ML20207E984

Errata & Addendum 1 to Edwin I Hatch Nuclear Power Plant, Unit 1,Reactor Pressure Vessel Surveillance Matls Testing & Fracture Toughness Analysis

ML20207E984
Person / Time
Site:	Hatch
Issue date:	05/31/1986
From:	GENERAL ELECTRIC CO.
To:
Shared Package
ML20207E982	List: ML20207E978 ML20207E984
References
NEDC-30997-ADD, NEDC-30997-ADD-01, NEDC-30997-ADD-1, TAC-60542, NUDOCS 8701050323
Download: ML20207E984 (17)

Text

.t .

. NUCLEAR ENER1Y BUSINESS OPER ATIONS o C ENERAL ELECTRtc COMPANY SAN JOSE, CALIFORNIA 96125 GENERAL $ ELECTRIC APPLICABLE TO:

N E -30997 Pu.uCATiON NO.

T.1. E. NO. SHEET Ti m Edwin I. Hatch Nuclear Power Plant. Unit 1. RPV Surveillance DATE Materials Testing and Fracture Toughness Analysis October 1985 NOTE: Correctal/ copies of the applicable 1SSUE DATE publication as specified below.

REFERENCES INSTRUCTIONS ITEM '(SECTioN, PAG E PA R AG R APH, UNE) (CORRECTIONS AND ADDITIONS)

1. Page 1-1 Replace with new page 1-1. .

I

2. Page 2-3 Replace with new page 2-3.

3. Page 2-5 Replace with new page 2-5.

4. Page 3-1 Replace with new page 3-1.

5. Page 6-2 Replace with new page 6-2.

6. Page A-11/12 Replace with new page A-ll/A-12.

7. Page B-48 Replace with new page B-48.

8. Page B-53 Replace with new page B-53.

(Change brackets in right-hanc margin indicate l

areas where report has been revised.)

8701050323 861226 PCR ADOCK 05000321 p PDR O

PAGE

. NEDC-30997

1. INTRODUCTION Part of the effort to assure reactor vessel integrity involv(s evaluation of the fracture toughness of the vessel ferritic materials. The key values which characterize a material's fracture toughness are the reference tempera-ture of nil-ductility transition (RTNDT) and the upper shelf energy (USE).

These are defined in 10CFR50 Appendix G (Reference 1) and in Appendix G or the ASME Boiler and Pressure Vessel Code,Section III (Reference 2). These docu-ments contain requirements that establish the pressure-temperature operating limits which must te met to avoid brittle fracture.

Appendix H of 10CFR50 (Reference 3) and ASTM-E185 (Reference 4) establish the methods to be used for surveillance of the reactor vessel materials. In ,

November, 1984 one of the surveillance specimen capsules required in Reference 3 was removed from the Edwir I. Hatch Nuclear Plant, Unit 1 (hereaf ter called Hatch 1) reactor after 10 fuel cycles of irradiation, or 5.75 effective full power years (EFPY) of operatior.. The surveillance capsule contained flux wires for neutron flux monitorinc and Charpy V-Notch impact test specimens and uniaxial tensile test specimens fabricated from the vessel core beltline mate-rials. The impact end tensile specimens were tested to establish material properties for the irradiated vessel materials. The results of surveillance specimen testing are presented in this report. The irradiated material prop-erties are compared to available unirradiated properties from earlier tests, and predictions of the RT NDT and USE at end of reactor life (E0L) are made for comparison with allowable values in Reference 1. Predictions of E0L properties based on surveillance test results were made using Regulatory Guide 1.99, Revision 1 (Refettuce 5).

Operating limits curves for the Hatch 1 reactor vessel are presented in this report. The curves account for recent new requirements of References 1 and 2. Geometric discontinuities and highly stressed regions, such as the feedwater nozzles and the closure flanges, are evaluated separately from the 1-1

~.

NEDC-30997 core beltline region. In the case of the closure flanges, special require-ments from Reference 1 are applied, adding conservatism at relatively low pressures.

