ML061810167

2006/06/23-Motion for Leave to Supplement the Petition and Petition to Add a New Contention, with Citizen'S Exhibits NC1 to NC10

ML061810167
Person / Time
Site:	Oyster Creek
Issue date:	06/23/2006
From:	Webster R Grandmothers, Mothers & More for Energy Safety, Jersey Shore Nuclear Watch, New Jersey Environmental Federation, New Jersey Public Interest Research Group (NJPIRG), Nuclear Information & Resource Service (NIRS), Rutgers Environmental Law Clinic, Sierra Club, New Jersey Chapter
To:	Abramson P, Anthony Baratta, Hawkens E Atomic Safety and Licensing Board Panel
Byrdsong A T
References
50-219-LR, ASLBP 06-844-01-LR, RAS 11886
Download: ML061810167 (446)

Text

{{#Wiki_filter:ii~a'i~ DOCKETED USNRC UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION June 23, 2006 (3:55pm) OFFICE OF THE SECRETARY OFFICE OF SECRETARY RULEMAKINGS AND ATOMIC SAFETY AND LICENSING BOARD ADJUDICATIONS STAFF Before Administrative Judges: E. Roy Hawkens, Chair Dr. Paul B. Abramson Dr. Anthony J. Baratta In the Matter of )

                                                                  )    Docket No. 50-0219-LR AMERGEN ENERGY COMPANY, LLC                                  )
                                                                 )     ASLB No. 06-844-01-LR (License Renewal for the Oyster Creek                        )

Nuclear Generating Station) ) June 23, 2006

                                                                 )

MOTION FOR LEAVE TO SUPPLEMENT THE PETITION Nuclear Information and Resource Service, Jersey Shore Nuclear Watch, Inc., Grandmothers, Mothers and More for Energy Safety, New Jersey Public Interest Research Group, New Jersey Sierra Club, and New Jersey Environmental Federation (collectively "Citizens" or "Petitioners") submit this Motion because AmerGen provided additional commitments and information as part of the license renewal process on June 20, 2006. The Atomic Safety and Licensing Board (the "Board") in its Order of June 6, 2006 invited Citizens to petition to add a new contention and directed Citizens to limit their argument to AmerGen's docketed commitment of April 4, 2006. LBP-06-16 at 9 (Jun. 6, 2006) (unpublished). Petitioners were given until June 26, 2006 to do so. Id. However, on June 20, 2006, AmerGen provided NRC with supplemental information concerning their aging management program for the Oyster Creek drywell shell during the license 1 Temp ov

4 renewal period as well as additional commitments. Letter from Gallagher to NRC, dated June 20, 2006. Citizens request leave to submit a supplement to their Petition to address AmerGen's new commitments and the new information provided in the June 20, 2006 letter. In addition, Citizens request the Board to order AmerGen and NRC Staff to respond to both the Petition and the supplement together ten days after the supplement is filed. This would allow the Board to evaluate the most current status of the dispute between AmerGen and Citizens, and would avoid needless duplicative rounds of filings. For the foregoing reasons, the Board should grant leave for Petitioners to supplement the current pleading and order AmerGen and NRC to respond to the Petition and the supplement at the same time. Respectfully submitted Richard Webster, Esq RUTGERS ENVIRONMENTAL LAW CLINIC Attorneys for Petitioners Dated: June 23, 2006 2

DOCKETED USNRC UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION June 23, 2006 (3:55pm) OFFICE OF THE SECRETARY OFFICE OF SECRETARY RULEMAKINGS AND ATOMIC SAFETY AND LICENSING BOARD ADJUDICATIONS STAFF Before Administrative Judges: E. Roy Hawkens, Chair Dr. Paul B. Abramson Dr. Anthony J. Baratta In the Matter of )

                                                              )    Docket No. 50-0219-LR AMERGEN ENERGY COMPANY, LLC                                    )
                                                              )    ASLB No. 06-844-01-LR (License Renewal for the Oyster Creek                          )

Nuclear Generating Station) ) June 23, 2006

                                                              .)

PETITION TO ADD A NEW CONTENTION PRELIMINARY STATEMENT Nuclear Information and Resource Service, Jersey Shore Nuclear Watch, Inc., Grandmothers, Mothers and More for Energy Safety, New Jersey Public Interest Research Group, New Jersey Sierra Club, and New Jersey Environmental Federation (collectively "Citizens" or "Petitioners") submit this Petition at the invitation of the Atomic Safety and Licensing Board ("ASLB") in its decision of June 6, 2006 in this proceeding. In accordance with that decision, Citizens now seek to add a new contention alleging that AmerGen Energy Co. LLC ("AmerGen") must set forth a monitoring program for the sand bed region of the drywell shell that ensures that adequate safety margins are maintained throughout the licensing period, and that it has so far failed to do so. Citizens request a hearing on this issue in accordance with 10 C.F.R. § 2.309. BACKGROUND This proceeding concerns the aging of the steel containment vessel of the Oyster Creek Nuclear Generating Station that is termed the drywell shell. The shell provides containment in the I

event of an accident. The lower portion of the shell is spherical with an inside diameter of 70 feet. Ex. NC 8 at 47. It is free standing from an elevation of 8 feet 11.75 inches from the bottom. Id. at

40. For around 3 feet 4 inches above that level to elevation 12 feet 3 inches, the exterior of steel liner used to have sand supporting it, but the sand was removed 1992. Id. at 47-48. This exterior portion of the drywell shell is termed the sand bed region. An interior floor is at elevation 10 feet 3 inches, id. at 47, and concrete curbs around the edge of the floor go up to the 11 foot elevation. Ex.

NC 10. In the sand bed region, the design thickness of the vessel was 1.154 inches. Ex. NC 8 at 40. Citizens initially contended that the testing of the extent of corrosion at all levels of the drywell shell proposed in AmerGen's license renewal application was inadequate to assure the continued integrity of this safety-critical structure for the period of the license extension. Petition at

3. To support this contention, Petitioners showed that the drywell shell is a safety-critical structure that acts both as a pressure boundary and as a structural support. Id. at 4. Petitioners then showed that water leakage onto the exterior of the drywell shell has caused significant corrosion, particularly in the sand bed region, where the N.R.C. regarded the corrosion as a "threat to drywell integrity." Id. at 4-6. Petitioners showed further that N.R.C. in 1986 regarded ultra-sonic testing of the sand bed region and other accessible areas of the drywell liner as "essential... for the life of the plant." Id. at 7.

