ML20137S733

Safety Evaluation Approving Use of Fuel Thermal Performance Code,Frosstey,For Analysis of LOCA Conditions at Low & Moderate Burnups

ML20137S733
Person / Time
Site:	Vermont Yankee
Issue date:	09/27/1985
From:	Office of Nuclear Reactor Regulation
To:
Shared Package
ML20137S718	List: ML20137S712 ML20137S733
References
NUDOCS 8509300488
Download: ML20137S733 (12)

Text

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

,s tes ]

UNITED STATES  ;

[ g NUCLEAR REGULATORY COMMISSION l

p W ASHINGTON, D. C. 20555

%,*...*/

! SAFETY EVALUATION BY THE OFFIEE OF NUCLEAR REACTOR REGULATION REVIEW OF FUEL THERMAL PERFORMANCE CODE VERMONT YANKEE NUCLEAR POWER CORPORATION VERMONT YANKEE NUCLEAR POWER STATION DOCKET NO. 50-271

1. INTRODUCTION By letters dated May 18, 1981 (Ref. 1) and July 27, 1981 (Ref. 2),.the Yankee Atomic Electric Company submitted two proprietary licensing reports, YAEC-1249P (Ref. 3) and YAEC-1265P (Ref. 4), describing the fuel thennal performance code, FROSSTEY, and by letter dated February 5, 1985, submitted nonproprietary versions of the above reports (identified as YAEC-1249 and YAEC-1265). Although the FROSSTEY code was submitted in support of the  ::

Vennont Yankee Cycle 9 and Cycle 10 (Refs. 5-6) reload applications, the submittal stated that the code is also intended for use in future safety analyses of the Yankee Rowe, Maine Yankee and Seabrook facilities. Our generic- review of the FROSSTEY code, described in the following sections, marks the first occasion in which a fuel performance code has been developed and submitted for NRC review without support from a major fuel vendor.

2. BACKGROUND l

The FROSSTEY (Fuel R0d 5,teady 5, tate T,hemal Effects) code is based on the GAPCON-THERMAL-2 code (Ref. 7) developed by Battelle Pacific Northwest Laboratory (PHL) fcr the Nuclear Regulatory Comission. Because our evaluation of the FROSSTEY code is based, in part, upon a technical review by the same contractor (PNL), our contractual arrangenents have included a routine conflict-of-interest detemination. Our conclusion that Battelle's

! objectivity has not been ccupromised in this review is based on a number of considerations, including: (1) NRC's GAPCON-THERMAL-2 code was placed in the public domain through the National Energy Softare Center; (2) Yankee Atmic's currently approved fuel perfomance code, GAPEXX (Ref. 8), is based on a cannon predecessor to the NRC code called GAPCON-THERMAL (Ref. 9);

(3) differences between FROSSTEY and GAPCON-THERMAL-2, as described in' 3

' Appendix D of YAEC-1249P. are significant; and (4) where similar models (e.g.,

fuel thermal conductivity) are used in both FROSSTEY and GAPCON-THERMAL-2, these models are the same for virtually all fuel perfomance codes used in the industry. 8509300400 e50927 DR ADOCK O 21  !

Input to this' review from Battelle has consisted of (1) PNL interrogatories (Ref.10) included in staff questions (Ref.11) to the licensee; (2) a pre-liminary technical evaluation (Ref.12) based only on the subject topical reports (Refs. 3-4); and (3) a final technical evaluation (Ref.13) based on the topical reports as well as licensee's responses (Refs.14-15) to staff @ estions.

t

3. PREVIOUS APPLICATIONS OF FROSSTEY As stated in the introduction, the FROSSTEY code was previously applied in the safety analysis of Vemont Yankee Cycle 9 and 10 reloads (Refs. 5-6).

In those applications, the code was used to calculate (a) incipient fuel centerline melt limits, (b) I percent cladding strain limits. (c) core- -

. average gap conductance and core-average fuel temperature for initializing non-LOCA (loss-of-coolant accident) transient analyses, and (d) average gap conductance and average fuel temperature of the peak bundle for initializing Previous analyses provided by General Electric hot channel calculations.

were used for the renainder of the fuel thermal and mechanical analysis, including LOCA initial conditions.

