ML20106B289
| ML20106B289 | |
| Person / Time | |
|---|---|
| Site: | Vermont Yankee File:NorthStar Vermont Yankee icon.png |
| Issue date: | 07/31/1992 |
| From: | Beyer C BROOKHAVEN NATIONAL LABORATORY |
| To: | Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20106B264 | List: |
| References | |
| CON-FIN-I-2009 NUDOCS 9210010103 | |
| Download: ML20106B289 (21) | |
Text
-- -
TECHNICAL EVALUATION REPORT TECHPICAL EVALUATION REPORT OF THE FR05STEY2 FUEL PERFORMANCE CODE C, E. Beyer July 1992 Prepared for the Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Washington, D.C.
20555 under Contract DE-AC06-76RLO 1830 NRC FIN 12009 Pacific Northwest Laboratory Richland, Washington
\\;k 9210010103 920924 g
DR ADOCK 0500 1
vr t-
\\
.... ~
i NOMENCLATURE BOC beginning-of-cycle BWR boiling-water reactor 1
CFR Code of Federal Regulations 4
DNBR departure from nucleate boiling ratio E0C end-of-cycle E0L end-of life GE General Electric Company LHGR linear heat generation rate LOCA loss-of-coolant accident LWR light water reactor i
MCPR minimum critical power
.tio NRC U.S. Nuclear Regulatory Commission PNL Pacific Northwest Laboratory j
PWR pressurized-water reactor i
SAR safety analysis review i
l SER safety evaluation report SRP safety review plan I
TER technical evaluation report VYNPC Vermont Yankee Nuclear Power Corporation YAEC Yankee Atomic Electric Company 1
- j i i i
)
4
= - -
w a
-c-,
-,,-.-+v.+a,
-c
,..q.
.--,s...-
g wen.-me e.,,t---~
w --
,e ye y
4---.',-
-,e r e-wwe
i
.i j
i CONTENTS 1
1.0 INTRODUCTION
3 2.0 FISSION GAS RELEASE...............................................
l 4
3.0 FUEL THERMAL CONDUCTIV:TY AND RELOCATION..........................
5 4.0 FLUX DEPRESSION...................................................
i 6
5.0 COMPARISON OF CODE THERMAL PREDICTIONS TO DATA....................
7 6.0 GAD 0LINI A BURNABLE POISON (U0 Gd 0 ) PROPERTIES..................
2 g3 8
7.0 APPLICATION OF CODE FOR LICENSING ANALYSES........................
Il 7.1 END OF LIFE INTERNAL R00 PRESSURES.
12 7.2 CLADDING STRAIN..............................................
12 7.3 FUEL CENTERLINE MELTING...............................-.......
7.4 GAP CONDUCTANCE FOR TRANSIENT ANALYSES.......................
13 7.5 FUEL TEMPERATURE FOR PHYSICS ANALYSES........................
15 15
8.0 CONCLUSION
S.......................................................
16
9.0 REFERENCES
l '
V
1.0 INTRODUCTION
The thermal and mechanical performance of fuel in a light water reactor (LWR) during its operational lifetime must be described in the safety analysis of the loss of coolant accident (LOCA) as well as for other accidents, transients, a nd normal operat an.
The determination of stored energy and rod pressures for the LOCA analysis and other analyses requires a fuel pin thermal performance model that is capable of calculating fuel and cladding behavior, including the gap conouctance between the fuel and cladding, as a function of burnup.
The parameters Ofecting fuel performance, such as fission gas- -
/elaase, cladding dimensional changes, fuel densification, fuel thermal expansion, and fuel swelling, should be accounted for in the modes.
The FROSSIEY2 fuel performance code has been submitted by Vermont Yankee Nuclear Power Corporation (VYNPC) to the Nuclear Regulatory Commission (NRC) for approval to apply this code to analysis of LOCA initial condititns, j
initial conditions for transient and end of life (EOL) limiting analyses.
]
The original FROSSTEt fuel performance code was submitted by the Yankee Atomic Electric Company (YAEC) in References ? and 3 and supplemented in i
Re'erence 4 However, the NRC approval of this code (Reference S) limited its use to non LOCA analyses at low to mcderate exposure ranges for both pressur-I i:ed-water reactor (PWR) and boiling water reactor (BWR) applications.
Therefore, the criginal code was not applincie to high burnup or LOCA applications.
The FROSS!EYP fuel performance code (Ref erence 1) is a reformulation of tne F 0SSIEY code and has been verified with high-burnup data in order to remove the earlier restrictions. Changes in the FROSSTEY2 fuel performance code, relative to the original submittal in References 2, 3, and 4, are 5
primarily in the models for fission gas release, fuel thermal conductivity, fuel relocation, and flux depression across the fuel pellet.
In addition, material properties for U0,-Gd;0 were included in FROSSTEY for burnable 3
poison rods but were not formally reviewed in the original FROSSTEY submittal (Reference 5),
these same material properties for burnable poison rods are 1
i 1
l included in FROSSTEY2 and approval for use is iecluded in the VYNPC submittal to the NRC. Therefore, based on the model changes and requested approval for applications, this Technical E,aluation Report (TER) of the FROSSTEY2 code is divided into seven major sections:
Fission Gas Release,. Fuel Thermal Conductivity and Fuel Relocation, Flux Depression, Comparison of Code Thermal Predictions to Data, Gadolinia Burnable Poison (L'O, Gd,0 ) Properties, 3
Application of Code for Licensing Analyses, and Conclusions.
