ML20137Q873

Summarizes Assessment of Potential Safety Concerns Re Tritium Production in Commercial Lwrs W/Emphasis on Matls Interactions & Oxidation Behavior Under Design Basis & Severe Accident Conditions

ML20137Q873
Person / Time
Issue date:	03/06/1997
From:	Cronenberg A Advisory Committee on Reactor Safeguards
To:	Advisory Committee on Reactor Safeguards
Shared Package
ML20137Q857	List: ML20137Q852 ML20137Q873
References
PROJECT-697 ACRS-GENERAL, NUDOCS 9704110157
Download: ML20137Q873 (16)

Text

,

jo tstg' g f.:

s j

V.

UNITED STATES e

1

.c NUCLEAR REGULATORY COMMISSION '

.I ADVISORY COMMITTEE ON REACTOR SAFEGUARDS WASHINGTON, D. C. 20555 March 6, 1997 MEMORANDUM tot'ACRS and Staff MEMORANDUM #:

AWC-106.97-DRAFT ZR2d:

August W. Cronenberg EUBJECT:-

Safety. Issues Concerning Tritium Production in 1

Commercial LWRs, j

Bummary:

Per request by Dana Powers, I have. summarized here observations from a " quick-looka assessment of potential safety concerns related.to tritium (H3) production in commercial LWRs, with j

emphasis on materials interactions and oxidation. behavior under design basis and severe accident conditions.

Questions related to the regulatory framework and appropriate CFR regulations that might ultimately govern licensing requirements are not considered.

i From an examination of phase-diagram information ani :hemical reaction i

energetics, it can be inferred that the Li-AlO targets clad in a 2

stainless-steel /Zircaloy composite present no increased degradation l

[

potential than that for typical PWR Zircaloy-clad A1 0 -B C burnable l

2 3 4

~

poison'roda.

Binary phase diagrams for Fe-Zr and FeO-ZrO indicate 2

eutectics at 1220 K for the Fe-Zr system and 1640 K for FeO-ZrO.

l 2

L since the Li-A10 target design entails concentric layers of i

2 stainless-steel and Zircaloy, eutectic melting can be expected for j

either metallic or oxidic conditions.

However, similar eutectic assembly degradation can be expected at the stainless-steel I

e spider /Zircaloy end-plug junction for typical PWR burnable poison rods,.and indeed was observed for the TMI-2 burnable poison y

. assemblies.

A comparison of simp 3e parabolic kinetics for Zircaloy-steam and steel-steam reactions indicates both a higher reaction heat and faster i-kinetics for the Zircaloy-steam reaction than for steel.

Thus, l

oxidation' associated degradation of typical Zry-clad PWR poison l

assemblies is predicted to be greater than that for the stainless-steel clad target rods.

These observations thus indicate no

]

increased overall degradation for target rods than that for standard burnable poison rods.. This reasoning is corroborated'by TMI-2 findings,: where the extent of degradation of Eircaloy-clad burnable poison rods was observed to be greater than that of the stainless-i steel clad Ag-In-cd control rods, i

r Although no *show stopper problems related:to use of Li-AlO /Zry-SS-I a

2 clad. targets.in commercial LWRs is evident from the limited ossessment provided here, a comprehensive examination of all aspects of/ safety will'be required for licensing of reload cores containing i

Li-A102. targets.

9704110157 970409 y

1 PDR PROJ Enclosure l

697 PDR Jg'

+

i l

l l

f

'la__ BACKGROUND

\\

The U.S. Department.of Energy (DOE) is currently pursuing tb."ee option. to assure continued tritium production capability for DOD j

-(Department of. Defense).

Under consideration are the accelerator j

production of tritium option,. tritium (H3) production in the Fast Flux i

Test Facility (FFTF),.as well as H3 generation from irradiation of i

Lithium-bearing targets in a commercial LWR.

In prior decades F3'was j

produced at the Savannah River Site (SRS), using heavy-water -

i moderator / coolant reactor technology with an'A1/U-alloy lead fuel-

]

design.and Li-Al alloy targets for tritium generation.

