NRC 2006-0090, License Amendment Request 251; Technical Specification 5.6.5, Reactor Coolant System Pressure and Temperature Limits Report

License Amendment Request 251; Technical Specification 5.6.5, Reactor Coolant System Pressure and Temperature Limits Report

ML063490041
Person / Time
Site:	Point Beach
Issue date:	12/14/2006
From:	Koehl D Nuclear Management Co
To:	Document Control Desk, Office of Nuclear Reactor Regulation
References
NRC 2006-0090
Download: ML063490041 (86)

Text

NMC>

Committed to Nuclear Excellence Point Beach Nuclear P l a a Operated by Nuclear Management Company, LLC December 14,2006 NRC 2006-0090 10 CFR 50.90 U.S. Nuclear Regulatory Commission ATTN: Document Control Desk Washington, DC 20555-0001 Point Beach Nuclear Plant, Units 1 and 2 Dockets 50-266 and 50-301 License Nos. DPR-24 and DPR-27 License Amendment Request 251; Technical Specification 5.6.5. Reactor Coolant System Pressure and Temperature Limits Report Pursuant to 10 CFR 50.90, Nuclear Management Company, LLC (NMC), hereby submits a proposed amendment to the Technical Specifications (TS) for Point Beach Nuclear Plant (PBNP), Units Iand 2.

The proposed amendment would revise TS 5.6.5, "Reactor Coolant System (RCS)

Pressure and Temperature Limits Report (PTLR)". The revision would add the FERRET Code as an approved methodology for determining RCS pressure and temperature limits. The NRC approved the existing PTLR methodology for PBNP on July 23,2001.

The FERRET Code was evaluated in Westinghouse Report WCAP-16083-NP, Revision 0, "Benchmark Testing of the FERRET Code for Least Squares Evaluation of Light Water Reactor Dosimetry". WCAP-16083 was approved by the NRC for referencing in plant-specific license amendments in NRC safety evaluation dated January 10,2006.

Enclosure 1 provides a description and analysis of the proposed change. Enclosure 2 provides the TS pages marked up to show the proposed change. Enclosure 3 provides revised (clean) TS pages. Enclosure 4 submits WCAP-16083-NP-A, Revision 0, dated May 2006 (Non-Proprietary).

NMC requests approval of the proposed license amendment by November 2007, with the amendment being implemented within 45 days.

This letter contains no new commitments or revisions to existing commitments.

In accordance with 10 CFR 50.91, a copy of this application, with attachments, is being provided to the designated Wisconsin Official.

6590 Nuclear Road Two Rivers, Wisconsin 54241 Telephone: 920.755.2321

,:,,;; ;-..( i= ,.+

I..

8 8 ' 8 , 8 ,

8 8 8 'L.  !"

8 .

Document Control Desk Page 2 I declare under penalty of perjury that the foregoing is true and correct.

Executed on December 14,3006.

Dennis L. Koehl

/

Site Vice-President, Point Beach Nuclear Plant Nuclear Management Company, LLC Enclosures cc: Regional Administrator, Region III, USNRC Project Manager, Point Beach Nuclear Plant, USNRC Resident Inspector, Point Beach Nuclear Plant, USNRC PSCW

ENCLOSURE 1 DESCRIPTION AND ANALYSIS OF CHANGE LICENSE AMENDMENT REQUEST 251 TECHNICAL SPECIFICATION 5.6.5 REACTOR COOLANT SYSTEM PRESSURE AND TEMPERATURE LIMITS REPORT POINT BEACH NUCLEAR PLANT, UNITS 1 AND 2

1.0 DESCRIPTION

This License Amendment Request (LAR) is made pursuant to 10 CFR 50.90 to revise Technical Specification (TS) 5.6.5, Reactor Coolant System (RCS)

Pressure and Temperature Limits Report (PTLR). The revision would add the FERRET Code as an approved methodology for determining RCS pressure and temperature limits. The NRC approved the existing PTLR methodology for Point Beach Nuclear Plant (PBNP) on July 23, 2001.

The FERRET Code was evaluated in Westinghouse Report WCAP-16083-NP, Revision 0, "Benchmark Testing of the FERRET Code for Least Squares Evaluation of Light Water Reactor Dosimetry". WCAP-16083 was approved by the NRC for referencing in plant-specific license amendments in NRC safety evaluation dated January 10, 2006.

2.0 PROPOSED CHANGE

The proposed amendment would support a change to the PBNP Final Safety Analysis Report (FSAR) by incorporating the FERRET Code as an analytical method that may be used to determine RCS pressure and temperature limits at PBNP. Nuclear Management Company, LLC (NMC), has concluded that the analysis contained in WCAP-16083-NP is applicable to PBNP.

The proposed amendment would also revise TS 5.6.5 to add the NRC safety evaluation (SE) approving use of the FERRET Code at PBNP, to the list of approved analytical methods contained in TS 5.6.5.

TS 5.6.5 is proposed for modification as follows (additions are double-underlined).

b. The analytical methods used to determine the RCS pressure and temperature limits shall be those previously reviewed and approved by the NRC, specifically those described in the NRC Letters dated October 6, 2000, and July 23, 2001, and [NRC SE date].

Page 1 of 7

3.0 BACKGROUND

FERRET is a least squares adjustment (LSA) code using calculated spectra weighting to minimize calculated value uncertainties. The least squares adjustment method uses neutron spectra adjustment, dosimeter spectral coverage, transport calculation uncertainties, measured reaction rates, and dosimeter cross sections and their uncertainties. This approach is endorsed by and is summarized in American Society of Testing and Materials (ASTM)

Standard E 944-02. In addition to the spectra, it also uses measured reaction rates and dosimetry cross sections and associated uncertainties. The results of the FERRET adjustment have been benchmarked by comparison to measurements in NIST-calibrated fission sources, the PCA simulated benchmark experiment, and the H. B. Robinson vessel dosimetry benchmark experiment.

The transport calculation and the dosimetry cross sections adhere to the guidance in RG 1.190, Calculational and Dosimetry Methods for Determining Pressure Vessel Neutron Fluence.

For the reasons stated above, the NRC staff determined that the FERRET code is acceptable to be referenced in operating plant licensing actions subject to the following limitation.

LSA is acceptable if the adjustments to the M/C ratios and to the calculated spectra values are within the assigned uncertainties of the calculated spectra, the dosimetry measured reaction rates, and the dosimetry reaction cross sections. Should this not be the case, the user should re-examine both measured and calculated values for possible errors. If errors cannot be found, the particular values causing the inconsistency should be disqualified.

4.0 TECHNICAL ANALYSIS

The technical justification for use of the FERRET Code is provided in the enclosed Westinghouse technical evaluation and in the referenced NRC safety evaluation. The justification for applicability of FERRET to PBNP is provided below.

Applicability to PBNP Best estimate projections of neutron fluence (E > 1.0 MeV) have been determined based on a least squares analysis of plant specific transport calculations and neutron dosimetry irradiated both at in-vessel surveillance capsule locations and at ex-vessel positions within the reactor cavity of Point Beach Units 1 and 2. These least squares analyses used the methodologies described in WCAP-16083-NP.

Page 2 of 7

As stated previously, the NRC SE found that the FERRET code application is acceptable subject to the following limitation on the Least Squares Adjustment (LSA):

LSA is acceptable if the adjustments to the M/C ratios and to the calculated spectra values are within the assigned uncertainties of the calculated spectra, the dosimetry-measured reaction rates, and the dosimetry reaction cross-sections. Should this not be the case, the user should re-examine both measured and calculated values for possible errors. If errors cannot be found, the particular values causing the inconsistency should be disqualified.

The dosimetry database used in the FERRET analyses consisted of three in-vessel and twenty ex-vessel data sets for Point Beach Unit 1 and four in-vessel and twenty ex-vessel data sets for Point Beach Unit 2. An examination of the FERRET analyses for the forty seven data sets included in the Point Beach Units 1 and 2 dosimetry data base showed that in no case did the adjustments provide results outside of the assigned uncertainties in the calculated spectra, measured reaction rates, or dosimetry reaction cross sections.

The 2 per degree of freedom (2/DOF) tests performed with each of the 47 evaluations indicated that the measurements and calculations were, in all cases, consistent within the assigned uncertainties. In addition, the final adjustments of the neutron flux (E > 1.0 MeV), the prime parameter of interest, were all within the assigned uncertainty of the stand-alone transport calculation.

For the flux (E > 1.0 MeV), 79% of the adjustments were within one standard deviation of the calculational uncertainty, 17% of the adjustments fell between one and two standard deviations, and 4% of the adjustments fell between two and three standard deviations. The maximum adjustment for any single data set was 2.3 standard deviations.

These comparisons of the data before and after application of the FERRET adjustment procedure indicate that the limitation as specified in the FERRET SE is met by the Point Beach Units 1 and 2 dosimetry data bases. Therefore, the best estimate values provided by the least squares evaluation are applicable to the Point Beach reactors.

Results and Conclusion Based on the above justification, implementation of the proposed Technical Specification change is consistent with the analysis and demonstrates that the FERRET Code methodology is applicable to PBNP. All applicable limits of the safety analysis will continue to be maintained.

Page 3 of 7

5.0 REGULATORY ANALYSIS

5.1 No Significant Hazards Determination In accordance with the requirements of 10 CFR 50.90, Nuclear Management Company (licensee) hereby requests amendments to facility operating licenses DPR-24 and DPR-27, for Point Beach Nuclear Plant, Unit 1 and Unit 2. The purpose of the proposed amendments is to revise Technical Specification (TS) 5.6.5, Reactor Coolant System (RCS) Pressure and Temperature Limits Report (PTLR). The revision would add the FERRET Code as an approved methodology for determining RCS pressure and temperature limits.

Nuclear Management Company (NMC) has evaluated the proposed amendment in accordance with 10 CFR 50.91 against the standards in 10 CFR 50.92 and has determined that the operation of the Point Beach Nuclear Plant in accordance with the proposed amendment presents no significant hazards. The NMC evaluation against each of the criteria in 10 CFR 50.92 follows.

1. Operation of the Point Beach Nuclear Plant in accordance with the proposed amendments does not result in a significant increase in the probability or consequences of any accident previously evaluated.

The proposed change revises Technical Specification (TS) 5.6.5, Reactor Coolant System (RCS) Pressure and Temperature Limits Report (PTLR), to add the FERRET Code as an approved methodology for determining RCS pressure and temperature limits.

The proposed change does not adversely affect accident initiators or precursors nor alter the design assumptions, conditions, or the manner in which the plant is operated and maintained. The proposed change does not alter or prevent the ability of structures, systems, and components from performing their intended function to mitigate the consequences of an initiating event within the assumed acceptance limits. Therefore, the proposed change does not significantly increase the probability of any accident previously evaluated.

There will be no change to normal plant operating parameters, engineered safety feature actuation setpoints, accident mitigation capabilities, or accident analysis assumptions or inputs. The proposed change does not affect the source term, containment isolation, or radiological release assumptions used in evaluating the radiological consequences of an accident previously evaluated. Further, the proposed change does not increase the types or amounts of radioactive effluent that may be released offsite, nor significantly increase individual or cumulative occupational/public radiation exposures.

Page 4 of 7

Therefore, the probability or consequences of any accident previously evaluated will not be significantly increased as a result of the proposed change.

2. Operation of the Point Beach Nuclear Plant in accordance with the proposed amendments does not result in a new or different kind of accident from any accident previously evaluated.

The proposed change incorporates the FERRET Code as an approved methodology for determining RCS pressure and temperature limits. The change does not impose any new or different requirements or eliminate any existing requirements. The proposed change is consistent with the safety analysis assumptions and current plant operating practice.

No new accident scenarios, transient precursors, failure mechanisms, or limiting single failures are introduced as a result of the proposed change.

Equipment important to safety will continue to operate as designed. The change does not result in any event previously deemed incredible being made credible. The change does not result in adverse conditions or result in any increase in the challenges to safety systems. Therefore, the proposed change does not create the possibility of a new or different type of accident from any accident previously evaluated.

3. Operation of the Point Beach Nuclear Plant in accordance with the proposed amendments does not result in a significant reduction in a margin of safety.

The proposed change incorporates the FERRET Code as an approved methodology for determining RCS pressure and temperature limits. The proposed change does not alter safety limits, limiting safety system settings, or limiting conditions for operation. The setpoints at which protective actions are initiated are not altered by the proposed change.

There are no new or significant changes to the initial conditions contributing to accident severity or consequences. The proposed amendment will not otherwise affect the plant protective boundaries, will not cause a release of fission products to the public, nor will it degrade the performance of any other structures, systems or components (SSCs) important to safety. Therefore, the requested change will not result in a significant reduction in the margin of safety.

Page 5 of 7

Conclusion Operation of the Point Beach Nuclear Plant in accordance with the proposed amendment will not result in a significant increase in the probability or consequences of any accident previously analyzed; will not result in a new or different kind of accident from any accident previously analyzed; and, does not result in a significant reduction in any margin of safety. Therefore, operation of the Point Beach Nuclear Plant in accordance with the proposed amendment does not result in a significant hazards determination.

5.2 Applicable Regulatory Requirements Point Beach was licensed prior to the 1971 publication of Appendix A, General Design Criteria for Nuclear Power Plants, (GDC) to 10 CFR Part 50. As such, Point Beach is not licensed to the Appendix A GDC. The Point Beach Final Safety Analysis Report (FSAR), Section 1.3, lists the plant-specific GDC to which the plant was licensed. The Point Beach GDC are similar in content to the draft GDC proposed for public comment in 1967. The Point Beach GDC addressing the reactor coolant pressure boundary are Point Beach GDC- 9, Reactor Coolant Pressure Boundary; GDC-33, Reactor Coolant Pressure Boundary Capability; GDC-34, Reactor Coolant Pressure Boundary Rapid Propagation Failure Prevention; and GDC-36, Reactor Coolant Pressure Boundary Surveillance. The applicable criteria for this system are discussed in FSAR Section 4.1, Reactor Coolant System - Design Basis.

Point Beach GDC-9, 33, 34 and 36 require, in part, that the reactor coolant pressure boundary be designed, fabricated, and constructed so as to have an exceedingly low probability of gross rupture or significant uncontrolled leakage throughout its design lifetime; be capable of accommodating without rupture the static and dynamic loads imposed on any boundary component as a result of an inadvertent and sudden release of energy to the coolant; be designed and operated to reduce to an acceptable level the probability of rapidly propagating type failures; and have provisions for inspection, testing, and surveillance of critical areas by appropriate means to assess the structural and leak tight integrity of the boundary components during their service lifetime.

Regulatory Guide RG 1.190, "Calculational and Dosimetry Methods for Determining Pressure Vessel Neutron Fluence," March 2001, which is based on Appendix A GDC 14, 30, and 31, describes the attributes of neutron transport methodologies that are acceptable to the NRC staff. RG 1.190 specifies that the neutron transport methods should be benchmarked to a statistically significant data base of measurement-to-calculation ratios (M/C) and that existing bias and uncertainties be estimated. In addition, the RG allows the use of suitably weighted averages of the M/C values.

Page 6 of 7

NMC concludes that the proposed changes to TS 5.6.5 will continue to assure that the design requirements of the reactor coolant pressure boundary are met.

The proposed changes continue to assure compliance with the requirements of the above criteria and adhere to the guidance in RG 1.190.

6.0 ENVIRONMENTAL CONSIDERATION

NMC has determined that the information for the proposed amendment does not involve a significant hazards consideration, authorize a significant change in the types or total amounts of effluent release, or result in any significant increase in individual or cumulative occupational radiation exposure. Therefore, NMC concludes that the proposed amendment meets the categorical exclusion requirements of 10 CFR 51.22(c)(9). Therefore, pursuant to 10 CFR 51.22(b), no environmental impact statement or environmental assessment need be prepared in connection with the proposed amendment.

7.0 PRECEDENT The FERRET Code was evaluated in Westinghouse Report WCAP-16083-NP, Revision 0, "Benchmark Testing of the FERRET Code for Least Squares Evaluation of Light Water Reactor Dosimetry". WCAP-16083 was approved by the NRC for referencing in plant-specific license amendments in NRC safety evaluation dated January 10, 2006.

8.0 REFERENCES

1. Westinghouse Report WCAP-16083-NP-A, Revision 0, "Benchmark Testing of the FERRET Code for Least Squares Evaluation of Light Water Reactor Dosimetry"

2. NRC letter, Final Safety Evaluation for Westinghouse Owners Group Topical Report WCAP-16083-NP, Revision 0, Benchmark Testing of the FERRET Code for Least Squares Evaluation of Light Water Reactor Dosimetry (TAC No. MC3974), dated January 10, 2006 Page 7 of 7

ENCLOSURE 2 PROPOSED (MARKED-UP) TECHNICAL SPECIFICATION CHANGES LICENSE AMENDMENT REQUEST 251 TECHNICAL SPECIFICATION 5.6.5 REACTOR COOLANT SYSTEM PRESSURE AND TEMPERATURE LIMITS REPORT POINT BEACH NUCLEAR PLANT, UNITS 1 AND 2 (1 page follows)

Reporting Requirements 5.6 5.6 Reporting Requirements

5.6.5 Reactor Coolant System (RCS) PRESSURE AND TEMPERATURE LIMITS REPORT (PTLR)

a. RCS pressure and temperature limits for heat up, cooldown, low temperature operation, criticality, hydrostatic testing, LTOP enabling, and PORV lift settings as well as heatup and cooldown rates shall be established and documented in the PTLR for the following:

(1) LCO 3.4.3, RCS Pressure and Temperature (P/T) Limits (2) LCO 3.4.6, RCS Loops-MODE 4 (3) LCO 3.4.7, RCS Loops-MODE 5, Loops Filled (4) LCO 3.4.10, Pressurizer Safety Valves (5) LCO 3.4.12, Low Temperature Overpressure Protection (LTOP)

b. The analytical methods used to determine the RCS pressure and temperature limits shall be those previously reviewed and approved by the NRC, specifically those described in the NRC Letters dated October 6, 2000, and July 23, 2001, and [NRC SE date].

c. The PTLR shall be provided to the NRC upon issuance for each reactor vessel fluence period and for any revision or supplement thereto.

5.6.6 PAM Report When a report is required by Condition B or F of LCO 3.3.3, "Post Accident Monitoring (PAM) Instrumentation," a report shall be submitted within the following 14 days. The report shall outline the preplanned alternate method of monitoring, the cause of the inoperability, and the plans and schedule for restoring the instrumentation channels of the Function to OPERABLE status.

5.6.7 Tendon Surveillance Report Abnormal conditions observed during testing will be evaluated to determine the effect of such conditions on containment structural integrity. This evaluation should be completed within 30 days of the identification of the condition. Any condition which is determined in this evaluation to have a significant adverse effect on containment structural integrity will be considered an abnormal degradation of the containment structure.

Any abnormal degradation of the containment structure identified during the engineering evaluation of abnormal conditions shall be reported to the

Point Beach 5.6-5 Unit 1 - Amendment No. 214 Unit 2 - Amendment No. 219

ENCLOSURE 3 REVISED (CLEAN) TECHNICAL SPECIFICATION PAGES LICENSE AMENDMENT REQUEST 251 TECHNICAL SPECIFICATION 5.6.5 REACTOR COOLANT SYSTEM PRESSURE AND TEMPERATURE LIMITS REPORT POINT BEACH NUCLEAR PLANT, UNITS 1 AND 2 (1 page follows)

Reporting Requirements 5.6 5.6 Reporting Requirements

5.6.5 Reactor Coolant System (RCS) PRESSURE AND TEMPERATURE LIMITS REPORT (PTLR)

a. RCS pressure and temperature limits for heat up, cooldown, low temperature operation, criticality, hydrostatic testing, LTOP enabling, and PORV lift settings as well as heatup and cooldown rates shall be established and documented in the PTLR for the following:

(1) LCO 3.4.3, RCS Pressure and Temperature (P/T) Limits (2) LCO 3.4.6, RCS Loops-MODE 4 (3) LCO 3.4.7, RCS Loops-MODE 5, Loops Filled (4) LCO 3.4.10, Pressurizer Safety Valves (5) LCO 3.4.12, Low Temperature Overpressure Protection (LTOP)

b. The analytical methods used to determine the RCS pressure and temperature limits shall be those previously reviewed and approved by the NRC, specifically those described in the NRC Letters dated October 6, 2000, July 23, 2001, and .

c. The PTLR shall be provided to the NRC upon issuance for each reactor vessel fluence period and for any revision or supplement thereto.