During the course of this review VYNPC changed the solid gap conductance model and this change will be discessed in the section entitled Comparison of Code Thermal Predictions to Data.
pacific Northwest Laboratcry (PNL) has acted as a consultant to the NRC in tnis review.
As a result of the NRC staff and their PNL consultant's review of the submitted report (Reference 1), a list of questions requesting clarification was sent by the NRC to VYNPC (References 6 and 7).
VYNPC responded to those questions in References 8 and 9.
A review of the responses concluded that VYNPC had not provided sufficient information on how they intended to apply the code for LOCA analyses nor did they adcquately address the problems with the code predictions of fission gas release and fuel temperatures (Reference 10).
S,nne of the questions from the original requests (References 6 and 7) were restated in Referense 10 to gain further information on FROSSTEY2 code applications for licensing analyses and identified those responses (References 8 and 9) that were unsatisfactory or incomplete.
VYNPC responded partially to these questions in References 11 and 12; however, complete responses to all questions that superseded those in References 11 and 12, were provided in Reference 13.
A review of these second round re-sponses concluded that the changes made by VYNPC had resolved the problems with the code predictions of fission gas release and fuel temperatures, but that the code was not being applied in a conservative manner consistent with previous NRC approv:'s for LOCA analysis methods using best estimate codes.
Therefore, a summary of different analytical approaches for maintaining conservatisms in code predictions of stored energy for input to LOCA analyses was provic1d by PNL-in Reference 14 for guidance to VYNPC.
In addition, a brief summary of the status of this review was provided in Reference 15, along with a restatement of the unresolved issue, i.e., the lack of conservatism in l
2
FROSSTEY2 applications to licensing analyses.
in order to try and resolve those unresolved issues, VYNPC submitted responses in References 16 and 17.
PNL has completed the review and written this TER based on References 1, 8, 9, 11, 12, 13, 16, and 17, 2.0 FISSION GAS RELEtSE The review of the original FROSSIEY2 submittal (Reference 1) revealed that the code underpredicted fission gas release data for particular con-ditions.
VYNPC was questioned about this underprediction and was asked to supply information on individual measured data and code predictions of these data (References 6 and 7).
A review of the fission gas release data supplied in VYNPC's responses revealed that the code significantly underpredicted both steady state and transient release data above a critical level of release.
Release values above this critical level are important for pred': ting EOL rod pressures at extended burnup levels.
VYNPC implied in their response (Reference 8) that this uncerprediction was acceptable because the uncertainty in '.he data did not allow for a more accurate prediction.
ML's rev'ew of this response indicated that other fuel vendor's fuel
- er'vmance coces nave been able to provide a much better prediction of this
- are 'ission gas release data than that provided by FROSSTEY2 (Reference 10).
Therefore, it was concluded by-PNL that the FROSSTEY2 underprediction of
:s mn gas release was unacceptable and, therefore, must be resolved before me :oce could be approved for licensing analyses of EOL rod pressures.
It was further concluded that VYNPC needed to account for uncertainties in both
- ne steady state and transient power histories that are used for licensing analyses, such as for calculating EOL rod pressures.
VYNPC was requested (Reference 10) to provide a second round of-responses in order to address the unresolved issues on fission gas release and power histories in the first round of responses.
VYNPC responded, based on the second round questions, by altering the i
fission gas release, solid gap conductance, fuel relocation, and fu., thermal conductivity models (Reference 13).
These changes resulted in the code, on 3
1
i l
the average, overpredicting fission gas release data under both steady state and transient, i.e., power bumping, rod power conditions.
The methodology proposed by VYNPC for determining the rod power histories for the EOL rod pressure analyses were also found to be conservative.
Therefore, PNL con-cludes that the methodology described by VYNPC for determining EOL rod pressures is satisfactory for licensing analyses.
However, because the verification of the model is limited to data with maximum rod average burnup levels of 60 GWd/MTM (approximate peak pellet is 66 GWd/MTM for a BWR), it is recommended that the code be limited to a rod average burnup level of 60 GWd/MTM for both BWR and PWR appitcations.
3.0 FUEL THERMAL CONDUCTIVITY AND RELOCA QOS The fuel thermal conductivity and relocatiori models are two of the most important for predicting fuel temperatures, along with the fuel to-cladding gap conductance model. These two models are generally two separate indepen-dent models in most fuel performance codes.
However, in the original FROSSIEY2 code (Reference 1) they ware empirically related to each other to give a best estimate prediction of a given set of measured centerline temper-ature data.
This originti version of FROSSTEY2 assumed that the fuel re-located to eliminate tM fuel-to cladding gap.
Tnis resulted in a high gap conductance, i.e., small temperature delta T across the gap.
The fuel thermal conductivity was then decreased in order to raise the delta-T across the fuel a corresponding amount and, thus, result in a match to the measured fuel centerline data.
Therefore, the FROSSTEY2 code assumed that most of the delta temperature drop existed across the fuel 9ellet and only a small delta temperature drop existed across the fuel-to cladding gap.