Petitioners asserted that the potential for ongoing corrosion means that ongoing comprehensive testing is required to ensure the remaining razor-thin safety margins are met throughout any extended life of the plant. Indeed, Petitioners' Exhibit 5 at pages 8 and 12 showed that while AmerGen reported the "current thinnest" area to be 0.8 inches in December 1992, the actual thinnest areas are less than 0.736 inches, which was the original basis for evaluation. Multiple measurements in bays 1 and 13 and isolated measurements in bays 11, 15, and 17 were below 0.736 inches. Id. at 12. 2

The ASLB admitted a narrowed version of the initial contention pertaining to the need for ultrasonic ("UT") testing of the drywell in the sand bed region. LBP-06-07, 63 NRC 188 (2006). In that decision, the Board decided that Citizens had adequately demonstrated representational standing. LBP-06-07 at 3-6. Because this issue is resjudicata, this Petition does not address this issue further, but relies upon Citizens' previous accepted demonstration of standing. The ASLB recently found that a new commitment made by AmerGen on April 4, 2006 to use UT testing to verify the thickness of drywell shell in the sand bed region every ten years had rendered the initial contention moot. LBP-06-16 (June 6, 2006). The ASLB also invited Citizens to submit a new contention concerning the adequacy of AmerGen's newly proposed UT testing regime for the sand bed region. Id. at 9. Information that has become available since Citizens since filed the initial contention has now clarified many issues. For example, AmerGen has recently reported that over 20 areas in the sand bed region are now thinner than 0.736 inches and these areas have an average thickness of 0.703 inches. Ex. NC 2 at 13. In addition, the thinnest single measurement to date is 0.603 inches. Ex. NC 1 at 7. Citizens have also been able to discover the various acceptance criteria that are proposed, more details about the spatial scope of the monitoring, and how the results would be analyzed. To avoid repetition, this Petition presents the details of the support for the new contention in the Section on basis. ARGUMENT The proposed new contention satisfies the regulatory requirements by providing a specific statement of the contention, an explanation of basis, a demonstration that it is within the scope of the proceedings, and a demonstration of material issues that are in dispute. In addition, the proposed new contention is timely, because it is based on highly significant new information, including AmerGen's newly proposed testing regime. 3

A. Specific Statement of the Contention In order to bring a contention before the Commissioners, Citizens must "[p]rovide a specific statement of the issue of law or fact to be raised or controverted." 10 C.F.R. § 2.309(f)(1)(i). The contention is: AmerGen must provide an aging management plan for the sand bed region of the drywell shell that ensures that safety margins are maintained throughout the term of any extended license, but the proposed plan fails to do so because the acceptance criteria are inadequate, the monitoring frequency is too low and is not adaptive to possible future narrowing of the safety margins, the scope of the monitoring is insufficient to systematically identify and sufficiently test all the degraded areas of the shell in the sand bed region, the quality assurance for the measurements is inadequate, and the methods proposed to analyze the results are flawed. B. Explanation of Basis

1. Legal Requirements At this preliminary stage, Citizens do not have to submit admissible evidence to support their contention, rather they have to "[p]rovide a brief explanation of the basis for the contention,"

10 C.F.R. § 2.309(f)(1)(ii), and "a concise statement of the alleged facts or expert opinions which support the ... petitioner's position." 10 C.F.R. § 2.309(f)(1)(v). This rule ensures that "full adjudicatory hearings are triggered only by those able to proffer ... minimal factual and legal foundation in support of their contentions." In the Matter of Duke Energy Cor. (Oconee Nuclear Station, Units 1. 2. and 3) CLI-99-11, 49 N.R.C. 328, 334 (1999) (emphasis added). The Commission has clarified that, "an intervener need not.., prove its case at the contention stage. The factual support necessary to show a genuine dispute exists need not be in affidavit or formal evidentiary form, or be of the quality necessary to withstand a summary disposition motion." In the Matter of Georgia Institute of Technology, CLI-95-12, 42 N.R.C. 111, 118 (1995). Thus, the Commission has indicated that where petitioners make technically 4

meritorious contentions based upon diligent research and supported by valid information and expert opinion, the requirement for an adequate basis is more than satisfied.

2. Factual Issues Already Addressed By The ASLB Citizens already demonstrated a basis for their initial contention about the lack of adequate UT testing. The initial petition and documents supporting that contention are incorporated into this pleading by reference. As recognized by the ASLB in its decision admitting the initial contention, Citizens had ample basis for the following points:

i) the drywell shell is a safety structure, LBP-06-07 at 26; ii) water intruded into the sand bed region causing severe corrosion; id_ at 33. iii) water either is intruding, or could intrude in the future, leading to corrosive conditions on the outside of the drywell shell, id_ at 36; iv) the epoxy coating that was applied to protect the sand bed is now beyond its rated life and may be deteriorating, id_.at 31, 36; v) corrosion could occur even if the epoxy coating had not visibly deteriorated, id. at 36-37;

3. Proposed Monitoring For The Sand Bed Region Of The Drywell Shell AmerGen has recently committed to perform visual inspections of the epoxy coating once before the end of the licensing period, and every ten years thereafter. Letter from Michael P.

Gallagher, AmerGen, to NRC (Apr. 4, 2006). In addition, AmerGen has committed to performing UT measurements in the sand bed region at the same locations where UT measurements were conducted in 1996 prior to any license extension and at ten year intervals thereafter. Id. Statistically significant deviations from the 1992, 1994, and 1996 UT results will result in: i) performing additional confirmatory UT testing; ii) notifying the NRC within 48 hours of the identified condition; iii) conducting visual inspection of the external surfaces where corrosion may be occurring; 5

iv) performing an engineering evaluation to assess the extent of corrosion and whether additional inspections are required to assure drywell integrity; v) performing an operability determination and justification for operation until the next inspection.

4. Deficiencies In the Proposed Monitoring Regime As outlined in the contention and discussed in more detail below, Citizens have identified many deficiencies in the proposed monitoring regime. The NRC Staff also recently raised some similar issues regarding the accuracy of the previous results and the time between inspections.

