Revision as of 17:02, 30 June 2018

Text

Official Exhibit - NYS000518-00-BD01 - Kirk, Mark and Sheng, Simon, Usnrc, Assessment of BTP 5-3 Protocols to Estimate Rtndt(U) and Use, (June 4, 2014) (ML14163A524)
	ML15331A202
Person / Time
Site:	Indian Point
Issue date:	06/04/2014
From:	; State of NY, Office of the Attorney General
To:	; Atomic Safety and Licensing Board Panel
	SECY RAS
References
	RAS 27910, ASLBP 07-858-03-LR-BD01, 50-247-LR, 50-286-LR
	Download: ML15331A202 (37)
	v • d • e

Assessment of BTP 5

-3 Protocols to Estimate RTNDT(u) and USE Mark Kirk Simon Sheng Senior Materials Engineer Senior Materials Engineer RES/DE/CIB NRR/DE/EVIB mark.kirk@nrc.gov simon.sheng@nrc.gov

NRC/EPRI Annual Materials Issue Program Information Exchange Meeting 4th June 2014 Rockville, Maryland, USA

Outline of Presentation Un-Irradiated RTNDT (RTNDT(u)) & Un-Irradiated Upper Shelf Energy (USE) definitions & estimates Background of questions concerning BTP 5-3 Staff Assessment Part I - Technical evaluation of BTP 5-3 estimation of RTNDT(u) & USE Staff Assessment Part I - Potentially Affected Plants Next steps

TemperatureNDT is the lowest temperature of "no

-break" performance No-Break: Fracture (darkened region) does not extend to the sides of the specimenBreak: Crack completely severs tension surface of specimen.

60,50/35TTMAXRTNDTNDT(u) Specimens notched transverse to RD Definitions: RTNDT(u) & USE per ASME NB

-2331 per ASTM E185

-82 USE average of all energies > 95% shear RTNDT(u) & USE Estimated by NUREG

-0800 BTP 5-3 60,50/35TTMAXRTNDTNDT(u) per ASME NB

-2331 per ASTM E185

-82 USE average of all energies > 95% shear Position 1.2 Positions 1.1(1) & 1.1(2)

Position 1.1(3)

Position 1.1(4)

Approximations

AREVA Letter (30 Jan 2014, AREVA Ref. NRC:14:004) & PVP Paper (PVP2014

-28897) claim Position 1.1(4) of BTP 5.3 is sometimes non-conservative for A508-2 forgings Literature search reveals 1983

EG&G report & 1985 IJPVP paper -Evaluation of BTP 5

-3 (then MTEB 5-2) for NRC

-Conclusions Always conservative

-Position 1.1(1): estimates TNDT -Position 1.1(2): estimates TNDT Sometime non

-conservative

-Position 1.1(3): estimates TCVE(50/35)

-Position 1.1(4): estimates RTNDT -Position 1.2: estimates USE Background of Questions Concerning BTP 5-3 AREVA Letter (30 Jan 2014, AREVA Ref. NRC:14:004) & PVP Paper (PVP2014

-28897) claim Position 1.1(4) of BTP 5.3 is sometimes non-conservative for A508-2 forgings Literature search reveals 1983

EG&G report & 1985 IJPVP paper -Evaluation of BTP 5

-3 (then MTEB 5-2) for NRC

-Conclusions Always conservative

-Position 1.1(1): estimates TNDT -Position 1.1(2): estimates TNDT Sometime non

-conservative

-Position 1.1(3): estimates TCVE(50/35)

-Position 1.1(4): estimates RTNDT -Position 1.2: estimates USE Background of Questions Concerning BTP 5-3 Part II: Assessment of applicability to plants Query RVID

-RTNDT(u): establishes BTP 5-3 use, but not which position was used -USE: establishes BTP 5

-3 use Search for documents

referenced by RVID in ADAMS legacy -Focus on plants closest to PTS

(50.61) limit, these being most prone to influence by potential non-conservatisms

-References establish which

position of BTP 5

-3 was used for RTNDT(u) Part I: Technical evaluation of BTP 5-3 estimation of RTNDT(u) and USE Data sources -Processed data (T50, USE, -) from 1983 EG&G report

-Raw data (CVE, MLE, temp) in

both specimen orientations from surveillance reports (stored in REAP) -Raw data (NDTT) from RVID refs.