In the early.

1990s, DOE made the decision to forgo restart of the Savannah River reactors, and to rely on a different technology for tritium i

production.

For LWRs the basic the SRS heavy-water technology, process is the same as used that using i

that is:

Li8+o 1

n

---> 3Li7 ---> 2H3 + 2 e*

H 3

3 H

(1/2-life) 12.3 yrs

=

3 DOD needs for fussionable tritium (gH) relate to weapons stockpile l

3 replenishment, due to its relatively short 1/2-life of -12.3 years.

To secure irradiation services from an existing commercial LWRs, will j

2 require NRC-approval and certification.

DOE has tasked.the Pacific j

Northwest Laboratory (PNL) to assess the qualification process for NRC 1

approval, where Li-Al alloy. targets would replace A1 0 poison rods presently used many commercial PWR cores. 3-B C burnable 2

4 For present discussions and comparative remarks, we shall assume basic Westinghouse 4-loop PWR design conditions.. The'following section briefly reviews the proposed design of the Li-Al alloy targets that l

would replace Al 0 -B C burnable poison rods, followed by a brief 2 3 discussion of material, core physics, and safety issues that may need i

to be considered in the licensing of such technology.

First however, i

I present my own thoughts on weapons H3 production in commercial LWRs.

i i

2.

PROPOSED Li-AlO TARGET DESIGN for LWR APPLICATIONS l

2 The overall approach is to replace several Westinghouse type A10 -B C

{

burnable poison rods with Li-A10 target rods for H3 production. 3All 2

2 Westinghouse fuel assemblies employ the same basic mechanical design, i

using a 17x17 array and differ only with respect to the types of contrc1, burnable poison, or instrument tubes that are part of each fuel assembly.

Figure 1 illustrates the basic features of a typical fuel assembly.

The fuel rods are connected to stainless-steel upper and lower end i

fittings, while the movable burnable poison rods in the assembly are t

-bolt attached to a stainless-steel spider-like structure which is chown in more detail in Figure 2.

Of interest here is differences between the Al 0 -B,C burnable poison rods that are normally contained 2 3

'within a licensed PW core, which are to be replaced by Li-Al alley targets for tritium production.

2 i

_~

j t=--~~l

_--.-c w s

n 4EiHBGEn.=

Z"N %/c5f

--3)

=T - C ~i mea 3...

• ~""~

j q

\\

',,..am 7

~ _ _

=::'.=

i L,

VL.

\\

• =- ** g P

0

)

Ena sims ano

,b ente v

, e;;y amums

_am

..1..

m EE a q

am m j

mm i

"'u kllla a a

$~#hh Figure 1. Illustration of a typical PWR fuel assembly incorporating a spider-like absorber assembly with control rods that move within guide tubes (thimbles).

s Most of the burnable poison rods in a typical Westinghous.e-PWR are a B,c compound dispersed in a ceramic Al 02 3 pellet material, which are clad with Zircaloy and attached to a stainless-steel spider element.

Motion of the A1 03 pellet stack is restrained by a hold-down spring, 2

while the ends of the Zirealoy cladding tubes are closed by welded Zircaloy plugs.

For Westinghou'se PWR designs, there are 16 burnable poison rods in each fuel assembly, which fit inside 16 Zircaloy guide tubes (thimbles), which provide an envelope which directs the movement of the rods.

The tubes ends are open to coolant flow ~at the top and bottom, but the amount of flow is small compared to a fuel rod chanhel.

The Li-A10 target design is shown in Figure 3, with similar 2

dimensional characteristics as the A10 -B c/Zry-clad burnable peisen 23 4

rods.

However, the Li-A10 target has a concentric / layered design.

2 The Li-A10 target is annular in form and is encased on both its inner 2

cnd outer surfaces with a nickel-plated Zircaloy barrier material.

The Ery-A10 -Zry annular composite is then clad with stainless-steel 2

to provide structural strength.

To prevent diffusion of hydrogen from the target, the inner surface of the cladding is coated with a 3

m

~

perueation raciatent cluminized barrier.

