ML100380002

Request for Additional Information Spent Fuel Pool Rerack Amendment

ML100380002
Person / Time
Site:	Beaver Valley
Issue date:	02/16/2010
From:	Nadiyah Morgan Plant Licensing Branch 1
To:	Harden P FirstEnergy Nuclear Operating Co
Morgan N, NRR/DORL, 415-1016
References
TAC ME1079
Download: ML100380002 (15)

Text

UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 20555-0001 February 16, 2010 Mr. Paul A. Harden Site Vice President FirstEnergy Nuclear Operating Company Beaver Valley Power Station Mail Stop A-BV-SEB1 P.O. Box 4, Route 168 Shippingport, PA 15077

SUBJECT:

BEAVER VALLEY POWER STATION, UNIT NO.2 - REQUEST FOR ADDITIONAL INFORMATION RE: SPENT FUEL POOL RERACK LICENSE AMENDMENT (TAC NO. ME1079)

Dear Mr. Harden:

By letter dated April 9, 2009, as supplemented by letters dated June 15, 2009, and January 18, 2010, FirstEnergy Nuclear Operating Company (licensee) submitted a license amendment for Beaver Valley Power Station, Unit NO.2 (BVPS-2). The proposed amendment would modify Technical Specifications (TSs) to support the installation of high density fuel storage racks in the BVPS-2 spent fuel pool.

The Nuclear Regulatory Commission (NRC) staff is reviewing the submittals and has determined that additional information is needed to complete its review. The specific questions are found in the enclosed request for additional information (RAI). The NRC staff is requesting a response to the RAI within 30 days of receipt.

The NRC staff considers that timely responses to RAls help ensure sufficient time is available for NRC staff review and contribute toward the NRC's goal of efficient and effective use of staff resources.

P. Harden

- 2 If you have any questions regarding this issue, please contact me at (301) 415-1016.

Sincerely, t/~-.

Nadiyah S. Morgan, Project Manager Plant Licensing Branch 1-1 Division of Operating Reactor Licensing Office of Nuclear Reactor Regulation Docket No. 50-412

Enclosure:

RAI cc w/encl: Distribution via Listserv

REQUEST FOR ADDITIONAL INFORMATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION RELATED TO THE SPENT FUEL POOL RERACK LICENSE AMENDMENT FIRSTENERGY NUCLEAR OPERATING COMPANY FIRSTENERGY NUCLEAR GENERATION CORP.

OHIO EDISON COMPANY THE TOLEDO EDISON COMPANY BEAVER VALLEY POWER STATION, UNIT NO.2 DOCKET NO. 50-412 By letter dated April 9, 2009 (Agencywide Documents Access and Management System (ADAMS) Accession No. ML091210251), as supplemented by letters dated June 15, 2009 (ADAMS Accession No. ML091680614) and January 18, 2010 (ADAMS Accession No. ML100191805), FirstEnergy Nuclear Operating Company (licensee) submitted a license amendment for Beaver Valley Power Station, Unit NO.2 (BVPS-2). The proposed amendment would modify Technical Specifications (TSs) to support the installation of high density fuel storage racks in the BVPS-2 spent fuel pool (SFP). In order to complete the review, the Nuclear Regulatory Commission (NRC) staff needs the following additional information:

Reactor Systems Branch/Oak Ridge National Lab Review Document reviewed: [Proprietary] Holtec Report No. HI*2084175, "Licensing Report for Beaver Valley Unit 2 Rerack"

1.

Does BVPS-1 and 2 share any resources related to fresh or spent fuel handling and storage? If so, it may be necessary to evaluate related normal and potential abnormal conditions.

2.

Table 2.5.1 provides some key storage rack design information, but does not provide enough information to assess the adequacy of the storage rack model. Provide the dimensions and tolerances from the bottom of the storage cell to the bottom of the storage rack absorber plates.

Enclosure

- 2

3.

In Section 3.4, on page 3-7, the first bulleted item states, "A decrease of no more than 5% in Boron-10 (B-10), as determined by neutron attenuation, is acceptable. (This is equivalent to a requirement for no loss in boron within the accuracy of the measurement.)" This statement implies that is acceptable to ignore a loss of up to 5% of the B-10 in the Metamic. It is not clear from the text provided in Section 4 that this acceptable loss of B-10 has been considered in the analysis.