The power level at which incipient fuel centerline melting and one percent cladding strain occur decreases with increasing burnup and gadolinia However, the power level at which these limits are reached f concentration.

l (approximately 20 kW/ft) is less limiting (i.e., higher) than other fuel For this reason, and design limits (e.g., the 13.4 kW/ft PLHGR limit).

because of the similarity between the licensee's results and those previously calculated by General Electric, these two applications of FROSSTEY (fuel centerline melt and one percent cladding strain limits) were previously found acceptable (Ref.16)

Core-average gap conductance and core-average fuel temperatures used for the

' Cycle 9 and 10 transient analyses were also calculated by the licensee in a l

conservative manner (in this c'ase, by using low values of gap conductance and

t i high average fuel taperatures). We recognize that the averaging technique (for core-average conditions) tends to eliminate many local uncertainties such as variations in the extent of fuel densification. In general, these l

limiting, core-average conditions tend to occur early-in-life as a result of l l

fuel densification effects. Peak bundle (hot channel) gap conductance, on the other hand, is conservative if a high value is used. This condition tends to occur late-in-life because of cladding creepdown and fuel swelling effects. As discussed in our Cycle 9 evaluation, the use of FROSSTEY to detemine both core-average and hot channel conditions was accepted on the  ;

basis of comparisons between FROSSTEY and the NRC audit code GAPCON-THERMAL-2.

It should also be noted that the hot channel calculations were accepted, even though they rely on analysis at higher burnups, because the licensee relied on 1

- a bounding value of hot channel gap conductance (based on previous GE analyses) :

and utilized FROSSTEY calculations only to confirm the applicability of the i bounding value.

The FROSSTEY code is also used in other areas of plant safety analysis such as establishing the Doppler coefficient contribution to the overall power defect. For steady-state boiling water reactor (BWR - e.g., Vemont Yankee) analysis, a large error in fuel.tenperature may lead to a large error in the Doppler reactivity. However, the Doppler coefficient contribution is not critical. The LOCA analysis, on the other hand, is used to detemine reactor operating limits. The use of FROSSTEY in LOCA analysis is discussed in Section 4.2. 1 i

4. EVALUATION OF FROSSTEY_

As discussed in Reference 13, we have concluded that the FROSSTEY code is l acceptable for providing best-estimate fuel tenperature, gap conductance, fission gas release and rod internal pressure predictions at low-to-

- intemediate burnups. Due to the lack of high burnup data in the licensee's submittal and the underprediction of additional data supplied by the staff,  !

we are unable to conclude that the code predictions are adequate for the f analysis of conditions at higher burnup.

l \

l ,

In addition, we have been unable to detemine if the code is appropriate for providing initial conditions (even at low burnup) for the, analysis of the loss-of-coolant accident. Because the loss-of-coolant accident analysis is used to detemine plant operating conditions (e.g., MAPLHGR limits), we have traditionally required sufficient conservatism in the fuel stored energy calculation (a FR0SSTEY output parameter) to bound the inherent uncertainties in the code prediction. A discussion of each of these two restrictions (high burnap and LOCA analysis) is provided in the following sections.

4.1 HIGH BURNUP ANALYSIS The high temperature fission gas release model in FROSSTEY is based on tne Beyer-Hann model (Ref.17) with a correction for burnup enhanced release.

~

Of those data used to derive the Beyer-Hann model, only a few were taken at burnups above 10 GWd/MtU and none were taken at burnups above 20 GWd/Mtu.

Recognizing the importance of fission gas release at higher burnups and the need for accurate fuel perfomance predictions at burnups of up to 50 GWd/MtU, we requested that the licensee confirm FROSSTEY's predictive ability at high burnup with additional data from Riso (Refs.18-19) and Zorita (Ref. 20) reactors. As discussed in the licensee's responses-(Ref.14), all of these additional data were underpredicted. Thus, we have been unable to conclude that the FROSSTEY code is adequate at high burnup and, at this time, cannot accept the results of this code as a basis for detennining conditions which are limiting at end-of-life (e.g., fuel rod internal pressure limits). This restriction does not apply to types of analysis previously reviewed and approved for Vemont Yankee (see Section 3 above).

Because the licensee has not yet utilized th'e FROSSTEY code to detemine end-of-life rod pressures or other limiting conditions at higher burnups, tnis restriction is (for the moment) moot. In the interest of eliminating the restriction, it should be pointed out that the issue of high burnup

~ fission gas release was resolved in our review of another recent fuel

~

perfomance code (Ref. 21) on'the basis of additional high burnup data (Ref.