AmerGen's response to Staff's concerns was filed on June 20, 2006. However, as instructed in the ASLB's June 6, 2006 decision dismissing the initial contention, Citizens have based this new contention on the April 4t commitment made by AmerGen. LBP-06-16 at 9. Because the June 20, 2006 AmerGen response amends AmerGen's commitments, Citizens are filing an accompanying motion to supplement this Petition in response to the new commitments. Turning to the substance, the proposed monitoring regime does not ensure that safety margins will be maintained throughout any renewed licensing period because the acceptance criteria are inadequate, the monitoring frequency is too low and is not adaptive to how close the shell thickness is to the acceptance criteria, degraded areas of the shell would not be systematically identified and sufficiently tested, the quality assurance for the measurements is inadequate, and the statistical techniques used in data analysis are flawed. This Section discusses these issues in detail. These identified deficiencies are safety-critical, because the sand bed region of the shell is severely corroded making margins of safety much thinner than when the plant was first built. To maintain safety, the monitoring regime must be able to predict how fast the metal could corrode to safety-critical levels, and must ensure that testing of areas that are closest to the margins occurs before there is any possibility that the metal has corroded too much. For example, in parts, over 0.5 inches of metal has corroded away from the steel drywell shell, leaving a metal thickness ofjust 6

over 0.6 inches. According to AmerGen, no part of the drywell shell in the sand bed region should be thinner than 0.49 inches. Thus, the monitoring regime must ensure that a thinning of around 0.1 inches would be detected to ensure that the corrosion could not threaten the structural integrity of the shell. Monitoring once every ten years is inadequate for this purpose because corrosion rates of more than 0.03 inches per year have been observed under corrosive conditions. a) The Acceptance Criteria Are Inadequate To first establish the thickness acceptance criteria, AmerGen used modeling of a 36 degree slice of the drywell shell (called a bay) that assumed the sand bed region had uniform thickness. Ex. NC 1 at 7-8. That model showed that if the shell at the sand bed had a uniform thickness of 0.736 inches, it would be able to support itself. Id. at 8. In addition, further modeling showed that one contiguous area of one square foot in each bay could be thinner than 0.736 inches, provided it was thicker than 0.536 inches. Ex. NC 3 at 9. Furthermore, analysis showed that areas 2.5 inches in diameter could be as thin as 0.49 inches. Ex. NC 1 at 9. To analyze the UT results, AmerGen initially analyzes whether the average wall thickness in each 6 inch by 6 inch monitored area is below 0.736 inches and whether each measurement is greater than 0.49 inches. Ex. NC 2 at 5. To evaluate areas where localized thickness is less than 0.736 inches AmerGen uses additional local wall acceptance criteria. Ex. NC 1 at 8. For small areas of less than 1 square foot, the mean thickness must be greater than 0.536 inches. Id. In addition, contiguous areas below 0.736 inches in average thickness should not exceed one square foot. Id. at 10.' The latter acceptance criterion did not fully reflect the limitations in the modeling that was used to derive the results. For instance, the modeling assumed only one area thinner than 0.736 The wording of AmerGen's response is slightly ambiguous in this regard. However, reference to the original calculation C-1302-187-5320-024, attached as Citizens' Exhibit NC 3, at Sheet 9 confirms that the modeling on which this criterion was based showed adequate strength if a 1 foot by 1 foot square area in each bay was 0.536 inches thick, and the rest of the bay was uniformly 0.736 inches in thick. 7

inches in each bay, but in bay 13 alone there are a total of at least nine areas that are below 0.736 inches. Ex. NC 3 at 26. In fact, AmerGen has recently reported that over 20 areas in total are now thinner than 0.736 inches and these areas have an average thickness of 0.703 inches. Ex. NC 2 at

13. AmerGen has also recently recognized that the minimum required linear distances between thin areas has not been calculated, but it has asserted that safety will be maintained if the total area under 0.736 inches in the sand bed region is less than one square foot. Id. at 11. Applying this criterion, AmerGen recently estimated that 0.68 square feet of the sand bed area are thinner than 0.736 inches.

Id. at 13. However, it is unclear how AmerGen derived this estimate and it is notable that no estimate of uncertainty was given. As discussed below, the area thinner than 0.736 inches is very sensitive to reductions in the thickness of the shell. Thus, it the uncertainty of this estimate must be high. Even this revised one square foot acceptance criterion is a misinterpretation of the modeling results. The model did not look at whether other geometries, such as a long thin gash, would lead to failure even if the thin area is less than one square foot. It also did not look at a situation which approximates the real condition, where the exterior of the drywell is more like a golf-ball with alternating thinner and thicker regions. Thus, AmerGen should either use the model to find the smallest area that could allow buckling to occur and compare that to the worst case total thin areas, or it should input comprehensive measurements into the model to show that the worst case that could occur before the next scheduled measurements could not allow buckling. In both cases, AmerGen should take full account of the uncertainty in the current thin area and the potential for future corrosion to rapidly expand that area. b) Monitoring Frequency Is Too Long And Monitoring Periods Must Adapt To Safety Margins AmerGen has stated that it derived the proposed one in ten year testing frequency from the standard in-service interval. Ex. NC 4 at 63. This is totally inadequate. As discussed in the 8

memorandum of Dr. Hausler, dated June 23, 2006 and attached to this Petition, the proposed visual inspections of the coating cannot substitute for UT testing, because they are too infrequent and corrosion could occur behind the coating without being noted visualy. Memorandum of Dr. R. Hausler, dated June 23, 2006 at 6. Furthermore, the current safety margins are, at best, razor thin. For instance, the thickness of small areas are now within around 0.083 inches of the safety margin, based on a measured thinnest point of 0.603 inches, a 0.03 inch allowance for uncertainty, and the acceptance criterion for such points of 0.49 inches. Id. The means of the six inch by six areas that are proposed to be measured again were within 0.07 inches of the safety margin in September 1994. Ex. NC 8 at 56. In addition, the acceptance criterion requiring the area per bay that is less than 0.736 inches thick to be less than one square foot in area would be violated if less than around 0.026 inches of corrosion occurs. Memorandum of Dr. Hausler, dated June 23, 2006 at 7. A reasonable estimate of the worst case potential corrosion rate that may occur could be obtained by analyzing the pre-1992 data. Id. at 6, 13. Observed corrosion rates to 1990 ranged up to 0.035 inches per year and were very uncertain. Ex. NC 9 at 7. As an illustration, even if the worst case corrosion rate were 0.02 inches per year and no corrosion has occurred since 1992, the drywell shell could exceed AmerGen's acceptance criterion for area below 0.736 inches in about one year. Memorandum of Dr. Hausler, dated June 23, 2006 at 7. Other criteria could be exceeded in around 4 years. The uncertainty in the worst case corrosion rate means that the measurements must be made at considerably shorter intervals than those calculated here to ensure that a measurement is taken before any of the acceptance criteria are violated. Thus, if a corrosive environment is present on the outside of the shell, UT measurements must be taken at least once every year, based on the current acceptance criteria. Id. Finally, the frequency of the measurements must be related to the time in which the shell could corrode beyond 9

