ML15252A505

NYS000569 - Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr. in Support of Contention NYS-26B/RK-TC-1B (Public, Redacted) (September 9, 2015)

ML15252A505
Person / Time
Site:	Indian Point
Issue date:	09/09/2015
From:	Lahey R Rensselaer Polytechnic Institute
To:	Atomic Safety and Licensing Board Panel
SECY RAS
References
RAS 28272, ASLBP 07-858-03-LR-BD01, 50-247-LR, 50-286-LR
Download: ML15252A505 (33)

Text

UNITED STATES 1 NUCLEAR REGULATORY COMMISSION 2 BEFORE THE ATOMIC SAFETY AND LICENSING BOARD 3 -----------------------------------x 4 In re: Docket Nos. 50

-247-LR; 50-286-LR 5 License Renewal Application Submitted by ASLBP No. 07

-858-03-LR-BD01 6 Entergy Nuclear Indian Point 2, LLC, DPR-26, DPR-64 7 Entergy Nuclear Indian Point 3, LLC, and 8 Entergy Nuclear Operations, Inc.

September 9 , 2015 9 -----------------------------------x 10 PRE-FILED SUPPLEMENTAL REPLY WRITTEN TESTIMONY OF 11 Dr. RICHARD T. LAHEY, JR.

12 REGARDING CONSOLIDATED CONTENTION NYS-26B/RK-TC-1B 13 On behalf of the State of New York ("NYS" or "the State"), 14 the Office of the Attorney General hereby submits the following 15 testimony by RICHARD T. LAHEY, JR., PhD. regarding Consolidated 16 Contention NYS-26B/RK-TC-1B.

17 Q. Please state your full name.

18 A. Richard T. Lahey, Jr.

19 Q. By whom are you employed and what is your position?

20 A. I am retired and am currently the Edward E. Hood 21 Professor Emeritus of Engineering at Rensselaer Polytechnic 22 Institute (RPI), which is located in Troy, New York.

23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 1

Q. Have you previously summarized your educational and 1 professional qualifications?

2 A. Yes, my education and professional qualifications and 3 experience are described in my Curricula Vitae and previously 4 filed testimony in this proceeding.

5 Q. I show you what has been marked as Exhibit ENT000616 6 and ENT000679. Do you recognize those documents?

7 A. Yes. They are copies of the pre-filed testimony of 8 the witnesses for Entergy on Contentions NYS-25 and NYS-26B/RK-9 TC-1B that were submitted in August 2015.

10 Q. I show you what has been marked as Exhibit NRC000197 11 and NRC000168. Do you recognize those documents?

12 A. Yes. They are copies of the pre-filed testimony of 13 NRC Staff witness that were submitted in August 2015. NRC000168 14 concerns Contention NYS-26B/RK-TC-1B, and NRC000197 concerns 15 Contention NYS-25. (I note that portions of those two USNRC 16 submissions also discuss Contention NYS-38/RK-TC-5, which I will 17 discuss separately.)

18 Q. Have you had an opportunity to review ENT000616, 19 ENT000679, NRC000168, and NRC000197?

20 A. Yes.

21 Q. Has Entergy's and the USNRC Staff's August pre-filed 22 testimony caused you to change the testimony and opinions that 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 2

you have previously submitted in this proceeding in connection 1 with Contention NYS-25 and Contention NYS-26B/RK-TC-1B?

2 A. In general, no. Entergy and the USNRC Staff have 3 failed to resolve the age-related safety concerns that I have 4 raised throughout this relicensing proceeding. They continue to 5 approach various aging mechanisms in silos, without addressing 6 the potential synergistic interactions between multiple 7 degradation mechanisms, and my related safety concerns.

8 Q. According to the USNRC Staff the 9 "Expanded Materials Degradation Assessment" (EMDA) (NYS000484A-10 B) and Light Water Reactor Sustainability (LWRS) Program 11 (NYS000485) do not apply to IP2 and IP3 because they are 12 associated with subsequent license renewal from 60 to 80 years.

