ML20138E282

Safety Evaluation Re Extended Burnup of C-E PWR Fuel. Methodologies Acceptable for Extended Burnup Application

ML20138E282
Person / Time
Site:	Calvert Cliffs
Issue date:	10/10/1985
From:	Office of Nuclear Reactor Regulation
To:
Shared Package
ML20138E277	List: ML20138E274 ML20138E282
References
NUDOCS 8510240652
Download: ML20138E282 (25)

Text

. . . . . - _ - . .- ..-- _ _=- _ __ _

g g UNITED STATES g NUCLEAR REGULATORY COMMISSION

& J WASHINGTON, D. C. 20655

\.../

SAFETYEVALUATIONBYTHEOFFICEOFNUCLEARREACTORREGUf.Ah!0N EXTENDED BURNUP OPERATION OF COMBU5 TION ENGINEERING PWR FUEL BALTIMORE GA5 AND ELEGIRIC COMPANY -

CALVERT CLIFF 5 NUCLEAR POWER PLANT, UNIT NOS.1 AND f DOCKET N05. 50-317 AND 50-318

1.0 INTRODUCTION

Economics and prudent utilization of resources have led utilities to seek more efficient use of current generation light water reactors (LWRs). Improved fuel utilization is one of the avenues being pursued for greater efficiency. One of i the greater improvements in fuel utilization is to increase the fuel discharge exposure which is currently at batch average burnups of approximately 28 mwd /kgM for BWRs and approximately 33 mwd /kgM for PWRs to. batch average burnups of

, .p__ approximately 40 mwd /kgM and 50 mwd /kgM or above, respectively. The higher dis-

. charge exposures result in a more complete consumption of the loaded U-235 and a better utilization of the plutonium produced in-reactor. The longer residence

~~

time in-reactor also . reduces spent fuel storage needs.

I In response to this trend for extended burnup fuel operation, we requested each -

fuel vendor to prepare and submit a topical report for review and approval that covers extended burnup experience, methods and test data to provide a generic b asis for operation at extended burnups.

--. Baltimore Gas and Electric Company (BG8E) submitted such a report2 for Ccmbustion

._e Engineering criteria and methods used for licensing their fuel to extended burnup levelsupto45 mwd /kgM(batchaverage). Approval of this report is only for ,

criteria and methods. Specific design approval for individual fuel designs at extended burnups will be addressed in design specific applications.

3 The original submittal by BG&E2 was updated and revised in 1984 in response to NRC questions. Because this reviged version answered only part of the staff questions, an additional response was submitted in August 1984.

• Because the purposes of each of the vendor's high burnup topical reports are slightly different, it is useful to quote CE's goal in preparing this report, as stated in Reference 3.

"This report describes the fuel performance parameters affected by increased fuel burnup (or core residence time) and the behavior phenomena governing the burnup dependence of these parameters. The models (or submodels) used by Combustion Engineering Inc. (CE) to represent these parameters are reviewed with emphasis placed on showing how burnup is included in the anatyses which incorporate these parameters. A review of the current and anticipated data base that support these models is made where appropriate to demonstrate the adequacy of the models up to batch average discharge burnups of 4!i' mwd /kgM 8510240652 951010 PDR ADOCK 05000317 P PDR

(maximum rod average burnups of 52 mwd /kgM). In this manner, the report pro-vides a basis for the generic licensing approval of CE's fuel performance models for operation of 14x14 and 16x16 fuel assembly designs to these target burnup values". "t We agree with the above goals for this review. J.

5 This safety evaluation follows the intent of the SRP to insure that all licensing requirements of the fuel system are reviewed with respect to extended burnup operation. The objective of Section 4.2 and this review are to provide assurance that as a result of extended burnup operation (a) the fuel system is not damaged as a result of normal operation and anticipated operational occurrences, (b) fuel system damage is never so severe as to prevent control rod insertion when it is

• required. (c) the number of fuel rod failures is not underestimated for postula-ted accidents, and (d) coolability is always maintained. "Not damaged" is defined

.: as meaning that fuel rods do not fail, that fuel system dimensions remain within omrational tolerances, and that functional capabilities are not reduced below tiose assumed in the safety analysis. This objective implements General Design F* Criterion (GDC) 10 of 10 CFR Part 50, Appendix A (" General Design Criteria for Nuclear Power Plants") and the design limits that accomplish this are called '

. Specified Acceptable Fuel Design Limits (SAFDLs), " Fuel rod failure" means that the fuel rod leaks and that the first fission product barrier (the cladding) has, therefore, been breached. Fuel rod failures must be accounted for in the dose analysis to demonstrate compliance with the offsite dose limits of 10 CFR Part 100("ReactorSiteCriteria")forpostulatedaccidents. "Coolability," which is sometimes tenned "coolable geometry," means, in general, that the fuel assembly

. retains.its rod-bundle geometric configuration with adequate coolant channels to permit ~ removal of residual heat after a severe accident. The general requirements

• T to maintain control rod insertability and core coolability appear repeatedly in

~

the General Design Criteria (e.g., GDC 27 and 35). Specific coolability require-m'ents for the loss-of-coolant accidents are given in 10 CFR Part 50.46 (" Accept-ance Criteria for Emergency Core Cooling Systems for Light Water Nuclear Power

> . Reactors").

In addition to Section 4.2 of the Standard Review Plan, we have reviewed those

" which are affected by high burnup and aspects of Section 4.3, " Nuclear Design,Section these are discussed in general terms in 5.0. A brief discussion of radiological consequences of operation with large burnup fuel is given in Section 6.0 and our conclusions and Regulatory Position are given in Section 7.0.

In order to meet the above stated objectives and follow the format of Section 4.2, this review covers the following three categories: (1) Fuel System Damage Mechanisms, which are most applicable to normal operation and anticipated operational occurrences; (2) Fuel Rod Failure Mechanisms, which apply to normal operation, anticipated operational occurrences and postulated accidents; and (3)

Fuel Coolability, which applies to postulated accidents. t As noted earlier, this review is intended to provide generic apprdval of the criteria and methods used by Combustion Engineering for licensing; analyses of 2

their fuel designs to extended burnups. The criteria sections in this review address limiting values for fuel damage that are acceptable under the three major categories of failure mechanisms listed above and in the SRP. These criteria, along with certain definitions for fuel failure, constitute the SAFDLs required by GDC 10. The purpose of this review is to determine if the Combustion Engineering criteria and their bases are applicable to extended burnup operatidn of CE fuel.

The evaluation sections review the methods that Combustion Engineering uses to demonstrate that the design criteria have been met for extended burnup operation and thus are reviewed with respect to their applicability to the proposed range of extended burnup operation. These methods and data may include operating ex-perience, prototypc testing and analytical techniques. The determination that specific Combustion Engineering designs meet the stated criteria is not addressed in this review but will be addressed in specific design reviews.

'l 2.0 FUEL SYSTEM DAMAGE .

The design criierta in this section should not be exceeded during nonnal operation i- including anticipated operational occurrences (A00s). The evaluation portion 6f each damage mechanism demonstrates that the design criteria are not exceeded during nonnal operation and A00s.

(a)DesignStress Bases / Criteria - The design basis for fuel assembly, fuel rod, burnable poison rod, and upper end fitting spring stresses is that the fuel system will be

_ functional and will not be damaged due to excessive stresses.