the safety margin. Thus, if the next round of measurements shows any deterioration, the monitoring frequency would have to be increased. Id. c) The Proposed Scope Of The Monitoring Is Too Narrow The spatial scope of the monitoring must be sufficient to allow meaningful comparison with the acceptance criteria that are to be applied to the results. In addition, the monitoring must look for all anticipated aging effects. Looking first at the spatial scope of the monitoring, at present the proposed monitoring only covers twelve 6 inch by 6 inch areas and seven 6 inch by I inch areas. Ex. NC 2 at 5. Thus, of the around 300 square feet in the sand bed region, 3 square feet, or around 1% is proposed to be monitored. Memorandum of Dr. Hausler, dated June 23, 2006 at 15. Furthermore, because the monitoring points were initially selected by measuring from the inside and around two thirds of the sand bed region is not accessible from the inside, the proposed monitoring regime misses out known areas of the shell that are below 0.736 inches in thickness. Id. at8. In addition, because there was no attempt to expand the spatial scope of the measurements when points below 0.736 inches were observed at the edge of the grids, the monitoring protocol only incompletely tracks the thin areas that it does monitor. Id. at 9. The proposed monitoring regime makes also fails to systematically survey the shell for new thin areas. Id. at 8-9. Because the area of each bay below 0.736 inches is an important acceptance criterion and is particularly sensitive to corrosion, it is critical that the monitoring regime systematically identify and track the thickness of all areas that are below 0.736 inches. Id. It is likely that this will require monitoring from the outside of the drywell. Id. at 9. In addition to expanding the area of monitoring, another type of UT testing must also be added, because the shell is vulnerable to fatigue cracking in pitted areas. Id. at 5. This could go undetected under the currently proposed testing regime. 10

d) The Quality Assurance For The Measurements Is Inadequate Recently, the NRC concluded that the 1996 UT testing results are anomalous because they show that the drywell shell got dramatically thicker between 1994 and 1996. Transcript of Meeting on June 1, 2006, attached as Citizens' Exhibit NC 4 at 28, 31. Despite this, AmerGen has continued to use these data to predict the thickness of the drywell shell during any license renewal period. See p& Citizens' Exhibit NC 1 at 19-30. This is wholly unjustifiable. To eliminate this possibility in the future, AmerGen must revise its quality assurance plans to identify flawed data soon after it is taken and must undertake to carry out replacement measurements if it finds that the original measurements are questionable. Memorandum of Dr. Hausler, dated June 23, 2006 at 9-10. e) Statistical Analysis Of Results Is Flawed As the NRC has recognized, uncertainty is the key issue when analyzing the UT results. Ex. NC 4 at 63-64. In fact, there are a number of uncertainties, all of which need to be taken into account in the design of the monitoring regime. The first is that the UT results themselves are subject to uncertainty. This uncertainty means that the thickness at the time the measurement is taken is uncertain and it also means that the rate of corrosion is uncertain. Adding to the uncertainty in the corrosion rate is that conditions may change over time. For example, coatings may deteriorate, or the volume and composition of the water reaching the corroded area may change. As Dr. Hausler discusses in detail in his memorandum, the current statistical techniques employed are inadequate to find either the worst case baseline from which corrosion could occur, or the worst case corrosion rate. Memorandum of Dr. Hausler, dated June 23, 2006 at 10-15. The key flaws identified by Dr. Hausler are: i) AmerGen has failed to use extreme value statistics to estimate the minimum current thickness of the drywell shell, id. at 5, 11, 14-15; ii) corrosion is assumed to be linear, whereas in reality the corrosion rate can increase rapidly in a non-linear fashion, id. at 3, 12; 11

iii) analyzing corrosion rates using the average of the individual measurements taken in each grid is an invalid approach that leads to artificially low estimates of uncertainty, id. at 12; iv) the thinnest points measured in the grids have sometimes been omitted from the means, leading to artificially high estimates of the current mean thickness, id. at 12-13; v) an estimate of corrosion rate to 95% confidence is not sufficiently conservative for safety-critical issues, because one in twenty times the corrosion would be worse than the estimated rate, id.; and vi) AmerGen has ignored previous analysis showing that at least four valid measurements are required to make a valid estimate of the corrosion rate and the confidence limits. Id. Thus, AmerGen must make comprehensive measurements of the current wall thickness as soon as possible. Id. at 15. It must also revise its statistical techniques to calculate worst case estimates for all the parameters that are to be compared to the acceptance criteria, and must also calculate a worst case corrosion rate, which can be used to determine the appropriate time before the next monitoring. C. The Scope of License Renewal Includes Corrosion Of The Drywell Liner Petitioners are required to demonstrate that the issues raised in their contentions are within the scope of the proceeding, 10 C.F.R. § 2.309(f)(1)(iii). After extensive briefing of this issue, the ASLB concluded that corrosion of the drywell shell is within the scope of license renewal proceedings. In the Matter of AmerGen Energy Company (License Renewal for Oyster Creek Nuclear Generating Station) LBP-06-07 (slip op. at 39-40) (February, 26, 2006). That finding directly applies to the current contention, because it concerns the very same issue. Thus, the issue of scope is currently res judicata in this proceeding and is not subject to further dispute. However, the decision to admit the initial contention is currently on appeal to the Commission. Therefore, should the Commission amend the ASLB's finding regarding scope in its review of the AmerGen's 12

appeal, Citizens request an opportunity to file a supplemental briefing addressing the Commission's findings. D. Showing of Materiality The regulations require Petitioners to "[d]emonstrate that the issue raised in the contention is material to the findings the N.R.C. must make to support the action that is involved in the proceeding." 10 C.F.R. § 2.309(f)(1)(iv). A showing of materiality is not an onerous requirement, because all that is needed is a "minimal showing that material facts are in dispute, indicating that a further inquiry is appropriate." Georgia Institute of Technology, CLI-95-12, 42 N.R.C. 111, 118 (1995); Final Rule, Rules of Practice for Domestic Licensing Proceedings - Procedural Changes in the Hearing Process, 54 Fed. Reg. 33,171 (Aug. 11, 1989). Similarly, in Gulf States Utilities Co. (River Bend Station, Unit 1), CLI-94-10, 40 NRC 43 (1994), the Commission stated that, at the contention filing stage, "the factual support necessary to show that a genuine dispute exists need not be in formal evidentiary form, nor be as strong as that necessary to withstand a summary disposition motion." 40 NRC at 51. Rather, the petitioner need simply make "a minimal showing that the material facts are in dispute, thereby demonstrating that an inquiry in depth is appropriate." Id. (internal quotation marks omitted). In admitting the initial Petition, the ASLB found that a genuine and material dispute existed about whether the then proposed aging management program, which did not include periodic UT measurements, would enable AmerGen to maintain safety margins during the term of any extended license. LBP-06-07 at 38-39. This new contention concerning AmerGen's April 4, 2006 commitment continues this material dispute, taking AmerGen's additional commitment into account. Furthermore, in this Petition, Citizens have shown by reference to exhibits and expert opinion that the proposed monitoring by AmerGen is too limited in scope and too infrequent to 13

