ML17229A042

Rev 1 to CEN-405-NP, Application of Reactor Vessel Surveillance Data for Embrittlement Mgt.

ML17229A042
Person / Time
Site:	Saint Lucie
Issue date:	09/30/1996
From:	ABB COMBUSTION ENGINEERING NUCLEAR FUEL (FORMERLY, ASEA BROWN BOVERI, INC.
To:
Shared Package
ML17229A041	List: ML17229A040 ML17229A042
References
CEN-405-NP, CEN-405-NP-R01, CEN-405-NP-R1, L-96-233, NUDOCS 9609260162
Download: ML17229A042 (86)

Text

/

,'t. Lucie Units 1 and 2

~ Docket Nos. 50-335 and 50-389

'-96-233 Enclosure A COMBUSllON ENGINEERING OWNERS GROUP CEN-405-NP Revision 1 APPLICATION OF REACTOR VESSEL SURVEILLANCE DATA FOR KMBRITTLKMKNTMANAGEMENT Prepared for the C-E 07>'NKRS GROUP September 1996 qz<>qoso1s2 9so+o PDR ADO DR P llew, QDQD ABB Combustion Engineering Nuclear Operations liQODOD Copyright l996 Combustion Engineering, Inc. hll rights reserved

LEGAL NOTICE This report was prepared as an account of work sponsored by the Combustion Engineering Owners Group and ABB Combustion Engineering.

Neither Combustion Engineering, Inc. 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 infringe privately owned rights; or e 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.

Combustion Engineering, Inc.

FOREWORD The purpose of Revision 01 of CEN-405-NP is to include portions of that document that are no longer considered by Combustion Engineering, Inc., to contain proprietary information. Portions made non-proprietary include the post-irradiation surveillance data and analysis results for Combustion Engineering fabricated reactor vessel surveillance materials, application of initial reference temperature uncertainty, and application of the Regulatory Guide 1.99 Position 2.1 reduction factor for reference temperature shift uncertainty. This information has become publicly available through industry standards and regulations or been accepted for common usage since the CE Owners Group project was initiated.

Revision 01 contains no technical changes. Its predecessor was issued in July 1993 and did contain extensive technical changes prepared in response to questions raised by the NRC Staff.

WINDSOR, CONNECTICUT TABLE F E

~ec i ~nN Title ~Pe

SUMMARY

INTRODUCTION BACKGROUND IV. RATIONALE 12

v. INTEGRATED SURVEILLANCE APPROACH 20 VI. MARGIN REDUCTION APPROACH 29 VII. CONCLUSIONS 37 VIII. REFERENCES 39 APPENDIX A STATISTICAL ANALYSIS, PLATE DATA BASE AND RESULTS A-1 APPENDIX B STATISTICAL ANALYSIS, WELD DATA BASE AND RESULTS B-1 APPENDIX C EFFECT OF WELD FLUX LOT ON WELD CHEMICAL CONTENT C-1

LI T FFI R

+i~m~r T~il Decision Tree for Selection of Integrated Surveillance Approach (ISA) or Margin Reduction Approach (MRA)

Predicted versus Measured Weld RTND~ Shift Results 16 P

Trend Curve for A533B Reference Material 24 Margin Reduction Approach 36 BL Summary of Basic Statistics for Charpy Shifts Difference Sample Sets 19

i. MMARY The surveillance program for Combustion Engineering (C-E) designed reactor vessels provides for the monitoring of irradiation damage in accordance with 10CFR50, Appendix H."'any of those programs, however, do not include the controlling vessel material as presently defined using Regulatory Guide 1.99, Revision 02.<" One consequence is that data from an approved Appendix H program can not be used for evaluation of vessel integrity issues. More specifically, direct application of Regulatory Guide 1.99, Position 2.1 to refine embrittlement predictions is not available for many vessels.

This report presents two approaches for C-E owners to apply Regulatory Position 2.1 in the specific case where a subject plant's surveillance data are credible in all respects except that the controlling material in the vessel gs not one of the surveillance program materials. If the controlling material of one reactor vessel is located in the surveillance program of another vessel, the Integrated Surveillance Approach {ISA) may be applied. When the controlling material of a vessel can not be traced to any other vessel's surveillance program, the Integrated Surveillance Approach can not be used and the Margin Reduction Approach

{MRA) is applicable. Figure 1 gives a decision tree defining under what circumstances the ISA or MRA is to be used.

In the Integrated Surveillance Approach, controlling material data for the subject vessel is available from another C-E fabricated vessel (the "host" vessel) surveillance program {e.g.,

from a Westinghouse vessel). ['

Once the preceding have been established, the chemistry factor and margin are determined for the subject vessel following Regulatory Position 2.1.

In the Margin Reduction Approach, plant-specific surveillance data are used to reduce the margin to be added to the predicted shift.

Rationale is provided to support use of the two approaches. [

Figure 1 Dec i s i on Tree for Se ect i on I

of Integrated Survei lance Approach ( SA)

I I or Mar g i n Reduct i on Approach (MRA)

Limiting Material in Surveillance Capsule YES Fo I I ow Regul atory Position 2. 1 Approach NO L i rni t. i ng Mater i a i n I

Sister Vessel Survei lance Progr I am Follow ISA per YES CEN-405P, Section V NO Follow MRA per CEN-405P,Section VI

iL R D TI The reactor pressure vessel is designed such that its fracture toughness will be sufficient to provide adequate margins of safety against brittle fracture during its service life. Thus, the original construction employed thick section low alloy steel base and weld materials which were inherently tough, as characterized by the initial reference temperature, RT>>~.

Particular attention was given to the reactor vessel beltline, the region of the reactor vessel that surrounds the effective height of the active core. This region is exposed to a relatively high level of neutron irradiation which, over time, will reduce the toughness of (i.e.,

embrittle) the base and weld materials. Each operating plant is required to have a reactor vessel surveillance program which monitors those irradiation induced changes in the toughness properties of the beltline materials.

Neutron irradiation embrittlement of the reactor vessel beltline is addressed for both normal operation and for design base accidents. Heat-up and cool-down limits on pressure and temperature are adjusted to account for the predicted irradiation induced elevation of RT~Y.

Accident analyses ascertain that vessel integrity will be maintained in the event of a postulated transient, such as pressurized thermal shock <'", despite predicted embrittlement of the vessel. In either case, predictions of irradiation embrittlement are based on Regulatory Guide 1.99<". In situations where "credible" surveillance program results are available, those predictions for establishing operating limits can be adjusted based on the surveillance data.

This report addresses the specific case where a plant's surveillance data are credible in all respects except that the controlling material in the vessel is not one of the surveillance program materials. Two prescriptive approaches are developed in order to maximize the value from plant specific surveillance measurements. In the Integrated Surveillance Approach, the chemistry factor and margin is adjusted using surveillance data from another Combustion Engineering (CE) fabricated reactor vessel of CE or Westinghouse design. In the Margin Reduction-Approach, plant-specific surveillance data are used to reduce the margin to be added to the predicted shift. Rationale is provided to support use of Regulatory

Position 2.1 of Regulatory Guide 1.99tu in these specific cases. Supplemental criteria are provided for demonstrating the viability of each approach.

Surveillance programs were designed to provide a means of omni orio; irradiation behavior of reactor vessel beltline materials. The concept in the 1960's was to measure the extent of embrittlement to verify the original design estimates. Surveillance capsules contained monitors to measure peak temperature and neutron flux, and test specimens to measure changes in strength and toughness. The test specimens were from materials selected to

~r~re~en the beltline materials, where selection criteria reflected the then current understanding of radiation embrittlement trends.

Presently, surveillance program requirements are given by 10CFR50, Appendix H"'nd, by reference, ASTM E185-82~'. The stated purpose is still to monitor property changes with the addition that the resultant data are to be used in support of 10CFR50, Appendix G analysis. Regulatory Guide 1.99<'> provides a means for predicting RT>>Y shift based on the chemical content of the vessel material and the neutron fluence. The Guide also presents a method, Regulatory Position 2.1, by which credible surveillance data can be used to refine the shift prediction and to reduce the uncertainty factor (margin) which must be added to the mean predicted shift.