22). In that case, the applicant denonstrated that the submitted code was

l

, (on the average) sufficiently close to a best-estimate predictor of the high burnup data. Vemont Yankee may wish to pursue a similar course of action or continue to rely on the fuel vendor for these calculations.

4.2 LOCA INITIAL CONDITIONS As stated above, we have concluded that the FROSSTEY code is essentially a best-estimate predictor of fuel thermal conditions at low and moderate bu rnups. Required and acceptable features of the code when used in the analysis of the loss-of-coolant accident are described in 10 CFR 50, Appendix K.

1 The steady-state temperature distribution and stored energy in the fuel

~

I before the hypothetical accident shall be calculated for the burn-up that yields the highest calculated cladding temperature (or, optionally, i

+he highest calculated stored energy). To accomplish this, the themal conductivity of the U02 shall be evaluuted as a function of burn-up and tanperature, taking into consideration differences in initial l density, and the thermal conductance of the gap between the U02 ""d *D' l cladding shall be evaluated as a function of the burn-up, taking into consideration fuel densification and expansion, the composition and pressure of the gases within the fuel rod, the initial cold gap dimension with its tolerances, and cladding creep.-(Ref. 23)

Although the regulations specify tolerances on initial gap size, fuel stored energy codes have traditionally shown a level of conservatism based upon other uncertainties as well (e.g., the fuel.densification assumptions in Regulatory Guide 1.126 - Ref. 24). Our reviews of other fuel performance codes (Ref. 25) have assessed this level of conservatism on the basis of comparison with the NRC audit code (with conservative options) and on the basis of statistical

< examination of predicted and measured fuel centerline temperature data.

l

1 In response to our request for infomation on FROSSTEY code conservatism (Question 6 of Reference 11), the licensee has stated that " application methodologies are devised such .that variable input parameters to the code (i.e., power levels,. distributions, and exposures) are chosen to provide a conservative output value" (Ref.14)._ Although conservatisms introduced through the choice of power level, power distribution snd exposure fom an important part of the overall margin in LOCA analysis, these parameters have not, in the past, been specifically used to account for uncertainties in the stored energy calculation. Based on infomation submitted by the licensee, we have been unable to detemine the magnitude of these predictive uncertainties or how they are accounted for in the proposed LOCA analysis.

j Because the licensee has not yet utilized the FROSSTEY code to detemine LOCA

- initial conditions, this restriction is also (for the moment) moot. It should .

be pointed out that there are two significantly different methods of resolving this issue: (1) Denonstrate that the conservatisms specifically employed in the FROSSTEY analysis (above and beyond thoso associated with power level, power distribution and exposure) are suffic.ent to account for the predictive uncertainties in the code; or (2) Demonstrate that the conservatisms employed in the overall LOCA analysis are sufficient to account for the overall predictive uncertainties for the event. The second approach has recently been proposed by In either case, General Electric (Ref. 21) and has been accepted by the NRC.

it is necessary tnat the licensee quantify the predictive undertainties in the FROSSTEY code, preferably on the basis of a statistical examination of the j

experimental data already submitted.

5. THE LICENSEE'S PROFICIENCY IN USING THE CODE

! The licensee's analysis of confimatory data (Refs.18-20) supplied by the

! staff would ordinarily be sufficient to demonstrate a proficiency in using the FRDSSTEY code for plant safety analysis (Generic Letter 83-11 (Ref. 26) requires that such a detemination be made).- As pointed out in Section 4.1, however, the licensee's analytical results underpredict the measured values Rather than base a proficiency detemination exclusively for all of these data.

l I

J on these underpredictions, we note that the licensee's predictions of

! low-to-moderate burnup fuel centerline thermocouple data (Ref. 27-30)

Based upon these predictions, l

supplied by the staff are much more reasonable.

and upon the technical content of the YAEC-1249P and YAEC-1265P, we conclude 4 that the intent of Generic Letter 83-11 has' been met.

