substitution of zircaloy for stainless steel

7/25/2019 SUBSTITUTION OF ZIRCALOY FOR STAINLESS STEEL

1/82

^ ^ r a i i f f i i i i

3 Li b 31 5=555 1

O R N L 3 4 7 6

UC 80 Reactor Technology

TID 4500 22nd ed .

SUM M ARY REPORT

A N EVALUATION O F TH E SUBSTITUTION O F

ZIRCALOY FO R STAINLESS STEEL IN N S

SAVA N N A H FUEL ELEMENT C ONTAINER S

T . D . An d er so n

E. E. G r o s s

H. C McCurdyL. D . Scha ffe rL. R. S h o b e

C L. Whitmarsh

C EN TR AL R ESEA RC H LIBRARY

DOCUMENT COLLECTION

L I R R Y L O N C O P Y

DO NOT TRANSFER TO ANOTHER PERSON

f you wish someone else to see thisdocument send in n me wi th document nd th e library will rr nge a loan

O K R I D G E N T I O N L L O R T O R Y

o p e r a t e d b yU N I O N C R B I D E C O R P O R T I O N

f o r t h e

U S T O M I C E N E R G Y C O M M I S S I O N


2/82

Printed in USA Price: 2.00 Available from theO f fi ce o f T e c h ni c al S e rv i c es

U S Deportment of Commerce

Washington 25 D C

L E G L N O T I E

This report was prepared as an account of Government sponsored work. Nei t he r t h e United States,

nor the Commission, nor any person acting op behalf of t h e C o mmi ss i on :

A Makes an y warranty or representation, expressed or i mp li ed , w it h respect to t he a cc ur a cy,

completeness, or usefulness of th e information contained in this report, or that the us e of

any i nf or ma ti on , a p pa ra tu s, m et ho d, or p r oc e ss d i sc l o se d in this r ep or t m ay n ot i nf ri ng e

privately o wne d r i gh t s; or

B Assumes any liabilities with respect to th e us e o f or f or damages resulting from the us e of

any information, apparatus, method , or process disclosed in this report.

As used in the above, person acting on behalf of th e Commission includes a ny e m pl oy e e or

contractor of t he C o mm i ss i on , or employee of such contractor, to th e ex ten t t ha t s uc h employee

or contractor of the Commission, or employee of such contractor prepares, disseminates , or

provides acces s to , a ny i nf or ma ti on p ur su an t t o h is e mp lo ym en t or c on tr ac t w it h t he C om m is si on ,or hi s emp lo y men t wi t h su ch contractor,


3/82

O R N L -3 4 7 6

Con tr ac t No. W-7 40 5-en g-26

R e a c t o r D i v i s i o n

S UM MA RY R EP OR T

AN EVALUATION OF THE SUBSTITUTION OF ZIRCALOY FOR STAINLESSSTEEL IN N.S. SAVANNAH FUEL-ELEM ENT C ONTAINER S

T. D. A n d e r s o n

E. E. G r o s s

H. C. McCurdy

L. D. S c h a f f e r

L. R. S h o b e

C. L. Wh i t m ar s h

Date I s s u e d

OCT 23 19 53

OA K R ID GE N AT IO NA L L AB OR AT ORY

Oak Ridge, Ten ne ss eeoperated by

UNION C AR BI DE C O RP O RAT IO N

f o r t h e

U.S. ATOMIC ENERGY COMMISSIONMARTINKWfSMIIMIinm

SYSTEM5LIBBAMES

l3 4M5L 03^52 1


4/82


5/82

l l

C O N T E N T S

Page

ABSTRACT v

1. INTRODUCTION 1

2. SUMMARY, CONCLUSIONS, AN D RECOMMENDATIONS 5

Results of Investigations 5

Physics Calculations 5

Economic Evaluation 7

Compat ibi l i ty Stud ies 9

Structural Analysis 9

Hazards Analysis 10

Conclusions and Recommendat ions 10

3. REACTOR PHYSICS 11

An al y ti c al M o de l 12

Treatment of Cross Sections in Analytical Model 12

Treatment of Geometry in A nalytical Model 14

Comparison of Model Results with CriticalE xp er im en t D at a 23

Analytical Results 23

Beginning-of-Life Mul tip l ica t ion Factors 23

P o w er D i st r ib u ti o ns 26

R e ac ti vi ty L if et im e 27

Burnup of U235 and Plutonium Production 28

Two-Fuel-Zone Loading 30

4. F U EL -C Y CL E E C ON O MI CS 31

Results of Cost Analysis 32

E co nom ics o f E x t e n d e d C ore L i f e 32

Ec on omi cs o f R ed uc ed F ue l E n r i ch m en t 3 4

E f fec t of C o s t V ar iab le s 34

Fuel-Cycle Cost Components 36

Fabr ica t ion 3 7

Net Burnup 38

Reprocessing 38


6/82

V

Page

U se Charge 38

Working Capital 39

5. C O M PAT I B I L I T Y O F Z I R C A L O Y C O N TA I N E R S W I T H R E A C TO RS Y S T E M 3 9

C o r r o s i o n 3 9

We a r 4 1

Fretting Corrosion 44

B et ti s T es ts 4 5

C ha lk R iv er Tests 45

Hanford Tests 46

Conclusions 4 7

Hydrogen Embritt lement 47

Long-Lived Activity 50

6. ST RU CT U RA L ANALYSIS 52

Op erat i ng C on di ti on s 53

Flow an d P re ss ur e C on di ti on s 53

Ship Motion 56

Th er m al Condi t ions 56

Design Criteria 58

Method of Analysis an d Results ' 59

Selection of Important Stress Models 59

Results 63

7. H AZ AR DS A NA LY SI S 65

Method of Analysis 65

Results 66

ACKNOWLEDGMENT 68

Appendix. DESCRIPTION OF N.S. SAVANNAH REACTOR 69


7/82

V

A B S TR A C T

The possibility of reducing fuel-cycle costs for the N.S. SAVANNAH

by replacing the s ta in less s tee l fue l-element conta iners in the permanent reactor core structure by similar containers of a zirconium alloy

was investigated. These containers, although not integral parts of the

fuel-bearing components, are located wit hin the active core and divide

the core into 32 separate channels into which the fuel elements are

placed. Areas of investigation included reactor physics, fuel-cycle

economics, materials compatibility, structural design, and reactor haz

ards. A summary of the method of analysis and results is given for each

area of i nves ti ga ti on .

Calculations indicated that the substitution of Zircaloy containers

would increase core reactivity about 6fo Ak and control-rod worth about

A Ak. Fuel-cycle costs would be reduced about 26 . Zircaloy-4 appearsto be compatible with the reactor system, except for some uncertainty

with respect to fretting corrosion, which can be resolved only by tests.

The substitution of cold-worked Zircaloy for stainless steel in the con

tainer assembly would necessitate only minor design modifications. Al

though this evaluation is strictly applicable only to the N.S. SAVANNAH

reactor, the results demonstrate the feasibility and advantages of usingzirconium alloys for in-core capital-cost components.


8/82

S UM MA RY R EP OR T

AN EVALUATION OF THE SUBSTITUTION OF ZIRCALOY FOR STAINLESSSTEEL IN N.S. SAVANNAH FUEL-ELEMENT CONTAINERS

T. D. Anderson L. D. SchafferE. E. Gross L. R. ShobeH. C. McCurdy C. L. Whitmarsh

1. IN T R O D U C T IO N

The N.S. SAVANNAH is the world's first nuclear-powered merchant ship.

This combination passenger and cargo vessel is powered by a pressurized-

water reactor that was designed and built by the Babcock & Wilcox Company.