5.6.6 PAM Report When a report is required by Condition B or F of LCO 3.3.3, "Post Accident Monitoring (PAM) Instrumentation," a report shall be submitted within the following 14 days. The report shall outline the preplanned alternate method of monitoring, the cause of the inoperability, and the plans and schedule for restoring the instrumentation channels of the Function to OPERABLE status.

5.6.7 Tendon Surveillance Report Abnormal conditions observed during testing will be evaluated to determine the effect of such conditions on containment structural integrity. This evaluation should be completed within 30 days of the identification of the condition. Any condition which is determined in this evaluation to have a significant adverse effect on containment structural integrity will be considered an abnormal degradation of the containment structure.

Any abnormal degradation of the containment structure identified during the engineering evaluation of abnormal conditions shall be reported to the

Point Beach 5.6-5 Unit 1 - Amendment No.

Unit 2 - Amendment No.

ENCLOSURE 4 WESTINGHOUSE REPORT, WCAP-16083-NP-A, REVISION 0, BENCHMARK TESTING OF THE FERRET CODE FOR LEAST SQUARES EVALUATION OF LIGHT WATER REACTOR DOSIMETRY, REVISION 0, DATED MAY 2006 (NON-PROPRIETARY)

(72 pages follow)

Westinghouse Non-Proprietary Class 3 WCAP-16083-NP-A May 2006 Revision 0 Benchmark Testing of the FERRET Code for Least Squares Evaluation of Light Water Reactor Dosimetry

NRC SAFETY EVALUATION UNrrEDSTATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C.

I January 10, 2006 I ECEIUf-:

Mr. Gordon Bischoff, Manager Owners Group Program Management Office JAN 13m Westinghouse Electric Company k O G PROJECT OFFICE P.O. Box 355 Pittsburgh, PA 15230-0355 u 8

SUBJECT:

FINAL SAFETY EVALUATION FOR WESTINGHOUSE OWNERS GROUP TOPICAL REPORT WCAP-16083-NP, REVISION 0, 'BENCHMARK TESTING OF THE FERRET CODE FOR LEAST SQUARES EVALUATION OF UGKF WATER REACTOR DOSIMETRY* (TAC NO. MC3974)

Dear Mr. Biichoff:

By letter dated July 30,2004, as supplemented by letter dated March 30,2005, the Westintthouse Owners G r o u ~ WOG) submitted To~icalRe~ortUR) WCAP-16083-NP.

~evisi&0, 'Benchmark ~e.&g of the FERRET Cdde for Gast &Ares Evaluation of Light Water Reactor Dosimetry,' to the U.S. Nuclear Regulatory Commission (NRC) staff for review.

On October 21.2005. an NRC draft safety evaluation (SE) regarding our approval of TR WCAP-16083-NP. Revision 0 was provided for your review and comment. By letter dated November 17.2005, the WOG provided comments on the draft SE. The NRC staff agrees with the WOG comments and the modifications suggested by the WOG have been made to the final SE, as discussed in the attachment to the final SE enclosed with this letter.

The NRC staff has found that TR WCAP-16083-NP, Revision 0 is acceptable for referencing in licensing applications regarding light-water reactor dosimetry to the extent specified and under the limitationsdelineated in the TR and in the enclosed SE. The SE defines the basis for acceptance of the TR.

Our acceptance applies only to material provided in the subject TR. We do not intend to repeat our review of the acceptable material described in the TR. When the TR appears as a reference in license applications, our review will ensure that the materialpresented applies to the specific plant involved. License amendment requeststhat deviate from this TR will be subject to a plant-specific review in accordance with applicable review standards.

In accordance with the guidance provided on the NRC website, we request that the WOG publish an acceptedversion of this TR within three months of receipt of this letter. The accepted version shall incorpoiate this letter and the enclosed SE between the title page and the abstract. It must be well indexed such that informationis readily located. Also, it must contain historical review information, such as NRC staff questions and accepted responses.

draft SE comments. and oriainal TR Dacles that were reDlaced. The acceDted version shall include a *- Ag (designating acceptd) fGllowing the TR Mentffiiationsymdol.

G. Bischoff If future changes to the NRC's regulatory requirements affect the acceptability of this TR, the WOG andlor licensees referencingit will be expected to revise the TR appropriately, or justify its continued applicability for subsequent referencing.

Sincerely, Richard P. Correia, Acting Deputy Director Division of Policy and Rulemaking Office of Nuclear Reactor Regulation Project No. 694

Enclosure:

Final Safety Evaluation cc wlencl:

Mr. James A. Gresham, Manager Regulatory Compliance and Plant Licensing Westinghouse Electric Company P.O. Box 355 Piisburgh, PA 15230-0355

UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, DE. 205550001 SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION TOPICAL REPORT WCAP-16083-NP. REVISION 0. 'BENCHMARK TESTING OF THE FERRET CODE FOR LEAST SQUARES EVALUATION OF LIGHT WATER REACTOR DOSIMETRY' WESTINGHOUSE OWNERS GROUP PROJECT NO. 694

1.0 INTRODUCTION

By letter dated July 30,2004, as supplemented by letter dated March 30,2005 (References 1 and 2. A g e r n . d e DocumentsAccess and Management System Accession Nos.

ML042160524 and ML050910119. respectively). the Westinghouse Owners Group (WOG) submitted Topical Report WCAP-16083-NP, Revision 0. 'Benchmark Testing of the FERRET Code for Least Squares Evaluation of Light Water Reactor Dosimetry,' to the Nuclear Regulatory Commission (NRC) staff for review.

The methodologyproposedin WCAP consists of three phases: (1) collection of a data base of benchmarkedplant-specific neutron transport calculations and corresponding dosimetry measurements at in-vessel and ex-vessel locations, (2) a least squares analysis involving the calculatedand measured data, and (3) use of the results to demonstrate consistency of measured and calculated values and to validate calculatedvalues at locations on the vessel inside diameter. The least squares adjustment method uses neutron spectra adjustment, dosimeter spectral coverage, transport calculation uncertainties, measured reaction rates, and dosimeter cross sections and their uncertainties. This amroach is endorsed bv and is summarized in American Society of Testing and ater ria is (ASTM) Standard ~k44-02 (Reference 3).

The purpose of this review is to describe the code, establish whether the method adheres to the guidance in Reaulatory Guide (RG) 1.I90 (Reference 4). examine the validation of the code and evaluate acceptability of the p r o p o d method h light water reactor (LWR) licensing actions.

2.0 REGULATORY EVALUATION

The basis for this review is RG 1.190 (Reference 4) that is based on General Design Criteria 14,30, and 31, and describes the attributes of neutron transport methodologieswhich are acceptable to the NRC staff. RG 1.I90 specifies that the neutron transport methods should be benchmarkedto a statistically significant data base of measurement-to-calculationratios (WC) and that existing bias and uncertaintiesbe estimated. In addition, the RG allows the use of suitably weighted averages of the M/C values.

3.0

SUMMARY

OF THE FERRET LEAST-SQUARESADJUSTMENT METHODOLOGY 3.1 Background The proposed least squares adjustment (LSA) method combines measurement data with corresponding neutron transport calculations to establish a best estimate spectrum and an estimate of the applicable uncertainties at the locationof the measurement. The spectrum is then used to calculate best estimate values of exposure quantities, such as activation rates, fluence, and iron displacements per atom. The FERRET code, which is a least squares adjustment, has been applied successfully in many reactor vessel applications. The ASTM promulgatedthe standard E 944-02 to address the application of neutron spectrum adjustment methods to reactor surveillance dosimetry. It is assumed that neutron transport is using the discrete ordinates method as in the DORT Code (Reference 5).

3.2 Applicationof the Methodology The aeneral obiective of an LSA method is to reconcile measured and calculated reaction rates.

dosir;;etry and iransport cross sections, and calculated neutron energy spectra within their corresponding uncertainties. In general, the following expression relates reaction rate R, to neutron energy spectrum q,andto dosimeter (group) reaction cross section a,, each with a corresponding uncertainty 6:

Application of the LSA method requires the following informationfor a specific measurement:

(1) a calculated spectrum and its uncertainty, (2) dosimeter measured reaction rate and uncertainty, and (3) dosimetry reaction cross sections and their uncertainty. The plant-specific neutron transport calculationsyielding the neutron energy spectrum should follow the guidance in RG 1.190.

3.3 Neutron Transport Calculations and Uncertainty The neutron transport calculation forms the basis for a reliable LSA. The flux synthesis method is used to calculate the three-dimensional neutron flux distribution 4(r, 8, z) as follows:

where @(r,8). @(r,z) and @(r)are the azimuthal, axial and radial flux distributions, respectively.

The WOG is using the DORT (Reference 5) discrete ordinates code and the BUGLE-96 (Reference 6) cross section library. An anisotropic scattering is treated with a minimum of a P, approximation and an S, minimum angular quadrature. As stated previously, transport calculations follow the guidance in RG 1.190. P, and S, are discussed in some detail in RG 1.190.

3.4 Geometric Modeling In developing the geometrical representation of the vessel, core, and internal components the effort is to use 'as-built* dimensions where available. Water temperatures (and thus water densities) are assumed at full power. The core is represented as a mixture of fuel, cladding, water, and structural materials at temperatures representing full-power operation. The choice of mesh size in the axial, radial, and azimuthal directions are chosen to achieve convergence in the inner iterations. In general, smaller intervals are chosen in areas where large flux gradients are anticipated. Normally, quarter core or octant core symmetry is applied. The core baffle, the former plates, and the thennal shield are represented as individualcomponents.

3.5 Neutron Source The source distribution is obtained from pin-wise power distributionfrom the two outer mw fuel assemblies. The fuel isotopic composition is accounted for as a weighting factor in the power-to-neutron conversion. he (r, 6) geometry transposition to (x. y) uses anarea weighting to assign source strength to each (x, y) cell from the corresponding (r. 6) cell(s).

3.6 Validation of the Transport Calculation The WOG used the transport method described in WCAP-14040-A (Reference 7) that has been approved by the NRC staff. The validation was based on the guidance in RG 1.1 90 and included comparison to the Oak Ridge Pool Critical Assembly (PCA), the H. B. Robinson dosimetry benchmark experiment, an experimental data base consisting of a large number of surveillance capsules from a variety of operating plants, and an analytical sensitivity study addressing the major uncertainty components.

The WCAP-16083validation includes three stages: (1) methods' validation addressn ig the adequacy of the transport calculation and associated dosimetry and cross sections. (2) validation of uncertaintiesthat are methods-related, and (3) validation addressing uncertainties that are related to lack of knowledge of code input parameters. The overdl calculational uncertainty is established from the above components.

3.7 UncertaintyInput to LSA The neutron energy spectrum in each measurement location is input as an absolute value.

Spectrum uncertaintyis obtained fmm plant-specifictransport calculations also at the location of the measurement. The spectrum input uncertainties should be consistent with the benchmarking results discussed in Section 3.6. The uncertainty matrix is constructedfrom the following relationship:

where R, is the overall fractional normalizationuncertainty, % and R . are groupwise uncertainties, and Pfo is a group correlation matrix. Analytic expressions for ,P , are also pmvided. The normalization uncertainty is related to the magnitude of the spectrum, while the groupwise uncertaintiesare related to the shape of the spectrum. WCAP-16083 provides specific numericalvalues for the uncertainties.

! 3.8 Reaction Rate Measurement and Uncertainties WCAP-16083 lists the standard dosimeters used by The WOG: Cu-63(n.a)C&60.

Ti-46(n,p)Sc-46, Fe-54(n,p)Mn-54, Ni-58(n,p)Co-58, U-238(n,f) fp (Cdcovered), Np-237(n,f) fp (Cdcovered), Co-59(n,y)Co-60 (with and without Cd cover). This dosimeter set provides adequate spectra! coverage. WCAP-16083 lists the ASTM standards relevant to the recommendedpractice for the use of these monitors. The analyticalexpressionto calculate the average dosimeter activationfor a given power level from the measured activation rate is given.

The section concludes with values of specific uncertainties and their justification.

3.9 Dosimetry Cross Sections and Uncertainties The activation cross sections and the associated uncertainties are obtained from the SNLRML library (Reference 8) that is based on the ENDFBVI file.

4.0 TESTING OF THE FERRET PROCESSING PROCEDURES As noted above, FERRET combines the dosimeter reaction rate measurements with the results of the neutron transport calculations, dosimetry reaction cross sections, and neutron spectra to calculate a best estimate fast neutronflux (E > 1.0 MeV) at the location of the measurement.

The process is divided into two steps: (1) processing of the calculated spectra and dosimetry cross sections and (2) application of the FERRET algorithm. Each of the steps is individually tested as outlined in the following paragraphs.

4.1 Data Comparison in the National Institute of Standards and Technology (NIST) U-235 Fission Field The SNLRML cross sections are collapsed 53 energy groups using the calculated energy spectrum as a weighting function. The FERRET report used the data in ASTM report E261-98 (keference 9) fission spectrum averaged cross sections applicable to U-235 and 6-252 spectra. The section lists numerous other comparisons with existing data to conclude that the SNLRML library and the FERRET processing result in accurate cross section values.

4.2 Evaluation of the PCA Simulator Benchmark RG 1.190 recommends benchmarking to the resutts of the PCA (Reference 10). In the past.

PCA has been analyzed by several researchers using least squares codes. The WOG updated the existing calculations using updated cross sections. Comparisonof the measured values to the updated calculated results demonstrates good-to-excellent agreement after the adjustment.

In addition, comparisons indicate consistency of the FERRET results from other analyses' methods and for all the measured locations.

4.3 Evaluation of the H.B. Robinson Benchmark The H.B. Robinson (Reference 11) vessel dosimetry measurementswere also used in the FERREF benchmark. The transport calculationswere carried out using the BUGLE-96 library based on the ENDFIB-VIfile, the P, anisotropic scattering, and the S, angular quadrature

approximations. The Robinson measurements consist of in-vessel and ex-vessel dosimetry.

The FERRET adjustment for both sets is very small and consistent with the uncertainty bounds.

5.0 FERRET SENSITIVITY STUDIES The purpose of the sensitivity study is to evaluate the impact of the spectral uncertaintyand of the foil composition on the LSA.

5.1 Composition of the Multiple Foil Sensor Set In this case, the spectral uncertaintieswere held constant as well as the uncertainties associatedwith the reaction rates. The base case consisted of a set of six dosimeters (Cu, Ti, Fe, Ni. U-238. and Np-237). Ten additional cases were constructed by dropping one or more dosimeters from the base case and calculating the adjustedlcalculated(NC) ratio. These were then compared to the base case. The results indicate that for minimum uncertainty the dosimeter set should include Fe. U-238, and Np-237 foils.

5.2 Input Uncertainties In this part of the study the reaction rate and the spectrum uncertainties were assigned high, medium. and low values. Considering the medium-medium case as the base-case the magnitude of the adjusted flux changes very little. However, the associated uncertainty changed considerably more, as expected.

6.0 TECHNICAL EVALUATION

6.1 Introductionand Historical Note Least squares adjustments have been applied for many years in dosimetry analyses. The ASTM Standard E 944 (Reference 3) includes an extensive list of codes and methods that have been adopted for dosimetry problems. FERRET, in particular, which was developed at the Hanford Engineering Development Laboratory (HEDL). has been used in the liquid metal fast breeder reactor and the NRC-sponsoredLWR pressure vessel surveillance dosimetry improvement program (LWR-PV-SDIP). The PCA benchmark experiment was part of the LWR-PV-SDIP program.

In the past, issues have been raised regardingthe consistency of the MIC data bases for LWR a~~lications. The WOG stated that variations due to neutron eneraies. dosimeter locations.

tr%tsport and activationcross sections, and time periods have beekimoved.

As stated earlier. a~~lication of the FERRET code reauires three tvDes of i n ~ uinformation:

t (1) calculated n&t& energy spectrum and uncertainty. (2) rneasiied r e a c h rates and uncertainties, and (3) energy-dependent dosimetry reaction cross sections. The following sections evaluate each input type.

I 62 NeutronTransport Calculations Although the required informationis the neutron spectra at the location of the measurements, an accurate neutron transport calculation is needed to obtain the spectra at given locations.

The method is based on the synthesis technique that combines two two-dimensional solutions in (r, 8) and (r, z) to produce a three-dimensiondl flux:

The transport calculation is carried out using the discrete ordinates, finite difference code DORT, using the BUGLE-96 cross sections, derived from the ENDFIB-Vl file. This calculation adheres to the guidance in R G 1.I90 and, therefore, it is acceptable.

6.3 Geometric Modeling The geometric modeling should be'designed to preserve the physical accuracy of the material regions. This is accomplished by using the appropriate number of mesh points. The descriptionof this model states that up to 250 radial points, 110 azimuthal, and 150 axial points may be used. The point distribution is judicious by accommodating areas of expected high flux gradients and high total cross section. Also, the inner iteration convergence criterion is set at 0.001. All of these features agree with the guidance in RG 1.190, therefore. the proposed geometrical model is acceptable.

6.4 Core Source Because neutron sources are volumetric and in (x, y) geometry, their transposition to (r, 8) geometry must preserve the fuel volume. In addiion, to assure that the energy spectrum is correct the isotopic composition of the fissionable nuclei must be represented correctly for the irradiation period represented in the calculation. Finally, the number of neutrons released per fission is also a function of the isotopic composition of the fissionable nuclei. The proposed method is designed to maintainthe source volume and estimate the fissionable nuclei through bumup. The review indicates that the source calculation is acceptable because its transposition maintains the volume and accounts for its isotopic composition assuring correctness of the energy spectrum and the number of neutrons produced per fission.

6.5 Validation of the Transport Calculation The validation process is based on the guidance in RG 1.I90 and includes comparisons with the PCA benchmark experiment, the H. B. Robinson measurements, an analytic sensitivity study, and comparison to an extensive data base consisting of surveillance capsule measurements from operating plants. The validation addresses the adequacy of the transport calwlational method, method related uncertainties, and uncertainties due to imperfect knowledge of the input data.

The results of the validation are well within the 20 percent (la) uncertainty prescribed in RG 1.190. In addition, the transport methodology is based on WCAP-14040-A that has been approved by the NRC. The NRC staff finds the validation acceptable because the methodology

has been approved, the validation process is as prescribed by RG 1.I90 and, the results are within recommended limits.

6.6 Uncertainty Input to the Least-squaresAdjustment The adjustment algorithm is basedon the absolute value of the neutron spectrum at the location of the measurement. The input is the spectrum uncertainty and is expressed as an uncertainty matrix that contains the nomlization uncertainty related to the magnitude of the spectrum and groupwise uncertainties. The values of the normalization and groupwise uncertaintiespresented in WCAP-16083 are within the range of similar values in the literature and well within the uncertaintiesspecified in the transport solution, therefore, the proposed method is acceptable.

6.7 Reaction Rate Measurement and Uncertainties Flux measurements in operating plants are accomplishedwith a set of dosimeters that assures

~ o o dspectral coveraae. Such a set was identified in Section 3.8 above. ASTM standards CE series) outline m e h s to optimize the efficiency and to maximizethe accuracy of the dosimeter measurements. WCAP-16083-NP states that the applicable standard is used for each dosimeter. In addition to the threshold detectors (as listed in Section 3.8). solid state track recorders that directly measure total (fluence) exposure are also mentioned in WCAP-16803-NP. Conventional dosimeters measure activation that is converted analvticallv ' to an irradiation rate and subsequently to fluence. WCAP-16083-NP outlines the specials procedures required for the fission dosimeters in particular. WCAP-16803-NP outlines several tests that demonstrate the historical improvement and evolution of dosimetry measurement accuracy. The values of the (la) uncertainties for the dosimeter set in Section 3.8 are similar to those found in the literature. In summary, the NRC staff finds the reaction rate measurement uncertaintyto be acceptable because the measurement process followed accepted standard procedures, becausethey have been benchmarked to existing standards, and because the values are comparable to those found in the literature.