Amend the analysis to consider the "acceptable" loss of B-1 0 or provide justification for not doing so.

4.

Section 4.5.3 describes the method used to calculate the axial burnup distribution used in the criticality analysis. From the text and Table 4.5.5, it is not clear whether this was done using core average axial burnup or if individual assembly burnup distributions were used. Use of core average axial burnup would minimize the importance of outliers that may have a significantly different axial burnup shape, and therefore, may not be appropriate.

Confirm that assembly-specific axial burnup distributions were utilized, or justify the use of core average axial burnup distributions.

5.

The analysis takes the lowest relative burnup at each node to create a fictitious profile for use. This makes the relative burnup for the entire fuel assembly less than 1.00. In this particular case, the total relative burnup is 0.96, which means the fuel assembly in the analysis actually has 4% less burnup than is stated. A fuel assembly with 30 GWD/MTU actually has 28.8 GWD/MTU. Attributing the k-effective (keff) of a fuel assembly with 28.8 GWD/MTU to one with 30 GWD/MTU should be conservative.

However, it is not clear that the derived profile is actually a limiting profile. It is not clear that one of the actual, or future, profiles may be more reactive, even with the slightly higher burnup. According to various tables in the analysis 1 GWD/MTU is worth roughly a half percent b.keff. According to NUREG/CR-6801 the axial profile can be worth several percent b.keff.

Demonstrate that the method of creating the profile is bounding.

6.

In Section 4.5.5, first paragraph, the last sentences states, "Neutron absorber panels are installed on all exterior walls facing other racks." This statement seems to imply that absorber panels are not installed on exterior walls that are not facing other racks.

The text in Section 4.7.13 stated that the mislocated assembly is modeled adjacent to two poisoned faces of the racks. A higher keff value may be calculated if the mislocated assembly is modeled next to a fuel assembly in a position that does not have an external as poison panel. Confirm that all external faces have poison panels or justify the model used for the mislocated assembly analysis conservatively bounding the other possible locations.

7.

The text in Section 4.5.6 stated that racks are separated by a minimum 1.5-inch gap, but does not clearly define this dimension. This might be from the outside of the sheathing that holds the Metamic in place or it might be from the outside of each stainless steel

- 3 box. Provide a clear definition of the1.5-inch gap between racks and confirm that this definition is used in the calculations.

8.

In Section 4.5.7, insufficient information is provided to justify the model used for the fuel rod storage basket (FRSB). Provide additional description of the FRSB and explain why the FRSB model used is conservative.

9.

While it is unlikely to prove limiting, a case with all assemblies moved away from the storage rack center should have been evaluated in Section 4.7.6. This would move the highest reactivity fuel (5 wt% fresh fuel) in four adjacent rack corners toward each other.

Expand the analysis to include this additional case or justify not doing so.

10.

Section 4.7.13.4 covers abnormal location of a fuel assembly. During the time when both old and new racks exist in the pool at the same time, there is the potential for the erroneous storage of a fuel assembly in the new racks that was intended for the old racks and for the erroneous storage of a fuel assembly in the old racks that was intended for the new racks.

Evaluate or provide justification for not including evaluation of potential abnormal conditions related to storage of fuel intended for the new racks in the old racks and to storage of fuel intended for the old racks in the new racks.

11.

Section 4.7.17, "Interim Configurations," described measures taken to isolate fuel stored in old racks from fuel stored in new racks. There is some potential that this administrative control will be violated. Evaluate or provide justification for not evaluating the erroneous placement of fuel in the prohibited two empty rows between old and new racks.

12.

Table 4.5.2 describes the design basis fuel assembly. Provide the distance and tolerance from the bottom of the fuel assembly to the bottom of the active fuel. If this distance varies with fuel design, provide design-specific information.

13.

In Table 4.5.2, the value used positive tolerance on fuel rod pitch appears to be crediting that the fuel rod pitch will vary randomly across a row of 17 fuel rods. Since there are only 16 pin pitches across a fuel assembly, shouldn't the tolerance have been 0.0075 divided by 16 rather than 17?