Proposed revisions of the Technical Specification and the Updated Final Safety Analysis Report (UFSAR) are included in Appendices A and B, respec-tively. The revisions are written to comply, to the extent practical, with the current requirements of fracture toughness analysis and surveillance as described in References 1 and 4.

l l

l 1-2

NEDC-30997

f. The maximum accumulated neutron fluence at the vessel end-of-life (EOL) (32 EFPY) was determined from the analytical model results at the peak 1/4 T location. The calculated fluence was conservative, compared to the fluence estimated with the flux wire test results and the analytical lead factor. The maximum 1/4 T vessel EOL fluence is 1.9 x 10 18 n/cm 2,

g. The surveillance Charpy V-Notch specimens were impact tested at temperatures selected to define the transition of the fracture tough-ness curves of the plate, weld, and HAZ materials. Measurements were taken of absorbed energy, lateral expansion and percentage shear.

Fracture surface photographs of each specimen are presented in Appendix C. From absorbed energy and lateral expansion results for the plate and weld materials the following values are extracted:

index temperatures for 30 ft-lb, 50 ft-lb, and 35-mil lateral expan-sion (MLE) values, USE and RT * ***** * **** ""'#8I NDT*

data are not curve fit because of the significant scatter in the test results.

h. The irradiated plate impact energy curve is compared to unirradiated data from earlier studies to establish the RT NDT irradiation shift for the surveillance program. The plate material shows a 47'F shift.

Decrease in USE was estimated for the plate material also, showing a 6% decrease.

1. The irradiated tensile specimens were tested at room temperature (76*F), reactor operating temperature (550*F), and estimated onset to upper shelf temperature (130*F or 185'F). The results tabulated for each specimen include yield and ultimate tensile strength (UTS),

uniform and total elongation, and reduction of area. The plate,

~

weld, and HAZ specimens behave similarly, showing lower properties at higher temperatures. The weld metal shows more temperature depen-dency in yield and reduction of area. .

2-3

NEDC-30997

j. The irradiated plate material tensile test results are compared to unirrrdiated data from the vessel fabrication test program. Based on the comparison, the tensile properties of the plate material are 1 unchanged as a result of irradiation.

k. As a part of the construction of the updated operating limits curves, the plate metal irradiation shift in RT NDT was compared to predict-ions calculated with Regulatory Guide 1.99, Revision 1 (Reference.5).

The surveillance test shif t of 47'F in plate material RT NDT I#*

17 n/cm2 (upper bound) is greater than the fluence of 3.0 x 10 predicted shift of 17'F. The method used to calculate irradiation shift versus full power years (from Reference 5) was modified to account for the surveillance test results of the plate material. As a result, the predicted EOL RT NDT shift is 174*F.

1. The USE at EOL is predicted using the methods in Reference 5. The weld metal USE is predicted to be 72 ft-lb at EOL. The miniaun plate ISE is 102 f t-lb longitudinal at EOL. Branch Technical Position MIEB 5-2 (Reference 6) recommends 65% of the longitudinal USE as an I estimate of transverse USE, so at EOL the plate USE would be 66 ft-lb transverse.

3

m. Operating limits curves were constructed for three reactor condi-tions: hydrostatic pressure tests, non-nuclear heatup and cooldown, and core critical operation. The curves are valid up to 16 EFPY of I

operation. The limiting regions of the vessel affecting the curves'

shapes are the et.re beltline (shifted to account for irradiation),

the feedwater nozzle, and the closure flange region. The bolt pre-j load and minimum permissible operating temperature on the curves of 76*F provides some additional margin in the closure flange region where a detectable flaw size of 0.24 inch is used instead of 1/4 T.