Tha irrcdistion design for the reference target le 550 effective full-power days, for one fuel cycle.

4 e

l~m G

?

A, r

4 d

o

% h

._.k.%,

~ -%.

=.,

ll:::,

~~

a sc Figure 2. Typical Westinghouse PWR spider-like absorber assembly with 16 rods that move within guide tubes (thimbles).

t 1

l A

Ei?

=?

l, M

E.c

-s.

'2 gd E

g*/l 20%::::

g,

=

%m E

~~

t9 ac Figure 3.

Illustration of proposed concentric-annular Li-A10 2 target design for tr.4 tium production in commercial LWRs.

l 4

3.

SAFETY' ISSUES for Li-AlO2= TARGETS in COMMERCIAL LWRs Table 1 presents prototypic material and dimensional data for the proposed tritium target cods and standard Westinghouse burnable poison rods.

Also included is data for Ag-In-Cd type control rods used in some PWRs (e.g..TMI-2).

As indicated, dimensional characteristics are essentially the same, though differences exist with regards to materials.

Although a Zirclaoy/ stainless-steel composite is proposed as the cladding for the Li-AlO absorber rods, as opposed to Zry-4 2

cladding for standard.PWR Al 0 -B C burnable poison rods, stainless i

3 2 3 4

steel is also the cladding for a number of licensed PWR reactors that y

use Ag-In-Cd type control rods.

i Table 1.

Comparison of Li-Alo Target Rods with Full-length Burnable 2

Poison (A10 -8 C/Zr-clad) and Ag-In-Cd Type Control Rods.

2 3 4

Parameter Al-Li Taratt heurnable Poj sen.

Control Rod I

{

Absorber Material L1-AlO A10 -s c Ag-In-Cd r;,verall Length

= 152 $n 15$ 3in 152 in 4

Absorber Length

= 134 in 142 in 142 in Rcd Weight (1 rod) 2.26 lb 1.9 lb

= 6 lb Rods / assembly 16 16 16 Absorber Clad CD 0.381 in 0.381 0.381 Absorber t1 adding 85-316 ll:irealoy-4 ss-304 Darrier Material Ni-Ery Guide Tube Thimbles tircaloy-4 Zircaloy-4 Eirealoy-4 A preliminary examination of some of the licensing and safety issues essociated with the use of Li-AlO target material in a commercial PWR

{

2 is presented in a recent PNL report [1].

The PNL report argues that the testing of the tritium production can be done by a commercial licensee under a 10CFR50.59 process.

A recent memo by Powers [2]

I however suggests that the 10CFR50.59 process may not apply because of a number of unreviewed safety cuestions associated with material interactions during design basis accident conditions, as well as changes in core neutronics associated with replacement of -

A10 '-B C/Zr-clad rods with Li-6 enriched Li-Alo /ss-clad rods, where 23 4

2

)

the full extent of differences in core response for rod drop accidents has not been assessed.

Additional concerns of Powers [2] include:

i lack of a comprehensiva assessment of material interactions and overall target assembly performance under design basis accident conditions the safety implications of steam reactions with stainless-1

steel, the safety implications of exothermic inter-metallic

)

reactions of Al with stainless-steel components (Fe, Cr, Ni)

^

adequate understanding of the potential for exothermic inter-metallic reactions of Zr with Ni

\\

5 1

=

potential intcrcetions of c Zr-Ni GutGctica with Li-AlO 2

alloy potential problems associated with cracking in the Li-AlO 2

alloy due to power cycling and shutdown transients.

lack of substantiated target vibrational analysis to preclude fretting of the Li-AlO2 targets stress corrosion cracking of assemblies and release of tritium to the coolant, may be issue requiring in-core detection for H3 In view of these concerns and safety questions, Dr. Powers has expressed the view [2] that NRC approval of a commercial PWR core reload, replacing Al 0 -B C burnable poison rod with Li-AlO /SS-clad 23 4

2 target rods, cannot be done under a 10CFR50.59 process.