Revise the analysis to correctly incorporate the fuel rod pitch tolerance or justify not doing so.

14.

Table 4.5.3 provides the core operating parameters used for CASMO depletion calculations. Was the analysis performed to show that these parameters are conservative and bounding? Note that consideration should be given to the full range of parameters experienced by all fuel currently stored and for fuel that will be stored in the future. Consider too that parameters that lead to spectral hardening and increased plutonium production also reduce depletion of thermal neutron absorbing fission products. It should not just be assumed that anything that hardens the spectrum is conservative.

- 4 Provide the ranges of operating parameters affecting the CASMO depletion calculations and provide better justification for bounding values selected.

15.

Table 4.5.4 lists the calculated burnup-dependent soluble boron concentrations for BVPS-2 Cycle 14. It is not clear from the text why using these calculated values is appropriate. More appropriate would be the measured boron letdown curve or the predicted boron letdown curve from the nuclear design report. Beyond that issue, consideration should also be given to the boron letdown curves from previous cycles.

The bounding value used for the depletion analysis should be bounding for all prior cycles and should also bound future cycles.

Confirm that the boron letdown curve provided in Table 4.5.4 bounds past measured boron letdown curves and expected future boron letdown curves.

16.

Table 4.5.10 provides key storage cell parameters and tolerances. Provide or justify not providing information on the following rack parameters:

a. Steel density, including tolerance.

b. Dimensions and tolerances associated with "corner angles" and "filler panels" (see figure 2.6.2), and associated with "developed cell" dimensions and tolerances. Note from Figure 2.6.4 that the exterior poison panels appear to be shorter than interior poison panels.

17.

Table 4.7.2 and Figure 4.7.2 both show a minimum required burnup of 0.00 GWD/MTU for fuel initially enriched to 2.0 wt%. The equation under Table 4.7.2 yields a minimum burnup of 0.52 GWD/MTU. The inconsistency between the equation under Table 4.7.2 and the data presented in Table/Fiqure 4.7.2 should be eliminated.

Revise Table 4.7.2 and Figure 4.7.2, the equation provided under Table 4.7.2 and any associated text to eliminate inconsistency in required minimum burnup.

18.

Section 4.2.1, first paragraph - the last sentence states: "This approach has been validated in [4.3] by showing that the cross sections result in the same reactivity in both CASMO-4 and MCNP4a."

This is not "validation". This cross-code comparison utilizing the same cross sections does not tell us anything about the potential composition and cross-section errors and associated biases introduced by the modeling of lumped fission products and use of lumped fission product cross sections.

What is the worth of the lumped fission products in the fuel storage racks? What fission products are included in the lumped fission products?

The "5% of the reactivity decrement" suggested by the Kopp memo does not cover modeling simplifications and approximations such as use of lumped fission products.

- 5 Eliminate or provide better justification for the use of lumped fission products.

Justification for the use of the "lumped fission product" modeling simplification should include quantification of associated bias and bias uncertainty.

19.

Section 4.2.1, second paragraph - The uncertainty stated in this paragraph (i.e. 0.0011) is the 95/95 confidence level uncertainty on the "mean" kef( value calculated for the critical experiments modeled in the validation study. Note that this is not the 95/95 confidence on the distribution of the sample population (the kef( values calculated for the critical experiments).

It looks like the variance of the population has not been included in the analysis. This is important because the requirement is that we have a 95% confidence with a 95%

probability that any single calculation that calculates as subcritical is indeed subcritical.

Thus, the sample variance about the bias is the relevant quantity, not the variance of the bias. The 95/95 uncertainty used to determine the acceptable kef( values would rise from 0.0011 to 0.0082.

Review the statistical approach used to determine the acceptable kef( values and either clarify how the variance of the sample population was included, or revise the statistical approach used to include consideration of the sample population variance, or justify not doing so.

20.

The text in Section 4.7.4, "Isotopic Compositions," describes modeling approximations related to calculation and use of burned fuel compositions. Provide additional detail in this section to clearly state the conditions used in the calculation of fuel compositions as a function of initial-enrichment and fuel burnup.