6.8 Dosimetry Cross Sections and Uncertainty Section 6.6 dealt with dosimeter uncertainties originating in the counting process. This section presents dosimeter activation cross section uncertainties. The uncertainties for the dosimeter set presented in Section 3.8 are part of the SNLRML library (Reference 8). These have been compiled from the most recent data and extensively tested for consistency and accuracy.

Because the SNLRML cross sections and their uncertainties are in general use for dosimetry work and becausethey have been subjected to extensive testing, they are acceptable for the proposed least squares adjustment for FERRET.

6.9 Data Comparison in the NlST U-235 Fission Field Measurementsof the dosimeter cross sections and their uncertaintiesare recorded in ASTM E 261-98, 'Standard Practice for Determining Neutron fluence, Fluence Rate, and Spectra by Radioactivation Technique' (Reference 11). Comparisons of calculated and measured values of the cross sections in the U-235 spectrum and the same from the PCA measurementsare shown in tabular form within ASTM E 261-98. Uncertainties documented in ASTM E261-98 are

within the (lo) range. The calculational method employed in ASTM E 261-98 is the same as that used by the WOG. therefore, the results are applicable. The same data are also available for the ~f-2b2spectrum with similar results. he& results support the claim for the value of the uncertainties and their suitability for the least squares analysis in FERRET and. therefore, the results are acceptable.

6.10 Evaluation of the PCA Simulator Benchmark RG 1.190 recommends the use of the results from the PCA experiment to compare and benchmarktransport calculationsand associated uncertainties. WCAP-16083-NP presents transmrt calculations for wsitions A, to A, reoresentina the inside surface of the thermal shield to the outside of the vessel: incl"din$ the inside the vessel thickness. The measured to calculated ratios fall in the range of 0.91 to 1.05. The adjusted values in t e r n of measured to adjusted ratios (MIA) are in the range of 0.94 to 1.06. The differences, the adjustments, and the uncertaintiesare small and consistent with the uncertainty bounds for the reaction rates and the neutron flux. The same conclusion is reached by analyzing similar calculations on PCA performed by HEDL, Oak Ridge National Laboratory (ORNL), and others.

In summary, analyses of the PCA benchmark experiment using the FERRET code yielded results that are consistent with prescribed uncertainty bounds. The uncertainty bounds become smaller when adjusted using the FERRET code. This supports the use of the FERRET code.

6.1 1 Evaluation of the H. B. Robinson Benchmark This k a case of laboratory quality surveillance applied to an operating plant. The analysis and evaluation were sponsored by the NRC, were performed by ORNL, a d are documented in NUREG/CR-6453 (Reference 9). A discrete ordinates code was used with the BUGLE-96 cross sections that'are based o; the ENDFIB-Vt file. The calculations used the P, inelastic scattering and the S, angular quadrature approximations. Review of the MIC ratios (before adjustment) indicates that they fall in the range of 0.95 to 1.11. The WA ratios adjusted individualdosimeter values fall in the range of 0.96 to 1.09. The FERREF code adjustment procedure reduced the uncertainty.

6.12 FERRET Sensitivity Studies Two studies examine the relative position of the threshold dosimeters to the in-vessel and ex-vessel spectrum and the effect of the coinpositionof the foil set in the accuracy of the results.

assuming that the full set of detectors results in the most accurate results. These studies are not a necessary part of the adjustment procedure but are instwctiveto the dosimetry analyst.

The first exercise indicates that in order to validate a calculation of the neutron flux, spectral weighting should be included in the calculations. The other indicatesthat to minimize the uncertaintyusing dosimeter measurementsthe dosimeter set should as a minimum include Fe, U-235, and NP-237.

6.13 Conditions for the Applicability of Least-squares Adjustment From the above discussion it is apparent that to successfully employ LSA, the measured and calculated values must be within their own uncertainty bounds. Should this not be the case,

both measured and calculatedvalues must be re-examined for possible errors and, if they cannot be found, the particular values causing the inconsistency should be disqualified.

WCAP-16803NP states that (1) in the past, data base consistency issues have been raised and (2) that the data base used in the FERRET benchmarking meets this condition.

7.0 CONCLUSION

SAND LIMITATION The WOG submitted the FERRET code for NRC staff review and approval. FERRET is a least squares adjustment code using calculated spectra weighting to minimize calculatedvalue uncertainties. In addition to the soectra, it also uses measured reaction rates and dosimetrv cross se&ons and associated uncertainties. The adjusted neutron fluxes could be usedto-form a data base to validate neutron transport calculations in accordance with the guidance in RG 1.1 90. The results of the FERRET adjustment have been benchmarkedby comparison to measurementsin NlST-calibratedfission sources, the PCA simulated benchmark experiment, and the H. B. Robinson vessel dosimetry benchmark experiment. The transport calculation and the dosimetry cross sections adhere to the guidance in RG 1.190.

For the reasons stated above, the NRC staff finds that the FERRET code is acceptable to be referenced in operating plant licensing actions subject to the following limitation:

LSA is acceptable if the adjustments to the MfCratios and to the calculated spectra values are within the assigned uncertainties of the calculated spectra, the dosimetry-measured reaction rates, and the dosimetry reaction cross sections. Should this not be the case, the user should re-examine both measured and calculated values for possible errors. If errors cannot be found, the particular values causing the inconsistency should be disqualified.

8.0 REFERENCES

1. Letter from F.P. Schiffley 11, Westinghouse Owners Group, to U.S. Nuclear Regulatory Commission. 'Transmission of WCAP-16083-NP. Revision 0, 'Benchmark Testing of the FERRET Code for Least Squares Evaluation of tight Water Reactor Dosimetry,' '

July 30.2004.

2. Letter from F.P. Schiffley 11, Westinghouse Owners Group. to U.S. Nuclear Regulatory Commission, 'Revision to WCAP-16083NP, Revision 0, 'Benchmark Testing of the FERRET Code for Least Squares Evaluation of tight Water Reactor Dosimetry,' '

March 30.2005. and Letter from S. Anderson Westinghouse Owners Group to Lambros Lois, U.S. Nuclear Regulatory Commission'Historical Perspective on Reactor Dosimetry Data Bases," August 22.2005.

3. ASTM Standard E 944-02. "Standard Guide for Application of Neutron Spectrum Adjustment Methods in Reactor Surveillance.' Annual Book of ASTM Standards, Section 12, Volume 12.02.2003.

4. Regulatory Guide 1.190, 'Calculational and Dosimetry Methodsfor Determining Pressure Vessel Neutron Fluence.' U.S. Nuclear Regulatory Commission, March 2001.

5. WORS 3.1. 'One-. Two-. and Three-Dimensional Discrete Ordinates NeutronlPhoton Cbde

~ r a n s ~ o r t ~~.&em,".~adiation Safety InformationComputationCenter, Computer Code Collection CCC-650. Oak Ridae National Laboratorv IORNL)

August 1996.

6. BUGLE-96, 'Coupled 47 Neutron, 20 Gamma Ray Group Cross Section Library Derived from ENDFIB-VI for LWR Shielding and Pressure Vessel Dosimetry Applications,'

Radiation Shielding lnformationCe-nter Data Library Collection DLC-185, Oak Ridge National Laboratory, March 1996.

7. WCAP-14040-A, Revision 3, 'Methodology Used to Develop Cold Overpressure MitigatingSystem Setpoints and RCS Heatup and Cwldown Limit Curves," by J. Andrachek, et. at., WestinghouseOwners Group, April 2002.

8. Data Library Collection DLG178. 'SNLRML Recommended Dosimetry Cross Section Compendium,' Radiation Shielding InformationCenter. Oak Ridge National Laboratory.

July 1994.

9. ASTM E261-98, 'Standard Practice for Determining Neutron Ruence, Fluence rate, and Spectra by RadioactivationTechnique,' ASTM Standards, Section 12, Volume 12.02, 2003.

10. NUREGICR-6454, 'Pool Critical Assembly Pressure Vessel Facility Benchmark,' by I.Rernek and F. B. K. Kam, Oak Ridge National Laboratory and U.S. Nuclear RegulatoryCommission, July 1997.

11. NUREGICR-6453. "H. B. Robinson Pressure Vessel Benchmark," by I. Remek and F. 6. K. Kam, Oak Ridge National Laboratory and U.S. Regulatory Commission, February 1998.

PrincipalContributor: Lambros Lois Date: January 10,2006

RESOLUTIONOF WOG COMMENTS ON DRAFT SAFETY EVALUATION FOR TOPICAL REPORT WCAP-16083-NP. REVISION 0.

'BENCHMARK TESTING OF THE FERRET CODE FOR LEAST SQUARES EVALUATION OF LIGHT WATER REACTOR DOSIMETRY' By letter dated November 17,2005, the Westinghouse Owners Group (WOG) provided comments on the safety evaluation (SE)for WCAP-16083-NP. Revision 0. 'Benchmark Testing of the FERRET Code for Least Squares Evaluation of Light Water Reactor Dosimetry.' The NRC staff agrees with the WOO comments and the modifications suggested by the WOG have been made to the final SE, as provided in the following table.

Table WOG Comments on the Draft SE for WCAP-16083 No. Draft SE WOG Comments NRC Staff Reference Resolution

1. Page 2. Line 11 Change "discrete elements" to "discrete Adopted ordinates*

2. Page 8. tines Insert blank line between lines 18 and 19. Adopted 18.19

3. Page 9. Line 1 Change "values should" to "values causing Adopted the inconsistency should"

4. Page 9. Line 20 Change "values should" to Values causing Adopted the inconsistency should"

WESTINGHOUSE NON-PROPRIETARY CLASS 3 WCAP-16083-NP-A Revision 0 Benchmark Testing of the FERRET Code for Least Squares Evaluation of Light Water Reactor Dosimetry S, L. Anderson Primary Components Asset Management Group May 2006 Approved: Electronicallv Approved Records Are Authenticated in the Electronic Document Management System John S. Carlson Primary Component Asset Management Approved: Electronicallv Approved Records Are Authenticated in the Electronic Document Management System Gordon C . Bischoff PWROG Program Management Office This work was performed under PWROG Project Number MUHP-7550 Westinghouse Electric Company LLC Energy Systems P.O. Box 355 Pittsburgh,PA 15230-0355 02006 Westinghouse Electric Company LLC All Rights Reserved

LEGAL NOTICE This report was prepared as an account of work performed by Westinghouse Electric Company LLC. Neither Westinghouse Electric Company LLC, nor any person acting on its behalf:

A. Makes any warranty or representation, express or implied including the warranties of fitness for a particular purpose or merchantability, with respect to the accuracy, completeness, or usefulness of the information contained in this report, or that the use of any information, apparatus, method, or process disclosed in this report may not infkinge privately owned rights; or B. Assumes any liabilities with respect to the use of, or for damages resulting from the use of, any information, apparatus, method, or process disclosed in this report.

COPYRIGHT NOTICE This report has been prepared by Westinghouse Electric Company LLC and bears a Westinghouse Electric Company copyright notice. As a member of the PWR Owners Group, you are permitted to copy and redistribute all or portions of the report within your organization; however all copies made by you must include the copyright notice in all instances.

DISTRIBUTION NOTICE This report was prepared for the PWR Owners Group. This Distribution Notice is intended to establish guidance for access to this information. This report (including proprietary and non-proprietary versions) is not to be provided to any individual or organization outside of the PWR Owners Group program participants without prior written approval of the PWR Owners Group Program Management Ofice. However, prior written approval is not required for program participants to provide copies of Class 3 Non Proprietary reports to third parties that are supporting implementation at their plant, and for submittals to the NRC.

May 2006 WCAP-16083-NP-A, Revision 0

PWR OWNERS GROUP MEMBER PARTICIPATION LIST* FOR PWROG PROJECT MURP-7550 Participant Utility Member Plant Site(s) Yes No AmerenUE Callaway (W) X American Electric Power D.C. Cook Units 1 & 2 (W) X Arizona Public Service Palo Verde Unit 1,2, & 3 (CE) X Constellation Energy Group Calvert Cliffs 1 & 2 (CE) X Constellation Energy Group Ginna (W) X Dominion Connecticut Millstone 2 (CE) X Dominion Connecticut Millstone 3 (W) X Dominion Kewaunee Kewaunee (W) X Dominion VA North Anna 1 & 2, Surry 1 & 2 (W) X Duke Energy Catawba 1 & 2, McGuire 1 & 2 (W) X Duke Energy Oconee 1,2,3 (B&W) X Entergy Nuclear Northeast Indian Point 2 & 3 (W) X Entergy Operations South Arkansas 2, Waterford 3 (CE), Arkansas 1 (B&W) X Exelon Generation Co. LLC Braidwood 1 & 2, Byron 1 & 2 (W), TMI 1 (B&W) X FirstEnergy Nuclear Operating Co Beaver Valley 1 & 2 (W) X FirstEnergy Nuclear Operating Co Davis-Besse (B&W) X Florida Power & Light Group St. Lucie 1 & 2 (CE) X Florida Power & Light Group Seabrook, Turkey Point 3 & 4 (W) X Nuclear Management Company Prairie Island 1 & 2 (W) X Nuclear Management Company Point Beach 1 & 2 (W) X Nuclear Management Company Palisades (CE) I X Omaha Public Power District Fort Calhoun (CE) X Pacific Gas & Electric Diablo Canyon 1 & 2 (W) X Progress Energy Robinson 2, Shearon Harris (W), Crystal River 3 (B&W) X PSEG -Nuclear Salem 1 & 2 (W) X Southern California Edison SONGS 2 & 3 (CE) X South Carolina Electric & Gas V.C. Summer (W) X South Texas Project Nuclear Operating Co. South Texas Project 1 & 2 (W) X Southern Nuclear Operating Co. Farley 1 & 2, Vogtle 1 & 2 (W) X Tennessee Valley Authority Sequoyah 1 & 2, Watts Bar (W) X May 2006 WCAP-16083-NP-A, Revision 0

TXU Power Comanche Peak 1 & 2 (W) X Wolf Creek Nuclear Operating Co. Wolf Creek (W) X PWR OWNERS GROUP INTERNATIONAL MEMBER PARTICIPATION LIST* FOR PWROG PROJECT M U W - 7 5 5 0 Participant Utility Member Plant Site(s) Yes No British Energy Sizewell B X Electrabel (Belgian Utilities) Doe1 1,2 & 4,Tihange 1 & 3 X Kansai Electric Co., LTD Mihama 1, Ohi 1 & 2,Takahama 1 X Korea Hydro & Nuclear Power Corp. Kori 1,2,3&4 X Yonggwang 1 & 2 (W)

Korea Hydro & Nuclear Power Corp. Yonggwang 3,4,5& 6 X Ulchin 3,4 ,5& 6 (CE)

Nuklearna Electrarna KRSKO Krsko X Nordostschweizerische Kraherke AG (NOK) Beznau 1 & 2 (W) X Ringhals AB Ringhals 2,3 & 4 X Spanish Utilities Asco 1 & 2,Vandellos 2, X Almaraz 1 & 2 Taiwan Power Co. Maanshan 1 & 2 X Electricite de France 54 Units X

• This is a list of participants in this project as of the date the final deliverable was completed.

On occasion, additional members will join a project. Please contact the PWROG Program Management Office to verify participation before sending documents to participants not listed above.

May 2006 WCAP-16083-NP-A, Revision 0

TABLE OF CONTENTS LIST OF TABLES vii LIST OF FIGURES viii 1.O INTRODUCTION

2.0 DESCRIPTION

OF THE LEAST SQUARES METHODOLOGY

2.1 BACKGROUND

2.2 APPLICATION OF THE METHODOLOGY 2.3 NEUTRON TRANSPORT CALCULATIONS AND UNCERTAINTIES 2.3.1 GEOMETRIC MODELING 2.3.2 CORE SOURCE 2.3.3 VALIDATION OF THE TRANSPORT CALCULATIONS 2.3.4 UNCERTAINTY INPUT TO THE LEAST SQUARES ADJUSTMENT 2.4 REACTION RATE MEASUREMENTS AND UNCERTAINTIES 2.5 DOSIMETRY CROSS-SECTIONS AND UNCERTAINTIES 3.0 TESTING OF THE FERRET PROCESSING PROCEDURES 3.1 DATA COMPARISONS IN THE NIST 2 3 5 FISSION

~ FIELD 3.2 EVALUATION OF THE PCA SIMULATOR BENCHMARK 3.3 EVALUATION OF THE H. B. ROBINSON BENCHMARK 4.0 FERRET SENSITIVITY STUDIES 4.1 COMPOSITION OF THE MULTIPLE FOIL SENSOR SET 4.2 INPUT UNCERTAINTIES 4.3 OPERATING POWER REACTOR COMPARISONS 5.0

SUMMARY

AND CONCLUSIONS

6.0 REFERENCES

M a y 2006 WCAP-16083-NP-A, Revision 0

vii LIST OF TABLES Table Title Page 3-1 Fission Spectrum Averaged Cross-Sections fiom ASTM E261-98 2 3 5 ~ 3-4 3-2 252~ ~ission f SpectnunAveraged Cross-Sections from ASTM E26 1-98 3-5 3-3 Comparison of Calculated 23?JFission Spectrum averaged Cross-Sections 3-6 3-4 Summary of Measurement Locations Within the PCA 12/13 Configuration 3-11 3-5 Least Squares adjusted Results for the PCA 12/13 Configuration Participating Laboratory Data from NUREGICR-33 18 3-6 M/C Comparisons for the PCA 12/13 Blind Test Experiment 3-7 Summary of Least squares Adjustment Results for the PCA 12113 Blind Test Experiment 3-8 FERRET Results for the PCA 12113 Blind Test Experiment 3-9 Comparison of Cment FERRET Least Squares Results with Prior Analyses - PCA 12113 Configuration 3-10 FERRET Results for the H. B. Robinson Benchmark Experiment 3-11 Summary of Least SquaresAdjustment Results for the H. B. Robinson Benchmark 3-12 M/C Comparisons for the H. B. Robinson Benchmark 4-1 Data Base Comparison for 104 In-Vessel Dosimetry Sets from 29 Reactors 4-2 Summary Comparison for Individual Foil Reactions May 2006 WCAP- 16083-NP-A, Revision 0

- 0 .

Vlll LIST OF FIGURES Figure Title Page 3-1 PCA 12113 Configuration - X,Y Geometry 3-10 4-1 Cumulative Sensor Response as a Function of Neutron Energy In-Vessel Surveillance Capsule Location 4-2 Cumulative Sensor Response as a Function of Neutron Energy Ex-Vessel Dosimetry Location May 2006 ~ ~ A F ' - 1 6 0 8 3 - N P - ~ ~ v i0s i o n

1.0 INTRODUCTION

In the assessment of the state of embrittlement of Light Water Reactor (LWR) pressure vessels, an accurate evaluation of the neutron exposure of the materials comprising the beltline region of the vessel is required. This exposure evaluation must, in general, include assessments not only at locations of maximum exposure at the inner diameter of the vessel, but also as a h c t i o n of axial, azimuthal, and radial location throughout the vessel wall.

In order to satisfy the requirements of 10CFR50,Appendices G and H['] for the calculation of pressureltemperaturelimit curves for normal heatup and cooldown of the reactor coolant system, fast neutron exposure levels must be defined at depths within the vessel wall equal to 25 and 75 percent of the wall thickness for each of the materials comprising the beltline region. These locations are commonly referred to as the 114T and 314T positions in the vessel wall. The 114T exposure levels are also used in the determination of upper shelf fracture toughness as specified in 10CFR50,Appendix G. In the determination of the pressurized thermal shock reference temperature (RTPTS)for comparison with the applicable screening criteria as defined in 1 0 ~ ~ R 5 0 . 6 1maximum

,[~] neutron exposure levels experienced by each of the beltline materials are required. These maximum levels will occur at the vessel inner radius. Furthermore, in the event that a probabilistic fracture mechanics analysis of the pressure vessel is performed, a complete neutron exposure profile is required for the entire volume of the pressure vessel beltline.

