greater shrinkage in sample length than In sample width. Discussions with BISCO personnel 1221 Indicate that the samples were cut from sheet material with the length being in the forming direction during the calendaring-like process used to manufacture Boraflex.

e The Co-60 date has been fitted to a linear relation of the form dL/L = (-1.4156 x 10~9)R - 0.0781 as shown inFigurej-4. Using all data points except the first at 1 x 10- reds, provides the best f!t with a correlation coefficient 99.925. The first data point i l

has been omitted owing to "ond effects". This linear I relationship at low doses has been observed in PDMS by others as discussed in Appendix B.

e The reggtor data contained in the box In Figure 4-3 at 5 x 10 rads and beyond may overstate the actual sample shrinkage owing to observed 12h1 rounding and erosion of the edges of the samples. It has been postulated 1211 that alpha particle damage may be responsible for this phenomenon. Water permeation and chemical effects may f

A- .

also be a factor.

l e The larger verlebility in sample length and width changes as well as loss in sample weight discussed sub-sequently at higher dosas tend to support this hypothe-sis.

e Olscussions with BISCO personnel 1221 have Indicated that data are not available which make it possible to relate either the length or width direction of the re-actor samples to a particular manufacturing process verleble.

l Figure 4-5 contains the average change in sample thickness for each series as a function of gamma dose. The large variability in the measurements make It difficult to Inter absolute not changes in thickness. The verlebility is likely due to trying to measure small dimensional changes in very thin samples.

Ov 4-12 1

___ w

l O

o

,.\

o 4 >

• - 0.5 -

sa. 1 1

E s i E

a 0

15 e -1.0 -

2 i

-1.5 -

o -

l . . . . . . . . l .

10s 108 Gamma Exposure, Rads Figure 4-4. Linear Fit of Percent Change in Sample Length versus Gamme Exposure. (Co-60 Irradiations Only).

1 O

l 4-13 1

O g i 3.0 - g 2.0 E O E 0 -

t 1.0 0 o 3 o 0.0

f. O .i P

$_ 9 0 o - -

g - 1.0 4) e O O. I '

O ,

c - 2.0 8

[ - 3.0 hcoSO Data f Reactor Data i i I l

108 10'0 10" 107 10s-Gamma Exposure, Rads Figure 4-5. Percent(Sources Change feradta+1on in Sample Thickness study af. versus Bora-Gamma Exposure. intarIm ,13s1 M , NS-1-050, hRev.

Neutron Abanchac 1 interim, Brand industrial Services, Inc. Park Ridge, Illinols, November 25, 1987.)

l O 4-14 a L

I s._,.v.,_.,..

e The function average of weight change in each group of samples is plotted as a Camme exposures In Figure 4-6 the Co-60 samples which were Irradiated in The open symbols are for small air. These samples show a not gain In weight up to an o.xposure of 1 x 10' rads. Based on gas evolutlots associated with gamma Induced crosslinking, one might expect are a not decrease evolved. Other in sample weight as hydrogen and hydrocarbons free radicals are investigators 111), however, have noted that as formed in PDMS during Irradiation, they may j recombine Appendix B. with constituents in air or with water vapor as discussed n i

Irradiation continues. This may explain the not increase in sample weight as A similar trend is noted in the reactor samples up to 2.5 x 10 10 reds.

These samples were Irradiated in an aqueous environment although the chemistry of the aqueous solution is not known. Beyond 5 x 1010 rads, the trend seems to reverse and at the higher doses th not decrease in sample weight. e samples, show a As this occurs, the variability of the data increases and it has been pos t u l ate d 1,2.9.1 th a t this is due to Boraflex which has been eroded from the edges of the samples. The Initial increase in sample weight is believed to be due to water absorption permeating inward. inftlally beginning at the edges of the samples and i

Shor. A and n Hardn.nz The the Co-60 Shore samplesA and 0 hardness was measured on all samples and reactor on after a cumulative exposure of 1 x 10 9 data are plotted in Figure 4-7 rads. The The Shore A hardness saturates rapidly with a value approaching full scale (100) at 1 x 10' reds.

The Shore D hardness tends to saturate beyond rads. However, exposures of I x 10 10 due to the brittle nature of Boraflex at these exposures, the samples tended to break when subjected to Shore D durometer measurements.

S 4 15

-~...;. . --

O 1

I 5.0 , ,  ;  ;

o l l

2.5 - k -

o

\

1 E o l

~

O O e 0 .0g- O g _

. i a l E -

)

u) e - 2.5 -

oi >

5 E - 5.0 - . )>- I ci l 2

E e - 7.5 -

< e E O Coso Data

- 10.0 -

O Reactor Data l 1 I I i 107 10e 108 1010 10" Gamma Exposure, Rads Figure 4-6. Percent Change in Sample Weight versus l

Gamma Exposure. (Sources leradfa+1on M al Bora-l 1.La.1 Neutron Abunebar tri+arta last D.a1A, NS-1-050, ,

Rev..I interim, Brand Industrial Services, Inc. Park Ridge, Illinois, November 25, 1987.)

O 4-16 m

\ O 100 g

t g

"} 11 00 o

a E  :

$ 8 90 -

h -) -

60 $

$ __ I

< a E

o ..

} $

5 z 80 :-

/1 20 l l l 10e 107 jos i

10s 1010 Gamma Exposure, Rads Figure 4-7. Shore A and Shore D Sample Hardness ver-sus Gamme Exposure. (Sources f e r a rf f = + 1 o n study af.

BermfIav Neutron Absorbar _fn+arim I.ggi agia, NS-1-050, Rev. 1 interim, Brand Industrial Services, Inc. Park Ridge, Illinols, November 25, 1987.)

O 4-17

, m

_________m__-- --

l ,

Wantenn AttanumtInn O Pro and post Irradiation neutron attenuation measurements have been q

reported for the specimens l irradiated in the test reactor. The average pre Irradiation attenuation for all samples was 0.979 +/ .002 and 0.0981 +/- .002 after Irradiation, thus indicating no change in neutron absorption within the limits of detectability of this type of '

measuring technique.

VfnumI A n n m a r a rn e n  !

At cumulative gamma exposures beyond 10 9 rads, Boraflex is reported M to have the appearance and feel of a ceramic material. The material is relatively strong in compression but subject to brittle fracture in tension. When frsctured, the material breaks into large pieces. At doses of I x 10 10 and beyond, a fine whitish-grey powder appeared on the surface of many samples. The occurrence of whitish grey powder was most prevalent on the edges of the samples extending inward for about 1/4 inch. The powder could be wiped off easily and thereby exposed Boraflex of. jet black color underneath.

Semelfte Grav1+v Mmmeuraman+m I )

l Subsequent to the issuance of Reference M, the measurement of samplo density or specific gravity has been reported M for some of the samples Irradiated in the test reactor. These data are shown in Figure 4-8.

The data Indicate a maximum change in specific gravity of about 20% at a gamma exposure of 1 x 10 II reds. Also plotted in Figure 4-8 Is the corresponding change in sample specific volume as a function of gamma dose. Chbnges in true sample volume were not l

computed since the pre and post Irradiation sample weights were not available.

The data in Figure 4-8 show considerably less variability than the corresponding earlier data (Figure 4-1) presumedly due to t

more carefully controlled experimental conditions. Nevertheless, the conclusions seem to be the samel that is, the change in specific volume seems to reach a maximum asymptotic value beyond 1.0 x 10 10 rads of about -175. Contrary to the results from the earf lor data, there II seems to be no Indication of sample swelling at exposures of 1 x 10 reds.

4 18

~ .

l l

l e

2.2  ;  ; , ,

e 0- [8 -

0.0 N O 3 2.0 -

- 2 8

• l B k )

A m 5 5 e

2 h - - 0.10 e e 4 A 2 F

g

-o e 9 Y O r$ 1.8 l

0%o j tt

- - 0.20 l

I I I I 1.6 10s 3os 1 010 10" Gamma Exposure, Rads Figure 4-8. Specific Gravity and Specific Volume of Boreflex versus Gemme Exposure. (Sources Reference 30.)

l e 4-19

~. __- a

3-- -- -

l I

j Dfacuttlnn nf Intacta Rannr+ rnnelue?nnt '

O Preliminary results of the recent 88500 Irraclation tests were On the reported in a BISCO interim Report in November of 1987 12A1.

basis of the test Irradiations, the Interim Report 1.211 cites some specific conclusions with respect to the behavior of Boraflex in a redletion environment. Certain of those conclusions described in the Interim Report differ somewhat from those in this EPRI study. These include, 1) the definition of those factors that contribute to the my life, 2) the extent of maximum i determination of total service shrinkage that may occur from Irradiation, and 3) the effect of l

Irradiation on the physical and chemical composition of Boraflex. l l

l l

Relative to the first point, it is acknowledged that the total j integrated radiation dose (gamma, alpha, and neutron) to which the reactor samples were subjected is likely to be well in excess of that expected during the 40 year design lifetime of spent fuel storage racks; however, the duration of the tests was only five to six months.

j The effects of long term exposure of Irradiated Boreflex to an aqueous environment cannot be assessed f rom the available data and theref ore ongoing survelliance should be performed to verify performance. Also, because the water chemistry of the test reactor was not reported, it i is not known whether it is representative of that normally found in a l spent fuel pool. In addition, the details of sample configuration, f other than that a " sandwich arrangement

• of nine samples, have not been presented and it is therefore uncertain whether all sides, or just the edges of the Boraflex samples, were exposed to the aqueous environment during the test Irradiations. Additionally, it is probably not practical to quantitatively proje'bt panel width allowance for edge oro.ilon effects from the data presented. l To the second point, the Interim Report conclud6s that the maximum shrinkage of Boraflex is between 2 to 2 1/25 due to radiation effects.

O 4-20

__ nw

Data presented in Section 3 of this report and discussed in Section 5 I

Indicates that the maximum linear shrinkage could be niore in the range of 3 to 45.

Differorces in the radiation environment and spectrum In-reactor versus the spent fuel pool may result in different radiation damage mecharisms and hence the difforence in shrinkage.

Relative to the third point, the Interim Report concludes that 8

radiation transforms Boraflex from a filled polymer to a " stable ceramic".

The Interim Report contains insufficient data to support this conclusion.

O 4-21 L

e O.

Section 5 DISCUS $10N i

The previous sections of this report have presented data from the utility surveillance programs and BISCO test Irradiations.

Additionally, included as Appendix A is a summary of the physical properties and characteristics of Boraflex and other polysiloxane polymers derived from the literature. Appendix B contains a review of the relevant literature on the effects of radiation on polysiloxane  !

polymers. In this section, the data from these various sources are combined and evaluated with the intent of quantifying the combined effects of gamma irradiation and long term exposure to the pool water on Boraflex. Potential factors influencing gap formation and subsequent growth are also discussed.

RADIATION DAMAGE MECHANISMS AND G-VALUES When Boraflex Is exposed to gamma radiation, the material undergoes I changes. These changes have been evident in the BISCO test Irradiations as well as experiments with materials similar to l Boraflex. The radiation Induced changes includes e Evolution of gases including hydrogen, methane and other hydrocarbons e Physical shrinkage, accompanied by an increase in specl-fic gravity and decrease In specific volume e Loss of rubberelasticity, increase in hardness and an increase in elastic modulus e Observed formation of blisters in some specimens be-lleved to be the result of gases which are formed and subsequently trapped in the polymer matrix when ex-posed to radiation l e Change in surface color from a shiny black to a whitish l gray and a tendency of the discolored material to l powder 5-1

- . - - - . . ~ . .. - -

p. e Permention of water believed to be the result of the development of open porosity in addition, there is evidence that Indicates environmental factors water and other agents may such as the presence of molecular oxygen, under this materlat be factors which influence the behavior of Irradiation.

I The radiation Induced changes outlined above are consistent, with the scisslocing as of crosslinking and radiation damage mechanisms low radiation d o s e's , the evidence described in Appendix B. At suggests that the predominant radiation changeismechanism in PDMS is the abstraction A consequence of crosslinking crosslinking.

and other radicals and the subsequent formation (relesse) of H', CH3

• Physically, the of crosslink bonds between adjacent polymeric chains.

crosslinks cause the adjacent chains to be pulled closer together thereby resulting in volume shrinkage and increase in the specific gravity of the material.

D

'Vh As the radiation dose is increased, the number of potential sites for f ormation of crosslinks is decreased and the transnational motion of the polymeric chains become restricted duecrosslinking to a relatively large tends to Accordingly, number of inter-cheln ties. increase in density) saturate and the physical changes (shrinkage and At associated with this damage mechanism approach asymptotic values.

this point the free radicals may continue to be generated without significant f ormation of crosslink bonds and scissioning is likely to Scissioning is the become the predominant r adiation damage mechanism.

severing of bonds in the spine of the polymer which results in a Evidence has reduction in molecular welght and eventual degradation.

been cited in Appendix B which suggests that the presence of oxygen, water and other agents may favor scissioning over crosslinking.

Gas evolution measurements from samples irradiated in air (see Figure I

4-2) Indicate that measurable gas evolution ceases s'/ or before a I x 10 10 reds, thereby suggesting gamma radiation exposure of 52 tw

i I

l crosslinking has saturated at or before this exposure. Since these experiments were conducted in a test reactor and not a pure gamma radiation field, it is useful to estimate the theoretical absorbed gamma dose required before crosslinking saturates. In Figure B-3, the G-value for crosslinking in PDMS is presented as a function of gamma Irradiation dose. At low doses, the G-value is about 2.8 and this diminishes to about .65 at a dose of 500 Mrads. The average G-value has been estimcted from data the of Delfdes and Shepherd by extrapolation and integration of G(E) and is in the range of .12 to

.20 to doses in excess of 1 x 1010 rads, in a previous report D.1, the total absorbed gamma dose required to crosslink all monomer units in the polymer has been estimated. It has been noted, however, that once crosslinks between chains have been 3

established that restricted transnational motion of the chains tends to preclude further crosslinking. For PDMS, the data from Delldes and Shepherd D.1.1 Indicate that, at saturation, at most only about 14% of the monomer units crosslink.

Using this saturation crosslink density and a range of G-values, it is .

possible to estimate the dose at which crosslinking in PDMS saturates.