13 (NRC000168, at A176, A178; Do you agree?

14 A. Absolutely not. The EMDA and LWRS programs study 15 aging degradation mechanisms that affect all licensed nuclear 16 reactors. The effects studied in the EMDA and LWRS programs are 17 also quite relevant to the safe extended operation of nuclear 18 reactors out to 60 years. The size and cost of these research 19 programs reveals the importance placed on studying the various 20 PWR aging concerns that I have previously raised.

21 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 3

Q. Do you agree with the USNRC Staff that 1 irradiation embrittlement can increase a component's fatigue 2 life? 3 A. There are some data which show that irradiation 4 embrittlement can increase fatigue life in certain situations, 5 especially where high cycle, low amplitude, fatigue occurs.

6 However, there are other data which show that for low cycle, 7 high amplitude fatigue the effects of embrittlement decreases 8 fatigue life by reducing the number of cycles to failure (N f). 9 As the USNRC Staff concedes, the data regarding the effects of 10 irradiation embrittlement on fatigue life are currently 11 incomplete and inconclusive. NRC000168, at A154; NRC000197, at 12 A196, A200. However, in the face of this uncertainty, Entergy 13 and the USNRC Staff simply assume that the effect can be 14 ignored, and that the Indian Point reactors can continue to be 15 operated until any synergistic degradation effects are directly 16 observed in the operating plants.

NRC000197, at A204 ("If 17 synergistic effects of aging mechanisms were to occur, the 18 resulting degradation will likely be found in at least one plant 19 in the fleet.") This is exactly what I mean when I say that 20 Entergy and the USNRC Staff have taken a "wait-and-see" 21 approach. Unfortunately, such an approach could lead to a 22 component failure and potentially serious consequences that 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 4

cannot be easily addressed after such a failure has occurred. A 1 much more prudent approach would be to apply an uncertainty 2 factor, or penalty, to the CUF en calculations, and to repair or 3 replace components with very high projected CUF en values. This 4 is a well-accepted engineering practice; indeed, the ASME Code 5 applies a similar uncertainty penalty to CUF en calculations 6 (i.e., a factor of 2 on stress or 20 cycles) to account for 7 uncertainties in test data, which were obtained from tests on 8 small scale, polished metal samples in air, rather than on 9 actual industrial structures, components and fittings in a 10 reactor environment (e.g., in a PWR).

11 12 13 14 A. 15 16 The operative document for management of aging 17 degradation effects on RVI components at Indian Point is the 18 Revised and Amended RVI Plan, which relies on MRP-227-A and 19 which was the subject of the Second Supplemental Safety 20 Evaluation Report for Indian Point (NUREG-1930, Supplement 2) 21 (Exh. NYS000507). I remain concerned with the adequacy of the 22 Revised and Amended RVI Plan, because it relies entirely on 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 5

inspections to determine the condition of RVI components. The 1 absence of detectable surface cracks does not necessarily mean 2 that the embrittled and fatigue-weakened structures, components 3 and fittings are not vulnerable to early (i.e., CUF en < 1.0) 4 failures. As MRP-227-A concedes, inspections cannot determine 5 the existence and extent of embrittlement. Accordingly, the 6 Revised and Amended RVI Plan does not account for the 7 possibility that embrittled and fatigue-weakened RVI components 8 could be subject to a shock load which would cause them to fail 9 suddenly.

10 Q. Do you agree with the USNRC Staff that 11 MRP-227-A inspections, coupled with environmentally assisted 12 fatigue calculations resulting in a CUF en of less than 1.0, are 13 adequate to manage the effects of aging on RVI components?

14 A. No. This is a perfect example of the type of "silo" 15 thinking that I am concerned about. According to the USNRC 16 Staff , embrittlement and the associated aging 17 effects can be managed through the inspection-based Revised and 18 Amended RVI Plan, while fatigue is managed through a separate 19 Fatigue Management Plan (FMP), and no further consideration of 20 the interaction between multiple aging mechanisms is necessary.