The stress limits for the fuel assembly components are provided in References 6 and 7. The design limit for fuel rod and burnable poison rod. cladding is that the maximum primary tensile stress is less than two-thirds of the Zircaloy yield strength as affected by temperature.

The design limit of the Inconel X-750 upper end fitting spring is that the cal-culated shear stress will be less than or equal to the minimum yield stress in shear.

Many of these bases and limits are used by the industry at large. Combustion Engineering has employed various conservatisms in the limits such as the use of 8 unirradiated yield strengths for zirconium-based alloys. The staff has concluded that the fuel assembly, fuel rod, burnable poison rod, and upper end fitting spring stress design bases and limits are acceptable for current burnup levels, i.e., approximately 33 mwd /kgM. Extended burnup operation does not reduce the applicability of these limits which therefore are found acceptable for use in extended burnup applications. ,

l

~L 3

Evaluation - The topical report states that the methods used to perform stress analyses will not change for extended burnup fuel. These analyse;siare performed using conventional engineering formulas which can be found in standard engineering mechanics textbooks and performed in accordance with ASME genegal (uidelines for analyzing primary and secondary stresses. The staff concluded that these stress analyses are acceptable for current burnup fuels. Extended burnup operation does not reduce the applicability of these methods. They therefore are also acceptable for application to extended burnup fuels.

(b)DesignStrain Bases / Criteria - With regard to fuel assembly design strain, the design basis for normal operation is that pemanent fuel assembly deflections shall not re-N sult in CEA insertion time beyond that allowable. This basis is satisfied by adherence to the stress criteria mentioned above.

. The topical report provides a design limit for fuel rod and burnable poison rod F. cladding circumferential plastic strain (due to cladding creep and pellet swelling and thermal expansion) of 1 percent as a means of precluding excessive cladding deformation. This value appears to be consistent with past practice and will allow CEA insertion within the required time, and is therefore acceptable. CEA insertion time is also verified by measure &r.ts after each reload as required by Technical Specifications.

The material property that could have a significant impact on the cladding strain criterion at extended burnup levels is cladding ductility. The strain criterion

._. could be impacted if cladding ductility were decreased, as a result of extended burnup operations, to a level that would allow cladding failure without the 1%

cladding strain criterion being exceeded in the Combustion Engineering analyses.

From examination of irradiated Zircaloy cladding ductility data, it has been con-cluded that ductility decreases with increasing fluence at low burnup levels, i.e.,

8 mwd /kgM, but asymptotically approaches either a constant value or a small fluence dependence beyond these low burnups. Consequently, cladding ductility has either little or no change for the increased burnup levels projected, i.e., from 33 mwd /kgM to approximately 45 mwd /kgM batch average burnups. In addition, Com-bustion Engineering has irradiated experimental and lead test assemblies with average burnups between 35 and 52 mwd /kgM with no adverse effects in cladding ductility.

From the above, we can conclude that the strain limits proposed by Combustion Engineering are applicable for extended burnup application.  ;

Evaluation - The analysis methods for fuel rod cladding strain for normal operation .

are those described in the FATES-3 documentation, e.g., the cladding creep and I fuel swelling models. This code and its applicable models have been approved f extended burnup application with modifications to the fission gas , release model.{0 l

4 I_-

Cladding strain for Combustion Engineering burnable poison rods is not expected to be a problem at extended burnup because nearly all of the boron-10 is burned-out (> 90%) by the time extended burnup levels are reached. Conseguently,the

~

major driv (r.g forces for poison rod strains such as poison pellet swelling and helium gas release (due to helium generation from the B-10 (n,9) reaction) are significantly reduced at extended burnups. e (c)StrainFaticue Bases / Criteria- The Combustion Engineering strain fatigue criterion3 ,6 is different from that described in SRP Section 4.2, viz., a safety factor of 2 on stress amplitude g of 20 on the number of cycles using the methods of O'Donnell and Langer . Instead, CE has proposed that the cumulative strain cycling usage (i.e., the sum of the ratios of the number of c

% effective strain range to the permitted number in that range)ycles in a given will not exceed 0.8. For Zircaloy cladding, the design limit curve has been adjusted to provide a strain margin-for the effects of uncertainty and irradiation. The resulting

-W curyg(of the criterion that is discussed in the SRP.given ment in References Therefore, the staff3con-and 6) bounds g

cluded that the Combustion Engineering criterion was acceptable for current

- burnup levels.

As noted for cladding strain, the material property that could have a significant effect on the strain fatigue criterion is cladding ductility. As discussed in the above section on design strain, extended burnuo operation has shown little or no observable effects on cladding ductility. From this it is concluded that

- extended burnup operation does not reduce the applicability of the strain fatigue limits and thus they are acceptable for use in extended burnup applications.

~

Evaluation - The fuel and cladding models used to determine fuel and cladding diametral changes g for the fatigue analysis are described in geference 9 which hashenapproved . As noted earlier, this updated version of the FATES code has been approved for extended burnup application and is thus acceptable for determining strain fatigue at extended burnups.

The power history used in the fatigue analysis includes daily power cycling between 10 and 100 percent power along with a total of 50 reactor heatups and cooldowns during the fuel lifetime. In addition, actual past power histories andassumed(basedonexpectedoperation)dailyloadcyclingarealsoconsidered.

This power history takes into account the extra duty cycle extended burnup operation will experience and thus is acceptable for detennining the strain fatigue margin of extended burnup fuel.

Based on the above and the fact that Zircaloy material properties are not adversely affected by extended burnup operation, we can conclude that the Combustion Engineering strain fatigue methodology is acceptable for extended burnup applications.  ;

5

(d) Fretting Wear i

Bases / Criteria - Fretting wear is a concern for fuel and burnable jioison rods and the Zircaloy guide tubes. Fretting, or wear, may occur on thefuel and/or burnable rod poison cladding surfaces in contact with the spacer grids if there is a reduction in grid spacing loads in combination with small amplitude, flow-indused, vibratory forces. Guide tube wear mey result when there is flow induced motion between the control rod ends and the inner wall of the guide tube.

While the Standard Review Plan (SRP) Section 4.2,5 does not provide numerical bounding-value acceptance criteria for fretting wear, it does stipulate that the allowable fretting wear should be stated in the safety analysis report and that the stress / strain and fatigue limits should presume the existence of this wear.

N Review of Combustion Engineering documentation 3 and responses to staff questions 4 indicates that a specific fretting wear limit is not used for fuel assembly components because it has not been a problem for current fuel designs. This same p__

argument is also used to explain why fretting wear is not accounted for in the fuel and burnable poison rod analyses for cladding stress and fatigue. In order i

-- to support this claim, Combustion Engineering provided fuel examination information from a large number of fuel assemblies with average burnups approximately up to the target burnup that show that no failures or significant wear gas observed on the surface of their fuel or burnable poison rods due to fretting Guide tube wear, however, was observed in several Combustion Engineering fuel assemblies in 1977. Since then a design change in the guide tubes has e liminated any noticeable guide tube wear for both 14x14 and 16x16 fuel assembly designs. Further confirmation of this lack of guide tube wear is anticipa3ed to be available from a demonstration program for the 16x16 fuel assembly

~ In view of the lack of observed fretting wear in the examination of a large number of Combustion Engineering fuel assemblies with extended burunups and existing fuel surveillance programs, we concur that there is no need for a fretting wear limit.