This assumption is significantly different from the previous FROSSTEY code (References 2 and 3), the GT2R2 code (Reference 18) used by NRC for auditing industry codes, and current fuel vendor codes.
It should also be noted that this assumption reduces the calculated-stored energy for LOCA. A fuel performance code entitled ESCORE (Reference 19), that has-been submitted to NRC for licensing applications, has made a similar assumption regarding fuel relocation and thermal conductivity.
The NRC Safety Evaluation Report 4
_______________....___.___._..._.J
(SER) (Reference 20) has concluded that the increase in fuel relocatica and j
l corresponding decrease in fuel thermal conductivity was not appropriate for licensing applicat!ons because of the high degree of uncertainty in the data that are claimed to support *.'1ese changes.
It was concluded (Reference 10) that the fuel relocation and thermal conductivity models in FROSSTEY2 were not appropriate for licensing applications for the same reasons as discussed in the NRC SER of the ESCORE code (Reference 20).
VYNPC responded in their second round of responses (Reference 13) by stating that the changes to the fuel relocation model were removed and the current FROSSTEY2 models were identical to the FROSSTEY models (References 2 anc 3).
VYNPC also claimed that these changes made the FROSSTEY2 code somew Mt conservative in predicting fuel temperatures.
However, examination 9 FRCSSTEY2 predictions of the latest experimental centerline temperature w.a from fuel rods typical of commercial fuel designs and at linear heat
- ew ration rates (LHGRs) important to licensing applications, i.e., >10 kW/ft.
4 se code aptes.'s to provice either a slight underprediction or a best estimate crediction of fuel centerline temperatures at low to moderate burnup levels (see Figur)s 2a 1, 2a 4, 2a 7, and 2a-10 of Reference 13).
The changes to the fuel relocation and thermal conductivity models (Reference 13) in the FRCSSTEY2 code are acceptaole for licensing applications if the code is applied using a conservative methodology for calculating fuel temperatures as discussed in Sections 5.0 and 7.0.-
4.0 FLUX DEPRESSION The FROSSTEY2 code includes the RADAR model for calculating the radial power profile across the fuel pellet due to flux depression (Reference 1).
The RADAR model was originally developed by British Nuclear Fuels Limited (Reference 21) and is currently used in the NRC audit code GT2R2 (Reference 18).
PNL h.s performed comparisons of RADAR calculated radial power distributions to those calculated with'more sophisticated and accurate physics codes. These comparisons have demonstrated that the RADAR code becomes less accurate as fuel burnup levels increase because it fails to accurately predict the plutonium buildup at the 'uel surface and, therefore, 5
4
RADAR underpredicts power production at the surface.
PNL has also evaluated the error in calculated fuel centerline temperature due to the underprediction and found less thar a 1% error at a rod average burqup level of 50 GWd/MTM.
Therefore, PNL concludes that the use of the RADAR model is satisfactory for licensing applications.
However, it is recommended that the FROSSTEY2 code be limited to a red average burnup level of 60 GWd/MTM for licensing appli-cations.
5.0 COMPARISON OF CODE THERMAL PRE.01CT10NS TO DATA The FROSSTEY2 submittal (Reference 1) provided a comparison of FROSSTEY2 predicted versus measured centerline temperatures but did not identify the specific data on the plots.
This prevented PNL and NRC reviewers from identifying how well the code predicted particular experimental fuel rod data.
The ability to make this distinction was important because some of the data are judged to be more applicable to verification of fuel performance codes used for licensing applications than other data.
Those specific data judged
/
to be more applicable are those from experimental fuel rods typical of today's
'uel designs, operating at LHGRs typically used for licensing calculations, and the best characterized and reliable experimental data, i.e., the most recent temperature data from Halden experimental programs.
Therefore, VYNPC was requested (References 6 and 7) to provide FROSSTEY2 code predictions of Halden measured centerline temperature data from experimertal assemblies IFA 432 and IFA 513.
VYNPC responded to this request by supplying FROSSTEY2 predictions of Rods 1, 2, 3, 5, and 6 of IFA-432 and Rods 1, 2, and 6 of IFA 513; however, the corresponding measured data were not supplied (References 8 and 9).
Examination of the FROSSTEY2 code predictions in Reference 8 indicated that FROSSTEf2 underpredicted the measured temperatures for Rod 3 of IFA-432 s
when burnup levels exceeded 15 to 20 GWd/MTM.
VYNPC was questioned on the reason for this uncerprediction (Reference 10).
In the second round of responses, VYNPC rt plied that the underprediction was due to unrealistically high interfacial p'essures predicted by the code between the fuel and cladding as burnup levels,ncreased.
Consequently, this led to unrealistically high 6
gap conductance values (Reft.Nnce 13). YYNPC has ieduced the maximum possible interfacial pressures in the code by a fractional amount of the original FROSSTEY2 calculational values. This has resulted in a conservative over-prediction of centerline temperatures for Rod 3 of IFA 432.
VYNPC has implied (Reference 13) that by changing the fuel relocation and thermal conductivity models in FROSSTEY2 to be the same as those in FROSSTEY (see discussion in the Fuel Relocation and Thermal Conductivity section of j
this report) that the code provides a more conservative prediction of fuel thermal conditions than for the original FROSSTEY2 code submitted in Reference 1.
PNL acknowledges that this may be true for volume average fuel temperatures, i.e., fuel stored energy, but can not be verified.