21 NRC000168, at A185. For example, the USNRC Staff argues that 22 the portions of my June 2015 testimony (NYS000530) relating to 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 6

embrittlement are not relevant to the management of metal 1 fatigue, even though I have repeatedly argued that embrittlement 2 effects should be considered when assessing component fatigue 3 life. NRC000168, at A168, A170. As long as CUF en is calculated 4 to be less than 1.0, the USNRC Staff apparently 5 believe that no critical structures, components, and fittings 6 will exhibit fatigue-induced surface cracking and therefore the 7 effects of embrittlement need not be considered in the fatigue 8 calculation. NRC000168, at A153. Their approach, however, 9 fails to recognize certain basic realities. In particular, the 10 CUF for a structure, component or fitting is defined as the 11 number of fatigue cycles it is expected to experience (N) 12 divided by the number of cycles to failure (N f). The N f for a 13 new, ductile material can be significantly greater than the N f 14 for a highly embrittled material; indeed, this is known to be 15 true for large amplitude, low cycle fatigue. See e.g. Kanaski, 16 et al., "Fatigue and Stress Corrosion Cracking Behaviors of 17 Irradiated Stainless Steels in PWR Primary Water," ICONE-5, at 18 2372 (May 1997) (Exh. NRC000177); Arai, et al., "Irradiation 19 Embrittlement of PWR Internals," Proceedings ASME/JSME 2d 20 International Nuclear Engineering Conference, Vol. 2, at 103 21 (1993) (Exh. NYS000564); Korth, G.E. & Harper, M.D., "Effects of 22 Neutron Radiation on the Fatigue and Creep/Fatigue Behavior of 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 7

Type 308 Stainless Steel Weld Materials at Elevated 1 Temperatures," Proceedings of the 7 th International Symposium on 2 the Effects of Radiation on Structural Materials, Gatlinburg, TN 3 (June 1974) (Exh. RIV000152). Additionally, embrittled and 4 fatigue-weakened structures may not be able to tolerate 5 significant seismic and shock loads as well as fully ductile 6 structures can. Thus, when the effects of embrittlement and 7 fatigue are considered together, there is a real risk that 8 components will fail before their calculated CUF en value reaches 9 unity. 10 Q. I show you a document marked as Exhibit [NYS000566]

11 that is entitled, Figure 1: "Comparison of Limit Line and Best 12 Estimate Predictions with Embrittlement and Without 13 Embrittlement." Are you familiar with this Figure?

14 A. Yes. 15 Q. How are you familiar with this Figure?

16 I developed Figure-1 in connection with my review of 17 Entergy's and the USNRC's Revised Statements of Position and 18 Revised Testimony in this proceeding.

19 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 8

Figure 1: Comparison of Limit Line (LL) and Best Estimate (BE) predictions, with embrittlement (BE e) and without embrittlement (BE ne). Note: (1) BE ne predicts possible failure at end of life (EOL).

(2) BE e predicts failure (CUFen = 1.0) before end of life.

(3) BE e predicts possible failure well before end of life.

Q. Why did you prepare this Figure?

1 A. I prepared this Figure to visually illustrate some of 2 the concerns that I have set forth in my testimony in this 3 proceeding. Specifically, this Figure shows three important 4 things: (1) a typical WESTEMS TM "Limit Line" prediction, (2) a 5 "Best Estimate" prediction of CUF en , the cumulative usage factor 6 under reactor operating conditions, and the associated 7 BE ne 1017 FLUENCE (n/cm 2) CUF en Limit Line (LL)

Best Estimate (BE)

(3) BE e 1.0 (2) (1) - EOL (PEO) TIME Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 9

uncertainty , and, (3) how the neutron fluence-1 induced embrittlement of a RVI structure, component, or fitting 2 can dramatically increase the possibility of a fatigue failure, 3 as represented by the cumulative usage factor under PWR 4 operating conditions, CUF en , being equal to unity. This figure 5 applies to RVI structures, components, and fittings made of 6 stainless steel or other materials. In summary, in Figure-1, I 7 have illustrated two different approaches for calculating a 8 time-dependent CUF en , namely the "Limit Line" and the "Best 9 Estimate" calculation. For the "Best Estimate" line, I also 10 11 effects of fatigue on embrittled (e) and on non-embrittled (ne) 12 RVIs. 13 Q. Please describe what is represented by the horizontal 14 x-axis (i.e., the abscissa) in Figure 1.