Focus on -Plates & forgings only No plants have used BTP 5

-3 for welds -Positions identified as sometimes

non-conservative in 1983 by EG&G report Position 1.1(3): estimates TCVE(50/35)

Position 1.1(4): estimates RTNDT Position 1.2: estimates USE NRC Staff Assessment Process

Part I: Technical Evaluation Overview While similar answers are expected from both sources -

Given the potential impact of this evaluation, going

back to the raw data was seen to be important.

Data sources

-Processed data (T50, USE, -) from 1983 EG&G report

-Raw data (CVE, MLE, temp) from surveillance reports (stored in REAP)

-Raw data (NDTT) from RVID

refs. Focus on -Plates & forgings only No plants have used BTP 5

-3 for welds -Positions identified as

sometimes non

-conservative by 1983 EG&G report Position 1.1(3): estimates

TCVE(50/35)

Position 1.1(4): estimates RTNDT Position 1.2: estimates USE 0255075100125150 tanh Fit Data0255075100125150-200-1000100200300 tanh Fit DataTemperature [F] Impact Energy [

ft-lbs] 130F 25% 28% 43% 47% A508-2: Heat 527536

Quotation If transversely

-oriented Charpy V

-notch specimens were not tested, the temperature at which 68 J (50 ft-lbs) and 0.89 mm (35 mils) LE would have been obtained on transverse specimens may be estimated by one of the following criteria:

-Test results from longitudinally

-oriented specimens reduced to 65% of their value to provide conservative estimates of values expected from transversely oriented specimens.

-Temperatures at which 68 J (50 ft-lbs) and 0.89 mm (35 mils) LE were obtained on longitudinally

-oriented specimens increased 11

°C (20 °F) to provide a conservative estimate of the temperature that would have been necessary to obtain the same values on transversely

-oriented specimens.

Position 1.1(3)

Tests Required Longitudinally oriented CVN specimens Clear Interpretation Note that this position applies only to conversion between longitudinal and transverse Charpy values.

There are two approximations. They may not produce the same results. They are as follows (a) ETRANS = 0.65xELONG, then calc TC(TRANS) MLETRANS = 0.65xMLELONG, then calc TC(TRANS-MLE) (b) TC(TRANS) = TC(LONG) + 20 °F TC(TRANS-MLE = TC(LONG-MLE) + 20 °F where ELONG is CVN energy measured by a longitudinally oriented specimen ETRANS is the estimated CVN for a transversely oriented specimen TC(LONG) is the temperature at which the minimum of three longitudinal CVN tests exhibits >35 mils AND >50 ft

-lbs TC(TRANS) is the estimated temperature at which the minimum of three transverse CVN tests exhibits >35 mils AND >50 ft

-lbs 1.Per the BTP, reduce longitudinal measurements to 65% of the measured values 2.Fit Charpy curves

-Energy vs. temperature

-Lateral expansion vs.

temperature 3.Determine MAX(T50ft-lb, T35mills) 4.Value from Step 3 estimates the transition temperature of transverse data Position 1.1(3)

Assessing(a): Trans

= 0.65Long Energy Temperature Longitudinal Transverse = 0.65xLongitudinal

0.00.10.20.30.40.50.60.70.80.91.0-100-50050100150200PercentileTransverse T 50Estimate Error [F]Non-conservative predictons: 33% Plates, 43% Forgings, 36% OverallPlateForgingBTP 5-3 Position 1.1(3) (+20F)

-100-50050100150200-50-250255075100125Transverse T 50Estimate Error [F]Trans T50est. from Longx0.65 [F]Non-conservative predictons: 33% Plates, 43% Forgings, 36% OverallPlateForgingBTP 5-3 Position 1.1(3) (+20F)Position 1.1(3)

Assessing(a): Trans

= 0.65Long Non-conservative Non-conservative Same data plotted two different ways Position is non-conservative about 36% of the time

-100-50050100150200-50-250255075100125Trans. T50ft-lbs&35millsEstimate Error [F]Trans. T50ft-lbs&35millsest. from Longx0.65 [F]Non-conservative predictons: 19% Plates, 48% Forgings, 30% OverallPlateForgingBTP 5-3 Position 1.1(3) (0F)0.00.10.20.30.40.50.60.70.80.91.0-100-50050100150200PercentileTransverse T50ft-lbs&35millsestimate error [F]Non-conservative predictons: 19% Plates, 48% Forgings, 30% OverallPlateForgingBTP 5-3 Position 1.1(3) (0F)Position 1.1(3)