- However, frcm examination of the fuel centerline predictions in Figures 2a 1 through 2a-12 of Reference 13, it is judged that the current FROSSTEY2 code provides a best estimate prediction of fuel centerline temperatures from experimental fuel rods typical of today's commercial fuel designs.
PNL, therefore, concludes that the FROSSTEY2 code is primarily a best estimate code and may be applied to predicting fuel thermal conditions in licensing applications if used in a conservative manner in the following two areas:
- 1) account for uncertainties in the code input (e.g., uncertainties in fuel dimensions, fuel rod dimensions, and operating conditions) and 2) uncertainties in the code predictions.
The conservatisms that must be applied in licensing applications are discussed further in the section entitled Application of Code for Licensing Analyses.
6.0 GA00LINI A BURNABLr, POISON ljlQg4Q ) PRODERTIES 3
VYNPC uses General Electric Company (CE) 00gGd 0 properties that have 3
previously been reviewed and approved by NRC (Reference 22).
Therefore, thase properties do not need to be reviewed again; however, the methodology of applying the FROSSTEY2 code and the UO -Gd 0 pror ties for evaluating 2
g3 burnable poison rod behavior for licensing analyses does need to be reviewed.
The application of FROSSTEY2 for evaluating burnable poison rod analyses is the same as for evaluating fuel rods and will be reviewed in the following section.
7
i i
7.0 APPLICATION OF CODE FOR LICENSING ANALYSES i
VYNPC was requested to supply information in the first round of questions l
(References 6 and 7) on how they planned to maintain their technical expertise in the use of FROSSTEY2 for licensing applications.
VYNPC responded that YAEC I
has maintained a fuel modeling function for over 15 years and during this time have maintained procedures for FROSSTEY and GAPEX cnde applications.
VYNPC 1
intends to establish similar procedures for FROSSTEY2 applications based on f
NRC's approval of the code. VYNPC has further indicated that currently more than eight engineers in their organization.have experience in. fuel performance i
code use.
PNL concludes that VYNPC has a plan to maintain technical expertise in the use nf FROSSTEY2 for licensing applications and that this plan is acceptable.
i The application of a best estimate code for licensing analyses requires, as noted above, that code input uncertainties and code calculational un-certainties must be applied in a conservative manner.
VYNPC was requested to provide FROSSTEY2 calculational example
- of how the code was to be applied for f
each licensing application and identify the conservatisms in their code input j
(References 6 and 7).
In addition, VYNPC was requested to crovide a descrip-
]
tion of the analysis methodology that demonstrates that the code application l
is conservative and bounding for each licensing application, e.g., uGCA, EOL rod pressure, cladding strain, fuel melting, minimum critical power ratio (MCPR) for BWR, and departure from nucleate boiling ratio (DNBR) for PWR I
l analyses, and to quantify conservatisms.
The analysis application for l
ournable poison rods is the same as for fuel rods.
The use of the same analysis approach for both burnable poison rods and fuel rods is acceptable.
VYNPC's response was reviewed (References 8 and 9) and it was concluded I
that the description of the conservatisms applied to the FROSSTEY2 code for i
determining the stored. energy for input to-the LOCA analysis were not ade-quately identified.
Therefore, VYNPC was again raquested to identify and quantify the conservatisms apolied to the code pred;ctions and code input for calculating stored energy for LOCA (Reference 10).
It was also suggested that VYNPC use the code comparisons to centerline temperature data for quantifying i
i I
I 1
1 l
i
the code calculational uncertainties and for quantifying the conservatisms i
that need to be applied to the code's best estinate prediction of fuel stored J
In the second round of responses VYNPC provided example FROSSTEY2 energy.
calculations for input to LOCA utilizing both nominal, i.e., best estimate, input values and licensingc i.e., conserveive, input values but provided no j
estimate of the code calculational uncertainties (Reference 13).
j 1
1 PNL provided a detailed description in Reference 14 of those conserva.
tisms that are required for utilizing best estimate fuel performance codes to l
calculate fuekstored energy. for_ LOCA._Inaummary,_those.conservatisms_are:.
- 1) use of maximum LHGRs allowed by plant technical specifications as code input: 2) use of worst case fuel rod dimensions as allowed by fabrication
{
specifications as code input; and 3) application of code calculational uncertainties, i.e., 95% upperbound probability at a 95% confidence level, to 2
i the code's best estimate crediction of stored energy.
l I
Following two conference calls with VYNPC, it became apparent that VYNPC only intended tc :;;,iy input uncertainties to FROSSTEY2 for determining fuel f
stored energy for input to LOCA analyses.
VYNPC believed that the conserva.
l tisms applied to the input for the FRCSSTEY2 code more than compensated for I
the code calculational uncertainties.
PNL prepared a letter (Reference 15) l tr.at examined VYNPC's claim that the FROSSTEY2 codes conservative input more j
than comDensated for the code's calculational uncertainties, i_.e., o code input n a code calculation.
PNL concluded (Reference 15) that VYNFC's claim of adecuate conser"atisms was not valid because the FROSSTEY2 code input un-urtainties were approximately equal to the code calculational uncertainties, i
1~
VYNPC provided a third round of responses (Reference 16) to address the t
FROSSIEY2 calculational conservatisms for licensing applications for LOCA.
j This third set of responses for LOCA application was reviewed and discussed j
-with staff _trom VYNPC and NRC, and NRC's consultant from PNL, at a meeting i
that was held at NRC Headquarters on April 7, 1992.