Evaluation - The lack of a significant amount of observed fretting wear in both in-reactor and out-of-reactor tests on fuel assemblies has been used to support the argument that analytical analyses for fretting wear are not needed. However, this does not answer the question of the impact even a small amount of fretting wear may have on safety analyses, e.g., LOCA and stress /strgin, because cladding thickness is reduced. Combustion Engineering has responded that the most limiting LOCA analysis is at early-in-11fe when stored energy is highest and fretting wear insignificant. We agree with this assessment. They have also perfonned cladding strain and fatigue analyses with a 2-m11 reduction in cladding thickness to show l that reductions in cladding thickness due to oxidation or wear are insignificant.

(These results will be discussed in more detail in the Oxidation and Crud Buildup l 6

section because cladding oxidation is believed to be the more dominant mechanism for cladding thinning at extended burnups.)

From these responses, it is concluded that Combustion Engineerind has provided substantial justification that fretting wear is not a significant f,uel damage mechanism for current designs at extended burnups. It is also concluded that fretting wear has an insignificant effect on LOCA and stress / strain ~ analyses.

Therefore, we concur that the inclusion of cladding fretting wear in LOCA, stress /

strain and fatigue analyses is not required for extended burnup evaluations.

As noted in the Criteria section, guide tube wear has been a problem in the past for Combustion Engineering assemblies. Design changes have been implemented by Combustion Engineering for both 14x14 and 16x16 assemblies to reduce guide tube wear. Both out-of-reactor and in-reactor confirmation tests have been perfomed to show that these design changes have resulted in a N significant decrease in guide tube wear. However, none of these tests were close to the in-reactor residence times expected for extended burnup assemblies. Ev'en though guide tube wear is not expected to be a problem at extended burnups for the newly designed guide tubes (based on the lack of wear-4 . at lower burnups) Combustion Engineering should obtain guide tube wear data at extended burnup levels to confirm the lack of wear at these burnups. In the meantime, this issue should be addressed in specific licensing applications for extended burnup operation.

(e)0xidationandCrudBuildup Bases / Criteria - Section 4.2 of the Standard Review Plan identifies cladding oxidation and crud buildup as potential fuel system damage mechanisms. General

, - - . mechanical properties of the cladding are not significantly impacted by thin

- oxides or crud buildup. The major means of controlling fuel damage due to claddirg oxidation and crud is through water chemistry controls, materials used in the primary system, and fuel surveillance programs that are all reactor specific. Because these controls are already included in the specific reactor design, a design limit on cladding oxidation and crud is not necessary.

This does not, however, eliminate the need to include the effects of cladding oxidation and crud in safety analyses such as for LOCA and mechanical analyses.

This is discussed further in the evaluation presented below.

4 Evaluation - Combustion Engineering has indicated that they do not explicitly include the effects of cladding oxidation in their thermal and mechanical analyses but that they arg implicitly considered in their steady-state fuel performance code, FATES-3 . There is a normal amount of oxidation expected during the irradiation of fuel rods that is dependent on time in reactor.

Therefore, extended burnups will result in thicker oxide layers that provide an extra thermal barrier affecting thennal analyses and cladding thinning 4that can affect the mechanical analyses. Combustion Engineering has intlicated that the most limiting LOCA analysis is early in life when thermal stored energy is highest and oxidation is insignificant. We concur with this assessment. For other thermally dependent analyses, e.g., fuel molting and interna'l rod pressures at extended burnup, there are other conservatisms such as peak linear heat generation rate and code conservatisms that are more than an order 7

of magnitude larger than the thermal effect of the oxide layer (approximately 25'C) at extended burnups. Consequently, these conservatisms more than compensate for the fact that cladding oxidation and crud are not)x licitly accounted for in these themal analyses. ,

CombustionEngineeringhasalsoperformedmechanicalcladdingstrainandfatigue analyses with and without a 2-mil reduction in cladding thickness (4-mil reduction ondiameter)toshowthattheeffectofcladdingoxidationisinsignificant.

Becausethemolevolumeoftheoxideis1.56thatofthemetal,thethickgss i of the metal consumed is approximately two-thirds of the oxide thickness.

Consequently, the 2-mil reduction in Zircaloy cladding thickness is equivalent to the formation of a 3-mil oxide layer which represents a conservative upper  !

. bound limit for PWR fuel rods with rod average burnups of approximately 60 mwd /kgM. The difference in the calculated cladding strain with and without

'- the 2-mil reduction was 0.785% and 0.780%, respectively. The difference in the

- calculated fatigue factor with and without the 2-mil cladding reduction was 0.535 and 0.52, respectively. The calculational uncertainty in these models is more than an order of magnitude greater than these differences. -

j

,4_

Consequently, the effect of normal cladding oxidation at extended burnup levels , i is found to be insignificant on both cladding strain and fatigue.

From the above, it is concluded that the effects of cladding oxidation on thermal and mechanical analyses are insignificant for rod average burnups to '

60 mwd /kgM and thus need not be incorporated in these analyses.

- (f) Rod Bowing T Bases / Criteria - Fuel and burnable poison rod bowing are phenomena that alter I

' the design-pf tch dimensions between adjacent rods. Bowing affects local nuclear i power peaking and the local heat transfer to the coolant. Rather than placing design limits on the amount of bowing that is permitted, the effects of bowing l areincludedinthesafetyanalys{s. This is consistent with the Standard Review Plan and has been approved for current burnup levels, and is judged to be acceptable for extended burnups. The methods used for predicting the degree of rod bowing at extended burnups are evaluated'below, i Evaluation - The methods used to account for the effect of fuel and poison rod bowingfny4x14and16x16fuelassembliesarepresentedinReferencT53 and CENPD-225 with its supplements. These methods have been approved for fcel and Type 3 poison rods to current burnup levels.

Reference 3 shows that the licensing model for 14x14 and 16x16 assemblies predicts conservative channel closure for assembly burnups to the burnup region in which the extended burnup assemblies are in general not limiting in terms of DNBR.

This is because the power peaking in extended burnup assemblies is dropping [

faster than the power penalties applied to account for rod bow, ljowever,this does not eliminate the need to confim whether specific licensigg, applications ,

for extended burnup operation are ONBR limged or not. It is coneiuded that the licensing methodology approved previously for current burnup is also applicable for extended burnup assemblies.  :

r

(g) Axial Growth Bases / Criteria - The core components requiring axial dimensional evaluation are the CEAs, burnable poison rods, fuel rods, and fuel assemblies. 7he CEAs are not included in this extended burnup review. The growth of the burnable poison rods and fuel rods is miinly governed by (a) the behavior of poison'and fuel pellets and their interaction with the Zircaloy-4 cladding and (b)'the irradi-ation-and-stress-induced growth of the Zircaloy-4 cladding (due to rod pres-sures being less than coolant pressure). The growth of the fuel assemblies is a function of both the compressive creep and the irradiation-induced growth of the Zircaloy-4 guide tubes. For the Zircoloy cladding and fuel assembly guide tubes, the critical tolerances that require controlling are (a) the spacing between the fuel rods and the upper end fitting (i.e., shoulder gap) and (b) the spacing between the fuel assemblies and the core internals. Failure to adequately design for the former may result in fuel rod bowing and, for the

'r-latter, may result in collapse of the holddown springs. With re adequately designed shoulder gaps,' problems have been reported References 16, (gard to in-17, 18, and 19) in foreign (Obrigheim and Beznau) and domestic (Ginna and e Arkansas Nuclear One, Unit 2 (ANO-2)) plants that have necessitated predischarge 7 modifications to fuel assemblies.