The operating limits curves for Hatch 1 are shown in Figures 2-1 through 2-3, 2-4 1

-~ - - - - - ,,--.----,---,,e,-,----e--m-,

w----,nvn---------w- ,-awy ,,,,,,ma. ,------,w---,-,,--~ .,-w -------.m +.mm~n- - - - - - - - - - . - , -

NEDC-30997 1

2.2 CONCLUSION

S

! The requirements of Reference 1 deal basically with EOL vessel conditions

{ and with limits of operation designed to prevent brittle fracture. Based on l the evaluation of surveillance testing, the following conclusions are drawn:

l a. The EOL value of RT NDT f r the plate material of 184*F is more i limiting than the weld metal EOL value and is below the Reference 1 4

annealing maximum limit of 200*F. The EOL values of USE for the i

plate and weld asterials are 66 ft-lb transverse and 72 ft-lb, respectively. These are above the Reference 1 minimum limit of 50 ft-lb for annealing. Therefore, provisions for annealing the reactor vessel before completing 32 EFPY of operation need not be considered.

1

b. Examination of the normal and upset operating conditions expected for the reactor shows that the worst pressure-temperature condition expected is 1180 psig at 250'F. Since this condition occurs af ter a j SCRAM and the steam done temperature is not changing significantly, Figure 2-1 is apr'.icable. The pressure limit at 250*F is over 1200 psig. Therefore, the only operating conditions for which the operat-ing limits are a concern are those involving operator control of the pressure and temperature, such as hydrotest.

c. The shift in RT measured for the irradiated plate material is f NDT greater than the shift predicted by Reference 5 by nearly a factor of 4 three. This underprediction is probably not representative because

! of the low fluence and small predicted shif t involved. However, the measured shift provides a valuable data point for shifts in the relatively low fluence range.

1

~

t i

i

< 2-5 t

NEDC-30997 1s00 VALID TO 16 EFFECTIVE FULL POWER YEARS OF OPERATION 1400 1200 ADJUSTED CORE BELTLINE, 1/4 T FLAW, RTNOT = 100F, IRRADIATION SHIFT = 1230F g

,I b

5 1

b p-2 800 E

3 E

A 600 VERTICAL LIMIT LINE FOR PRES $URE AsOVE 20% HYDROTEST (312 psig),

400 - B ASED ON 10CFR50 APPENDIX G REQUIREMENT OF (RTNDT

• 900F),

FLANGE REGION RTNDT

• 160F m

s BO'.T PRELOAD TEMFERATURE OF

, # 760F B ASED ON RECOMMENDED (RTNDT + 600F) FOR 0.24 IN. F LAW IN CLOSURE FLANGE REGION, RTNDT

• 10*F 0

0 100 200 300 400 600 600 MINIMUM VESSEL METAL TEMPER ATURE (*F)

Figure 2-1. Pressure versus Minimum Temperature for Pressure Tests, Based on Surveillance Test Results 2-6

NEDC-30997

3. SURVEILLANCE PROGRAM BACKGROUND 3.1 CAPSULE RECOVERY The Hatch 1 reactor was shut down in September 1984 for refueling and maintenance. The accumulated thermal power output was 5,109,600 mwd or 5.75 EFPY. The reactor pressure vessel (RPV) originally contained three surveil-lance capsules, at 30 , 120 , and 300-des azimuths at the core midplane. The specimen capsules are held against the RPV inside surface by a spring loaded specimen holder. Each capsule receives equal irradiation because of core .

symmetry. In November 1984, Capsule 1 at 30 degrees was removed. The capsule was cut from the holder assembly and shipped by a 204 series cask to the General Electric Vallecitos Nuclear Center in Pleasanton, California.

Upon arrival at Vallecitos, the capsule was examined for identification.

The reactor code of 61 and the basket code of 13 from Reference 7 were con-firmed on the capsule, as shown in Figure 3-1. The capsule contained three Charpy specimen packets and four tensile specimen tubes. Each Charpy packet contained 12 Charpy specimens and 3 flux wires. The four tensile specimen tubes contained eight specimens. The specimen sage sections were protected by aluminum sleeves, and during removal of the sleeves, the threaded ends of the specimens were slightly damaged. The threads were later chased with a die-hex rethreading tool. The sage sections of the tensile specimens were not damaged during removal.