From the text in Section 4.7.4, it appears that assembly average compositions are used rather than pin-by-pin compositions. This minimizes the impact on reactivity of removing WABA rods from an assembly. During depletion, the WABA rods depress powerlburnup locally until the WABA rods are removed. Provide additional detail describing and justifying the use of modeling approximations associated with modeling burned fuel.

Where appropriate, include biases and bias uncertainties associated with the modeling simplifications and approximations.

21.

Section 4.5.4 provided a discussion of fuel assembly reactivity control devices. Confirm that integral (e.g. IFBA) and non-integral (e.g., WABA, WDR, RCCAs, etc.) reactivity control devices are modeled during depletion and removed prior to restart in rack geometry.

22.

As described in Section 4.7 and in other places in the report, CASMO-4 was used to provide rack kef( sensitivity information. Considering that the storage rack is constructed of "fabricated" cells and "developed" cells (see Fig. 2.6.2), it is not obvious what the CASMO model looked like. Did CASMO model simplifications affect the results?

Describe the CASMO model used for fuel storage rack models. Where appropriate, include biases and uncertainties associated with CASMO model simplifications.

- 6

23.

Section 4.7.1 included a discussion justifying not modeling spacer grids. Table 4.7.5 provided some burnup-dependent results for models with and without spacer grids. The initial enrichment used was not clear from the text. Provide additional information regarding the initial enrichment(s) used for these calculations. Describe the impact of varying initial enrichments on the results.

The text in Section 4.7.1 stated, "it is conservative to neglect the fuel assembly spacer grids..." Conservative is not the right word. One might argue that the impact of the grids is negligibly small, but not conservative. The last sentence in Section 4.7.1 implies that not modeling the spacer grids yields a conservatism of approximately 0.01 L\\k. As was pointed out earlier in Section 4.7.1, the effect is overestimated in the study, and the conservatism for fresh fuel is likely significantly less than 0.01 L\\k. Revise the text in Section 4.7.1 to more accurately describe the impact of not modeling spacer grids.

24.

Section 4.7.2 addressed the reactivity effect of fuel assembly reactivity control devices.

There are three issues associated with the analysis described in this section:

a.

It does not appear that depletion with control rods present was considered.

Frequently, second-and third-cycle fuel is placed under control rods, some of which may be used to control reactor power. Thus a realistic fuel depletion scenario could include first cycle depletion with burnable absorbers and second cycle of depletion with partially inserted control rods. Some plants have also included part-length absorbers rods in some peripheral locations to reduce neutron flux to reactor vessel welds. Tin low-leakage loading patterns, these peripheral locations frequently hold fuel that is being used for a third cycle.

Provide justification for modeling reactivity control devices throughout fuel assembly life.

b.

It appears that IFBA and WABA were not modeled during depletion as being present in the same assembly. Does some mechanical feature or technical specification prevent them from being in the same assembly? Provide justification for not modeling both in the same assembly.

c.

Some plants have used standard Pyrex glass burnable absorbers. These absorbers deplete more slowly and can result in increased plutonium generation.

Confirm that standard Pyrex glass burnable absorbers have not been used at BVPS-2. If they have been used, revise the analysis to address such use.

25.

The last sentence of Section 4.7.4 is "For conservatism, the isotopic concentrations are determined at zero cooling time, without xenon and [.... J." Provide a reference or justification for the assertion that use of such isotopic concentrations is conservative.

26.

Section 4.7.5 addresses uncertainty in depletion calculations. There is no basis in the Kopp memo to suggest that coverage of changes in fuel geometry during irradiation was intended. Uncertainty in kef( due to random fuel geometry changes should be handled as variations in the parametric ranges of the analysis and included along with the other uncertainties. Variation in kef( due to anticipated geometry changes during irradiation should be handled as a bias.

- 7 Confirm that the impacts of fuel geometry changes during irradiation are properly handled.

27.

Generic issue on calculation of uncertainties using MCNP: Where the reactivity worths of an uncertainty are calculated using MCNP, the reactivity worth should also include allowance for the Monte Carlo uncertainty in the calculation of the reactivity worth.

Revise the analysis to properly incorporate keff uncertainties calculated using MCNP.

28.