15 A. S t (where 16 (n/cm 2-s) and t is the time of operation of the reactor) the x-17 axis tracks both the fluence ( and the time (t). As time 18 passes, each RVI reaches the reactor's End of Life (EOL) for the 19 Period of Extended Operation (PEO). Also, as we move from left-20 to-right along the x-axis (i.e., as the time of reactor 21 operation increases), we see an increase in the number of 22 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 10 fatigue cycles (N) that the various RVIs (one of which has been 1 chosen to be displayed in Figure-1) have been subjected to.

2 Q. Please describe what is represented by the vertical y-3 axis (i.e., the ordinate) in Figure 1.

4 A. The y-axis displays CUF en , the cumulative usage factor, 5 considering PWR environmental conditions. At a nuclear power 6 plant, the maximum number of fatigue cycles that can be 7 experienced by any structure, component or fitting must always 8 result in a CUF en of less than 1.0. That is, the number of 9 actual fatigue cycles (N) should always be less than the number 10 of allowable cycles (N f) to avoid failure of the RVI.

11 Q. In Figure-1 what does the "Limit Line" represent?

12 A. The "Limit Line" represents 13 a supposedly-conservative 14 calculation of CUF en for a hypothetical RVI component or 15 structure.

16 17 The "Limit Line" shown in Figure-1 includes 18 Entergy's implicit assumption that a stainless steel RVI 19 structure, component, or fitting remains perfectly ductile, and 20 thus shows that the CUF en for a RVI will increase with an 21 increasing number of fatigue cycles, but that it is not affected 22 by the fluence. The "Limit Line", therefore, uniformly 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 11 increases in CUF en even after the fluence exceeds 10 17 n/cm 2 , 1 where significant embrittlement begins to develop (i.e., it is 2 assumed that the number of cycles to failure are not a function 3 of fluence). Figure-1 shows that, as the RVI component 4 approaches the end of life (EOL) for the period of extended 5 operation (PEO), "Limit Line" will approach 6 (but still be below) CUF en = 1.0. This is depicted as location 7 (1). 8 Q. In Figure-1 what does the "Best Estimate" line 9 represent?

10 A. The "Best Estimate" plot depicts what would be, for 11 example, a WESTEMS prediction of CUF en in which conservative 12 assumptions are made and my concerns 13 14 have been 15 addressed. In addition, the "Best Estimate" plot in Figure-1 16 also includes a prediction of a "propagation of errors" type 17 analysis (i.e., an uncertainty analysis), which can, and should, 18 be done.

19 Q. What does a "propagation of errors" analysis include?

20 A. This type of analysis considers important parameters 21 that have some uncertainty associated with them; for example, 22 the coarseness of the computational mesh, uncertainties in F en , 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 12 the "best estimate" local heat transfer coefficients, etc.;

1 uncertainties which, in turn, impact the RVI's actual fatigue 2 life (i.e., CUF en). The net uncertainty intervals (

3 account for uncertainties in the number of transients 4 (including, for example, seismic events), and synergistic 5 effects of radiation (i.e., the fluence) and stress corrosion 6 effects on metal fatigue. Accounting for the uncertainty in 7 these parameters allows one to estimate the overall uncertainty 8 of the "Best Estimate" CUF en prediction, so that we can see 9 if the "Limit Line" predictions are indeed 10 sufficiently conservative. That is, comparing the "Best 11 Estimate" results, p predictions 12 reveals whether the "Limit Line" approach 13 is adequate or 14 not. 15 Q. Can you indicate how one might perform a "propagation 16 of error" analysis?