Methods acceptable to the NRC staff for the determination of the neutron exposure of LWR pressure vessels are described in Regulatory Guide 1.190, "Calculational and Dosimetry Methods for Determining Pressure Vessel Neutron ~luence."[~I This Regulatory Guide requires that the exposure projections be completed with a calculation and that measurements be used to qualify the calculational methodology as well as to identify biases in the calculations and to provide reliable estimates of the uncertainties in the exposure projections. The guide also stipulates that, in the determination of potential biases and uncertainties, measurement to calculation (MIC) comparisons should include a suitably weighted average of individual M/C values that accounts for the spectral coverage of the sensor set and the uncertainties in the measurements, the dosimetry cross-sections, and the neutron energy spectrum.

The methodology described in this report is based on the application of benchmarked plant specific neutron transport calculations providing neutron exposure maps throughout the reactor geometry including the pressure vessel wall, as well as all in-vessel and/or ex-vessel measurement locations. Evaluation of the dosimetry from the in-vessel andlor ex-vessel irradiations is subsequently accomplished with a least squares analysis including both the calculation and measurement data to establish best estimates of exposure parameters with reduced uncertainties at the locations of the measurements. The results of the least squares evaluations are used to demonstrate consistency among the various measurements and the baseline calculation at all measurement points and to validate the calculated results within the pressure vessel wall.

The use of a least squares adjustment method to perform dosimetry evaluations represents a rigorous approach to weighting the spectral coverage of individual sensor measurements and to combining the uncertainties associated with the transport calculations, measured reaction rates, and dosimetry cross-sections. Further, the results of the least squares evaluations provide comparisons of the best estimates of exposure parameters of interest (Fluence [E > 1.0 MeV and iron atom displacements [dpa]) with the corresponding calculated results before adjustment. It is these damage exposure parameters and their uncertainties rather than individual sensor reaction rates that are of the ultimate interest.

May 2006 WCAP-16083-NP-A, Revision 0

This approach is summarized in ASTM E 94-4-02, "Standard Guide for Application of Neutron Spectrum Adjustment Methods in Reactor ~urveillance,"~~~ which states the following:

"Adjustment methods provide a means for combining the results of neutron transport calculations with neutron dosimetry measurements in order to obtain optimal estimates for neutron damage exposure parameters with assigned uncertainties. The inclusion of measurements reduces the uncertainties for these parameter values and provides a test for the consistency between measurements and calculations and between different measurements."

It is important to emphasize that the least squares adjustments described in this report are used to combine multiple foil sensor measurements and plant specific calculated neutron spectra to determine best estimate exposure parameters only at in-vessel and ex-vessel measurement locations. This approach allows for appropriate spectral weighting of the measurements obtained from the multiple foil sensor sets and provides a rigorous treatment of the various measurement and calculational uncertainties associated with the process. Thus, the application of the least squares procedure allows the validation of the transport calculations to be based on comparisons of the exposure parameters of interest such as (P(E > 1.0 MeV) rather than on some combination of individual sensor M/C ratios.

h this application, the FERRET code[5sqis used for the least squares analysis of the dosimetry data sets. That is, the FERRET code is used solely to determine the best estimate exposure at locations where calculations and measurements coincide. The relationship between the best estimate exposure at the measurement locations and that at the pressure vessel wall, where no measurements exist, is not addressed internal to the FERRET code. Rather, the results of the plant specific neutron transport calculations are relied on to relate the neutron exposure at the pressure vessel wall to that at the various measurement locations within the reactor downcomer and or the ex-vessel cavity. This is analogous to an approach where one might use spatial calculations throughout the reactor core coupled with measurements at the center of selected he1 assemblies to generate detailed core power distribution maps which include locations removed from the measurement positions.

This is in contrast to the more sophisticated LEPRICON code systemt7]that performs simultaneous adjustments of the calculated neutron spectra at the various measurement locations, as well as at the pressure vessel wall in order to determine the best estimate exposure with associated uncertainties within the pressure vessel itself.

The purpose of this report is to describe the application of the FERRET code to the least squares analysis of typical LWR dosimetry sets irradiated at either in-vessel or ex-vessel locations and to summarize the testing of the FERRET approach in both benchmark and power reactor neutron fields. The results provided in this report are intended to validate the FERRET code and the least squares procedure for use in the analysis of LWR dosimetry.

May 2006 WCAP-16083-NP-A, Revision 0

2.0 DESCRIPTION

OF THE LEAST SQUARES METHODOLOGY

2.1 BACKGROUND

Least squares adjustment methods provide the capability to combine measurement data with the results of neutron transport calculations to establish a best estimate neutron energy spectrum with associated uncertainties at the measurement locations. Best estimates for key exposure parameters such as neutron flux, +(E > 1.0 MeV), or iron atom displacement rate, dpafs,along with their uncertainties are then easily obtained fiom the adjusted spectrum. The use of measurements in combination with detailed transport calculations results in a reduced uncertainty in the calculated spectrum and provides a method to identify any biases or inconsistencies that may exist in the baseline transport calculation or in the measured data.

The application of least squares adjustment methods in LWR dosimetry analysis is common throughout the dosimetry community. The American Society for Testing and Materials (ASTM) has addressed the use of adjustment codes in ASTM E944 "Application of Neutron Spectrum Adjustment Methods in Reactor Surveillance," and many industry workshops have been held to discuss the various applications. For example, the ASTM-EURATOM Symposia on Reactor Dosimetry holds workshops on neutron spectrum unfolding and adjustment techniques at each of its bi-annual conferences.

In ASTM E944, a comprehensive listing of available adjustment codes commonly employed in reactor surveillance programs including STAY'SL[*],L S L - M ~ ~FERRET, ~], and LEPRICON is provided. Each of these codes is publicly available from the Radiation Safety Information Computational Center (RSICC) at the Oak Ridge National Laboratory (ORNL).

The FERRET code was initially developed at the Hanford Engineering Development Laboratory (HEDL) and has had extensive use in both the Liquid Metal Fast Breeder (LMFBR) program and the NRC sponsored Light Water Reactor Surveillance Dosimetry Improvement Program (LWR-PV-SDIP). Examples of the prior use of the FERRET code for LWR applications include: 1) a re-evaluation of the dosimetry fiom commercial pressurized water reactor surveillance capsules['21and 2 ) an evaluation of the dosimetry included in the PCA blind test experiments.['31Both of these applications were completed in support of the LWR-PV-SDIP program.

The former evaluation was carried out to establish an updated surveillance capsule dosimetry data base for use in the establishment of improved trend curves defining the radiation induced shifi in reference nilductility transition temperature and the drop in upper shelf energy versus neutron fluence or displacementsper atom (dpa). These updated correlations were later used in the development of the trend curve data appearing in Regulatory Guide 1.99, Revision 2.[14*Is* 16*In The latter evaluation was completed to provide estimates of key exposure parameters (Fluence E > 1.0 MeV and dpa) for use in performing blind tests of neutron transport calculations in the PCA faci1ity.f12. 13] This, in turn, allowed blind test comparisons of calculated exposure parameters directly with the least squares results in addition to the comparisons with measured reaction rates obtained from the multiple foil sensor sets included in the PCA irradiations.

As a result of participation in several cooperative efforts associated with the LWR-PV-SDIP, the FERRET approach was adopted by Westinghouse in the mid 1980's as the preferred approach to the evaluation of LWR surveillance dosimetry. The least squares methodology was judged to be superior to the previously employed spectrum averaged cross-section approach that is totally dependent on the accuracy of the shape of the calculated neutron spectrum at the measurement May 2006 WCAP-16083-NP-A, Revision 0

locations. Further, the application of the least squares methodology allowed for a rigorous treatment of the uncertainties associated with the dosimetry evaluation process.

2.2 APPLICATION OF THE METHODOLOGY In general, the least squares methods, as applied to LWR dosimetry evaluations, act to reconcile the measured sensor reaction rate data, dosimetry reaction cross-sections, and the calculated neutron energy spectrum within their respective uncertainties. For example, relates a set of measured reaction rates, Ri, to a single neutron spectrum, @

,, through the multigroup dosimeter reaction cross-section, oi,, each with an uncertainty 6. The primary objective of the least squares evaluation is to produce unbiased estimates of the neutron exposure parameters at the location of the measurement.

The application of the least squares methodology requires the following input:

1 - The calculated neutron energy spectrum and associated uncertainties at the measurement location.

2 - The measured reaction rates and associated uncertainty for each sensor contained in the multiple foil set.

3 - The energy dependent dosimetry reaction cross-sections and associated uncertainties for each sensor contained in the multiple foil sensor set.

For LWR dosimetry applications, the calculated neutron spectrum is obtained fiom the results of plant specific neutron transport calculations that follow the guidelines specified in Regulatory Guide 1.190.[~'

The sensor reaction rates are derived fiom the measured specific activities obtained using established ASTM procedures. The dosimetry reaction cross-sections and uncertainties are obtained fiom the SNLRML dosimetry cross-section library. Each of these critical input parameters and associated uncertainties are discussed in this section.

23 NEUTRON TRANSPORT CALCULATIONS AND UNCERTAINTIES An accurate plant specific neutron transport solution with its associated uncertainty represents the foundation upon which the least squares analysis of in-vessel and ex-vessel dosimetry sets builds.

Therefore, in this section, the methodology employed by Westinghouse to perform neutron and gamma ray transport calculations for LWR applications is described in some detail. Also, included is a discussion of the benchmarking of the analysis approach and the resultant uncertainties associated with the calculational results.

For most routine analyses, the calculation of the neutron and gamma ray environment within the reactor geometry is completed on a fie1 cycle specific basis using the following three-dimensional flux synthesis technique:

May 2006 WCAP-16083-NP-A, Revision 0

where 4(r,B,z) is the synthesized three-dimensional neutron flux distribution, 4(r,B) is the transport solution in r,B geometry, $(r,z) is the two-dimensional solution for a cylindrical reactor model using the actual axial core power distribution, and $(r) is the one-dimensional solution for a cylindrical reactor model using the same source per unit height as that used in the r,B two-dimensional calculation.

All of the transport calculations are carried out using the DORT discrete ordinates code Version 3. l[lol and the BUGLE-96 cross-section library["]. The BUGLE-96 library provides a 67 group coupled neutron-gamma ray cross-section data set produced specifically for light water reactor application. Anisotropic scattering is treated with a minimum P3legendre expansion and the angular discretization is modeled with a minimum Sg order of angular quadrature. Energy and space dependent core power distributions as well as system operating temperatures are treated on a fuel cycle specific basis. The synthesis procedure combining the 4(r,8), @(r,z),and $(r) transport solutions into the three-dimensional flwdfluence maps within the reactor geometry is accomplished via post-processing of the scalar flux files generated as a part of the DORT output.

In some extreme cases were three-dimensional effects reduce the accuracy of the synthesis approach, a fully three-dimensional analytical approach may be used to perform the transport analysis. In these instances, the TORT["] three-dimensional discrete ordinates code, included as a part of the DOORS 3.1 code package, may be used in either x,y,z or r,B,z geometry to effect the three-dimensional solution. As in the case of the transport calculations used in the synthesis analysis, the TORT calculations use the BUGLE-96 cross-section library with a minimum P3 scattering approximation and a minimum S8order of angular quadrature.

2.3.1 GEOMETRIC MODELING In developing an analytical model of the reactor geometry of the LWR reactor geometry, nominal design dimensions are normally employed for the various structural components. In some cases as-built dimensions are available; and, in those instances, the more accurate as-built data are used for model development. However, for the most part as built dimensions of the components in the beltline region of the reactor are not available, thus, dictating the use of design dimensions.

Likewise, water temperatures and, hence, coolant density in the reactor core and downcomer regions of the reactor are normally taken to be representative of full power operating conditions.

The reactor core itself is treated as a homogeneous mixture of fuel, cladding, water, and miscellaneous core structures such as fuel assembly grids, guide tubes, etc. Sensitivities of the analytical results to tolerances in the internals dimensions, as well as to fluctuations in water temperature are discussed and quantified in Section 2.3.4.

The r,B geometric mesh description of the reactor model is normally accomplished using from 150 to 250 radial intervals and 60 to 110 azimuthal intervals depending on the overall size of the reactor and on the complexity required to model the core periphery, the in-vessel surveillance capsules, and the details of the reactor cavity. Mesh sizes are chosen to assure that proper convergence of the inner iterations is achieved on a pointwise basis. The pointwise inner iteration flux convergence criterion utilized in the r,B calculations is set at a value of 0.001.

The mesh selection process results in a smaller spatial mesh in regions exhibiting steep gradients, in material zones of high cross-section (C3,and at material interfaces. In the modeling of May 2006 WCAP-16083-NP-A, Revision 0

in-vessel surveillance capsules, a minimum set of 3 radial by 3 azimuthal mesh are employed within the test specimen array to assure that sufficient information is produced for use in the assessment of fluence gradients within the materials test specimens, as well as in the determination of gradient corrections for neutron sensors. Additional radial and azimuthal mesh are employed to model the capsule structure surrounding the materials test specimen array. In modeling the stainless steel baffle region at the periphery of the core, a relatively fine spatial mesh is required to adequately describe this rectilinear component int,8 geometry. In performing this x,y to r,8 transition, care is taken to preserve both the thickness and volume of the steel region in order to accurately address the shielding effectiveness of the component.

It should be noted that, although the example cited in this section is based on an octant symmetric model, the method itself is generally applicable even when such r,B symmetry does not exist. The only change in the approach would be the use of a larger geometric model in the calculation.

As in the case of the r,8 model, the r,z model is also based on the use of nominal design dimensions and 111power coolant densities. In this case, the homogenous core region is treated as an equivalent cylinder with a volume equal to that of the active core zone. The stainless steel former plates located between the core baffle and core barrel regions are also explicitly included in the model. The r,z geometric mesh description of the reactor model normally consists of from 130 to 230 radial intervals and fkom 90 to 150 axial intervals depending on the size of the reactor.

Again, spatial mesh sizes are chosen to assure that proper convergence of the inner iterations was achieved on a pointwise basis. The pointwise inner iteration flux convergence criterion utilized in the r,z calculations is also set at a value of 0.001.

The one-dimensional radial model used in the synthesis procedure consists of the same radial spatial mesh array included in the r,z model. Thus, radial synthesis factors are easily determined on a mesh wise basis throughout the entire geometry.

2.3.2 CORE SOURCE The spatial variation of the neutron source is generally obtained from a burnup weighted average of the respective power distributions from individual fuel cycles. These spatial distributions include pinwise gradients for all fuel assemblies located at the periphery of the core and typically include a uniform or flat distribution for fuel assemblies interior to the core. The spatial component of the neutron source is transposed from x,y to r,8 geometry by overlaying the mesh schematic to be used in the transport calculation on the pin by pin array and then computing the appropriate relative source applicable to each r,8 interval.

The x,y to ryetransposition is accomplished by first defining a fine r,8 mesh working array. The Ar and A8 of the fine mesh are usually chosen so that there is at least a 10x10 array of fine mesh over the area of each fuel pin at the core periphery. The coordinates of the center of each fine mesh interval and its associated relative source strength are assigned to the fine mesh based on the pin that is coincident with the center of the fine mesh. In the limit as Ar and A8 approach zero, this technique becomes an exact transformation.

Each space mesh in the r,8 transport geometry is checked to determine if it lies totally within the area of a particular fine r,8 working mesh. If it does, the relative source of that fine mesh is assigned to the transport space mesh. If, on the other hand, the transport space mesh covers a part of one or more fine mesh, then the relative source assigned to the transport mesh is determined by an area weighting process as follows:

May 2006 WCAP-16083-NP-A, Revision 0

where: Pm = the relative source assigned to transport mesh m.

Ai =the area of fine working mesh i within transport mesh m.

Pi = the relative source within fine working mesh i.

The energy distribution of the source is determined by selecting a fuel burnup representative of conditions averaged over the irradiation period under consideration and an initial fuel assembly enrichment characteristic of the core designs used over the applicable period. From this average burnup, a fission split by isotope including 2 3 s ~2 3, 8 ~ 2, 3 8 2~ 3, 9 ~ 2~ ?, l ? and

~ , 2 4 1 is~ derived;

~

and, from that fission split, composite values of energy release per fission, neutron yield per fission, and fission spectrum are determined. These composite values are then combined with the r,8 spatial distribution to produce the overall absolute neutron source for use in the transport calculations.

2.3.3 VALIDATION OF THE TRANSPORT CALCULATIONS The transport methodology described in Sections 2.3.1 and 2.3.2 of this report is identical to that described in the NRC approved version of WCAP-14040, Revision 3, Methodology Used to Develop Cold Overpressure Mitigation System Setpoints and RCS Heatup and Cooldown Limit The validation of the transport methodology is based on the guidance provided in Regulatory Guide 1.190. In particular, the validation consists of the the following stages:

1 - Comparisons of calculations with benchmark measurements from the Pool Critical Assembly (PCA) simulator at the Oak Ridge National Laboratory (ORNL).

2 - Comparisons of calculations with surveillance capsule and reactor cavity measurements from the H. B. Robinson power reactor benchmark experiment.

3 - An analytical sensitivity study addressing the uncertainty components resulting from important input parameters applicable to the plant specific transport calculations used in the exposure assessments.

4 - Comparisons of calculations with a measurement data base obtained fiom a large I

number of surveillance capsules withdrawn from a variety of pressurized water reactors.

At each subsequent application of the methodology, comparisons are made with plant specific dosimetry results to demonstrate that the plant specific transport calculations are consistent with the uncertainties derived from the methods qualification.

The first stage of the methods validation addresses the adequacy of basic transport calculation and dosimetry evaluation techniques and associated cross-sections. This phase, however, does not test the accuracy of commercial core neutron source calculations nor does it address uncertainties in operational or geometric variables that impact power reactor calculations. The second stage of the validation addresses uncertainties that are primarily methods related and would tend to apply May 2006 WCAP-16083-NP-A, Revision 0

generically to all fast neutron exposure evaluations. The third stage of the validation identifies the potential uncertainties introduced into the overall evaluation due to calculational methods approximationsas well as to a lack of knowledge relative to various plant specific parameters.

The overall calculational uncertainty is established from the results of these three stages of the validation process.

The following summarizes the uncertainties determined from the results of the first three stages of the validation process:

PCA Benchmark Comparisons 3%

H. B. Robinson Benchmark Comparisons 3%

Analytical Sensitivity Studies Internals Dimensions Vessel Inner Radius Water Temperature Peripheral Assembly Source Strength Axial Power Distribution Peripheral Assembly Burnup Spatial Distribution of the Source Other Factors 5%

The category designated "Other Factors" is intended to attribute an additional uncertainty to other geometrical or operational variables that individually have an insignificant impact on the overall uncertainty, but collectively should be accounted for in the assessment.

The uncertainty components tabulated above represent percent uncertainty at the lo level. In the tabulation, the net uncertainty of 11% from the analytical sensitivity studies has been broken down into its individual components. When the four uncertainty values listed above (3%, 3%,

11%, and 5%) are combined in quadrature, the resultant overall lo calculational uncertainty is estimated to be 13%.

To date, the methodology described in Section 2.2.1 coupled with the BUGLE-96 cross-section library has been used in the evaluation of dosimetry sets from 82 surveillance capsules from 23 pressurized water reactors. These withdrawals consisted of 2-5 capsules from individual reactors.

The comparisons of the plant specific calculations with the results of the capsule dosimetry are used to further validate the calculational methodology within the context of a lo calculational uncertainty of 13%.

2.3.4 UNCERTAINTY INPUT TO THE LEAST SQUARES ADJUSTMENT The neutron spectrum input to the least squares adjustment procedure is obtained directly from the results of plant specific transport calculations for each sensor location. The spectrum at each location is input in an absolute sense (rather than as simply a relative spectral shape). Therefore, within the constmints of the assigned uncertainties, the calculated data are treated equally with the measurements. The input uncertainties associated with the calculated spectrum must be consistent with the benchmarking results discussed in Section 2.3.3.