allow the current razor-thin safety margins to be maintained. In addition, Citzens have demonstrated that AmerGen has proposed to use flawed acceptance criteria and statistical methods to determine whether the results are significant and to project how quickly corrosion to safety critical levels could occur in the future. Thus, Citizens contend that the proposed program would fail to ensure that safety margins would continue to be met during any license renewal period. In contrast, AmerGen has stated that the committed monitoring regime will ensure that it can maintain safety margins throughout any extended license term. AmerGen Motion to Dismiss The Admitted Contention at 8. It has also stated that it made the additional commitments to "provide assurance that the drywell shell will remain capable of performing its design functions throughout the license renewal period." Letter from Michael P. Gallagher, AmerGen, to NRC (Apr. 4, 2006). Thus, at a high level the dispute is about the adequacy of the commitments to ensure that safety margins are maintained. At the more detailed level, Citizens have identified a myriad of flaws in AmerGen's approach, such as the failure to consider deterioration of the epoxy coating, the assumption that corrosion will be linear, and the failure to measure all the identified degraded areas. Thus, many more detailed material issues are also in dispute. Finally, because safety of the reactor hinges on the outcome of this dispute, it must be resolved before the NRC can issue any extended license. E. This Request Is Timely Petitioners may add new contentions after filing their initial petition, so long as they act in accordance with 10 C.F.R. § 2.309(f)(2). Entergy Nuclear Vermont Yankee, L.L.C. (Vermont Yankee Nuclear Power Station), LBP-05-32, 62 NRC 813 (2005). The Commission's regulations allow for a "new contention" to be filed upon a showing that: (i) The information upon which the amended or new contention is based was not previously available; 14

(ii) The information upon which the amended or new contention is based is materially different than information previously available; and (iii) The amended or new contention has been submitted in a timely fashion based on the availability of the subsequent information. 10 C.F.R. § 2.309(f)(2)(i)-(iii). In Vermont Yankee, the Board first admitted a contention of omission challenging an applicant's failure to perform structural and seismic analyses. The applicant subsequently performed structural and seismic analyses, after which it filed a motion to dismiss the contention as moot, which the Board granted. See Vermont Yankee, LBP-05-32, 62 NRC 813, 820. The Board gave the petitioner 20 days to file a new contention. Id. In response, the petitioner filed a contention challenging the sufficiency of the structural and seismic analyses. Id. In admitting the new contention, the Board held that the analyses were clearly information that was "not previously available" because it filled a prior omission, and that they were "materially different than information previously available" because something is obviously different than nothing. Vermont Yankee LBP-05-32, 62 NRC 813, 820; 10 C.F.R. § 2.309(f)(2)(i)-(ii). The facts of the present case directly parallel the facts of Vermont Yankee. First, the Board admitted a contention challenging AmerGen's failure to provide a plan for periodic UT testing in the sand bed region of the drywell. AmerGen subsequently docketed a commitment to adopt aging management procedures that included performing visual and UT testing every 10 years over the 20-year relicensing period, after which it filed a motion to dismiss the contention as moot. Just like Vermont Yankee, the Board granted the mootness motion, but also invited Citizens to file a new contention concerning the adequacy of the new commitment within 20 days. Licensing Board Memorandum and Order (Contention of Omission is Moot, and Motions Concerning Mandatory Disclosure are Moot), LBP-06-16 at 2 (Jun. 6, 2006) (unpublished). In accordance with the Board's 15

Order, Citizens are now seeking to add this contention challenging the sufficiency of the proposed monitoring regime. Thus, like Vermont Yankee, the ASLB should now find that the new contention is based upon information that was "not previously available," and that is "materially different than information previously available." 10 C.F.R. § 2.309(f)(2)(i)-(ii). Further supporting the conclusion that the April 4, 2006 commitment is materially different information is that the Board decided that it made Citizens' previously admitted contention moot. Thus, it made a material difference to this litigation. Such a conclusion is further reinforced by noting that "something" (a UT testing plan) cannot be materially the same as "nothing" (no UT testing plan at all), meaning that the newly announced UT plan is "information ... materially different than information previously available." 10 C.F.R. § 2.309(f)(2)(ii). See Vermont Yankee. LBP-05-32, 62 NRC 813, 820. In addition, at the time the initial Petition was submitted, Citizens had limited information about the drywell corrosion issue. For example, Citizens did not know what the 1996 measurements showed because, despite diligent efforts, Citizens had been unable to obtain those measurements. It was also unclear how AmerGen had changed the acceptance criteria for measurements that showed that the steel shell was already thinner than the initial 0.736 inch criterion. The Exhibits attached to this contention and upon which Dr. Hausler has based his June 23, 2006 memorandum have now clarified these issues, but they were not available to Citizens at the time the initial Petition was submitted. More specifically, Exhibits NC 1 and NC 2 were created in April 2006, Exhibit NC 4 was created in June 2006, and Citizens obtained the rest of the Exhibits from AmerGen through the document disclosure process. Thus, material new information has allowed Citizens to now submit a much more specific new contention, which therefore satisfies 10 C.F.R. § 2.309(f)(2)(i)-(ii). 16

Finally, because this contention is being filed within the timeframe specified by the Board's Order of June 2, 2006, it satisfies 10 C.F.R. § 2.309(f)(2)(iii). Furthermore, the Order also makes clear that "if NIRS satisfies the remaining factors in section 2.309(0(2) - the parties need not address the requirements under 10 C.F.R. § 2.309(c), which apply to 'nontimely filings.P" Licensing Board Memorandum and Order (Contention of Omission is Moot, and Motions Concerning Mandatory Disclosure are Moot), LBP-06-16 at n. 12 (Jun. 6, 2006) (unpublished). CONCLUSION For the foregoing reasons, the ASLB should grant this Petition to add the proposed new contention. Respectfully submitted Richard Webster, Esq RUTGERS ENVIRONMENTAL LAW CLINIC Attorneys for Petitioners Dated: June 23, 2006 17