The reactor has a normal operating power rating of 63 Mw(thermal).From the standpoint of economics, the inherent features of a nuclear

propulsion system are high capital costs and potentially low fuel costs.

Although the N.S. SAVANNAHis not competitive, in an economic sense, with

conventional vessels, it does represent a substantial first step on whicha sound nuclear-powered ship program could be based. Economic nuclear

power for ships can be attained only by reducing capital costs and, at

the same time, exploiting the potential of low fuel-cycle costs. The

work summarized here was directed toward reducing fuel-cycle costs forthe N.S. SAVANNAH by replacing part of the permanent stainless steel

reactor core structure by a similar structure made of a zirconium alloy.

The parts of the reactor core structure considered were the fuel-element

containers, which are shown in Fig. 1. These containers, although notintegral parts of the fuel-bearing components, are located within the

active core. They provide 21 control-rod channels and 32 separate chan

nels into which fuel elements are placed. To give a clearer understand

ing of the makeup of the core, a brief description of the reactor isgiven in the Appendix. A more comprehensive description was presentedin the Final Safeguards Report for the N.S. SAVANNAH.1

1Babcock & Wilcox Company, Nuclear Merchant Ship Reactor FinalSafeguards Report - Volume I, Description of the N.S. SAVANNAH, USAECReport BAW-1164, June 1960.


9/82


10/82


11/82

P R O P O S E D C H A N G E IN N. S . SAVANNAH R E A C TO R

S U B S T I T U T I O N OF Z IR C AL O Y F O R

STAINL E SS S T E E L IN F UE L E L E M E NT CONTAINERS

A R E A S O F I NV ES T IG ATI ON

F U E L C Y C L E

EC O N O M I C S T U D Y

P H Y S I C S

C AL C UL AT IONSS T R U C T U R A L

A N A LY S I S

S T U D I E S OFC OM PA TI BI LI TY O F Z IR CA LO Y

WIT H R E A C T O R E NVIR ONM E NT

E C O N O M I C

EVALUATION T E C H N I C A L EVA LU ATI O N

O V E R - A L L EVA LU ATI O N OFT H E S U B ST I TU T IO N O F Z IR CA L OY

FO R S TA IN L ES S S T E E L IN

F UE L E L E M E N T CONTAINERS

U N C L A S S I F I E D

O R N L - L R - D W G 6 0 5 5 5 R

H A Z A R D S

A N A LY S I S

Fig. 2. Diagram Showing Scope of Zircaloy Evaluation.

A general ground rule of the evaluation was that no changes in the

reactor system s ho ul d be c on si der ed except those absolutely essential

for the Zircaloy substitution. On this basis the only method of reactor

control allowed was by control rods of the existing type. Similarly,

for the physics and economic portions of the evaluation, the only fuel

elements considered were those of the core-I configuration. In the

structural investigation, however, consideration was given to require

ments that might be imposed by future fuel-element designs. Both Zirca

loy-2 and Zircaloy-4 (low-nickel-content Zircaloy-2) were considered in

the compatibility portion of the evaluation. For the other parts of the

evaluation, it was assumed that the p ro pert ie s of these two zirconium

alloys are identical.

S S >Ss1SS*5--p*.


12/82

2. SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS

Results of Investigations

Physics Calculations

The nuclear characteristics of a reference core made up of stainless

steel fuel-element containers and core I-type fuel elements were computed

to serve as a basis for judging the changes associated with th e use of

Zircaloy fuel-element containers. It was, of course, intended that the

reference core be representative of core I. Th e main d if fe re nc e w as that

the reference core had a single-zone enrichment of 4.2 wt , whereas core

I has two enrichment zones (4.2 and 4.6 wt ). The difference between the

two was shown, however, to be inconsequential. A summary of the important

nuclear characteristics of the various cases studied is given in Table 1.

The reference core (case S-420), composed of stainless steel fuel-

element containers and 4.2 wt enriched fuel, gave a reactivity lifetime

of about 1.63 full-power years (full thermal power, 63 Mw). The worth of

all control rods in the reference core was about 16 Ak, giving a shut

down margin of a clean, cold core of something over 2 Ak. Since it was

fe l t t h a t a mo re s en si bl e s hu td ow n c ri te ri on w o u l d b e b a s e d o n th e m o s t

important rod stuck out of the core, the shutdown m ar gi n w it h the c en tr al

rod out was computed; the one-stuck-rod shutdown margin was 1.4 Ak.

Th e effect of substituting Z ir ca lo y f or stainless steel in the fuel-

e lement con ta in er s was well demonstrated by case Z-420. This core was

identical with the reference core, except for the use of the Zircaloy con

tainers. It ma y be noted that the reactivity lifetime i nc re as ed f r om 1.63

full-power years (fpy) to 3.20 fpy as a result of the Zircaloy substitu

tion. In addition to the increase in core reactivity, the control-rod

worth increased from 16 to 20 Ak. The increase in rod worth was not suf

ficient, however, to offset the 6 Ak increase in core reactivity. The

shutdown margin was less than 1 Ak, and the cold, clean core could not

be shut down with the central rod out. Thus case Z-420 would not be prac

t ical us in g the p re se nt c on tr ol rods.

A case of more practical interest is case Z-386. The enrichment of

this core, which had Zircaloy containers, was reduced to 3.86 wt so that


13/82

Table 1. Summary of Calcula ted Nuclear Characteri s ti cs

Case S-420: Reference case, core I-type fuel elements (uniform enrichmentof 4.20 wt ) in stainless steel containers

Case Z-420: Same as reference case except Zircaloy containers

Case Z-386: Core I-type fuel elements (uniform enrichment of 3.86 wt )in Z ir ca lo y containers; e nr ic hm en t a dj us te d to give sameshutdown margin with c en tr al r od out as reference case

Case Z-338: Core-I type fuel elements (uniform enrichment of 3.38 wt )in Zircaloy containers; enr ichment adjus ted to give samel i f e t i m e as r e f e r e n c e c a s e

O ne-zon e enr ichm en t, wt 2 3 5U'

Mul tip l i ca t ion fac tor of clean corewith rods out (followers in)

C o l d

H o t

Multiplication factor of clean corew i t h r o d s i n

C o l d

H o t

Cold worth of all control rods, Ak

Shutdown margin of cold, clean corewith central rod out, Ak

Lifetime, years at thermal power of63 M w

S t a i n l e s s S t e e l

Containers,R ef er en ce C as e

S - 4 2 0

4 . 2 0

1 . 1 3 2

1 . 0 7 9

Zircaloy Fuel-Element Containers

Ca s e Z - 4 2 0 C a s e Z 3 i Ca s e Z - 3 3 8

4 . 2 0 3 .8 6 3 .38

1 . 1 9 5 1 . 1 7 2 1 . 1 3 5

1 . 1 3 8 1.114 1.079

0 . 9 7 2 0 . 9 9 1 0 .9 6 9 0 . 9 3 4

0.882 0 . 9 1 0 0.886 0 .848

1 6 2 0 20 2 0

1. 4 Supercritical 1. 4 4. 9

1 . 6 3 3 . 2 0 2 . 5 2 1 . 6 3

c^


14/82

7

the shutdown margin with the central rod out was the same as for the

reference case, that is, 1.4 Ak. The computed reactivity lifetime of

this core was 2.52 fpy. Thus, using the same control rods as for the

reference case, an 8 reduction in enrichment gave the same shutdownmargin (with the central rod out) and a 55 increase in life relative tothat of t he r ef er en ce core.