We note that: -

1 Megarad = 6.25 x 10I9 ev/g (5-1) i and that for a given G value, we can write the number of crosslinks/g induced by 1 Megared absorbed dose as:

1 Megarad ----> G x 6.25 x 10II crosslinks/g (5-2) l Since the molecular weight of the PDMS monomer unit is 74: I

{

! )

g 53

m. _. _ -

II 1 Megarad ----> G x 4.62 x 10 crosslinks/ mole (5-3) p 1

23 units, the absorbed dose Since one mole consists of 6.023 x 10 4 required to crosslink all monomer units is (1/G) (1.302 x 10 )

f max the fractional number of monomer units Megarads. If is crosslinked at saturation, then the dose to crosslink all possible units is

/G) (1.302 x 10 )4 Megarads (5-4)

D,,9 = (f Using G-values in the range of .12 to .20 and a value of .14 for fmax*

It is possible to use Equation 5-4 to provide upper and lower estimates of the absorbed dose to crosslink all potential sites. This 9

exerciso provides a dose to crosslink all potential sites of 9 x 10 10 to 1.5 x 10 reds.

While It is acknowledged that the polymer matrix In Boreflex may not be identical to the PDMS studied by Delldes and Shepherd, it is believed that the estimated dose required for crosslinking to saturate would be approximately the same. It is believed that Boraflex in its as-produced condition has some degree of initial crosslinking. The degree of Initial crosslinking is probably Induced chemically during the vulcanization process and is Indicated by the initial values of Shore A hardness and elastic modulus of the material. Initial crosslinking would tend to reduce f max, the maximum fractional crosslinking density which can be Induced by subsequent Irradiation.

l It has been noted that the Boraflex polymer probably contains radiation stabilizers in the form of free carbon and $102. These stabilizers are believed to prevent crosslinking after abstraction of either H* or CH 3 from the polymer.

\ G( %

5-4

Referring to Equation 5-4, it is likely that additives and modiffers which may be present in Boraflex relative to PDMS have the effect of reducing f,,, as well as the G-value.

Accordingly, the not effect may be approximately the same estimated dose to crosslink all possitie sites as for PDMS. A saturation dose for crosslinking of about 1 x 10 10 rads gamma seems consistent with the gas evolution measurements from the early BISCO test program as well as with specific gravity measurements from the latest BISCO test program.

Chenees in Meacific Gravl+v and Snactfte Volume Vertue Dona The specific gravity measurements from the latest BISCO tests provide a means to estimate the change in specific volume as a f unction of absorbed gamma dose. This provides not only an estimate of the maximum change in sample specific volume but also the functional dependence of the change In specific volume with dose.

Bopp and Sismen 1121 have noted that in crosslinking polymers the decrease in specific volume may be fitted to the equations dV(R) =

dV,,, (1 - o-cR) (5-5) where:

R = gamma dose, Rads dV(R) = fractional change in specific volume dV,,9 = asymptotic fractional change in specific volume c = constant, Rads ~I The latest BISCO measurements of specific volume (reciprocol of specific gravity) versus gamma dose have been fitted to Equation 5-5 with c = 1.735 x 10-10 R~I and dV,,9 = .17. This resulting curve is compared to the data points In Figure 5-1. This comparison 5-5

O I I I

0.0.

e a

E e

5 -

& - 0.1 - $

E A 0 W

$ \ 8- __.

8 m

I I I

- 0.2 10" 108 10' 1 010 Gamma Exposure, Rads Figure 5-1. Specific Volume Change of Boraflex versus r Gamme Exposure from the Latest Bisco Reactor Irradiation Tests.

O 56 Mw

l i

i i

Indicates that the latest DISCO data are well represented by Equation 5-5. In considering the early BISCO data (Figure 4-1), while it does exhibit considerable variability it tends to confirm the maximum l magnitude of speelfic volume change as provided by the latest test l data.

I It should be noted that this change in specific volume shown in Figure l 5-1 reflects changes in the "true" sample volume since the specific gravity measurements on which they are based are via immersion f

techniques. These volume changes represent changes in envelope dimensions (geometric) or shrinkage of the samples as well as any l l

porosity open to. water permeetion which may have developed.

I 1

Channac in Rnenflav Otmanatana Varnum Onta Changes in Boraflex dimensions versus gamma exposure can be inferred l from a number of data sources. These data sources include the latest BISCO Co-60 and test reactor Irradiations, some data from the utility '

)

coupon programs (where the dose is known or has been estimated) as well as from the Quad Cities neutron radioassey measurements of gaps.

A composite of all data relative to length or width changes in  ;

Boraflex collected as part of this study is presented in Figure 5-2. l Perhaps the most precise low doso (up to 10' rads) data are from the BISCO Co-60 Irradiations. These data Indicate an average shrinkage i (of ten specimens) in coupon length of 1.5065 with a relatively Icw deviation of +/- 0.0435 at the one sigma level. At the same dose the average shrinkage in width is .9575 +/ .1915. Changes in sample thickness are dif ficult to ascertain owing to the very thin Boraflex specimens. The average measured change in thickness was .565 with a relatively large deviation of +/- 1.325. While the method for determining specimen thickness is not reported, it is noted that a 15 change in thickness of a sample w'eth Initial thickness of 0.075 in. Is only 0.00075 In.. This change is probably less than or equivalent to the precision of the measuring technique.

O 5-7

O ,

I I I I

~~

E 0.0l """thfY% 9 *

< i 2 l' is ~~

g - 1.0 g

5 0C g , O II

.i* - 2.0 --

c qp C'---

P int Bea h Coupons \ j g-0 6

Point Beach Panel 1 ,

l g

- 4.0 - Prairie Island 1 yr. Coupon !_ _ _ _

4 in. Quad Cities Gap 46

- 5.0 -

i I I I l-----!J 107 108 10' 1010 10" Gamma Exposure, Rads Co60 Irradiations: Reactor Tests:

e width 4 > width O length length Figure 5-2. Change in Boreflex Length and/or Width versus Gamme Exposure.

l l

l l

O 5-8

'-M8" 9apme @ e-@=W***-

, ,,,gg , m $

-- ------ - _ _ _ __ _ _ _ _ _ _ _ _ _ _ _ J

in any case, the length and width measurements do provide quantitative data and suggest that the shrinkage may be entsotropic. Discussion with BISCO personnel 1221 Indicates that the length (long) direction of the samples is parallel to the forming direction applied during the manufacture of Boraflex which may account for the variation in shrinkage rate. These data seem to suggest that the magnitude of shrinkage is SOS greater in length than in width.

The data from the BISCO reactor Irradiation program are also plotted in Figure 5-2 and show a greater variability due to the small size of the samples as well as edge erosion effects as discussed in Section 4 The data enclosed in the box represent those coupons where edge erosion may have affected the determination of dimension changes.

Because of the large verlability of the data, it is not possible to confirm the anisotropic nature of shrinkage as Indicated by the Co-60 data. Also, discussion with BISCO personnel 1221 Indicates that data are not available to relate sample length and/or width to any particular manufacturing parameter. Excluding the data where sample edge effects may influence the measurements, the reactor data suggest  ;

g a maximum shrinkage In length or width of less than 35 at a dose of up to 1.12 x 10 lI reds. Changes in sample thickness cannot be assessed due to the large varlebility of the measurements.

In addition to the BISCO data, length and width changes from the one year Pratrie Island coupon and the one year Point Beach coupons are plotted. It Is noted that data for only two of the three Point Beach coupons is plotted since the third was broken during handling. The Point Beach data Indicate a maximum shrinkage of -3.105 although the measurement uncertainty is rather large due to the small size of these coupons. The one year Pralrle Island coupon was relatively large (8" x 12") and showed a shrinkage in length of -3.335 and about -15 in wldth. This would again suggest anisotropic shrinkage with length changes being about 3 times greater in this case than width changes.

It is possible that the degree of anisotropy may depend on design parameters such as thickness and filler composition.

O 59 e

v in addition to the coupon data, the shrinkage corresponding to ais4 in, gap as determined by neutron the radloessay measurements at Quad Cities measurement technique was taken also plotted. The resolution of A4 as +/-0.5 in. and the uncertainty in the dose estimate Is +/-50%.

In. gap corresponds to a shrinkage of 2.6% in length.

The data presented in Figure 5-2 provide a means for projecting As was changes in Boraflex dimensions as a function of gamma dose.

previously demonstrated, in Equation 5-5 provides an accurate means of sample specific volume as a function representing observed changes of absorbed gamma dose. It is therefore not unreasonable to assume l f

changes In sample length can be represented by an equation of similar forms (5-6) l dl/l = (df/l,,9) (1-e-CR)  :

l where O = gamma dose, rads R  ;

C

= constant, rads"I 1 dl/l,,9 = asymptotic value of shrinkage The constant C, in this case, is not necessarily the same as in Equation 5-5 since Equation 5-6 represents the change in specific volumein envelope which dimension and Equation 5-5 the change  !

includes porosity effects.

In order to develop estimates of C and df/l,39, the previously to saturate are estimated doses for crosslinking (and hence shrinkage) used.

in addition, the average semple length change at a gamma dose It was of 1 x 10' rads f rom the BISCO Co-60 Irradiations is used.

noted that the gamma dose required for crosslinking to esturate was 10 rads. If we use the estimated to be between 9 x 10' and 1.5 x 10 O 5-10

y l

conditions: 9 1 dl/l = -1.5065 8 1 x 10' rads (5-7a)

(dl/l)/(df/l,,,) = .999 8 9 x 10' rads (5-7b) we obtain the following upper estimate of changes in Boreflex length versus gamma doses dt/l = (-2.8%) (1 - exp (-7.67 x 10-10)R) (5-8)

If the conditionst

}

dl/ I = -1.5065 8 R = 1 x 10' (dl/l)/(di/l,,9) =

.999 8 R = 1.5 x 10 ads rads (5-9a)

(5-9b)

)h l

l are used, than the following lower estimate equation is obtained l l

I dl/l = (-4.15) (1 - exp(-4.61 x 10-10)R) (5-10)

These upper and lower estimate curves are plotted in Figure 5-2 to provide an estimate of the range of length / width changes projected for Boraflex as a function of gamma dose. ,

almost It is noted that the curves fit exactly the Co-60 data and bound within experimental uncertainty the coupon and gap data. The exceptions are the date enclosed in the box in Figure 5-2 where it has been noted that erosion offects may overstate the extent of shrinkage. In summary, the date in Figure 5-2 would Indicates e

Some samples showed more shrinkage in one direction than in the other.

5-11

e The maximum shrinkagT0 in either direction saturates at rods and is in the range of 3-45.

approximately 1 x to Chanean in snactfle V Q g ?arnum Ganmatric Voluna The date from the BISCO speelfic gravity measurements Indicate a maximum change in specif ic volume of about -175. The data based on r geometric measurement of coupon length and width measurements can be used to estimate the change In geometric volume of the coupons for comparison to the specific volume change. In order to estimate the change in geometric volume, the upper and lower estimate curves in Figure 5-2 for length change versus gamma dose have been used. Using the Co-60 data at 1 x 10' reds, en average sample volume change of

~3.0 5 is computed with a maximum of -3.95 and alnimum of -2.15 at the one sigma level. The relatively large range in maximum and minimum volume Is introduced by the large uncertainty in the thickness measurements.

Of the 3.05 average volume $snge, about one half (1.55) is attributed to shrinkage in the length direction of the coupons. This provides a basis for converting length changes into volume changes. Figure 5-3 contains a plot of change in volume versus gamma exposure. The change i in specific volume Indicative of true volume changes Is plotted as well as the upper and lower estimate geometric volume change derived from the corresponding length change curves in Figure 5-2. The lower estimate curve minus 1 sigma and the upper estimate curve plus 1 sigma are also plotted.

The measured geometric volume changes from the BISCO test Irradiations as well as the utility coupon data are also plotted. The geometric volume change corresponding to a 4 In. Quad Cities gap measurement has been computed assuming volume change is equivalent to two times the length change. It is noted that the geometric volume changes are generally bound by the upper estimate curve plus 1 sigma and the lower estimate curve. The maximum geometric volume change as projected by the lower estimate curve is about -8.55 which compares with -17.05 maximum change in true sample volume. The difference between the two O 5-12 h __ ._________.__.___m_ _ _ . _ _ _ _ _ _ _ _ _

~

1 l

l O

l i l i l

0.0 m . A, g ** %

{% %

-- g- g z-6 s,*% .

g \ y Range of 3 Geometric .)

7:i g%g C Volume l

Change

-0.10 -

O% . .-----~ I 8

G Specific 0-4 Volume cnange

-0.2 g i I Da 10s 1010 10" Gamma Exposure, Stads Figure 5-3. Change in Boreflex Geometric and Specific Volume versus Gemme Exposure. ,

I i

9 O

5-13 s

values is Delleved to be open ' porosity which is inheront in tho immersion measurement of volume.

10' rads both the spec 1fic and geometric It is noted that at 1 x ,

volume changes are the same and at higher doses they begin to diverge.

The dif f erence between the true volume change and the geometric value change can provide en estimate of the development of open porosity as a function of gamma dose. Figure 5-4 provides an estimate of the range In open porosity versus gamma dose based on the bounding curves of geometric volume changes in Figure 5-3.

to confirm While no measurements of open porosity have been reported appearance of both irradiated and these estimates, the visual untrradiated Boreflex qualitatively confirms the existence of numberopen porosity, the edge surf ace appearing to contain a significant of volds (see Figure 5-10). The radioassey measurements of the l unirradlated and Irradiated panels examined at Point Beach further ,

suggest a radiation Induced mechanism whereby water permention is f accelerated relative to the untrradiated material.

0 l The development of surface crazing, or microcracks, in the material l

may be related to either chain scissioning Chain or Internal stresses scissioning and l attributed to other radiation mechanisms.

Internal stresses are potential effects which could result in the formation of porosity. Erosion of 84 C and other filler particles from the surface of the Boraflex Is another suchAs mechanism which would discussed subsequently, contribute to the f ormation of porosity.

the cut edge of a g radiated Boraflex contains significant initial porosity as confirmed by scanning election alcroscopy.

Can Fnrentinn and Can Growth An essential factor determining whether Boraflex sheet will develop gaps appears to be the existence of a mechanism for restraint of the sheet. In the Point Beach fuel racks, the sheets of Borsfiex are held O 5-14 I

O I i 12.0 -

~

~

- ,e

• 9.0 -

2 ....

2 6.0 -

c -

O 3.0 - ~

~

l j

0.0 I  !

10s 108 10'O 10" l Gamma Exposure, Rads Figure 5-4 Gamma Exposure. Estimated Range of Boraflex Open Porosity versus 5 15 4

- ~ -

I In the a pair of Y-shaped grooves (see Figure 2-6) in place between sheets were removed for sheathing. When those Stainless steel Presumably, as the Boraflex sheet was examination, they were intact.

subjected to radiation-Induced shrinkage, the frictional restraint provided by the V grooves was not suffIclent to result in local stresses to cause the material to fear.

restraint' provided either through the use of In other rack designs, h adhesives or by mechanical means appears to be sufficient to cause t e formation of tears or gaps. is In the Quad Cities racks it has been held in place with an adhesive (Dow noted that the Boreflex sheet The design drawings for during manufacture. l Silicone Seelant No. 999) these rocks have been reviewed and there appears to be no additiona l I t h a s b e e n p o s t u l a t e d b y B I S CO ,[.10.1, means for mechanical restraint.

however, that distortions in the stainless steel enclosure introduced during the welding process may be sufficient to cause substantial r e st r a i nt .