A dichotomy exists given surveillance programs designed in the 1960's and early 1970's and the current Regulatory Guide 1.99u'. The means by which surveillance materials were selected for encapsulation differs substantially from the method currently prescribed in the Guide. For example, for the Palisades reactor vessel, the beltline plates were compared on the basis of drop weight NDTT and the Charpy impact test results;"'he plate with the highest NDTT and the highest temperature at the 30 ft-lb Charpy impact value was selected for inclusion in the surveillance program. The assumption was that differences in the initial toughness would be retained after irradiation for plates purchased to the same specification.

It was not recognized at the time that small differences in residual chemistry content could

~

result in a large difference in irradiation sensitivity (NDTT shift). ~ If the same plates were evaluated using the current Regulatory Guide 1.99"', a more rigorous analysis would be

~

~

performed on the basis of RT>>Y and irradiation induced changes in the transition

temperature (shift) and upper shelf energy. The plate selected on this basis for the surveillance program would very likely be different from the one originally selected.

Therefore, the post-irradiation surveillance plate data would not be credible because it would not meet the first Regulatory Guide 1.99 credibility criterion "Materials in the capsules should be those judged most likely to be controlling with regard to radiation embrittlement according to the recommendations of this guide."'"

Hence, the dichotomy exists; the non-credible reactor vessel surveillance data cannot be used to compute the best-fit chemistry factor, and any further use of the surveillance data must be justified to the NRC even though the surveillance program complies with the version of ASTM E185 in effect at the time the program was designed.

This situation exists for many reactor vessels because of differences in methods to define initial toughness and to predict shift. RT~Y was used in the surveillance material selection process for those vessels built to the Summer 1972 Addenda to the ASME Boiler and Pressure Vessel Code.<"> For earlier vessels, ND'IT or a Charpy index temperature was used to differentiate initial toughness properties of the candidate materials. Shift predictions for material selection were based on ranges of copper content; separate trend curves were used for 0 to 0.10% Cu, 0.11 to 0.15% Cu, and greater than 0.15% Cu. Typically, the beltline plates from one vessel would all fall within one range of copper content and, therefore, be predicted to exhibit the same shift. The beltline welds would have a similar situation. In contrast, Regulatory Guide 1.99<" predictions are based on explicit values of copper and nickel content such that each beltline plate and weld would have a unique predicted shift. Given the major differences in methods used to select the surveillance material and present day criteria"'or identifying the controlling material, the probability of having the precise controlling material in the surveillance capsule is low. As a consequence, will severely limit th'e number of reactor vessels to which Regulatory Position present rules 2.1iu can be applied.

e

eactor vessels designed and built by Babcock & Wilcox encountered a problem with the surveillance program which prompted the establishment of an integrated surveillance program (see for example Reference 6). The problem involved both the forced removal of surveillance capsules from several reactor vessels and the need for continued surveillance of the beltline materials from those vessels. The B&W situation was the impetus for Section II.C of 10CFR50, Appendix H<'> which detailed requirements and acceptance criteria for establishing an integrated surveillance program. The basic approach entails irradiation of representative material in a host reactor for use by other reactors having similar design and operating features. Presumably, that data could also be used in accordance with Regulatory Position 2.1<" for Appendix G<'> analyses.

In the B&W situation, numerous vessels were fabricated using similar materials and processes, including those vessels made for use in Westinghouse designed PWRs. Similar to B&W, Combustion Engineering also fabricated vessels for Westinghouse. Therefore, surveillance materials from Westinghouse designed vessels represent a potential source of data on specific heats and types of vessel beltline materials. Establishment of an integrated surveillance program between two vessels supplied by the same fabricator could, in numerous cases, provide surveillance data on the controlling vessel material for one or both of those vessels.

In contrast to the B&W situation, where some vessels had to have the surveillance capsules removed, all of the C-E designed PWRs still have surveillance capsules, Therefore, measurements of neutron flux, irradiation temperature, and surveillance material irradiation sensitivity can be obtained for each C-E vessel. This provides for the monitoring required by Appendix HP and implementation of an integrated surveillance program could provide the data on the controlling material as input for analyses required by Appendix G<4> and 10CFR50.61."

'11-

tV.

The objective of this report is to establish two approaches for implementing Regulatory Position 2.1"'or C-E designed PWRs for which the surveillance program is credible in all respects except that the controlling material in the vessel is not in the surveillance program.

In the Integrated Surveillance Approach, data for the controlling material is obtained from another reactor vessel surveillance program. In the Margin Reduction Approach, the plant-specific data are used when data on the controlling material are unavailable. Both approaches are intended to address the Regulatory Position 2.1"'ase in which "...

surveillance data are credible in all respects except that the (surveillance) material does not represent the critical material in the vessel..." The primary purpose is to realize as fully as reasonable the benefits from the surveillance program once certain criteria have been satisfied. Those criteria include the surveillance data credibility factors given in Regulatory Guide 1.99"'nd, where

~ applicable, the criteria given in 10CFR50, Appendix EP'or an

~ ~

integrated surveillance program. The added certainty obtained through surveillance capsule

~

of the benefits from application of Regulatory Position

~

measurements justifies realization 2.1 <n The purpose of this section is to describe the rationale used in establishing the Integrated Surveillance and Margin Reduction Approaches.

'Note that the second and third rationale elements are directed primarily at the Integrated Surveillance Approach.) The three elements are detailed below.

FIGURE 2 PREDICTED VERSUS MEASURED WE TNDT SEEFT RESULTS 400 0 SURVEILLANCE DATA 350 EXPERIMENTAL DATA (WELD-1, LCP) 300 0

250

~r ra 200 a

Z 150 100 POSITIOH 2.1. RG 1.99 REV. 2 CURVE FIT TO SURVEILLANCE WELD DATA WITH 1rr = 28F BOUNDS 50 0.1 0.2 0.4 0.6 0.8 1.0 2.0 4.0 6.0 8.0 10.0-FLUENCE (E > 1 MeV). 10 n/cm Table 1

SUMMARY

OF BASIC STATISTICS FOR CHARPY SHIFT DIFFERENCE SAMPLE SETS Predicted Minus Measured Shift Sample Statistics (Degrees F)

~Sam ie Set Plates:

Number of Data Standard

~evi a tie Range

.,... e C-E 28 4.18 16.94 -32 34 Westinghouse 74 -0.82 19.09 36 Pooled 102 0.55 18.54 36 Welds:

C-E 14 7.57 21.40 Westinghouse 30 4.87 27.53 64 Pooled 5.73 25.79

V. I E RATED R EILLA E APPR The purpose of this section is to establish the criteria which need to be addressed and the procedure to be followed in order to utilize surveillance data from another reactor vessel in support of a Regulatory Position 2.1<" analysis. The approach combines the credibility criteria of Regulatory Guide 1.99"'ith the concept of integrated surveillance programs defined in 10CFR50, Appendix H."'rior to applying Regulatory Position 2.1 per this approach as shown in Section V.E., the following must be established: traceability of the controlling material (Section V.A.), the credibility of the subject vessel (Section V.B.) and host vessel surveillance data (Section V.C.), and the similarity of the irradiation environment of both vessels (Section V.D.).

A. ~M'I  % -A '

h 'di h h<<

program is equivalent to the controlling material in the subject vessel.

ll

V 1 rv il e D r I

'l' The surveillance data from the subject reactor vessel must be credible in all respects except that the surveillance material does not represent the critical material in the vessel. This is established by satisfying the following five criteria which are taken from Reference 1 with modifications to items 1 and 5:

1. Materials in the capsule shall be representative of the. reactor vessel beltline materials, including both base metal and weld metaL[

2. Determination of the 30 ft-lb index temperature and the upper-shelf energy shall be done unambiguously for both the irradiated and unirradiated Charpy data.<'>

3. Two or more sets of post-irradiation surveillance data for both base and weld metal shall be available from the subject reactor vessel, and a Regulatory Position 2.1 analysis shall be performed. The measurements shall be within J lo~ of the mean curve of the actual surveillance data, where a~ is 17'F for base metal and 28'F for weld metal."'f the fluence range is two orders of magnitude or greater, the measurements must be within +2o,.<"

4 The irradiation temperature of the Charpy specimens shall be within J25'F of the vessel wall temperature at the cladding/base metal surface.'" The Charpy specimen temperature shall be estimated based on evaluation of the temperature monitors included in the surveillance capsules or from heat transfer calculations. The vessel wall temperature shall be based on cold leg or vessel wall measurements.