May 2006 WCAP-16083-NP-A, Revision 0

The uncertainty matrix for the calculated spectrum is constructed from the following relationship:

where % specifies an overall fractional normalization uncertainty and the fractional uncertainties Q and $ specify additional random groupwise uncertainties that are correlated with a correlation matrix given by:

where The first term in the correlation matrix equation specifies purely random uncertainties, while the second term describes the short range correlations over a group range y (8specifies the strength of the latter term). The value of 6 is 1.0when g = g' and 0.0otherwise.

The normalization uncertainty pertains primarily to the magnitude of the spectrum, whereas, the groupwise uncertainties deal with the shape of the spectrum relative to energy.

A typical set of parameters defining the input uncertainties for the calculated spectrum is as follows:

Flux Normalization Uncertainty (Rn) 15%

Flux Group Uncertainties (Rg, Rg')

(E> 0.0055 MeV)

(0.68eV < E < 0.0055MeV)

(E < 0.68 eV)

Short Range Correlation (8)

(E > 0.0055 M e V )

(0.68eV < E < 0.0055 MeV)

(E < 0.68 eV)

Flux Group Correlation Range (y)

(E> 0.0055MeV) 6 (0.68eV < E < 0.0055 MeV) 3 (E < 0.68 eV) 2 These uncertainty assignments provide an input covariance matrix are consistent with the calculational uncertainties defined through the benchmarking process.

May 2006 WCAP-16083-NP-A, Revision 0

2.4 REACTION RATE MEASUREMENTS AND UNCERTAINTIES Measurements at operating power reactors are generally accomplished with comprehensive multiple foil sensor sets including radiometric monitors (RM) and, in some instances for ex-vessel monitoring, solid state track recorders (SSTR). In general, sensor sets employed in Westinghouse dosimetry programs include materials in which the following reactions can be measured:

In-Vessel Ex-Vessel 63~ u ( n , a ) ~ c o 63

~ u ( n , a ) ~ C(Cd o Covered) 4&Ti(n,p)46~~ 4&Ti(n,p)46~c(Cd Covered) 54~e(n,p)54Mn 54~e(n,p)54Mn(Cd Covered) 58~i(n,p)58~o 58~i(n,p)58~o (Cd Covered) 2 3 8 ~ ( n , f (Cd

) ~ ~Covered) 2 3 8 ~ ( n , f(Cd

) ~ ~Covered) 2 3 7 ~ p ( n , f (Cd

) ~ ~Covered) 2 3 7 ~ p ( n , f(Cd

) ~ ~Covered) 59~ o ( n , ~ ) ~ ~ C o 59~o(n,y)60~o 59~ o ( n , ~ ) ~(CdC oCovered) 59~ o ( n , y ) ~ C(Cd o Covered)

These sensor sets provide adequate spectrum coverage in the fast neutron energy range above approximately 0.5 MeV and also include bare and cadmium covered cobalt sensors to provide an assessment of the thermal neutron flux at the measurement locations. These sensor sets are fblly consistent with the guidance specified in Section 2.1.1 of Regulatory Guide 1.190. Similar sensor set designs are also utilized by other vendors of LWR dosimetry programs.

Following irradiation, the specific activity of each of the radiometric sensors is determined using the latest version of ASTM counting procedures for each reaction. In particular, the following standards are applicable to the radiometric sensors typically used in LWR programs:

E523 Standard Test Method for Measuring Fast Neutron Reaction Rates by Radioactivation of Copper.

E526 Standard Test Method for Measuring Fast Neutron Reaction Rates by Radioactivation of Titanium.

E263 Standard Test Method for Measuring Fast Neutron Reaction Rates by Radioactivation of Iron.

E264 Standard Test Method for Measuring Fast Neutron Reaction Rates by Radioactivation of Nickel.

E704 Standard Test Method for Measuring Fast Neutron Reaction Rates by Radioactivation of Uranium-238.

E705 Standard Test Method for Measuring Fast Neutron Reaction Rates by Radioactivation of Neptunium-237.

E481 Standard Test Method for Measuring Neutron Fluence Rate by Radioactivation of Cobalt and Silver.

El005 Standard Method for Application and Analysis of Radiometric Monitors for Reactor vessel Surveillance.

May 2006 WCAP-16083-NP-A, Revision 0

E48 1 Standard General Methods for Detector Calibration and Analysis of Radionuclides.

Following sample preparation and weighing, the specific activity of each sensor is determined using a germanium gamma spectrometer. In the case of multiple foil sensor sets, these analyses are usually completed by direct counting of each of the individual sensors, or, as is sometimes the case with 2 3 8 and

~ 2 3 7 ~ pfission monitors from in-vessel irradiations, by direct counting preceded by dissolution and chemical separation of cesium from the sensor.

For ex-vessel dosimetry irradiations, gradient chains or wires are often included with the multiple foil sensor sets. For these gradient measurements, individual sensors are obtained by cutting the chains into a series of segments to provide data at appropriate intervals over the extent of the beltline region of the pressure vessel. The determination of sensor specific activities in these segments then proceeds in the same fashion as for individual foils from the multiple foil sensor sets. In general, data from the following reactions are obtained from the gradient chain measurements.

These data can be used in conjunction with high purity foil measurements to provide mappings of the neutron environment external to the reactor pressure vessel.

Solid state track recorders (SSTR) have also been used for the measurement of fission reaction rates at ex-vessel dosimetry locations. The determination of reaction rates fiom these sensors is performed in accordance with the following ASTM standard:

E854 Standard test Method for Application and Analysis of Solid State Track Recorder (SSTR) Monitors for reactor Surveillance In the SSTR analysis, the individual track recorders are optically scanned to determine the total number of fissions that occurred during the course of the irradiation. Since the scanning procedure results in an integral quantity representative of the entire irradiation, no radioactive decay corrections are required to determine the sensor reaction rates.

The individual radiometric and SSTR measurement data obtained fiom the counting laboratories, the physical characteristics of the sensors, and the operating history of the reactor are used to determine full power reaction rates characteristic of the irradiation period experienced by the foil sets. Generally, the reactor operating history data are obtained on a monthly basis for the sensor irradiation period. For the sensor sets used in both in-vessel and ex-vessel monitoring, the half-lives of the product isotopes are long enough that a monthly histogram descniing reactor operation has proven to be an adequate representation for use in radioactive decay corrections.

For the radiometric sensors used in LWR irradiations, reaction rates referenced to full power operation are determined fiom the following equation:

May 2006 WCAP- 16083-NP-A,Revision 0

where:

=measured specific activity (dpslg).

=sensor reaction rate averaged over the irradiation period and referenced to operation at a core power level of PXf(rpslatom).

=number of target element atoms per gram of sensor (atornlg).

=weight hction of the target isotope in the target material.

=number of product atoms produced per reaction.

=average core power level during irradiation period j (MW).

=maximum or reference core power level of the reactor (MJV).

=calculated ratio of #(I3 > 1.0 MeV) during irradiation period j to the time weighted average $(E > 1.0 MeV) over the entire irradiation period.

=decay constant of the product isotope (s-I).

=length of irradiation period j (s).

=decay time following irradiation period j (s).

and the summation is camed out over the total number of monthly intervals comprising the irradiation period.

In the above equation, the ratio P j k f accounts for month by month variation of power level within a given fuel cycle. The ratio Cj is calculated for each fuel cycle using the neutron transport methodology described in Section 2.1 of this report and accounts for the change in sensor reaction rates caused by variations in flux level due to changes in core power spatial distributions from fie1 cycle to fuel cycle. For a single cycle irradiation, Cj = 1.O. However, for multiple fuel cycle irradiations, particularly those using low leakage fuel management, the additional Cj correction must be utilized. This additional correction can be quite significant for sensor sets that have been irradiated for many fuel cycles in a reactor that has transitioned from non-low leakage to low leakage fuel management.

Since SSTR sensors are integrating devices not susceptible to radioactive decay of a product isotope, measurements of fissions per target atom, A, are converted directly to reaction rates using the following equation:

where the denominator in the above equation represents the total effective fill power seconds (EFPS) of reactor operation during the irradiation period of the solid state track recorders.

Prior to using the measured reaction rates for direct comparison with the results of transport calculations or as input to the least squares adjustment procedure, additional corrections must be made to 2 3 8 measurements

~ to account for the presence of 2 3 5 impurities

~ in the sensors and for the build-in of plutonium isotopes over the course of the irradiation. These corrections are location and fluence dependent and are obtained from a combination of calculated data from the May 2006 WCAP-16083-NP-A, Revision 0

plant specific discrete ordinates analysis described in Section 2.1 of this report and, when available, measurements made with U 5 foils ~ or solid state track recorders.

In addition to the corrections for competing neutron induced reactions in the 2 3 8 sensors,

~

corrections must also be made to both the 2 3 8 and

~ 2 3 7 ~ psensor reaction rates to account for gamma-ray induced fission reactions that occur during the irradiation. These photo-fission corrections are also location dependent and are obtained fiom the transport calculational methodology discussed in Section 2.1.

Typical corrections to the measured fission rates at in-vessel and ex-vessel sensor locations are summarized as follows:

m i c a 1 Correction In-Vessel Ex-Vessel

=%J Impurities 8-12% 0-15%

Pu Build-In 7-20% <I%

• " ~ ( ~ 9 f )

4-6% 3-5%

237N 9

1-2% 4 %

It should be noted that the corrections listed are typical values for a PWR plant with in-vessel capsules mounted on the outer radius of the thermal shield. These values can not be used with a plant specific sensor set. Rather, the appropriate corrections must be determined for each sensor set based on the actual plant specific geometry and irradiation history.

Along with the reaction rates themselves, the uncertainty associated with each of these measurements is also an important input to the least squares adjustment procedure. The overall uncertainty in the measured reaction rates includes components due to the basic measurement process, the irradiation history corrections, and the corrections for competing reactions. A high level of accuracy in the reaction rate determinations is assured by using laboratory procedures that conform to the ASTM National Consensus Standards listed earlier in this Section. In all cases, the latest available versions of the applicable standards are used in the dosimetry evaluations.

From these standards, it is noted that the achievable uncertainties in the measured specific activities of each of the sensors comprising typical LWR multiple foil sensor sets are as follows:

Reaction Precision Bias 63cu(n,a)"Oco 1% 3%

'"ri(n,~)'~sc 1% 3%

54~e(n,p)54Mn 1% 3%

S8~i(n,p)58~o 1% 3%

238~(n,f)FP 1% 5%

237~p(n,f)~~ 1% 5%

59Co(n,y)"Co 1% 5%

These uncertainties included the effects of counting statistics, sample weighing, detector calibration, source/detector geometry corrections, and product nuclide branching ratios.

May 2006 WCAP-16083-NP-A, Revision 0

In determining reaction rates from the measured specific activities, the following additional uncertainties are incurred:

Fission Product Competing Reaction Yield Half-Life Reactions 63cu(n,a)"co 0.02%

46Ti(n,p)46~~ 0.2%

54~e(n,p)s4Mn 0.2%

58~i(n,p)S8~o 0.2%

238~(n,f)~~ 1% 0.1% 4%

237~p(n,f)F? 2% 0.1% 1%

59~o(n,y)"~o 0.02%

After combing all of these uncertainty components, the sensor reaction rates derived fkom the counting and data evaluation procedures typically result in the following net uncertainties associated with the sensor reaction rates that are input to the least squares evaluation:

Reaction Rate Reaction Uncertainty (lo) 63cu(n,a)"~o 5%

46Ti(n,p)46~c 5%

s4~e(n,p)s4Mn 5%

58Ni(n,p)s8~o 5%

238~(n,f)~~ 10%

237~p(n,f)~~ 10%

59C O ( ~ , ~ ) ~ ~ C O 5%

In addition to the adherence to ASTM National Consensus Standards in the evaluation of sensor reaction rates, the procedures used by Westinghouse have been periodically tested via round robin counting exercises included as part of the NRC s-ponsoredLight Water Reactor Surveillance Dosimetry Improvement Program (LWR-SDP) as well as by evaluation of fluence counting standards provided by the National Institute of Science and Technology (NIST). A summary of the results of these counting validations is as follows:

1980 Round robin counting of the foil sets irradiated at the Thermal Shield Back (TSB) and Pressure Vessel Face (PVF) positions of the PCA simulator.

1981 Round robin counting of additional foil sets included in the first metallurgical simulated surveillance capsule, also irradiated in the PCA benchmark mockup.

These two counting exercises involved direct comparisons with measurements obtained by the Hanford Engineering Development Laboratory (HEDL). At the time of these irradiations, HEDL was a prime contractor providing measurement services for the PCA benchmark and was cross calibrated with NIST and the MOL laboratory in Belgium.

1985 Counting and evaluation of 4&Ti(n,p)46~~, 54~e(n,p)54Mn, and 58~i(n,p)58~o certified fluence standards supplied by NIST.

May 2006 WCAP-16083-NP-A, Revision 0

Comparisons with fluence standards involve the determination not only of the reaction rate of each foil, but also of the spectrum averaged cross-section in the NIST 2 3 5 irradiation

~ facility Thus, the comparisons with certified fluence standards test both the measurement process and the energy dependent reaction cross-sections used in the evaluation.

1992 Counting of NIST foils irradiated in an ex-vessel dosimetry experiment at the Trojan power reactor.

This exercise involved duplicate counting of a subset of irradiated foils by both Westinghouse and NIST to assure adequate cross-calibration of the laboratories so that data could be confidently mixed in the overall fluence evaluations performed by NIST and the Oak Ridge National Laboratory (ORNL).

1998 Round robin counting of 2 3 8 and

~ 2 3 7 ~certified p fluence standards irradiated by NIST in the MDRF facility at the University of Michigan.

As in the case of the 1985 radiometric sensor evaluations, the fluence standard involved the determination of the reaction rate of each sensor, but also of the spectrum averaged cross-sections in the MDRF facility.

The results obtained from these counting comparisons are summarized as follows:

1

[West]/@EDL] [West]/[M[ST]

1980 1981 1985 1992 1995 Average 63~ u ( n , a ) ~ c o 1.041 1.018 0.969 1.009 4&Ti(n,p)46~~ 1.036 1.012 1.030 1.026 54~e(n,p)54Mn 1.006 1.008 1.011 1.056 1.020 58~i(n,p)58~o 1.006 0.990 1.028 1.029 1.013 2 3 8 ~ ( n , f ) ~ ~ 1.014 1.014 1.017 1.015 2 3 7 ~ p ( n , f ) ~ ~ 1.006 1.017 1.097 1.040 59C O ( ~ , ~ ) ~ C O 1.017 1.017 1.017 These comparisons demonstrate that the procedures used by Westinghouse in the determination of sensor reaction rates have produced accurate and stable results over an extended period of time.

The crosscomparisons with HEDL and NIST support the reaction rate uncertainties used by Westinghouse in performing LWR fluence evaluations.

In addition to these periodic comparisons, laboratory calibrations with NIST supplied sources are also carried out on a routine basis.

2.5 DOSIMETRY CROSS-SECTIONSAND UNCERTATNTIES The third key set of input data for the least squares procedure includes the reaction cross-sections for each of the sensors included in the multiple foil dosimetry packages. The reaction rate cross-sections used by Westinghouse are taken from the SNLRML This data library provides reaction cross-sections and associated uncertainties, including covariances, for 66 dosimetry sensors in common use. Both the cross-sections and the uncertainties are provided in a fine multi-group structure for use in least squares adjustment applications.

May 2006 WCAP-16083-NP-A, Revision 0

These cross-sections were compiled f?om the most recent cross-section evaluations including ENDFB-VI and J R D F - ~ oand ~ ~ have

~ ] been tested with respect to their accuracy and consistency for least squares analyses. Further, the library has been empirically tested for use in fission spectra determination as well as in the fluence and energy characterization of 14 MeV neutron sources. Detailed discussions of the contents of the SNLRML library along with the evaluation process for each of the sensors is provided in Reference 23.

For the sensors of interest to LWR dosimetry applications, the following uncertainties in the fission spectrum averaged cross-sections are provided in the SNLRML documentation package:

Reaction Uncertainty (lo) 63~ u ( n , a ) ~ ~ c o 4.08416%

4?i(n,p)46~c 4.5 14.87%

54~e(n,p)54Mn 3.05-3.11%

58~i(n,p)58~o 4.49456%

238~(n,f)~~ 0.54-0.64%

237~p(n,f)~~ 10.32-10.97%

59C O ( ~ , ~ ) ~ C O 0.79-3.59%

These tabulated ranges provide an indication of the dosimetry cross-section uncertainties associated with typical sensor sets used in LWR irradiations.

May 2006 WCAP- 16083-NP-A,Revision 0

3.0 TESTING OF THE FERRET PROCESSING PROCEDURES As noted in Section 2.0 of this report, the FERRET least squares adjustment code is employed by Westinghouse to combine the results of plant specific neutron transport calculations, dosimetry reaction cross-sections, and multiple foil reaction rate measurements to determine the best estimate values of damage exposure parameters (fluence [E > 1.0 MeV] and iron atom displacements [dpa]) along with their associated uncertainties at the measurement location.

As implemented in the FERRET analysis, the least squares evaluation of a given data set can be subdivided into the following two stages:

1 - A pre-adjustment procedure performed by the SAND module that processes the calculated neutron spectrum and the SNLRML dosimetry cross-sections into the 53 energy group structure required by the FERRET module.

2 - The subsequent application of the least squares algorithm in the FERRET module itself.

These two stages of the overall least squares adjustment procedure were individually tested by data comparisons in standard and benchmark fields as well as by analysis of a series of in-vessel and ex-vessel irradiations at operating power reactors.

In this section, results of least squares evaluations of dosimetry results from irradiations in the National Institute of Standards and Technology (NIST) Standard 2 3 s thermal ~ fission field, the NIST Standard 2 5 2 ~ spontaneous f fission field, the PCA simulator benchmark, and the H. B.

Robinson power reactor benchmark are presented and discussed. In addition, results of the least squares evaluation of a database of in-vessel and ex-vessel power reactor measurements are described. These power reactor evaluations include data from a variety of different reactor designs and for several repeat measurement campaigns at individual reactors.

3.1 DATA COMPARISONS IN TEE NIST 23% FISSION FIELD The sensor reaction cross-sections and associated uncertainties play a key role in the least squares evaluation of dosimetry data sets. Therefore, it is important to assess both the overall accuracy of these cross-sections and the impact on that accuracy of any data processing included in the evaluation procedure.

The least squares approach used by Westinghouse makes use of the SNLRML dosimetry cross-section library. This comprehensive library has been recommended in ASTM E l 0 18-01, "Standard Guide for Application of ASTM Evaluated Cross-Section Data ~ile"['~] for use in LWR applications. The library is provided by the Radiation Safety Information Computational Center (RSICC) in a 640 neutron group format spanning an energy range from thermal to 20.0 MeV.

Prior to use in the least squares adjustment, this fine group library is collapsed to a broad group structure consisting of 53 groups using the calculated neutron spectrum at the measurement location as a weighting function. The data comparisons from the standard field irradiations were used to determine the level of accuracy of the base cross-section library, as well as to demonstrate the adequacy of the collapsing procedure used to generate the 53 group library.

In ASTM E261-98, "Standard Practice for Determining Neutron Fluence, Fluence Rate, and Spectra by Radioactivation ~echni~ues,"['~~ fission spectrum averaged cross-sections applicable to the 2 3 s thennal

~ fission field and the 2 s 2 ~spontaneous f fission field are provided for a variety May 2006 WCAP-16083-NP-A, Revision 0

of threshold activation detectors that are used in power and research reactor irradiations. In this data compilation, both calculated and measured spectrum averaged cross-sections are provided along with their evaluated uncertainties. The magnitude of errors in the processed dosimetry cross-section library can be judged by the observed disagreement between the calculated spectrum averaged cross-sections and the corresponding measured values for the standard 2 3 5 ~

and 2 5 2 ~fields.

f The data listed in Tables 3-1 and 3-2 have been extracted from Table 3 of ASTM E261 and are representative of the foil sets used in power reactor irradiations and in the PCA benchmark irradiations. This subset of the ASTM E261 information includes all of the threshold reactions typically used in Light Water Reactor (LWR) surveillance capsule and ex-vessel dosimetry irradiations.