.-- For fuel and burnable poison rods, allowances are made to ensure adequate (non-zero) shoulder gap clearance (at a 95% confidence level) to the upper fuel a'ssembly end fitting such that the clearance is maintained throughout the design lifetime of the fuel. For fuel assembly axial growth, Combustion Engineering has a design basis that sufficient clearance between the fuel assembly and the upper guide structure should exist throughout the expected lifetime of the fuel assembly.

- This basis allocates a fuel assembly gap spacing which will accomodate the

. maximum axial growth when establishing the design minimum initial fuel assembly

- clearance with respect to the core internals. These design bases and limits dealinggithaxialgrowthpreventmechanicalinterference. They have been approved for current burnup levels. These design bases and limits will ensure that contact is prevented and thus are judged to be also applicable to extended burnup operation.

Evaluation.TFeCombustionEngineeringmethodsggodelsusedforpredicting fuel rod and assembly growth have been submitted for limited exposures on a plant specific basis. This evaluation will discuss 1) the models used to pre-dict assembly growth and thus gap spacing between the fuel assembly and core internals,and2)themodelsusedtopredicttheshouldergapspacingsbetween the fuel rod and the upper end fitting.

J "L

9

._ = _ - - -- .. . _ _. -- _ _ - . _ - _ .

Review of the information and data from Reference 3 indicates that fuel assembly length changes are conservatively predicted, using the upper 95%  !

probability curves, for 14x14 fuel assemblies with recrystallizat16n annealed (RXA) guide tubes.

The data1x10 fluences up to approximately from2 ghe extended burnup) report has assembly nyt (E :>0.821 MeV which corresponds to ,

assembly average burnups up to approximately 50 mwd /kgM. This appears to cover I the burnup range of extended burnup fuel. For 16x16 fuel designs with both RXA and stress predicted forrelief assembly annealed (SRA) fluences up toguide 6x10 tubg,nyt (E > 0.821 MeV) which islength chan nearly half the fluence level expected for extended burnup assemblies. This points to the need for higher fluence data for the 16x16 design before the 16x16 assembly growth model can be approved for extended burnup application.

,. Review of the shoulder gap spacing data for 14x14 fuel designs with RXA guide tubes (upperindicates equation-in that the 4-7 Table upper 955 bounding) of Referenegg3 curve, Sounds usingallthe nearly of the original RXA model fuel rod

, data with fivences to approximately 9x10 n n (E> 0.821 MeV). Before this .

P- model can be approved for extended burnup appication, additional data at higher fluences, i.e., epproximately 1.3x10 nyt, are needed for this design.

. For the 16x16 fuel design, Reference 3 has presented an interim model for cal-culating the shoulder gap spacings for those assemblies with RXA guide tubes at extended burnups. This interim model by its nature 2 s conservative up to the nyt(E>0.821MeV)since fluencelimitofthedata,i.e.,approximately7xylor16x16fuelde it bounds all data. A separate model is proposed f SRA guide tubes that is al g conservative up to the fluence limit of the data,

- i.e., approximately 6.5x10 nyt (E > 0.821 MeV). Before either model (for

! the RXA or SRA guide tube) for the 16x16 fuel design can be approved for extended i burnup application, additional data at higher fluences are needed.  !

Consequently, the only axial growth model ap3 roved for extended burnup application is the one used for predicting tie 14x14 fuel assembly (with RXA l

> quidetubes)lengthchangesandthusthespacingbetweentheassemblyandcore

<nternals. The remaining models for determinin spacings and shoulder gap spacings for the and14x1g assembly-to-core 16x16 (with RXA and SR internal g guide tubes) fuel assembly designs need extended burnup data to confinn their applicability to extended burnup levels. In the meantime, this issue will be  ;

addressed in specific licensing applications for extended burnup operation. l f

I t

i

. l 10 l i

l

i l

i (h)RodPressure ,

~

Bases / Criteria - Rod internal pressure is a potential driving forcI for, rather than a direct mechanism of, fuel system damage that could contribut~e to the loss of dimensional stability and cladding integrity. The Standard Review Plan presents an acceptance criterion that is sufficient to preclude fuel damage in this regard and has been widely used by the industry; it states that rod in-ternal gas pretsure should remain below the nominal system pressure during normal operation unless otherwise justified.

4 The Combustion Engineering response to question 5 of this review has indicated that the current criterion for internal rod pressures is that rod pressure remain below system pressure. This is consistent with the SRP and thus is

'y found acceptable for extended burnup operation.

Evaluation - The models and methods used to predict internal rod pressures are evaluatedinghissection. The models1 gsed are described in the FATES-3 code -

7' documentation which has been approved with a modification on its application at high burnups. The modification is imposed on the fission gas release model by limiting the grain size as a function of burnup. No further restrictions have been placed on the application of this code to extended burnup fuels.

Consequently, the FATES-3 code is approved for predicting internal rod pressures for extended burnup fuel.

In response to a staff question on the methods used to calculate rod pressures.

i.e., code input, Combustion Engineering has indicated that the power histories

~-..

input to the code are based on a power history which includes both steady-state and transient operation.

This methodology is found to bound all possible rod power at any given burnup level which in turn bounds any real rod in any given reactor core.

Consequently, this methodology is found to be acceptable for determing the power histories to be used in predicting rod internal pressures.

(1)Assemblyt.iftoff Bases / Criteria - The Standard Review Plan calls for the fuel assembly holddown capability (wet weight and spring forces) to exceed worst-case hydraulic loads for nomal operation, which includes anticipated opergtfonal occurrences. The Combustion Engineering Extended Burnup Topical Report applies this design basis. This is found acceptable for application to extended burnup assemblies, i

4

• L 11

Evaluation-CombustionEggineeringliftoffanalyseshavebeensummarizedin Reference 24 and approved . The fuel assembly liftoff forces are a: function of plant coolant flow, spring forces, and assembly dimensional chang,es. Extended burnup irradiation will result in additional holddown spring relax 4 tion and assembly length increases which will have opposing effecg on assembly holddown forces. Combustion Engineering has provided a reference for the methods used in determining holddown spring relaxation with irradiation. Combustion Engineer-ing has also stated " spring relaxation increases by a small amount for the additional fluence associated with extended burnup". This is consistent with industry experience. Assembly length increases due to extended burnup ir-radiation are found to have the dominant effect on holddown forces and, because this phenomenon will increase these forces, assembly liftoff is not judged to be a problem at extended burnups.

(j)ControlMaterialLeaching ,

Bases / Criteria - The Standard Review Plan and General Design Criteria require

, that reactivity control be maintained. Rod reactivity can sometimes be lost -

P- by leaching of certain poison materials if the cladding of control-bearing material has been breached.

Evaluation - Reactivity loss from burnable poison rods at extended burnup levels is found to be insignificant because nearly all of the reactivity controlling boron-10 is burned out at these burnup levels. Consequently, reactivity loss due to leaching of burnable poison rods at extended burnups is not significant. No further evaluations are needed in this area.

3.0 FUEL ROD FAILURE

~

In the following paragraphs, fuel rod failure thresholds and analysis methods for the failure mechanisms Ifsted in the Standard Review Plan are reviewed.