5. One of the surveillance capsules used in the evaluation should include Charpy specimens from a standard reference material. (CEOG vessel surveillance programs use HSST01 for reference material.) The measured shift for the standard reference material shall fall within the scatter band (+2ag of the data base for that material as given in Figure 3.""

C. H st Reac r urveillan D redibilit - The principal interest in the host reactor data is the one surveillance plate or weld representing the subject reactor's controlling material. The following criteria are based on the five criteria from Reference 1, with modifications to items 1, 4 and 5. Each of the following criteria are to be satisfied to establish the credibility of data for that one material:

1. The traceability between the controlling material from the subject reactor and surveillance material from the host reactor is to have been established as detailed in V.A.

2. Determination of the 30 ft-lb index temperature and the upper-shelf energy shall be done unambiguously for both the irradiated and unirradiated Charpy data.<'>

3. Two or more sets of post-irradiation surveillance data for the controlling material shall be available from the host reactor vessel, and a Regulatory Position 2.1 analysis shall be performed. The measurements shall be within Jla~ of the mean curve of the actual surveillance data, where o~ is 17'F for base metal and 28'F for weld metal."'f the fluence range is two orders of magnitude or greater, the measurements must be within

+2o~."'he 4, irradiation temperature of the Charpy specimens shall be within +25'F of the vessel wall temperature at the cladding/base metal surface."'he Charpy specimen temperature shall be estimated based on evaluation of the temperature monitors included in the surveillance capsules ifavailable. The vessel wall temperature shall be based on cold leg or vessel wall measurements.

5. Standard reference material (SRM) is not available in some host reactor vessel surveillance programs.

However, ifSRM is available for one or more host reactor capsules, the measured shift for the SRM shall fall within the scatter band (+2'f the database for that material as given in Figure 3.""

iZ5 ZOO v v I ioo v

CI i50 v

M 4. 75 Ol m

Cfl a

ioo / till ~ a 50 a

I o

50  % Survolllanco Capouloo HSSTO?.

0 Survolllanco Capaulos HSSTOi Z5 h BSR Irradlat tons (HSSTO?)

V ORR-PSF Irradiat lone (HSST03) 0.00 0.50 i.00 i.50 Z.OO Z.50 "'.00 3.50 4.00 4.50 5.00HE19 Fluanco (6 > i.O Hav)

FIGURE 3 TREND CURVE FOR A533B REFERENCE MATERIAL Embrittlement of the A533B reference material relative to the draft Reg. Guide 1.99, Revision 2. The values for HSST01 and HSST03 plates are adjusted relative to HSST02 plate to account for differences in chemistry. The upper and lower curves are the 34'F uncertainty bounds (20) specified by Reg. Guide 1.99.

(Source: Reference 15)

D. I i i n nvir nm n m ri - Ascertain that the irradiation environment for the host reactor surveillance capsule is comparable to that for the subject reactor surveillance capsule using the following factors and provide a qualitative ranking of the two capsules in terms of the significance of any differences on RT>>r shift.

1. Reactor Coolant Inlet Temperature 7

2. Neutron Flux E. A i i n fR 1 P i i n 2 l- Once the four preceding items have been satisfactorily addressed, the host reactor'surveillance data for the controlling material may be used to determine the adjusted reference temperature for that material in accordance with Regulatory Position 2.1"'s follows:

The overall equivalence of the subject reactor vessel's controlling material and the host reactor vessel's surveillance material was established in Section V.A.

If the reported copper and nickel content of both materials is identical, then proceed to step 2. If, however, they are not identical, then the measured values of shift, AT>>>, shall be adjusted by multiplying them by the ratio of the chemistry factor for the vessel material to that of the surveillance material in accordance with Regulatory Position 2.1.

Fit the surveillance data to obtain the relationship of d,RT>>T to fluence using the following equation:

QgZ NDT (gp) f (0.28-0.10 2+gE) where:

CF = chemistry factor f = neutron fluence(10" n/cm2, E ) 1 Mev)

To do so, calculate the chemistry factor, CF, for the be"t fit by multiplying each dRT>>~ (or the adjusted values from step 1) by its corresponding fluence factor, summing the products, and dividing by the sum of the squares of the fluence factors. The resultant value of CF is then to be entered into equation 1 for calculating d,RT>>T.

Note: If the host reactor surveillance data are less than predicted, but the subject reactor surveillance data are greater than predicted, then application of the CF derived from the host reactor could be non-conservative for the subject reactor. In this situation, the host reactor calculated chemistry factor can be adjusted as follows:

CF(H) x QF j~) = Adjusted CF CF (S,P)

where:

CF(H) is the calculated CF for the host reactor from step 2 CF(S,C) is the calculated CF for the corresponding plate or weld for the subject reactor following step 2 CF(S,P) is the, predicted CF for the same subject reactor material based on Table 1 or 2 of Reference 1 II A

The larger of the two values, Adjusted CF or CF(H) should then be entered into equation 1.

3. Calculate margin as follows:

Margin = 2 a~+a<2 e

where:

o; = standard deviation for the initial RTNDY a~ = standard deviation for hRTNDY In cases where the initial RTND~ has been determined for the subject material in accordance with the ASME Boiler and Pressure Vessel Code"", o; = 0. In cases where a generic value of -56'F is assumed for submerged arc welds fabricated using Linde 0091, 1092, or 124 flux, o; = 17'F."~ (In cases where a generic value is assumed for SA 533B Class 1 or SA 302B (Modified) plates, o; is to be determined based on the variance of the data used to derive the generic value.)

0

The standard deviation for shift, ocan be reduced to 8.5'F for base metal .

and 14'F for welds if the surveillance data credibility has been established in accordance with sections V.B and V.C per Regulatory Position 2.1tu and the irradiation environments have been established to be comparable in accordance with Section V.D.

4. Calculate adjusted RTND~(ART) as follows:

ART = RT~ + d,RT~~ + Margin where:

RTNDY initial RT>>~, measured or generic value d,RTNDg = shift, using revised chemistry factor in equation 1 Margin = value using equation 2 including values of r, and 0~ from step 3

5. Document results of the evaluation following Section V.A through V. E, including assumptions employed and conclusions reached.

VI. MAR I RED TI APPR A H The previous section addressed the special case in which an integrated surveillance program approach could be used to augment plant-specific surveillance data. The purpose of this section is to address the case in which surveillance data on the controlling material is not available from either the plant-specific program or another reactor vessel, but the surveillance data are credible in all other respects. An approach is given for determining how much the standard deviation for ART>>~, 0~, can be reduced based on the degree of credibility of the surveillance capsule measurements or, as phrased in Regulatory Position 2.1<'>, "depending on where the measured values fall relative to the mean calculated for the surveillance materials".

The following procedure presents a recommended approach for satisfying Regulatory Position 2.1<" in order to reduce the value of 0~.

The procedure which follows addresses two issues, representative materials and predictability of the surveillance measurements, as the basis for reducing the value of 0~.

'

A. v Each surveillance program for a C-E designed reactor vessel includes encapsulated Charpy specimens from a plate, weld, heat-affected-zone, and standard reference material (SRM). In accordance with the edition of ASTM E185 in effect at the time, the surveillance materials (exclusive of SRM) were selected to represent the reactor vessel beltline. The surveillance plate was selected from one of the stx (typical) betttine plates. [

The surveillance weld was fabricated using portions of two beltline plates following the same procedure and weld consumables as one of the beltline welds .

In summary

~dii - gf Gdfd*999"'9

• ff* f jdg g credibility of surveillance data. The four Regulatory Guide criteria dealing with

1) scatter of the Charpy data, 2) scatter about the hRTND~ versus neutron fluence best fit curve, 3) capsule irradiation temperature, and 4) SRM test results shall be evaluated. Once those criteria are satisfied data predictability needs to be evaluated to provide the basis for reducing the value of 0~.