UNITED STATES OF AMERICA BEFORE THE NUCLEAR REGULATORY COMMISSION OFFICE OF THE SECRETARY In the Matter of )

                                                             )    Docket No. 50-0219-LR AMERGEN ENERGY COMPANY, LLC                                   )
                                                             )    ASLB No. 06-844-01-LR (License Renewal for the Oyster Creek                         )

Nuclear Generating Station) ) June 23, 2006

                                                             )

CERTIFICATE OF SERVICE I hereby certify that the foregoing Petition with attachments and motion was sent this 23rd day of June, 2006 via email and U.S. Postal Service, as designated below, to each of the following: Secretary of the Commission (Email and original and 2 copies via U.S Postal Service) United States Nuclear Regulatory Commission Washington, DC 20555-0001 Attention: Rulemaking and Adjudications Staff Email: HEARINGDOCKET@-N.R.C..GOV Administrative Judge E. Roy Hawkens, Chair (Email and U.S. Postal Service) Atomic Safety and Licensing Board Panel Mail Stop - T-3 F23 United States Nuclear Regulatory Commission Washington, DC 20555-0001 Email: erh@nrc.gov Administrative Judge Dr. Paul B. Abramson (Email and U.S. Postal Service) Atomic Safety and Licensing Board Panel Mail Stop - T-3 F23 United States Nuclear Regulatory Commission Washington, DC 20555-0001 Email: pbaa~nrc.gov Administrative Judge Dr. Anthony J. Baratta (Email and U.S. Postal Service) Atomic Safety and Licensing Board Panel Mail Stop - T-3 F23 United States Nuclear Regulatory Commission Washington, DC 20555-0001 18

Email: ajb5anrc.gov Law Clerk Debra Wolf (Email and U.S. Postal Service) Atomic Safety & Licensing Board Panel Mail Stop - T-3 F23 U.S. Nuclear Regulatory Commission Washington, DC 20555-0001 DAW1 @nrc.gov Office of General Counsel (Email and U.S. Postal Service) United States Nuclear Regulatory Commission Washington, DC 20555-0001 Email : OGCMAILCENTER@N.R.C..GOV Mitzi Young (Email and U.S. Postal Service) U.S. Nuclear Regulatory Commission Office of the General Counsel Mail Stop: 0-15 D21 Washington, DC 20555-0001 E-mail: may@nrc.gov Alex S. Polonsky, Esq. (Email and U.S. Postal Service) Morgan, Lewis, &Bockius LLP 1111 Pennsylvania Avenue, NW Washington, DC 20004 Email: apolonsky@morganlewis.com Kathryn M. Sutton, Esq. (Email and U.S. Postal Service) Morgan, Lewis, &Bockius LLP 1111 Pennsylvania Avenue, NW Washington, DC 20004 Email: ksuttonamorganlewis.com Donald Silverman, Esq. (Email and U.S. Postal Service) Morgan, Lewis, & Bockius LLP 1111 Pennsylvania Avenue, NW Washington, DC 20004 Email: dsilvennanamorganlewis.com J. Bradley Fewell (Email and U.S. Postal Service) Exelon Corporation 200 Exelon Way, Suite 200 Kennett Square, PA 19348 bradley.fewell@exceloncorp.com John Covino, DAG (Email and U.S. Postal Service) State of New Jersey 19

Department of Law and Public Safety Office of the Attorney General Hughes Justice Complex 25 West Market Street P.O. Box 093 Trenton, NJ 08625 E-mail: john.corvino@dol.lps.state.nj.us Paul Gunter (Email and U.S. Postal Service) Nuclear Information and Resource Service 1424 16th St. NW Suite 404 Washington, DC 20036 Email: pgunteranirs.org Edith Gbur (Email) Jersey Shore Nuclear Watch, Inc. 364 Costa Mesa Drive. Toms River, New Jersey 08757 Email: gburl @comcast.net Paula Gotsch (Email) GRAMMIES 205 6th Avenue Normandy Beach, New Jersey 08723 paulagotsch@verizon.net Kelly McNicholas (Email) New Jersey Sierra Club 139 West Hanover Street Trenton New Jersey 08618 Email: Kelly.McNicholasesierraclub.org Suzanne Leta (Email) New Jersey Public Interest Research Group 11 N. Willow St, Trenton, NJ 08608. Email: sleta@,niyirg.org Peggy Sturmfels (Email) New Jersey Environmental Federation 1002 Ocean Avenue Belmar, New Jersey 073 19 Email: psturmfelsecleanwater.org Michele Donato, Esq. (Email) PO Box 145 Lavalette, NJ 08735 Email: mdonato@micheledonatoesq.com 20

Signed: ______________ Richard Webster Dated: June 23, 2006 21

CORRO-CONSULTA 8081 Diane Drive Rudolf H. Hausler Kaufinan, TX 75142 Telk 972 962 8287 (office) rudyhau@msn.com Fax: 972 932 3947 Tel. 972 824 5871 (mobile) Memorandum To: Richard Webster, Esq. June 23, 2006 From: Rudy Hausler

Subject:

Discussion of Corrosion Monitoring Methodologies At Oyster Creek Nuclear Plant Dry Well

SUMMARY

The corrosion on the outside of the Oyster Creek drywell steel liner, particularly in the former sandbed region, is of great concern in regards to the structural integrity of the liner. Various structural integrity calculations had been performed by Amergen/Exelon in the past to arrive at various wall thickness criteria. Subsequently these criteria were compared to actual measurements of remaining wall thicknesses. Going forward, continuing corrosion rates have been discussed, and times at which possible minimum wall thickness, as defined by the criteria, have been derived by the operator. This study critically reviews what is known about the corrosion in the sandbed area, the way the corrosion measurements had been evaluated, and the conclusions that had been drawn. As it turns out, only a very small fraction of the total sandbed area had been examined, which poses the problem as to whether in fact the most severely corroded areas had been observed, and whether, extrapolation of these observation to the entire surface are justified. The measurements were performed with a 6inch by 6inch template and consisted of point measurements at one-inch spacings. As a consequence assessments of corrosion flaws could only be made in the z-direction (depth) while the x/y dimensions of the flaws remained unexplored. However, acceptance criteria are based on spatial dimensions, which consequently had to be guessed at. It had been assumed that pitting corrosion rates in the sand bed area would be constant in time. This assumption is not justified based on an analysis of the corrosion mechanism. It had also been assumed that the pit distribution would be Gaussian, and that therefore the deepest measured pits which were beyond 1