When the failure thresholds are applied to normal operation including antici-pated operational occurrences, they are used as limits (and hence SAFDLs) since fuel failure under those conditions should not occur according to the traditional conservative interpretation of General Design Criterion 10. When these thresholds are used for postulated accidents, fuel failures are permitted, but they must be accounted for in the safety analysis dose calculations. The basis or reason for establishing these failure thresholds is thus established by GDC 10 and 10 CFR Part 100 and only the threshold values and the analysis methods used to assure that they are met are reviewed below.

(a)Hydriding Bases / Criteria - Internal hydriding as a cladding failure mechanism is precluded bycontrollingthelevelofhydrogenimpuritiesduringfabrication,.)Themoisture level in the uranium dioxide fuel is limited (to a proprietary value by Com-bustionEngineeringtolessghan20 ppm,andthisspecificationis; compatible with the ASTM specification 2 which allows two micrograms of hy.dro, gen per gram 12

of uranium (i.e., 2 ppm). This is the same as ghe limit in the Standard Review Plan. It has previously been found acceptable and continues to be acceptable for extended burnup application.

A specific design basis and limit for external hydriding has been f,ound un-necessary. As justification, Reference 3 cites data that indicate-(a) hydrogen absorption of up to 800 ppm in cold worked Zircaloy with circumferentially orientated hydrides has no effect on cladding ductility at 572 F, and (b) hydriding remains below a level which could cause deleterious effects up to a proprietary value of burnup in the extended burnup region. The references that contain this infonnation have not been reviewed but they are consistent with the evidence of other fuel vendors. Also, we are not aware of any LWR fuel failures due to external hydriding at either current or extended burnup levels. The staff has previously found that the lack of a design limit on external hydriding was acceptable for current burnups and this is also

} acceptable for extended burnups. ,

Evaluation - The above discussion indicates that there is reasonable evidence 4_ that hydriding will not be a likely failure mechanism at or near batch average-

. burnups of 45 mwd /kgM.

~

(b)CladdingCollapse, Bases / Criteria - If axial gaps in the fuel pellet column were to occur due to densification, the cladding would have the potential of collapsing into a gap (i.e., flattening). Because of the large local strains that would result from collapse, the cladding is assumed to fail. It is a Combustion Engineering

' design basis that cladding collapse is precluded during the fuel rod and burnable poison rod design lifetime. This design basis is the same as that in

- the Standard Review Plan and has been approved . It is also found to be applicable to extended burnup applications.

Evaluation - The longer in-reactor residence times associated with extended-burnup fuel will increase the amount of creep of unsupported fuel cladding.

Extensive postirradiation examinations reported in the Topical Report do not show any evidence of cladding collapse or large local ovalities in CE fuel designs. This is primarily the result of use of prepressurized fuel rods and

. stable fuel in current generation designs.

Reference 3 discusses a new methodology for calculating cladding collapse 27 ,

This new methodology is based on a prohgbility for collapse of an unsupported tube and uses the earlier creep models' . This method is still conservative because it assumes a gap has formed and the tube is unsupported. This method will account for the longer in-reactor residence times associated with extended burnup fuel and is thus found acceptable for extended burnup application.

J

~L 13

(c) Overheating of Cladding Bases / Criteria - The design limit for the prevention of fuel failures due to overheating is that there will be at least 95% probabilit confidence level that departure from nucleate boiling (DNB) y atoccur will not M!95%on a fuel rod having the minimum DNBR during normal operation and anticipated operational occurrences. Thisdesignlimitisconsistentwiththeethegmal margin criterion of SRP Section 4.2 and thus has been found acceptable for use at current burnup levels. It is also judged to remain acceptable for extended burnup applications.

Evaluation - As stated in SRP Section 4.2, adequate cooling is assumed to exist when the thermal margin criterion to limit departure from nucleate boiling (DNB) in the core is satisfied. The method employed to meet the DNB

• design basis is adequate to encompass extended burnup applications.

J (d) 0verheating of Fuel Pellets

, Bases / Criteria - As a second method of avoiding cladding failure due to over -

1- - heating, combustion Engineering avoids centerline fuel pellet melting as a design limit. This design limit is the same as given in the Standard Review s

, - Plan and has been approved for use at current levels. This is also found to be acceptable for extended burnup applications.

Evaluation - The design evaluation of the fuel centerline melg limit is performed with the approved CE fuel performance code, FATES-3 . This code is also used to calculate initial conditions for transients and accidents.

- As not.ed earlier, this code is acceptable for fuel performance calculations on extended burnup fuels with the modification to the fission gas release model.

~

I,n applying the FATES-3 code to the centerline melting analysis, the melting .

temperature of the U0 is assumed to be 5080*F unirradiated and is decreased by 58'F per 10 GWd/t.2 This dependence on burnup is based on the mag con-

> servative of two sets of UO melting data generated by Christiansen . The more conservative data base2 consists of newer melting data with burnups up to 52 mwd /kgM. Christiansen has also presented a second data set that is based on older U0,, melting data with burnups to > 95 mwd /kgM. This older data set shows that the U02 melting temperature at burnups of 95 mwd /kgM is nearly equivalent to the melting temperature of the newer data at burnups of 50 mwd /kgM.

There is alo evidence that the mixed oxide melting point decrease with burnup may be less than the 58'F per 10 mwd /kgM proposed by Christiansen.

Thus the decrease extended in U0, ion. melting temperature with burnup is acceptable for burnup applicat (e) pellet / Cladding Interaction .

Bases / Criteria - As indicated in SRP Section 4.2, there are no generally applicable criteria for pellet / cladding interaction (PCI) failure.' However, two acceptance criteria of limited application are presented intie SRP for PCI: (1) less than 1% transient-induced cladding strain, and (2) no centerline 14

fuelmegting. Both of these ligits are used in Combustion Engineering fuel '

designs and have been approved for current burnup applications. _CE proposed that they also be applied for extended burnup application. We fimi this acceptable. T 9

Evaluation - Combustion Engineering uses the FATES-3 code to show their fuel meets both the cladding strain and fuel melt criteria. As noted earlier, this code is acceptable with a modification for application to extended burnup fuel.

Reference 3 also indicates that particular design characteristics of CE fuel help to mitigate the effects of PCI in addition to plant operating guidelines.

These design characteristics are prepressurized fuel rods, pellet dishing and an appropriate pellet length. In addition to these factors, it is noted that at extended burnup the fuel will experience a reduction in power capability

,- that reduces the possiblity of PCI occurrence, Therefore, the above factors mitigate the possibility and effects of PCI and

, thus are acceptable for application to extended burnup fuels.

. y (f)CladdingRupture Bases / Criteria - In the LOCA analysis of Combustion Engineering fuel designs, an empirical model is used to predict the occurrence of cladding rupture. The failure temperature is expressed as a function of differential pressure across the cladding wall. There are no specific design limits associated with cladding rupture other than the 10 CFR 50 Appendix K requirement that the incidence of

rupture not be underestimated. The rupture model is an integral portion of the approved ECCS evaluation model. This is found acceptable for extended burnups.