3.2 RPV MATERIALS AND FABRICATION BACKGROUND 3.2.1 Fabrication History The Hatch 1 RPV is a 218-in. BWR/4. It was constructed by Combustion Engineering to the 1965 ASME Code with Addenda up to and including winter 1966. The shell and head plates are ASTM A533 Grade B, Class 1 low alloy steel (LAS). The nozzles and closure flanges are A508 Class 2 LAS and the 3-1

NEDC-30997 closure flange bolting materials are A540 Crade 524 LAS. The fabrication process employed quench and temper heat treatment immediately af ter hot fort-ing, then submerged are welding and post-weld heat treatment. The post-weld heat treatment was typically 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> at 1150' +25'F. The arrangement of plates and welds relative to the core beltline and various nozzles is shown in Figure 3-2.

3.2.2 Material Properties of RPV at Fabrication A search of General Electric Quality Assurance (QA) records was made to determine the chemical properties of the plates and welds in the RPV beltline.

Table 3-1 shows the chemistry data obtained for the beltl'ine materials. All data shown for the beltline plates, except copper content, were taken from QA records. The copper content of the plates was provided by Lukens Steel (Reference 8). Chemical composition of the beltline welds was obtained from Combustion Engineering (Reference 9), but as-welded copper data for the longi-tudinal seas welds was not available. Data for a weld with the same weld process and weld wire, but with a different flux, is included in Table 3-1.

Since the surveillance weld metal specimens were fabricated with the same weld procedure as was used in the longitudinal seas welds, samples from the speci-mens were analyzed to obtain a representative chemistry. The results are pre-seated in subsection 3.2.3.

A search of QA records was made to collect results of certification sechanical property tests performed during RPV fabrication, specifically tensile test, Charpy V-Notch and dropweight impact test results. Some results were not reported or could not be located. In such cases, the requirements of the purchase specification (Reference 10) are applied since no deviations from the specification were reported regarding RPV material properties. Properties of the beltline materials and other locations of interest are presented in Table 3-2.

3-2

! NEDC-30997 l

I

! 6. TENSILE TESTING Eight round bar tensile specimens were recovered from the surveillance capsule. Uniaxial tensile tests were conducted in air at room temperature, RPV operating temperature, and onset of upper shelf temperature. Tests were conducted in accordance with ASTM E8-81 (Reference 13).

l l

6.1 PROCEDURE All tests were conducted using a screw-driven Instron test frase equipped with a 20-kip load cell and special pull bars and grips. Heating was done with a Satec resistance cleashell furnace centered around the specimen load l train. Test temperature was monitored and controlled by a chronel-alueel thermocouple spot-welded to an Inconel clip that was friction-clipped to the surface of the specimen at its midline. Before the elevated temperature tests, a profile of the furnace was conducted at the test temperature of interest using an unitradiated steel specimen of the same geometry. Thermocouples were spot-welded to the top, middle, and bottom of a central 1-in. sage of this specimen. In addition, the clip-on thermocouple was attached to the midline of the specimen. When the target temperatures of the three thermocouples were within +5'F of each other, the temperature of the clip-on thermocouple was noted and subsequently used as the target temperature for the irradiated specimens.

All tests were conducted at a calibrated crosshead speed of 0.005 in./ min until well past yield, at which ti=* the speed was increased to 0.05 in./ min until fracture. A 1-in, span knit ,e extensoneter was attached directly to each specimen's central ange region and was used to monitor sage extension during test.

The test specimens were nachined with a minimum diameter of 0.250 inch at the center of the sage length. The three specimens each of base metal and HAZ were tested at room temperature (RT = 76*F), onset of upper shelf temperature (estimated at 130'F and 185'F, respectively), and RPV operating temperature 6-1 I

{

NEDC-30997 (550*F). The tvs weld metal specimens were tested at upper shelf temperature (185'F) and 550'F. The yield strength (YS) and ultimate tensile strength (UTS) were calculated by dividing the nominal area (0.0491 in.2) into the 0.2% offset Idad and into the maximum test load, respectively. The values listed for the uniform and total elongations were obtained from plots that recorded load versus specimen extension and are based on a 1-in. sage length.