Assessing(a): Trans

= 0.65Long Non-conservative Non-conservative Same data plotted two different ways Position is non-conservative about 30% of the time T50ft-lbs determines the value of T50ft-lbs&35mills 81% of the time for longitudinal specimens 92% of the time for transverse specimens

-100-50050100150200250-50-250255075100125Transverse T 50[F]Trans. T50from Longx0.65[F]Mean: T50T= 0.72xT50(e)+ 12.5 Bound: T50T= 0.72xT50(e)-27.5Plate & ForgingOutlier (excluded)MeanLower BoundPosition 1.1(3)

Assessing(b): Alternative TC(TRANS) Estimates Preliminary analysis suggests that alternative formulae could be developed to convert longitudinal to transverse T50 values in a manner that is always conservative.

-100-50050100150200250-50-250255075100125Transverse T50ft-lbs&35mills

[F]Trans. T50ft-lb&35millsest. from Longx0.65 [F]Mean: T50&35T= 0.66xT50&35(e)+ 13 Bound: T50&35T= 0.66xT50&35(e)-34Plate & ForgingMeanLower BoundPosition 1.1(3)

Assessing(a): Alternative TC(TRANS) Estimates Preliminary analysis suggests that alternative formulae could be developed to convert longitudinal to transverse T50 values in a manner that is always conservative.

Position 1.1(3)

Assessing(b): TC(TRANS)

= TC(LONG) + 20 F Non-conservative Non-conservative Same data plotted two different ways

-100-50050100150200-50-250255075100125Transverse

-Longitudinal T 50[F]Longitudinal T 50[F]Non-conservative predictons: 70% Plates, 50% Forgings, 64% OverallPlateForgingBTP 5-3 Position 1.1(3) (+20F)0.00.10.20.30.40.50.60.70.80.91.0-100-50050100150200PercentileTransverse

-Longitudinal T 50[F]Non-conservative predictons: 70% Plates, 50% Forgings, 64% OverallPlateForgingBTP 5-3 Position 1.1(3) (+20F)Position is non-conservative about two

-thirds of the time 0.00.10.20.30.40.50.60.70.80.91.0-100-50050100150200PercentileTransverse

-Longitudinal T50ft-lbs&35mills

[F]Non-conservative predictons: 63% Plates, 57% Forgings, 60% OverallPlateForgingBTP 5-3 Position 1.1(3) (+20F)-100-50050100150200-50-250255075100125Trans. -Longl. T50ft-lbs&35mills

[F]Longitudinal T50ft-lbs&35mills

[F]Non-conservative predictons: 63% Plates, 57% Forgings, 60% OverallPlateForgingBTP 5-3 Position 1.1(3) (+20F)Position 1.1(3)

Assessing(b): TC(TRANS)

= TC(LONG) + 20 F Non-conservative Non-conservative Same data plotted two different ways Position is non-conservative about two-thirds of the time T50ft-lbs determines the value of T50ft-lbs&35mills 81% of the time for longitudinal specimens 92% of the time for transverse specimens

Position 1.1(3)

Assessing(b): Alternative TC(TRANS) Estimates Preliminary analysis suggests that alternative formulae could be developed to convert longitudinal to transverse T50 values in a manner that is always conservative. -100-50050100150200250-50-250255075100125Transverse T 50[F]Longitudinal T 50[F]Mean: T50T= 0.58xT50L+ 44.7 Lower Bound: T50T= 0.58xT50L+ 2.0Plate & ForgingOutlier (excluded)MeanLower Bound

-100-50050100150200250-50-250255075100125Transverse T50ft-lbs&35mills

[F]Longitudinal T50ft-lb&35mills

[F]Mean: T50&35T= 0.68xT50&35L+ 35 Low Bound: T50&35T= 0.68xT50&35L-17Plate & ForgingMeanLower BoundPosition 1.1(3)

Assessing(b): Alternative TC(TRANS) Estimates Preliminary analysis suggests that alternative formulae could be developed to convert longitudinal to transverse T50 values in a manner that is always conservative.

