ML12340A798

Official Exhibit - BRD000002-00-BD01 - Presentation of Dr. Joram Hopenfeld to the ASLB in Response to Dr. Horowitz Regarding Checworks

ML12340A798
Person / Time
Site:	Indian Point
Issue date:	10/15/2012
From:	Riverkeeper
To:	Atomic Safety and Licensing Board Panel
SECY RAS
References
RAS 23614, 50-247-LR, 50-286-LR, ASLBP 07-858-03-LR-BD01
Download: ML12340A798 (14)

Text

United States Nuclear Regulatory Commission Official Hearing Exhibit Entergy Nuclear Operations, Inc.

In the Matter of:

(Indian Point Nuclear Generating Units 2 and 3)

ASLBP #: 07-858-03-LR-BD01 Docket #: 05000247 l 05000286 Exhibit #: BRD000002-00-BD01 Identified: 10/15/2012 Admitted: Withdrawn:

Rejected: Stricken:

Other:

Presentation of Dr. Joram Hopenfeld to the ASLB in Response to Dr.

Horowitz Regarding CHECWORKS Indian Point License Renewal Proceeding Adjudicatory Hearing October 15, 2012

I FUNDUMENTALLY DISAGREE WITH DR. HOROWITZ REGARDING HOW THE CHECWORKS PROGRAM WAS DESIGNED AND HOW IT WORKS MAIN POINTS OF DISAGREEMENT Subject CLAIMED BY ENTERGY Disagree/Agree NOTES Background Need for single Disagree Not the main need phase inspection Models technical All known data Disagree Selective data was used basis Mathematical Inputs represent Disagree Component Cr. Content Analogue known quantities, is unknown. Incorrect (mass transfer rates relation between metal Cr, geometry) loss and mass transfer coefficient Designed to Can handle changes Disagree No correlation before handle changes in in flow and after flow changes Plant Conditions Model predictions Validated and Refined Disagree 40-60% of the time predictions are not conservative

BACKGROUND

• CHECWORKS was introduced by EPRI 7 months after the Surry accident. To meet schedule, EPRI developed its own definition of FAC:

i. EPRIs Definition: FAC = corrosion (controlled by diffusion mass transfer) ii. Common definition: Erosion/ Corrosion = Corrosion + Erosion + Synergy Note: Corrosion is relatively easy to predict; corrosion + synergy is very difficult to predict

• To support its FAC definition, EPRI did the following:

i. Ignored scholarly papers offering different theories ii. Did not provide any technical basis iii. Used data selectively, regardless its applicability iv. Misrepresented the validity of model predictions

CHECWORKS underlying assumptions are wrong

• CHECWORKS Assumption #1: Non-conservative predictions, 50% of the time, are acceptable

- Not acceptable: model must be recalibrated when non conservative (NUREG-1801, Rev. 2)

• CHECWORKS Assumption #2: Component Chromium content is known

- a factor of six in Chromium uncertainty leads to a factor of 10 in FAC uncertainty

• CHECWORKS Assumption #3: Wall thinning by FAC is controlled solely by chemical dissolution

- Existing theoretical and experimental data indicate that FAC is controlled by both dissolution and erosion (for example, see reference: Digby D. Macdonald, The Point Defect Model for the Passive State, J. OF THE ELECTROCHEMICAL SOCIETY, Vol. 139, Issue No. 12 (Dec. 1992)); NRC findings in LR-ISG-2012-01 (Draft License Renewal Interim Staff Guidance, Wall Thinning Due to Erosion Mechanisms, Exhibit ENT000573) and its own data at IPEC.

• CHECWORKS Assumption #4: Average mass transfer coefficients are applicable

- ASME code requirements are based on maximum local rates (Tcrit - is the minimum local thickness)

• CHECWORKS Assumption #5: Corrosion rate equations derived from tests of electrochemically driven copper dissolution in hydrofluoric acid are applicable (reference: Bryan Poulson and Russel Robinson, The Use of A Corrosion Process to Obtain Mass Transfer Data, SCIENCE, Vol. 26, No.4, pp. 265-280 (1986))

- Such tests are not applicable to carbon steel in the reactor environment (reference: G.J. Bignold, K.

Garbett, R. Garnsey, & I.S. Woolsey, Erosion/Corrosion in Nuclear Steam Generators, in WATER CHEMISTRY OF NUCLEAR REACTOR SYSTEMS 5, Paper 1 (British Nuclear Engineering Society, 1980)).

CHECWORKS Predictions:

Non-conservative by as much as a factor of 10 CHECWORKS predictions are presented in a manner that hides the degree of non-conservatism, resulting in misleading plots.

• As shown in Figures 1A and 1B, interchanging the y and the x axes , increases the number of non-conservative points, exceeding the 50% line, by a factor of 6.