Reduction of area (RA) values were determined from post-test measurements of the necked specimen diameters using a calibrated blade micrometer and employ-ing the formulat A -A RA(%) =

• 100.

o ,

After testing, each broken specimen was photographed end-on showing the frac-ture surface and lengthwise showing fracture location and local necking behavior.

6.2 RESULTS Tensile test properties of YS, UTS, RA, and uniform and total elongation (TE) are presented in Table 6-1. Shown in Figure 6-1 is a stress-strain curve for a 550*F base metal specinen typical of the stress-strain characteristics of all the specimens tested. Shown graphically in Figures 6-2 and 6-3 are the data in Table 6-1. Photographs of fracture surfaces and necking behavior are given in Figures 6-4, 6-5 and 6-6 for base, weld and HAZ specimens respectively.

The base, weld, and HAZ materials generally follow the trend of decressing properties with increasing temperature. The three materials behave very simi-larly, as seen in Figures 6-2 and 6-3, with the weld metal YS and 550'F RA showing more temperature dependency than the other materials.

6-2

l .. .,

l NEDC-30997 analysis location gives a conservative estimate of nazimum 1/4 7 depth flux of 1.86 x 109 (n/cm 2_,,,),

The first capsule containing test specimens was withdrawn in ,

November 1984 after 5.75 EFPY of operation. The specimens were .

tested according to ASTM E135-82 and the results are in GE report NEDC-30997. The curves of Figures 3.6-1 through 3.6-3 include the findings of the test report related to the copper phosphorus content of the RPV core beltline asterials, the flux wire test and fluence distribution analysis results, and the Charpy V-Notch specimen test results.

C. Reactor Vessel Head Stud Tensionina ,

The requirements for cold bolt-up of the reactor vessel closure are based on the RT NDT temperature plus 60*F which is derived from the requirements of the ASME Code to which the vessel was built.

The maximus RT NDT of the closure flanges, adjacent head and shell material and stud material is 16'F. The minimu.a temperature for bolt-up is therefore 16 + 60

• 76*F. The neutron radiation fluence j at the closure flanges is well below 10 17 nyt (> 1 Hev) and therefore radiation effects will be minor and will not influence this temperature.

A-11/A-12

NEDC-30997 i

GENER AL h ELECTRIC CMI TRANSMITTAL

, No. 85-212-0008 ,

NUCLEAR FUEL AND SPECIAL PROJECTS DIVISION CORE MATERIALS TESTING AND ANALYSIS TEST REPORT DETERMINATION OF FAST NEUTRON FLUX DENSITY AND FLUENCE:

MATCH 1 NUCLEAR POWER PLANT March 11, 1985 Prepared By: h. C . W\ub- 3[/J//[

G. C. Martin Date Core Materials Testing & Analysis Reviewed By: Y ee ld J/d//r Date L. K. Kessler Core Materials Testing & Analysis Approved By: / 4"E' # 4I*

R. B. Adamson, Manager

)//3/dI Date Core Materials Testing & Analysis B-47

NEDC-30997 DETERMINATION OF FAST NEUTRON FLUX DENSITY AND FLUENCE:

MATCH 1 NUCLEAR POWER PLANT SIMtARY The fast neutron flux density and fluence (integrated neutron flux) at a capsule near the reactor vessel wall of the Match 1 Nuclear Power Plant of the .

Georgia Power Company have been determined to be ,

1.3x10' n/cm2 .s >l MeV full-power flux density 2.1x10' n/ce2 *s >0.1 MeV full-power flux density 2.3x1017 n/cm2 >l MeV fluence 3.6x1017 n/cm2 >o,g y,y ggy,,c, following the analysis of irradiated copper flux dosimeters, in accordance with the GE CN&S Method No. 10.1.6.0 R3.