The surveillance data to be evaluated are the measured shifts for the base material and weld. These measurements are to be compared to the mean predicted shifts based on Regulatory Position 1.1'", Equation 2. The value of 0~ is then reduced by an amount "depending on where the measured values fell relative to the mean calculated (shift) for the surveillance materials" in accordance with Regulatory Position 2.1tu The rationale to be used in this comparison is that if surveillance measurements consistently fall at or below the quantity (Mean Predicted Shift + 0~/2), then the Regulatory Guide 1.99 shift prediction technique is accurate or conservative, for that reactor vessel's beltline material. Therefore, the shifts predicted for the controlling vessel material will also be accurate or conservative such that the value of 0~ can be reduced to a limit of one half the prescribed value. If measurements are above the quantity (mean predicted shift + 0~/2), then the amount by which 0~ can be reduced will depend on the extent of this variation. The criteria to be used for determining the amount by which 0~ may be reduced for the controlling material are as follows:

criteria

1) M-P ~ 0.50~ for all measurements 0.5

2) M-P ~ 0.60~ for all measurements 0.6

3) M-P ~ 0.70~ for all measurements 0.7

4) M-P < 0.8r~ for all measurements 0.8

5) M-P ~ 0.90~ for all measurements 0.9 where:

P predicted shift based on Equation 2 of Regulatory Guide 1.99

measured shift for base metal (longitudinal or transverse orientation) and weld.

a~ = standard deviation for d,RTNDY which is 17 F for plates and 28'F for welds"'31-

It should be noted that for all situations in which predictions exceed measurements (i.e., M-P < 0), no additional credit for o~ reduction is provided; that is, the reduction factor is not to be less than 0.5.

0 The margin reduction approach is shown graphically in Figure 4. Iffor each of the surveillance plate and weld measurements the M-P value is equal to or less than 0.5 o~ (the cross-hatched region in Figuie'4), the reduction factor to be applied for the o~

of the controlling material would be 0.5. Smaller reductions of r~ would apply ifone or more of the surveillance M-P values fell above 0.5 o~ up to a limit of 0.9 r~ as illustrated by the dashed lines in Figure 4.

If the most restrictive measurement (largest M-P) differs substantially from the other measurements, a supplemental evaluation must be performed before that single measurement is disregarded. A supplemental evaluation is permitted only if the datum meets the following criteria:

M-P does not exceed 1 o~.

2. It is from the lowest fluence capsule.

3. It is a different product form from the controlling vessel material.

4. M-P for all other measurements is less than 0.5 n~.

C. A licati n f Re ul P i i n 1.1 with Mar in R c ion - Once the two preceding items have been satisfactorily addressed, the known information on the controlling material may be used to determine the adjusted reference temperature for that material in accordance with Regulatory Position 1.1"'s follows:

Obtain the value of hRTND~ based on the chemistry factor of the controlling material and the fluence using the following'equation:

(gg f028-0.10 log/)

where:

CF = chemistry factor f = neutron fluence (10" n/cm', E ) 1 Mev)

To do so, calculate the chemistry factor, CF, for the controlling material using the known chemical composition as prescribed in Regulatory Position 1.1 of Reference 1. If the exact chemical composition is not known, use the guidelines of Reference 1 to determine the appropriate chemistry factor.

2. Calculate margin as follows:

Margin = 2 o> + (RFaz) (2) where:

0; = standard deviation for theinitialRTND~

0'~ = standard deviation for ARTNDT RF = reduction factor from Section VI.B and as described below In cases where the I

initial RTNDY has been determined for the controlling material in accordance with the ASME Boiler and Pressure Vessel Code"", 0;

= 0. In cases where a generic value of -56'F is assumed for submerged arc weld fabricated using Linde 0091, 1092, or 124 flux, e, = 17'F."~ (In cases where a generic value is assumed for SA 533B Class 1 or SA 302B (Modified) plates, 0; is to be determined based on the variance of the data used to derive the generic value.)

The standard deviation in BRTNDY, 0~, as defined in Regulatory Guide 1.99 is:

0~ = 28'F for welds and 0~ = 17'F for base metal. The Margin Reduction Approach may then be applied to the controlling material based on the results of testing of the available credible surveillance data to reduce the required value of 0~ as follows:

grit~eri Reducti n Factor RF

1) M-P ~ 0.50~ for all measurements 0.5

2) M-P ~ 0.6~~ for all measurements 0.6

3) M-P ~ 0.70~ for all measurements 0.7

4) M-P ~ 0.8r~ for all measurements 0.8

5) M-P ~ 0.9o~ for all measurements 0.9 where:

P = predicted shift based on Equation 2 of Regulatory Guide 1.99"'

= measured shift for base metal longitudinal or transverse orientation) and weld 0~ = standard deviation for dRTND~ which is 17 F for plates and 28'F for welds"'

3. Calculate adjusted RT>>Y(ART) as follows:

.ART = RTD+ hRTND+ Margin where:

RTm~ initial RTNDr, measured or generic value ARTND7 shift, using equation 1 Margin value using equation 2 including values of 0; and 0~ from step 2

-Yh I fhp Rig' bd&.N should include demonstration that the surveillance materials are representative of the reactor vessel beltline materials as described in Section VI.A, The surveillance data credibility shall be established using data sets from two or more surveillance capsules and the four Regulatory Guide 1.99 criteria identified in VI.B. Finally, the surveillance material predictability and the resultant 0~ reduction factor shall be established following the procedure given in Section VI.C.

FrGURE 4 MARCax aEDUCnON APPROACH 0 b~ y~~ O~

@ ~ 6 o o o o o

// /i //

Q Q Q Q Q 1.0 0.0 0.0 0.V

OA PREDICTED SHIFT (P)

Advancements in technology for predicting the extent of neutron irradiation embrittlement in pressure vessel steels have caused significant changes in the definition of the "controlling" vessel beltline material. One consequence is the surveillance capsules do not contain that controlling material. Regulatory Position 2.1 of Regulatory Guide 1.99'" permits the use of data from the reactor vessel surveillance program to refine shift predictions but only for.

those programs which include the controlling vessel material.

This report provides the justification for employing surveillance data in support of vessel embrittlement analyses when the surveillance data are credible in all respects except the controlling material is not one of the surveillance program materials. Approaches are provided for Integrated Surveillance and Margin Reduction to maximize the value gained from plant-specific surveillance measurements.

In the Integrated Surveillance Approach, data for the controlling material from a host reactor vessel surveillance program is used in conjunction with plant-specific data to adjust the chemistry factor and reduce the required standard deviation for shift. This data sharing between Westinghouse and Combustion Engineering reactor vessel surveillance programs is justified based on the following rationale:

The Integrated Surveillance Approach identifies criteria to be addressed and a procedure to follow in order to utilize the host reactor surveillance data in support of a Regulatory Position 2.1<u analysis for a C-E vessel.

In the Margin Reduction Approach, the plant-specific surveillance data are used to reduce the margin for predicted shift of the controlling material based on the predictability of the surveillance measurements.

The two approaches presented in this report are applicable to surveillance programs for C-E designed reactor vessels. Their use is not intended for other vessel designs and surveillance programs because the approaches were based in part on practices and design characteristics unique to C-E.

RIII . R~RRRNN

1. USNRC Regulatory Guide 1.99, Revision 2, "Radiation Embrittlement of Reactor Vessel Materials," May 1988.

2. American Society for Testing and Materials, "Standard Practice for Conducting Surveillance Tests for Light-Water Cooled Nuclear Power Reactor Vessels" ASTM E185-82, July 1982.

3. 10CFR Part 50, Appendix H, "Reactor Vessel Materials Surveillance Program Requirements," Federal Register, Vol. 48, No. 104, May 27, 1983.

4. 10CFR Part 50, Appendix G, "Fracture Toughness Requirements," Ibid.

"Summary Report on Manufacture of Test Specimens and Assembly of Capsules for Irradiation Surveillance of Palisades Reactor Vessel Materials," Combustion Engineering Report P-NLM-019, April 1971.

"Babcock & Wilcox Owners Group Program for Evaluation of Reactor Vessel Properties," BAW-1474, Rev. 4, December 1986.