From this meeting, the NRC staff and their PNL consultant concluded that the basic approach used by VYNPC for determining experimental uncertainties and FROSSTEY2 code calcula-I tional uncertainties was acceptable with two exceptions The first exception 9
3
recomended a deletion and en addition to the experimental rods used in the VYNPC fuel temperature data base.
The second exception recomended that VYNPC include the response surface uncertainty in their estimate of FROSSTEY2 code uncertainty. At tht conclusion of this meeting, VYNPC was requested to provide a fourth set of responses with these revisions to their calculation of LOCA stored energies and to provide a description of how FROSSTEY2 would be applied for non-LOCA applications.
The fourth set of VYNPC responses (Reference 17) with the above requested revisions in oroer to resolve the two exceptions for code application to LOCA-stored energy was found to be satisfactory.
The responses for non-LOCA application have one particular common problem with the methodology that will be discussed at this time, ihose code methodologies for non-LOCA applications that are unique to the particular analysis will be discussed in the sub-4 sections for those analyses, e.g., EOL rod pressure, cladding strain, fuel melting, gap conductances for transient analyses, and fuel temperatures for physics analyses.
The common VYNPC methodology for all non LOCA analyses is the use of nominal fabricated dimensional values for input to FROSSTEY2.
This use of nomin&1 f abricated input is satisf actory for calculating core average con-ditions such as for core average fuel temperatures for physics analyses and core average gap conductances for departure from-nucleate boiling analyses.
However, this is not satisfactory for " hot channel or high power rod" analyses, e.g., for EOL rod pressure, cladding strain, fuel melting, and hot channel gap conductances for transient analyses.
Previously approved NRC methodologier for the " hot channel or high power rod" analyses from the fuel vendors have required that bounding f abricated dimensions be used to provide con'servative output results for the specific
- analysis, Therefore, it is recommended that for hot channel or high power rod analyses that VYNPC use either bounding fabricated specifications for input to these analyses or account for the fabrication uncertainties in the output results as done for the LOCA analyses.
Therefore, the code predictions of EOL rod pressure, cladding strain, fuel melting, and hot channel gap conductances 10
j i
for transient analyses are to be appropriately conservative based on the sncertainties in the input values.
This specific issue for each application 2
is also discussed in the following Sections 7.1, 7.2, 7.3, and 7.4.
l 1
7.1 END OF LIFE INTERNAL ROD PRESSVRES The input power history for calculating EOL internal rod pressures is. "-
of the most important input parameters for this analysis.
VYNPC has proposed to use a maximum expected bounding rod power history with a nominal axial
]
power shape based on physics reload analyses for the plant / cycle in question.
In order to simulate transient power operation, VYNPC superimposes a signifi-l cant number of transient axial power shapes throughout the irradiation life of the bounding power rod that allows the peak node to be at the maximum LHGR technical specification lirrit (MAPLHGR for a BWR and the F) limit for a PWR) for a brief period for each reload analyses.
This VYNPC power history allows l
for a conservative prediction of fission gas relea:s for the peak operating rod in the core and, therefore, is satisfactory for this licensing appli-i j
cation.
l The FROSSTEY2 code-calcuiated parameters that are important to the determination of tne rod inttrnal pressures are fission gas release and internal rod void volume.
As noted earlier in Section.2.0 Fission Gas Release, the FROSSTEY2 code provides a conservative overprediction of fission gas release.
PNL has concluded that the code's overprediction of fission gas release covers the code's calculational uncertainties of this prediction, i
i I
The original comparison of FROSSTEY2 predictions to internal rod void volume data as measured from high burnup fuel-rods, provided in Reference 13 i
per NRC's request in Reference 10, demonstrated that the code significantly underpredicted internal rod vo;d volumes.
This underprediction results in a significant degree of conservatism in the FROSSTEY2 rod pressure calculation.
However, as a result, VYNPC made a correction to the calculation of dish volume in the FROSSTEY2 code that provides a much better comparison to measured data (as shown in Figure 3.11 in Reference 16).
VYNPC has claimed that the FROSSTEY2 code still provides 4 sli, t-('onservasive) underprediction g
11 1
of internal void volume (Reference 17).
PNL concludes that the use of bounding f abricttad values for plenum and dish volumes as recommended above will adeouately cover tne code calculational uncertainties in void volumes.
In addition, the use of lower or upper bounding values of fuel to cladding gap size for the rod pressure calculation should be based on those gap size values based on uncertainties in rod fabrication that provide the most conservative (highest) prediction of rod pressures.
Therefore, PNL concludes that FROSSTEY2 calculational uncertainties in internal rod pressures are adequately covered by inherent conservatisms in the code, the power history (steady-state and transient), and the above recom-mended uncertainties in fabrication for input to the code.
7.2 CLADDING STRAIN The input steady-state power history used for the FROSSTEY2 calculation of cladding plastic strain is the same as the bounding power history used for the EOL rod pressure analysis with power ramping up to the power level that results in 1% plastic strain at periodic burnup levels.
This technique i used to determine the power level of the 1% plastic strain limit, per the hi.