Since the condition of the Dow Sealant, in particular its adhesion properties as a function of gamma dose are not known, uncertainty in exists with respect to the manner and degree of Boraflex restraint Nevertheless's, it is useful to postulate three the Quad Cities racks.

bounding scenarlos which includes ('

o The adhesive bond provided by the Dow sealant complete-  ;

ly breaks down at low doses of gamma exposure.

is e The bond between the Boraflex and stainless steelis completely unifo

" perfect", thatcol properties of the Boraflex are uniform along its onttre length.

e The (i.e., bond reglonisexpected intact at the ends of to receive thethe Boreflex lowest dose) sheet and has failed or partially failed in the central region.

In the first scenario, since the bond is postulated to have failed along the entire length of Boreflex, the shoot is unrestrained and would be expected to shrink in a stress free condition as irradiation proceeds.

A net shrinkage would occur but tearing of the sheet and 5-16 E

subsequent gap formation would not be expected.

If the bond were " perfect", as in the second scenarlo, high local stresses would develop in the sheet as the material tries to shrink and one might expect the material to tear at many locations forming aany small gaps along the length of the sheet.

In the third scenarlo, with the Boraflex sheet restrained at the ends, and as the material shrinks, the greatest accumulation of local stresses would be expected in the central position of the sheet.

Therefore, the sheet may pref erentially tear in the central region.

This scenario seems to be supported at least partially by the Quad Cities gap measurements of gap occurrence versus axial elevation (see Figure 3-6).

Assuming a mechanism of restraint, whether It is provided by the cdhesive or by mechanical means, it is useful to estimate the accumulation of stresses as a function of gamme exposure relative to the tensile stress of Boraflex. In Section 4 it was shown that Boraflex shrinkage from the latest BISCO Co-60 Irradiations could be aall represented as a linear function of dose up to an exposure of 1 x 9

10 rads (see Figure 5-5 for r'esulting strain versus dose). Data were presented in Section 4 (Table 4-3) from early BISCO test Irradiations with the "NS-1" polymer (believ;d to be unfilled Boraflex) which provided an estimate of changes In elastic modulus as a f unction of 8

exposure to 7 x 10 reds. The data presented in Table 4-3 have been fitted as a linear function of dose as shown in Figure 5-6 over one decade (1 x 10 0 to 1 x 10' reds). The work of Bueche 13,3J. described in Appendix B has Indicated that the elastic modelus of PDMS can be represented as a linear function of exposure over such ranges.

If the strain Induced in a restrained panel of Boraflex as it tries to shrink is a linear function of dose and the elastic modulus increase es a linear function of dose, the product, the stress, will increase os the square of the dose. To est imate the bulldup of stresses in a 5 17

restralned panel of Boreflex, the elastic modulus and strain as a

/

G function of exposure shown in Figure 5-5 and 5-6 are used. Figure 5-7 shows the results of that calculation and Indicates that over one decade in dose, stresses approaching 600 psi are estimated.

The tensile strength of unirradlated Boreflex has been reported in Appendix A as being 200 pst which is somewhat lessOver than for the the range unfilled NS-1 polymer reported in Table 4-3 (510 psi).

polymer in exposure studied, the tensile strength of the NS-1 increases to 896 psi (or about 755). If it is assumed that the tensile strength of Borsflex increases in a similar manner, say to 350 psi, over this exposure Interval it can be seen that the estimated stress is of the same magnituda as the tenslie stress.

The presence of chemical agents such as water, free oxygen and other agents may reduce the stress to tear relative to what would be in air or vecuum. In summary, these estimates tend to confirm the bulldup of stresses in a restrained panel of Boraflex to levels sufficient to cause tears.

Once a tear occurs, the growth of the resulting gap can be estimated using the shrinkage dato contained in Figure 5-2. In Figure 5-8, the cumulative gap in terms of percent of the initial Boreflex panel length is plotted as a function of samma exposure. For a given fuel rock design, the maximum cumulative gap size (percent) In this Figure is multiplied by the length of the Boraflex panel to obtain the cumulative gap size in Inches. Should more than one gap form in a panel, the sum of the Individual gaps would be equel to the cumulative gap size. Figure 5-8 Indicates for a Boraflex panel 144" long, the range in cumulative gep size would be between 4 in. and 6 In.

Water Parmentinn and Onunnn Waleht Chennet it has been noted that gamme radiation seems to change Boreflex in such a menner as to make it susceptible to some degree of water b

v 5 18

1 l

l O i 1

I l

i e i i e i e i i 1.4 - _ )

1.2 -

1.0 -

?

(0.8 l

1 u)

O.S 0.4 -

0.2' i e i i i i e i 1x10s 1x10' Gamma Exposure, Rads Figure 5-5.

ShrinkageInducedSt.aininaRestragnedPanelof Boratlex Over the Exposure Range of 1x10 6 to 1x10 Reds. I O

5-19

. . . . . . i e n -

o' 30 M

9 y

S I20 -

5 3 -

3 3 10 l O *-

• 1x108 1x10' Gamma Exposure, Rads Figure 5-6. Elastic Ngdulus of the 9 NS-1 ' Polymer Over the Exposure Range of 1x10 to 1x10 Rads.

O 5-20

___m _ _ _ _ _ _ _ _ _ . - _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . . _ . _ _ _ _ . _ _ _ _ _ _

s O

600 . . . . . . . .

400 -

3

o. ,

2 -,

Di 200 -

0 8 ' ' ' ' ' '

1x108 1x108 j Gamma Exposure, Rads Figure 5-7. Estimated Stress Ig a Restrgined Panel of Boreflex Over the Exposure Range of 1x10 to 1x10 Rads. j l

4 l

5-21 I

J ---_ xI

O i j

s g

5.0 g

~

Upper

• Estimate er 4.0 -

E

\N Lower C 3.0 - Estimate

.E 5s -

l 2.0 O' ?m - ,

ha 1.0 2

I I

1 010 10" 0:s 10 108 Gamma Exposure, Rads Maximum Cumulative Gap Size as a Percentage of Figure 5-8, Initial Boratlex Panel Length versus Exposure.

O 5 22 i

g

permention. This has been suggested by radioassay measurements of the Irradiated and untrradiated panels inspected at Point Beach. $1milar ll measurements of the Quad Cities coupons for pool cont,aminants tend to confirm this findlag. The Irradiated Point Berch panel as well as surveillance coupons further Indicate that water permention starts at has been cut to SIIe after the edges where the panel / coupon manufacture and progresses inward. Edge erosion has also been noted in the latest 81500 reactor Irradiation and was attributed to thermal neutron radiation damage 1211. Water permeation from the cut edge of the coupons could have also been a factor based on the as-manufactured condition of the edge as described subsequently.

I In order to determine whether there are differences between the flat, as hiny" surface of Boreflex and the edge where It has been cut to size, mounts of both surfaces were prepared from fuel rack grade Boraflex for scanning electron microscopy (son). Figure 5-9 shows the scanning electron micrograph of the flat, " shiny" surface of Boraflex at magnifications of 270X and 2800X. The surface appears relatively smooth, probably characteristic of the polymer matrix with minor Imperfections including "rlpples" and some scattered " debris" on the ]f surface. The former may be an Image.of similar Imperfections in the carrier sheet used during the forming process. The latter could be filler particles which have not been completely sealed by the polymer.

In general, the flat surface appears more or less completely " sealed" )

by the polymer matrix.

l Figure 5-10 shows the cut edge of the same specimen of Boreflex at 280X and 2800X magnif icat ion. In contrast to the flat surface, the cut edge contains relatively deep volds (compared to the flat surface) and what appears to be exposed particles of filler material. Some of the volds or pores may have been caused by filler particles which were extracted from the polymer matrix during the cutting process. The pores shown In Figure 5-10 are probably deeper than they appear in the A

micrograph since som tends to " flatten" the depth of field.

comparison of Figures 5-9 and 5-10 confirms that there are distinct ,

differences in surface morphology (flat versus cut edge) which could f make the cut edge more susceptible to the onset of water permeation.

O 5-23 i

• b i ___

m j

O c .: #.: ,e x *- ..

.: . , . A

. v, - y,s

.5 5

f f4.4 - ;fs .

o "

4 .. -

7.

?

, g, [j

. l' ~

.p ,,

.. , ffy Y' .

Figure 5-9. Scanning Electron Microscopy of the Flat Surface of Boraflex at 270x and 2800x Magnification.

l O 5-24 i

I

.a

~.

O 1

l i

, ,u y k

~ E),,. .. -

.". .f *

.3}&j , '(d?"g Vt, I

Lg

• , ,t -s  ;

,4;.).

__ a. a h -

I

~

'k '

'3e .% , ' * * . ' i r< -

" I {}t

[EY a- ., - a.

f, . ,

Figure 5-10 Scanning Electron Microscopy of the Cut Edge of Boraflex at 280x and 2800x Magnification.

O 5 25

l O (he T

see Figure increase 4-6) up in to asample gamma doseweight of 2.5 x in

,10 the latest BISCO reacto IO rods is further l

l Indication of water permention. Beyond this exposure the samples show At a weight decrease which has been attributed to edge erosion 12A1.

exposures of 10' rads and less, sample weight gain is less than .5%

This result is generally and increases to 3 to 45 at 2.5 x 109 reds.

consistent with the postulated development of open porosity discussed previously.

It is interesting to note that samples irradiated In air in the Co-60 in weight up to an exposure of I x 10' tests show a similar increase reds. If crossil'nking were the only mechanism at play, one would expect a decrease In sample weight due to the generation and evolution It wrs noted in Appendix 8 that the of H' and hydrocarbon radicals. (humidity) in the presence of molecular oxygen and perhaps solsture air combine with free radicals remaining in the polymer. This offers l

one potential explanation for the cpparent increase in sample weight f measured in Co-60 tests conducted' in alr.

It is noted that such effects may continue after the polymer is removed from the 0, chemicalradiation field as long as free radicals are present.

Taa+ Dame +nr verana Knan+ Fual Pnnt Radin+ tan Pandt+tana in order to accumulate Integrated gamma exposures in a reasonable time frame, BISCO conducted test reactor Irradiations. It has been noted I

that significant differences between the test reactor radiation spectrum and that in the spent fuel pool exist with the total energy deposited in the test reactor balng significantly greater ,than the In the test reactor, fast fuel pool for the same gamma exposure.

neutron radiation as well as recoli alpha and LI particles from thermal neutron reactions with Boron-10 (In the4BC filler) are present and are not factors In the pure gar.ma spectrum of the spent fuel pool.

A rough estimate of the effect of fast neutron and recoli particles f u 151 has can be made. The experimental data of Kilns and Jacobs 5-26 a

been used to estimate the energy deposited by fast neutrons.

I I Corrections to the neutron scattering cross shettons have been applied to account for differences between Boraflex and the polymers studied.

The estimated energy deposited by recoll particles from neutron Induced reaction in Boron-10 depends on a number of verlables including the thermal flux depression factor in the samples, boron carbide particle size and test reactor spectrum. Since the exact sample configuration and boron carbide particle size distributions are not known, only ar. order of magnitude estimate is possible.

Nevertheless, such an estimate Indicates that the combined effect of fast neutron and recoli particle energy deposition is at least an order of magnitude greater than th comme dose alone. That is to say, for a gamme exposure of 1 x 10 rads in the reactor, the total oxposure is probably in excess of 1 x 10 12 reds. Of this additional exposure, this estimate suggests approximately one third is attributable to fast neutrons and two thirds to recoli particles.

)

Bernfiar servlee Ltfa The ultimate service life of Boraflex will be determined by the ability of the polymer matrix to perform its Intended function--to retain the boron carbide filler motorist. That ability is likely to be influenced by not only the Integrated gamma dose received during the service lif e (estimated 13.Q) to be in the range of 2.5 x 10 10 9, 3 x 10 lI reds) but also by the offacts of long term exposure to the cqueous pool environment. It is noted that In the BISCO reactor tests the samples were subjected to perhaps an order of magnitude greater dose (when fast neutron and recoli particle energy deposition is considered) than that expected during the service lif etime in a pool.

Ths test Irradiation envircoment la therefore conserv4tive relative to

^

the Integrated gamma exposure projected over forty years in a spsat ,

I fuel pool. However, the tests were relatively short in duration (five ,

to six months) and therefore do not provide complete data of the long term effect of environmental factors, guch as pool water chemistry, temperature, etc.. Accordingly, while there has been no data available to this study which would suggest the onset of a rapid or 5 27 h -

gross degradation mechanism leading to the loss of boron carbide from Boraflex, it appears prudent to continue adequate surveillance O programs to monitor the long term performance on a plant by plant basis.

Factors which may influence the service life of Boreflex includst e Integrated gamma dose o Pool water chemistry e Pool water temperature e Rock design specific conditions such as the potential for local flow around the Boraflex e Bor aflex thickness e Boron carbide and other filler composition in the Bora-flex l

I W!th regard to the first four fe-tors, Information is presented in Appendix B which Indicates that the environment (i.e., presence of l molecular oxygen, water, etc.) can have an influence on the behavior j of materials, such as Boraflex, under Irradiation. These effects may 1 be long term In developing since the rate at which agents in the pool  ;

water enter Boraflex is likely to be a diffusion governed mechanism.

As such, pool temperature, the presence of open porosity (believed to l be a function of gamma doso), and the availability of chemical agents j (potential for Iccal flow around the Boraflex) may all be important factors. Furthermore, there is some Indication from the Point Beach coupon train that local flow can cause erosion from the surface of Irradiated Boraflex 111.

The Initial thickross of the Boraflex sheets is likely to be a factor which influences service life. Boraflex is used in spent fuel racks over a range of thickness from 0.040 In. to in excess of 0.100 in..

If, as has been suggested, water permeation and chemical effects are important, then they may have a lesser effect (on a percentage of Borsflex volume bests) on thick sheets than on thin sheets for the same exposure conditions. .

5 28

. . , L

- _ - _ _ - ~ ~

e in addition, the volume percentage polymer versus volume perc filler (i.e., boron carbide loading) may also effect sarviceentage life. For of Boraflex with high boron carbide loadings, the volume percent of polymer is less and there la physically less matrix to retain the boron carbido.

In summary, the data currently available, and in particular that for the Point Beach panel which had been exposed to the pool environment for 6-7 years, has demonstrated that Boraflex has performed its Intended function of this period of time.

retalning the boron carbide filler material for However, since this period of experience is relatively short relative to the design service life of 30 to 40 years, caution must be exercised service life and it in projecting the ultimate allowable is recommended that the performance of Boraflex be monitored, vertfled and documented.

no Indications From the available data there are aqueous envir onment result that combined exposure to gamma radiation and the pool in the onset of rapid or gross degradation of the polymer matrix that would cause the loss of boron carbloc.

Factors have been Identified wh!ch are likely to effect service life 4 and some of these will very from plant to plant.

formal Accordingly, a surveillance program at each plant appears a prudent course to verify the continued serviceability of the Boraflex.

5-29

_ _ _ _ _ _ - - - - - - ~ - - - - '

- ._ . , , _ .. _ _ _ - - - _ - - - _ - - - - - - - - -------m__

l 1

O Section 6 j 1

CONCLUSIONS AND RECOMMENDATIONS, l

When Boreflex is subjected to a gamma radiation field it undergoes changes as been evidenced by data from the 815C0 test Irradiations, utility surveillance programs and from other sources identified In the literature. 'The data from these sources has been evaluated as part of l this project and the observed changes In physical characteristics of Boraflex are what would be expected for this class of polymers based on reports identified in the litersture. In summary, evidence of j radiation Induced changes includes e Evolution of gases including hydrogen and hydrocarbons e Physical shrinkage accompanied by an increase in spect-fic gravity and decrease in specific volume e Loss of rubberelasticity, increase in hardness Lnd an increase in elastic modulus '4 e Observed formation of blisters in some specimens e Change in surface color from shiny black to a whitish l grey and a tendency of the discolored material to 1

powder e Permeation of water believed to be the result of develop-ment of open porosity The first four of the above are consistent with the radiation damage mechanism of crosslinking as described in Appendix B. The color change and formation of porosity may be a result of scissioning and/or chemical effects attributed to the aqueous service environment.

Physical shrinkage seems to very from sample to sample--perhaps a result of manufacturing verlables such as sheet thickness and filler content. Some samples have shown greater degrees .of anisotropic shrinkage than others (i.e., more shrinkage in one direction than In O

l 6-1

- - _ _ _ _ _ _ . .__ _ _ ._ _J

another) and this may again be due to manuf acturing variables. The maximum shrinkage in either direction seems to be limited to 3-45.

These estimates provide a basis for projecting a maximum cumulative gap size for a given rack design should a particular design be subject i l

to gap formation.

l l

While neutron radloessay measurements detected the presence of gaps in the Quad Cities racks 121, inspection of a full length panel from Point Beach 11.1 and neutron radioassay measurements at Turkey Point 1221 Indicate that gaps have not formed in these rack designs. A key factor determining whether e particular rock design is susceptible to gap formation appears to be the manner in which the Boraflex sheets are restrained in the rack structure. In a tightly restrained panel of Boraflex, stresses approaching the tensile stress of the material f over a relatively small increment of exposure (10 8 to 10' rads) have been estimated, thus suggesting a mechanism for gap formation.

The recent test reactor Irradiations sponsored by BISCO have subjected samples of Boraflex to Integrated radiation doses well in excess of those expected during the 30 to 40 year service life of fuel storage racks. The gamma dose alone in these tests was approximately equal to that expected during the normal service life of the racks. The energy deposited by fast neutron and recoli perficlea results In an equivalent dose estimated to be an order of magnitude greater than the gamma dose alone. These tests Indicate no loss of Boron-10, as determined by neutron attenuation measurements, but slight edge j

erosten of the Boraflex in some samples. The latter has been attributed to thermal neutron /recoli particle ef fects 12Al although the presence of water cannot be discounted as e confrlbuting factor.

In summary, the BISCO test Irradiations demonstrate Integrity of the Boraflex to radiation doses well in excess of those expected for the material in spent fuel racks. It has been noted, however, that while the tests were conducted in en aqueous environment, the duration of these tests was only on the order of five to six months. Accordingly, material Integrity under the combined effects of gamma radiation and O 62 R

l long term exposure (30 to 40 years) to the aqueous pool environment Data available to this study would Indicate j remains undemonstrated.

no mechanism to suggest a rapid or gross degradation of Boraflex that could result in the loss of boron carbide.

l l

Evidence from the literature has been presented which suggests that I

the presence of chemical agents including air, molecular oxygen and cater can influence the behavior of polydimethylsfloxanos and other materials similar to Boreflex. The rate at which these agents penetrate the Boraflex is likely to be a diffusion governed mechanism.

Differences between the true volume of irradiated Boraflex samples and the geometric volume as determined by dimensional measurements suggest c radiation Induced mechanism which produces open porosity in the Boreflex. This is further supported by the penetration of radioactive pool contaminants from the pool water in the irradiated Point Beach panel relative to the unirradiated panel. The development of open porosity would be expected to accelerate the permention of pool water into the Boraflex.

Because of these factors, the ultimate long term service life of Boraflex subjected to gamma radiation concurrent with long term oxposure to the pool aqueous environment is somewhat uncertain from the data currently available. Owing to the nature of the material and the mechanisms at play, demonstration will probably awalt long term oxpoaure In the spent fuel environment. Accordingly, continued surveillance of Boraflex currently in place in spent fuel racks around the country would seem to provide the best source for developing these long term data.

As a result of these conclusions , the following recommendations have been developed with the Intent of Improving the quality of data generated in the future relative to the performance of Boraflex in spent fuel racks e The IrradiatedpanelexaminedatPolnf0 Beach has recolv-ed an exposure estimated to be 1 x 10 rads accumulat-ed over a period of approximately 7 years in the spent 63

fuel pool. Consideration should be given to the proper-etion of large coupons from this material for reinser-f- tion into a spent fuel pool. Characterization of the coupons should be as recommended in Appendix C where pos-slble. Some coupons should be exposed to f urther irredi-ation in the pool and others only to pool water without further Irradiation, e Consideration should be given to establishing lead PWR and BWR surveillance programs with surveillance coupons prepared and monitored per the guidelines outlined in Appendix C. For these programs some additional charac-terization and measurements may be desirable.

e in llou of further sample Irradiations in test reac-fors, consideration should be glven to an accelerated exposure program in a utility spent fuel pool. This may be conducted as part of the lead programs described above.

e Determine whether some physical property of Boraflex can i be correlated with gamma exposure to serve as a dost-meter measurement of Integrated dose. Alternatively, ,

i develop methodology to rapidly and with reasonable de-gree of atcuracy calculate the Integrated gamma expo-sure of surveillance coupons.

e Continue to collect and evaluate data from utility surveillance programs as they become evallable. These

/~ data can be used to supplement the data collected as

(,,}/ part of this study and to further verify the estimates of Boraflex performance.

e For utility coupon programs initiated in the future, the manufacturer's recommended surveillance procedures should be supplemented with those outilned in Appendix C. .

1

\

l The following recommendations are made relative to the design and manufacturing of new spent fuel racks which utillze Boraflext l e No means of mechanical or adhesive restraint should be used so that the material can undergo shrinkage in a stress free condition thereby precluding the potentle'l for gap formation.

e Oversize panels should be provided to compensate for the effect of shrinkage on the reactivity of the fuel / rack system.

I O ... .

6 A

O Se>; tion 7 REFERENCES

1. C. W. Fay (Wisconsin Electric Power Co.). Letter to George Lear (U. S. Nuclear Regulatory Commission), 'Results'of Examination of Poison insert Assemblies Removed from the Point Beach Nuclear Plant, Units 1 and 2,* Docket nos. 50-266 and 50-301, February

( '. 1987.

2. hu~ etat w a u +r an A++ anum + tan Taa+ #ne unch nai '+v snan+ Fumt RacktlWat.1. Mounta in V low, CA: National Nuclo r Corporation for Commonwealth Edison Co., December 1987.

3. Pratletnarv Aaammamant af Raraftav Parfarmanca in 4hm Dund Et+ tan Kn n+ rumi C+nrana ameka. Kingston,.New York: Northeast Techno-logy Corp, April'1987. NET-042-1, Rev. O.

4 nahavtne af utch-nanal+v snan+ . runt s+neman amekm.. Palo Alto, Calif.: Electric Power Research Institute, August 1986. NP-4724

5. W. G. Hairston, Ill (Alabama Power Co.). Letter to K. Lindquist .

(Northeast Technology Corp), " Joseph'N. Farley Nuclear Plant-- -

Units 1 and 2, EPRI Project RP-2813-4,* NS-87-0237, August 20, 1987. -

6. Fwamtam+lan nf Enant Funt mark turvatttanea cananta J2 Frne MfAl-a+nna 2 Windsor, CT: Combustion Engineering, Inc., October 29, 1987. 70385-9351-Q-013.

7. 1"==arv Ranar+ nn rv m i nm+ t an n # ornnan mara#inv turvatttane.

Cananna. Palm Harbor, FL: Nusurtec, July 20, 1987.

8. P. H. Kohn (Wisconsin Electric Power Co.). Teleph6ne converse-tion with K. Lindquist (Northeast Technology Corp), January 21, 1988. ,

9. H. 0. Nelson (Northern State Power Co.). Letter to K. Lindquist (Northeast Technology Corp), April 24, 1987.

10. Fwaminm+ tan af inan+ Fuat Pan t C+nrana mark Mm+artat. Lynchburg, Va.: Babcock and Wilcox, January 5, 1983. RDD: 83: 5194-01: 01.

11. rw minm+ tan n# tarnna s+nemna mark um+metat cannnn. Lynchburg, Va.: Babcock and Wilcox, January 17, 1984 RDD: 84: 7278-01:

01.

12. A. J. Blamey (Commonwealth Edison Co.). Letter to K. Lindquist (Northeast Technology Corp), May 22, 1987.

I

13. A. J. Blamey (Commonwealth Edison Co.). Letter to K. Lindquist (No,theast 1.chnolog, Co,,), 0. combe, 29, 1987. l l :d 7-1 b_____----------------

Letter to K. Linquist

14. A. J. Blamey (Commonwealth Edison Co.).

(Northeast Technology Corp), January 29, 1988.

Letter to

15. G. L. Perez (Sacramento Municipal Utility District).

K. Lindquist (Northeast Technology Corp), May 29, 1987.

Letter to K. Lind-

16. A. W. Tyson (Carolina Power and Light Co.).NF-87-493, July 30, 1987.

quist (Northeast Technology Corp),

Letter to J. Scarborough

17. R. R. Burn (University of Michigan). June 18, 1986.

(Carolina Power and Light Company),

18. "Model 30GL A Scale Durometer." Pacific Transducer Corp., Dada l Rutla+1n 9111-4, los Angeles, Ca.

Pacific

19. " Instructions for Nodels 306 and 306L Type A Durometer."

Transducer Corp., Los Angeles, Ca.

20. Riarkamen T==+ tan nf Rnemflaw in Katme+ad catin af +ha Turkav Poin+ if n t + 3 (nan + Fuat K+ncana Racks, Palm City, FL, Nusertec inc., August 1987.

Telephone converse-

21. R. Gouldy (Florida Power and Light Company). March 7, t ion with K. Lindquist (Northeast Technology Corp),

1988.

22, Irradtm+ Inn K+ndu af.Anrmfley Wan+rnn Khtaldinn Mm+arfalz, Park Ridge, IL, Brand industrial Services Inc., August 12, 1981, 748-10-1 Rev. 1.

R. R. Burn and G. Blessing, " Radiation Effects on Neutron Shield-O 23.

Ing Materials, Temne.

pp. 48-49.

Am. Nort. Cne , vol. 32, Suppl. 1, 1979, 24 R. R. Burn and G. Blessing, " Radiation Effects on Spent Fuel ene_,

Am. wort.

Storage Rock Neutron Shleiding Materials", Trant-Vol. 38, 1981, pp. 429-431.

25. Rarmflav Rut +mht11+v Rs.gaci, Park Ridge, IL, Brand industrial Services Inc., May 5, 1978, 1047-1 Rev. 1.

26. Tat + s 748-91-1. A Final Rannr+ nn +he Fffme+< af Winh Ta mn ar m-in RitPO Rnemftaw Nan +ran A h e nr h e r Fwnneurm tura Rnrm+ad Wa+me Wa+artml, Park Ridge, IL, Brand Industrial Services Inc., August 25, 1978, N-2.

McGraw-

27. R. Juran, Ed., Mndarn Plat +tra Fnevrinnadta, New York:

Hill, 1987, p 124.

28, Irradfattan K+udv nf Rnemflav Nan +rnn Ahenrhar fntarlm Tat + D a9 m ,

Park Ridge, IL, BISCO Products Inc., November 25, 1987, NS-1-050 (InterIr) Rev. 1.

29. K. Lindquist (Northeast Technology Corp), Telephone conversation with L. Detrich (BISCO Products Inc.), December 1987.

30. S. Turner, Presentation at the EPRI Boraflex Workshop, Washing-ton, D. C., December 17, 1987.

O 72

l

31. G. G. Delldes and 1. W. Shepherd, " Dose Effects in the Crosslinking of irradlated Polystloxano", Radim+ tam Phwaten a nd chamta+rv, Vol. 10, 1977, pp. 378-385. >

33. C. D. Popp and D. Sissman, " Radiation Stability of Plastics and Elastomers", Ruc h , Vol. 13, 1955, p. 28.

33. A. M. Bueche, "An investigation of the Theory of Rubber Elasti-city Using Irradiated Polydimethylsiloxane", Janennt nf.Pntw.

mar metanem, Vol. 19, 1956, p. 297.

34. D. E. Kilne and A. Jacobs, " Calculations of the Rate of Energy Deposition in Polyethylene by Reactor Radiation", Janenat ef Annifad Phwatem Vol. 30, 1959, pp. 1741-1747.

35. A. Jacobs and D. E. Kilne, " Energy Deposition in Polymers by ,

Reactor Radiation", fanenal af Anelfad Pn f umm e. Relanca, Volf VI, l 1962, pp. 605-612.

I 1

0!

7-3 I

O APPENDIX A PROPERTIES AND CHARACTERISTICS OF POLYSILOXANE POLYMERS 1

0

O CONTENTS ser+ ten l

GENERAL MOLECULAR STRUCTURE OF POLYSILOXANE A-1 CHARACTERISTICS OF ELASTOMERS A-2 i j

BORAFLEX ELASTOMER A-3 REFERENCES FOR APPENDlX A A-7 I i

1 1

l l

{

O A-iii i

1 l

1 Appendix A PROPERTIES AND CHARACTERISTICS OF POLYSILOXANE POLYMERS GENERAL MOLECULAR STRUCTUPE OF POLYSILOXANE Polymers are large molecules, typically long chains comprised of rather simple units repeated many times. Generally, they have a I

distribution of molecular weights described mathematically by various

1. averages". In the solid state, they can be amorphous or crystalline or a combination of both. The amorphous regions soften over a range of temperatures, as in " glass" transition, whereas the crystalline roglons, if present, have a more clearly defined molting " point" which characterizes them.

Most polymers have carbon in the backbone of the chain with side groups which distinguish one polymer from another, but the chain can also include other atoms such as N, 0, and St. The polystloxano polymers, along with a relatively few other polymers have an Inorganic backbone chain. In the polysiloxane case, this is characterized by:

A A A I I i

- SI - 0 -

SI - 0 - $1 -

1 I i B B B i Where A and B are typically hydrocarbon side groups. The most common silicone polymer has a " repeat unit" where the form of A and B Is CH3 and is commonly referred to as silicone rubber. In the literature, sillcone rubber is also referred to as polydimethyl siloxane (PDMS) and dimethyl polysiloxane. Other common types include a phenyl or vinyl group for A and/or B.

A1 O

i

. _ _ _ _ _ _ _ _ . _ _ _ _ _ . _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ m

l polymers are characterized by a temperature (O) These

, service silicono range classes which is very of large compared to many other polymers IA:11 The $1-0 backbonc is degraded or oxidized only slowly at about (

400 C or higher and typical service temperatures extend up to 300*C.

0 l At the same time these polymers are flexible at temperatures as low as l

-100'C, a property which is probably partly due to the flexible oxygen l

linkage and limited crystaillnity. Without significant crosslinking between chelns, the polymer chains can range in size and form from l oils to viscous liquids to elastic solids. Although the formulation of Boreflex is proprietary, the general form is a polysiloxane polymer believed to be slmlier to polydlmethyl s!!oxane.

l l

CHARACTERISTICS OF ELASTOMER $

Generally the description of an ideal elastomer is 1A:11 l e it should be rapidly stretchable to very high elonge- l 1

tions (up to 5005 or more).

e It should possess comparatively high tensile strength when fully stretched, l.a., the tensile restoring force -

l I

continues to increase as the elongation increases to high extension. l

{v~'{

e it should snap back (rapidly) when the stress is re-leased.

1 l

e It should retract completely with no permanent set.

When the above behavlor is observed, certain molecular and environmental conditions usually exist. They are 1A 11:

e The material is a polymer.

e The material is amorphous.

e The temperature is above Tg , the glass transition temperature.

e The material is lightly crosslinked.

It will become apparent later that many of the above ere notable In the case of Boreflex, and that changes In Boreflex properties In the service environment may be correlated with the descriptions and conditions of effects reported in the literature.

O A-2

The rubberelasticity of polymers is provided by stretching and I rearranging the chains under stress, where the system is in a very flexible state. Return to the shape of the original state is provided by points which act as ties between chains and these, in principle, can be crystalline regions, chain entanglements, or, most commonly, i crosslink bonds between chains. 'Crosslinks are typically made by '

chemical ac+1ons sometimes called " curing" or " vulcanization"; but, as will be noted later, nuclear Irradiation can also yloid crossi, Inks.

For slight crosslinking or a small number of tio points, the chain j length between crosslinks is large and the elongation that can be obtaired under stress is very large, perhaps 1000$ or more. However, the rubberelastic modulus is correspondingly low because the system is  !

Inherently not very rigid. If the crosslinking density is increased, the system becomes more rigid. The modulus rises, but the elongation which is possible decreases. Concurrent with increased crosslinking, shrinkage typically occurs probably because the chains are brought l closer together and the free volume decreases.

i As in the cose of polymers in general, rubber elastic materials are l considerably affected by fillers and additives. It is very common to cdd fillers as a method to change the proporties. In other cases, the filler is needed to make the elastomer useful in a special way, and, as a result of the filler and the service environment, the physical properties, etc., change very noticeably.

Such is the case with Boraflex where 48 C is the filler used to provide reactivity control, and, in turn, the filler tends to affect the elastomer's properties.

BORAFLEX ELASTOMER Boraflex ls reported J.A,11 to be a primarily methylated polysiloxane clastomer filled with finely divided boron carbide powder, 8 4C, (100 to 400 mesh). The polymer is believed to be very similar to dimethyl polysiloxane as reported in the literature, but it is assumed that verlations in the structure, possibly at the methylated site, might be present. For example, R e f er e nc e 1,A .11 c i te s the addition of phenyl A-3 u .

_ _ - _ - - - - - - _ _ _ _ - - - - - _ - - - - - - - ~

groups, in the amount of 0.1 to 0.6 mole present, for methyl groups to vulcanization rate and the Initial crosslink Increase both the density. Phenyl groups are sometimes substituted to improve low temperature performance, to enhance radiation resistance and to lower resillence. Carbon and silica fillers have been addedIt to hasthese been polymers which have been vulcanized with radiation 1A=A1 in Boreflex for the reported 1A-11 that carbon and sillca are present purpose of increasing radiation resistance of the material. .

Polysiloxane polymers are generally reinforced with allica fillers increase the tensile strength of such 1A=11. The use of silica can polymers from 50 pst for the nonreinforced polymer to 1500 ps! for the l Fumed and/or precipitated silica is generally reinforced materiel.

with a number of hydroxyl groups on the used and consists of $10 2 surface of the filler aggregate. It has been postulated 1A=11 that the increased tenslie strength results from hydrogen bonding and/or l

van der Waals forces between the hydroxyl radials on the filler and I oxygen atoms in t,he spine of the polymer.

O The PDMS polymer is basically the matrix for the composite Boreflex 1A=11, with the B4 C serving as a filter with a large neutron thermal absorption cross section. It is because of the B-10 Isotope in the naturally occurring boron that it is useful in nuclear fuel storage The polymer is racks for providing the required react'Ivity control.

manuf actured by Dow Corning for 88500 and has been ref erred to as NS-1 polymer 1A=11 Boraflex can be manufactured with ranges of 8C 4 content (percentages) and dimensions (thickness). Reference 1A=ll cites a typical composition a for spent fuel storage applications as noted in Table A-1. The B4 C powder can very in sizecomposite from 100 to 400 1A=11 f on the B-10 loading of the mesh depending l

Presumably, a finer mesh powder is used at the higher B-10 loadings to l sssure complete encapsulation of all or most of the B4C particles.

According to reference 1A=11, "The NS-1 polymer acts to provide a waterproof coating for the majority of the boron carbide in the matrix, thus reducing any chance of water contact." ,

O A-4 g

. l Based on the elemental compositions cited in Table A-1, .t Is possible to estimate the weight percent polymer matrix and filler in the composite Boreflex material. Such an estimate yleids 35-40 w/o polymer, 40 w/o 84 C and 20-25 w/o silica with perhaps a small quantity of carbon. Accordingly, in the composite material a little more than one third (on a welght percent basis) is comprised of the matrix and a little less than two thirds filler.

In addition to the constituents identifled in Table A-1, It is likely l that Boraflex contains small quantitles of other materials. Because i of its form (elastomer versus oil or gol) the polymer matrix In j Boraflex is likely to contain some initial degree of crosslinking. 1 accomplished chemically in PDMS during the .

This is can be manufacturing process by the use of vulcanizing agents such as f i

peroxides .LA 11 and platinum believeit to be used in the case of As discussed subsequently, the presence of such residual f Boraflex.

f catalysts can influence the nature of radiation damage. l f

2 As-produced Boreflex with a B-10 loading of 0.020 gm B-10/cm and 2.5 mm thickness exhibits the properties shown in Table A-2 .(.A 31 It is to be noted that elastomers in perilcular and polymers in general are typically affected to a large extent by fillers. The properties summarized In Table A-2 would be expected to depend upon the nature of l the fil!sr, the filler properiles, the filler size, the service and i environment in which it is used, etc. . In addition to the changes in the polymer matrix with service, the filler properties .and bonding might also change, particularly If the environment includes nuclear radiation.

l I

\

t

\ O! l A5 j __-_____/ l

Table A-1 TYPICAL BOP.AFLEX ELEMENTAL COMPOSITION (5) l O )

Element Composition Boron 31.5 w/o Carbon 19.0 Silicon 24.5 0xygen 22.0 Hydrogen _242 TOTAL: 100.0 (1005)

Source:

Bornflav Neutrnn shlaldinn Mn+ertal Pendue+ Performance Data, Park Ridge, IL, August 25, 1981, N-38, 748-30-2.

Table A-2 i

~"\ . BORAFLEX PROPERTIES (Y

Property Amount Modulus of Elasticity 1000 psi .

l l

Tensile Strength 200 pst 1.7 g/cc 3 Speelfic Gravliy Har dness 75 Shore A 200 CC# j Temperature Stability

1. Note: aintrum Rnenflaw without variable Neu4rnn distoriion Khfaldtrn Mntorial Produc+ Performance Onta, Source:

Park RidDe, IL, August 25, 1981, N-38, 748-30-2. 3 1

I

)

l A6 i

i

)

I REFERENCES {

A-1. D. J. Cornellus and C. M.14onroe, "The Unique Properties of Silicone and Fluorosilicone Elastomers", Palvewr rnntnaartna ~ ~

and selanca, Vol. 25, 1985, pp. 405-473. l 1

A-2. F. W. Billmoyor, Tex + bank of Palvrer setenea, New Yorks inter- 1 Science Publishers, 1982.

A-3. Dnraitav Nau+rnn thfaldinn Materfat. Pendue+ Parformanca Data, l l

Park Ridge, IL, Brand Industrial Services Inc., August 25, 1981, N-38, 748-30-2.

A-4 F. A. Bovey Fffnets of Innfiten RadIn+1an on Natural and svn=

thafte Polvmart, New York: Interscience Publishers Inc. 1958.

I A-5. Bnraflaw sut+ahIIftv Rannr+, Park Ridge, IL, Brand industrial l I Services Inc., May 5, 1978, 1047-1 Rev. 1. l l

l 4

A7

-~- .

APPENDIX B RADIATION EFFECTS IN POLYSILOXANE POLYMERS l

1 i O

e - O . m m CONTEl4TS Smetinn GENERAL EFFECTS IN POLYMERS B-1 EFFECTS IN POLYSILOXANE: A REVIEW OF TFE LITERATURE C-4 CrosslinkI'ng and G-Yalues 34 Ef f ect of Radiation on flie Properties of Polysiloxane g.g EnvIronmentei Effacts B-11 REFERENCES FOR APPENDlX B g.j7 l t

B-tii

, ._m ._. . . . v Appendix B RADIATION EFFECTS IN POLYSILOXANE POLYMERS It has been noted in Section 4 that data are not available relative to the long terni performance of Boreflex une.r simultaneous exposure to gamma radletion and aqueous the spent fuel pool environment.  !

Accordingly, the published literature of radiation studies with the PDMS polymer has been reviewed. The Intent of this review is to document the relevant literature and to provide an understanding ofs i

e The predominant radiation damage mechanisms in PDMS l e The offact of radiation damage on the physical proper-ties of PDMS ,

o The effect of environment (i.e., presence of air, oxy-  !

gen, water, etc.) on the physical properties of Irra- '

disted PDMS A search of the literature has Indicated that the effects of Ionizing rodf ation on the polys t loxane polymers, and In particular PDMS, have bcon studied experimentally quite extensively starting in the 1950's.

While the data reported are, in many cases, for relatively low dose experiments (up to 5 x 100 rads) as compared to the Integrated dose expected during the design service life of Boraflex, the experiments were conducted under various yet carefully controlled conditions.

Thus, the effect of radiation and service environment can be examined olther Independently or in combination. The following Appendix of this report contains a summary of the published literature relative to experimental studies of radiation effects in the PDMS polymer.

GENERAL EFFECTS IN POLYMERS When a polymer is subject to a radiation field, changes in the atomic / molecular structure occur ML::M. Radiation can result In the severing of atomic bonds and subsequent formation of new bonds B1

(crosslinking) between atoms in adjacont polymoric cholns as alli bo subsequently. Radiation may also result in described in detall another mechanism, referred to as scissioning, in which radiation causes the severance of atomic bonds in the main chain-of the polymer but crosslinking does not occur. 1 i

The relative rate (por unit of energy absorbed) at which a particular A #G value" is a process occurs is characterized by a "G-value".

description of the number of events of a given type that occur when

= 2 Implies 100 eV of energy is deposited. For example, G(scission) l occur for each that two scission events (two atomic bonds are severed) 100 eV absorbed.

For crosslinking of polymer molecule repeat units, which are already incorporated in molecular chains, Cf %U) can be units which have been defined as the number of polymer repeat crosslinked to others for each 100 eV 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.

I it has been noted (B-1.e-21 that polymers can be classified into two groups according to which radiation damage mechanism is predominant o Crosslinking polymers have chains linked following Irradiation.

This leads to higher molecular weight and/or branching, and eventually to formation of an insoluble network.

e Scissioning polymers have chains which are broken during exposure to radiation. This results in a decrease in the average molecular weight of the polymer.

I Both processes can occur simultaneously In many polymers, and the classification of polymer type depends on the predominant process.

Table B-1 presents some examples. It is believed that crosslinking is the predominant mechanism in dimethyl polysiloxane et low doses of radiation exposure in vacuum fB-1.e-21 The experimental evidence at high dose levels, crosslinking tends to DL-21 suggests that saturate and scissioning may be the predominant mechanism.

B-2

Table 8-1 SOME EXAMPLES OF POLYMER CLASSIFICATION Crosslinking Types Degrading Types Polystyrene Polymethyl methacrylate Polyethylene Polyvinylchloride Polypropylone Polytetrafluoroethylene Nylon Poly-alpha-methylstyrene Natural Rubber Polyisobutylene Polydimethylsiloxane Polyesters Source: Adapted from A. Chapiro, Radtm+ tan cha=f=+rw af - Patvaarte 19<+===, New Yorks interscience Publishers, Inc., 1962.

Table B-2 COMPOSITION (IN PERCENT) 0F EVOLVED GASES FROM POLYSILOXANE POLYMERS DURING 1RRADIAT10N Polymer H CH 4 CN 0H Investigator 2 26 66 Dimethylsiloxane 41 47 12 -- Charlesby Octamethylcyclotretrast-loxane 34 60.5 4.5 --

Warwich Octamethylcyclotretrast-loxane 27 57 16 -- Kantor Dlmethylsiloxane 31 47 22 --

St. Pierre et.al.

Methylphenylsiloxane 31 28 8 --

Prober Source: Adapted from A. Chapiro, Radtm+ tan chanta+rw av pniv.. rte Evn+ama, New Yorks interscience Publishers, Inc., 1962.

I i

B3

l l

Presence of molecular oxygen and other chemical agents tends to l l

encourage scission as compared to crosslinking in many polymers.

0xygen is believed to combine with radicals formed in the polymer due to chain scission or evolution (abstraction) of hydrogen or other 1 gases. In order f or oxygen to have a significant ef f ect, there must be an adequate supply and oxygen must diffuse into the polymer at a The rate compareble to its rate of reaction within the polymer. l offact of Irradiation environment on PDMS is discussed subsequently. l EFFECTS IN POLYSILOXANE: A REVIEW OF THE LITERATURE CrottlinkInn and G-Valuan Polyslloxenes were among some of the first polymers studied by Sissman and Bopp (B-3.B-4) and ot her s 12-11 w it h regard to determining the behavior of physical properties of materials subjected to nuclear radiation, in the 1950's and 1960's they were further studied by a j

number of Investigators (B-5--B-11), a primary goal of this work being i

to develop Information about the crosslinking process.

l As in the case of Boraflex, a characteristic of the redlolysis of the organic silicone polymers Investigated is that gases are evolved ,

during radiation. In a summary of data from several sources (B-12) It .

l has been noted that when silicones are Irradiated in the absence A of 02 large (In vacuum), the evolved gases contain 30 to 405 hydrogen.

fraction of the gas is methane, and in addition, ethane, etc., is present in appreciable amounts.

Irradiation of polymethylphenyl-slloxane and octamethylcyclotetrasiloxano yleids gases which differ from those of PDMS. The composition of evolved gases from these l various polysiloxane is shown in Table B-2.

1 i

These results lead to the conclusion that both C-H and SI-C bonds were i fractured leading to the formation of hydrogen atoms and methyl and phenyl radicals. Accordingly, it has been noted fB-12) that at least two primary radiation events must be assumed to take place as shown in Figure B-1. Subsequently, two primary crosslink bonds may be formed B-4 I

CH'3 + Sl*- o  ;;1 l CH 3 l CH 3 l

==e Si - o l CH2*

CH 3 l H' + SI - O ll2; Polydimethylsiloxane --

l CH3 Figure B-1. Primary Gamma-Induced Radiation Events in Polydimethylsiloxane.

CH 3 CH 3 CH 3 CH 3 I I I l SI - O - St Si o - Si l Or CH i

CH 2 (I

.. Sl O l

SI ...

I I I l CH 2 CH 3 CH 3 CH 3 l l

.SI St.

l l CH 3 CH 3 Figure B-2. Subsequent Crosslink Bond Formation.

B-5

es shown in Figure B-2.

O Several early studies have been reported in which the G-value for crosslinking was determined experimentally (B-5.B-7.R-8.a-13.9-14),

in this work, the behavior of PDMS at low radiation doses was investigated over a range of molecular weight from IIquids to gets to elastomers. Dolldes and Shepherd in a somewhat more recent publication (R-151 have noted that this early work could be summarized r as follows:

l e For PDMS, the degree of crosslinking is proportional to the radiation doso, at least up to a reasonably high dose.

e The degree of crosslinking is independent of the molecular weight.

e The degree of crosslinking depends primarily on the ,

dose received and is Independent of both the type and Intensity of radiation.

e Some discrepancies remained. For Instance, G(XL) <

varied from 1.6 to 4.5 (R-7.R-16.B-1.71.

(h Delldes and Shepherd (B-15.R-18) subsequently Investigated Irradlated PDMS as a function of dose up to 500 Mrad and founds gamme e The crosslink density dependence on dose was linear to about 160 Mrad and exponential for higher doses, tending to a constant value for extremely high doses.

i (See Figure B-3).

l e For doses less than 160 Mrad, they estimated.a G(XL) of 2.8 +/- 0.3 which was in good agreement with some i of the other earlier work.

e At 500 Mrad, they estimate G(XL) = 0.6 to 0.7 and as-cribe the decrease from the value of 2.8 to hindrance of transnational motion of the polymer chains due to crosslinks. They also note that there is a decrease ,

in the number of monomer units available to produce radicals.

e Evolved gas measurements also suggest a saturation effect as the doses become very large.

e it has been noted that the presence of oxygen is known to alter the results. Also, gases produced during Irra-dietion can be trapped in the polymer and must be re-moved for accurate analysis. Gases in crosslinked spe-A l

U B6 l

6 m g

l l I I l j f g+

f14 - -

14 3 g2 -

Evolved Gases -

12 I 10 - -

10 $

Crosslink Density 8

j 6 -

G Values 6}-3f w j j 4 4 o-2l g .

E2 - ~

2 5-1>

[l 50 I

100 e I 200

. I 300 e I 400 500 I

f i Radiation Dose (Mrad)*

Figure B-3. Crosslink Density, Evolved Gases and G-Values versus Gamma Radiation Exposure for PDNS.

(Sources C. G. Delldes and I. W. Sheperd. " Dose Effacts in the Crosslinking of irrediated Polysi-l o x a n e", E.aAlatl.QA Phvmica g chamts+rv, vol. 10, 1977, pp. 379-385.)

I I I l l h 6 - -

G-

!5 - -

$4 _

/ _

2 3 - -

3 - 520,000 l S 2 ', O - 890,000 Fraction _

g - I - 439,000 Fraction Ei 41 - 220,000 Fraction l O

0 ,9 I I I I l 0 10 20 30 40 50 60 Rad!ation Dose, r x 10-e.

Figure B-4. Elastic Modull of PDNS versus Electron Radle'-

tion Dose. (Source: A. N. Bueche. "An investigation of the Theory of Rubber Elasticity Using Irradiated Polydl-m e t h y l s i l ox a n e s ', Jau.c nal af Po l v ma r Eclanca, vol. 19, 1956, pp. 297-306.

B-7

cimens have been observed to produce bubbles in the polymer iB-18). .

O-Y A summary of the G values for crosslinking, G(XL), has been presented by Chaplro fR-12) which includes work of a number of Investigators. i Table B-3 summarizes the G-values for crosslinking as determined by l verlous techniques and with various radiation sources for PDNS and f other similar materials.

1 Referring again to G values, Charlesby ML11 point s out that G(XL) 0 varies with temperature and cites the example of G(XL) = 2.6 at -78 C up to G(XL) + 4 7 at + 150'C. Others (R-171 have observed the same effect in PDMS and in the copolymer dimethyldtphonyl polysiloxane. t i

Fffnet of Radiatten nn +he P e n n e r + 1 e n _, n f Pntvettnwnnen l

exposure to radiation changes the molecular structure and i As composition of polystloxanes, the physical properties of these i 1

materials also change. Of interest in the current study is the effect i

of redletion on the following properties:

e Specific Volume d

e Hardness o Elongation to break 1 6 Tensile Strength e Elastic Modulus A summary of the available literature relative to those properties with radiation exposure is discussed subsequently.

1 Specific volume, or alternatively, density, of polymers changes with Irradiation as summarized by Charlesby .LfL-11. In the case of crosslinking polymers, the distance between chelns Is shortened by the crosslinking bond. For instance, in polyethylene which crosslinks and which has been studied in some detall, the speelfic volume at 160 C decreases from about 1.29 cc/gm at 4 dose to 1.135 cc/gm at about 750 d

f B8

'9  ! .4 e .

Table B-3

SUMMARY

0F G(XL) FOR VARIOUS POLYSILOXANE Description G(XL) References Polydimethylsiloxane infusibility 3.4 (B-14) 2.8 gel-point determination (B-14) j elastic modulus 3.1 (B-14) swelling & elastic modulus 2.4--2.7 (B-5,8-7)

'M by cryoscopy 2.5 (B-8,B-25) gSI-point determination 2.2--2.5 Barnes, et. al.

gol-point determinations In vacuo 2.8 Okamura in air 2.5 Okamura gel-point determinations in vacuo 2.1 Davison in air 1.1 Davison Octamethylcyclotetrasiloxanes determination of dimer 2.0 Kantor Polysethylphenylsiloxanes i M, by cryoscopy 0.8 Prober eFor reference with no number, refer to .Chapiro 1B-121 , p. 480. l l

4 4

B9

Mrad samme equivalent. This is a decrease of about 125. The rate of decrease is larger at lower doses. The contraction in dimensions is A attributed to the formation of crosslinks between carbon atoms which results in a decrease in the Intermolecular distance from about 4A to 1.54A at the crosslink site.

Since polystloxenes have been shown to crosslink, it is to be expected that the specific volume in this case will also decrease with dose, rather rapidly at low dose with an ultimate saturation as the density of crosslinks becomes greater. Decreases in dimensions with dose in which a shrinkage of 8 to 9% in thickness in a polysiloxane film Irredleted by 10-15 Key electrons has been noted fB-19). Lateral shrinkage was negligible, possibly because of edherence to a rigid substrate. Subsequent heat treatment caused further changes in dimensions. Roberts (B-19) also investigated the changes In sensitivity to Irradiation in some polystloxanes by changing the l organic groups. When phenyl groups were substituted for methyl groups, a dose 6 times greater was required to achieve the same radiation-Induced change in physical property.

A U

The effect of Irradiation on the mechanical properties of polystloxanes has been studied quite extensively fe-5.e-6.m-7.

B-8.B-20). In early studies, Harrington ( B - 9 . B - 12.1 investigated the offacts of Co-60 redletion on a wide range of the commercially available elastomers. In particular, he studied a series of Dow j Corning elastomers (manufacturer of base Boreflex polymer used by Bl5CO) which included dimethylpolysiloxane and various copolymers I nc l u d t r.g methyl / vinyl, methyl /phenyl and methyl / vinyl /phenyl. Both filled and unfilled polymers were included in the metrix of test specimens studied. ,

The change in hardness, elongation to break, tensile strength and color of the Dow Silastic meterials as a function of increasing gamma dose con be summarized as follows:

e Hardness: Shore A hardness increases with dose.

' B 10  ;

1 j

o Elongationato breaks Decreases significantly with dose.

e Tensile strengths initially most specimens show an in- 1

{

tresse In tensile strength at low dose. Subsequently, some showed further increase while others showed a de-crease over the range of doses studied.

o Embrittlement: At highgr doses, all specimens broke when subjected to a 180 bend test.

e Color changes: Some samples showed color changes when  !

subjected to gamma radiation. l l

The above observed material changes with gamma radiation exposure tend to confirm and verify the behavior observed with Boraflex as described in Sections 3 and 4 Elastic modull of polysiloxanos tend to increase linearly with dose as a determined by a number of Investigators (R-5.R-7.R-fdl, in the Case j

of Buecho's work 11-21, the dose range was 60 Mrad (electron i

irradiation), with deviations from linearity only at low doses where  !

ond effects apply as shown- In Figure B-4 These results are essociated with, or complimentary to, mechanical property studies of Irradiated filled-polymers where the crosslinking causes changes in tonslie strength, Shore A hardness and elongation-to-break as discussed previously.

Fnvfrommental Fffae+a As noted previously, the environment, namely the presence of molecular oxygen as well as other agents, (water vapor, peroxides, etc.) is an Important element which may influence the behavior of polysiloxane in o redletion field. In addition to chemical agents, temperature may be on important factor.

It should be noted that Irradiated polysiloxane are very different materials from unirradiated polysiloxane in that the former contain maoy free radicals. It is believed that many, if not most, of the polymers as defined as crosslinking types in Table B-1 would undergo degradations If sufficient free access of oxygen acre provided 18::2.1.

In order for oxygen to have a significant offect, it must diffuse into the polymer at a rate comparable to its B-11

v e

rate of reaction in the polymer. If diffusion is slow relative to the rate of radical formation, then the effects of oxygen are likely to be O confined initially to the surfaces and edges of the polyrer.

oxygen diffusion proceeds, the effects of oxygen would progress inward As from the surface. The rate of oxygen diffusion into the polymer would be expected to be a function of temperature, increasing as the temperature Increases.

The primary reaction of oxygen is thought to be (B-2):

R* + 0 ----> R0 2 2 where the radical R' has been formed by a chain scission or abstraction of a hydrogen atom. These Investigators did not determine l whether oxygen effects the rate of polymer degradation by accelerating scissioning or by retarding crosslinking 11L.21.

O Other agents have also been observed to accelerate chain scissioning in silicone rubber. Osthoff, Buecho and Grubb (B-211) have shown that some of these agents are water, carbon dioxide, alkaline l polymerization catalysts and peroxide residuos (Introduced during the i vulcanization process). These Investigators studied the effect of chemical stress relaxation in stressed semples of PDMS. Samples of l PDMS were Initially elongated and the load ( g r ams,) required to malotain constant elongation as a function of time was measured.

Experiments were conducted in verlous atmospheres and temperatures.

Figure B-5 Illustrates the effect of dry N ,2 N2 + water vapor, air and dry CO 2 on the load required to maintain constant elongation. The i

Introduction of water vapor at 50 hours5.787037e-4 days <br />0.0139 hours <br />8.267196e-5 weeks <br />1.9025e-5 months <br /> causes a rapid decrease in the required load. It was postulated (B-21) that this rapid decay in stress in too presence of water vapor Involves a scission of the siloxane bond with the subsequent formation of silanols: ,

1 I I

- $1 SI + HO 2

---

> 2 - $10H l i 1 O B 12

l l

i 1

j I

il l l l l l (N

- \ -

100 g - e'g -

80 I r 70 lI '-

e, -

60 50 M

Dry N *N ~

E e 40 - -

6 NgrH2 O g

c 30 -

k -

e

.3 Air N 20 - -

Dry CO2 Or'N l

'o -

i 10 i

20

. i 30

'VI 40

~

50 60 Time in Hours i FigureB-5.EffactofWater,AirengCarbonDioxide on Stress Relaxation of PDMS at 130 C. (Sources R.

C. Osthoff, A. M. Bueche and W. T. Grubb. " Chemical Stress-Relaxation of Polydimethylsiloxane Elastomers",

Journal Amartean chnetent secte+v, vol. 76, 1954, pp. 4659-4663.)

a 1

B-13 r

Figure B-5 also Indicated that air and dry CO2 can also accelerato stress relaxation in PDNS. Contrary to the previous study cited, pure dry oxygen was found to have no of f act (stress relaxation curve was the same as for dry N 2 ). Similar offects have been observed in other I rrad i at ed pol ymer s ( B-22--D-24 ) . In these cases the effect has been termed oxidative scission.

The effect of residual agents from chemical vulcanization was also ,

studied.

The stress relaxation of PDMS crosslinked with 1.655 benzoyl peroxide and by 800 key electron radiation are compared In Figure B-6.

The p-esence of water vapor appears to have a strong influence on in l stress decay in the chemically crosslinked material but no effect j

the sample crosslinked with radiation. The crosslink density for the l

I two samples is not given. If the crosslink' density of the Irradiated sample were signifIcantly greater than the chemically crosslinked sample, the effect of chemical stress relaxation might not be evident due to a large number of tio points between chains.

The effect of air on silicone and other elastomers Irradiated under was studied by Harrington 18 ,21 In these stressed conditions '

experiments, tensile samples of silicone rubber and other materials were restrained at the ends and were subjected to gamma radiation. '

The samples Some samples were irradiated in air and others in vacuum.

trradlated in air to a dose of 5 x 10 roentgen developed cracks and 7

in some cases the samples broke. The companion samples Irradiated in vacuum did not have cracks. It was postulated 1EL-21 that ozone formed by the redlolysis of air was the prime agent causing the cracks in the stressed elastomers.

This mechanism for crack initleilon appears analogous to stress l For example, corrosion cracking which can occur in some alloys.

oxygen attack of stainless steel at the grain bounderles can cause crack initiation and growth at stress levels well below the tensile ,

strength of the material. The effect of radiation environment on stressed Boreflex and mechanisms of gap formation is discussed in Section 5 cf this report.

B 14

1 I

1 s s e e a 4 00 Dry N, 200 -

'No Crosslinked with 1.65%

= Benzoyl Peroxide N2/H2 O E '

o i e

6 Dry N2 and N 21H2O 8 +O0 O=O OM

" Crosslinked with High Energy Ciectrons

. e 100 a a i 0 10 20 30 40 50 60 70 Time in Hours Figure B-6. Stress-Relaxation of PDMS Crosslinked with Benzoyl Peroxide and Electron Radiation.

(Source: R. C. Osthoff, A. M. Busche and W. T. Grubb.

" Chemical Stress-Relaxation of Polydimethylsiloxane Elastomers", Journal American Chemical Eneta+v, vol.

76, 1954, pp. 4659-4663.)

B-15 l

Environmental effects in elastomers may continue for some period after they are removed from the radiation field since lens and electrons created as a result of the radiation persist for considerable periods ML.21. It has been noted that gamma radiation causes the creation of a variety of free radicals in the polymer. The diffusion of chemical agents into 1he polymer and the combining with free radicals would be expected to continue after the gamma field is removed.

O .

i i

i O B-16 o

O

1 l

l REFERENCES FOR APPENDIX B B-1. A. Charlesby, A+6mte Radin+ ten and Polvmara, London: Pergamon Press, 1960.

B-2. F. A. Bovey, Effac+a of tentrtne nadtatten en Natural and Synthatle Palvmara, New Yorks interscience Publishers Inc.,

1958. 4 B-3. C. D. Bopp and O. Sissman, ' Radiation Stability of Plastics and Elastomers", wuetaanten, vol. 13, 1955, p. 28 D-4 C. D. Bopp and O. Sissman, 'How Radletion Changes Polymer Nechanical Properties", Nuelannten, Vol.13, 1955, pp. 51-55.

]

B-5. E. L. Warwich, ' Effects of Radiation on Organopolysiloxanes", t' indum+rtal and Fnetnaartne chanta+rv, Vol. 47, 1955, pp. 2388-2393'.

B-6. E. L. Warwich , W. A. Plecoll, F. O. Stark, " Melt Viscosities i of Dimethylpolysiloxane", Journal of Amartean chamleal Eneta+v, Vol. 72, 1955, p. 5017.

B-7. A. M. Buecho, "An investigation of the Theory of Rubber Elasti-city Using feradiated Polydimethylsiloxane", Journal of Petv-mar setanea, Vol. 19, 1956, p. 297.

B-8. L. E. St. Pierre, H. A. Dewhurst, A. M. Bueche, " Swelling and Elasticity of Irradiated Polydimethyl Siloxane*, Journal of Palvmar Relanca, Vol. 36, (1959) pp. 105-111.

B-9. R. Harrington, ' Elastomers for Use in Radiation Fields", Rubbar Asa, Vol. 81, 1957, pp. 924-980. ,

l B-10. R. Harrington, " Elastomers for Use in Radiation Fields, Part  !

l II Effect of Gamma. Radiation on Heat Resistance Elastomers",

Ruhhar Aga, Vol. 82, 1957, pp 461-470.

B-11. A. N.. Bueche, "The Curing of Sillcone Rubber with Benzoyl Peroxide", Journal of Palvmar tetanea, Vol. 15, 1955, pp.

105-120.

B-12. A. Chapiro, Radta+1en chamtm+rw of Petvmarte Eva+ama, New Yorks interscience Publishers Inc., 1962.

B-13. A. Charlesby, "Effect of Molecular Wolght on Crosslinking of Silicones by High-Energy Radiation", Natura, Vol. 173, 1954, pp. 679-680.

B-14 A. Charlesby, " Changes in Silicone Polymeric Fluids Due to High-Energy Radiation", Peneandtnen of the Rnval teete+v, Vol.

A230, 1955, p. 120.

B-15. G. G. Delldes and I. W. Shepherd, " Dose Effects in the -

Crosslinking of irradiated Polysiloxane , nRadta+ ten Phvaten and chamlatev, Vol. 10, 1977, pp. 378-385.

B 17

- r B-16. M. Kloke and A. Dunno, " Radiation Effects on Dimethyl-dipnenyt Siloxane Copolymer. I Protective Effect of Phenyl Radical on Crosslinking", Journal of +he Phvalcal tacta+v af Jagan, Vol.

15. 1960, pp. 1501-1508.

B-17. M. Kloke, ' Radiation Effect in Dimethyl-Diphenyl $lloxane Copolymer. II Effect of Temperature on Crosslinking During irradiation", Journal of the Phvalcal teetatv af Janan, Vol.

18, 1963, pp. 387-396.

B-18. G. G. Delldes and I. W. Shepherd, ' Crystallization In Polydi-methyl Siloxane Networks Formed by Gamma Radiation", Palvear, Vol. 18, 1977. pp 97-98.

B-19. E. D. Roberts, "The Preparation and Properties of a Polysiloxane Electron Resist,a Journet of +he reme+rnehamicai Koeta+v, Vol. 73, 1973, pp. 1716-1721.

B-20. R. K. Traeger and T. T. Castonguay, 'Effect of Gemma Radiation on the Dynamic Nechanical Properties of $llicone Rubbers",

Journal of Annited Potvmar setanea, Vol. 10, 1966, pp. 535-551.

B-21. R. C. Osthoff, A. M. Buecho and T. C. Grubb, "Chemica!-Stress Relaxation of Polydimethylsiloxane Elastomers", Journat of +ha Amartean chemical Caeta+v, Vol. 76, 1954, pp. 4659-4663.

B-22. R. L. Clough and K. T. Gillen, " Radiation-Thermal Degradation I i

of PE and PVCS Nechanism of Synergies and Dose Rate Effects", ,

Radlm+1on Phvalca and chaelafrv, Vol. 18, 1981, pp. 661-662. l l

O B-23. R. L. Clough and K. T. Gillen, " Occurrence and implication of Radiation Dose-Rate Effects for Material Aging Studies", ERA 1A-

+ Ion Phvalen and chamta+rv, Vol. 18, 1981, pp. 679-687. 1 i

B-24. R. L. Clough and K. T. Gillen, " investigation of Cable l Deterioration inside Reactor Containment", Nuclear Technetenv, Vol. 59, 1982, pp. 344-354  ;

B-25. F. W. Billmeyer, Tav+heek of Folvmar Eclanca, New Yorks inter-science Publishers, 1982. l i

I 8-18

APPENDIX C GUIDELINES FOR A STANDARD BORAFLEX COUPON SURVEILLANCE PROGRAN

CONTEhTS Kact1en M 1.0 PURPOSE AND SCOPE C-1 3.0 SURVEILLANCE PROGRAM OVERVIEW C-2 3.0 COUPON PREPARATION C-3 3.1 Scope C-3 3.7 Coupon Selection C-3 3.7.1 Standard Coupons C-3 3.2.7 Special Coupons C-3 3.3 Coupon Size and Orientation C-4 3.4 Coupon Encapsulation C-4 3.5 Coupon Installation C-5 .

3.6 Related Documents C-6 3.6.1 ASTN Standerds C-6 4.0 SURVEILLANCE FREQUENCY C-6 4.1 Scope C-6 4.2 Standard Coupons C-7 4.3 Special Coupons C-7 4.4 Verlations in Sampling Frequency . C-7 9.0 TESTS FOR STANDARD AND SPECIAL COUPONS C-8 6.0 COUPON CHARACTERIZATION C-6 6.1 Scope C-C 6.2 Coupon Conditioning C-6 6.3 Visual Inspection C-9 6.4 Dimensions C-9 6.5 Coupon Weight C-9 6.6 Shore A Hardness C-12 C-iit

CONTENTS (continued)

O M e e t I c,n g 6.7 Radioassay C-12 6.8 Specific Gravity and Coupon Volume by Immersion C-12 6.9 Boron Determination C-14 6.10 Releted Documents C-14 6.10.1' ASTN Standards C-14 7.0 ACCEPTANCE CRITERIA C-16 8.0 ENVIRONMENTAL CONDITIONS C-17 8.1 Epent Fuel Pool Water Chemistry C-17 C-17 I 8.2 Fuel Placement and Gamma Exposure l

l l

< 1 Q i 4

i l

i i

i i

C-iv

O Appendix C GulDELINES FOR A STANDARD BORAFLEX COUPON SURVEILLANCE PROGRAM Nottem The following are Intended to be used as guidelines in establishing Boreflex Coupon Surveillance Programs for Individual fuel rock Installations. As such, they are not Intended to supplant elements of programs recommended by the manuf acturer of either the fuel storage rocks or Boraflex. The Intent, rather, is to serve as a supplement to tests and procedures as recommended by the Boraflex or fuel rack supplier. Accordingly, only specific elements of the Guldelines may be selected for implementation at a particular facility.

1.0 PURPOSE AND SCOPE These guidelines have been developed to provide the utility with a basis for specifying recommended elements of a Boraflex Coupon Program. While these guidellnes are Intended principally for plants initiating new surveillance programs, there may be elements which can be used to benefit existing programs. The guidelines have as their objectives a standardized coupon surveillance program which wills e Provide data in sufficient quantity and of sufficient quellty (pre and post Irradiation characteristics) to yloid accurate and meaningful information on the per-formance of Boraflex In the spent fuel pool environment, o Provide coupon data which can be readlly extrapolated to tracking and projecting the performance of Bora-flex in the spent fuel racks.

e Provide Industry-wide date which is uniform In quality and accuracy.

e Be applicable to both PWR and BWR pools with racks utt-Itzing Boraflex for criticality control.

To meet these objectives the following sections of this document i address methods for coupon material selection, coupon configuration  ;

and quantity, frequency of surveillance and methods for pre and post Irradiation of characterization of specific physical attributes.

C1

Various ASTM Standards for the measurement of the specific physical attributes have been referenced that may be useful in developing techniques for obtalning coupon date of uniform quellty and sufficient accuracy. In some cases the ASTM Standards will not be practical to apply exactly as prescribed and may require adaptation to Boreflex. However, the standards can be useful In understanding the Intent of the measurement and the sensitivity to measurement variables. Since the spent fuel pool water condition may be an important element in the overall service life of this material, recommendations for monitoring pool water chemistry and feeperature are also provided.

2.0 SURVEILLANCE PROGRAM OVERVIEW While a coupon program consisting of many large coupons scheduled for f frequent inspection can provide useful data, it is recognized that j certain limitations exist with respect to coupon size, number and  ;

frequency of surveillance. For this reason the program outlined here  ;

utilizes two types or sets ,of coupons. The first type consists of j coupons characterized with respect' to certain physical attributes  !

C which can be readily examined in the station chemistry laboratory. j These coupons, termed the standard coupons, are designed for periodic '

removal, inspection and, if feasible, are reinserted in the pool for  ; l subsequent exposure. The recommended inspections and measurements are  !

l nondestructive and have been selected to provide an Indication of the i general condition of Boraflex and any Indication of gross or unusual  ; J degradation. The standard coupons are scheduled, for frequent f

Inspection, typically on a schedule dependent on refuelings. It is l l recognized that in certain situations that reinsertion of coupons may l l l

prove impractical and each reactor program will have to be adjusted as necessary to meet any local constraints.

l 1

The second type or set of coupons will be procharacterized with respect to the same set of physical attributes as the standard coupons but will be scheduled for more extensive post Irradiation examination.

These coupons, termed the special coupon set, will probably have to be sent to an Independent laboratory for the post Irradiation O C-2 l

examination. Since the post Irradiation tests of the special coupons may involve certain destructive testing, these coupons are not scheduled for reinsertion in the pool. The frequency for removal and Inspection of coupons from the special set is less frequent than for I the standard coupons, typically once every five years.

3.0 COUPON PREPARATION 3.1 Ir.nna This section provides recommendations for sampling production lots of Boraflex and coupon material selection. Coupon configuration j and methods of encapsulation are also addressed.

3.2 Counen Enlac+ fen j

)

3.2.1 S+mndard counena-Approximately 20 samples to serve a3 standard coupons should be ~

selected. For the purposes of selecticq, the production lots from which the coupons are selected should be randomly identlfled and no two samples should be taken from the samt production lot. Coupon unterlal from the same production lot should be selected randomly consistent with BISCO's quality assurance procedures. In addition, one archive sample should be randomly selected from each production lot and retained. Each coupon and sample should to identitled as to l

the production lot from which it was taken and Ossigned a unique identification number. Serialization documentat ion for full traceability from raw material to installed post?lon in the fuel racks should be retained. Material certification of Isotopic necifications and chemical compositions for each production lot should b1 retained.

3.2.2 snacfal connene.

Approximately 10 coupons and archive samples should be selected as in paragraph 3.2.1 to serve as special coupons. Each coupon should be C-3

=

clearly identified as to its production lot and assigned a unique identification number. Documentation and certification specified

() under 3.2.1 also appIles.

3.3 Onunen Siva and Orlantatten Coupons should be of certain minimum dimensions according to the dimensions of Boraflex used in the spent fuel storage racks. Coupon l

width should be equal to the nominal width of the Boreflex sheet used in the fuel racks. Coupon length should be approximately twice its width.

i A small hole should be punched in one corner of the coupon. The hole will serve to orient the coupon and also serve as a convenient means to suspend the coupons for specific gravity measurements. The short edge of the coupon nearest the hole is considered the top of the coupon. The long edge nearest the hole is considered the right edge. '

With the coupon oriented so that the hole is In its upper right hand I

corner, the facing surface is consider 9d the front of the coupons the ,

(h oppostte surface, the back of the cou'pon. The edges of the coupon cut during manufacture should be Identifled and recorded. The edges which have been cut during coupon preparation should be similarly identified l and recorded. <

3.4 Counnn Encannula+1en i i

Each coupon should be encapsulated in a clad material Identical to l that used to capture and retain the Boraflex In the spent fuel storage <

racks. For many applications, this material is stainless steel. Vent holes should be provided to allow the escape of offges produced when  ;

Boraflex is Irradiated. The vent holes should allow the pool water to  !

enter the capsule and contact the Boraflex. The size and location of vents should be selected so as to simulate, as closely as possible, .,

the water exchange rate between the Interior of the capsule and the l pool as well as local flow conditions which prevall in the spent fuel  !

storage racks. Since different means for venting are used in various I

C-4 l o

m___ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ .

fuel rack designs, each will have to be evaluated separately.

Means should be provided to open or close the capsule (for the purposes of Boraflex removal) without causing damage to the coupon.

Such means may include, but need not be limited to, mechanical fasteners such as machine screws. Welding and crimping are to be svolded since these would generally require cutting the clad and the potential for mechanical damage to the Boraflex.

Each coupon should be placed in a capsule criented such that the top I edge of the Boreflex corresponds to the top edge of the capsule with the small hole located in a position corresponding to the upper right hand corner. When the capsules are suspended in the pool, the top edge of each Boraflex coupon should be oriented toward the pool surface. One flat surface of each capsule cladding should be clearly marked with the unique coupon Identification number. The marking j

should be accomplished by stamping, etching or welding in letters 1 and/or numerals at least 1/2" high. r i

3.5 cannnn In=+= tim +tna l

The set of standard coupons and set of special coupons should be affixed to separate trains for placement In different locations In the spent fuel pool. The maximum length of the trains should be restricted to approximately one half of the active fuel f.ength of fuel stored in the pool. The elevation of ths coupon trains in the pool ,

should be selected so that the midpiano of the trains corresponds to l the midplane of the active fuel stored adjacent to the trains. This I is to assure that' all coupons In a given train receive uniform gamma exposure.

{

i i

The rack location of the coupon train containing the standard coupons j should be selected so that freshly discharged fuel is stored on all  !

four adjacent sides of the train after each refueling outage. In this i manner, the standard coupons will receive accelerated gamma exposure i

C-5 1

relative to the overage exposure of the storage racks.

O The rock location for the special coupons should be selected so that gamma exposure is accumulated at a rate corresponding to that of an average storage cell In the racks.

Specific requirements for the coupon train will vary from pool to pool and the specific design details should be developed by the utility and coupon supplier. Similarly, refueling procedures very from utility to utility so that the details of coupon train placement must be developed on a site specific basis.

3.6 nata +ad noeuman+m I

3.6.1 AETH E+andards, e D 1898-68 Practice for Sampling of Plastics )

Q i i

4.0 SURVEll. LANCE FREQUENCY 4 .1 insta This section provides guidelines for establishing schedules for coupon  ;

the E and One schedule is recommended for removal surveillance.

standard coupon set and another, less frequent schedule, for the special coupons. Provision is made for more frequent sampling and more detalled examination of the standard coupons should Indication of I degradation be observed. The recommended unusual Boraflex survolliance frequencies are based on the assumption that the standard coupons will be subjected to accelerated gamme exposure from freshly discharged fuel assemblies after each refueling outage. This will f assure that the coupons will lead the average rock exposure by a substantial margin.

O I

C-6 4

a

4.2 S+andard enunent

{

The schedule for removal of the standard coupons should be keyed to the cumulative gemma exposure received by the coupons. Since the gamma exposure is to be accelerated by placing freshly discharged fuel adjacent to the coupons after each refueling, the following schedule is recommended. After initial coupon placement, a minimum of one standard coupon should be removed just prior to the next ref ueling ourage. Coupon removal should be repeated in a steller manner just prior to the next two ref ueling outages. Thereafter, the frequency H for coupon removal and Inspection should be just prior to every other i refueling. After Inspection per Sections 5 and 6, the coupons are to be reassembled and returned to the spent fuel pool.

l 4,3 Enaelat counont One coupon each should be removed and tested every five years  ;

following coupon installation. The special coupons are not to be l returned to the pool. I 4.4 Vnefattent in Samnifna Franuanew Should inspections of standard coupons Indicate unusual degradation of the Soraflex, increased sampling frequency and Inspection should be considered. For these purposes, unusual degradation of the Boraflex is defined as any observable or measurable change in the, condition of the material which would result in loss of the boron carbide.

Examples of such changes include:

o Excessive thinning of the material particularly along the edges.

e Changes in coupon length or width in excess of those .

expected from normal shrinkage of Boraflex. I o A significant reduction in coupon weight.

e Softening of the Boreflex.

o Bilstering, cracking or spelling of the surface  !

of the coupon.  !

I C-7 i

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ -._____.____m 1

. l j

In such case, the standard coupon should be inspected as per the ,

testing requirements for the special coupons. Results of these tests 1 1

O and measurements should be compared to the Acceptance Criteria, guidelines for which are provided in Section 7.0.

5.0 TESTS FOR STANDARD AND SPECIAL COUPONS l

All coupons should be procharacterized with respect to those physical in Sections 6.3 through 6.9, and chemical attributes specified exclusive of 6.7. After removal from the pool, the standard coupons are to be characterized with respect to those attributes outlined in i Sections 6.3 through 6.7. The special coupons are to be subjected to testing per Sections 6.3 through 6.9.

1 i

)

6.0 COUPON CHARACTER 12AT10N q

6.1 Scasta This section provides guidelines as to which physical and chemical attributes are to be characterized pre and post Irradiation.

Appropriate ASTM Standards have been cited which provide some guldance )

In developing measurement procedures. Because the properties of Irradiated Boraflex may differ from the materials addressed by the Standards, the procedures may require adaptation for application to 1 Irradiated Boraflex. ,

i i  !

]

6.2 cauenn candI+ ten 1no f l

For procharacterization of as-produced Soraflex, the coupon should be l l

conditioned at standard laboratory temperature and standard laboratory j atmosphere. The purpose of sample conditioning is to bring the material into equilibrium with normal average room conditions and to obtain reproducible measurement results.

i O m.

- - - - - - - - - - - - _ - - - - - - - - - - - - - - - - - - -- a

For coupons which have been exposed to the spent fuel pool water, the coupons should free of surface moisture and conditioned at standard laboratory temperature in a desiccator for 48 hours5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br />. The coupon should be then conditioned to the laboratory atmosphere.

6.3 vfnumf inanne+ fen The general overall appearance of the coupon should be noted and recorded. Noticeable qualitative changes in surfaces, outline and general appearance should be recorded. These changes include color, l surface Irregularities, edge erosion, cracks, etc.. Changes should also be noted as they occur, especially those which alter the shape so that Intended dimensions are no longer significant. A color j photograph, of sufficient resolution to distinguish areas of different I color and shades, should be taken and retained.

1 l

l l 6.4 Diman<fane l

l Coupon thickness should be determined. The thickness of each coupon should be determined at nine locations es shown in Figure C-1 Coupon length and width should be determined at three locations each as shown in Figure C-2. Measurements of length and width should be made using a method providing an overall reproducibility of +/- 0.50% or less.

l Care must be exercised in measuring to account for ease erosion that may have developed in the coupons.

6.5 cannnn wmfnh+

Coupon weight should be determined using an Instrument capable of weighing accurately to 0.1% of the coupon weight. Coupons which have been exposed to the pool water should be weighed before and after conditioning. The pre and post conditioning welght should be recorded.

C-9

O g l

, O O i

1 r

4 1

1 4 I O 1 i

li U2 i

I

-3 '

I 1 in.

I.  ! . i v.

I l .11n.j i r I I. : Wl2 - r. "

)

Figure C-1. Locations for Coupon Thickness Measurements.

I 1

C-10

l l

- l 1

i 1

'e I

Wl2 r' i

$1 f I f

.! i J6 d6 O -

2 in. .

l W

if l l

l U2 1 1

l I

l

__!' . y e 1

4 l

l l

l

?

Figure C-2.

t tt Locations for Coupon Length and Width Measurements.

l l C 11 1

1 1

1 _ _ _ _ - .-__

y -

6.6 shern A Hardnatt Shore l Shore A hardness should be determined using a Type A durometer.

f A measurements are to be taken at the locations as shown In Figure #'

1 C-3. Shore D hardness measurements are not advisable since The the use of  ;

s Shore D durometer may result in destruction of the coupon. the l Intended to provide a rough estimate of Shore A measurements are f gamma exposure and to detect any gross softening of the material. ,l Coupons of sufficient thickness or layers of suf ficient thickness, as described in Section 3.0 of this report, should be used.

6.7 Radteattav l l

In order to provide an Indication of the extent of water permeation into the coupons, radionssay of the surface of the coupon for beta and Radioassey should be performed gamma radiation should be perf ormed. sensitive to beta and gamma with a suitable detector having a window radiation and of dimensions not larger than one sixth the width of the are coupon. Measurements of the activity on contact with the coupon to be made midway between the top and bottom edges as follows:

e On the right and left edges of the coupon. I e One quarter of the distance In from the right and left edges.

e Midway between the right and left edges. I The beta and gamma acilvities (Mr/hr) as well as the gross activity as  ;

a function of detector location are to be recorded and retained.

6.8 pnactfte cravt+v and counen veteem bv immareten (nemetal eeueens M

Specif ic gravity and coupon volume by lamersion should be determined For the purposes of computing the volume of by a displacement method. the Immersion, the volume should be defined as the coupon by reciprocal of the density multiplied by the sample weight.

l C-12 k

O l

t

- _ 7-l l

1 t>

1 1

1 i i

l U2 I

i __ ___

n .

g l

I I

i 2 in. -

1 I

I 1 i, e o i 1 8

i I

1 i i ln.1 1

'= '

I W/2 -:

Figure C-3. Locations for Shore A Hardness Measurements.

Ik C-13 -

l i

e  !

6.9 Rnenn Dm+meminm41en finactal enunant nnfwi Boron verification measurements should be made via neutron attenuation ,

O measurements. The purpose of the neutron attenuation measurements is to verif y the unif ormity of B-10 loading at several locations on the coupon. Neutron attenuation measurements should be taken at three locations er shown in Figure C-4 and compared with measurements taken on a control coupon. if a significant decrease in neutron attenuation is measured relative to the control coupon, boron determination via chemical analysis should be completed. At the J location having the lowest measured neutron attenuation characteristics, a sample of suitable site should be removed from the should be subjected to analyses for filler j coupon. The sample I

composition (B4 C) and total boron content. The B-10 loading in grams, B-10/cm2 , should be computed and compared with the Acceptance Critoria l

In Section 7.0. l 6.10 Refm++d u nr uin a n + m -

N 6.10.1 ASTM K+andmedi.

While there are no ASTM Standards specifically for Boreflex, there I

exist several Standards for the general class of materials, plastics

}

and rubber. A list of related Standards which can provide some  ;

guidance In developing specific procedures for inspecting and test,Ing Boreflex coupons includes:

e D 1349-87 Standard Practice for Rubber--Standard Temp-eratures for Testing.

e D 616-61 Standard Method of Conditioning Plastics and Electrical Insulating Materials for Testing.

o D 756-78 Standard Practice for Determination of Weight and Shape Changes of Plastics under Accelerat-ed Service Condition.

e D 3767-84 Standard Pr'actice for Rubber-Measurement of Dimensions.

e D 1042-83 Standard Test Method for Linear Dimensional Changes of Plastics under Accelerated $srvice C-14

l

• [s l

x 3--  ;

i ln. O

.t _ _ _

U2 l

l l

l l

l l

l l

l I

.l _ _ . g.

1 ln.

!  ! _ _t. _

l 'diny i i 1 l= W/2 :j Figure C-4 Locations for Neutron Attenuation Hessurements.

C-15

Conditions.

/'Ng o D 792-66 Standard Methods of Test cf Specific Gravity y ,/

and Density by Displacement.

e D 2240-86 Stenderd Test Method for Rubber Property--

Durometer Hardness.

e D 297-81 Stenderd Methods for Rubber Products--Chem.-

cel Analysis, e D 8838-66b Stenderd Definitions of Terms Relating to Plastics.

e C 791-63 Stenderd Methods for Chemical, Mass Spectro-metric, and Spectrochemical Analysis of Nu-clear-Grade Boron Carbico.

7.0 ACCEPTANCE CRITERIA Evaluation of the coupon date described in Section 6.0 should be completed on a plant by plant basis. This is required since the criticality analysis for each fuel rack design is conducted on a plant speelfic basis. Each analysis contains implicit assumptions as to the nominal Boron-10 loading and the extent of verletion in B-10 loading,

/91 Boreflex width and thickness. As-produced verlations in B-10 loading, Boreflex width and thickness are provided by the manufacturer and their effect on the reactivliy of the fuel and rack is determined by a rocctivity calculations. The acceptance series of differential criteria are therefore determined by the range in variation of these quantitles assumed f or the criticality analysis. Coupon measurements siiou l d be compared with the following assumptions used in the criticality analysis e Minimum Boron-10 loading.

e Mirilmum sheet width.

e Minimum sheet length.

e Minimum sheet thickness .

The results of the cospon measurements are acceptable If the measured coupon dete exceed the minimum riange implicit in the criticality i

analysis.

I O U C-16

" - m l

8.0 ENVIRONMENTAL CONDITIONS 4 ,

8.1 Snant Fuel Pont Water Chamit+rv s

The spent fuel pool water chemistry should be monitored periodically and everage values of the following recorded.

e Water ph.

e Temperature, everage, maximum and minimum.

e Boric acid concentration (ppm Boron).

e Conductivity (onlon and cation).

e Activity levels (Isotopics).

8.2 Fuel Placement and Camma Fwnotura The location of the coupon train should be recorded as to position and date of residence at that position. In addition, the fuel assembly luoniification number of assemblies located in storage cells adjacent to the train should be recorded. Fuel management records should be available to provide at a a l n i cum' the following data for those assemblies located adjacent to the trains e Fuel assembly operating power exposure history prior to discharge.

o Reactor shutdown time and date. ,

o Time and data of fuel placement in the racks.

The above data should be available for all assemblies located adjacent to the coupen train.

l The Integrated gamma exposure to the coupons may be calculated by suitable means or measured. Sultable calculational methods include simulation of the fuel operating history and shutdown cooling history to determine the gamma photon source terms as function of cooling C-17 s_

time. A second photon transport calculation is then required to provide the gamma dose rate at the coupon location as e function of O fuel cooling time.

\

s l Alternatively, the gamma exposure may be calculated from a " point in time' measurement of the gamma dose rate at the coupon location l obtained with suitable survey Instrumentation. The dose rate versus i time may then be estimated using a standard prescription for the decay of fission product gamme activity.

i Irrespective of which method is selected, en estimate of the encertelnty in the Integrated dose should be determined.

l l

O C-18

_ _ _ _ _ . _ . _ . _ _ _ _ _ . _ _ _ _ . . _ _