For the comparisons shown in Tables 3-1 and 3-2, the authors of ASTM E261 based the calculated spectrum averaged reaction cross-sections on data from the recommended SNLRML library. Since the Westinghouse methodology and the evaluations provided in ASTM E261-98 are both based on the same dosimetry cross-section library, the calculated spectrum averaged cross-sections produced by the SANDEERRETprocessing procedure should closely match the calculated values cited in the standard. Significant differences between the two sets of calculated spectrum averaged cross-sections would indicate errors in the processing procedure. Comparisons of the calculated spectrum averaged cross-sectionsfrom ASTM E261 with the corresponding cross-sectionsprocessed by Westinghouse are listed in Table 3-3.

The calculation to measurement comparisons given in Table 3-1 indicate that for the 2 3 5 thermal~

fission field, all of the C/M ratios except 46Ti(n,p)46~~ fall within one standard deviation of the combined uncertainty in the calculation and measurement. For these reactions, the agreement between calculation and measurement is within 5%. In the case of the 46Ti(n,p)46~~ reaction, the C/M ratio falls within two standard deviations of the combined uncertainty with the calculation falling within 11% of the measured value.

In the case of the 2 5 2 ~spontaneous f fission field, the comparisons provided in Table 3-2 show that, with the exception of the, 2 3 8 ~ ( n , f"SIn(n,n')"5m,

)~~, and 4&Ti(n,p)46~~reactions, all of the C/M comparisons fall within one standard deviation of the combined uncertainty in the calculations and measurements. The C/M ratios for the 2 3 8 ~ ( n , f 1"In(n,n')"5m

)~~, reactions fall within two standard deviations and the C/M ratio for the 4&Ti(n,p)46~~ reaction falls within three standard deviations of the combined uncertainty. For all reactions other than 46Ti(n,p)46~~ the agreement between calculation and measurement is within 6%. The calculated spectrum averaged cross-section for the 4?i(n,p)46~~ reaction falls within 11% of the measured value.

The comparisons provided in Tables 3-1 and 3-2 demonstrate that the SNLRML dosimetry cross-sections as processed by the authors of ASTM E261 produce accurate representations of the spectrum averaged cross-sections in the NIST standard fission fields. The Westinghouse least squares approach uses this same base dosimetry cross-section library, but a somewhat different processing procedure.

To compare the cross-section processing procedure used in the Westinghouse approach to expand the calculated input spectrum, spectrum weight the dosimetry cross-sections, and re-collapse the spectrum and dosimetry cross-sections to the FERRET 53 energy group structure, the ASTM E261 calculations for the 2 3 5 thennal

~ fission field were duplicated for the foil reactions contained in both the power reactor sensor set and PCA sensor set. In performing this calculation, the ENDF/B-VI 2 3 5 fission

~ spectrum supplied with the BUGLE-96 cross-section library was May 2006 WCAP- 16083-NP-A, Revision 0

input to the SANDmERRET procedure as the calculated spectrum. The dosimetry cross-sections for were taken directly from the SNLRML library.

Comparisons of the Westinghouse processed spectrum averaged cross-sections with the calculated values from ASTM E261-98 are listed in Table 3-3 for both the power reactor and PCA sensor sets. An examination of the data given in Table 3-3 shows that the spectrum averaged cross-sections calculated by Westinghouse using the SAND pre-processing module are essentially identical to the calculated values given in ASTM E261-98, with the largest difference being at the 1% level.

The comparison results summarized in Table 3-3 coupled with the CiM results listed in Tables 3-1 and 3-2 demonstrate that using the SNLRML dosimetry cross-section library coupled with the algorithms included in the SAND pre-processing module to produce a spectrum weighted broad group library, results in an appropriate cross-section representation for use in the FERRET least squares adjustment algorithm.

May 2006 WCAP-16083-NP-A, Revision 0

Table 3-1 2 3 5 Fission

~ SpectrumAveraged Cross-Sections fiom ASTM E26 1-98 m i c a 1 Power Reactor Sensor Sets Spectrum Average Cross-Section (millibarns)

Reaction Calculation Measurement C/M 63~u(n,a)60~o 0.521 (2.85%, 6.05%) 0.50 (11.O%) 1.042 (12.87%)

46~i(n,p)46~~ 10.43 (2.46%, 5.4%) 11.6 (3.45%) 0.899 (6.86%)

54~e(n,p)54Mn 80.18 (2.17%, 4.69%) 80.5 (2.86%) 0.996 (5.91%)

58~i(n,p)58~o 105.69 (2.43%, 4.52%) 108.5 (5.0%) 0.974 (7.16%)

238~(n,f)FP 306.23 (0.53%, 4.21%) 309.0 (2.6%) 0.991 (4.98%)

237~p(n,f)~~ 1330.1 (9.33%, 4.31%) 1344.0 (4.0%) 0.990 (11.0%)

PCA Sensor Sets Spectrum Average Cross-Section (millibarns)

Reaction Calculation Measurement C/M 27~l(n,a)2%a 0.727 (1.40%, 6.95%) 0.706 (3.97%) 1.030 (8.13%)

58~ i ( n , ~ ) " C o 105.69 (2.43%, 4.52%) 108.5 (5.0%) 0.974 (7.16%)

"s~n(n,n')"5m~n 186.35 (2.17%, 4.17%) 190.3 (3.84%) 0.979 (6.07%)

103~h(n,n')103mRh 706.02 (3.1%, 4.14%) 733.0 (5.2%) 0.963 (7.33%)

238~(n,f)~~ 306.23 (0.53%, 4.21%) 309.0 (2.6%) 0.991 (4.98%)

237 Np(n,f)FP 1330.1 (9.33%, 4.3 1%) 1344.0 (4.0%) 0.990 (11.O%)

Notes: 1 - The tabulated data were taken fiom Table 3 ofASTM E261-98.

2 - For the calculated values, the cross-section and spectrum components of the uncertainty, respectively, are shown in parentheses.

3 - The measurement uncertainty is also shown in parentheses.

4 - The uncertainty in the C/M ratio represents a sum in quadrature of the measurement and calculational uncertainty.

May 2006 WCAP- 16083-NP-A, Revision 0

Table 3-2 2 5 2 Fission

~f Spectrum Averaged Cross-Sections from ASTM E26 1-98 m i c a 1 Power Reactor Sensor Sets Spectrum Average Cross-Section (millibarns)

Reaction Calculation Measurement C/M 63~ u ( n , a ) ~ C o 0.678 (2.83%, 1.38%) 0.689 (1.98%) 0.984 (3.72%)

4&Ti(n,p)46~~ 12.56 (2.45%, 1.18%) 14.09 (1.76%) 0.891 (3.24%)

54~e(n7p)54Mn 88.12 (2.14%, 0.79%) 86.92 (1.34%) 1.014 (2.65%)

58~i(n,p)"co 115.31 (2.40%, 0.73%) 117.6 (1.3%) 0.981 (2.83%)

238~(n,f)~~ 315.39 (0.53%, 0.4%) 325.0 (1.63%) 0.970 (1.76%)

237~p(n,f)FP 1335.0 (9.2%, 0.23%) 1361.0 (1.58%) 0.981 (9.43%)

PCA Sensor Sets Spectrum Average Cross-Section (millibarns)

Reaction Calculation Measurement C/M 27~ l ( n , a ) ~ ~ N a 1.04 (1.36%, 1.61%) 1.017 (1.47%) 1.019 (2.57%)

58~ i ( n , p ) ~ ~ c o 115.31 (2.40%, 0.73%) 117.6 (1.3%) 0.98 1 (2.83%)

l1sh(n,n')"5mh 189.8 (2.16%, 0.38%) 197.6 (1.3%) 0.961 (2.55%)

'03~h(n,n')103mRh 714.45 (3.08%, 0.27%) 757.0 (4.0%) 0.944 (5.06%)

23%(n,f)~~ 3 15.39 (0.53%, 0.4%) 325.0 (1.63%) 0.970 (1.76%)

237~p(n,f)~~ 1335.0 (9.2%, 0.23%) 1361.0 (1.58%) 0.981 (9.43%)

Notes: 1 - The tabulated data were taken from Table 3 of ASTM E261-98.

2 - For the calculated values, the cross-section and spectrum components of the uncertainty, respectively, are shown in parentheses.

3 - The measurement uncertainty is also shown in parentheses.

4 - The uncertainty in the C/M ratio represents a sum in quadrature of the measurement and calculational uncertainty.

May 2006 WCAP-16083-NP-A,Revision 0

' Table 3-3 Comparison of Calculated 2 3 5 Fission

~ SpectrumAveraged Cross-Sections m i c a 1 Power Reactor Sensor Sets Spectrum Average Cross-Section (millibarns) Ratio Reaction ASTM E261-98 SAND/FERRET FERRETE261 63~u(n,a)60~o 0.52 1 0.523 1.004 4vi(n,p)46~c 10.4 10.3 0.990 54~e(n,p)54Mn 80.2 80.3 1.001 58~ i ( n , p ) ~ ~ ~ o 106 106 1.OOO 238~(n,f)~~ 306 306 1.OOO 237~p(n,f)~ 1330 1330 1.OOO PCA Sensor Sets I Spectrum

- Average Cross-Section I (millibarns) Ratio Reaction ASTM E261-98 SAND/FERRET FERRETE261 27~l(n.a12"Na 0.727 0.729 1.003 May 2006 WCAP- 16083-NP-A,Revision 0

3.2 EVALUATION OF THE PCA SMULATOR BENCHMARK The guidance provided in Regulatory Guide 1.190[~] recommends using measured data from the Pool Critical Assembly (PCA) pressure vessel simulator['3**' 211 to benchmark neutron transport calculational methods for application to LWR pressure vessel fluence determination. The documentation describing the PCA experimental program and the subsequent evaluation of the dosimetry data obtained from the simulator irradiations also affords the opportunity to compare the results from the Westinghouse least squares methodology using the FERRET code with similar analyses completed by other laboratories. These comparisons are valuable in that they highlight any differences that may occur due to the use of different input or different least squares adjustment codes.

In Reference 20, several least squares evaluations of dosimetry data from the PCA 12/13 configuration were documented. These evaluations were completed by Rolls Royce and Associates (RR&A), Hanford Engineering Development Laboratory (HEDL), and Oak Ridge National Laboratory (ORNL), each using a different least squares adjustment code. The results of these least squares analyses were used to establish recommended values of key neutron exposure parameters ($(E > 1.0 MeV) and dpa) for this blind test configuration.

A plan view of the PCA reactor and pressure vessel simulator showing materials characteristic of the core axial midplane is shown in Figure 3-1. The configuration shown in Figure 3-1 was developed from dimensional information provided in Reference 13 and reflects the latest available geometric data for the simulator. During the PCA experiments, measurements were taken at several locations within the mockup to provide traverse data extending from the reactor core outward through the pressure vessel simulator and on into the void box. The specific measurement locations are illustrated on Figure 3-1 and listed in Table 34. All of the measurements of interest were obtained on the lateral centerline of the mockup at an elevation opposite the axial midplane of the simulator.

The measurement locations specified in Table 3 4 provide data sufficient to generate calculation/measurement comparisons throughout the entire 12/13 configuration. The data afford the opportunity for comparisons over a wide attenuation range with a changing neutron energy spectrum within the carbon steel simulator wall. Thus, this simulator benchmark experiment provides an excellent test for the evaluation of transport calculation and dosimetry evaluation methodologies for both in-vessel and ex-vessel irradiations.

The least squares evaluations documented in Reference 20 included data fiom a subset of the measurement points listed in Table 34. In particular, least squares comparisons were provided at locations A2, A4, A5, A6, and A7. These data locations were intended to simulate an in-vessel surveillance capsule mounted on the outer diameter of the thermal shield, the 1/4T, 1/2T, and 3/4T locations interior to the pressure vessel wall, and an ex-vessel dosimetry location in the reactor cavity external to the pressure vessel wall.

The Reference 20 analyses were completed using three different least squares adjustment codes (SENSAK[~~], FERRET['], and LSL-M~~']). The calculated input spectra for the SENSAK analysis were based on discrete ordinates calculations using the EUROLIB-4 cross-section library. The FERRET and LSL-M2 analyses likewise used calculated spectra from discrete ordinates calculations, but based on the use of the ENDFB-IV SAILOR cross-section library. All of these least squares evaluations used the recommended measurement data from the PCA irradiations and dosimetry reaction cross-sections from the ENDFB-V data files.

May 2006- WCAP-1 608=-A, Revision 0

The least squares results provided in Section 7.1 of Reference 20 are reproduced in Table 3-5. The data included in Table 3-5 show that, in terms of adjusted neutron flux (E > 1.0 MeV), the RR&A and HEDL analyses produce essentially identical results while the ORNL evaluation yields adjusted results that are lower by 4.0-8.5 percent. However, all of the reported results are consistent within the combined uncertainties quoted for the adjusted results. Thus, these prior least squares evaluations of the PCA benchmark data did not indicate any significant differences in the adjustments produced by these three codes.

As a test of the current Westinghouse least squares methodology using the FERRET code, these PCA evaluations were repeated using updated input based on the now available ENDFIB-VI neutron transport and dosimetry reaction cross-section libraries. The intent of this updated evaluation was to test the capabilities of the FERRET code as applied by Westinghouse using current methodology and data libraries.

Due to the relative small size of the PCA configuration, calculations using the typical 3D synthesis approach tend to break down at locations toward the back side of the simulator (A6, A7). Therefore, the spectnun input to the least squares analysis was taken fi-om a filly three-dimensional calculation using the TORT code[lolrun in X,Y,Z geometry. The TORT calculations were run using the BUGLE-96 cross-section libraryr"] with a P3 scattering cross-section expansion and an S8angular quadrature. The uncertainty associated with the calculated spectrum was based on the formulation described in Section 2.3 of this report. The measured reaction rates used in the least squares evaluation of dosimetry sets from locations A1 through A7 within the PCA 12113 configuration were the recommended values from Reference 20. The reaction rate uncertainties used in the analysis were 5% and 10% for the non-fission and fission sensors, respectively. The dosimetry reaction cross-sections and cross-section uncertainty data were obtained from the SNLRML library.[231 Results of the FERRET least squares evaluations of the PCA dosimetry from locations A1 through A7 are given Tables 3-6 through 3-9. In Table 3-6, comparisons of the measurement to calculation ratios for each sensor are listed before and after application of the least squares procedure. The comparisons before adjustment (MIC) show that the baseline calculation and the reaction rate measurements are in good agreement for all reactions at all measurement locations with the M/C values falling in a range of 0.91-1.05. The linear average of the M/C data at each measurement location ranges fiom 0.97-1.02, with the standard deviations in these averages varying fi-om 2.048%. All of these M/C comparisons fall within the 15% standard deviation ascribed to the unadjusted calculation. In aggregate, these comparisons indicate reasonable consistency between the calculated and measured reaction rates and imply that any adjustments to the calculated spectra should be relatively small.

The comparisons after adjustment (M/A) show that the adjustments are indeed small, but do result in improved agreement between the calculated and measured data. After adjustment, the MIA data for the individual sensors fall in the range of 0.94 to 1.06, while the linear average of the MIA ratios at each measurement location ranges from 0.99-1.01. The standard deviations in these linear averages have been reduced, ranging from 1.O-3.6%.

In Table 3-7, the neutron flux (E > 1.0 MeV) at each measurement location is provided before and after adjustment. The data in Table 3-7 show that the net adjustment in the fast neutron flux was 3% or less, depending on location and that the inclusion of the measurement information has reduced the uncertainty in the magnitude of the fast flux from 15% to 4%. This improved uncertainty in the fast neutron flux is consistent with the corresponding improvement in the calculated sensor reaction rates.

May 2006 WCAP-16083-NP-A, Revision 0

Detailed comparisons of the data adjustments for each measurement location are given in Table 3-

8. These comparisons show that in all cases, the adjustments performed by the FERRET analysis are small and consistent with the input uncertainty bounds for the reaction rates and calculated neutron flux.The X2 per degree of freedom associated with each of the analyses indicate good data consistency for all cases.

Finally, in Table 3-9, the results of the current FERRET evaluation of the PCA data are compared with the prior analysis listed in Table 3-5. The newer FERRET results are essentially identical to the previous results reported by RR&A and HEDL, but with an improved uncertainty. The improvement in the uncertainty is attributable to greater accuracy in the input calculated spectra which were based on ENDFB-VI transport cross-sectionsrather than on the ENDFB-IV data libraries that were available at the time of the initial evaluations.

The analyses summarized in Tables 3-6 through 3-9 demonstrate that the FERRET code as applied by Westinghouse yields results consistent with those produced by other least squares adjustment codes for the PCA benchmark analyses. Further, the PCA evaluations show that the FERRET least squares analyses produced consistent results with no anomalous behavior for a wide range of neutron spectra characteristic of in-vessel and ex-vessel dosimetry locations as well as for locations interior to the carbon steel pressure vessel wall.

May 2006 WCAP-16083-NP-A, Revision 0

Figure 3-1 PCA 12/13 Configuration - X,Y Geometry M a y 2006 WCAP-16083-NP-A, Revision 0

Table 3-4 Summary of Measurement Locations Within the PCA 12113 Configuration Measurement Location ID Y(cm)

Core Center A0 -20.57 Thermal Shield Front A1 12.0 Thermal Shield Back A2 23.8 Pressure Vessel Front A3 29.7 Pressure Vessel 1/4T A4 39.5 Pressure Vessel 1/2T A5 44.7 Pressure Vessel 3/4T A6 50.1 Void Box A7 59.1 Table 3-5 Least Squares Adjusted Results for the PCA 12/13 Configuration Participating Laboratory Data from NLTREGICR-33 18 RR&A HEDL ORNL Location +(E> 1.0)  % std. +(E> 1.0)  % std. $(E > 1.0)  % std.

A2 4.0 1e-07 9 3.85e-07 8 A4 4.50e-07 7 4.58e-08 7 4.22e-08 4 A5 2.21e-07 7 2.21e-08 7 2.03e-08 4 A6 9.73e-09 8 9.82e-09 7 8.91e-09 4 A7 2.88e-09 43 May 2006 WCAP-16083-NP-A, Revision 0

Table 3-6 W C Comparisons for the PCA 12/13 Blind Test Experiment Comparisons Before Adjustment M/C Ratio A1 A2 A3 A4 A5 A6 A7 2 7 ~ l ( n , a ) 2 (Cd 4 ~ a) 1.OO 1.OO 0.95 0.96 0.97 0.98 58~i(n,p)5%o (Cd) 1.04 1.02 0.99 1-00 1.01 0.96 115~n(n,n~)~'5mJn (Cd) 1.04 1.02 0.96 0.96 0.99 0.99 1.04 103~h(n,n7)103mRh (Cd) 1.01 0.97 0.97 0.96 1.02 1.05 1.05 2 3 8 ~ ( n , f(Cd)

)~~ 0.91 1.OO 1.02 1.05 1.04 2 3 7 ~ p ( n , f(Cd ) ~ )~ 1.05 0.99 0.98 1.01 0.96 Average 1.02 1.OO 0.97 0.98 1.OO 1.01 1.02

% Standard Deviation 2.0 2.4 4.8 2.1 2.1 3.7 4.1 Comparisons After Adjustment

- M/A Ratio A1 A2 A3 A4 A5 A6 A7 1.01 1.01 1.01 1.oo 0.98 0.98 0.99 1.02 1.01 1.02 1.01 1.01 1.oo 1.OO 1.01 0.99 0.99 0.99 1.06 0.98 0.98 0.94 1.OO

- 0.95 - 1.02 - 1.02 - 1.OO - 1.04 Average 0.99 1.OO 1.OO 1.OO 1.OO 0.99 1.01

% Standard Deviation 1.O 1.6 3.6 1.7 1.5 2.7 2.2 Table 3-7 Summary of Least Squares Adjustment Results for the PCA 12113 Blind Test Experiment

+(E> 1.0 MeV)

[n/cm2-s]

Location ( Calculated 1 Adjusted A1 1 3.77e-06 (15%) 1 3.88e-06 (4%)

Note: Numbers in parentheses represent one standard deviation.