Powers (2) argues that the safety analysis will have to be far more comprehensive and persuasive than that provided in the anbject PNL report [1].

Although no attempt is made here to address all of the above indicated safety concerns, the following information is considered useful in providing some physical basis to assess target material behavior under operational and design basis accident conditions.

Areas of primary investigation discussed here, relate to material interactions and oxidation behavior under off-normal /high-temperature (LOCA) and core uncovery conditions.

4.

MATERIAL-INTERACTION & SAFETY _f0NSIDERATIONS To assess potential chemical and metallurgical processes that may influence target behavier under accident conditions, infortnation is 1

presented comparing thermophysical property and phase-diagram information for the Li-AlO /SS-Zry composite target rods as opposed to 2

the Al 0 -B C/Zry-clad burnable poison rods.

Heavy reliance is made 23 4

on prior investigations [1-El related to material interactions noted from post-accident examination of the TMI-2 core debris, since the TMI-2 core incorporated A10 -B C/Zry-clad burnable rods.

23 4

i 4.1 Comoarison of Thermal / Physical Pronerties Most of the burnable poison material in the TMI-2 core was Al 0 -B.C pellets which were clad in Zircaloy-4.

About 1.4 wt-%

of 2 3 the W.I-2 pressed and sincered burnable poison pellet is B C.

The theoretical density of burnable poison pellets is about 248 lb/ft 3 (3.97 g/cm3).

The melting point of the Al O2 3 + B C mixture is about 4

3700*F (2310 K), while the heat of fusion can be approximated as that of Al 0 is 46 Btu /lb (25.5 cal /g).

23 j

1 Figure 4 shows the thermal conductivity, specific heat, and thermal cxpansion properties of Al 0 As indicated, 3 and B C [1]

an a function of 2

4 the thernal conductivity of Al 03+BC temperature.

2 4

decreases with increasing temperature, although it is somewhat higher than that of U0 -

2 6

1 4

20

'l c

e---e At:0,0% possity 3

j i

( 16

1. """O Al o,,230% porosity s

j O B,C unirrad!sted

3,C Irradiated

?

a l10 l

N J

k 0

0-2

~0 Al,O 8,C a

Malting g

i i

i Poin' i

._J 0

1000 2000 3000 4000 J

I g.36

.G

}

/*#

,p_ -

t j.u l

e---e Al:03 Extrepolated

.16 0

1000 2000 3000 4000 i

'2 2-3 Y

f 6

e-o A1:03 e==-4 B,C 1

/*

l0 ed s

8

~

0' 1000 2000 3000 4000 llwnperatum('F)

M#1 1

Figure 4. Temperature dependent thermophysical properties of Al 0 and B C.[1].

23 4

7

.~

~. _

_ _ _ _ _ ~.. _

t 1

As previously indicated.the Li-A10 target (see Figure 3) ' has a 2

concentric / layered design where.the stainless-steel cladding.is lined on its inside surface with a-Zircaloy-Ni barrier material,.so that the Zr-Fe binary phase diagram is of interest.

As shown in Figure 5, the..

spider /end-plug junction of standard PWR burnable poison' assemblies consist-of a Zirceloy end-plug attached to the stainless-steel spider.

Thus the Zr-Fe binary phase diagram is'likewise of interest and is.

j alco presented in Figure 5.

This phase diagram indicates.a eutectic j-

.at.1577 K for 0.088 atom. percent (a/o) Zr and another eutectic at 1220 1

^

.K for 0.76 m/o Zr.

Thus at temperature above about 1600 K, eutectic

• associated degradation of hoth PWR burnable poison and Li-AlO2 target rods can be expected.

Figt.re16 presents ti,a binary phase diagram.for the FeO-Ero J

system.[1], indice. ting a cutectic at about 1370*C (1643 K)mfor 51w/o

.Zr0.

Therefore, whether in a metallic or oxide state, Zr-Fe eutectic 2

interaction can be expected during severe core damage conditions for either Zircaloy-clad A10 -B C burnable poison or Li-AlO /SS-Zry 3

target assemblies.