Section 4.7.8 discussed temperature and water density effects. Clarify the source of the bias, the method by which the bias is calculated, and the justification for using CASMO to estimate the bias. Confirm that the calculational results shown in Table 4.7.15 reflect changes in both water density and temperature. Discuss how the temperature bias changes with the presence of soluble boron. If appropriate, include a revised temperature bias for soluble boron crediting.

29.

Section 4.7.13.4.1 described the analysis of misloaded fresh fuel assemblies. Provide the misloaded fuel assembly locations that were evaluated. If a limited set of positions were evaluated, justify use of the selected locations.

30.

Section 4.7.13.4.2, second paragraph, first sentence stated that a periodic boundary condition was used. Use of a periodic boundary condition with the model shown in Figure 4.7.3 is not appropriate. Confirm that an appropriate boundary condition was used or provide justification for use of the periodic boundary condition.

From the text in Section 4.7.13.4.2, it is not clear if any variations of the mislocated assembly model were considered. What is the impact on keff of moving the three assemblies, within their cells, closer to the mislocated assembly? What is the impact of moving the mislocated assembly a cm or two away from the storage racks? Describe what work was done to ensure that the mislocated assembly model used was the most reactive.

31.

Section 4.7.15 and others - The second paragraph of Section 4.7.15 provided the minimum soluble boron concentration required under normal conditions. A value of 472 parts per million (ppm) was obtained by linearly interpolating between 0 and 800 ppm.

The uncertainty associated with calculation of the required boron concentration is not provided and, due to the way in which it is calculated, may be significant. The relevant requirement is from 10 CFR 50.68(b)(4).

For the required soluble boron concentrations determined in the report, determine the bias and bias uncertainty associated with the method used to determine the required boron concentration. Incorporate this bias and bias uncertainty into the report.

- 8

32.

Table 4.7.7 in the rows for "manufacturing tolerances uncertainty" and "fuel tolerances uncertainty" - Each row cites footnote, which says "These tolerance uncertainties are the maximum from all burnups and enrichments." In Table 4.7.6, this was implemented by listing the largest value for all columns. For example, 0.0037 was listed for "manufacturing tolerances uncertainty" for all enrichments and 0.0074 was listed for "fuel tolerances uncertainty" for all enrichments. Table 4.7.7 shows a different value for each enrichment. Confirm that the rest of Table 4.7.7 used the correct uncertainty values.

33.

It appears that the purpose of Tables 4.7.8 and 4.7.10 is to produce an estimate for the bias introduced by not modeling the IFBA and the IFBAs with WDRs during the depletion calculations. Since the storage racks do not take credit for the presence of IFBA, it would seem that this table should show only the impact of the presence of IFBA during the depletion calculations. All of the results should have been restarts in the rack geometry with IFBA B-10 and WDRs removed. Consequently all "Ref." values and associated k; values at zero burnup should have been the same.

It also appears that the columns for IFBA with four and eight WDRs includes the IFBA B

10. This creates large negative i1k values in the last column.

So this approach uses the IFBA B-10 to reduce the i1k values, thus indirectly taking credit for IFBA B-10. This is probably not appropriate without further justification. Justify crediting the residual IFBA B-10 in the calculation of the depletion bias.

The calculations were done only with fuel having initial enrichment of 3.4 wt%.

Justify not performing the calculations for the range of initial enrichments and IFBA loadings permitted.

34.

Table 4.7.11 has a problem that is similar to that observed in Table 4.7.10. The k:

values presented from 0 to 15 GWd/MTU appear to include the worth of the WABA rods.

In Table 4.7.11, it is clearly not appropriate to include the effects of the presence of the WABA rods in the bias calculations. Table 4.7.11 should have shown the impact of depletion with WABAs present, but the k; values presented should have been from restarts in rack geometry with WABAs removed.

Revise the analysis to correctly calculate the bias associated with depleting fuel with WABA present or justify not doing so. Secondly, justify not performing the calculations for the full range of initial enrichments and WABA loadings permitted.

35.