Case Z-338 demonstrated the effect of the Zircaloy substitution on

enrichment. In this case it was possible to reduce the enrichment to

3.38 wt , a reduction of nearly 20 , and yet maintain the reference corelifetime of 1.63 fpy. Also, this core was just as reactive as the ref

erence core, and yet the shutdown margin was 4.9 Ak, as compared with

the 1.4 Ak of the reference core. This again demonstrated the signifi

cant increase in control-rod worth resulting from the Zircaloy substitut ion .

A calculation in x-y geometry of the power distribution for the ref

erence core with the control rods out (followers in) gave an overall ra

dial peak-to-average power ratio of approximately 2. Similar calculations

for the core with Zircaloy containers showed that the local peaking was

more severe but the gross radial distribution was more uniform than in

the reference core. These two effects combined to give a lmost identically

the same overall radial peak-to-average power ratio for the cores withZircaloy containers as for the reference case.

E co n o m i c E v a l u a t i o n

A steady-state fuel-cycle cost was calculated for the reference core

(case S-420) to provide a basis for comparing fuel-cycle cost reductionsresulting from the use of Zircaloy fuel-element containers. The reference

core was selected for this purpose in preference to core I to maintain

consistency with the physics analysis. It was found, however, that the

calculated fuel-cycle cost savings resulting from the Zircaloy substitu

tion were li ttl e affected by the choice between core I and the one-zone

reference core. A summary of the calculated cost reductions for the cores

with Zircaloy fuel-element containers is given in Table 2.

The cost savings for case Z-420 show the effects of the direct sub

stitution of Zircaloy for stainless steel. The results indicate that a


15/82

Table 2. Cost Reductions Resulting from Substituting Zircaloy' fo r Stainless Stee l in Fuel-Element Containers

Cost Reduction ( ) Based on Cost

for R efe r en ce C ase(case S-420)a

Cost ComponentCase Z - 4 2 0 C; ase Z-386 C a s e Z - 3 3 8

(: Lifetime, lifetime, Lifetime,

3, .20 fpy)

4 9

2 .52 fpy)

3 8

l, .63 fpy)

F a b r i c a t i o n 9

Ne t burnup < 1 4 2

Reprocessing 4 9 3 7 7

Uranium us e charge 7 15 2 6

Working cap ital 2 5 9

To ta l f uel cycle 3 2 2 6 1 0

aSee Table 1 for description of cases.

reduction in total fuel-cycle cost of about 32 could be realized for

this case. As pointed out before, however, case Z-420 is marginal with

respect to shutdown capability. Thus, it appears that different control

rods or a burnable poison would be necessary before the indicated cost

saving could be achieved in practice.

Case Z-386 could be controlled with the present control rods andwithout the use of burnable poisons. The estimated fuel-cycle cost re

duction for this case is 26 . This cost reduction is brought about by

both an increase in lifetime and a reduction in enrichment relative to

the reference core. Extension of the lifetime is, however, the more im

portant effect.

The economic effect of utilizing Zircaloy containers to reduce en

richment (keeping lifetime constant) is illustrated by case Z-338. De

creasing the enrichment from 4.2 to 3.38 wt resulted in a fuel-cyclecost saving of 10 . Although substantial, this cost reduction does not

compare favorably with the saving obtainable by utilizing the Zircaloy

containers to give increased reactivity lifetime.

It should be noted that most previous studies comparing Zircaloy

and stainless steel as core structural materials were of a rather general


16/82

nature and, consequently, were incapable of demonstrating the influence

of structural materials on control-rod worth. If the more general approach

had been used in this study, case Z-338 would have been the only one con

sidered; and, as a result, the economic potential of the Zircaloy substit ut io n w ou ld have been underestimated.

Compatibility Studies

Zircaloy-2 and Zircaloy-4 were evaluated for long-term service in

the core on the basis of a literature survey of basic properties of the

materials, an evaluation of the reactor water chemistry, and operating

experience with other pressurized-water reactors. It should be empha

sized, however, that existing data were obtained in relatively short-

duration tests, and uncertainties inherent in extrapolation of the data

are present. For 20-year service, the following conclusions were reached:

1. Corrosion resistance is excellent, with an indicated maximum

penetration of less than 2 mils on each exposed surface.

2. Based on rather limited data, no major wear problem is expected

as a result of using Zircaloy fuel-element containers.

3. The susceptibility of Zircaloy to fretting corrosion is uncertain,

and to resolve this problem, a proof test is needed.

4. Zircaloy-4 is recommended over Zircaloy-2 because of a lessertendency to pick up hydrogen. Hydrogen concentrations at 20 years of

261 and 104 ppm for Zircaloy-2 and Zircaloy-4, respectively, were cal

culated for 140-mil-thick plates.

5. Long-lived crud activity would not exceed the activity resulting

from the stainless steel components for which the primary coolant decon

t amin at io n sys tem was designed.

Structural Analysis

It was found that only minor design modifications of the fuel-element

container assembly would be required as a result of the Zircaloy substi

tution if 15 cold-worked Zircaloy were used. The most significant de

sign change would be an approximately 30 increase in thickness of the

c o n t a i n e r w a l l s .


17/82

1 0

Hazards Analysis

The only aspect of the Zircaloy substitu tion that might add to the

operational hazards would be the possibility of a zirconium-water reac

tion during the maximum credible accident (mca). It was conservativelycalculated that about 21 of the total zirconium in the core would be

oxidized during the mca. The energy released from the reaction would

not, however, contribute to the peak containment-vessel pressure because

of the time delay between the start of the mca and the initiation of the

zirconium-water reaction. It is concluded, therefore, that the Zircaloy

substitution would not significantly alter the consequences of the maxi

m u m cred ib l e acc iden t .

C o n c l u s i o n s a n d R eco m m en d a t i o n s

It is apparent from the results of this study that the use of Zircaloy

fuel-element containers would significantly reduce the fuel-cycle cost

of the reactor. It must, of course, be recognized that there are some

uncertainties in the evaluation given here. Recognition of these un

certainties forms the basis for a ddi ti ona l work which should be accom

plished prior to the use of Zircaloy containers. The two major areas in

which additional work is required are discussed below.1. Although the physics calculations were quite detailed, there are

uncertainties in such calculations. In order to obtain the m ax im um b en e

fit from the Zircaloy containers with the first new core loading, it is

thought that a critical experiment should be performed. The critical ex

periment would be useful as a desi gn tool to specify the maximum enrich

ment, and therefore lifetime, consistent with the minimum acceptable shut

down margin. Nevertheless, it should be noted that a critical experiment

would not be essential if something other than the optimum enrichment were

acceptable. Using the results of the physics calculations given here, it

should be possible to construct an acceptable core. Of course, if the

zero-power test of the core should indicate an insufficient shutdown mar

gin, reliance would have to be placed on reactivity shimming by means of

lumped poison.


18/82

1 1

2. With respect to the compatibility of Zircaloy with t he sys tem,

it was found that there was a possibility of the occurrence of fretting

corrosion between the fuel-element bundles and fuel-element containers.

It would be necessary to perform a loop test using a full-scale mockupof a fuel-element container and a fuel bundle to resolve this uncertainty.

3. R EA CTOR P HY SI CS

There is another important impetus, aside from the expected reduc

tion in U235 inventory, for investigating the substitution of Zircaloyfor stainless steel in the fuel-element container structure. Because of

the proximity of the containers to the control rods, substituting Zirca

loy for stainless steel in this region should result in an increase in

control-rod worth, since the flux in this region is not as depressed by

Zircaloy containers as by stainless steel containers. This effect was

demonstrated earlier with aluminum containers in the critical experiments. 2

An increase in control-rod worth would permit an increase in core

reactivity and consequently an increase in core reactivity lifetime.