Suchbr4Fe eutectic interaction was indeed 2

i~

.obperved from examination of.the TMI-2' core [3, S-6).

1 4

4 I

weght fraction 2r

.i.:..s a.s a

.7 a ss.e.es I

' ' y ',,,a,'

' pas,.

sIma,hn '

Om.=who - +

4 4 4 4,,,,,

swa g

i 800

[

t g

l-p uom i

e i

e s.

,l hM W j L

i R

l e

e l

e op a,,

2r oss plus L

=--

=

j= L

'.\\\\,.j:

i n

f

gs i

Alsos M

~l om w s.c.

• * * ' = "

,,y g00 1.

- l 4

[

' h. fel

[a oI de 8' 8"*'

• os at p.

M L

.,m Asomic tression 2r i

Figure 5. I31ustration of the material composition of a A10 -B C burnable poison assembly at the upper-end fitting 23[1]4, and the Fe-Zr binary phase diagram [2].

?-

a-I l.

I I

[

l I

{

Liquid l

l-w0o -

Uguid + Zro (solid solution) i s

I I

l u00 -

4 I

I i

1 s

FeO 20 40 60 SO ZrO Figure 6.

Binary phase diagram for the FeO-Zro2 system.

Al E -2r fo) Svatem 2 3 Since the Al 0 cladding, the phase diagrams of interest are for the A10 -Zr and+ B C abso 2 3 4

A1 0 -ZrO2 systems.

23 These binary phase systems are not available in 2 3 the common literature, so that one can not definitively judge the potential for eutectic melting of such materials.

However, the fact that the melting point of Al 023 (2310 K) is significantly higher than its metal (Al = 933 K), and that both Zr and Zrom have relatively high melting points (Zr = 2125 K, ZrO2 = 2950 K), indicates that the A1 0 -Zr or A1 0 -ZrO 23 23 2 systems, probably would not form eutectics melta at temperatures lower than that for the Fe-Zr system (eutectic melt at 1220 K).

Thus, the most likely mode of eutectic melt failure is via Fe-Zr (spider /end-plug) interaction, rather than Al 0 -pellet /Zr-cladding interaction.

23 4

4 9

,~

4.2 Erothermic Reactions 1

In. addition to phase-diagram con'siderations and the attendant-potential'for formation of low melting point eutectics, high-temperature exothermic reactions may also increase structural degradation of the burnable poison and target assemblies.

Although the potential exists for numerous chemical reactions between core constituents, reactions that are significant. contributors to core degradation-are essentially limited to tsechanisms in which one of the i

reactants is in the gas phase, for example steam-metal reactions.

, Solid-molid or liquid-solid reactions are considered of secondary importance, because the kinetics of such ion-exchange reactions are largely' dictated by reactant diffusivities for the condensed phase ),

which are much lower than that for the gas-phase.

Thus, the relevant-exothermic reactions of interest for absorber rod degradation are essentially limited to s' team-metal (i.e., gas-solid) rather than absorber-cladding (i.e. solid-solid) reactions.

Zirealey cladding,. stainless: steel, and UO fuel. exhibit some tendency 2

tol react with steam [11-17).

From-a core degradation standpoint, Zircaloy and stainless-steel reactions with steam are of primary importance, because of their relatively high reaction heats and fast reaction kinetics.

With respect to the actual absorber materials, the i

burnable poison pellets (B C dispersed in ceramic-like A10 ) would 4

2 3 j

not be expected to experience significant oxidation by steam, since A1 0 is itself a stable oxide.- Therefore, the primary chemical 2 3 reactions of interest center on the steam induced oxidation reactions of Zircaloy and. stainless steel.

The principal thermodynamics properties for.these reactions are summarized in Table 2.

As indicated, the heat of reaction for Zircaloy is about 2-1/2 times greater than that of stainless steel, at the respective melt temperatures.

i l

Table 2.

Thermodynamic Properties of ZircaLoy and Stainless Steel Reaction with Steam e

i Melt Principal Heat of Reaction mesetant Temperature oxides Formed with Oxygen with staam (11auidi (K)

(enl /a)

(cal /a) l.