May 2006 WCAP-16083-NP-A, Revision 0

Table 3-8 FERRET Results for the PCA 12/13 Blind Test Experiment Location A1 (x2/Degreeof Freedom = 0.03)

Reaction Rate [rpslatom] Adjustment Reaction Measured Calculated Adjusted  % of Calc.

27~ ( n , a ) ~ ~(CdN aCov) 5.48e-33 5.49e-33 5.53e-33 0.7 58~i(n,p)58~o (Cd COV.) 6.3 1e-3 1 6.07e-3 1 6.23e-3 1 2.6 "51n(n,n')"5"'In (Cd Cov.) 1.05e-30 1.01e-30 1.04e-30 3.O

'03Rh(n,n')'03"'Rh (CD Cov.) 4.06e-30 4.00e-30 4.10e-30 2.5 2 3 8 ~ ( n , f(Cd

) ~ COV.)

~

2 3 7 ~ p ( n , f(Cd

) ~ ~Cov.)

$(E >1.0 MeV) [n/cm2-s] 3.77e-06 3.88e-06 2.9 Location A2 (x2/Degreeof Freedom = 0.08)

Reaction Rate [rps/atom] Adjustment Reaction Measured Calculated Adjusted  % of Calc.

27~l(n,ct)2"Na (Cd Cov) 7.16e-34 7.14e-34 7.20e-34 0.8 58~ i ( n , p ) ~ (Cd~ ~Cov.)

o 6.72e-32 6.57e-32 6.64e-32 1.1 1'5~n(n,n')"5mIn(Cd Cov.) 1.14e-31 1.12e-31 1.13e-31 0.9

'03~h(n,n')'03""Rh(CD Cov.) 4.50e-3 1 4.64e-3 1 4.60e-3 1 -0.9 2 3 8 ~ ( n , f(Cd

) ~ COV.)

~

2 3 7 ~ p ( n , f(Cd

) ~ ~COV.)

$(E >1.0 MeV) [n/cm2-s] 4.26e-07 4.26e-07 0.2 Location A3 (x2/Degreeof Freedom = 0.13)

Reaction Rate [rpslatom] Adjustment Reaction Measured Calculated Adjusted  % of Calc.

27Al(n,a)2"Na(Cd Cov) 3.13e-34 3.28e-34 3.15e-34 -4.0 58~ i ( n , ~ (Cd ) ~Cov.)

~ ~ o 2.50e-32 2.53e-32 2.45e-32 -3.2 "51n(n,n')"S% (Cd Cov.) 3.68e-32 3.84e-32 3.72e-32 -3.1 103Rh(n,n')103"Rh(CD Cov.) 1.47e-31 1.52e-3 1 1.47e-3 1 -3.3 2 3 8 ~ ( n , f(Cd

) ~ COV.)

~ 5.91e-12 6.47e-32 6.27e-32 -3.1 237~p(n,f)FP (Cd COV.) 3.05e-3 1 2.90e-3 1 2.91e-31 0.3

$(E >1.0 MeV) [n/cm2-s] 1.42e-07 1.38e-07 -2.9 May 2006 WCAP-16083-NP-A,Revision 0

Table 3-8 (Continued)

FERRET Results for the PCA 12/13 Blind Test Experiment Location A4 a2/Degree of Freedom = 0.06)

Reaction Rate [rpslatom] Adjustment Reaction Measured Calculated Adjusted  % of Calc.

27~l(n,a)2%a(Cd Cov) 7.15e-35 7.46e-35 7.23e-35 -3.1 58Ni(n,p)58~o(Cd COV.) 5.69e-33 5.70e-33 5.58e-33 -2.1 11sIn(n,n')115% (Cd Cov.) l.lle-32 1.15e-32 1.13e-32 -1.7

'03Rh(n,n')'03"Rh (CD COV.) 5.67e-32 5.89e-32 5.70e-32 2 238~(n,f)FP (Cd Cov.) 1.79e-32 1.79e-32 1.75e-32 -2.2 237~p(n,f)FP (Cd COV.) 1.20e-31 1.21e-32 1.19e-32 -1.7

$(E >1.0 MeV) [n/cm2-s] 4.73e-08 4.62e-08 -2.4 Location A5 (x2/Degreeof Freedom = 0.04)

Reaction Rate [rps/atom] Adjustment Reaction Measured Calculated Adjusted  % of Calc.

2 7 ~ l ( n , a ) 2 4(Cd

~ a Cov) 2.92e-35 3.01e-35 2.94e-35 -2.3 58~i(n,p)58~o (Cd Cov.) 2.25e-33 2.24e-33 2.23e-33 -0.4 11sIn(n,n')115mIn(Cd Cov.) 5.20e-33 5.27e-33 5.27e-33 0.0 103Rh(n,n')103"Rh(CD COV.) 3.24e-32 3.17e-32 3.22e-32 1.6 2 3 8 ~ ( n , (Cd f ) ~COV.)

~ 7.88e-33 7.70e-33 7.70e-33 0.0 237~p(n,f)FP (Cd Cov.) 6.56e-32 6.71e-32 6.66e-32 -0.7

$(E >1.0 MeV) [n/cm2-s] 2.24e-08 2.25e-08 0.2 Location A6 (x2/Degreeof Freedom = 0.04)

Reaction Rate [rps/atom] Adjustment Reaction Measured Calculated Adjusted  % of Calc.

"~l(n,a)~%a(Cd Cov) 1.12e-35 1.15e-35 1.12e-35 -2.6 5 8 ~ i ( n , p ) 5 8(Cd

~ o COV.) 7.99e-34 8.29e-34 8.14e-34 -1.8 11sIn(n,n')115mIn (Cd Cov.) 2.23e-33 2.26e-33 2.25e-33 -0.4 103Rh(np')'03"Rh (CD Cov.) 1.67e-32 1.60e-32 1.66e-32 3.7 2 3 8 ~ ( n , f(Cd ) ~COV.)

~ 3.26e-33 3.09e-33 3.07e-33 -0.6 237~p(n,f)FP (Cd Cov.) 3.46e-32 3.43e-32 3.47e-32 1.2

$(E>1.0 MeV) [n/cm2-s] 9.87e-09 9.88e-09 0.1 May 2006 WCAP- 16083-NP-A, Revision 0

Table 3-8 (Continued)

FERRET Results for the PCA 12/13 Blind Test Experiment Location ~ 7 ( ~ ~ / D eof~ Freedom ree = 0.04 Reaction Rate [rpslatom] Adjustment Reaction Measured Calculated Adjusted  % of Calc.

27~l(n,a)2%a(Cd Cov) 58~i(n,p)58~o (Cd Cov.)

llsIn(n,n')llSmIn(Cd Cov.) 6.43e-34 6.20e-34 6.40e-34 3-2 103~h(n,n')103"'Rh (CD Cov.) 4.83e-33 4.60e-33 4.80e-33 4.3 2 3 8 ~ ( n , f(Cd

) ~ ~COV.) 8.65e-34 8.33e-34 8.61e-34 3.4 237~p(n,f)FP (Cd Cov.) 9.60e-33 9.97e-33 9.98e-33 0.1

$(I? >1.0 MeV) [n/cm2-s] 2.72e-09 2.81e-09 3.2 Table 3-9 Comparison of Current FERRET Least Squares Results with Prior Analyses PCA 12113 Configuration FERRET RR&A HEDL ORNL 4 p 1 . 0 ) % std. 4QDl.O)  % std. 4 p 1 . 0 ) % std. $(E>1.0) % std.

A1 3.88e-06 4 A2 4.26e-07 4 4.01e-07 9 3.85e-07 8 A3 1.38e-07 4 A4 4.62e-08 4 4.50e-07 7 4.58e-08 7 4.22e-08 4 A5 2.25e-08 4 2.21 e-07 7 2.21e-08 7 2.03e-08 4 A6 9.88e-09 4 9.73e-09 8 9.82e-09 7 8.91e-09 4 A7 2.81e-09 4 2.88e-09 43 May 2006 WCAP-16083-NP-A, Revision 0

33 EVALUATION OF THE H. B. ROBINSON BENCHMARK In performing the benchmarking of the neutron transport methodology described in Section 2.3 of this report, comparisons of calculations using ENDFB-VI cross-sections with measured reaction rates fkom both in-vessel and ex-vessel dosimetry sets irradiated in the H. B. Robinson Cycle 9 benchmark experiment were used. This direct comparison of the calculated reaction rates with measurements showed excellent agreement. As a further demonstration of the application of the least squares procedure using the FERRET code, an evaluation of the in-vessel and ex-vessel data from the H. B. Robinson benchmark was also completed.

The neutron transport calculations input to the least squares analysis were completed using the standard synthesis approach described in Section 2.3 of this report. The calculations were run using the BUGLE-96 cross-section libraryr"' with a Pg scattering cross-section expansion and an Sg angular quadrature. The uncertainty associated with the calculated spectrum was likewise based on the formulation described in Section 2.3 of this report. The measured reaction rates used in the least squares evaluation of the dosimetry were the recommended values from Reference 25.

The reaction rate uncertainties used in the analysis were 5% and 10% for the non-fission and fission sensors, respectively. The dosimetry reaction cross-sections and cross-section uncertainty data were obtained fi-om the SNLRML library.[231 Results of the least squares evaluation of the H. B. Robinson in-vessel and ex-vessel sensor sets are provided in Tables 3-10 through 3-12. Detailed comparisons of the data adjustments for each measurement location are given in Table 3-10. These comparisons show that in both cases the adjustments performed by the FERRET analysis are small and consistent with the input uncertainty bounds for the reaction rates and calculated neutron flux. The x2per degree of fi-eedom associated with each of the analyses indicate good data consistency for all cases.

In Table 3-1 1, the neutron flux (E > 1.0 MeV) at each measurement location is provided before and after adjustment. The data in Table 3-11 show that the net adjustment in the fast neutron flux was 1% for the in-vessel dosimetry and 2% for the ex-vessel data set. Further, the inclusion of the measurement information has reduced the uncertainty in the magnitude of the fast flux from 15%

to 5% and 7% at the in-vessel and ex-vessel locations, respectively. This improved uncertainty in the fast neutron flux is consistent with the corresponding improvement in the calculated sensor reaction rates and with the FERRET analysis performed for the PCA data sets.

In Table 3-12, comparisons of the measurement to calculation ratios for each sensor are listed before and after application of the least squares procedure. The comparisons before adjustment (MIC) show that the baseline calculation and the reaction rate measurements are in good agreement for all reactions at all measurement locations with the M/C values falling in a range of 0.95-1.1 1. The linear average of the M/C data at both measurement locations is 1.03, with the standard deviations in these averages being 4.4 and 5.2% for the in-vessel and ex-vessel data, respectively. All of these M/C comparisons fall within the 15% standard deviation ascribed to the unadjusted calculation. In aggregate, these comparisons indicate reasonable consistency between the calculated and measured reaction rates and imply that any adjustments to the calculated spectra should be relatively small.

The comparisons after adjustment M A ) show that the adjustments are indeed small, but do result in improved agreement between the calculated and measured data. After adjustment the M A data for the individual sensors is improved and falls in the range of 0.96 to 1.09 while the linear average of the M/A ratios at both measurement location is 1.01. The standard deviations in May 2006 WCAP-16083-NP-A, Revision 0

these linear averages have been reduced, ranging from 3.6% at the in-vessel location to 5.0% in the reactor cavity.

The comparisons summarized in Tables 3-10 through 3-12 support the analyses previously presented for the PCA benchmark experiment and M e r demonstrate the applicability of the FERRET code and the least squares procedure used by Westinghouse for analysis of LWR dosimetry data.

May 2006 WCAP- 16083-NP-A, Revision 0

Table 3-10 FERRET Results for the H. B. Robinson Benchmark Experiment In-Vessel (x2/Degreeof Freedom = 0.23)

Reaction Rate [rpslatom] Adjustment Reaction Measured Calculated Adjusted  % of Calc.

63~(n,a)60~o 3.98e-17 3.85e-17 3.97e-17 3.1 4&Ti(n,p)46~~ 6.49e-16 6.07e-16 6.30e-16 3.8 54~e(n,p)54Mn 3.83e-15 3.79e-15 3.84e-15 1.3 58~i(n,p)58~o 4.88e-15 5.13e-15 5.10e-15 -0.6 238~(n,p)~~ 1.80e-14 1.68e-14 1.69e-14 0.6

"'N~(~,~)FP 1.20e-13 1.18e-13 1.18e-13 0.0

$(E >1.0 MeV) [n/cm2-s] 4.60e+10 4.57e+10 -0.7 Ex-Vessel (x2/Degreeof Freedom = 0.44)

Reaction Rate [rpslatom] Adjustment Reaction Measured Calculated Adjusted  % of Calc.

63~~(n,a)60~o 4.01e-19 3.88e-19 4.09e-19 5.4 46~i(n,p)46~~ 6.17e-18 5.54e-18 5.82e-18 5.1 54~e(n,p)54Mn 3.59e-17 3.53e-17 3.67e-17 4.0 58~i(n,p)58~o 5.29e-17 5.32e-17 5.49e-17 3.2 238~(n,p)~~ 2.72e-16 2.44e- 16 2.50e-16 2.5

$(E >1.0 MeV) [n/cm2-s]

9.65e+08 9.80e+08 1.6 Table 3 - 11 Summary of Least SquaresAdjustment Results for the H. B. Robinson Benchmark 1 I 4.(-E > 1.0 MeV) 1 1

[n/cm2-s]

Location Calculated Adjusted A/C In-Vessel 4.60e+10 (15%) 4.57et-10 (5%) 0.99 Note: Numbers in parentheses represent one standard deviation.

May 2006 WCAP- 16083-NP-A, Revision 0

3-19 Table 3-12 MIC Comparisons for the H. B. Robinson Benchmark Comparisons Before Adjustment Reaction M/c In-Vessel Ex-Vessel 63~u(n,a)"~o 1.03 1.03 4&Ti(n,p)46~~ 1.07 1.11 s4~e(n,p)54Mn 1.01 1.02 58~ i ( n , p ) ~ ~ C o 0.95 0.99 238~(n,p)~~ 1.07 1.11 237~p(nyp)~ 1.02 Average 1.03 1.03

% Standard Deviation 4.4 6.5 Comparisons After Adjustment M/C Reaction In-Vessel Ex-Vessel 63~u(n,a)"co 1.OO 0.98 4&ri(n,p)46~~ 1.03 1.06 s4~e(n7p)54Mn 1.OO 0.98 s8~i(n7p)s8~o 0.96 0.96 238~(n,p)~~ 1.07 1.09 237~p(n,p)~~ 1.02 Average 1.01 1.01

% Standard Deviation 3.6 5.0 M a y 2006 WCAP-16083-NP-A, Revision 0

4.0 FERRET SENSITIVITY STUDIES The information discussed in this section is intended to provide an understanding of how the composition of the multiple foil sensor sets and the input values for the uncertainties in the calculated neutron spectrum and measured reaction rates impact the results of the least squares analysis in terms of both the magnitude and uncertainty of the adjusted spectrum.

The threshold foils comprising typical LWR sensor sets respond to different portions of the neutron energy spectrum. These multiple foil measurements should be thought of as a set of partial measurements of the flux or fluence rather than as a group of complete and independent determinations. Of particular interest for weighting and averaging threshold detector

. measurements is the spectrum coverage of the individual foils which recognizes that such measurements do not form an equivalent observation set and hence are not easily matched to the principle of maximum likelihood.

This response is highlighted in Figures 4-1 and 4-2 for the neutron spectra characteristic of an in-vessel surveillance capsule location and an ex-vessel dosimetry location. The graphical representations shown in Figures 4-1 and 4-2 provide response profiles for the 63~u(n,a),

54 Fe(n,p), 238~(n,f), and 237~P(n,f) threshold reactions, as well as for the neutron flux (E > 1.0 MeV). The 46~i(n,p)and 58~i(n,p) reactions exhibit behavior similar to the 63~u(n,a), 54~e(n,p) reactions, respectively.

From Figures 4.1 and 4.2, it is evident that the response of the higher threshold reactions exhibits significantly different behavior than does the neutron flu (E > 1.0 MeV), while the fission monitor response shows a better match to the spectral behavior of the neutron flux. This behavior suggests that in order to validate a calculation of the neutron flux (E > 1.0 MeV), significant spectral weighting of measured reaction rates should be included in the comparisons. The least squares approach allows this spectral weighting to be included in a rigorous manner. The data fiom Figures 4-1 and 4-2 also indicate that the makeup of the foil set could have an impact on the final results of the dosimetry comparisons.

4.1 COMPOSITION OF THE MULTIPLE FOIL SENSOR SET In order to assess the impact of the makeup of the sensor set on the final solution of the least squares adjustment, a parametric study was performed for a typical LWR dosimetry data set. In the parametric study, the calculated neutron spectrum and uncertainty was held constant along with the uncertainties associated with the measured reaction rates. The base case consisted of an evaluation including all of the threshold reactions while the variations in the analysis were accomplished by deleting foil reactions individually and in combination. The 11 cases analyzed in the parametric study are summarized as follows:

M a y 2006 WCAP-16083-NP-A, Revision 0

Foils Included in the EERRET Analysis 2 3 8 ~ 237N Case Cu Ti Fe Ni 1 X X X X X X 2 X X X X X 3 X X X X 4 X X X 5 X X 6 X X 7 X X 8 X 9 X 10 X 11 X The results of the least squares evaluations for each of these 11 cases are as follows:

$(E> 1.0 MeV)  % Diff

[n/cm2-s] From Case Calculated Adjusted A/C Case 1 1 4.45e+10 (15%) 4.64e+10 (6%) 1.04 2 4.45e+10 (15%) 4.64e+10 (6%) 1.04 0.0 3 4.45e+10 (15%) 4.75e+10 (6%) 1.07 2.4 4 4.45e+10 (15%) 4.76e+10 (6%) 1.07 2.6 5 4.45e+10 (15%) 4.69e+10 (7%) 1.05 1.1 6 4.45e+lO (15%) 4.78e+lO (7%) 1.07 3.O 7 4.45&10 (15%) 4.79e+lO (8%) 1.08 3.2 8 4.45e+10 (15%) 4.60e+10 (12%) 1.03 -0.9 9 4.45e+10 (15%) 4.84e+10 (9%) 1.09 4.3 10 4.45e+10 (15%) 4.69e+10 (9%) 1.05 1.1 11 4.45e+10 (15%) 4.80e+10 (12%) 1.08 3.4 Note: Numbers in parentheses represent one standard deviation.

This data tabulation shows that in terms of the magnitude of the adjusted solution the results for all cases are within the uncertainty of the base case (Case 1). However, the uncertainty in the adjusted flux increases as the content of the foil set is reduced with the highest uncertainties occurring when only a single foil is used in conjunction with the transport calculation.

The data in the tabulation M e r shows that for a minimum uncertainty solution the foil set should consist of at least Fe, U, and Np foils (Case 4). The addition of Cu, Ti, and Ni do not enhance the capability of the foil set. This is due to the fact that the very high threshold of the 63 Cu(n, a)and 46~i(n,p)reactions places their response well above the important energy range for the neutron flux (E > 1.0 MeV) and the 58~i(n,p) response is so similar to 54~e(n,p) that the reaction is redundant. Further reductions in the foil set fiom the Fe, 2 3 8 2~3, 7 ~ ppackage results in an increased uncertainty in the adjusted flux.