the 2s limit could be dropped from consideration. This is considered an unprofessional approach for two reasons. First: no measurements should ever be excluded from consideration (on statistical considerations only) unless it can be demonstrated that such measurements are flawed technically. Second: Pit distributions are not Gausian, but exponential, hence the deepest pits are of vital importance. Amergen/Exelon evaluated corrosion rates based on average remaining wall thickensses. However it is well known that structures do not fail by averages but rather by extremes, namely where due to corrosion the wall thickness had become thinnest. Consequently, evaluation of the available data by extreme value statistics demonstrated that the most probable deepest pits (corrosion anomalies) were deeper than those assessed by the operator or Oyster Creek. At this point in time, there are no valid assessments of possible corrosion (pitting) rates. The operator assumed that conditions might have been constant over time and would remain constant in the future. However, this assumption cannot be justified under any condition. It is, therefore, considered of primary importance that a) the entire drywell surface be examined with UT technology capable of assessing corrosion anomalies spatially. It is furthermore essential that the coating, which is well past its useful lifetime be examined with methodology other than just visual, in order to completely assess whether it is still protective. Programmatic aging surveillance must include such measurements much more frequently than every 10 years, because deterioration of the coating is not linear in time either. I. Background It is well established that serious corrosion occurred over the years on the outside of the drywell containment of the nuclear reactor at Oyster Creek 1). While corrosion occurred in all areas on the outside of the drywell, which experienced temporary or permanent wetting due to water leaks, the most severe damage was observed in the sandbed region 2). In 1992 the sandbed was removed and the corroded areas were coated with an epoxy coating. The coating was specified to have "an estimated life of 8 to 10 years". Subsequently three UT inspections were performed in 1992, 1994 and 1996. Based these inspection results, projections were made to the effect that no corrosion would occur over the next 10 years. There are many concerns within this paradigm, which need further examination and discussion. The most striking are:

) see for instance e-mail correspondence from George Beck (Exelon Corp.) to Donnie Ashley (dial Onrc.gov) 4/5/06) 2)see for instance GPU Nuclear Corporation letter to US NRC September 15, 1995 2

        "   The assertion of no further corrosion based on the '92, '94 and '96 UT measurements was erroneously based on the assumption that conditions would remain constant, i.e. the epoxy coating of the dry well liner and concrete floor, as well as the elastomer used to seal the crevice between the floor and the drywell liner, all part of the sandbed region, would not deteriorate with time.
        " The analysis of the results erroneously assumed that averaging 49 individual UT measurements, which were conducted over a 6x6 inch grid at 1 inch spacings would adequately represent the corrosion damage occurring in each bay, and hence these averages could therefore be used to conduct the necessary structural integrity calculations.

Embedded in these major concerns are a number of issues dealing with basic assumptions made in the evaluation of the corrosion measurements. These are listed below and will require some discussion:

         " Amergen/Exelon have assumed that the growth of the observed localized attach (pitting) would be linear with time, hence the corrosion rate) pit penetration rate) would be constant with time. The known pitting mechanisms will not support this assumption
         " It has been furthermore assumed and so stated in many supporting documents that the pit size distribution would be normal (Gaussian). This assumption has then led to a number of conclusions and actions, which must be revised.
         " It has also been assumed that averages of observed pit sizes would be representative of the corrosion processes, and that such averages from observations over time could be used to extract the "corrosion rate" (or more precisely, the pit penetration rate).

II. Some Comments Regarding the Corrosion Mechanism A simple model as follows is being considered in order to delineate the major processes and parameters, which control them: 3

The fact is that the sandbed was essentially soaked with water, either periodically or permanently. This water was initially aerated which caused corrosion, even if the pH is above 7. As corrosion in the wet sandbed continues, the wet environment in the sand becomes depleted of oxygen. However, there is an almost inexhaustible reservoir of oxygen just above the sandbed - the air space. As a consequence, the steel surfaces embedded in the sand become anodic, while the cathodic reaction takes place on the areas which are richer in oxygen - the typical situation for crevice corrosion. The anodic reaction is not uniform, but pits will be forming. Initially, there will be a plethora of small shallow pits. Eventually some grow deeper than others, in fact at the cost of others. The frequency distribution of the pits is not normal, rather one can observe an exponential distribution - the frequency of pit depth decreases exponentially with pit depth. The fact that often a normal distribution is observed is an artifact, simply because the smaller pits are not normally measured, but are attributed to surface roughness and hence not included in the histogram. For this reason it is not proper to evaluate pit depth distribution on the basis of Gaussian statistics, and it is even less proper to discard deep pits outside the 95% confidence limits as atypical. Rather, deep pits, which have been measured, are a fact of life and must be included in any statistical evaluation, unless the measurement can be shown to be faulty for technical (not statistical) reasons. We will therefore show below the application of extreme value statistics to some data obtained from Oyster Creek. Since pits are anodic areas where iron ions are being generated it stands to reason that anions must migrate into the pit, generally through a corrosion product layer, which fills the pit, such as iron hydroxide (two valent), or iron oxyhydroxide (three valent). The anions, which are present in the water at the highest concentration, are most likely to accumulate at the bottom of the pit where iron ions are being generated. The water in the sand bed is said to have contained as much as 500 ppm of chloride ions. This is more than the concentration the hydroxide ion at a pH of 7 or 1.sx10" moli/ CI vs. 10-7 mol/L for OH-. Chloride therefore will accumulate at the bottom of the pit. This will cause the pH in the pit to decrease to perhaps as little as 1 or 2. (This chemistry is well known and has been described in the literature many times). Lowering of the pH in the pit will accelerate pit growth, provided that the mass transport of water into the pit can sustain a higher corrosion rate at the bottom of the pit.

      " It is therefore no forgone conclusion that the pit growth rate is constant with time. In fact, depending on the nature of the corrosion product in the pit, the mass transport into the pit can either be shut down, or sustain an accelerated corrosion rate due to the lower pH.
      " Organic coatings will greatly reduce the transfer of both water and oxygen to the pit area. However, as the coating ages, such mass transfer is again accelerated. The unverified assumption that the coating will shut down pit growth for all eternity is totally unjustified. Furthermore, the unverified 4

assumption that visual observation of the coated areas is sufficient to assert that no corrosion occurs is also unjustified. The assumption that if the coating held for 10 or 15 year it will hold for another 20 years is also unjustified and contradictory to general observations. (Coating life has been specified for 8 to 10 years).