Tests Required Limited longitudinally oriented CVN tests at a single temperature Interpretation Define: TTEST = the temperature at which limited longitudinally oriented CVN tests were conducted CV = absorbed energy observed at TTEST IF CV 45 ft-lbs then RT NDT = TTEST ELSE RTNDT = TTEST + 20 °F Quotation If limited Charpy V

-notch tests were performed at a single temperature to confirm that at least 41 J (30 ft-lbs) was obtained, that temperature may be used as an estimate of the RTNDT provided that at least 61J (45 ft-lbs) was obtained if the specimens were longitudinally oriented. If the minimum value obtained was less than 61 J (45 ft-lbs), the RTNDT may be estimated as 11

°C (20 °F) above the test temperature

. When assessed using data sets for which full Charpy energy curves are available, EG&G interpreted Position 1.1(4) as having 2 possible meanings:

-RTNDT = T45(LONG), and -RTNDT = T30(LONG) + 20 F These might not produce the same result. Therefore, both were assessed.

Position 1.1(4)

Assessed for Forgings Top & bottom panel of each pair are the same data plotted two different ways -150-100-50050100150-150-100-50050100Estimated RTNDT= T30L+20 [F]"Official" RTNDTper ASME NB

-2331 [F]SA-508-2 Forging. RTndt per BTP 5

-3 1.1(4)0.00.10.20.30.40.50.60.70.80.91.0-100-50050100150200PercentileError in Estimated RTNDT[F]93% of data under

-predicted. Max =

-95F-150-100-50050100150-150-100-50050100Estimated RTNDT= T45L[F]"Official" RTNDTper ASME NB

-2331 [F]SA-508-2 Forging. RTndt per BTP 5

-3 1.1(4)0.00.10.20.30.40.50.60.70.80.91.0-100-50050100150200PercentileError in Estimated RTNDT[F]93% of data under

-predicted. Max =

-86FNon-Conservative Non-ConservativeNon-ConservativeNon-ConservativeForgingSA-508-2RTNDT = T30(LONG) + 20 F RTNDT = T45(LONG) Position is non-conservative about 90% of the time 0.00.10.20.30.40.50.60.70.80.91.0-100-50050100150200PercentileError in Estimated RTNDT[F]38% of data under

-predicted. Max =

-27F-150-100-50050100150-150-100-50050100Estimated RTNDT= T45L[F]"Official" RTNDTper ASME NB

-2331 [F]SA-533B-1 Plate. RTndt per BTP 5

-3 1.1(4)0.00.10.20.30.40.50.60.70.80.91.0-100-50050100150200PercentileError in Estimated RTNDT[F]38% of data under

-predicted. Max =

-38F-150-100-50050100150-150-100-50050100Estimated RTNDT= T30L+20 [F]"Official" RTNDTper ASME NB

-2331 [F]SA-533B-1 Plate. RTndt per BTP 5

-3 1.1(4)Non-Conservative Non-ConservativeNon-ConservativeNon-ConservativePlateSA-533B-1Position 1.1(4)

Assessed for Plates Top & bottom panel of each pair are the same data plotted 2 different ways RTNDT = T30(LONG) + 20 F RTNDT = T45(LONG) Position is non-conservative about 40% of the time Quotation For the beltline region of reactor vessels, the upper shelf toughness must account for the effects of neutron radiation.

Reactor vessel beltline materials must have Charpy upper shelf energy, in the transverse direction for base material and along the weld for weld material according to the ASME Code, of no less than 102 J (75 ft-lbs) initially and must maintain Charpy upper shelf energy throughout the life of the vessel of no less than 68 J (50 ft-lbs). If Charpy upper shelf energy values were not obtained, conservative estimates should be made using results of tests on specimens from the first surveillance capsule removed.

If tests were only made on longitudinal specimens, the values should be reduced to 65% of the longitudinal values to estimate the transverse properties.