Safety Review Plan (SRP) (Reference 23).
PNL concludes that this analysis methodology is satisfactury when conservatively bounding uncertainties due to fabrication inputs to FROSSTEY2 are applied, per the recommendations given above.
7.3 FUEL CENTERLINE MELTJES The input power history used for the FROSSTEY2 calculation of fuel melting is similar to that used for cladding strain except the periodic ramped powers are taken to the point of fuel melting.
In 4ddition, VYNPC takes into account the uncertainties in FROSSTEY2 calculated fuel centerline temperatures at a 95% probability with a 95% confidence lev 91 for fuel melting in the same manner as the FROSSTEY2 application to LOCA-stored energy with the exception that nominal fabrication input data are used.
As noted earlier, PNL recom-mends that VYNPC also utilize conservatively bounding uncertainties due to 12
- 1 i
t 1
fabrication input to FROSSTEY2 are applied for the fuel melting calculation as done for LOCA applications.
PNL concludes that this analysis methodology is satisfactory when conservatively bounding uncertainties due to fabrication inputs to FROSSTEY are applied per the recommendations in the above paragraph.
7.4 GAP CONDUCTANCE FOR TRANSIENT ANALYSES Core reload analyses require that all Safety Analysis Review (SAR) transients are analyzed to determine the MCPR limits for BWRs and DNBR limits for PWRs at various exposures for each cy>:le.
These transient analyses are f
d1vided into core wide system responses and a hot channel response that require estimates of the core wide gap conductance and the hot channel gap conc.ctance.
At this point, i t should be stressed that the FROSSTEY2 gap 1
conductance model itself is satisf actory for this application, however, the 1
issue is in the code application that involves the use of a nomdnal f abri-cation input for hot channel analyses, and a constant ax;al power sh]oe for all exposures at steady-state initial conditions prior to the transient (for BWRs).
It is noted that the FROSSTEY2 code is only used for initial steady-state conditions prior to the transient.
However, it sn0uld also oe noted that the use of a constant axial power shape and axial gap conductance during the transient for delta CPR analyses may also not be' appropriate for some BWR fuel designs.
l The VYNPC methodology for the use of FROSSTEY2 for determining the core-wide gap conductance is based on nominal fabrication input values and a volumetrically-weighted average gap conductance of each of the fuel types in the core and a constant axial power shape for both BWRs and PWRs.
The use of nominal fabricatior, input values is acceptable for determining core wide gap conductance for both BWR and OWR applications because this is an average valve for the whole core.
However, the use of constant power shapes is_ judged to be acceptable only for PWR applications.
The use of constant axial power shapes is not acceptable for all BWR designs because the steady-state core axial power shape can change significantly from beginning-of-cycle (BOC) to end-of-
^
13
,.--,.----.n y.
,--n,-
,.,-,,,_..,m,
,<p.-m.4
,,w,.,,,,.-,m-.,vg e,y---
r
,.,,,,-r-
cycle (E0C) and this change may be different for different fuel designs, it is also noted that the axial power shape will also change during the course of a transient.
The magnitude of this change in axial power shape from steady-state operation during the cycle is dependent on the BWR fuel design, axial variations in enrichment, core loading pattern, and control blade withdrawal pattern.
The change in axial power shape during a transient is also dependent on these same factors, but in addition the power magnitude and the duration of the transient is depen 6ent on the initial steady-state axial shape and the typt of transient.
The NRC is aware that core and hot bundle axial power shape changes can significantly imprct delta CPR calculational results for some BWR fuel designs and transient types.
Based on this NRC experience, PNL recommends that VYNPC include the changes in* steady state axial power shape during the cycle exposure for initialization of the transient and also the change in core axial power shape during the transient for determining delta-CPR limits with exposure for BWRs.
i j
The VYNPC methodology for determining the hot channel gap conductance is l
to use nominal as fabricated input values, and assume that the peak node is j
operating at the maximum average planar linear heat generation rate (MAPLHGR) limit for the BWR and that the peak node and rod radial powers are at the technical specification limits for a PWR.
The axial power shapes for both f
BWRs and PWRs are a constant chopped cosine.
This is acceptable for PWR applications, however, this may mat be conservative for determining delta-CPR
~
i limits for BWRs because, as noted above, the BWR change in axial power shape for initial steady state and transient conditions can impact delta CPR results.
Therefore, PNL recommends that VYNPC include the changes in steady-1 state and transient axial power shapes and the associated change in axial gap j
conductance during steady-state and transient conditions with exposure in their hot channel analyses of delta CPR.
For the hot chaanel gap conductance at each exposure level, VYNPC calculates an axially dependent gap conductance l
for each rod type in a BWR bundle and each axial rod type segment is averaged to produce an axially dependent gap conductance.
The'VYNPC methodology for
[
calculating the hot channel gap conductance for a PWR is based only on the limiting rod ir the bundle and is a power-weighted average of each of the I
axial segments in the limiting hot rod.
Both of these averaging techniques 14 i
r_rm..
---m.
c.._.y
.,m.w,,~,
m,-
.,-r,m.,
r.. e. ~ -
..m._,,,--
m,..
.--.,-w,-4.,..o_.,
J for gap conductance of the limiting bundle and rod are acceptable for BWR and i
J PWR applications, respectively.
