CORRRfdT1JE CORRESPO d(R 728E cWC -.

. .

:

.

SOCIETY / COMMITTEE: ADDRESS CORRESPONDENCE TO:,

ANS-5.4 R. O. Meyer

SUBJECT: U.S. NUCLEAR REGULATORY COMMISS:ON'

Fuel Plenum Gas Activity

.iGENDA ITEM:'

FILE NO.: N/A DATE: f[8 2g |379

TO: C. E. Beyer L. D. f'ohleWestinghouse Hanford A/59 General Electric Comoany, M/C 138Hanford Engineering Development Lab. 175 Curtner AvenueP. O. Box 1260 San Jose, California 95125Richland, Washington 99352

M. J. F. Notley

B. J. Buescher Atomic Energy of Canada, Ltd.The Babcock & Wilcox Company Chalk River, Ontario

P. O. Box 1260 Canada, K0J1JOLynchburg, Virginia 24505

Chang S. RimR. J. Klotz Korea Atomic Energy Research InstituteDepartment 9492 P. O. Box 7, Cheong RyangCombustion Engineering, Inc. Seoul, KoreaWindsor, Connecticut 07085

R. L. RitzmanR. A. Lorenz Science Applications, Inc.Oak Ridge National Laboratory 2680 Hanover StreetP. O. Box X Palo Alto, California 94304

Oak Ridge, TennesseeS. E. Turner

W. Leech Southern Science Applications, Inc.Nuclear Fuel Division, W Corp. P. O. Box 10-

33528P. O. Box 355-

Dunedin, FloridaPittsburgh, Pennsylvania 15230

Dear Group Members:

Enclosed is a draft of the introduction to the ANS-5.4 support documentas requested at the November 8,1978 meeting. Your comments are welcome.

Sincerely,

O'# M790322O L/6 6 Ralph 0. 'Meyer, Leader

Reactor Fuels SectionCore Performance Branch

snc ronMgl sure: As stated00 h)

.

.

I. Introduction (R.0. Meyer, NRC)

ANS Werking Group 5.4 has met a number of times since its inception

in June 19/4 to examine fission product releases from U0 f"'I* H#"#2

calculations of reactor behavior require a knowledge of the gradual

release of fission pro 6 cts from the ceran.ic U0 pellets. This is2

especially true in accident analysis where the inventory of radioactive

volatiles ed mes released from fuel pellets, but retained by the fuel

cladding (plenum), defines a source term for plant-release calculations.

The scope of ANS-5.4 is thus narrowly cer;ned to study such

releases and includes the following:

1. Review available experimental lats on release of volatile

fission products from UO nd mixed-oxide fuel.2

2. Survey existing analytical models currently being applied

to light-water reactors.

3. Develop a standard analytical model for volatile fission

product release to the fuel rod void space. Emphasis is

placed on obtaining a model for radioactive fission

product releases to be used in assessing radiological

consequences of postulated accidents.

The volatile and gaseous fission products of primary significance

are krypton, xenon and iodine. The radioactive isotopes of interest

are, by their nature, unstable, i.e. , they have finite half lives.

Ironically, for krypton, xenon and iodine there are no radioactive

isotopes with half lives greater than 8 days except Kr (10.7 yr)

and I (1.6x10 yr).

1-1

.

.

phenomenological model, the Booth diffusion-type model (1-5), and has

fitted this model empirically to a selected data set (6), whose virtues

will be described below.

The Booth model describes diffusion of fission-produc+ atoms in a

sphere of fuel material. The governing equation is

3C/at = B - AC - VJ, (1)

where C is the isotope concentration (atoms /cd ), B is the production

or birth rate (atoms /cm3 sec), A is the decay constant (sec-l), and J

is the local mass flux (atoms /cm2 sec). This equation is fundamental

and applies to isotopes of any chemical species with any half life. It

simply says that the rate of concentration change in a region is equal

to the rate of production minus the rate of decay minus the rate of

loss by mass flow out of the region. Equation 1 says nothing about the

mechanism of mass flow. The apparent diffusion coefficient D is contained

in the flux term, which is given by

J = -D VC. (2)

This basic diffusion equation, like Eq. 1, contains no information

about the diffusion mechanism and merely assumes that a net flow of

matter occurs because of the existence of a concentration gradient

and that the flux is proportional to that gradient.

The production rate B and decay constant A is known for all iso-

topes, but the diffusion coefficient is unknown and must be determined

.I - 3

enpirically from experimental data. From a general knowledge of atomic

migration (7_) it is known that the diffusion coefficient of a species in

a host material depends on the properties of that material and its inter-

action with the diffusing species. These interactions are primarily

electronic in nature so that different atoms (elements) would have

different diffusion coefficients. Because the valence and ionic pro-

perties of krypton and xenon are similar, it is not surprising to

discover that their diffusion coefficients in UO2 are similar, but

there is no reason to expect them to behave like iodine. Therefore, it

must be presumed that different elements migrate and are released at

different rates.

On the other hand, the diffusion behavior of a chemical species

can be expected to be the same for all isotopes of that species. While,

strictly speaking, there is a diffusion isotope effect that is dependent

on isotopic mass (8), this effect is very small, has only been detected

in a few precise experiments using isotopes with large mass differences,

and such small differences in diffusion behavior would be totally

imperceptable in the context of fission gas release.

Admittedly, the Booth diffusion model is an over-simplification

of the physical process, and the effective diffusion parameters that

are determined by empirically fitting the Booth model to gas release

data are not the diffusion coefficients for atomic diffusion of

krypton and other chemical species in pure U0 . Atomic diffusion,2

gas bubble nucleation, bubble migration, bubble coalescence, inter-

action of bubbles with structures and irradiation resolution are all

I-4

.

mechanisms that are involved in fission gas release. Some of these

processes, like bubble migration, are relatively well understood.

The microscopic parameters that govern these mechanisms are, in turn,

dependent on the atomistic materials prcperties, such as diffusion

coefficient, heats of vaporization, etc., which are independent of

isotopic makeup. It, therefore, seems appropriate to assume that the

overall release kinetics are the same for all isotopes of the same

chemical species regardless of the complicated nature of the release

mechanisms.

There is a recognized pitfall in the method chosen by the Working

Group. The Booth equations describe a smooth continuous release process

and, therefore, do not show discontinuous releases or bursts. To the

extent that burst releases affect the relatively small sub-population

of radioactive gases, an error is incurred. It is considered beyond the

state of the art to model burst releases in a quantitative manner and,

therefore, such errors must be tolerated.

Finally, a temperature--independent recoil mechanism is also

expected to be important for radioactive gas releases. As with the

temperature-dependent diffusion-type model, the release fraction will

depend on the isotopic half life. Because of the mechanical nature

of the recoil process, however, all chemical species are treated alike.

1-5

.

.

References, Section I

1. A. H. Booth, Chalk River Report, CRDC-721 (1957).

2. A. H. Booth, Chalk River Report, DCI-17 (1.357).

3. A. H. Booth, and G. T. Rymer, Chalk River Report, CEDC-72D (1958).

4. S. D. Beck, Battelle Report, BMI-1433 (1960).

5. B. Lustman, in J. Belle (ed.), " Uranium Dioxide: Properties and

Nuclear Applications," (USAEC, Washington,1961) p. 431.

6. C. E. Beyer and C. R. Hann, Battelle Report, BNWL-1875 (1974).

7. P. G. Shewmon, " Diffusion in Solids," (McGraw-Hill, New York,1963).

8. N. L. Peterson, Solid State Physics j![, 409 (1968).j

1-6