EXPERIMENTAL Wires of iron, nickel and copper (three each) were irradiated in a GE pressure vessel capsule holder at Itatch 1 from November 2,1974 (startup) to September 29, 1984. Each wire was removed from the capsule, cleaned with dilute acid, weighed, mounted on a counting card, and analysed for its radio-activity content by gamma spectrometry. The copper wires, after cleaning, were still coated with a dark layer containing fission products Ce-144, sb-125 Eu-152. Ru-106, Cs-137, and Cs-134 These wires were purified by partial dissolution with concentrated nitric acid. Each iron wire was analysed for Mn-54 content, each nickel wire for Co-58, and each copper wire for Co-60 at a calibrated 4 or 10 cm source-to-detector distance with 80-ce Ge(L1) and 35-ce Ge(L1) detector systems .

From daily thermal power genera: ion histograms and history summary tables, the irradiation time periods (cycles) were evaluated as indicated in Table 1.

B-48

• { ...

NEDC-30997 RE-DETERMINATION OF FAST NEUTRON FLUX DENSITY AND FLUENCE USING LOCALI"ED F0WER HISTORY:

MATCH t NUC . EAR POWER PLANT

$1291ARY The fast neutron flux density and fluence (integrated neutton flux) at a capsule near the reactor vessel wall of the Match 1 Nuclear Power Plant of the .

Georgia Power Company have been determined to be 1.3x10' n/ca tas >1 MeV full.-pcwor flux density 2.1x10' n/ca 2es >0.1 MeV full-power flux density 2.4x1017 n/ca2 >g y,y gg,,,,,

3.gx1017 n/ce2 >0.1 MeV fluence Localised power history (as opposed to reactor power history) has been utilised for the dostaetry determinations for this report.' Compared to results of the original (3/11/05) report, no change in full-power fast-neutron flux density was apparent; an increase of approximately 52 in fluence has resulted.

DISCUSSION Evaluation of the neutron flux in the vicinity of the Match 1 pressure vassa.1 dosimetry capsule as a function of cycle-time has been made by Shielding and Radiological Engineering. Based on bundle exposure data for edge bundles nearest the capsule, and adjusting for void differences, a more realistic power history (localised vs. reactor - see Table 1) has improved the relative fast-neutron results from the three dosimeter types (copper, iron, nickel).

Table 2 gives the re-determined flux density and fluence results at the dosimetry capsule location.

The previous report (CNT Transmittal No. 85-212-0008) gave flux density differences of 20% (iron vs. copper) and 30% (nickel vs. copper) which indicated an inconsistent power history over the ten-year irradiation.

Because of the improved consistency in the results from the asasured shorter-lived (312-d) Mn-54 (fros iron) and (71-d) Co-58 (from nickel) compared to (5.3-y) Co-60 (from copper), increased confidence in the results from utilisation of the localised power history values is warranted.

Results from this report supersede results from report number 85-212-0008. No change was evident in the reported full-power flux density results: an increase of approximately 5% in fluence has resulted.

B-53

t.')

NEDC-30997 ,

TABLE 1. Match 1 Irradiation Cycles Localised Reactor Full-Fower Weighted Full-Fower Between Cycle Cycle

• Days Fraction ** Relative Flux ** Fraction **** Time (Days) 1 860 0.516 1.29 0.666 72 2 284 0.768 1.13 0.868 45 3 371 0.723 1.14 0.824 129 4 548 0.702 0.87 0.611 110 5 114 0.801 0.95 0.761 48 6 149 0.604 1.03 0.622 49 7 119 0.772 1.03 0.795 135 8 257 0.808 0.97 0.784 36 9 63 0.801 0.74 0.593 24 10 206 0.822 0.74 0.608

• Refer to March 11, 1985 report for cycle dates.

• • Full power was 2436 MW

• • • ForbundlesneardosinItercapsule.

• • • • This report.

B-54