4 i
3 In surtcary, PNL concludes ; hat the VYNPC methodology for determining hot channel gap conductance for PWRs is acceptable when conservatively bounding i
uncertainties due to fabrication are applied per the above recomendations.
I The VYNPC methodology for calculating hot channel gap conductance for MCPR l
limits for BWRs is clso found to be acceptable when conservatively bounding l
uncertainties due to fabrication are applied and the hot channel axial gap
)
conductance is determined based on the change in axial power shape during f
steady-state and the transient for the delta CPR analysis.
In addition, PNL I-recomTends that VYNPC model the change in both initial s'aady state and transient axial power shape with exposure for the delta CPR analysis.
i I
)
l 7.5 FUEL TEM'ERATURE FOR PHYSICS ANALYS,fj i
j The VYNPC physics code which calculates the three dimensional core f
response to reactivity char.ges uses a volume average fuel temperature re-j lationship versus power at various exposure intervals.
This volume average fuel relationship versus rod power at different exposures for different fuel types is to be calculated with FROSSTEY2 using an average of axial power shapes from previous applicable cycles.
PNL concludes that this analysis methodology is satisfactory for application to physics calculations because high accuracy in fuel temperatures is not required for this calculation.
8.0 CONCLUSION
S I
PNL has reviewed the documentatiore on the opplication of the FROSSTEY2 fuel per$ormance code to LOCA and non-LOCA licensing analyses (References 1, 8, 9, 11, 12, 13, 16, and 17) in accordance with Section 4.2 of the SRP.
PNL.
concludes that the FROSSTEY2 code is acceptable for licensing applications with the following recommended conditions.
a 15 W'r&
a'Trry
=---yog:py-g-y ps e q -
p3--s ery 9--p9r 9gww Ig-,e a-a--*W-rrm et as - Mm"be1--+--mes'ry'gw -mre,w gr= q v ger 4 t N"g='1'-
yv' TN rwv4r w *W e w-T w
-*'*y UPN-'
F-F F-
l
(,
1 1.
The FROSSTEY2 code is to be used for licensing applict.tions only up to a j
maximum rod average burnup level of 60 GWd/MTM per recommendations in Sections 2.0 and 4.0 for both SWRs and PWRs.
2.
Non LOCA applications that involve the hot channel or peak power rod shall use either bounding fabricated specifications for input or, as done in the VYNPC LOCA analysis, account for the uncertainties in fabrication in a mnner conservative to the analysis results, i.e., internal rod i
pressure, cladding strain, centerline temperature, and gap conductance i
results per Section.7.0.
j 1
l 3.
For calculation of core wide and hot channel gap conductances for determining BWR MCPR limits, VYNPC needs to include the effect of changes in axial power shape with exposure on delta CPR limits for each BWR design utilized b.v VYNPC per Subsection 7.4.
The FROSSTEY2 code is only used for steady-state calculations ar.d. therefore, the FROSSTEY2 initial-i ization of fuel temperatures and gap conductance for the delta-CPR analysis needs to include the change in axial power shape with exposure during steady state operation.
9.0 REFERENCES
t l
1.
Letter, L. W. Capstick (VYNPC) to V. L. Rooney (NRC).
December 16, 1987, i
Subject:
Vermont Yankee LOCA Analysis Method FROSSTEY Fuel Performance Lode (FROSSTEY2).
Letter number FVY 87-116, Vermont Yankee Nuclear Power Corporation, Vermont.
2.
Letter, D. E. Vandenberg (YAEC) to Office cf Nuclear Reactor Regulation (NRC).
May 18, 1981.
Subject:
Submittal of Documentation of YAEC Fuel Performance Code (FROSSTEY) YAEC-1249P, entitled " Methods for the Analysis of 0xide Fuel Rod, Steady-State Thermal Effects (FROSSTEY)."
Letter number FVY 81-82.
Yankee Atomic Electric Company, Framingham, Massachusetts.
3.
Schultz, S, P., and K. E. St. John.
June 1981. Methods for the Analysis of Oxide fuel Rod Steady-State Thermal Effects (FROSSTEY) Code C'valifi-i cation and Application.
YAEC-126SP, Yankee Atomic Electric Company, Framingham, Massachusetts.
l 16
,,,,-,p.
-e.-,-,,-., -. _...., -, >., - -., -.. -
.e,-
.-m r-
-,e
,,,,n n,-g
,, - - + - -.,....,.
y<
n- - ---.,
i, 4
1 4
Letter, L. H. Heides-(YAEC) to D. G. Eisenhut (NRC).
March 17, 1983.
Subject:
Response to Questions on Yankee Atomic Electric Company Fuel Performance Code (FROSSTEY).
Letter number FVY 83-22, Yankee Atomic J
Electric Company, Framingham, Massachusetts.
I 5.
Letter, D. B. Vassallo (NRC) to R. W. Capstick (VYNPC).
September 27, i
1985.
Subject:
Approval of Use of Fuel Performance Code FROSSTEY.
U.S.
j Nuclear Regulatory Commission, Vashington, D.C.
6.
Letter, M. B. Fairtile (NRC) to R. W. Capstick (VYNPC).
May 2, 1989.
Subject:
Request for Additional Information - FROSSTEY2 Fuel Performance l
j Code (TAC No. 68216).