May 2006 WCAP-16083-NP-A, Revision 0

4.2 INPUT UNCERTAINTIES A second sensitivity study was completed to evaluate the impact of the input uncertainties in the measured reaction rates and the calculated neutron flux on the adjusted solution and the final uncertainties determined by the FERRET least squares procedure. In performing this sensitivity study, the following uncertainty matrix was evaluated for the reaction rates and calculated spectra:

Reaction Rate Uncertainties Uncertainty Category Reaction 'Qpe High Medium Low Non-Fission 10% 5% 2.5%

Fission 20% 10% 5.0%

Neutron Spectrum Uncertainties Uncertainty Category Reaction m e High Medium Low Normalization 30% 20% 10%

Spectrum Groups 1-53 30% 20% 10%

Spectrum Groups 28-48 50% 25% 10%

Spectrum Groups 48-53 100% 50% 25%

In this sensitivity study, the "High" category would tend to overstate achievable uncertainties, the "Medium" category would represent a routinely achievable case, and the "low" category would be equivalent to values achievable in the laboratory or benchmark environment.

The results of the input uncertainty sensitivity study are summarized as follows:

Adjusted $(E> 1.0 MeV) and % Standard Deviation Flux Spectrum Reaction Rate Uncertainty Uncertainty High Medium Low High 2.05e+10 2.11e+lO 2.14e+10 (16%) (14%) (13%)

Medium 1.99e+10 2.04e+10 2.10e+10 (12%) (10%) (9%)

Low 2.00e+10 1.98e+10 2.03e+10 (7%) (6%) (5%)

M a y 2006 WCAP-16083-NP-A,Revision 0

Considering the "Medi~m~~-~'Medium" case as the baseline, the magnitude of the adjusted flux varies by less than 5% for all of the cases evaluated, indicating that the magnitude of the adjusted flux is dependent primarily on the magnitude of the inputs, rather than on the input uncertainties.

However, the associated uncertainty varies in a predictable trend from 5% for the "Low"-"Low" case to 16% for the "High"-"High".

The indications from this sensitivity study indicate that the FERRET least squares algorithm is operating as anticipated.

M a y 2006 WCAP-16083-NP-A, Revision 0

Figure 4-1 Cumulative Sensor Response as a Function of Neutron Energy In-Vessel Surveillance Capsule Location

• -*=

1.1 -------------- ---"."- *--A -.-

I I

1.o N-0.9

/ 0.8 #

u I 0

E 0.7 I r'

2 0.6 , 8 B 0

.2 0.5 I a

u Q 0 Y 0.4 -

E 5 e 7;:

0.3 - 8 8

0.2 0I I I I I I I I I 0.5 2.0 3.5 5.0 6.5 8.0 9.5 11.0 12.5 14.0 15.5 17.0 18.5 Neutron Energy (MeV)

May 2006 Revision 0 WCAP-16083-NP-A,

4-6 Figure 4-2 Cumulative Sensor Response as a Function of Neutron Energy Ex-Vessel Dosimetry Location o~0.8 - , X,____

0X

- i.-Cu-63(n,a)

X -Ar - Fe-54(n,p)

CO >, -x,- U-238(n,f) 4P - Np-237(n,f)

E03 > A~

6.5 8.0 9.5 11.0 12.5 14.0 15.5 17.0 18.5 Neutron Energy (MeV)

WGAP-16083-NP-A, Revision 0 May 2006 May 2006 WCAP-16083-NP-A, Revision 0

43 OPERATING POWER REACTOR COMPARISONS In addition to the sensitivity studies described above, the transport methodology and the least squares dosimetry evaluations have been extensively compared with data fiom operating power reactors. These comparisons are intended to provide support for the validation of the transport calculation itself as well as validation for the uncertainties assigned to the results of those calculations.

One ,concern that has been raised in regard to measurement/calculation comparisons fiom operating power plants is that the uncertainties in the positioning of the dosimetry within the reactor could lead to large uncertainties in the measurement/calculation data base that would, in turn,tend to reduce the value of these comparisons. The data comparisons provided in this section are intended to demonstsate that the dosimetry locations within operating power reactors are known within the manufacturing tolerances and that, when calculations and dosimetry processing are completed on a consistent basis, the uncertainties in the measurement~calculation data base are less than those associated with the stand alone calculation.

In Section 2.3 of this report, it was noted that the combination of benchmarking comparisons and analytical sensitivity studies resulted in an evaluated uncertainty of 13% (lo) in the calculation of neutron flux or fluence (E > 1.0 MeV). Based on the analytical sensitivity studies using allowable manufacturing tolerances for the reactor components as limits, the component of the uncertainty in measurement/calculation comparisons due to miss-positioning of surveillance capsule dosimetry is in the range of 4-5% (1o).

In order to demonstrate the validity of these uncertainty assessments, comparisons using the methodologies described in Section 2.0 of this report were applied to a large measurement data base consisting of 104 surveillance capsule dosimetry sets withdrawn fiom 29 reactors having Westinghouse or Combustion Engineering as the Original Equipment Manufacturer (OEM) of the NSSS. The data base includes surveillance capsules attached to the pressure vessel wall, capsules mounted on the external surfaces of thermal shields or neutron pads, and accelerated capsules mounted on the core barrel. The irradiation times included in the data base range fiom 1 to more than 20 fuel cycles. The number of capsules withdrawn from individual reactors ranges fiom two to six. The overall data base includes five basic Westinghouse internals configurations and two Combustion Engineering intemals designs.

In Table 4-1, the comparisons of the adjusted results with the original calculations of the neutron flux (E > 1.0 MeV) are provided as [adjusted]/[calculated] (AIC) ratios for each of the 104 surveillance capsule dosimetry sets included in the data base. Also included in the tabulation are average A/C values for the 29 individual reactors and for the data base as a whole. From the data listed in Table 4-1, it is noted that the overall data base average A/C is 0.99 with an associated standard deviation of 7%. This data shows that the stand alone transport calculations are essentially unbiased. Further, the 7% standard deviation associated with the A/C data base is approximately half of the 13% uncertainty assigned to the calculation alone. It is evident fiom this data base that lack of knowledge of dosimeter positioning does not introduce large scatter and correspondingly high uncertainty into the comparisons of dosimetry results with calculations.

Although the data comparisons listed in Table 4-1 compare the results of the least squares adjustment with the original calculated neutron flux (E > 1.0 MeV), similar conclusions regarding the effects of dosimetry positioning can be drawn from a comparison of

[measurement]/[calculation] (MIC) ratios for individual foil reactions. These comparison applicable to the 104 capsule data base are listed in Table 4-2 along with the previously discussed May 2006 WCAP- 16083-NP-A,Revision 0

A/C results fi-om the adjustment analysis. From Table 4-2, it is noted that none of the comparisons for the individual foil reactions show any excessive scatter that could be attributed to a large uncertainty in the dosimeter positioning.

It must be emphasized that a key factor in completing the comparisons of calculation and measurement is that, for all data points, both the transport calculations and the dosimetry evaluations must be done using the same methods, cross-sections, and basic nuclear data. Use of older published data based on different methods and assumptions will lead to distortions in the data base. This could be due to one or more of the following:

1- Changes in treatment of the core source (conservative vs best estimate) 2- Changes in transport cross-sections (ENDFB-TV vs ENDFA3-VI) 3- Changes in dosimetry cross-sections (ENDFA3-IV vs ENDFA3-VI) 4- Changes in dosimetry evaluation methods (spectrum averaged cross-sections vs least squares adjustment)

These artificial distortions could introduce data scatter that would lead to a misinterpretation of the validity of the data base.

Consider the following example for Reactor 15 from Table 4- 1:

The Reactor 15 data base consists of 5 surveillance capsule withdrawals during the 20 year interval between 1985 and 2004. The original documentation of each of those capsule analyses was based on the methodology and basic nuclear data accepted at the time of the evaluations.

These are summarized as follows:

1- Completed in 1985.

Spectrum averaged cross-section dosimetry evaluation.

Design basis power distribution, intended to be conservative.

ENDFA3-IV transport cross-sections (SAILOR).

ENDFA3-V dosimetry cross-sections.

2- Completed 1988.

Spectrum averaged cross-section dosimetry evaluation Plant specific power distribution, intended to be best estimate.

ENDFB-IV transport cross-sections (SAILOR).

Axial peaking based on core power distribution.

ENDFB-V dosimetry cross-sections.

3- Completed 1991.

Least squares dosimetry evaluation.

Plant specific power distribution, intended to be best estimate.

ENDFB-IV transport calculations (SAILOR).

Axial peaking based on core power distribution.

ENDFB-V dosimetry cross-sections.

May 2006 WCAP-16083-NP-A, Revision 0

4- Completed 1998.

Least squares dosimetry evaluation.

Plant specific power distribution, intended to be best estimate.

ENDFB-VI transport cross-sections (BUGLE-96).

Axial peaking based on core power distribution.

ENDFB-VI dosimetry cross-sections (SNLRML).

5- Completed 2004.

Least squares dosimetry evaluation.

Plant specific power distribution, intended to be best estimate.

ENDFB-VI transport cross-sections (BUGLE-96).

Axial peaking based on 3D synthesis.

ENDFB-VI dosimetry cross-sections (SNLRML).

The methodology used for the latest evaluations for Reactor 15 are identical to those described in Section 2.0 of this report.

The following tabulation provides a comparison of the M/C or A/C ratios taken from the original reports for each of the capsule withdrawals from Reactor 15 with the latest AIC comparisons based on the methodology described in Section 2.0.

Original Latest Capsule MJC or AJC AJC 1 0.86 0.93 2 1.03 0.94 3 1.11 0.96 4 0.98 0.98 5 0.95 0.95 Average 0.99 0.95

% std dev 9.4 2.0 This comparison shows that the use of consistent methodologies for calculation and dosimetry evaluation reduces the standard deviation in the data base by almost a factor of five. This, clearly demonstrates that changing methodologies can introduce significant scatter in the data base and highlights the importance of re-evaluating dosimetry from previously withdrawn capsules each time a new surveillance capsule from a given reactor is analyzed.

May 2006 WCAP-16083-NP-A, Revision 0

Table 4-1 Data Base Comparison for 104 In-Vessel Dosimetry Sets fi-om 29 Reactors AIC Ratio - Neutron Flux (E > 1.0 MeV)

Capsule Capsule Capsule Capsule Capsule Capsule Reactor 1 2 3 4 5 6 Average  % std 1 1.10 1.01 1.05 0.93 1.02 7.0 2 1.03 0.96 0.97 0.92 0.97 4.7 3 1.OO 0.89 0.86 0.92 8.0 4 1.03 0.94 1.05 1.03 1.01 4.9 5 1.08 0.99 0.94 1.OO 7.1 6 1-01 0.99 0.97 0.95 0.98 2.6 7 0.94 1.06 1.06 1.02 6.8 8 1.01 1.02 1.02 1.01 1.02 0.6 9 0.99 1.04 1.01 0.93 0.99 4.7 10 0.85 1.03 1.04 0.97 0.97 8.7 11 1.16 0.92 1.04 16.3 12 0.95 1.02 0.91 0.99 0.97 4.9 13 1.01 0.99 0.93 0.84 0.96 0.95 7.0 14 0.97 0.98 0.95 0.97 1.6 15 0.93 0.94 0.96 0.98 0.95 0.95 2.0 16 0.94 0.97 0.93 0.81 0.88 0.91 6.9 17 0.94 0.86 0.89 1.04 0.85 0.92 8.5 18 0.94 0.91 0.93 2.3 19 1.06 0.94 0.88 0.99 0.97 7.9 20 0.98 0.96 1.01 1.04 1.OO 3.5 21 1.07 1.02 1.14 1.04 1.07 4.9 22 1.04 1.07 1.06 2.0 23 0.99 1.01 1.OO 1.4 24 1.06 1.01 1.07 1.05 3.1 25 1.05 0.95 1.06 1.02 6.0 26 1.13 1.05 0.97 1.05 7.6 27 0.99 0.96 1.01 0.98 1.08 1.09 1.02 5.3 28 1.12 0.94 0.99 1.02 9.1 29 0.93 1.14 1.04 14.3 Average 0.99 7.0 May 2006 WCAP-16083-NP-A, Revision 0

Table 4-2 S m a r y Comparison for Individual Foil Reactions Data Base Average Reaction M/C or A/C  % std dev 63~u(n,a)60~o 1.08 7.2 54~e(n,p)54Mn 0.98 7.9 58~i(n,p)58~o 0.98 7.9 238~(n,p)~~ 1.01 11.4 U7~~(n,~)m 1.04 11.7 4(E > 1.0 MeV) 0.99 7.0 May 2006 ~ ~ A P - 1 6 0 8 3Revision

- ~ ~ ,0

5.0

SUMMARY

AM) CONCLUSIONS In this report, the use of the FERRET adjustment code for the least squares evaluation of LWR surveillance dosimetry has been described and benchmarked. The ability of the least squares procedure to combine calculations with available measurements to determine the best estimate spectrum with reduced uncertainties at the measurement locations has been demonstrated by benchmark comparisons in the NIST 2 3 5 and ~ 2 3 7 ~ pstandard fission fields, the PCA simulator benchmark, and the H. B. Robinson power reactor benchmark.

The importance of the key input parameters (Neutron spectrum, measured reaction rates, and dosimetry cross-sections) and their associated uncertainties have been discussed and the values used in the Westinghouse least squares procedure have been described and justified. The sensitivity of the adjustment procedure to the makeup of the foil set and the input uncertainties in the calculated spectrum and the measured reaction rates has been provided.

The conclusions from these studies are as follows:

1- The FERRET code operates as intended as a least squares adjustment code for reactor dosimetry analyses.

2- The input parameters (calculated neutron spectrum, measured reaction rates, and dosimetry reaction cross-sections) and their associated uncertainties are appropriate for use in LWR dosimetry evaluations.

3- The adjusted neutron flux (E > 1.0 MeV) from the FERRET evaluations can be used to develop a database for use in validating plant specific neutron transport calculations for LWR pressure vessels. This adjustment process provides a rigorous spectrum weighting of individual reaction rate MIC comparisons.

The comparison database developed from the least squares evaluation of the dosimetry sets can be used along with the guidance in Regulatory Guide 1.190 to validate neutron transport calculations and to determine any calculational biases that may be present.

M a y 2006 W C A P - ~ ~ O ~ ~ -Revision N P - A0,

6.0 REFERENCES

1- Code of Federal Regulations Title 10 Part 50, "Domestic Licensing of Production and Utilization Facilities," Appendix G, "Fracture Toughness Requirements" and Appendix H, "Reactor Vessel Material Surveillance Requirements,"

2- Code of Federal Regulations Title 10 Part 50.61, "Fracture Toughness Requirements for Protection Against Pressurized Thermal Shock Events,"

3- Regulatory Guide 1.190, "Calculational and Dosimetry Methods for Determining Pressure Vessel Neutron Fluence," U.S. Nuclear Regulatory Commission, OEce of Nuclear Regulatory Research, March 200 1.

4- ASTM Designation E944-02, "Standard Guide for Application of Neutron Spectrum Adjustment Methods in Reactor Surveillance," Annual Book of ASTM Standards, Section 12, Volume 12.02,2003.

5- RSIC Code Package PSR-145, "FERRET Least Squares Solution to Nuclear Data and Reactor Physics Problems," Radiation Shielding Information Center, Oak Ridge National Laboratory (ORNL), January 1980.

6- F. Schrnittroth, "FERRET Data Analysis Code," HEDL-TME-79-40, Hmford Engineering Development Laboratory, September 1979.

7- RSIC Code Package PSR-277, "LEPRICON PWR Pressure Vessel Surveillance Dosimetry Analysis System," Radiation Shielding Information Center, Oak Ridge National Laboratory (ORNL), June 1995.

8- RSIC Code Package PSR-113, "STAY'SL Least Squares Dosimetry Unfolding Code System," ,"Radiation Shielding Information Center, Oak Ridge National Laboratory (ORNL), December 1991.

9- RSIC Code Package PSR-233, "LSL-M2 Least Squares Logarithmic Adjustment of Neutron Spectra," Radiation Shielding Information Center, Oak Ridge National Laboratory (ORNL),

February 1991.

10- RSICC Computer Code Collection CCC-650, "DOORS 3.1, One- Two- and Three-Dimensional Discrete Ordinates NeutrodPhoton Transport Code System," Radiation Safety Information Computational Center, Oak Ridge National Laboratory (ORNL),

August 1996.

11- RSIC Data Library Collection DLC-185, "BUGLE-96, Coupled 47 Neutron, 20 Gamma Ray Group Cross-Section Library Derived from ENDF/B-VI for LWR Shielding and Pressure Vessel Dosimetry Applications," Radiation Shielding Information Center, Oak Ridge National Laboratory (ORNL), March 1996.

12- R. L. Simons, et. al, "Re-Evaluation of the Dosimetry for Reactor Pressure Vessel Surveillance Capsules," in NUREGICP-0029, "Proceedings o the Fourth ASTM-EURATOM Symposium on Reactor Dosimetry," F. B. K. Kam, Editor, July 1982.

May 2006 WCAP-16083-NP-A, Revision 0

13- W. N. McElroy, et. al, "LWR PV Surveillance Dosimetry Improvement Program: PCA Experiments and Blind Test," NUREGICR-1861, July 1981.

14- Regulatory Guide 1.99, Revision 2, "Radiation Embrittlement of Reactor Vessel Materials,"

U. S. Nuclear Regulatory Commission, Office of Nuclear Regulatory Research, May 1988.

15- G L. Guthrie, "Uncertainty Considerations in Development and Application of Charpy Trend Curve Formulas," in REACTOR DOSIMETRY Dosimetry Methods for Fuels, Cladding, and Structural Materials, Proceedings of the Fifth ASTM-Euratom Symposium on reactor Dosimetry, GKSS Research Centre, Geesthact, F. R. G, J. P Genthon and H Rottger, Editors, D. Reidel Publishing Company, 1985.

16- G L. Guthrie, "Charpy Trend Curve Development Based on PWR Surveillance Data," in NUREGICP-0048 Proceedings of the U. S. nuclear Regulatory Commission Eleventh Water Reactor Safety research Information Meeting, Volume 4 - Materials Engineering Research,"

U. S. Nuclear Regulatory Commission, Office of Nuclear Regulatory Research, January 1984.

17- G. L. Guthrie, "Charpy Trend Curves Based on 177 PWR Data Points," NUREGICR-3391, Volume 2, LWR-PV-SDIP Quarterly Progress Report, April 1983.

18- ASTM Designation El 0 18-01, "Standard Guide for Application of ASTM Evaluated Cross-Section Data File," Annual Book ofASTM Standards, Section 12, Volume 12.02,2003.

19- ASTM Designation E261-98, "Standard Practice for Determining Neutron Fluence, Fluence Rate, and Spectra by Radioactivation Techniques," Annual Book of ASTM Standards, Section 12, Volume 12.02,2003.

20- W. N. McElroy, et. al, "LWR Pressure Vessel Surveillance Dosimetry Improvement Program:

PCA Experiments, Blind Test, and Physics-Dosimetry Support for the PSF Experiments" NUREGICR-3318, September 1984.

21- I. Remec and F. B. K. Kam, "Pool Critical Assembly Pressure Vessel Facility Benchmark,"

NUREGICR-6454, July 1997.

22- A. K. McCracken, "Few Channel Unfolding in Shielding- The SENSAK Code," in Proceedings of the 3* ASTM-EURATOM Symposium on Reactor Dosimetry, Ispra, Italy, October 1-5, 1979, EUR 6813, Commission of the European Communities, Vol. 11, P. 732, 1980.

23- RSIC Data Library Collection DLC-178, "SNLRML Recommended Dosimetry Cross Section Compendium," Radiation Shielding Information Center, Oak Ridge National Laboratory, July 1994.

24- N. P. Kocherov, et. al, "International Reactor Dosimetry File (IRDF-go)," International Atomic Energy Agency, Nuclear data Section, IAEA-NDS-141, Rev. 0, August 1990.

25- I. Remec and F. B. K. Kam, "H. B. Robinson Pressure Vessel Benchmark,"

NUREGICR-6453 (ORNLRM-13204), February 1998.

May 2006 WCAP-16083-NP-A, Revision 0

26- J. D. Andrachek, et. al, "Methodology Used to Develop Cold Overpressure Mitigating System Setpoints and RCS Heatup and Cooldown Limit Curves," WCAP-14040, Revision 3, April 2002.

May 2006 WCAP- 16083-NP-A,Revision 0