Eircaloy

.2125 Er0

-2,883

-15so f

2 stain-steel 1sso Feo, er os

-1,opa

-sss r

For Zircaloy, the governing stoichiometric reaction can be expressed as:

2r <= E.> 2H2 + Er02 + 1560 cal /g-Er 2H 0(steam) +

2 1

where k cxidizab,isthereactionrateconstant.

For each mole of Zircaloy

{

two moles of hydrogen are generated and a'significant amount of reaction heat is released.

10 l

1 1

I

.*e For. stainless steel, the~reac'tions are more complex, since a number of c/

j the' constituents of the alloy material _can oxidize._. Stainless steel-304.contains about 70 w/o iron, 18 w/o chromium, 8 w/o nickel, and lower fractions of other elements.

stainless steel-316 contains

'about 2:w/o more of nickel and 2 w/o less of, chromium.

For present l purposes, the chemical reaction suggested by Baker [11-13] can be used to approximate the reaction of stainless steel (SS) with steam:

k, 2Cr + Fe + 4H2 < = = = = > FeO [Cr20 ) + 4H

+ 596 cal /g-SS 3

2 1

The ratios for various equilibria are shown in Figure 7.

The figure L

indicates that chromium would tend to-react when-in the presence of a.

.large excess of hydrogen to form Cr2 3 However, the exceptional-0 stability of the FeO-Cr2 3 oxide compound, suggests that it would be 0

the likely product in a hydrogen-rich atmosphere because of the higher weight percent of iron in the steel alloy as compared to chromium.

However, the remainder of the iron should form FeO only when the local steam to hydrogen ratio is greater than about 0.5.

Any comparison of-the overall impset of oxidation on structural degradation, involves not only a com speed at which such reactions occur,parison of reaction heats but also that is, reaction kinetics.

For n-intact geometry, the oxidation reactions of metal structures in a steam environment can be approximated based by parabolic kinetics

[12), where the mass of metal reacted per unit surface area (W) is proportional to the square root of time, i.e.

W= (k t)L5 p

where t is time in seconds and b is the parabolic rate constant i

(mg-Zr reacted /cm )2/s.

Such pafabolic kinetics are governed by the 2

oxygen diffusional characteristics in the metal, so that the rate constant A (analogous to the diffusion coefficient) can be expressed

~;

in terms of the temperature-dependent diffusional properties of the oxide surface layer.

The temperature dependence of k can be described by the Arrhenius equation of the form:

p kp = Aexp(-Q/RT) where A is a scaling constant, Q is the activation energy,:R is the universal gas constant, and T is the absolute temperature.

f Table 3 presents the principal expressions developed for % for Zircaloy and stainless steel, where results are compared at the common temperatures of 1200 K and 1600 K.

As indicated, the Zircaloy-steam reaction exhibits faster overall kinetics then stainless steel at equivalent temperatures.

The fact that burnable poison rods are clad in Zircaloy, which is.more reactive with steam then stainless-steel clad rods assembly, indicates standard burnable poison assemblies would experience a higher degree of oxidation and structural degradation when exposed to high-temperature steam.

G 11 p

w

-,m-

-r

-y w

' " =

7m 800 1000 ano 1400 seco sooo I.

gg, 1 I

I s

s kA d

w t

w So Pg Pu,o 1 No w

N f.19 m

4 p

e D

D f

I f

f I

y to e

as at as as c4 inorrtio Figure 7. Thermochemical equilibria data for oxides of stainless-steel constituents.

Table 3.