Table 4.7.16 includes seven subsections, each of which is for one Region 3 enrichment/burnup combination. It appears that the burnup-dependent "total corrections" for Region 3 were used for every subsection of this table. For example, the first subsection was for Region 3 with 2 wt% with zero burnup and should have used a Region 3 "total corrections" value of 0.0194 for all columns in the subsection. The target keff for interpolation in the last column in the first subsection should have been calculated as 0.945 - maximum (0.0229, 0.0194) = 0.9221, interpolating to this value gives a "Soluble Boron Requirement" of 427.9 ppm. Instead, it appears that the Region 3 "total corrections" value used in the last column was 0.0238, the value for 5 wt% with 45 GWd/MTU burnup. Consequently, of the 49 "Soluble Boron Requirement" values presented in Table 4.7.16, 35 values appear to be incorrect.

- 9 Review the calculations used to prepare Table 4.7.16 and correct the table and associated text as needed. Note that some required soluble boron concentrations are repeated in other places in the report.

36.

In Tables 4.7.17 and 4.7.18, it appears that the wrong "Target keff' value was used.

According to the footnote on the bottom of the table, the largest total correction value for Region 2 and 3 at any burnup and enrichment should have been used as the "corrections". This value is 0.0238 from Table 4.7.7. So the target value should have been 0.945 - 0.0238 = 0.9212 not 0.9221. This caused each of the interpolated boron concentrations to be incorrect. Further, interpolating between 0 and 2500 ppm is probably not good. Linear interpolation over this large range may result in a significant underestimate of the soluble boron requirement.

Review the calculations used to prepare Tables 4.7.17 and 4.7.18 and correct the tables and associated text as needed. Note that some required soluble boron concentrations are repeated in other places in the report.

37.

Table 4.7.21 included the maximum "soluble boron requirement" from Table 4.7.16. In response to comment No. 35 on Table 4.7.16, this value may be revised. If it is, revise Table 4.7.21 accordingly.

38.

Either in Section 4 or in Appendix A to Section 4, the following validation issues should have been addressed:

a.

State the ranges of parameters that the safety analysis fits within. For example, the minimum and maximum values for 235U enrichment, EALF, soluble boron concentrations, Pu content and composition, etc.

b.

Describe other features present in the safety analysis models that require validation.

c.

State the ranges of parameters covered by the critical experiments used in the validation study.

d.

Discuss the applicability of the validation study to the safety analysis models.

e.

Discuss gaps within the parametric range covered by the validation and, if necessary, additional margin adopted to cover interpolation.

f.

Discuss extrapolation beyond the parameter range covered by the validation study and, if necessary, additional margin adopted to cover extrapolation.

g.

Discuss validation gaps (e.g., fission product validation) and, if appropriate, additional margin adopted to cover validation gaps.

- 10

39.

Section 4A.1, Eq. 4A.2 - As is accurately described in the text, this equation yields the standard deviation of the mean. The requirements in 10 CFR 50.68 require that we have a 95% probability with a 95% confidence level that a calculation that calculates as being subcritical actually is subcritical. The 95/95 requirement is not on the mean, but rather on a single calculation. Thus, what is needed in Section 4 is the variance of the population about the mean, not the standard deviation of the mean. The difference is an extra factor of "n" in the denominator of Eq. 4A.2. Again, it is not that Eq. 4A.2 is incorrect. Instead, the population variance about the mean needs to be calculated in Appendix A and used in Section 4.

Revise Appendix A to also provide the variance of the population about the mean.

Revise Section 4 to correctly incorporate this variance.

40.

Sections 4A.2 and 4A.3 include discussion of comparisons between MCNP and KENO results. From Section 4A.2:

"Since it is very unlikely that two independent methods of analysis would be subject to the same error, this comparison is considered confirmation of the absence of an enrichment effect (trend) in the bias."

These two methods are not independent. They both use data derived from the same cross section measurements and, in part, from the same ENDF/B-V evaluation.

Consequently, they include the same nuclear data measurement errors and evaluation errors and would be expected to respond similarly to these errors. The comparisons provided serve only to confirm that both codes respond to the same data, erroneous or not, in the same way.

The conclusions in Section 4A.2 and 4A.3 on enrichment and B-10 biases based on the KENO/MCNP comparison should not be given any credit. Revise the text in Sections 4A.2 and 4A.3 to remove use of code-to-code comparisons.

41.