7. American Society for Testing and Materials, "Standard Practice for Characterizing Neutron Exposures in Ferritic Steels in Terms of Displacements per Atom (DPA),"

ASTM E693-79, August 1979.

W. N. McElroy, "LWR Pressure Vessel Surveillance Dosimetry Improvement Program: LWR Power Reactor Surveillance Physic-Dosimetry Data Base Compendium," NUREG/CR-3319, prepared by Hanford Engineering Development Laboratory, HEDL-TME 85-3, August 1985.

9. R. K. Nanstad, et.al, "Effects of 50'C Surveillance and Test Reactor Irradiations on Ferritic Pressure Vessel Steel Embrittlement," presented at the 14th International ASTM Symposium on the Effects of Radiation on Materials, Andover, Massachusetts, June 1988.

10. "Evaluation of the First Maine Yankee Accelerated Surveillance Capsule," Effects Technology, Inc., Report CR75-317, August 15, 1975.

11. "Maine Yankee Nuclear Plant Reactor Pressure Vessel Surveillance Program-Capsule 263," Battelle Columbus Laboratories Report BCL-385-21, December 12, 1980.

12. "Analysis of the Maine Yankee Reactor Vessel Second Accelerated Surveillance Capsule," Westinghouse Report WCAP-9875, March 1981.

13. J. R. Hawthorne, J. J. Koziol, and S. T. Byrne, "Evaluation of Commercial Production A533-B Steel Plates and Weld Deposits with Extra-Low Copper Content for Radiation Resistance," NRL Report 8136, October 21, 1977.

14. J. R. Hawthorne, "Notch Ductility Degradation of Low Alloy Steels with Low-to-Intermediate Neutron Fluence Exposures," NRL Report 8357 (NUREG/CR-1053),

January 14, 1980.

15. F. W. Stallman, "Analysis of the SA302B and SA533B Standard Reference Materials in Surveillance Capsules of Commercial Power Reactors," NUREG/CR-4947, prepared by Oak Ridge National Laboratory, ORNL/TM-10459, January 1988.

16. American Society of Mechanical Engineering, Section Components," of ME B il r n Pr r V 1, III, "Nuclear Power New York.

Plant

"Evaluation of Pressurized Thermal Shock Effects Due to Small Break LOCA's with Loss of Feedwater for the Combustion Engineering NSSS," Combustion Engineering Report CEN-189, December 1981.

18. F. W. Stallman, et.al., "PR-EDB: Power Reactor Embrittlement Data Base, Version 1," NUREG/CR-4816, prepared by Oak Ridge National Laboratory, ORNL/TM-10328, June 1990.

19. 10CFR Part 50, Section 50.61, "Fracture Toughness Requirements for Protection Against Pressurized Thermal Shock Events," Federal Register, Vol. 56, No. 94, May 15, 1991.

0 APPENDIX A STATISTICAL ANALYSIS PLATE DATA BASE AND RESULTS

APPENDIX A A.1 DATA BASE The reactor vessel surveillance plqte data base used in the statistical analysis is listed in Tables A-1 and A-2 for Combustion Engineering and Westinghouse NSSSs, respectively.

The data base is limited to vessel materials originating in the C-E Chattanooga facility and to correlation monitor materials included in the capsules.

The "Material Heat ID" refers to the designation in the ORNL Engineering Data Base (see ORNL Report NUREG/CR-4816, June 1990). The reported. chemical content, specimen orientation, neutron fluence and measured shift (BCv30) in Table A-1 are the values judged to be most representative of the surveillance materials. The data in Table A-2 are the as--

reported values from ORNL-EDB or the individual post-irradiation evaluation report; one exception is that A302B data were assigned 'a nickel content of 0.20% if no value was reported (see Table A-3). For both tables, the listed neutron fluence reflects, in order of preference, the recomputed values from NUREG/CR-3319 (August 1985), the most recent computed value, or the value given in the original capsule report. Values given in parentheses are the chemistry, fluence or shift reported on ORNL-EDB which were changed as described above.

A.2 STATISTICAL ANALYSIS RESULTS The results of the statistical analysis were summarized in Table 1 of the report. That analysis was performed using "predicted minus actual" (P-A) shifts for the two sets (i.e.,

Tables A-1 and A-2) using Regulatory Guide 1.99, Revision 2 to predict shift. The following presents these results graphically.

Figures Al, A2, A3 and A4 are plots of P-A versus copper content, nickel content, fast neutron fluence and flux. The plotting symbols differentiate C-E from Westinghouse NSSS data. There is no apparent bias for the four variables for the pooled results or for the vendor-specific data. The majority of the outliers are for nickel-alloyed plates (A533B and A302B-Modified) (Figure A2) and for fluences below 10"n/cm'Figure A3).

- A Figures A5 and A6 present the data in frequency histograms for C-E and Westinghouse data, respectively. Figure A7 is a histogram of the pooled data. Normal probability density

. functions for the two data sets are compared in Figure A8, and the pooled results are shown in Figure A9. The Westinghouse,.data center on P-A = -0.82, and the C-E data center on P-A = 4.18 as shown in Figure A8. The pooled. data center on P-A = 0.55 as shown in, Figure A9.

Several tests were performed to determine whether the C-E and Westinghouse data sets were comparable:

a) One-Way Analysis of Variance - For a ratio of variances F = 1.480, variability between groups is a random occurrence at the 0.2267 significance level.

b) Two-Sample Analysis - For a "t" statistic = 1.216, the difference between group central values (means) is a random occurrence at the 0.2267 significance level.

c) Kruskal-Wallis Analysis - For a test statistic = 1.056, the difference between central values is a random occurrence at the 0.3041 significance level.

d) Kolmogorov-Smirnov Two Sample Test -'or an estimated overall statistic DN =

0.1631, the combined difference. between central values and variances is a random occurrence at the 0.9996 significance level.

In summary, the individual C-E 'and Westinghouse plate surveillance data sets are not sufficiently different to indicate a significant difference in response to fast neutron irradiation. Therefore, surveillance plate data from a Westinghouse designed NSSS will, on average, be representative of surveillance plate data from a C-E NSSS, and vice-versa, for reactor vessels manufactured in the C-E Chattanooga facility.

The latter two tests (c and d) were performed as an overcheck on the more traditional F and t tests.

e

- A REACTOR VESSEL PLATE SURVEILLANCE DATA COMBUSTION ENGINEERING NSSS Reactor Vessel a sule Identit Arkansas Nuclear One Unit 2 Material Heat iD Spec.

Qrien ~Cu ~o Chemistry

~Ni %

Neutron Fluence

~lD" I Neutron

~lO' Flux

'PShift P-A W-97 PAN201 LT .08 .60 3.41 6.39 21 15 W-97 PAN201 TL .08 .60 3.41 6.24 50 ~ -14

. Calvert Cliffs Unit 1 W-263 PCC103 LT . 12 .64 6.00 6.47 60 12 W-263 SHSS01 LT .18 .66 5.90 6.36 88 29 Calvert Cliffs Unit 2 W-263 PCC202 LT .14 .66 8.06 5.58 84 12 W-263 SHSS01 LT .18 .66 8.14 5.64 128 1 Fort Calhoun W-225 PFC101 LT .10 .48 5.83 (4.5) 5.60 60 -5 W-225 SHSS01 LT .18 (NA) .66 (N/A) 5.83 (4.8) 5.60 124 -8 W-265 PFC101 LT .10 .48 8.30 4.78 74 -12 W-265 PFC101 TL .10 .48 8.70 5.01 70 -8

Table A-1 (continued)

Reactor Vessel a ule Identit Millstone Unit 2 Material Heat ID Spec.