17 A. Yes, uncertainty analyses can be performed in a number 18 of ways. One of the most commonly used methods by engineers is 19 the method proposed by Kline & McClintock [ASME J. Mechanical 20 Engineering, Vol.75, No.1, 3-8, Jan. 1953] (Exh. NYS000514).

21 For the case in which the "Best Estimate" value of CUF en involves 22 "N" parameters, each having some uncertainty associated with 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 13 them (e.g., the Dittus-Boelter convective heat transfer, 1 has an inherent uncertainty of about 25%), 2 the net unce given by:

3 4 5 This approach is widely used by engineers since it is relatively 6 easy to evaluate and it gives an acceptable approximation of the 7 "error bars" in a "Best Estimate" evaluation.

8 Q. In Figure-1, the "Best Estimate" line is below the 9 "Limit Line." Does this mean that everything is OK?

10 A. Not necessarily. As shown in Figure-1, a "Best 11 Estimate" prediction plus the 12 possible RVI structure, components or fitting failures (i.e., 13 CUF en = 1.0), sooner than the "Limit Line" would. Moreover, the 14 effect of embrittlement can make the situation much worse. The 15 problem is that no one knows how much conservatism, if any, is 16 in the "Limit Line". As a consequence, we can have no 17 confidence that "Limit Line" results near CUF en = 1.0, are 18 sufficiently conservative to bound all the possible 19 uncertainties, as illustrated in Figure-1 by . 20 Q. In Figure-1, the "Best Estimate" line is shown for the 21 case in which there is no effect of embrittlement on fatigue, 22 BE ne , and the case in 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 14 which embrittlement degrades fatigue life, BE

e. Can you explain 1 what these lines mean?

2 A. Yes, the plot depicts two "Best Estimate" predictions 3 for a hypothetical RVI component: one under conditions of no 4 embrittlement (BE ne), for which the "Best Estimate" CUF en line 5 continues to increase as the fatigue-inducing cycles (N) 6 accumulate with time, and the other "Best Estimate" prediction 7 (BE e) which includes the effect of embrittlement on the (now 8 time-dependent) maximum allowable cycles (N f). For the latter 9 case, increases in fluence result in more embrittlement and thus 10 a more rapid increase in CUF en than in the former case. This is 11 because after a fluence of about 10 17 n/cm 2 , significant 12 irradiation-induced damage and embrittlement begin to occur.

13 Embrittlement results in the loss of facture toughness and the 14 loss of ductility. Even though the data taken to date have been 15 inconclusive, there is ample evidence that this embrittlement 16 may reduce the number of fatigue cycles to failure (N f) and thus 17 increase the corresponding CUF en , which is defined as CUF en =. 18 Q. What does location (1)in Figure-1 show?

19 A. Location (1) shows that, even if embrittlement is not 20 considered, RVI fatigue failure may occur by the end of life for 21 the period of extended operation. This is because the sum of a 22 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 15 "Best Estimate" prediction plus the uncertainty in 1 associated with this prediction would exceed CUF en = 1.0. In this 2 case the "Limit Line" result is clearly non-conservative.

3 Q. May I now direct your attention to location (2) in 4 Figure-1. What does this location show?

5 A. Location (2) shows that embrittlement can result in 6 the physical failure of the RVI well before the end of life for 7 the period of extended operation.