Evaluation - The topical report3,4 has concluded that those parameters input to the LOCA licensing analysis and dependent on extended burnup levels, such as stored energy, fission gas release, fuel swelling, and irradiation growth are explicitly modeled and no further consideration is necessary. These parameters are input to the LOCA analysis from the FATES-3 code . This code hasbeenappgvedwithamodificationtothefissiorgasreleasemodelat high burnups with no further restrictions with respect to its application to extended burnups. It should also be noted that this code has also been verified against data with rod average burnups to approximately 55 mwd /kgM. Consequently, the parameters important for LOCA analyses at extended burnups are founri to be adequately modeled and verified to extended burnup levels.

15 C __ _ - _ _ _ _ . _ _ . _ _ _ . _ . _ _ . _

I l

l Zircaloy property changes up to the extended burnup levels analyzed in Reference 3, i.e., batch average burnups of 45 mwd /kgM, have been found to be insignificant (see Section 2.0 (b)). In addition, irradiation effects in the fuel rod cladding i would be annealed out at the temperatures obtained during a LOCA.' 4 i

~

l In Reference 3 it is also concluded that "no new burnup consideratfons need to be modeled for extended burnup analyses of cladding defomation and rupture during l a LOCA", and that CE analyses have shown that extended burnup is not a limiting condition for LOCA. The above evaluation concurs with these conclusions. There-fore, it is concluded that the methodology used for detennining cladding deformation and rupture during a LOCA is applicable to extended burnup fuel.

(g) Mechanical Fracturing Bases / Criteria - Mechanical fracturing of a fuel rod could potentially arise i from an externally applied force such as a hydraulic load or a load derived 3 from core-plate motion. To preclu~de such failure, the topical report states that fuel rod fracture stress limits shall ie in accordance with the criteria

._ given in Table 9-1 of CENPD-178, Revision 1 -

The review of CENPD-178, Revision 1 and the criteria given in Tabg 9-1 has been completed and CENPD-178 Revision 1 has been found acceptable for current burnup levels. The effect of extended burnup on Zircaloy ductility has been discussed earlier in Section 2.0(b) and found to be small or non-existent at the batch average extended burnup levels proposed, i.e., 45 mwd /kgM. Con-sequently, these criteria are also found to be acceptable for extended burnup application, e

Evaluation - The analysis methods for detennining hydraulic and core-plate motion

' .:.. loadsarealsopregntedinCENPD-178, Revision 1andwerefoundacceptablein the earlier review . These methods are also found acceptable for application to extended burnup assemblies. -

4.0 FUEL C00 LABILITY For accidents in which severe fuel damage might occur, core coolability must be maintained as required by several General Design Criteria (e.g., GDC 27 and 35).

In the following paragraphs, limits and methods to assure that coolability is maintained for the severe damage mechanisms listed in the Standard Review Plan are reviewed with respect to high burnup.

(a) Fragmentation of Embrittled Cladding Bases / Criteria - The most severe occurrence of cladding oxidation and possible fragmentation during a design basis accident is a result of a significant degree of cladding oxidation during a LOCA. In order to reduce the effects of cladding oxidation for a LOCA, Combustion Engineering uses the acceptance criteria of 2200*F on peak cladding temperature and 17% on maximum cladding oxidatioreas prescribed by 10 CFR 50.46. .;

16

Some non-LOCA accidents (i.e., steam-line break, locked rotor, and CEA ejection) may lead to departure from nucleate boiling (DNB) and produce high A ladding tem-peratures, which might involve substantial cladding oxidation and embrittlement  !

and thus challenge core golability. Combustion Engineering has provided analyses  !

and experimental results to show that the time and temperature fo'r these accidents  !

are significantly less severe than those that would challenge coreroolability.

Evaluation - The cladding oxidation model used to detennine that the above ,

crf terta are met is not affected by extended burnup operation; however, the steady-state operational input provided to the LOCA analysis is burnup de-pendent. Those burnup dependent parameters from steady-state operation importanttotheLOCAanalysissuchasstoredenergyandfissigngasrelease '

are provided by the FATES-3 steady-state fuel pergnnance code . This code has been approved for extended burnup application as noted earlier. Con-N sequently, the effects of extended burnup on these analysis methods have been adequately addressed by CE. -

4 e It should be noted that Combustion Engineering has indicated that its -

7 analyses with the above LOCA methodology have shown that extended burnup fuel is not LOCA limited.

(b) Violent Expulsion of Fuel Material Bases / Criteria - In a CEA ejection accident, large and rapid deposition of energy in the fuel could result in melting, fragmentation, and dispersal of fuel. The mechanical action associated with fuel dispersal might be sufficient

- to destroy fuel cladding and the rod-bundle geometry and to provide significant pressure pulses in the primary system. To limit the effects of CEA ejection, '

1 Regulatory Guide 1.77 guidelines state that the radially-averaged energy deposition at the hottest axial location be restricted to less than 280 cal /g.

This limit has not been explicitly stated by Combus on Engineering in this .

review; however, it has been stated in past reviews i

> . r Evaluation - The methods used by CE to calculate energy deposition as a result i of reactivity insertion accidents are conservative. Radial average enthalpies  !

for reactivity insertion accidents calculated by CE for operating reactors do  ;

not exceed 220 to 230 cal /gm. Because of this, and because extended burnup fuel would be expected to have a lower enthalpy rise than new fuel, it is concluded  ;

that extended burnup fuel would meet the 280 cal /gm criterion with considerable i margin.  !

(c)CladdingBallooningandFlowBlockage Bases / Criteria - In the CE LOCA analysis, empirical models are used to predict the degree of cladding circumferential strain and assembly flow blockage at the time of hot-rod and hot-assembly burst. These models are functions of differential pressure across the cladding wall. There are no spec,1fic design limitsassociatedwithballooningandblockage,andtheballooniygandblockage models are integral portions of the ECCS evaluation model.

17 i

Evaluation- The cladding ballooning and flow blockage models used in the combustion Engineering for cladding rupture LOCA-ECCS temperature analysis and burst strainare directly coupled (discussed to 3.0 in Section the modelg).

The cladding gfonnation, rupture and flow blockage models are t(ote proposed in NUREG-0630 . These models have been approved by the NRC for. current burn-up level.

J 3

CE stated that cladding oxidation at extended burnup levels may result in re-duced cladding strains from those predicted by the NUREG-0630 models. However, lower cladding strains will result in less blockage and thus the possible effects ,

of extended burnup are found to make the current cladding ballooning and LOCA  !

analysis more conservative. In addition, the high cladding temperatures associated with the LOCA analysis will anneal out any irradiation damage effects on clad-ding properties. Consequently these models are found to be acceptable for ex-tended burnup application.

N -

The steady-state operational input'that is provided to the LOCA analysis from the~ FATES-3 fuel performance code is burnup dependent. As noted earlier (see c Section 3.0(f)), the FATES-3 code has been approved for extended burnup appli--

~7 cation for the LOCA analysis.

(d) Structural Damage from External Forces Bases / Criteria - To withstand the mechanical loads of a LOCA or an earthquake, the fuel assembly is designed to satisfy the stress criteria listed in Section 2.0(a), and guide-tube defonnation is limited such as not to prevent CEA in-sertion during tg safe shutdown earthquake (SSE). These criteria have been found acceptable for current burnup fuel and are also found acceptable for

. .. extended burnup fuel because yield strengths will only increase with increased

7. irradiation and the decrease in cladding ductility is negligible as noted in Section2.0(a).