Thus, a core with Zircaloy containers offers the attractive possibilityof reducing fuel-cycle costs by simply increasing core lifetime for a

given control rod shutdown requirement. It is to be emphasized that the

magnitude of this effect (i.e., control-rod worth dependence on container

material) depends very much on the details of this very heterogeneouscore and cannot be evaluated by a general type of study.

These considerations led to analyt ica l methods which were somewhat

detailed in both physics and geometry and which made extensive use of

available machine codes. Descriptions of the physical and geometrical

models employed in the nuclear evaluation of the container materials are

given below. The results obtained by using these models are then compared with critical experiment data, and, finally, results comparing

Zircaloy and stainless steel as container materials are presented.

R. M. Ball, A. L. MacKinney, and J. H. Mortenson, Marty CriticalExperiments - Summary of 4 Enriched U02 Cores Studied for NMSR, USAECReport BAW-1216, Babcock . Wilcox Co., May 1961.


19/82

1 2

Analytical Model

Treatment of Cross Sections in Analytical Model

In order to consider the two-dimensional aspects of the core designand to adequately include the effects of the control rods, a four-energy-

group method was used, since the av aila ble two-dimensional codes had afour-energy-group limit. (This limitation of two-dimensional codes has

recently been alleviated by the 18-energy-group ANGIE code.3) The group

cross sections for each materia l were f lux-averaged cross sections. The

thermal group cross sections were flux averaged over a hardened Maxwell-

Boltzmann distribution using an effec tive neutron tempera ture based on

the work of Brown.4 The water and control rod followers in the channels

between the fuel-element containers were assumed to be in an unhardened

Maxwell-Boltzmann flux distribution when the followers were present in

the channels. The dependence of the average thermal cross sections on

the moderator temperature was based on the work of Petrie, Storm, and

Zweifel.5

A multi-group slowing-down calculation using the GNU code6 was usedto obtain flux-averaged cross sections over the three nonthermal groups.

A typical slowing-down flux versus lethargy distribution giving the

limits of the three nonthermal groups is shown in Fig. 3. In the slowingdown calculation, the core was assumed to be homogeneous with respect

to the nonthermal neutrons. The very heterogeneous effect of resonance

absorption in U238 was accounted for by generating multigroup U238

3Stuart P. Stone, 9-ANGIE - A Two-Dimensional Multigroup NeutronDiffusion Theory Reactor Core for the IBM 709 or 7090, USAEC ReportUCRL-6076, University of California Radiation Laboratory, Oct. 28, 1960.

4H. D. Brown, Neutron Energy Spectra in Water, USAEC Report DP-64,Du Pont de Nemours (E.I.) & Co., February 1956.

5C. D. Petrie, M. L. Storm, and P. F. Zweifel, Calculation ofThermal Group Constants for Mixtures Containing Hydrogen, Nuclear Sci.Eng., 2: 728 (1957).

6J. M. Bookston, C. L. Davis, and B. E. Smith, GNU, A MultigroupOne-Dimensional Diffusion Program for the IBM-704, General MotorsCorporation Report, April 8, 1957.


20/82

6


ORNL-LR-DWG 5 9 8 4 8

5

f4

| \Z3 . 1 1 GROU 1 GROL ip ? BROUPl Jr c C 3LUcc

^-2

i

i

i

4

/0 8 10

u, LETHARGY12 14 16

Fig. 3. Slowing-Down Flux Versus Lethargy D i st ri b ut io n f o r Core I at 68F.

18


21/82

1 4

absorption cross sections that were consistent with a Dancoff-corrected7resonance integral8 appropriate to the N.S. SAVANNAH reactor geometry.

Other important physical effects accounted for in the determination of

the slowing-down flux versus lethargy distribution (Fig. 3) were(l) Goertzel-Selengut hydrogen slowing down, (2) inelastic-scattering

slowing down, (3) leakage from the core, (4) absorptions in all materials

on the basis of available resonance integrals,9 and (5) nonthermal fis

sions in U235 consistent with differential cross-section data and consist

ent with the measured fission resonance integral.9

Treatment of Geometry in Analytical Model

The four-group cross sections so obtained were used to determine

flux distributions throughout the reactor. For purposes of the physics

calculations, the nature of the core geometry is well illustrated by the

horizontal section through a ferrule plane shown in Fig. 4. The geometry

of Fig. 4 was, of course, much too complicated and needed to be greatly

simplified before gross flux distributions throughout a full-size core

could be computed.

The first step in simplification was to determine flux distributions

across a so-called typical fuel-pin cell. The objective here was to re

duce the complexity of the core by replacing the fuel-pin region by anequivalent homogeneous medium. The repetitive arrangement of fuel pins

suggested defining a unit fuel-pin cell whose area was the fuel-pin pitch

squared. For an infinite array of fuel pins the net current across a cell

boundary should be zero, and a reasonable approximation to this boundary

was the surface of a cylinder whose axis was coincident with the fuel-

pin axis. The geometry of a fuel-pin cell giving details of the pellet,

gap, and cladding sizes for a core I fuel pin is shown in Fig. 5. The

7J. A. Thie, A Simple Analytical Formulation of the Dancoff Correction, Nuclear Sci. Eng., 5: 75 (1959).

8E. Hellstrand, Measurements of the Effective Resonance Integralin Uranium Metal and Oxide in Different Geometries, J. Appl. Phys.,28: 1943 (1957).

9R. L. Macklin and H. S. Pomerance, Resonance Capture Integrals, Progr. in Nuclear Energy, Vol. 1, p. 179, McGraw-Hill, New York, 1956.


22/82

ZIRCALOY

CONTROL ROD FOLLOWER

^ E

0 . 3 7 5 in.

- 0.187 in.

STAINLESS STEEL

^SPACER BAR(FLOW BLOCK)

FUEL ELEMENT

CO N TA IN ER

8 . 9 8 8 in.

CONTROL ROD (BORON STAINLESS STEEL)

Fig. 4. Details of Fuel Cell.

UNCLASSIFIED

O R N L- LR - D WG 4 3 8 8 9 R

n U i


23/82

S TA IN L E S S

S T E E L

1 6


O R N L - L R - D W G 5 9 8 4 6

PI N C E L L

B O U N D A RY

W A T E R

Fig. 5. Details of G eo me tr ic al M od el U se d fo r Fuel-Pin Cell.

f ou r- gr o up f lu x d is tr ib ut io ns s at is fy in g t he a bo ve a ss um pt io ns a re shown

in Fig. 6. The data presented were obtained using the SN G transport

theory code10-'11 in the S-4 approximation, since diffusion theory was

not expected to a pp ly w el l in t he s tr on gl y a bs or bi ng UO2 region.

The data of Fig. 6 indicate that the fuel-pin region can be replaced

by a single homogeneous region. A similar calculation with an additional

10B. G. Carlson, The Sn Method and the SNG and SNK Codes, USAECReport LASL T-l-159, Los Alamos Scientific Laboratory, Jan. 26, 1958.

11B. G. Carlson and G. I. Bell, Solution of the Transport Equationby the Sn Method, Proceedings of the Second United Nations InternationalConference on the Peaceful Uses of Atomic Energy, Geneva, 1958, Vol. 16^p. 535, United Nations, New York, 1958.