Section 4A.4.1 includes the following statement:

"The tendency toward over-prediction at close spacing means that the rack calculations may be slightly more conservative than otherwise."

This also means the critical experiment calculations are too high, which is non conservative. The rack calculations are not "more conservative" because the computational bias also includes this trend. Revise the text to remove claims that the tendency toward over-prediction at close spacing is conservative.

- 11 Document reviewed: FirstEnergy Nuclear Operating Company letter L-09-162, "Beaver Valley Power Station, Unit No.2, License No. NPF-73, Additional Technical Information Pertaining to Licensing Amendment Request No.08-027 (TAC No. ME1079)," dated June 15, 2009.

The licensee's response to the second RAI question was as following:

Section 4.7.5 of HI-2084175 states that the depletion uncertainty is intended to encompass the following calculational uncertainties: lack of critical experiment data of spent fuel storage rack geometries containing both actinides and fission products, uncertainty in actual versus calculated isotopics, and changes in fuel geometry (clad creep, pellet densification, etc.) during irradiation. However, this appears inconsistent with the magnitude of the isotopic uncertainty in Appendix 6E of Holtec Report HI 951251. Provide clarification on the magnitude of these effects, such that the Nuclear Regulatory Commission (NRC) staff may evaluate whether or not 5% of the reactivity decrement associated with the burnup of interest is sufficient to encompass these effects.

The licensee's response provided does not include sufficient detail to permit NRC staff to reach a conclusion on whether or not the 5% of the fuel depletion reactivity decrement uncertainty has been used appropriately in the analysis supporting the original license amendment request.

The following additional information is requested:

42.

The analysis described in the licensee's response to the original RAI was based on "accurate measurements of critical reactor parameters" at 99 points from five cycles of BVPS-1 and 2. Provide additional details for these 99 points so that NRC staff may reach a conclusion as to the appropriateness of using this data to assess the fuel depletion uncertainty. Include as much of the following information as possible, such as:

unit, cycle number, cycle burnup, % of full power, measured and predicted soluble boron concentration, control rod bank positions, % axial offset (l.e., (PT - PB)/(PT + PB)*100%;

where PT is the power in the top half of the core and PB is the power in the bottom half of the core).

43.

The analysis relies on results from core follow calculations that were performed using CASMO-4 cross-section data. Describe the core follow calculations (e.g., codes used, input data, calibration with core measurements, etc.) and the cross-section data (e.g., 2 group macroscopic, homogenized node average, etc.) used.

44.

Two methods were used to estimate the depletion uncertainty. The first method uses the difference between the highest and lowest eigenvalues to estimate uncertainty. The second method apparently uses population variance around a linear least-squares fit of calculated eigenvalues to estimate uncertainty. Neither method appears to include consideration of systematic deviation from expected values that might indicate a bias.

The 5% of the reactivity decrement suggested by the Kopp memo was intended to cover both bias and bias uncertainty in fuel composition and keff calculations. Describe how the two methods presented include consideration of burned fuel composition and keff calculation biases.

- 12

45.

It is likely that the bulk of the 99 sets of measured critical reactor parameters are at hot full power (HFP) conditions. At HFP, the eigenvalue is most sensitive to the fuel compositions away from the core periphery. For highly burned fuel in storage racks, the eigenvalue is most sensitive to the fuel compositions in the top 24 inches of the fuel assembly. Due to axial flux redistribution, it is not clear how the uncertainty in kefffor a reactor at HFP is related to the uncertainty in kefffor highly-burned fuel in spent fuel storage racks. Additionally, the reactor will include a mixture fresh, once-, twice-and, sometimes thrice-burned fuel. Since fuel is generally reused until it is highly burned, the uncertainty in kefffor the highly-burned fuel in fuel storage racks could be significantly higher than the uncertainty for the mixture of fuel in the core. Provide justification for application of uncertainty in reactor core eigenvalues to highly-burned fuel in spent fuel storage racks.

IVIL100380002 OFFICE DORLILPLI-1/PM DORLILPLI-1/LA DSS/SRXB/BC DORLILPLI-1/BC NAME NMorgan SUttle GCranston NSalgado DATE 2/16/10 2/16/10 2/16/10 2/16/10