Qrien ~u~~o Chemistry

~Ni ~o Neutron Fluence

~IO "n/cm Neutron

~IO" Flux I B Shift W-97 PML201 LT .14 .61 3.75 70 3 W-97 PML201 TL .14 .61 3.67 3.87 96 -24 Maine Yankee A-25 PMY01 LT .15 .59 17.60 (13.0) 43.0 120 7 A-25 SHSS01 LT .18 (NA) .66 (NA) 17.60 (13.0) 43.0 150 8 A-35 PMY01 LT .15 .59 77.30 (83.9) 61.4 185 -22 A-35 PMY01 TL .15 .59 77.30 (85.2) 61.4 195 -32

. W-263 PMY01 LT .15 .59 5.67 (6.6) 4.70 97 -4 W-263 PMY01 TL .15 .59 5.67 (6.6) 4.70 93 0 Palisades A-240 PPAL01 LT .25 .53 60.60 (44.0) 62.0 205 34 A-240 PPAL01 TL .25 .53 60.60 (45.0) 62.0 205 34 W-290 PPAL01 LT .25 .53 11.00 7.01 175 -5 W-290 PPAL01 TL .25 .53 11.30 7.20 155 17 St. Lucie 1 W-97 PSL101 LT .15 .57 5.40 3.67 68 22 W-97 PSL101 TL .15 .57 5.40 (5.54)'.96 3.67 70 20 St. Lucie 2 W-83 PSL201 LT ~ 61 1.62 4.60 35 3 W-83 PSL201 TL .61 1.63 4.63 21 18 San Onofre Unit 2 W-97 PSO201 LT .10 .60 5.07 (5.54) 4.80 55 (51.1) -2 W-97 PS 0201 TL .10 .60 5.07 4.80 35 (44.5) 18

Table A-2 REACTOR VESSEL PLATE SURVEILLANCE DATA WESTINGHOUSE NSSS Neutron Neutron Reactor Vessel a sule Identit Material Heat ID Spec.

Qrien Chemistry

~Cu % ~Ni ~o Fluence

~10 "n/cm Flux

~lD" I PP.

Shift P-A Beaver Valley Unit 1 U PBV101 LT .20 .54 6.54 5.79 120 6 U PBV101 TL .20 .54 6.54 5.79 135 -9 V PBV101 LT .20 .54 2.91 (2.55) 7.92 130 ~ -35 V PBV101 TL .20 .54 2.91 (2.55) 7.92 140 -45 W PBV101 LT. .20 .54 9.49 5.11 150 -9 W PBV101 TL .20 .54 9.49 5.11 185 '44 D. C. Cook Unit 1 T PCK101 LT .14 .49 2.71 (1.8) 6.79 60 2 T PCK101 TL .14 .49 2.71 (1.8) 6.79 70 -8 T SHSS02 LT .14 ~ 68 2.71 (1.8) 6.79 60 6 Y PCK101 LT .14 .49 13.40 8.59 105 -1 Y PCK101 TL .14 .49 10.60 6.80 115 -17 Y SHSS02 LT .14 .68 12.00 7.69 110 -3 Callaway Unit 1 U PCL101 LT .07 .59 3.27 0 30 U PCL101 TL .07 .59 3.27 30 0

Table A-2 (continued)

Neutron Neutron Reactor Vessel a sule Identit Material Heat ID Spec.

Qrien ~u~o Chemistry

~Ni o Fluence

~ltl" I ~l" I Flux

'PP.

Shift Haddam Neck A SASTM LT .20 .18 3.16 (2.07) 6.04 85 -17 D SASTM LT .20 .18 22.20 6.68 140 -18 F SASTM LT .20 .18 6.06 (4.04) 7.92 80 6 H SASTM LT .20 .18 20.00 (17.9) 8.37 127 -8 Diablo Canyon Unit 1 S PDC103 LT .077 .46 2.98 7.51 0 (-2) 33

'S SSHS02 LT. .14 .68 2.98 7.51 66 2 Diablo Canyon Unit 2 U PDC201 LT ;15 .67 3.51 11.2 65 15 U PDC201 TL .15 .67 3.51 11.2 73 7 Farley Unit 1 U PFA101 LT .14 .55 16.50 17.3 115 -3 U PFA101 TL .14 .55 16.60 17.3 90 22 X PFA101 LT .14 .55 28.30 (28.0) 14.7 135 -10 X PFA101 TL .14 .55 28.30 (28.0) 14.7 105 20 Y PFA101 LT .14 .55 5.83 16.3 85 -2 Y PFA101 TL .14 .55 5.83 16.3 55 28 Farley Unit 2 U PFA201 LT .20 .60 5.61 16.2 103 22 U PFA201 TL .20 .60 5.61 16.2 133 -8 W PFA201 LT .20 .60 15.40 12.5 165 2

,

W PFA201 TL .20 .60 15.40 12.5 165 2 0

Table A-2 tinued)

Neutron Neutron Reactor Vessel Material Spec. Chemistry Fluence Flux Shift a ule Identit Heat ID Qrien ~u~o ~Ni ~o ~IO" / ~lD"hl ~oF H. B. Robinson Unit 2 S PHB201 LT .12 .20 (NA) 3.91 (3.69) 9.29 30 (0) 20 S PHB202 LT .10 .20 (NA) 3.91 (3.69) 9.29 20 (0) 23 S PHB203 LT .09 .20 (NA) 3.91 (3.69) 9.29 15 (0) 24 S SASTM LT .20 .18 3.91 (3.69) 9.29 70 4 T PHB203 LT .09 .20 (NA) 41.10 18.1 75 -3 T SASTM LT .20 .18 41.10 18.1 150 -14 V PHB202 LT .10 .20 (NA) 7.24 (4.51) 6.90 45 (0) 8 V SASTM LT. .20 .18 7.24 (4.51) 6.90 70 (0) 21 Indian Point Unit 2 Y PIP203 LT .14 .57 4.72 6.40 145 -67 OO I Y SASTM LT .20 .18 4.72 6.40 70 9 Indian Point Unit 3 T PIP301 LT .18 .50 3.23 (2.92) 7.67 89 -2 T PIP304 LT .24 .52 3.23 (2.92) 7.67 137 -26 T PIP304 TL .24 .52 3.23 (2.92) 7.67 118 -7 Y PIP304 TL .24 .52 8.05 8.15 150 1 Y SHSS02 LT .14 .68 8.05 8.15 140 -44 Z PIP303 LT .19 49 10.70 6.11 150 -17 Z PIP304 LT .24 ,52 10.70 6.11 170 -6 Z PIP304 TL .24 .52 10.70 6.11 155 9 Kewaunee R SHSS02 LT .14 .68 20.70 14.5 140 -18 V SHSS02 LT .14 .68 6.41 (5.59) 15.8 95 -6

Table A-2 (continued)

Neutron Neutron Reactor Vessel Material Spec. Chemistry Fluence Flux Shift a s le Identit Heat ID Orien ~Cu ~o N~i% kilo" I ~10'9/ ~F McGuire Unit 1 U PMC101 LT .087 .60 4.14 14.2 45 -3 U PMC101 TL .087 .60 4.14 14.2 50 -8 X PMC101 LT .087 .60 13.80 10.1 45 16 X PMC101 TL .087 .60 13.80 10. 1, 65 -4 Salem Unit 1 T PSA101 LT .22 .53 2.84 (2.56) 8.29 100 0 T PSA102 LT. .23. .54 2.84 (2.56) . 8.29 100 4 T PSA103 LT .22 .52 2.84 (2.56) 8.29 75 24 T SHSS02 LT .14 .68 2.84 (2.56) 8.29 60 7 Y PSA103 LT'T ;22 .52 8.91 8.33 110 36 Y SHSS02 .14 .68 8.91 8.33

'25

-26 Salem Unit 2 PSA201 LT .10 .61 2.56 6.99 50 -9 PSA201 . TL .10 .61 2.56 6.99 70 -29 San Onofre Unit 1 A PS 0103 LT .18 .20 (NA) 28.60 49.1 100 18 A SASTM LT .20 .18 28.60 49.1 120 8 D PSO101 LT .17 .20 (NA) 56.20 63.3 140 -15 D PS 0102 LT .18 .20 (NA) 56.20 63.3 110 21 D PS 0103 LT .18 .20 (NA) 56.20 . 63.3 130 1 D SASTM LT .20 .18 56.20 63.3 150 -8 F PSO102 LT .18 .20 (NA) 57.30 23.5 120 11 F SASTM LT .20 .18 57.30 23.5 130 13

Table A-2 ntinued)

Reactor Vessel a ule I enti Wolf Creek Unit 1 Material Heat ID Spec.