• The upper line is labeled +50% instead a +100%, misrepresenting CHECWORKS accuracy

Figure 1A -Shows how almost all the non-conservative data points appear to be bound by the -50% line 0 150 300 Measured Wear ( mils)

(

Reference:

RIV000111 - Hopenfeld Re-plotted CHECWORKS data)

Figure 1B - By interchanging the x and the y axes, the number of non-conservative points in Figure 1A, exceeding the 50%

line, was increased from 2 to 12 i.e., by a factor of 6 Predicted Wear (Mils)

(

Reference:

RIV000111 - Hopenfeld Re-plotted CHECWORKS data)

Local variation wall thinning shows that FAC is controlled by both corrosion and erosion: wrong Input to model Definitions: MTCRE: Maximum elbow to pipe mass transfer coefficient ratio = Maximum elbow to pipe metal loss MTCRE criteria can determine whether FAC is controlled by chemical dissolution or erosion

• Chemical dissolution controlled: MTCRE < 1.6

• Erosion controlled: MTCRE > 1.6 Calculations:

Elbow. Max. metal loss: (pt1) 1.446-1.325=0.121 (Pt2) 1.547-0.513=1.034 Straight pipe, Max. metal loss : 1.399-1.379=0.020 Metal loss Ratio, (Elbow /straight pipe):(0.121/.02) &

(1.034/.02) i.e. 6 to 52

Reference:

Entergy Ultrasonic Examination Report, IPEC00020853, at batestamp IPEC00020856

==

Conclusion:==

If metal loss were controlled by dissolution alone it would have not exceeded 1.6X; in fact it exceeded it by 6X to 52X, meaning that FAC is controlled by both chemical dissolution and erosion.

Reference:

Jianrong Wang, Siamack A. Shirazi, A CFD based correlation for mass transfer coefficient in elbows, INTL J. OF HEAT AND MASS TRANSFER, 44 (2001) 1817-1822.

CHECWORKS Methodology Results in Overly Large Grid Size Abrupt local thickness despite Entergy claim that FAC is uniform:

Reference:

Entergy Ultrasonic Examination Report, IPEC00020853, at batestamp IPEC00020858

CHECWORKS was designed to predict average thinning rates by corrosion, not local rates by either corrosion or erosion

• Elbows thickness varied from 0.257 to 0.059, (77%), more than 30 component grids exhibit large local variations, yet Entergy does not consider FAC as a local phenomenon (reference: Testimony of Entergy Witnesses Regarding Contention RK-TC-2, ENT000029, at p.54)

• Unpredictable component to component local thinning variation (geometry changes with time)

• CHECWORKS comparison of local thinning predictions with measurements was never provided

(

Reference:

RIV000049 - Entergy Indian Point U3, FAC, 3RF13 Outage, 2005)

CHECWORKS non-conservative predictions and the definition of FAC caused confusion; NRC issued new guidance

• NUREG-1801, Rev. 2, at XI M17-2, now requires that, when measurements show the predictions to be non-conservative, the model must be re-calibrated using the latest field data.

• Draft License Renewal Interim Staff Guidance, LR-ISG-2012-01, Wall Thinning Due to Erosion Mechanisms (Exhibit ENT000573),

revised the definition of wall thinning to include erosion mechanisms such as cavitation, flashing, droplet impingement, and solid particle impingement.

CHECWORKS is Not Used for Steam Generator Components

• 10 CFR § 50.49(b)(1) Requires AMP for steam generators

• Steam generators components such as inlet piping, nozzles, blow down piping, and distribution ring, are highly susceptible to FAC

• Entergy approach to Steam Generator AMP, When FAC Program inspections reveal wall thinning, that data is evaluated against the appropriate design loading conditions, including seismic loads is in direct violation of Part 50

Summary Issue CLAIMED BY ENTERGY FACTUAL NOTES ACCURACY +/- 50% + /- a factor of 2 As much as a factor of with respect to LCF (observed /predicted) 10 FAC Mechanism Dissolution only Consequences: More Probable:

wrong input Combined Chemical parameters and Mechanical Conservatism Conservative results 40-60% results are Code was not designed not conservative to be conservative Meet NUREG- YES NO Cannot be recalibrated 1801, Rev. 2 10 CFR § 50.49 Compliance by NO evaluating conseq. on discovering Tw<Tdes

CONCLUSION:

CHECWORKS does not meet Part 54 and NUREG-1801 requirements

• CHECWORKS can neither predict wall thinning for establishing inspection frequencies nor can it be used to identify new components for inspections.

• Because of its inherent design CHECWORKS can not be recalibrated:

- CHECWORKS predictions for the years 2000-2011 show that 40-60% of the time the code produced non-conservative results by as much as a factor of 10.

- CHECWORKS was designed to over-predict and under-predict data by 50%.

- Uncertainties in Chromium content input, would prevent any meaningful recalibration of the code.

- After 20 years of operations CHECWORKS remains uncalibrated.

- The inputs on the effects of geometry and velocity on FAC rates are incorrect because they were based on tests of copper in an acid. In aggressive acids, the dominant damage is by chemical dissolution; these tests are not applicable to steel in water.

• Entergy cannot meet Part 50, Part 54 and NUREG-1801 requirements because it does not have a predictive methodology as required by NUREG-1801, Rev. 2:

- Monitoring and Trending: CHECWORKS or a similar predictive code is used to predict component degradation in the systems conducive to FAC;

- Acceptance Criteria: Inspection results are input for a predictive computer code, such as CHECWORKS, to calculate the number of refueling or operating cycles remaining before the component reaches the minimum allowable wall thickness;

- NSAC 202L-R3 The purpose of quantitative analysis of using predictive methodology such as CHECWORKS is to predict the FAC wear rate and to determine the remaining service life for each piping component, including uninspected components.