• More disturbing, however, the fact that pit depth of 600 mils can easily be demonstrated statistically. This corresponds locally to a remaining wall thickness of about 550 mils or close to the 490 mil criterion for small areas. (This criterion, as we understand it is based not on buckling considerations, but on pressure calculations.) If isolated pits of that size exist, and extreme value statistics predict such pits with a high probability (see below), then the specter of chloride induced fatigue cracking is raised.

Again, the danger is based on the fact that chloride is present in the base of the pit (and has actually been found there), that the pH in the pit is low, and that the stress at the pit tip is approaching a limiting value. All this contributes to stress corrosion and/or fatigue cracking. It will therefore be necessary to examine the corroded areas, and in fact all areas susceptible to corrosion, for the possible existence of cracks in the dry well liner wall. III. Monitoring Frequency Is Too Long And Monitoring Periods Must Adapt To Safety Margins Because the sand bed region is severely corroded, margins of safety are now much thinner than when the plant was first built. For example, in parts, over 0.5 inches of metal have corroded away from the steel drywell shell over areas larger than just single pits, leaving a metal thickness ofjust over 0.6 inches. According to AmerGen, no part of the drywell shell in the sand bed region should be thinner than 0.49 inches. Thus, to maintain safety, the monitoring regime must be able to predict how fast the metal could corrode to safety-critical levels, and must ensure that testing of areas that are closest to the margins occurs before there is any possibility that the metal has corroded too much. The monitoring regime proposed does not achieve this goal because the monitoring frequency is too low and is not adaptive to how close the shell thickness is to the acceptance criteria, degraded areas of the shell would not be systematically identified and tested, the quality assurance for the measurements is inadequate, and the statistical techniques used in data analysis are flawed. This Memorandum discusses these issues in detail.

1. Overview AmerGen has stated that it derived the proposed one in ten year testing frequency from the standard in service interval. Ex. NC 4 at 63. This is totally inadequate.

5

To insure that margins of safety are maintained, AmerGen must predict the worst case corrosion rate that could occur before the next scheduled round of monitoring. The monitoring regime should show that in the worst case the acceptance criteria will continue to be met. Interestingly, in the past the reactor operator has recognized this need to some extent. For example, in 1992 a calculation estimated that with 95% confidence, the mean thickness of area 13A would not go below 0.736 inches before June 1995. Ex. NC 7 at 9. The operator also predicted the minimum mean thickness at the 95% confidence level at the date of the next scheduled monitoring to verify that it was less than the acceptance criterion. Id. at 10. However, more recently AmerGen has not estimated the corrosion rate at the sand bed because it has assumed that it is zero, which, far from being the worst case, is actually the best possible case. See NC 1 at 19 to 30. Furthermore, although the reactor operator used to provide 95%ile confidence limits for its predictions, AmerGen has ceased to do this for the sand bed region, id., while continuing to do this for the upper drywell. Ex. NC 6 at 8. AmerGen attempts to justify this on the basis that visual inspection of the sand bed is sufficient. Ex. NC 1 at 32. However, the coating could deteriorate between inspections, because it is already well past its 8 to 10 year expected life. Ex. NC 8 at 56. In addition, corrosion behind the coating could occur and not be noted visually. Furthermore the committed visual inspection period is once every ten years, the same as the UT testing period. Therefore, visual inspections will not provide any information on changes in conditions between UT tests. In addition, because past analyses relied on prediction of the mean thickness, they failed to apply a corrosion rate to the measurements at individual points to ensure that even in the worst case they will remain thicker than the 0.490" acceptance criterion before the next scheduled monitoring. Furthermore, they failed to predict the rate of growth of the areas below 0.736 inches in each bay to ensure that they will also remain less than one square foot before the next scheduled monitoring. At present, AmerGen has insufficient data to predict the worst case corrosion rate without sand. As discussed in more detail below, one reasonable approach to resolve this problem would be to use results taken before the sand was removed, derive a statistically valid worst case corrosion rate, and see how soon acceptance criteria could be violated using that rate. For example, AmerGen has stated that the thinnest individual result that has been measured is 0.603 inches. Ex. NC 1 at 7. The acceptance criterion for individual points is 0.490 inches. The uncertainty in each measurement is around 5% or 0.03 inches meaning that the thinnest real condition consistent with the measurement is around 0.573 inches.3 This yields a current margin of safety of approximately 0.083 inches. The second highest long term corrosion rate estimated was 0.017 inches per year. Ex. NC 1 at 20. Thus, assuming that the next round of monitoring shows no further deterioration, and that 3 As discussed below, AmerGen should make a more rigorous estimate of this parameter using appropriate statistical measures. 6

the worst case corrosion rate could be around 0.020 inches per year, further testing would be needed in approximately four years. Turning to the area below 0.736 inches, bay 13 was closest to the safety margin when measurements were taken from the outside in 1992. The results showed that nine areas below 0.736" were widely scattered over a large area in this bay. Ex. NC 3 at Sheet 26-29. The outside of the shell was found have indentations from a thickness of around 0.800 inches that were "about 12 to 18" in diameter... at about 12 inches apart." Id. at Sheet 24. Measurements of nine one to two inch diameter areas at the thinnest parts of these indentations showed thicknesses ranging from 0.618 inches to 0.728 inches. Id. at Sheets 26, 28. The areas below 0.736 inches were "not more than 1 to 2 inches in diameter," except for location 7 which could have been 6 inches square with an average thickness of 0.677 inches. Id. at Sheet 26. Applying the one square foot below 0.736 inches acceptance criterion to these measurements, the total area measured below 1 square foot was around 0.3 square feet. However, this area is very sensitive to additional corrosion because in a length of around 5 inches, the thickness changed from around 0.736 inches to 0.800 inches. Assuming that the edge of the hole is a straight line, this means that a change of 0.064 inches in depth occurs over about 5 inches in length. Thus, for the radius of the thin area to change by two inches, the depth would have to change by only 0.026 inches. If this occurred the total area below 0.736 inches would be approximately 1.6 square feet, well beyond the current acceptance criterion. Assuming a worst case corrosion rate of 0.020 inches per year shows that the area acceptance criterion could be violated in around a year, even if the thin areas have not grown bigger since they were last measured in 1992. These results show that the currently proposed monitoring interval of ten years is far too long. If the worst case corrosion rate is around 0.020 inches, the total area under 0.736 inches could increase beyond the safety margin in about a year. Thus, monitoring would be needed at least once per year. