ML20212J012

Forwards Response to 861223 Request for Addl Info Re 860520 Application for Amend to License R-67.Info Supports Designation of Triga Low Enriched U,Containing 30 Weight Percent U,As Std Triga Fuel

ML20212J012
Person / Time
Site:	General Atomics
Issue date:	01/22/1987
From:	Asmussen K GENERAL ATOMICS (FORMERLY GA TECHNOLOGIES, INC./GENER
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM), Office of Nuclear Reactor Regulation
References
67-1015, NUDOCS 8701280071
Download: ML20212J012 (17)

Text

o .. ..

GA Technotogies Inc.

PO. BOX 85608 SAN DIEGO. CAUFORNIA 92138 (619) 455 3000 January 22, 1987 67-1015 Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Washir:gton, D.C. 20555 Attention: Document Control Desk

Subject:

Docket No. 50-163: License No. R-67; Proposed Technical Specification Change - Response to Request for Additional Information (3 copies) ,

References:

(1) Asmussen, K. E. (GA) letter no. 67/9000 to office of Nuclear Regulatory Regulation, dated May 20, 1986.

(2) Asmussen, K. E. (GA) letter no. 67-9037 to Office of Nuclear Reactor Regulation, dated September 5, 1986.

Dear Sir:

GA Technologies Inc. (GA) made application in May 1986 (Ref. 1) for a license amendment for GA's TRIGA Mark F Reactor R-67. Supplemental information in support of our application was provided in September 1986 (Ref. 2). By NRC letter dated December 23, 1986, Dr. Robert E. Carter requested written responses to some additional questions which arose during NRC review of our application, The requested responses are enclosed herein. The enclosed information answers the above mentioned questions and continues to support the designation of TRIGA LEU fuel containing up to 30 wt. % uranium as a standard TRIGA fuel.

We trust that the material which has now been submitted is sufficient to support the issuance of the requested license amendment. If you should have questions regarding the enclosed answers, please contact me, Dr .

William Whittemore, or Mr. Gordon West at (619) 455-2823, (619) 455-3277, or (619) 455-2292 respectively.

Very truly yours, 8701280071 870122 Licensing, S fety and PDR ADOCK 05000163 P PDR ,

Nuclear Compliance KEA/mk

Enclosure:

" Answers to NRC Request for Additional Information, dated December 23, 1986," dated January 21, 1987 M cc: Dr. Robert E. Carter, NRC (2) 10955 JOHN JAY HOPKINS DR . SAN DIEGO. CAUFORNIA 92121

e .

2 STATE OF CALIFORNIA ) -

) ss COUNTY OF SAN DIEGO )

On this the 22nd day of January 1987, before me,2aew$/e &#res ,

the undersigned Notary Public, personally appeared Keith E. Asmussen, Manager, Licensing, Safety and Nuclear Compliance, proved to me on the basis of satisfactory evidence to be the person whose name is subscribed to the within instrument, and acknowledged that he executed it.

WITHNESS my hand and official seal.

~~:::::::::::::.::.

OFFICIAL SEAL LORRAINE .

caupota.4

@ noraar rusuc . ROGER $

. PNinGPAL OFFICs IN y un == counn . Notary's Signature Mr Comanssion Expiess Jan. 30,1988

i '

ENCLOSURE 1/21/ 87 GA Technologies Inc. ,

Docket No. 50-163

. LEU TRIGA Fuel' Answers to NRC Request for Additional Information Dated December 23, 1986

1. With the increased uranium loading in 20, 30 and 45 weight-percent LEU fuel, there might be an enhanced potential for formation of eutectic reactions between the uranium and components of the cladding, such as iron and nickel. Please discuss further any evidence of such reactions in your fuel tests, the potential .for such reactions under accident conditions that 4

could lead to high fuel and cladding temperatures, and the potential consequences to integrity of fuel cladding.

Answer:

The post-irradiation examinations (PIE) of TRIGA-LEU fuel from the high burndp tests in the ORR showed no evidence of formation of eutectics between the uranium and components of the cladding. However, GA has previously performed fuel quench tests with the LEU fuel which show that eutectics will not form and cause unexpected melting below about 105000.

The results of these tests are documented in Section 2.12 of General Atomic I Report E-117-833 (February 1980) which was a part of the submission for  ;

this License Amendment. Enclosed, as Appendices A thru F to this letter, are the phase diagrams for both uranium and zirconium with nickel, chromium and iron, the main constituents of the Incoloy 800 or stainless steel clad used on TRIGA fuel (or the Inconel 600 thermocouple sheath). )

l The quench tests consisted of heating fuel samples, with an Inconel 600 l sheathed thermocouple inserted, to temperatures between 800oC and 12000C and then quenching them in cold water. The results of the quench tests g --e.. , - - -.--m----,-. -n, , ,.-,,-,--,,,--n,-- , - . - , , - . , , _ - , . _ . , , , -

showed no melting due to eutectic formation at temperature below 105000. l At 105000, localized melting did occur as a result of eutectics formed between the uranium and the Inconel 600 (mainly nickel). The enclosed phase diagrams show that eutectics with melting points as low as 740oC and 72500 can be formed between uranium and nickel and uranium and iron. ,

l Eutectics with higher melting points, 96000 and 93400, are formed between zirconium and nickel and zirconium and iron. However, as the quench tests have shown, for the high, 45 wt-5 uranium loading in zirconium, the eutectic melting temperature is about 10500C. The composition is much too uranium rich to approach the minimum melting temperature of 7400C requiring about 10 wt-% nickel with uranium.

The combination of nickel + iron is nearly the same for Inconel 600 and the Incoloy 800 and 304 stainless steel clad materials used for TRIGA fuel. .

The amounta of iron or nickel needed to produce the minimum eutectic melting temperatures are also about the same (-10 wt-5). The minimum melting temperatures of 74000 and 72500 are nearly the same for eutectics of uranium and nickel or iron. And, the thickness of the thermocouple sheath was nearly the same as the clad thickness used on TRIGA fuel. (The quench test samples were unclad). Thus, the interaction of the Inconel 600 thermocouple sheath with the TRIGA fuel containing 45 wt-% uranium is an excellent demonstration of the interactions which could be expected between the fuel and the standard clad materials for TRIGA fuel. No eutectic formation and melting was observed in several tests below about 105000, i

It should also be pointed out that for longer heating times at lower l

temperatures, the eutectics with zirconium would form preferentially to those with uranium and these eutectics would have higher melting temperatures than the eutectics with uranium. The zirconium eutectics l would form preferentially because of the much greater volume of zirconium in the fuel mix'ture compared to uranium. The Zr/U ratio is about 3 5 to 4 i for 45 wt-5 U fuel and is even larger for the 30 and 20 wt-% fuel. It may j be noted that the quench tests were done only with the 45 wt-% U fuel.

However, fuels with lower U content, such as the 30 wt-5 U fuel applied for i

e

,-,,-r_.-.., . _ . . , - , , , , , . _

___,,,y _ . . - _ . . , _ , _ . _ _ - _ _ _ _ - , _ _ , , _

in the license amendment under consideration, would have less susceptibility to low melting point uranium eutectic formation because of the competition with zirconium.

From the quench test results, it is concluded that for conditions where the ,

clad temperature can approach the fuel temperature for several minutes or jonder, eutectics can form due to interaction between the fuel and clad and

~

melting of this eutectic occurs at about 105000. However, this is about 10000 higher than the temperature at which calculations show that internal hydrogen pressure can rupture the clad, should both clad and fuel be at the same temperature.

During pulsing operations, where a safety limit of 115000 exists for the fuel temperature, the clad is in contact with the water coolant and never reaches any temperatures at which eutectics could form with the fuel. Any fuel elements equipped with sheathed thermocouples would also not be susceptible to eutectic formation and melting at 10500C because the hottest fuel during a pulse is toward the outside diameter of the fuel element.

The fuel in the central part of the fuel element, where the thermocouple is located is at temperatures 2000C to 18000C lower than the peak fuel temperature.

r l

l

2. As fuel burn-up increases, chainges in morphology, such as porosity distribution, might change parameters in the fission product diffusion process. With sufficiently large quantities of fission products escaping into the fuel-clad gap, internal gas pressures at high temperature might -

i add significantly to the hydrogen-only pressure. Please discuss your measurements related to gaseous fission product escape frqa the fuel meat

} both during extensive burn-up and while the fuel is hot as a result of a transient, and implications for integrity of cladding at the usually accepted safety limit of 115000. Do the gamma scan burn-up results of fuel rod Number 1086 have any bearing on this issue?

Answer 4

The results of measurements of fission product release performed on the TRIGA-LEU fuel after extensive burnup in the ORR are summarized in Table i 8-2 of GA-A18599 (which is a part of the Application for License Amendment). They indicate an upper limit release fraction of about 2 x 10-3 and a best estimate value of about 4 x 10-4 The best estimate value correlates well with expected fission product release rates based on tests with unirradiated fuel and using the temperature distribution in the fuel during the test. In Table 1 are summarized the calculated accident ccadition pressures resulting from stable fission product gases which have l

beent released to the free volume within a fuel element during its entire lifetime assuming either the best estimate or the upper limit release fraction. The table shows results for the nominal 0.5 inch diameter fuel used in the ORR tests and the nominal 1.5 inch diameter fuel which is the topic of this application for License Amendment. As seen in Table 1, the pressure contribution from fission product gases is nearly the same in the

. 0.5 and 1.5 inch diameter elements and the pressures are never more than i

about 0.7 to 1 atmos. for the upper limit release fraction and 1/5 of those values (0.13 to 0.19 atmos) for the best estimate release fraction. At l most, this is less than a 15 perturbation on the 138 atmos pressure caused 1

! by the hydrogen at 11500C (fuel temperature safety limit when clad l

1 e

-4 l

-- - - . ~ - - - - .

TABLE 1 Pressure Contribution g Stable Fission Product Gases .

.gn Cladding g TRIGA Iggl Elements 0.D. (inch) 1.475 0.542 Wt-% U _ Enriched 30 - 20 45 - 20 Burnup (%) 50 60 Free volume inside clad (cc) 31.5(1) 8.5(2)

Weight U-235 (gm) 162 55 Stable Fission Product Gases Created (moles) 0.092 0.037 Fission Product (Best (Upper (Best (Upper Release Fraction

• Estimate) Limit) Estimate) Limit) 4x10-4 2x10-3 4x10-4 2x10-3 Pressure 6 11500C (atmos) 0.13 0.67 0.19 0.96 Pressure 6 9500C (atmos) 0.12 0.60 0.18 0.88 v

(1) 0.5 inch axial gap + 0.040 inch radial gap between graphite end reflectors and clad.

(2) 2.5 inch plenum above fuel.

temperatures are below .5000 0) and only a 5% perturbation on the 13.4 atmos. hydrogen pressure at 95000 (fuel temperature safety limit when clad and fuel are at the same temperature). The pressure contribution from the equilibrium concentration of unstable fission products in the free volume of a fuel element is only a very small fraction of the pressure resulting from the stable fission product gases and thus would be essentially

.

• l l

inconsequential as far as pressure is concerned. It should be noted here i that 'the oxygen and nitrogen from the approximately one atmosphere air inside the fuel element are absorbed by the zirconium in the early part of fuel life. Thus the fission product gases released are in essense just renlacina the pressure contribution of the air. Also, it only takes a very ,

few degrees change in temperature .to change the hydrogen pressure by 55.

During pulse (transient) operation of TRIGA fuel, significantly high temperatures last for only a few seconds and essentially no diffusion of fission products occurs over this short time scale, and thus there is no additional pressure contribution from fission products aside from those released during normal operation.

i For instances where the temperature can go to high values on a relctively .

slow transient, such as af ter a loss of coolant accident (not applicable to this Application for License Amendment) where peak temperatures occur af ter about 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br />, the diffusion of additional fission products can occur.

However, it seems fairly evident that the release times are far too short,

! even for highly elevated, short - term release rates, to significantly increase the fission product inventory which has been released during the i entire burnup lifetime of the fuel element during which 505 to 60% of the U235 has been depleted. A fuel temperature excursion to -950cc following a loss" of coolant will have fuel temperatures above normal operating temphratures for only about 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />. The maximum release rate for only a portion of this time might be as much as about 100 times the release rat,e -

for normal , operation. However, it took 10 years' or more of normal operation at 1 MW to release the quantities of fission products shown in Table 1. Thus, the additional release of fission products during a temperature excursion to 950oC adds only a small amount to the already released fission product inventory. This would cause a directly corresponding small increase in the pressure due to the fission products.

It is important to emphasize here, however, that the diffusional release rates for fission products from highly burned up TRIGA fuel should be lgag than for low burnup fuel because of the morphology changes, such as o

increased porosity, which have occurred. The reason for this is that the low solubility- fission product gases will be trapped in the increased number of internal voids existing in more highly burned fuel (1) (2).

No correlation is seen between the gamma scan burnup results of fuel rod .

Number 1086 (see Table 2-1 of GA-A18599) and the possibility that fission i

product release values from highly burned up TRIGA fuel are sizably ' greater than expected. It is assumed that the question raised is whether . the unexpectedly low burnup result for fuel red 1086 is actually due to the -

release of fission products from this ruptured element. Thus, the other fuel elements which were gamma scanned were all intact vs rod 1086 which had ruptured and red 1086 gave a gamma scan result which showed its burnup to be lower than expected compared to all other rods which were gamma scanned. In other words, could it be that the fission product release from the fuel to the free volume within the clad is much greater than expected, and the loss of those fission products from the ruptured fuel rod 1086 lead to a deduced lower burnup (but it fact, not really a lower burnup)?

i With reference to Table 2-1 of GA-A18599, the four fuel rods which were gamma scanned had burnup values which were derived from calculations normalized to experimental values. The time-dependent burnup of each

] specific fuel rod in the specific geometry of the fuel cluster was calculated over the duration of the test (as shown in Figure 1-6 of GA-Ak8599). The relative time dependent burnup of each rod, compared to each other rod, was determined by this calculation. Absolute burnup values were determined . by normalizing' this' set of calculated curves to the experimentally determined burnup values for two of the nine 45 wt-5 U rods.

The experimental burnup values were determined by isotopic analysis of the

, U235 and U238. Both experimental values gave essentially the same (1) M. T. Simnad, " Fuel Element Experience in Nuclear Power Reactors; ANS-AEC j Monograph, USAEC,1971.

(2) D.R.T. Frost, Nuclear Engl Elements, Pergammon Press, N.Y. 1982

_7_

l

normalization to the calculated , set of burnup curves. Fuel rod 1086 (one of the four gamma ccanned rods) also had its burnup determined explicitly by isotopic analysis. It did not correlate with the value derived from the normalized calculations. One could say that all the calculated burnup values should be re-normalized to the experimentally determined isotopic analysis burnup value for fuel rod 1086. However, this would mean that the ign previously determined normalization values would no longer fit the derived set of burnup curves. It was concluded that:

1. The burnup values for the 45 wt-5 U fuel rods are much more strongly supported by the previously determined Ann isotopic analysis values;

2. The set of burnup curves should not be changed (except for rod 1086);

3 The experimentally determined burnup for rod 1086 is correct; and

4. The derived burnup for rod 1086 is wrong.

Based on these conclusions, one can use the burnup values to calculate the relative burnup between the four rods which were gamma scanned. The fuel rod with highest burnup, rod 1090, is given a value of 1.0, as shown in Table 2-1 of GA-A18599 The relative values of the gamma scan results can also.,be determined based on a value of 1.0 for rod 1090. As shown in Table 2-1 of GA-A18599, the relative values from the burnup results agree reasonably well with the values from the gamma scan results. From this comparison, it is concluded' that the gamma scan results are confirmed by the burnup results, and essentially no gamma contribution is lost due to the rupture of the clad on fuel rod 1086.

Results of the last tests to be completed on the burned up TRIGA-LEU fuel are just now becoming available. These last two tests involved the determination of the fission product gases remaining in the fuel matrix.

Initial results(1) show that the, quantities of fission products in the fuel matrix are as expected from the quantities produced during irradiation life of the fuel. These measurements, of course, cannot be of the accuracy needed to determine fission product release rates, but they already confirm that no relatively large release rates have occurred. .

(1) Personal communication, J. R. Snelgrove, ANL 4

-g-

- - - ,e , . , -- , _ . - . - .-.

MT

~II -

1 IS-li -

o t

i y

4 APPENDIX A NICKEL-URANIUM PHASE DIAGRAM

• v s

WElGHT PER CENT NICKEL 2 4 6 4 to 15 20 25 30 35 40 45 ie 50 60 i70e i80i 90 e e i i e i e i e i t 1455' 1400

.j' kk5 l

• THERM AL ARREST l g a UQUt0 PRESENT $ 5

+ NO UQUID l i305' I

1300 - 6 mC' 126 e

Y l 1200 j M li i 8

j

/ I S l g

• u / l 3

• n 2 s* i i i titoa  !

• k L

diiOO a

5 I

/

• l d-sum-q

< ~

a,  ; t. c~ t I l

\

l

% i 31 i I I i

h1000 [ g EE[ 985' s $,8

\s H e/ i i l ,

I "O

i \ '

/ i er 's

\

/ ro-l' l 800 d2 to 51 I-790* .d JW :'so '

L'jn26)!

. N/'fre s. m 33 go,3i s

l l

l i

l 700 ,7n. i i sto inAns; e

., j 600 O 10 20 30 40 50 60 70 80 90 100 u ATOMIC PER CENT NicxEL Ni Fig. 577. Ni-U i

l 1

Reference:

Hansen, Max, Constitution of Binary Alloys, 2nd ed., ,

prepared with the cooperation of Kurl Anderko, McGraw-Hill Book Co., '

New York, 1958. l A-1

'f5 p t

. o .

%e t

L f -

x t

l h

APPENDIX B CHROMIUM-URANIUM PHASE DIAGRAM

• WEIGHT PER CENT CHROMlUM 5 10 20 50 1200

/

n s 2+

f' 1100 h ,

/

/

1000 ,

A* i f' 900 -

& 3 y[k=4

\ '

859t90' ,,

j 19.4 (5.0)

I a 800  ;

f 752 t10* ,,

700 -

-i.5 t-0.n f

e L 640t10' ,,

[ 600

1. . Edt i

500 O to 20 30 40 50 60 70 80 90 10 0 U ATOMIC PER CENT CHROMIUM Cr Fig. 3294. Cr-U

Reference:

Hansen, Max, Constitution of Binary Alloys, 2nd ed.,

prepared with the cooperation of Kurl Anderko, McGraw-Hill Book Co.,

New York, 1958.

B-1 2 ,

As

r-4

+-

ii E

t i

r t.

1 APPENDIX C IRON-URANIUM PHASE DIAGRAM

• WElGHT PER CENT URANtdh 10 20 30 40 50 60' 70 80 05 90 92 94 '

96 98' 99 '

1600 I I I I is34 . n E F. i wr /. r.

y" a nEF. 2 0.2 0.4 1500 1000 ,

' \

d r. 'g =- REF. 2

,,00 -q s r

=

,00_ \_

g

\ g 1300 8007e0*

\  % / 7s5*

yy,, \ 1233' 8u-

• o.s 1200, s

"%' 700 s75*

r I - /

\ \ L.,

(iin .

,o00.. '/

- \

,i Sa o

, -l. ".)

'e

- > Ar.-% Fe j

{

1000 \ y f

=,, - r r. t t /

. . sio- g j 900

\ ri o

~" E~,_- ,

\ _!14'_

,oo  ;

m aGN. inANSF.

72s*

\-_sA - . .

so

'#j O .ys . ,73 d u' 600 500) 0 to 20 30 40 50 60 70 80 90 10 0 Fe ATOMIC PER CENT URANIUM U Fig. 399. Fe-U

Reference:

Hansen, Max, Constitution of Binary Alloys, 2nd ed.,

prepared with the cooperation of Kurl Anderko, McGraw-Hill Book Co.,

New York, 1958.

C-1

,i J

I. .

1:

, a I

t -

4 k

j

(

r.

I

, APPENDIX D NICKEL-ZIRCONIUM PHASE DIAGRAM

• WEIGHT PER CENT NICKEL 20 40 60 80 2000 S l3 1860' 1800 f5

,,,ng, n I #N i a' / 9, ,

1600 h ii d. /'

I I ,g

v. l\  ; -'?  ! ian:

al1400 ', $ [-j:'f -

- - ~ ~

',.?M

= i g , 89 11i I84I 1

E I t

} /

-1200* $$

O

'2"  ; i n r .,

U kl# Zrf 's a ll 1000 J 96i' E sp2n f, TR 2.90.9) g N 90e' t 800 g g <i,33 ,

,as-Ir 600 0 20 40 60 80 100 Zr ATOMIC PER CENT NICKEL Ni Fig. 581. Ni-Zr

Reference:

Hansen, Max, Constitution of Binary Alloys, 2nd ed.,

prepared with the cooperation of Kurl Anderko, McGraw-Hill Book Co.,

New York, 1958.

D-1 1

4

l 5

?

4 APPENDIX E CHROMIUM-ZIRCONIUM PHASE DIAGRAM

• WEIGlif PER CEsT CHROMlUM 5 10 15 20 25 30I 35 f 40 45 f 50 55 60 i TO 80f 90f f f f f f f f 1 f e REF 2

-1850* x REF3 tte6g 3" '

1800 s N s, ~} , /  ;

-1675* ,' ,

s /~P' - / ~1630' ,

1600 's / ,-

amrol >s8.5 s N (971 l

\ ,, (Crik 5' '

\ / ,

,. Zrcra (H) j 0 281181  !

{200 / F.8l4.!)

g1 ,

= I 1000 __ ___

_ _ _ ..___.. _ _ _ ..100,,Q,, _ _

,,__,1 8

TRAN$F

'is- i r,C,, u 800 _-t.7 til ,

I en I 600 O to 20 30 40 50 60 TO 80 90 10 0 Zr AToulc PER CENT CMRoululf Cr Fig. 332. Cr-Zr

Reference:

Hansen, Max, Constitution of Binary Alloys. 2nd ed.,

prepared with the cooperation of Kurl Anderko, McGraw-Hill Book-Co.,

New York, 1958.

E-1 s

k.

s l

l f APPENDIX F

)

4 IRON-ZIRCONIUM PHASE DIAGRAM

• E I

t

r

-[ <

WElGHT PER CENT ZlRCONIUM 10 20 30 40 50 60 70 80 85 90' 95

3900,

.L .- L. I _a i 1 ' t

! iss0*

h

?+

igno .__ . ..__. _.

_. .. . .. p

)

i t . . n t F. 1 si 7 " sf i , = a tF. 7 f l l l l 7 i

ii I

1 00; m, ,

! lq 1

I y

'g 34' I g f, 1500 , ' I i  %

i

$ 8 r. .. 's *

\ . t ,I

< \. *

\  !

Jbid/ -,,,0

~/ l 1300 't 6 \ [ s'

/

\ i O #

=1200

, \ ' l a ,s ,

l I g U l \ l g I y tt00- l I f '

I -

's , e Tre t ,

s i I

,000 - t

. '; / -

934 t s'

' Or tio* [ \# Z

- 76 900 i t-841 ~ 7,73 ig $3 g

-e35*

R .

"0

\/

600' Tai na R-'~ (' 53 70 0; W ar.

600l 0 to 20 30 40 50 60 70 80 90 100 Fe ATOMIC PER CENT ZlRC0Nluu 2r Fig. 404. Fe-Zr

Reference:

Hansen, Max, Constitution of Binary A11ovs. 2nd ed.,

prepared with the cooperation of Kurl Anderko, McGraw-Hill Book Co.,

New York, 1958.

F-1 1