Position 1.2 Tests Required Longitudinally oriented CVN specimens tested on the upper shelf. Clear Interpretation USETRANS = 0.65 x USELONG where USELONG is CVN energy measured by longitudinally oriented specimens on the upper shelf USETRANS is the estimated CVN energy for transversely oriented specimens on the upper shelf

0.00.10.20.30.40.50.60.70.80.91.00.00.20.40.60.81.01.21.4PercentileTransverse / Longitudinal USE RatioNon-conservative predictons: 20% Plates, 14% Forgings, 18% OverallPlateForgingBTP 5-3 Position 1.2 (0.65)0.00.20.40.60.81.01.21.4050100150200250Transverse / Longitudinal USE RatioLongitudinal Upper Shelf Energy [ft

-lbs]Non-conservative predictons: 20% Plates, 14% Forgings, 18% OverallPlateForging BTP 5-3 Position 1.2 (0.65)Position 1.2 Assessing: USETRANS = 0.65 USELONG Non-conservative Non-conservative Same data plotted two different ways Position is non-conservative about 18% of the time 0.00.10.20.30.40.50.60.70.80.91.00.00.20.40.60.81.01.21.4PercentileTransverse / Longitudinal USE RatioNon-conservative predictons: 13% Plates, 33% Forgings, 21% OverallPlateForgingBTP 5-3 Position 1.2 (0.65)0.00.20.40.60.81.01.21.4050100150200250Transverse / Longitudinal USE RatioLongitudinal Upper Shelf Energy [ft

-lbs]Non-conservative predictons: 13% Plates, 33% Forgings, 21% OverallPlateForging BTP 5-3 Position 1.2 (0.65)Position 1.2 Assessing: USETRANS = 0.65 USELONG Non-conservative Non-conservative Same data plotted two different ways Position is non-conservative about 21% of the time Position 1.2 Assessing: Alternative USETRANS Estimates Preliminary analysis suggests that alternative formulae could be developed to convert longitudinal to transverse USE values in a manner that is always conservative. 050100150200250050100150200250Transverse Upper Shelf Energy [ft

-lb]Longitudinal Upper Shelf Energy [ft

-lb]Mean: Tuse = 0.61xLuse + 21.2 Lower Bound: Tuse = 0.61xLuse

-22.1PlateForgingMeanLower Bound 050100150200250050100150200250Transverse Upper Shelf Energy [ft

-lb]Longitudinal Upper Shelf Energy [ft

-lb]Mean: Tuse = Luse

-36 Lower Bound: Tuse = Luse

-65PlateForgingMeanLower BoundPosition 1.2 Assessing: Alternative USETRANS Estimates Preliminary analysis suggests that alternative formulae could be developed to convert longitudinal to transverse USE values in a manner that is always conservative.

Summary on Part I - Technical Evaluation Positions 1.1(3) and 1.2

-Results of the two studies are similar

-Staff analysis confirms non

-conservatism Position 1.1(4)

-EG&G report demonstrates position is non

-conservative

-Awaiting NDTT data from Archives to complete staff assessment Position of BPT 5-3 Forging Non-Conservative Prediction Rate Plate Non-Conservative Prediction Rate EG&G Data Raw Data EG&G Data Raw Data 1.1(3) (a) TRANS

= 0.65LONG 43% 48% 33% 19% (b) TC(TRANS)

= TC(LONG) + 20 F 50% 57% 70% 63% 1.1(4) RTNDT = T45(LONG) 93% TBD 38% TBD RTNDT = T30(LONG) + 20 F 93% TBD 38% TBD 1.2 USETRANS = 0.65 USELONG 14% 33% 20% 13%

Part II: Assess Potentially Affected Plants

- Position 1.1(3) Plant Identification Search RVID for plants using BTP 5

-3 to determine plate (forging) RTNDT(u): 20 operating plants Rank plates (forging) according to the difference between RTPTS at 32 EFPY and 270

°F: Eight plants have their limiting plates or forgings using BTP5

-3 with difference less than 100

°F.

Plant-specific evaluation results The majority of the plants did not specify which BTP 5

-3 B1.1 position was used in determining their RTNDT(u) values Details of calculation of RTNDT(u) values are not available.

One plant has full transverse Charpy data and the staff confirmed that BTP 5-3 was not used , so it will be dropped from the list A few plants have full longitudinal Charpy data - The staff's RTNDT(u) values using lower bound Charpy data and linear interpolation between two temperatures are lower than the licensee's value by 10 °F A few plants may have PTS concern because the RTPTS values are below 270 °F by less than 75

°F - In one case, the longitudinal Charpy data for one plate are significantly higher than other plates, indicating potential mislabeling

-100-50050100150200-50-250255075100125Trans. -Longl. T50ft-lbs&35mills

[F]Longitudinal T50ft-lbs&35mills

[F]Non-conservative predictons: 63% Plates, 57% Forgings, 60% OverallPlateForgingBTP 5-3 Position 1.1(3) (+20F)

Why do we use the Selecting criterion of 75

°F?