~* t t s i ^ S r f f l ^ H B i t M t M aK M a r t o ^ l M u ^


24/82

1 7

U N C LA S S I F I EDO R N L - L R - D W G 6 2 8 1 7 R

7 .0

6 . 0

.

h r

V --


25/82

1 8

in the core. Although a fuel-element container cell was obviously two

dimensional, man y features of this cell are well illustrated by the re

sults obtained fro m a one-dimensional slab geometry treatment. Agai n

using symmetry arguments, solutions can be obtained for the flux distributions in an infinite array of slab fuel elements. The resulting

S-5 flux distributions for the case when control rods were present in

the channels between containers is shown along with the slab geometry

in Fig. 7. These slab geometry results should be representative of the

flux distributions several mean free paths (thermal mean free path, ~1 cm)

from a corner, and the effect of the control rod on the four-group fluxes

is well demonstrated. The strong nonthermal absorption properties of the

control rods revealed by Fig. 7 may also be illustrated by the energy

distribution of n e ut ron a bsorpt io ns in the control rod as shown in

Table 3. It is to be noted that, according to these results, only about

one-fourth of the neutrons absorbed by the control rods are thermal neu

t r o n s .

Aside from d es cr ib in g the g en er al c har ac te ris ti cs of the control

rods, the data of Fig. 7 served the important function of providing

2 4 6 8 10

x DISTANCE FROM CONTROL ROD CENT ERL INE ( c m )

U N C LA S S I F I EDORNL - L R- DWG 4 9 8 6 1 R 3

Fig. 7. S-5 Flux Distributions at 68F w ith Control Rods Presenti n Ch ann el s B et we en F ue l- El em en t Containers.


26/82

1 9

Table 3. Energy Distr ibut ion ofContro l-Rod Absorpt ions

C o n t r o l - R o d

Group Energy Interval Absorptions

1 9.1 x 103 ev < E < 107 ev 8

2 15.17 X ev < E < 9.1 x 103 ev 29

3 0.15 ev < E < 15.17 ev 37

4 0 < E < 0.15 ev 26

transport-theory boundary conditions at the surface of the control rod.13

These boundary conditions were then used to replace the control rods in

a two-dimensional diffusion-theory calculation of a fuel-element-container

cell (an x-y geometry transport theory code was not available for these

calculations). This procedure provided a consistent method for treatingthe transport properties of control rods, with due respect for the hetero

geneities in the control-rod neighborhood.

Similar results for Zircaloy followers in the channels between con

tainers are shown in Fig. 8. These data illustrate the severe thermal

flux peaking caused by the water channels, and the important effect of

this peaking on the flux level in the container material is evident. The

use of thermal cross sections averaged over a Maxwell-Boltzman distribu

tion for the channel region and thermal cross sections averaged over a

hardened Maxwell-Boltzman distribution for the container and fuel regions,

as described above, may not give the best results for the thermal flux

distributions in the channel area.14;15 This method was consistently

applied, however, and the effect on thermal flux and reactivity of

13B. W. Colston, E. E. Gross, and M. L. Winton, Heterogeneous Control Rod Studies, Proc. ANPP Reactor Analysis Seminar, October 11-12, 1960,USAEC Report MND-C-2487, Martin Co., Nuclear Division, January 1961.

14W. B. Wright and F. Fenier, Note on Position Dependent Spectra,Nucl. Sci. Eng., 6: 81 (1959).

15G. P. Calame, A Few-Group Theory of Water Gap Peaking, Nucl.Sci. Eng., 8: 400 (i960).


27/82

3 2-e -

- Z IR C A L O Y F O L L OW E R

- WAT E R C H A N N E L

- S TA I NL E S S S T EE L C O N TA I N E R

2 0

U N CL A SSIFIE D

O R N L- LR - D WG 5 0 3 7 8 R 3

- H O MO G E N IZ E D F U E L -

, 15.17eu


28/82

MINA-I LE R

2 1

?

iZZZZZ

UNCLASSI FI EDO RN L -L R-D WG 4 3 8 9 0 R 2

mJ ~ 7

' Z I RC A L O Y F O L LO W E R

- - S TA I N LES S S T E E L ORZIRCALOY C O N TA I N ER

H O M O GE N I ZE D F U E L

PIN REGION

- Z IR C A L OY F O LL O WE R

i -4- STAINLESS S TEEL ORZIRCALOY C O N TA I N ER

?

STAINL E SS S T E E L

Fig. 9. Ge omet r ic al M od e l for One Quadrant of a Fuel-Element Co ntainer Cell with Control-Rod Followers in the Channel Region.


O R N L - L R - D W G 4 9 8 6 2 R

REGIONS:

HOMOGENIZED

HOMOGENIZEDF U E L

HOMOGENIZEDC H A N N E L WAT E RAN D CONTAINER

FOLLOWER ORC O NT RO L R O D

(5) WATER REFLECTOR

Fig. 10. Ge ome tr i ca l M o de l for One Quadrant of Core I.


29/82

2 2

neutrons in th e a xi al d ir ec ti on wa s a cc ou nt ed f or by specifying a group-

independent axial buckling, which was o bt ai ne d from the cri ti ca l exp er i

ments.16 The results obtained from a calculation in the geometry of

Fig. 10 were used to homogenize the ent ir e core, and a calculation ofth e full-sized core in r- z geometry wa s then performed. I n this wa y

leakage from all sides of the core was treated explicitly.

One wa y of illustrating th e detail contained in th e method just out

l ined is shown in Fig. 11, which presents t he average f ou r- gr ou p fl ux

levels in each of th e regions of core I at th e operating temperature.

All fluxes were normalized to unity in the UO2 region. It is to be

n o t e d t h a t t h e t h e r m a l f l u x l e v e l i n t h e s t a in l es s s t e e l c o n t a i n e r is

16 NMSR Quarterly Technical Report, July-September 1959, BAW-1180,pp. 23 ff.

0. 5

_l

NO F E R R U L E

R E G I ON C L AD D I NG

NO F E R R U L E

R E GI O N WATE R

F E R R U L E R E G I O N

C L A D D I N G

S TA I N LES S S T E E L

F E R R U L E

F E R R U L E REGION

WATER

S TA I N L E S S S T E E L

C O N T A I N E R

S T A I N L E S S S T E E L

F L O W B L O C K

C H A N N E L WATER

Z I R C A L O Y

F O L L O W E R

A L U M I N A F I L L E R

S TA IN L E SS S T E EL

C E N T R A L T U B E

C E N TR A L T U BE

WAT E R

L l

I I

1.5

I

U N CL A SSIFIE DO R N L - L R - D W G 5 9 8 4 7 R 2

FAST, 9 .1 X1 0 ev < < 1 0 ev

Vf , 15.17ev < < 9.1 x 103evEPITHERMAL,

0 15ev


30/82

2 3

considerably higher than the thermal flux level in the stainless steel

cladding.

Comparison of Model Results with Critical Experiment Data

The four-group method, which is described in more detail elsewhere,17

was applied to some critical experiment data2 that compared aluminum fuel-

element containers with stainless ste el containers in a core with 16

identical fuel elements. A comparison of measured and calculated multi

plication factors for various control-rod patterns in Marty II (a core

with stainless steel containers) and Marty III (a core with aluminum con

tainers) is shown in Table 4. The comparison shows the sizeable increase

in both core reactivity and control-rod worth resulting from the substi

tution of aluminum for stainless steel. The ability of the analytical

method to estimate the effects of material exchange is also evident from

Tab l e 4.

Ana ly ti ca l R esul ts

Beginning-of-Li fe Mul tip l ica t ion Factors

Ap pli cat ion of the fo ur- grou p method to core I fuel elements within

stainless steel containers and within Zircaloy containers yielded the

multiplication factors shown in Fig. 12. These data are for single-fuel-

zone cores at 68F. The effect of two-fuel-zone loading was not impor

t an t in th i s evalua t ion an d is d iscussed later. C al cu la ti on s o f t w o e n

richments (3.0 and 4.2 wt U235) were made for the Zircaloy containers.