8 Q. May I direct your attention to location (3) in Figure-9 1. What does this location show?

10 A. At location (3), the "Best Estimate" plus uncertainty 11 e) predicts 12 possible failure of the hypothetical RVI component well before 13 the end of life for the period of extended operation, even 14 though the "Limit Line" prediction indicates substantial margin 15 to failure at this point. That is, at location (3) the CUF en for 16 the "Best Estimate" line, accounting for embrittlement and for 17 all uncertainties (i.e., BE e

s unity. Moreover, as I 18 have stated in my previous testimony, highly embrittled RVIs may 19 fail even sooner than by fatigue alone if significant seismic or 20 shock loads occur. That is, fatigue-weakened and embrittled 21 structures cannot tolerate large impulsive seismic and shock 22 loads like fully ductile structures can. Moreover, other 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 16 synergistic aging effects, such as the thermal embrittlement(TE) 1 of some CASS RVIs (the upper part of the core support columns) 2 would may move the BE e curve to the left of its position shown in 3 Figure-1 (i.e., the RVI would be predicted to reach failure, 4 CUF en = 1.0, even sooner than the cases (2) and (3) in Figure-1).

5 Q. Can any conclusions about the "Limit Line" and "Best 6 Estimate" lines be drawn from Figure 1?

7 A. Yes. Because of Entergy's failure to explicitly 8 account for uncertainties, and for the effect of embrittlement 9 on fatigued RVIs, we can have no confidence in the "Limit Line" 10 type of CUF en predictions 11 12 13 14 15 16 17 18 19 Q. In response to Dr. Hopenfeld's testimony, the USNRC 20 Staff stated that "CUF or CUF en analyses are not required for 21 safety assessments of DBA events . . . because CUF and CUF en , 22 which are indicators of possible fatigue crack initiation, are 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 17 not a significant contributor to safety during DBA events."

1 NRC000168, at A145. How do you respond?

2 A. Again, this is a perfect example of "silo" thinking.

3 The USNRC Staff refuses to assess the risk that a highly 4 fatigued system, component or fitting which has been fatigue-5 weakened and embrittled as a result of neutron fluence, for 6 example, could fail, relocate, and degrade core cooling, when it 7 is subjected to a DBA LOCA or some other significant shock load.

8 This approach ignores the reality of a reactor environment, 9 where multiple aging mechanisms act simultaneously on RVIs and 10 other components, and that this age-related degradation needs to 11 be taken into account in plant safety analyses.

12 Q. 13 14 15 16 17 18 19 A. As I have stated above, ignoring the 20 effects of accident loads on aging RVI components is not at all 21 appropriate. While it is probably true that fatigued and 22 embrittled RVI components may operate normally during steady-23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 18 state operations, they can fail suddenly and catastrophically 1 when subjected to a significant shock load. Without considering 2 accident-induced loads, Entergy cannot provide adequate 3 assurance that the RVIs will continue to perform their functions 4 during the period of extended operation (PEO).

5 Q. Have there been RVI component failures at nuclear 6 reactors in the past?

7 A. Yes. Baffle former bolts and clevis insert bolts have 8 failed at several PWRs in the past, clearly demonstrating that 9 inspections alone will not detect aging effects prior to 10 component failure.

11 Q. With respect to the clevis insert bolt failures, do 12 you agree with the USNRC Staff's testimony that "[t]he failed 13 bolts were detected visually" (A290)?

14 A. Partially. According the SSER2 at 3-25 (Exh.

15 NYS000507), only 7 out of 29 damaged bolts, or about 24%, were 16 detected visually. The vast majority were not detected 17 visually. Nonetheless, Entergy has proposed to conduct only 18 visual inspections of clevis insert bolts. Exh. NYS000496, at 19 51. 20 Q. 21 22 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 19 1 2 3 A. the RVI 4 AMP does not consider the reality of conditions within the 5 reactor. All of the components in and around the clevis insert 6 bolts are also undergoing a range of aging mechanisms which may 7 affect their functionality or their ability to withstand a 8 sudden shock load.

9 10 11 12 13 14 15 16 Q. 17 18 19 20 A. I do not believe that this is a safe or 21 responsible approach to aging management. Although, due to 22 baffle bolt redundancy, the plant might be able to function 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 20 properly during steady-state operations, when a large percentage 1 of the incore bolting has failed, a significant shock load could 2 cause many of the remaining bolts to suddenly fail, resulting in 3 the relocation of core components and the possible loss of a 4 coolable core geometry.