U.S. Nuclear Regulatory Commission, Washington, i
D.C.
l 7.
Letter,'M. BEFairtile~ (NRC) ~ to RTW7Capstick-(VYNPC); May 24,--1989. ~
~
Subject:
Second Request for Additione Information - FROSSTEY2 Fuel 4
Performance Code (TAC No. 68216).
U.S. Nuclear Regulatory Commission, i
Washington, D.C.
l 8.
Letter, R. W. Capstick (VYNPC) to Document Control Desk (NRC).
July 14, 4
1989.
Subject:
Response to NRC Request for Additional Information on the FROSSTEY2 Fuel Performance Code.
Letter number BVY 89 65, Vermont l
Yankee Nuclear Power Corporation, Vermont.
9.
Letter R. W. Capstick (VYNPC) to Document Centrol Desk (NRC).
August 4 j
1989.
Subject:
Supplemental Information on the FROSSTEY2 Fuel Per-formance Code.
Letter number BVY 89 74, Vermont Yankee Nuclear Power Corporation, Vernont.
10.
Letter. M. B. Fairtile (NRC) to L. A. Tremblay (VYNPC).
March 9, 1990.
Subject:
Request for Additional Informtion FROSSTEY2 Fuel Performance Code (TAC No. 68216).
U.S. Nuclear Regulatory Commission, Washingtor.,
D.C.
11.
Letter. L. A. Tremblay (VYNPC) to Document Control Desk (NRC).
April 19; l
1990.
Subject:
Responses to Request for Additional Information on i
FROSSTEY2 Fuel Ferformance Cede.
Letter number BVY 90 045, Verrront Yankee Nuclear Power Corporation, Vermont.
12.
Letter, L. A. Tremblay (VYNPC) to Document Control Dest (NRC).
May 10, 1990.
Subject:
Supplemental Information to VYNPC Apr','. 19, 1990 Response Regarding FROSSTEY2 Fuel Performanesi Code.
Letter number c
EVY 90-054, Vermont Yankee Nuclear Power Corporation, Vermont.
13.
Letter, L. A. Tremblay (VYNPC) to Document Control Desk (NRC).
March 6, 1991.
Subject:
Responses to Request for Additional Information on FROSSTEY2 Fuel Performance Code.
Letter number BVY 91-024, Vermont Yankee Nuclear Power Corporation, Vermont.
14.
Letter, C. E. Beyer (PNL) to S. L. Wa (NRC).
August 19, 1991.
Pacific Northwest Laboratery, Richltnd, Was: ngton.
17
.,m.
.m
.~._._,<m-,-
m...._,
l.
15.
Letter, C. E. Beyer- (PNL) to S. L. tiu (NRC).
December 23, 1991.
Pacific Northwest Laboratory, Richland. Washington.
16.
Letter, L. A. Tremblay, Jr. (VINPC) to Document Control Desk (NRC).
March 27. 1992.
Subject:
LOCA-Related Responses to Open Issues on FROSSTIY2 Fuel Performance Code.
Letter number BYY 92 39 Vermont,'ankee Nuclear Power Corporation, Vermont.
17.
Letter, L. A. Tremoley, Jr. (VYNPC) to Document Control Desk (NRC).
May 15, 1992.
Subject:
FROSSTEY2 Fuel Performance Code - Vermont Yankee Response _to Remaining Concerns.
Letter number BVY 92-54, Vermont Yankee Nuclear Power Corporation, Vermont.
18.
Cunningham, M. E., and C. E. Beyer.
1984.
GT2R2:
An Updated Version of GAPCON THERMAL-2. NUREG/CR 3907 (P1L 5178), Pacific Northwest Labor 6 tory, Richland, Washington.
19.
ElectHc Power Research Institute.
February 1997 ESCORE The EPRI Steady State Core Reload Evaluator Code:
Genersi Description.
EPRI NP 5100 (Final Report), prepared by 1. B. Fiero, Program Nanager, Combustie i Er.gineering, Inc., for the Electric Power Research Institute, San Jose. Califernia.
20.
Letter, A. C. Thadant (NRC) to C. R. Lehmann (PPLC).
May 23, 1990.
Subject:
A:ceptance for Referencing of Licensing Topical Report EPRI NP 5100.
U.S. Nuclear Regulatory Commission, Washin9 ton, D.C.
21.
Palmer,1. D., K. W. Hesketh, and P. A. Jackson.
1982.
"A Model for Predicting the Radial Power Profile in a Fuel Pin."
Presented at the IAEA Specialists Meeting on Water Reactor Fuel Element Performance Computer Modelling, Preston, United Kingdom.
22.
Akerlund, S. O et al.
June ), 1984, "The GESTR-LOCA and SAFER Models for the Evaluation of the loss of the loss of Coolant Accident -
Volume 1: GESTR LOCA A Modci for the Predi-t;on of Fuel Rod Performance." NEDE 32785 1-P-A, Appendix B.
23.
U.S. Nuclear.'.egulatory Commission.
July 1981.
"Section 4.2, Fuel System Design."
In :tandard Review Plan for the Review of Safety Analysis Reports for Nuclear Pouer l'iants - LWR Edition.
NUREG 0800, Rev. 2, U.S. Nuclear Regulatory Commission, Washington, D.C.
18
_ _ _ _ _ _ _ _ _.