Arrhenius correlations for the Parabolic Rate constant Zirealov/ Steam:

g f(mo-2r reacted /em 322 /s Immoerature, K Reference

2. 94 E+ 6 [exp (-3 9,94 0/RT) )

T < 1800 Cathcart-Pawel 8. 79E+ 5 [exp (-33, 000/RT) )

T > 1650 Urbanic-Heidrick i

3 3. 3E+ 6 [exp (-4 5,500/RT) )

T > 1500 Baker-Just (*)

26. 8E+ 7 [exp (- 52,350/RT) )

T > 1770 Prater-Courtright Stainless Steel / Steam:

g I(me-2r reacted /cm ;2/m Temeeratu W Referenet a

2. 4 E+12 [exp (- 84,3 00/RT) )

1273 <T< 3650 White

3. 0E+ 07 [exp (-53,000/RT) )

1200 <Tc 1700 Baker (*)

Resultse Parabolic Rate k at 1200 K k at 1600 K CQDgtant in Steam (ma-mekal reacted /cm } 2 (mo-metal reacted /em 172 2

p Zircaloy 0.172 20.3 Stainless-Steel 0.023 4.43 R = 1.987 cal /g-mole K o = Reaction Rate Constan!: (k ) used here p

I 12

.., _. - _ _ ~ _. - _ _ _ __ _

7 1

5.

INSIGHTS ON TRITIUM PRODUCTION USING Li-ALO RODS in LWRs 2

!sased'on a limited examination of phase-diagram-information and-

-cherical reaction energetics, it can be inferred that for design basis

'and. severe accidents, Li-A10 /ss-Zry-clad targets present no increased 2

degradation potential under steam-oxidation conditions than that.for=

i typical Zircaloy-clad Al 0 -8 C burnable poison rods. - It is noted 22 4

that=the binary phase diagrams Jar the Fe-2r and FeO-ZrO2 systems.

1 indicate low melting point eutectics.

For the Fe-tr system, a.

eutectic as. low a,s 1220 K can occur, while for the oxide FeO-ZrO 2

L system a eutectic is indicated at about 1640 K.

At the stainless-steel'apider/zircaloy and-plug junction of burnable. poison assemblies, autectic' melt failure can be expected at the temperatures indicated..similar behavior would be' expected for the Eircaloy/ Stainless-steel concentric target cladding.

t A comparison'of the-Zircaloy-steam oxidation reactions for standard PWR burnable poison' rods (with Zircaloy cladding) and tritium production target roda (stainless-steel cladding) assemblies, indicates both a higher reaction heat and faster kinetics for the Zircaloy-steam reaction versus the steel-steam reaction.

Thus, eteam-induced oxidation of the zircaloy. clad poison assemblies can be i

. cxpected to be greater for typical PWR burnable poison rods than for the stainless-steel clad target rods.

These observations thus indicate no increased overall degradation for target rods than would occur for standard burnable poison rodst.rather Zirealoy-clad burnable poison rods indicate greater degradation potential.

This reasoning is corroborated by TMI-2 findings, where the extent of degradation of L

Zircaloy-clad burnable poison rods was observed to ee greater than that.of the stainless-steel clad Ag-In-cd control rods [1].

t Y

Although the limited comparison here does not indicate any "show otopper" problems related to use of Li-AlO /Zry-SS-clad targets, other 2

considerations may be noteworthy.

Though the use of commercial LWRs to produce tritium for fusion weapons replenishment may be the most cost efficient of the three options proposed, particularly when compared against the spallation / accelerator option (several billions s

of dollars in research/ development / construction costs), I would nevertheless expect that use of commercial LWRs for this purposa to draw extensive negative public reaction.

Recent surveys appear to indicate a slow but steady increase in public j

occeptance of nuclear power, following low-points after the TMI-2 and Chernobyl accidents.

I believe the nuclear power industry would be well advised to weigh heavily public opinion in this case, and may

. best forgo any short-term gains related to weapons tritium production.

As in the past, the production of weapcas material I believe is best 4

cecomplished at a dedicated government facility, where the best i

cpproach appears to lie with the FFTF ontion.

If indeed the government-owned FFTP option should prevail, as a reactor facility, FFTF conversion / restart would most likely be subject to NRC jurisdiction, as this agency moves to oversight of the DOE complex.

s 13 s

~

~

~ _.,

L,

-6; REFERENCES 1.

Pacific Northwest National Laboratory, Report.on the

-Evaluation of the Tritium Producing Burnable Absorber Rod Lead Test Assembly, PNNL-11418, (Nov. 1996).