The shape of the curve between 3.0 and 4.2 wt U235 enrichments was

taken from the results of a previous survey18 using a somewhat simpler

17E. E. Gross, B. W. Colston, and M. L. Winton, Nuclear Analysesof the N.S. SAVANNAH R ea ct or w it h Zircaloy or S ta in le ss S te el as Fuel-Element Containers, USAEC Report ORNL-3261, Oak Ridge NationalLaboratory, April 24, 1962.

l8Memo from J. A. Bast et al., Massachusetts Institute of TechnologyEngineering Practice School, to E. E. Gross, Oak Ridge National Laboratory,

Comparison of Zircaloy and Stainless Steel as the Structural Materialsin the Core of the N.S. SAVANNAH, Nov. 9, 1959.


31/82

2 4

Ta b le 4 . Compar i son o f C a l c u l a t e d and E x p e r i m e n t a l M u l t i p l i c a t i o nF a c t o r s f o r M a r t y I I an d M ar ty I I I C r i t i c a l E x p e r i m e n t s

U N C LA SSIFIEDORNL-LR-DW G 5 2 9 7 0 R 2

M A RT Y I I M A RT Y II I

R OD PAT TE RN EXP s CALC EXP s CALC

1 . 0 4 3 + 0 . 0 0 1 1 . 0 5 5 1 . 1 6 + 0 . 0 2 1 . 1 8 0

1 . 0 2 4 + 0 . 0 0 1 1 . 0 3 0 1 . 1 3 + 0 . 0 2 1 .1 4 8

O i D D

qpn *

1 . 0 0 11 + 0 . 0 0 0 1 1 . 0 0 0 1 . 0 9 + 0 . 0 1 1 . 0 9 9

D i D O DD l I D l ]

DjnnjDm m

1 . 0 1 3 + 0 . 0 0 1 1.021 1 .11 + 0 .0 1 1 . 1 2 6

analytical method. It is apparent from Fig. 12 that substitution of

Zircaloy containers for the present stainless s teel containers would in

crease the reactivity of 4.2 wt fo U235 core I fuel elements by about

6 Ak.

The worth of the control rods may be obtained from Fig. 12 by subt ra ct in g t he m ul ti pl ic at io n f ac to r of th e core with control rods from

t he m ul ti pl ic at io n f ac to r of th e core with control-rod followers. The

rod worth obtained in this way is remarkably independent of enrichment

in a core with Zircaloy containers, at least in the enrichment range of

3.0 to 4.2 wt U235. Although the 21 control rods appear to be worth


32/82

0 . 9 5

0 . 9 0

2 5


O R N L - L R - D W G 5 9 8 4 0 R

3 .8 4 .0

FUEL ENRICHMENT ( w t U )

Fig. 12. Calculated Multiplication Factors at 68F for Core I TypeFuel Elements in Zircaloy and in Type 304 Stainless Steel Containers atBe gin ni ng of Life.

A Ak more in the core with Zircaloy containers than in the core with

s ta in less s teel co ntaine rs, this ro d worth is barely enough to control

4.2 wt enriched core I fuel elements. In order to obtain the same

shutdown margin as is present in the core with stainless steel containers,

the core with Zircaloy containers would require 3.9 wt fo enriched fuelelements. Since the control rods are worth more in a core with Zircaloy

con ta ine rs , a better sh ut d own c ri t er i on would be to r equ ir e t he same

shutdown margin with the central rod stuck out of th e core as there is

in t h e co re w i th s t a in le s s s t ee l containers . B a s e d o n t h e d at a of Tab l e 4

and Fig. 12, the one-stuck-rod criterion for the shutdown margin would be

satisfied with an enrichment of about 3.86 wt U235.

Calcu lat ed mul tip l i ca t ion fac tor s fo r full-power conditions at th e

beginning of core life are shown in Fig. 13. By comparing Fig. 13 with

Fig. 12, it may be seen that there is an overall temperature deficit

(loss of reactivity in going from room temperature to full power) of

5.3fo Ak for the core with stainless steel containers and 5.7 Ak for the

core with Zircaloy containers. Also, it is apparent from Fig. 13 that

a fuel enrichment of 3.38 wt U235 in a core with Zircaloy containers


33/82

1 .15

1 . 0 5

^ 0.95

0 . 9 0

0 . 8 0

3 .0 3 .2 3 .4

2 6


O R N L - L R - DW G 5 9 8 3 9 R

' S TAI N L E SS S T EE L CONTAINERSW IT H R OD S

3 .6 3. 8 4 .0 4 .2

F U EL E N RI C HM E NT ( w t 7 U )

Fig. 13. Calculated Multiplication Factors for Core I Type FuelElements in Zircaloy and Type 304 Stainless Steel Containers at Beginningof Life for Full-Power Operation at 508F Without Xenon.

would have the same initial hot reactivity (and therefore about the same

reactivity lifetime) as 4.2 wt enriched fuel in a core with stainless

s t e e l c o n t a i n e r s .

P o w e r D i s t r i b u t i o n s

The c al c ul a te d p owe r d is tr ib ut io ns of th e cores with s ta in le ss s te el

and Zircaloy containers are compared in Fig. 14. Since Zircaloy is a much

weaker thermal-neutron absorber than stainless steel, the power peak is

higher near the water channels for the Zircaloy containers than for the

stainless steel containers. The increased local power peaking caused by

the Zircaloy containers is offset, however, by a gross radia l p ower flat

tening. Power profiles along directions other than that shown in Fig. 14

have similar characteristics. Although the results of Fig. 14 are en

couraging, the three-dimensional power peaking for partially inserted

rods needs to be investigated.


34/82

2 7


O R N L - L R - DWG 5 9 8 4 9 R 2

2 0 3 0 4 0 5 0 6 0

x, DIS TANC E FR OM R EACTOR CENTER LINE (cm)

Fig. 14. Comparison of Power Distribution for 4.2 wt fo EnrichedCore I Fuel Elements in Zircaloy and in Type 304 Stainless Steel Cont a iners at 508F.

Reactivity Lifetime

Simple one-dimensional radial burnup calculat ions were performed

using the CANDLE 2 code19 to estimate the reactivity lifetime of a core

with Zircaloy containers relative to core I. Four-group cross sections

representing a core at the beginning of life were obtained from homoge

nization of the results for a full-sized core (geometry of Fig. 10) con

taining only control-rod followers. For the burnup calculations the core

was controlled by a uniformly distributed poison, and the microscopic

cross sections of the core materials were assumed to be constant in time.

The ful l-p ower mu lt i pl i ca t io n fac to r as a function of t im e c al cu la te d

with CANDLE 2 (with the above assumptions) for an initial loading of4.2 wt fo enriched fuel in both stainless steel and Zircaloy containers

is shown in Fig. 15. After equilibrium xenon and samarium have built

up (~0.3 year), the multiplication factor as a function of time ap pearsto be remarkably linear, with essentially the same slope for both cores.

19D. J. Marlowe and P. A. Ombrellaro, CANDLE - A One-DimensionalFew-Group Depletion Code for the IBM-704, USAEC Report WAPD-TM-53,Westinghouse Atomic Power Division, May 1957.


35/82

2 8

U N C L A S S I F I E DO RN L -L R-D WG 5 9 8 4 1 R 3

I

i^^ZIRCALOY CONTAINERS

^ ^-STAINLESS STEELCONTAINERS

M A X I M U M X E N O N \V OVERRIDE -____~^\

i ^N^J0 0.5 1.0 1.5 2.0 2.5 3.0

FULL-POWER YEARS O F O PERATION AT 63 Mw

Fig. 15. Calculated Reactivity Lifetime of 4.2 wt fo Enriched Core IFuel in Zircaloy and in Type 304 Stainless Steel Containers at a PowerL e v e l of 63 Mw .

After about two years of burnup at full power, the buildup of plutonium

begins to affect the burnup curve slightly.

A xenon override option in CANDLE 2 reveals a maximum xenon over

ride of O.lfo Ak over and above the reactivity tied up in equilibrium

xenon (l.2 Ak). Assuming that the end of core life occurs when maximum

xenon cannot be overridden, the core with s ta in less s teel conta iners

would last 1.63 full-power years, while the core with Zircaloy containers

would last 3.2 full-power years, or about twice as long as the stainless

s t e e l c o r e .

As pointed out earl ier, a core with Zircaloy containers would not

have as large a shutdown margin as a core with stain less steel containers

with the same fuel enrichment. In order to obtain the same shutdown

margin with a central rod removed as is now available in core I would

require about 3.86 wt fo U235 in Zircaloy containers. Based on the data

of Fig. 13, which provide an estimate of the initial hot reactivity for

this enrichment, and the slope of the burnup curve of Fig. 15, sucha

core is estimated to have a reactivity lifetime about 0.9 full-power

years longer than core I.

Burnup of U235 and Plutonium Production

The burnup of U235 and the production of plutonium are also impor

tant considerations in fuel-cycle economics. The U235 inventory as a


36/82

2 9

function of core l ife ( fu ll -power years) is shown in Fig. 16. T hese data

apply to an initial one-zone loading of 4.2 wt fo enriched fuel in either

stainless steel or Zircaloy containers. The burnup of U235 is fairly

linear, with a slope of about 25 kg per year. The buildup of Pu239 andPu241 for both cores is shown in Figs. 17 and 18. The buildup of Pu239occurs at a rate of about 8 kg per year during the first year and at the

rate of about 6 kg per year during the second year of full-power opera

tion. The buildup of Pu241 is quadratic, depending as it does on the

buildup of intermediate isotopes of plutonium.

The estimate of the buildup of plutonium depends directly on the

resonance integral employed for U238 and therefore contains the usual

uncertainties encountered when evaluating the resonance integral of lumps

3 0 0

2 8 0

-2 2 6 0

LU

o r

oo

2

if 2 4 0r oCM

3

2 2 0

2 0 01.0 1.5 2.0

F U L L P O W E R Y E A R S

U N C L A S S I F I E DORNL-LR-DWG 5 9 8 4 3

2. 5 3.0

Fig. 16. Burnup of U235 in 4.2 wt fo Enriched Core I Fuel Elementsa t a P o w er Level o f 63 Mw.


37/82

LUrr

ou

enr o

Q .

3 0

UNCLASSIFIEDO R N L - L R - D W G 5 9 8 4 4

2 5

2 0

1 5

1 0

5

n3.0

0 0 . 5 1.0 1. 5 2 .0

FULL POWER YEARS

2.5

Fig. 17. Buildup of Pu239 in 4.2 wt fo Enriched Core I Fuel Elementsat a Power Level of 63 Mw.

of U238 with nonuniform temperature distributions in a heterogeneous lat

tice. The same resonance integral (20.35 barns at full power) was used

in both burnup calculations, however, and therefore the production rates

of plutonium in both cores are on the same basis.

Two-Fue l-Zon e Lo ad i ng

The calculations described above were based on cores initially con

taining fuel elements with a single fuel enrichment. In the actual

core I, the inner 16 fuel elements initially contained 4.2 wt fo U235,whereas the outer 16 fuel elements contained 4. 6 wt fo U235. This two-

zone loading arrangement helped to flatten the radial power distribution

and provided a method of trimming core reactivity upward, if necessary.

By repeating some of the calculations for two-zone loading, it was

fou nd that the cold reactivity of the core wit h sta inles s st ee l cont ainers


38/82

3 1

1 0 0 0


O R N L - L R - D W G 5 9 8 4 2

8 0 0

EDk_

6 0 0UJ

c c

ou

5 40 0CM

0 .

2 0 0

n

0 0 . 5 1. 0 1. 5 2 . 0

FUL L P O W E R Y E A R S

2 . 5 3 . 0

Fig. 18. Buildup of Pu241 in 4. 2 wt fo Enriched Core I Fuel Elementsa t a P o w e r L e v e l of 6 3 Mw .

was increased by about Ak = 0.003 w hen compared with one-zone loading.

Thus the reactivity data for uniform fuel lo ad in g w ou ld be little af

fected by con side ratio n of two-zone fuel loading. Consequently, since

t wo -z on e l oa d in g would not play an important role in evaluat ing conta iner

materials, evaluation on the basis of a uniform fuel loadi ng ap peare d to

be justified.

4 . FUEL-CYCLE ECONOMICS

In order to evaluate the s av in g t ha t would result from the Zircaloy

substitution,20 it was first necessary to establish a steady-state fuel-

20C. L. Whitmarsh, Potential Fuel-Cycle Cost Savings Resulting fromTechnological Changes in the N.S. SAVANNAH Reactor, USAEC Report ORNL TM-144, Oak Ridge National Laboratory, April 6, 1962.


39/82

3 2

cyc le cost fo r a reference core. T he r ef er en ce core, designated case

S-420, consisted of core I-type fuel elements (one-zone enrichment of 4.2

wt fo in stainless steel fuel-element containers. This core was selected

in preference to core I (two fuel zones) to maintain consistency with thephysics analysis. Calculations showed, however, that the estimated fuel-

cycle cost savings resulting from the Zircaloy s ub st it ut io n w er e l it tl e

affected b y whether t he o ne -z on e core or core I was chosen as the refer

ence c o r e

Zircaloy containers ca n be utilized to red uc e fu el -c yc l e costs by

either increasing the core lifetime or decreasing enrichment (keeping

lifetime constant). In order to examine the relative merits of these two

approaches, a cost analysis of three different cores with Z ir ca lo y co n

tainers was necessary. The results of the fuel-cycle cost analysis fo r

t hes e t hree cases ar e presented below. Following th e presentation of re

s u l t s is a d i s c u s s i o n o f t h e e f f e c t s o f v a r i o u s c o s t v a r i a b l e s a n d u n c e r

tainties. Finally, the basis on which each cost component was obtained

is given.

Results of C os t A na ly si s

E co no mi c s o fE x t e n d e d

C o r e L i f e

Results of the physics calculations (see Section 3) indicated that

core lifetime would be extended from 1.63 to 3.20 full-power years if no

changes were made in the core other than the su bst itu tio n of Zircaloy for

stainless steel. This material change, in addition to i nc re as in g exce ss

reactivity, a lso i nc rea se d control-rod worth; however, the increase in

control-rod worth was not sufficient to give the same degree of shutdown

safety as the reference core. In order to give the same shutdown margin

with the c en tr al c on tr ol ro d out as would obtain with th e reference core,

the fuel enrichment to be used with the Zircaloy containers should be re

duced to 3.86 . For this case, the core lifetime would be 2.52 full-power

y e a r s .