Evaluation - As noted above, the material properties of the fuel assembly used

=-

in the 55E-and-LOCA loading analysis either improve or do not change signifi-cantly at the extended burnups proposed by Combustion Engineering; i.e., 45 mwd /kgM (batchaverage). In addition, the methods used to detennine that the above criteria have been met are found to be independent of irradiation exposure and fuel burnup. These methods have been approved for current burnup fuel and, based on the above, are found acceptable for extended burnup applications.

5.0 NUCLEAR DESIGN Typical extended fuel burnup and increased fuel cycle length core designs utilize higher fuel enrichments, low leakage patterns, burnable poison rods and/or axial blankets. Higher fuel enrichment is required to reduce the number of feed assemblies and offset the reactivity loss resulting from the higher fission product inventory. The core neutron economy is impvoved by re-

. ducing the radial leakage using low leakage loading patterns in which the high l burnup fuel is located on the core periphery. Axial blankets ar,p psed to flatten the axial burnup distribution and improve fuel utilizatiot. The in-creased power peaking resulting from the larger reactivity differences between 18

the fresh and high burnup fuel and the use of low leakage loading patterns is generally controlled using burnable poison rods. ,p

~

Thesefeaturesaffectthephysicscharacteristicsofhighburnupckrede-signs. The increased fuel depletion in high burnup cores results in an in-crease in the plutonium fission fraction and the fisson product inientory, the higher plutonium fission fraction in turn hardens the neutron spectrum and increases the neutron production per unit energy. The increased fission prod-uct inventory and use of burnable absorbers tends to increase absorption and also harden the neutron spectrum.

While the increased fuel burnup does affect the core physics characteristics, the changes are relatively small and the physics parameters are determined using standard calculational methods and procedures. The high burnup neutronic effects

, enter through the microscopic cross sections and fuel assembly lattice group constants. The present calculations of these parameters account for substantial levels of plutonium, fission products and burnable absorbers, and these methods

, are expected to adequately treat the neutronics changes associated with extended fuel burnup. The depletion methods used to track the plutonium and fission product isotopics and various normalization procedures are also expected to be equally

, __ valid for high burnup fuel configurations.

The high burnup fuel physics characteristics and core configuration affect the '

core nuclear safety > parameters. The major effect is to increase the power in the low burnup and/or centrally located fuel assemblies and to decrease the power in the high burnup and/or peripherally located fuel assemblies. The resulting

- increase in the number and power of the peak powered rods is typically controlled by use of burnable poison rods.

~

The increased fission product inventory and use of burnable absorbers increases thermal and epithennal absorption and hardens the core neutron spectrum. These factors combine to reduce the boron and control rod worth, prompt neutron life-time and Doppler coefficient. The moderator temperature coefficient may in-crease or decrease depending on the particular high burnup design, and is also controlled using burnable absorbers as in present core designs. The delayed neutron fraction is also reduced as a result of the increased plutonium fission fraction.

ce In addition to improving the neutron economy, the low leakage patterns reduce the pressure vessel damage fluence by shifting the power toward the center of the core and away from the vessel. This fluence reduction is partially offset, however, by the harder neutron spectrum and increased neutron production (per MeV) of the high burnup fuel.

The calculation of the high burnup core safety parameters is carried out using the same core and lattice methods and procedures used for present cure designs.

The changes in the core safety parameters resulting from the higher fuel burnup designs tend to be relatively small as a result of the low relatise;importance of the high burnup fuel and the tendency for the increase in pluton'fum fission 19

rates and fission product inventory to saturate. These calculated safety param-eters provide the core neutronics input to the required plant transient and accident analysis. (L As the above discussion indicates, the effect of high burnup on the~ physics design is expected to result in relatively small changes in the pridicted cnaracteristics of the core, and also relatively small extensions in range of the methods used to calculate the characteristics. Because high burnup fuel is not subject to limiting duty and because of its low relative importance in detemining the core characteristics, we conclude that present methods are ad- -

equate for high burnup designs. To provide added assurance that these methods are adequate, we recomend that the fuel vendors and licensees pay special

. attention to comparisons of predicted and measured physics parameters (partic-ularly power distributions) which are monitored during the reactor cycle. A i systematic pattern of deviation between predictions and measurements would pro-vide an indication of potential pr~oblems. We intend to take an active role in following these' comparisons. ,

7 6.0 RADIOLOGICAL CONSIDERATIONS OF POSTULATED ACCIDENTS WITH EXTENDED BURNUP OPERATION s Discussion of Present Analysis Procedures To ensure that accidents involving the movement of fuel do not constitute an offsite health and safety issue, design events are assessed. Analyses of fuel handling accidents assume release of the entire volatile radionuclide fuel assembly gap ~and plenum inventory under nominally 23 feet of water after the

g. assembly has cooled substantially (usually at least 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> for BWR assemblies, I e:- 72 or 100 hours0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br /> for PWR assemblies). For assemblies with burnup up to 38,000

' mwd /t batch average at discharge, Regulatory Guide 1.25 assumptions are used.

Thesestipulateaninventoryof10percentofthetotaj5 fuel assembly iodines and noble gases (with the exception of 30 percent for Kr) in the gap and ,

plenum volumes released upon clad perforation. An iodine decontamination factor (DF) of 100 (" Evaluation of Fission Product Release and Transport for a Fuel Handling Accident," G. Burley, USAEC, Revised Oct. 5,1971) is assumed for 23 feet of water cover, and appropriate airborne radionuclide filtration / mixing, if any, is applied in the analysis before release to the atmosphere. The decontamination factor is based, in part, on an analysis of work presented in WCAP-7518-L, " Radiological Consequences of a Fuel Handling Accident," M. J.

Bell, et al., June 1970, NES Proprietary Class 2." j For fuel handling accident offsite radiological consequence evaluations involving fuel assemblies with burnup) greater than 38,000 mwd /t batch average at discharge (

Guide 1.25 assumptions, but with modified gap and plenum fractiona] volatile radionuclide inventories. The fractional inventories range from a-few percent (less than the R. G. 1.25 10 percent recommendation) to as much au 40-50 percent for certain high burnups/radionuclide combinations. The gap and. plenum fractional inventories for the highest-power assembly are computed as a function of at least burnup, and at most time, temperature, and burnup using the GAPCON-THERMAL-2 computer code in conjunction with the ANS 5.4 fission gas release standard 20 l

(model) proposed by the American Nuclear Sc:iety in " Radioactive Gas Release from LWR Fuel", C. E. Beyer, draft NUREG CR-2715, April 1982. In generating these estimated fractional inventories, the conservative assumption of fuel assembly operation at a constant maximum-allowed peak linear heat generation rate (LHGR) for PWRs or MAPLHGR for BWRs is made. This assumption appears to be conservative within a factor of 2-3 for gap and plenum volatile inventories.

In addition to the conservative assumption regarding fuel assembly power operation noted above, there are two other significant sources of conservatism in the staff's analysis. The iodine decontamination factor (DF) assigned to the pool is taken to be a factor of 100. It can be inferred from the report upon which this factor is based (WCAP-7518-L) that this value is probably conservative by about a factor of three. Finally, plateout of volatile iodine released from the fuel into the gap and fuel rod plenum has been entirely neglected. Although not well quantified, a tentative estimate suggests that about 10 percent or less of the iodine released

into the gap will remain volatile at the fairly low temperatures after the fuel has-been allowed to cool for about a day or more.