- Identify raw data with the greatest conservatism and non

-conservatism Watts's Bar 1 - greatest non-conservatism Non-conservative Watt's Bar 1 Millstone 2 - greatest conservatism

Closer Look at the Charpy Data with the Greatest Non

-Conservatism Longitudinal Data T50flb [oF] T35mill [oF] 31 - 0255075100125150-200-1000100200300Impact Energy [ft

-lbs]Temperature [F]tanh FitData0255075100125150-200-1000100200300Impact Energy [ft

-lbs]Temperature [F]tanh FitDataT50flb [oF] T35mill [oF] 114.5 83.0 Transverse Data T50flb [oF] T35mill [oF] -15.5 -9.4 Longitudinal Data X .65

Determine the RTNDT for the Raw Data with the Greatest Non

-Conservatism Nil-ductility transition temperature (NDTT): -22 °F Official RTNDT(u) value: = 54.5 °F (114.5 °F - 60 °F) RTNDT(u) based on BTP 5-3B1.1(3)a: -22 °F (At 31

°F, the equivalent Charpy energy (.65 x longitudinal data) is 50 ft

-lb; RTNDT = NDTT) RTNDT(u) based on BTP 5-3B1.1(3)b: -22 °F (At -15.5 °F, the Charpy energy is 50 ft-lb; since the adjusted temp is (-15.5 °F

+ 20 °F ), less than (-22°F + 60 °F), RTNDT = NDTT)

Summary on the Study Focusing on the Raw Data with the Highest Non-Conservatism The highest non

-conservative raw data is about 75

°F RTNDT determination is not sensitive to whether B1.1(3)a or B1.1(3)b is used for this case RTNDT determination is affected by whether curve fitting of the entire Charpy data or hand calculations based on Charpy data at two temperatures are used

Part II: Assess Potentially Affected Plants

- Position 1.2 Plant Identification

-45 operating plants identified in RVID as using Position 1.2 -RVID clearly identifies Position 1.2 as

UNIRR_USE_METHOD=65% -Spot-checking of RVID references to confirm accuracy still TBD Non-conservatism

-Data analysis shows the Position 1.2 estimate to be

non-conservative between 13% and 33% of the time

Next Steps NRC Complete technical analysis

-Need NDTT data from Archives to complete assessment of Position 1.1(4)

-Investigate GE RTNDT(u) procedure

-Document findings Complete plant assessment

-Need to assess the impact to Pressure

-temperature limits

-Recommend to NRC management regarding use of interim conservatism in defining RTNDT(u) for the plants which may need to update their PTS evaluations Communicate findings to affected plants

-Precise means TBD May need to revise BTP 5

-3 in Standard Review Plan Industry Assess the impact of reported potential non

-conservatism including the need to redefine the RTNDT(u) on pressure

-temperature limits and PTS evaluations

BACKUP SLIDES

MethodNDTT [F]Meas. T50 [F]Trans. T50 [F]RTNDT [F]Non Conservatism

[F]NB-2331-22114.5114.554.5

---1.1(3)a: Energy*0.65-223131-2276.51.1(3)b:

T50(LONG)+20 F-22-15.54.5-2276.5NB-2331-22109.5109.549.5

---1.1(3)a: Energy*0.65-224141-1968.51.1(3)b:

T50(LONG)+20 F-22020-2271.5 Charpy Fit Method: tanh (mean)Charpy Fit Method: Interpolate lower bound dataMain Points There is some effect of tanh fitting versus lower

-bound interpolation

-Interpolation can produce higher or lower transition temperature values than tanh fitting Using either Charpy fitting method,

BTP 5-3 Position 1.1(3) is non

-conservative Comparison of Charpy Fit Methods 0.65 0255075100 tanh FitData0255075 tanh FitDataCharpy Energy [ft

-lbs] 0255075100125150-200-1000100200300 Temperature [F]tanh FitDatatanh fit interpolate lower bound data Assessment of BTP 5