5 Q. When discussing aging management of the baffle-former 6 bolts, the USNRC Staff describes the number of baffle-former 7 bolts in "[t]hree-loop Westinghouse design PWRs like IP2 and 8 IP3[.]" NRC000197, at A243. Is it accurate to say that IP2 and 9 IP3 are three-loop PWRs?

10 A. No. IP2 and IP3 are actually four-loop Westinghouse 11 designed PWRs. It is not clear why the USNRC Staff believes the 12 IP2 and IP3 reactors use a three-loop design. Perhaps this 13 reference is a "cut and paste" remnant from another proceeding 14 for a different facility that USNRC Staff has reused here.

15 Q. 16 17 18 19 20 21 22 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 21 A. Virtually all metals (embrittled or not), 1 experience a decrease in ductility as the temperature decreases.

2 However, irradiated carbon steel undergoes a very sharp decrease 3 in ductility at a fluence-dependent temperature commonly called 4 the nil ductility temperatures (NDT), while stainless steel does 5 not. Nevertheless, as I have said before, sufficiently strong 6 shock loads can lead to failure of highly fatigue-weakened and 7 embrittled stainless steel RVIs.

8 Q. 9 10 11 A. 12 13 14 15 16 17 18 19 20 21 22 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 22 1 2 3 4 The USNRC 5 Staff concedes that "[g]iven the variability in assumptions made 6 by different analysts, it is difficult to explicitly quantify 7 the exact overall safety margin present in fatigue 8 calculations." NRC000168, at A210.

9 10 However, given the critical 11 siting of the Indian Point plants in the New York metropolitan 12 area, it is especially appropriate to identify and verify the 13 assumptions and calculations. Indeed, as distilled by the 14 concept of "trust but verify," verification precedes trust.

15 Q. Do you agree with the USNRC Staff that a 16 "propagation of errors" analysis (i.e., an uncertainty 17 analysis), is not needed for EAF calculations?

18 A. No. the USNRC Staff claim that an 19 uncertainty analysis is not necessary because the EAF 20 calculation is "deterministic" and contains adequate 21 conservatisms. NRC000168, at A171; 22 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 23 1 2 3 4 5 6 7 8 9 10 Q. So far you have been talking about reactor vessel 11 internals (RVIs). Do you have some similar concerns about the 12 fatigue evaluations for primary pressure boundary components and 13 fittings?

14 A. Yes, except for the fact that radiation-induced 15 embrittlement (IE) is not normally an issue, virtually all of my 16 previously discussed concerns are the same.

17 18 19 20 21 22 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 24 1 2 Q. the USNRC Staff state that Entergy will 3 compare the actual transients experienced at Indian Point with 4 the transients anticipated in the EAF calculations. Does this 5 approach address your concerns?

6 A. No. The number of fatigue cycles is obviously 7 important; however, my principal concern has never been that 8 components will experience a greater number of transient cycles 9 than anticipated in the EAF calculation. In contrast, my real 10 concern for the integrity of primary pressure boundary 11 components is that a significant shock load - caused by a 12 seismic, LOCA, or other event - could cause the fatigue-weakened 13 component to suddenly fail well before their associated CUF en 14 value reaches 1.0.

15 16 17 18 19 20 21 22 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 25 A. In my opinion, an observable surface crack of 1 3mm or greater, which would be expected when CUF en reaches 1.0, 2 is quite 'significant'. However, microscopic cracks exist and 3 propagate within the metal structures, components and fittings 4 even when CUF en < 1.0. These microscopic cracks cannot be 5 detected by non-destructive testing (NDT), but they can weaken 6 the components and make them more susceptible to failures during 7 a significant seismic or pressure/thermal shock load event, 8 especially when coupled with embrittlement due various thermal 9 embrittlement or corrosion mechanisms.