The fuel-cycle cost savings are presented in Table 5. If the total

excess reactivity is utilized for lifetime extension (case Z-420), a re

duction of 32 in fuel-cycle cost may be realized. If the shutdown margin


40/82

3 3

Table 5. Effect of Ex te n de d L ifet ime onFuel-Cycle Costs8-

Cos t R e d u c t i o n { Based onC o s t f o r R e f e r e n c e

(Case S-420)bC a s e

C a s e Z- 420c Casi e Z-386d(lifetime,

3.20 fpy)(lifetime,

2.52 fpy)

F a b r i c a t i o n 4 9 3 8

Ne t b u r n u p < 1 4

Reprocessing

Uranium use charge

4 9

7

3 7

1 5

Working capital 2 5

To ta l fu el cycle 32 2 6

aAssumed load factor of 0.67.

^Case S-420: Reference case, core I-type fuelelements.(uniform enrichment of4.20 wt fo in stainless steelc o n t a i n e r s .

cCase Z-420: Same as reference case exceptZircaloy containers.

dCase Z-386: Core I-type fuel elements (uniform enrichment of 3.86 wt fo inZircaloy containers; enrichmentadjusted to give same shutdownmargin with central rod out asr e f e r e n c e case .

(with central rod out) of the core utilizing Zircaloy containers is ad

justed to be equivalent to the shutdown margin of the reference core, the

cost saving is 26 . The additional cost of Zircaloy containers compared

with stainless steel containers is not included in the above analysis. A

cost estimate indicated that the Zircaloy containers would cost about 130,000 more than the stainless steel containers. This differential is

attributable to higher material and labor costs. When amortized over the

20-year service life, however, this additional cost for Zircaloy appears

insignificant compared with annual fuel-cycle costs of 700,000 to

1,000,000.


41/82

3 4

E co no mi cs o f R ed uc ed F ue l E nr ic hm en t

As an alternative approach , the initial fuel enrichment may be re

duced and still maintain the reference core lifetime. The reduced en

richment, as indicated in Section 3, was calculated to be 3.38 . As shown

in Table 6, the resulting reduction in fuel-cycle cost was 10 of the ref

erence fuel-cycle cost. As in the previous analysis, the additional cost

of the Zircaloy c onta iner s was not included.

Table 6. Effect of R e du c ed E n ri c hm e nton Fuel-Cycle Costs8

Cost Reduction \-/o

for Case Z-338bB a s e d o n C o s t fo r

R ef er en ce C as e

Fa b r i c a t i o n 9

Net burnup 2Reprocessing 7Uranium use charge 26Wor ki ng c ap it al 9

Total fuel cycle 10

Ass umed load factor of 0 .67

Case Z-338: Core I-type fuel elements(uniform enrichment of3.38 wt fo in Zircaloycontainers; enrichmentadjusted to give samelifetime (1.63 fpy) asr e f e r en ce case.

E ff e c t of C os t Var ia bl es

The relative importance of each cost component to the total fuel-cycle cost for the reference core is presented in Table 7. It is to be

noted that greater than 60 of the reference fuel-cycle cost is repre

sented by components with costs that are inversely proportional to core

lifetime, that is, fabrication and reprocessing. Thus it appears that

lifetime extension represents the greatest potential of any single change


42/82

3 5

Table 7. Estimated Relative Importance of Fuel-Cycle Cost Components

Cost ( of total fuel cycle)

Zircaloy Fuel-Element Container CasesRefe renceCase S-420 Cage Z_A2Q CaQe z_3g6 Case z_33g

F a b r i c a t i o n

Ne t b ur nu p

Reprocessing

Uranium use charge

Wor ki ng c ap it al

3 9

1 6

23

1 5

7

2 9

2 3

1 7

2 1

1 0

3 3

2 1

2 0

1 7

9

3 9

1 8

2 4

1 2

7

for reducing the reference fuel-cycle cost. Use charge, an enrichment-

dependent cost, represents 15 of the total cost; burnup and working

capital are virtually independent of any changes under consideration.

The load factor, fabrication costs, and uranium burnup represent the

greatest uncertainties in fuel-cycle cost calculations. The load factors

can only be estimated on the basis of past experience with similar situa

tions. In many respects the SAVANNAH is unique; and, consequently, any

estimated load factor will be primarily a guess. The effect of the load

factor on cost reduction is shown in Fig. 19 to be small; for example,

a reduction of the load factor from 0.67 to 0.3 reduces the Zircaloy corefuel-cycle saving from 32 to 26 . Fabrication costs will become more

definitive as industry gains experience in this field. At present, data

must be extrapolated from similar operations to obtain unit costs, which

are then combined to obtain total core cost. These costs are sensitive

to many factors, such as dimensional tolerances, chemical specifications,

and hardware complexity. As also shown in Fig. 19, the effect of core

procurement cost is small; for example, halving the fabrication cost re

duces the Zircaloy core fuel-cycle saving from 32 to 29 .

Burnup values may be refined by improved calculational techniques but

can be determined precisely only by actual reactor operation. Increased

burnup will decrease the relative importance of the lifetime-dependent

cost components - fabrication and reprocessing. Lifetime calculations

are complex, however, and the resulting change in fuel-cycle costs can

not be evaluated quantitatively without extensive additional work.


43/82

3 6


O R N L - L R - D W G 6 5 5 1 7

0 . 43 6

CORE PROCUREMENT COST (FRACTION OF REFERENCE CORE)

0.6 0.8 1.0 1.2 1.4 1.6 1.8 2. 0

3 4

oo

w 3 2

o

00 o

< 30CO

LU

P ooo zO LUo cc

LU

i_ 28

5s- 2 6

2 4

C0 F E PROCUREMENT COST E F F E C T- - .

^ LOAD FACTOR EFFEC T

R F F F R F N C F C O R F

LIFETIME = 1.63yr AT 63 Mw

CORE WITH ZIRCALOY CONTAINERSLIFETIME = 3. 2yr AT 63 Mw

0.2 0.3 0.4 0.5 0.6 0.7

L OA D F AC TO R

0. 8 0. 9 1.0

Fig. 19. Effects of Load Factor and Fabrication Cost Estimates onFuel-Cycle Cost Savings Resulting from Use of Zircaloy Fuel-Element Cont a i n e r s .

Indications are that, for longer reference core l ifetimes (>1.63 full-

power years), the estimated relative full-cycle cost saving for the

Zircaloy cases would be reduced; conversely, for shorter reference core

lifetimes, the relative cost saving would be increased.

Fuel-Cycle Cost Components

In general, core data from the preliminary safeguards report21 were

used to establish an equilibrium fuel cycle. Information on reactivity

lifetime, U235 burnup, and plutonium production was obtained, however,

Babcock Wilcox Co., Nuclear Merchant Ship Reactor Project, Preliminary Safeguards Report, USAEC Report BAW-1117, Vol. I (Rev. l),December 1958.


44/82

3 7

from the burnup calculation outlined in Section 3. Components considered

in a typical fuel cycle are shown in Fig. 20. The ground rules and as

sumptions used to evaluate the economics are summarized below. All cal

culations are based on one core loading plus four spare fuel elements.For calculational purposes, it was considered that the spare fuel was kept

on standby during reactor operation and then reprocessed with the spent

c o r e

UNCLASSIFIED

O R N L- LR - D WG 6 0 5 5 3 R

IIF, -FABRICATION FRESH CORE

C O N V E RT

TOU02U0 2

PELLETIZIN GCORE

FABRICATION

N E W F U EL

S H I P M E N T

P L US SPA R ES

t I I 1SALVAGE 1 SALVAGE

WITHDRAWAL

CHARGE

LOSS LOSS

UJ

_

>

O(M

0 . 3

3 4 5

EXPOSURE TIME (days)

0 . 2

0. 1

(MO3)

substitution of zircaloy for stainless steel

Documents