Qrien ~~~o Chemistry

~HI %

Neutron Fluence

~IO"n cm ~lo'l Flux

'c.

Neutron Shift P-A c'e U PWC101 LT .07 .62 3.39 12.0 30 U PWC101 TL .07 .62 3.39 12.0 25

Table A-3 A302 B PLATE NICKEL CONTENT mVmal eri1I nifi i n ikl no o Yankee Rowe Upper Shell Plate 0.21 Test Material YA9 0.19 N/A HSST Plate 0.18 Big Rock Surveillance Plate 0.18 H. B. Robinson W-9807-3 0.10 W-9807-5 0.10 W-9807-9 0.15 W-10201-1 0.11 W-10201-2 0.25 W-10201-3 0.08 W-10201-4 0.09 W-10201-5 0.12 W-10201-6 0.09 Mean Value: 0. 14% Ni Range: 0.08 to 0.25% Ni Standard Deviation: 0.055% Ni

(F RED-ACT) CVT SHRIFT us CU CON~ENT C-E 8- Mest'.inghouse Plates 48 28

++ + c:

0 0 0

0 0

0

-28

-48

-68

-88 Bo 86 8~1 8~ 16 8.2 8 ~ 26 Cu Cont ent'% )

C-E

-A + Mest,inghouse

FlGURE A-2 t.'PRED-ACT) CVT SHIFT us NX CONTENT C-E 8 Mast inghouse Pj.ates 1 I t

4e RB Gl Gl D 0+

Qp B

g 0

I C

I -ae

+

~

Q 0 Cl +

-4B

~

Q I

Z

-ee e.a B~ 4 e.e e.8 Ni. Content (%)

C-E

+ Masti.nghousa

'IGURE A'-3 (r ~ED-ACT) CVT SHRIFT ua FAST F~UENC~

C-E 8. 4leat inghuuae Platea 4e BB 28

-4e

-ee

-8B ae 1ee Faat Fluent=e (x18 18 n/c:rn "2)

C-E

+ Weatinghouae

-A FIGURE A-4 CPRED-ACT) CVT SHIFT us NEUTRON FLUX C-E a Westingt ouse Plates 4e

++

0 -----++t-----

ae 0 .----'++

0 0 0 l

(

8 '0 g 0

@I-+II-D I

C Q -ae

~

Q Q

-4e

~

Q Z

0 -ee

-ee 18 188 Flux (x1.8 "18/am" 2 sea) 0 C-E

+ Westinghouse

FIGURE A-5 (PRED-ACT) CUT SHIFT FOR C-E PLATES Fr equenc:g Hi~t.agr am

-4e -ae 28 4e CUT Shit't'., Dif'f'erence (Deg. F)

-A FIGURE A-6 (PRED-ACT) CVT SHIFT FOR MEST PLATES F l" mguenc:g Hi mt ug r mm

-ee -ee 4B 4B CVT Shit't Dif'f'er ence (Deg. F)

-A FIGURE A-7 (PRED-ACT3 CVT SHIFT FOR ALL PLATES Fr equencg Histogram 12 48 -2B 2B 4B ee CVT Shit't Dif'f'er ence (Deg. F )

-A FIGURE A-8 (PRED-ACT7 CVT SHIFT FOR PLATES C-E 8 4laat.. Nar mal Pr ab. Dan+i.t LI Fane (x e.eel) ae 3l

~

Q 0 I C I 5 I D I I

3l xa I

~

Q I J

~

Q I I

Xl I GJ I D

Q.

-wee -ee'ae zee CVT Shi f't'. Di:f'f'aranca (Dag F)

C-E

--. Maetinghuuee

FIGURE A-9 (PRED-ACT) CVT SHIFT FOR PLATES Normal Probability DensitLN Funct ion (x 8.881) 28

~

Q ill 16 Gl Cl 12

-1ee -68 -28 6e 188 CVT Shif't'. Dif'f'erence (Deg. F)

-A APPENDIX B STATISTICAL ANALYSIS WELD DATA BASE AND RESULTS

APPENDIX B B.1 DATA BASE The reactor vessel surveillance weld data base used in the statistical analysis is listed in Tables B-1 and B-2 for Combustion Engineering and Westinghouse NSSSs, respectively.

The data base is limited to vessel materials originating in the C-E Chattanooga facility.

The "Material Heat ID" refers to the designation in the ORNL Engineering Data Base (see ORNL Report NUREG/CR-4816, June 1990). The reported chemical content, neutron fluence, and measured shift (B,Cv30) in Table B-1 are'the values judged to be most representative of the surveillance materials. The data in Table B-2 are the as-reported values from ORNL-EDB or the individual post-irradiation evaluation report. For both tables, the listed neutron fluence reflects, in order of preference, the recomputed values from NUREG/CR-3319 (August 1985), the most recent computed value, or the value given in the

~

original surveillance capsule report. Values given in parentheses are the chemistry, fluence

~

or shift reported in ORNL-EDB which were changed as described above.

B.2 STATISTICAL ANALYSIS RESULTS The results of the statistical analysis were summarized in Table 1 of the report. That analysis was performed using "predicted minus actual" (P-A) shifts for the two sets (i.e.,

Tables B-1 and B-2) using Regulatory Guide 1.99, Revision 2 to predict shift. The following presents those results graphically.

Figures Bl, B2, B3 and B4 are plots of P-A versus copper content, nickel content, fast neutron fluence and flux. The plotting symbols differentiate C-E from Westinghouse NSSS data. There is a tendency for over-prediction of shift for low copper (less that 0.1%) welds in Figure B1, and for over-prediction of shift for all but four of the C-E surveillance weld data at all levels of copper content. From Figure B2 it can be seen that the over-prediction tendency also applies to the entire range of nickel contents of the C-E data. However, there is no apparent'trend with copper or nickel content for the pooled data to suggest the

- B Chemistry Factor from Regulatory Guide 1.99 is inaccurate. The trend with neutron fluence in Figure B3 is for the greatest scatter to occur below 10"n/cm, where the range of P-A is

-46'F to +53'F; above 10"n/cm, the range is -24'F to +64'F. (Note that the plate data in Figure A3 of Appendix A exhibited a similar trend.) This trend with fluence indicates that data sharing between vessels of different design would be most reliable above 10"n/cm .

There is no obvious trend with neutron flux in Figure B4 which suggests that one order of magnitude variation in flux is insignificant to'data sharing.

Figures BS and B6 present data in frequency histograms for C-E and Westinghouse data, respectively. Figure B7 is a histogram of the pooled data. Normal probability density functions for the two data sets are compared in Figure B8, and the pooled results are shown in Figure B9. The Westinghouse data center on P-A = 4.87, and the C-E data center on P-A = 7.57, as shown in Figure B8. The pooled data center on P-A = 5.73, as shown in Figure B9.

Several tests were performed to determine whether the C-E and Westinghouse data sets were comparable:

a) One-Way Analysis of Variance - For a ratio of variances F = 0.105, variability between groups is a random occurrence at the 0.7510 significance level.

b) Two-Sample Analysis - For a "t" statistic = 0.3240, the difference between group central values (means) is a random occurrence at the 0.7475 significance level.

c) Kruskal-Wallis Analysis - For a test statistic = 0.7343, the difference between central values is a random occurrence at the 0.3915 significance level.

d) Kolmogorov-Smirnov Two Sample Test - For an estimated overall statistic DN =

0.3476, the combined difference between central values and variances is a random occurrence at the 0.1989 significance level.

- B In summary, the individual C-E and Westinghouse weld surveillance data sets are not sufficiently different to indicate a significant difference in response to fast neutron irradiation. Therefore, surveillance weld data from a Westinghouse designed NSSS will, on average; be representative of surveillance weld data from a C-E NSSS, and vice-versa, for reactor vessels manufactured at the C-E Chattanooga facility.

The latter two tests (c and d) were performed as an overcheck on the more traditional F and t tests.

- B Table B-1 REACTOR VESSEL WELD SURVEILLANCE DATA COMBUSTION ENGINEERING NSSS Reactor Vessel a sule I en it Arkansas Nuclear One Unit 2 Material Heat ID Qi~o Chemistry Nii~o Neutron

~10'4n/cm'eutron Fluence

~l ~ "/Flux 'B.