10 11 12 13 14 15 A. what I 16 mean by not reducing design and safety margins as the plant ages 17 is that component fatigue life calculations should retain the 18 original design conservatisms and that the components should be 19 repaired or replaced if they exceed acceptable design margins 20 during the period of extended operations.

21 22 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 26 In my opinion, removing 1 modeling conservatisms as IP2 and IP3 exceed 40 years of 2 operation, and the components become more and more degraded, is 3 both irresponsible and dangerous.

4 Q. 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 27

Q. I show you what has been marked as Exhibit NYS000563.

1 Do you recognize it?

2 A. Yes, it is an excerpt from an USNRC training manual, 3 entitled "Reactor Concepts Manual - PWR Systems." It was 4 prepared by the USNRC Technical Training Center.

5 Q. What information did you draw from this USNRC 6 document?

7 A. This document identifies the relative location of 8 various components in a 4-loop, Westinghouse, pressurized water 9 reactor (PWR). On page 4-25, it shows that one of the 10 accumulators uses the same nozzle as the residual heat removal 11 (RHR) system, the low pressure coolant injection (LPCI), and the 12 intermediate pressure coolant injection (IPCI) safety systems.

13 In contrast, the high pressure coolant injection (HPCI) system 14 has a separate connection on the same cold leg as the 15 RHR/Accumulator nozzle.

16 Q. Why is that relevant?

17 A. 18 19 20 21 22 the components could fail 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 29 if subjected to a significant seismic event or shock load. As I 1 have pointed out in some of my previous ASLB testimony, the 2 failure of this important primary pressure boundary structure 3 could lead to a LOCA event in which the Accumulator, LPCI and 4 IPCI systems are not be able to inject water into the reactor's 5 cold leg, and thus into the core, which, in turn, could lead to 6 core melting.

7 Q. Do you have particular concerns with respect to any 8 other components?

9 A. Yes. 10 11 12 13 14 or 15 16 17 18 19 20 21 22 23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 30 1 2 3 4 5 6 7 8 9 10 11 12 13 the USNRC 14 Staff has indicated that if transients accumulate "at a rate 15 greater than the rate assumed in the fatigue calculation,"

16 Entergy would be permitted to conduct yet another "more refined 17 analysis." NRC000168, at A106. In short, there is apparently 18 no limit to the number of times CUF en can be recalculated to 19 obtain a CUF en result less than unity, and there is no standard 20 which defines the amount of conservatism that must be retained 21 in these calculations.

22 Q. Does this complete your testimony?

23 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 31 A. Yes, it does. I do, however, reserve the right to 1 supplement my testimony if new information is disclosed or 2 introduced.

3 4 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 32 UNITED STATES 1 NUCLEAR REGULATORY COMMISSION 2 BEFORE THE ATOMIC SAFETY AND LICENSING BOARD 3 -----------------------------------x 4 In re: Docket Nos. 50

-247-LR; 50-286-LR 5 License Renewal Application Submitted by ASLBP No. 07

-858-03-LR-BD01 6 Entergy Nuclear Indian Point 2, LLC, DPR-26, DPR-64 7 Entergy Nuclear Indian Point 3, LLC, and 8 Entergy Nuclear Operations, Inc.

September 9 , 2015 9 -----------------------------------x 10 DECLARATION OF RICHARD T. LAHEY, JR.

11 I, Richard T. Lahey, Jr., do hereby declare under penalty 12 of perjury that my statements in the foregoing testimony and my 13 statement of professional qualifications are true and correct to 14 the best of my knowledge and belief.

15 Executed in Accord with 10 C.F.R. § 2.304(d) 16 17 _________________________

18 Dr. Richard T. Lahey, Jr.

19 The Edward E. Hood Professor Emeritus of Engineering 20 Rensselaer Polytechnic Institute, Troy, NY 12180 21 (518) 495-3884, laheyr@rpi.edu 22 Pre-filed Supplemental Reply Testimony of Richard T. Lahey, Jr.

Consolidated Contention NYS-26B/RK-TC-1B 33