2..

Dana Powers, DOE Proposal to Make Tritium-in LWRs, memo to

&CES Members & Staff, (Feb. 1997).

3.

A. W. Cronenberg, " Effects of Burnable Poison Rods on TMI-2 Core. Degradation, Idaho National Engineering Laboratory a

Report,. EGG-TMI-8233, (Nov. 1988).

4 4.

M. L. Russell, "TMI-2 Accident Evaluation Program Sample Acquisition and Exa'mination Plan", Idaho National Engineering Laboratory Report, EGG-TMI-7132, (Jan. 1986).

S.

E. R. Carlson and B. h. Cook, " Chemical Interaction Between Core and structural Materials," Proceedinos of ist International Information Meetino on the TMI-2 Accident, Germantown, MD, (Oct. 21, 1985).

6.

X. Yinjamuri, D. W. Akers, and R. R. Hobbine, " Examination of HB and BB Leadscrews From Three Mile Island Unit 2 (TMI-2)", Idaho National Engineering Laboratory Report, EQTcTMI-6685, (April 1985).

7.

M. Hansen, Constitution of Binary A11ov.g, McGraw-Hill, New York, 1958.

8.

R.

P. Elliott, Constitution of Binarv A11ovs: 1st surelement, McGraw-Hill, New York, 1965.

9.

E. M. Levin et al., National Bureau of Standards, Phase Dinarams for Ceramista, American, Ceramic Society, 1969.

10.

S. M. Walas, Reaction Kinetics for Chemical Encineers, McGraw-Hill, New York, 1959, pp. 126-148.

11.

L. Baker and R. C. Liimatainen, " Chemical heactions,"

i 4

Chap. 17, The Technoloev of Nuclear Reactor Safetv, Ed..T.

J. Thompson and J. G. Beckerley, The MIT Press, Cambridge, MA, 1973.

12.

L. Baker, "An Assessment of Existing Data on Zirconium 0xidation Under Hypothetical Accident Conditions in Light Water Reactors", Argonne National Laboratory Report ANL/ LWR /SAF-83-3, (1983).

4 13.

L. Baker and L. C. Just, " studies of Metal-Water Reactions at High Temperature: III. Experimental and Theoretical Studies of Zirconium-Water Reaction", Argonne National Laboratory Report, ANL-6548, (1962).

]

1 14 i

Y o r:

.s 14.

J. V. Cathcart and R. E. Pawel, Zirconium Metal-Water Oxidation Kinetics: IV Reaction Rate Studies, Oak Ridge National Laboratory Report, ORNL-NUREG-17, (August 1977).

15.

V. F. Urbanic and'T..R. He'idrick, "High-Temperature Oxidation of Zircaloy-2 and Zircaloy-4 in Steam",

J, Nucl.

Mater.

(75), pp. 251-261, (1978).

16.

J. T. Prater and E. L. Courtright, "High Temperature oxidation of Zircaloy-4 in Steam and Steam-Hydrogen Environments, Pacific Northwest Laboratory Report, NUREG/CR-PEL-5558, (Sept. 1985).

17.

J. F. White et al., " Seventh Annual Report: AEC Puels and Materials Development Program," GEMP-1004, (March 1968).

e 1

l 15 1

.. l

o

.s.

J

> q.

i Ppoject No. 697' DOE 1 Tritium Program i

cc:

Max Clausen 1

Office of Commercial Light-Water Reactor Production i

Tritium Project Office'

+

U.S. Department of Energy 1000 Independence ~ Avenue, SW Washington, DC 20585 DP-60' Records Management Office of Commercial' Light-Water.

)

Reactor Production

' Tritium Project Office

- U.S. Department of Energy 1000 Independence Avenue, SW Washington, DC 20585:

i Jerry _L. Ethridge, Sr. Program Manager Environmental Technology Division Pacific Northwest National Laboratory l'

Battelle Blvd.

P.O. Box 999 Richland, WA 99352 f'

5 4

>