ShiA W-97 WAN20 .04 .08 3.34 6.26 10 f2 Calvert Cliffs Unit 1 W-263 WCC101 .24 .18 6.10 6.58 59 44 Calvert Cliffs Unit 2 W-263 WCC201 .20 .04 7.97 5.52 69 15 Fort Calhoun W-225 WFC 101 .35 .60 5.83 (4.2) 5.60 205 (238) -25 W-265 WFC101 ..35 .60 8.00 4.61 221 -22 Millstone Unit 2 W-97 WML201 .30 .06 3.77 3.98 76 23 Maine Yankee A-25 WMY01 .36 .78 17.60 (13.0) 43.0 270 10 A-35 WMY01 .36 .78 77.30 (86.9) 61.4 345 13 W-263 WMY01 .36 .78 5.67 (6.80) 4.70 222 -18

Table B-1 (continu Neutron Neutron Reactor Vessel Material Chemistry Fluence Flux Shift P-A a le Identit ~Hea ID g~o ~Ni % ~l'~num~ Ella'Sl ' LPP ~OF Palisades A-240 WPAL101 .22 1.27 60.60 (46.0) 62.0 350 33 W-290 WPAL101 .22 1.27 10.30 6.56 290 -22 St. Lucie Unit 1 W-97 WSL101 .23 .11 5.30 3.60 74 16 St. Lucie Unit 2 W-83 WSL201 .05 .07 1.62 4.60 17 San Onofre Unit 2 W-97 WSO201 .03 .12 5.07 (5.54) 4.80 '15'(7.2) 10

Table B-2 REACTOR VESSEL SURVEILLANCE DATA WESTINGHOUSE NSSS Neutron Reactor Vessel Material Chemistry "n/cm'eutron Fluence Flux - Shift P-A a ule Iden i Heat ID QiJt~o Nii~o ~/0 ~O'Sl ~F Beaver Valley Unit U

V W

1 WBV101 WBV101 WBV101

.26

.26

.26

.62

.62

.62 6.54 2.91 (2.55) 9.49 5.79 7.92 5.11

'55 150 185 6

-29 D.C. Cook Unit 1 T WCK101 .27 .74 2.71 (1.8) .6.79 80 53 Y WCK101 .27 74 10.60 6.80 ~

200 9 Callaway Unit 1 U WCL101 .06 ~ 07 3.27 )2.1 70 -46 Haddam Neck A. WCTY01 .22 .046 3.16 (2.07) 6.04 95 -27 D WCTY01 .22 .046 22.20 6.68 110 10 Diablo Canyon Unit 1 WDC101 .21 .98 2.98 7.51 110 41 Diablo Canyon Unit 2 U WDC201 .22 ~ 83 3.51 11.2 174 -28

Table B-2 (continu Neutron Neutron Reactor Vessel Material Chemistry Fluence Flux Shift P-A a le Identit ~Hca ID g~u% ~Ni % ~l "c cm~ 10'/cm'c oF . ~oF Farley Unit 1 U WFA101 .14 .19 16.50 17.3 80 9 X WFA101 .14 .19 28.30 (28.0) 14.7 100 0 Y WFA101 .14 .19 5.83 16.3 80 -14 Farley Unit 2 U %FA201 .03 .90 5.61 16.2 10 24 W WFA201 .03 .90 15.40 12.5 10 36 H. B. Robinson Unit 2 T WHB201 .34 .66 41.10 18.1 . 285 11 V WHB201 .34 .66 7.24 (4:51) 6.90 '75 22 Indian Point Unit 2

%IP201 .23 1.06 5.89 7.99 195 14 Indian Point Unit 3 T %IP301 .15 1.02 3.23 (2.92) 7.67 143 -11 Y WIP301 .15 1.02 8.05 8.15 180 0 z WIP301 .15 1.02 10.70 6.11 220 -24 Kewaunee R %KWE01 .20 20.70 14.5 235 -9 V WKWE01 .20 6.41 (5.59) 15.8 175 -10

Table B-2 (continued)

Neutron Neutron Reactor Vessel Material Chemistry Flux Shift P-A a sule Identit Heat ID ~u~% ~Ni ~o Fluence (10 "n/cm'sec) ('F) ( F)

(10"n/cm~)

McGuire Unit 1 U WMC101 .21 .88 4.14 14.2 160 -1 X WMC101 .21 .88 13.80 10.1 165 64 Salem Unit 1 WSA101 .16 1.26 8.91 8.33 165 43

'p Salem Unit 2 WSA201 .23 .71 2.56 6.99 155 -37 San Onofre Unit 1 A WSO101 .19 .20 (NA) 28.60 (12.0) 49.1 80 48 F WSO101 .19 .20 (NA) 57.30 (51.4) 23.5 '45 -6 Wolf Creek Unit 1 U WWC101 .04 .09 3.39 12.0 20

CVT Shit't Dif'f'eI"enaa (Deg. F) 0 XI O ..-+ +-

+ 0; C3 0 I I

C l9 N'o Q 3

+"' 0 Q

0 L

W

+-++ m m

C 0

0 OO,'O D OO z z

FIGURE B-2 CVT SHIFT 'PRED-ACT)

'ua Nj: CONTENT C-E 8 Meatinghouaa Mazda 88 Be 48 28 0

+ 0

+ + +0

+

-++

-ae 0

~

Q 0 M

-48 U

-ee Be 2 8~4 e.e 8~ 6 le 2 1~4 Ni Content (%)

0 C-E r

-B + 4Jest.inghouse

FlGURE 8-3 (PRED-ACT) CVT SHIFT us FAST FLUENCE C-E a Westinghouse Walds 8B 8B 0

4B

-aB

-4B

-8B 1B 1BB Fast Fluence (xi,e 18 n/cm"2)

+ Westinghouse

-B FIGURE B-4 (PRED-ACT) CUT SH1FT ua NEUTRON FLUX C-E 8r Meatinghauae Meld+

88 6e 4e I0 5

I C

28 Gl

~

Q 0

-48 O

68 xe 188 Flux (xie "18 n/am 2 mac:)

o C-E

+ Mant.inghuuat

-B PIGVRE B-S (PRED-ACT) CVT SH1FT FOR C-E MELDS Fraciue'ncLI Histogram

-4e -2e ae 4e 6e

-B FIGURE B-6 (PRED-ACT) CUT SHIFT FOR MEST. MELDS Frequency Hiatagram

-6B -4B -2B 4e 6e 8B CVT Shif't Dif'f'erence (Deg. F)

FlGURE 8-7 (PRED-ACT) CUT SHIFT FOR ALL MELDS FrequencLj Histogram

-ee -4e -ae ae 4e 6e 8e

-B FIGURE 8-8 (PRED-~CT) SHRIFT FOR MELDS C-E 8 Meat. Normal Prob. Density Fnca

e. Ba ei 816 3l

~

Q Ul \

C Gl ..L....

e. 81a \

\

3l I I

I

( 1

~ Q I \

I

e. Bee "'""'I't I

I 0 I

\

\

1 I

I \

EL 8 '84 \~ '

I I

I I

a/

I

/

-14e -Se -4e 18 6e 11e 16e CUT Shi f't. Dil. l et"enae (Deg. F)

C-E 4lest inghouse

-B FIGURE B-9 (PRED-ACT) SH1FT FOR MELDS C

Normal Pr ogabilitLj Density Function

<x e.e.e1>

16 3l 12

~

I Q

L Gl Cl lU 0

4 CL

-13e -Be -3e 12e CUT Shif't Dif'f'erence (Deg. F)

APPENDIX C EFFECT OF WELD FLUX LOT ON WELD CHEMICAL CONTENT

APPENDIX C C. I INTRODUCTION C.2 WELD FLUX TYPE

- C C.3 ~D FLUX LOT

- C

]

- C TABLE C-1 Expected Effect of Flux Type on the As-Deposited Weld Properties Using the Same Weld Wire

TABLE C-2 As-Deposited Weld Chemistries as a Function of Flux Type

- C TABLE C-3 As-Deposited Weld Chemistries as a Function of Flux Lot (Weld Wire Heat ¹4P7869)

- C-7-