With regard to demonstrating the licensee's proficiency in using the FROSSTEY '

First, the licensee's i code, two additional factors should be mentioned.

reanalysis (Ref.15) of Rod JBYO97 from Maine Yankee Core 1 appears to be the only credible analysis of this rod fran anyone in the nuclear industry, including the Electric Power Research Institute (Ref. 31). Second, we note I that one of the authors of YAEC-1249P and YAEC-1265P (Schultz) has published

In

' a doctoral thesis on the subject of fuel perfomance analysis (Ref. 32).

j

- our' opinion, neither one of these observations are necessary to conclude that the licensee's agent (Yankee Atomic Electric Company) has demonstrated a However, depth of understanding and expertise in fuel perfonnance analysis.

they do provide substantial credibility to the FROSSTEY methodology in spite' l

of the fact that the code was developed and submitted without support from a

! major fuel vendor, t

, 6. CONCLUSIONS We have reviewed the Yankee Electric Power Company reports YAEC-1249P and i

YAEC-126SP and have concluded that the FROSSTEY methodology described therein

) is acceptable for reference in plant safety analysis. However, because of ,

l j . the lack of high burnup data in the licensee's submittal and the under-l prediction of additional data suppplied by the staff, we restrict our approval l

to the analysis of conditions which are not limiting at end-of-life. This

! restriction applies to the analysis of fuel rod internal pressure limits, but g

not to the analysis of these conditions previously approved for Vemont Yankee

. (i.e., one percent cladding strain, fuel centerline melt, and hot channel

}

conditions) for non-LOCA transient and accident analysis.

i i

I l

As discussed in-Section 4.2, we also do not accept the use of FROSSTEY to determine initial conditions (even at low burnup) for the analysis of the loss-of-coolant accident.

Principal Contributor: J. Voglewede Dated: September 27, 1985

REFERENCES 1.

D.E. Vandenburgh (YAEC) letter to T. A. Ippolito (NRC) on," Submittal of Documentation of YAEC Fuel Perfomance Code (FROSSTEY)," May 18,1981.

2.

D.E. Vandenburgh (YAEC) letter to T. A. Ippolito (NRC) on "Suhaittal of f 27, 1981.

Documentation of YAEC Fuel Perfomance Code (FF.0SSTEY)," July 3.

S.P. Schultz and K.E. St. John, " Methods for the Analysis of 0xide Fuel Rod Code /Model Description Manual,"

Steady-State Themal Effects (FROSSTEY):

Yankee Atmic Electric Company Report YAEC-1249P (Proprietary), April 1981.

4. S.P. Schultz and K.E. St. John, " Methods for the Analysis of Oxide Fuel Rod Steady-State Themal Effects (FROSSTEY): Code Qualification and Application," Yankee Atomic Electric Company Report YAEC-1265P '

(Proprietary), June 1981. f

5. A. A.F. Ansari et al ., "Vemont Yankee Cycle 9 Core Perfomance Report,"

Transmitted Yankee Atmic Electric Conpany Report YAEC-1275, August 1981.

by R.H. Groce (VYNPC) letter to the Office of Nuclear Reactor Regulation (NRC) on " Reload 8 Licensing Submittal" dated Septenber 2,1981. .

6. B.G. Baharynejad et al., "Vemont Yankee. Cycle 10 Core Performance Report,"

Yankee Atanic Electric Company Report YAEC-1342, January 1983. Transmitted by L.H. Heider (VYNPC) letter to D.B. Vassallo (NRC) on " Cycle 10 Core )

Perfomance Analysis" dated February 22, 1983.

7. C.E. Beyer et al ., "GAPCON-THERMAL-2: A Computer Program for Calculating the Themal Behavior of an Oxide Fuel Rod," Battelle Pacific Northwest Laboratory Report BWL-1898, Novenber 1975.

8. P.A. Bergeron, " Maine Yankee Fuel Themal Perfomance Model," Yankee Atomic Electric Company Report YAEC-1099P (Proprietary), February 1976.

e 4

e

9. G.R. Horn and F.E. Panisko, " User's Guide for GAPCON: A Computer Program to Predict Fuel-to-Cladding Heat Transfer Coefficients in Oxide Fuel Pins,"

Hanford Engineering Development Laboratory Report HEDL-THE 72-128, September 1972.

29, 1981.

10. C.E. Beyer (PHL) letter to J.C. Voglewede (NRC) dated December 22, 1982.

11. D.B. Vassallo (NRC) letter to R.L. Smith'(VYHPC) dated February

12. C.E. Beyer (PHL) letter to J.C. Voglewede (NRC) dated May 20, 1982 (Proprietary).

13. C.E. Beyer (PNL) letter to J.C. Voglewede (NRC) dated March 28, 1984 '

(Proprietary).

14. L.H. Heider (VYNPC) lette- FVY 83-22 to D.G. Eisenhut (NRC) on " Response to Questions on Yankee Atomic Electric Company Fuel Perfomance Code (FR0SSTEY)" dated March 17, 1983.