$- Because of the significance of these conservatisms, the staff intends to study' and quantify them in more detail and to use the results of such evaluations to

- - - appropriately revise the staff's Standard Review Plan (SRP), NUREG-0800. In the interim, the staff concludes that consideration of all three factors together noted above may pennit a significant reduction of estimated thyroid doses com-pared to existing analyses. Adequate justification by licensees on a case-by-case basis, or by vendors on a generic basis, are likely to provide sufficient bases for departing from SRP criteria until such time as detailed changes can be made.

A reduction by a factor of two is likely to be appropriate and conservative.

i Consequently, with regard to evaluation of thyroid doses for fuel-handling accidents involving extended-burnup fuel (> 38,000 mwd / tonne) and, pending SRP revision, it is likely that justification can be provided for lower estimates of thyroid doses from fuel handling accidents by a factor of two in departures L from SRP review criteria.

7.0 REGULATORY POSITION The review of Baltimore Gas and Electric Company's submittal for gxtended burnup operation as described in CENPD-263-P, Revision 1-P and responses to questions has been completed. As noted in Section 2.0(g) above, the methods to predict shoulder gap spacings (between the fuel rod and top of the assembly) for 14x14 and 16x16 assemblies, and the evaluation methods used to predict the gap spacing between 16x16 fuel assemblies and core internals require data at higher fluences before they can be approved. Until this is done, predictions using the current CE models will have to be justified in individual applications for extended burnup operation. .

Apart from these exceptions, the criteria and methods used to evaluate extended burnup operation have been found acceptable to predict fuel perfohnance and physics characteristicstoburnuplevelsof45 mwd /kgM(batchaverage).?Ihetopical Report (CENPD269PRevision1)maybereferencedinlicensingactionsinvolving the design of extended burnup cores.

21

We reconnend that the licensees and CE should monitor fuel cycles with extended burnup fuel and should inform the staff of any significant deviation between prediction and measurement of various physics parameters.;[

Because of known conservatisms in the analysis of the Fuel Handling Accident, pending revision of the SRP, it is likely that justification can be provided for lower estimates of thyroid doses from fuel handling accidents by a factor of two.

'c - -

• e g h-en -

O

= 4 4 g.

G a -

e

' T

. 22

6.0 REFERENCES

1. Letter, L. S. Rubenstein, USNRC, to A. E. Scherer, Combustion . Engineering, June 2, 1981.  ;[

2. External Burnup Operation of Combustion Engineering PWR Fuel, CENPD-269-P, combustion Engineering, Inc., April 1982. -

3 Extended Burnup 03eration of Combustion Engineering PWR Fuel, CENPD-269-P, l Revision 1-P, Com)ustion Engineering, Inc., July 1984.

4. Letter from A. E. Lundvall, BG&E, to J. Miller, NRC, " Responses to Questions on Combustion Engineering CENPD-269-P," August 1984.

5. Standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants--LWR Edition, NUREG-0800, Section 4.2, " Fuel System

- Design," Rev. 2, July 1981. -

, 6. System 80 Standard Safety Analysis Re3 ort, Final Safety Analysis Re3 ort.

1- (CESSAR F5AR), 5TN-50-470F, Combustion Engineering, Inc., October 19/3.

, _ _ _ 7. Structural Analysis of Fuel Assemblies fc; Seismic and Loss of Coolant Accident Loading, CENPD-178-P, Rev. 1-P, Combustion Engineering, Inc., August 1981.

8. Letter from D. Eisenhut, NRC, to A. E. Scherer, CE, " Approval of CESSAR",

December 21, 1981.

9. Improvement to Fuel Evaluation Model, CENPD-161(B)-P, Combustion i Engineering, Inc., July 1981,

10. Letter from R. A. Clark, USNRC, to A. E. Lundvall, Jr, BG&E, " Safety EvaluationofCEN-161(FATES-3)," March 31,1983.

11. W. J. O'Donnell and B. F. Langer, " Fatigue Design Basis for Zircaloy Components," Nuc. Sci. Eng., Vol. 20, p. 1 (1964).

12. CE Evaluation Model Topical Report, CENPD-139, Combustion Engineering, Inc., July 1974.

13. H. Stehle, W. Kaden and R. Manzel, " External Corrosion of Cladding in PWRs," Nucl . Eng. and Design, Vol . 33, p.155,1975.

14 Fuel and Poison Rod Bowing, CENPD-225, Combustion Engineering, Inc.,

October 1976.

15. L.S.Rubenstein,NRC,memorandumforT.M.Novak,"SERsforiestinghouse, Combustion Engineering, Babcock & Wilcox, and Exxon Fuel Rod , Bowing Topical Reports," October 25, 1982. . , ,

23

16. H. Schenk, " Experience from Fuel Performance at KWO," SM-178-15, International Atomic Energy Agency, October 1973. .

!i

17. K. Kuffer and H. R. Lutz, " Experience of Commercial Power PfaR~t Operation in Switzerland," Fifth Foratom Conference, Florence Italy (1973).

18. " Robert Ennett Ginna Nuclear Power Plant, Unit 1 Final Safety Analysis Report," Docket Number 50-244, p.103, Rochester Gas and Electric Corporation, 1972.

19. Letter, J. R. Marshall, Arkansas Power & Light Company, to W. C. Seidle, NRC, Licensee Event Report No. 82-030/01T-0, October 6, 1982.

20. Jn-Reactor Dimensional Changes in Zircaloy-4 Fuel Assemblies, CENPD-

- 198-P and supplements, Combustion Engineering, Inc., December 1975.

21.- A) plication of CENPD-198 to Zircaloy Component Dimensional Changes,

, C N-183(B), Combustion Engineering, Inc., September 1981. -

1- --

22. Deleted

23. Arkansas Nuclear One, Unit 2 Cycle 4 Shoulder Gap Evaluation, CEN-261(A),

Combustion Engineering, Inc., November 1983.

24. Letter, A. E. Scherer, CE, to D. G. Eisenhut, NRC, Number LD-83-035, April 26, 1983.

25. B. Z. Hyatt, " Degradation of the Stress Relaxation Properties of Selected I91~ Reactor Materials in a Fast-Neutron Flux," WAPD-TM-881(L), Bettis Atomic Power Laboratory, March 1973.

26. Standard Specifications for Sintered Uranium Dioxide Pellets, ASTM

- Standard C776-76, Part 45 (1977).

27. Statistical Approach to Analyzing Cree) Collapse of Oval Fuel Rod Cladding Using CEPAN, CEN-182(BJA, Com)ustion Engineering, Inc., September 1981.

28. Deleted

29. CEPAN Method of Analyzing Cree) Collapse of Oval Cladding, CENPD-187, Combustion Engineering, Inc., March 1976.

30. Christensen, J. A., et al., " Melting Point of Irradiated Uranium Dioxide," TR ANS AM.NUCL. S0C., Vol 7 (2) 1964

~

31. Letter from H. Bernard, NRC, to A. E. Scherer, CE, "Acceptanch for Referencing of Topical Report CENPD-178(P)," August 6, 1982. y uL

, 24

32. Letter from A. E. Scherer, CE, to D. G. Eisenhut, NRC, Number LD-82-003, January 11, 1982.  :

- ?_ _.'

33. D. A. Powers, R. O. Meyer, Cladding Swelling and Rupture Models for LOCA Analyses, NUREG-0630, April 1980. ,

Y - -

• e

, N-ew

• . ,e 4

^ 't g' .

.- L r

.s 1

25