15. L.H. Heider (VYNPC) letter FVY 83-26 to D.G. Eisenhut (NRC) on " Additional Response to Questions on Yankee Atomic Electric Company Fuel Perfomance Code (FROSSTEY)" dated April 8,1983.

16. V.L. Rooney (NRC) letter to R.L. Smith (VYNPC) on Anendmrnt 70 to Facility License No. DPli-28 dated November 27, 1981.

17. C.E. Beyer and C.R. Hann, " Prediction of Fission Gas Release from 002 Fuel," Battelle Pacific Northwest Laboratory Report BNWL-1875, November 1974.

, . 18. P. Knudsen and C. Bagger, " Power Ramp and Fission Gas Perfonnance of Fuel Pins M20-18, M2-2B and 19-3B," Riso National Laboratory (Dersnark) Report RISO-M 2151 December 1978.

t  !

./

l

19. C. Bagger, H. Carlsen and P. Knudsen, " Details of Design, Irradiation and

' Fission Gas Release for the Danish UO2-Zr Irradiation Test 022," Riso l National Laboratory (Denmark) Report RISO-M-2152, December 1978.

20. M.G. Belfour et al., "Zorita Research and Development Program: Volume' l -

Final Report [and] Volume 2 - Data Summary," Westinghouse Electric

Corporation Report WCAP-10180 (Non-Proprietary) September 1982.

L

21. C. Thomas (NRC) letter to J. Quirk, " Acceptance for Referencing of Licensing Topical Report NEDE-23785-1(P) entitled 'The GESTR-LOCA and SAFER Models for Evaluation of the Loss-of-Coolant Accident, Volume 1, i

GESTR-LOCA: A Model for the Prediction of Fuel Rod Thermal Performance,'"

dated November 2,1983.

22. M.G. Belfour, W.C. Chubb and R.F. Boyle "BR-3_High Burnup Fuel Rod Hot l

Cell Program Volume 1 - Final Report [and] Volume 2 - Data Sissnary,"

I Westinghouse Electric Corporation Report WCAP-10238 (Non-Proprietary). .

Novenber 1982.

23. Code of Federal Regulations, Title 10 - Energy, Part 50 - Domestic j Licensing of Production and Utilization Facilities, Appendix K - ECCS I Evaluation Models, Office of the Federal Register, National Archives and Records Service, General Services Administration, January 1,1984.

i l 24. ~ "An Acceptable Nodel and' Related Statistical Methods for the Analysis of.

I Fuel Densification," U.S. Nuclear Regulatory Commission Regulatory Guide

! 1.126. Revision 1, Mr.rch 1978.

l

25. D.B. Vassallo (NRC) letter to J.B. Sinclair (VYNPC) on " Previous Staff Safety Evaluations on Fuel Performance Codes" dated ,

April 12,198.

4 l

26. D.G. Eisenhut'(NRC), Generic Letter 83-11 to All Operating Reactor Licensees on " Licensee Qualification for Performing Safety Analyses in Support of Licensing Actions" dated September 26, 1983.

27. C.R. Hann et al., " Test Design, Precharacterization and Fuel Assembly Fabrication for Instrumented Fuel Assemblies IFA-431 and. IFA-432," Battelle Pacific Northwest Laboratory Report NUREG/CR-0332 (PNL-1988), November 1977.  !

28. C.R. Hann et al ., " Data Report for the NRC/PNL Halden Assembly IFA-432,"

j Battelle Pacific Northwest Laboratory Report NUREG/CR-0560 (PNL-2673),

April 1978.

E.R. Bradley et al., " Data Report for the NRC/PNL Halden Assembly IFA-432:

29.

April 1978 - my 1980," Battelle Pacific Northwest Laboratory Report NUREG/CR-1950 (PNL-3709), April 1981.  !

4 ~

30. E.R. Bradley et al., " Final Data Report for the Instrumented Fuel Assmbly (IFA)-432," Battelle Pacific Northwest Laboratory Report NUREG/CR-2567 (PNL-4240), June 1982.

l

31. " Light Water Reactor Fuel Rod Modeling Code Evaluation," Electric Power Resean:h Institute Report EPRI NP-369, prepared by Combustion Engineering, Inc., 1977.

32. S.P. Schultz, "The Utilization of Fuel Integrity Prediction in Light Water .

Reactor Fuel 'Managenent," ScD Thesis, Massachusetts Institute of f Technology,1977.

t

[

i 1

l I

I l

r

- - __ , - ___ ._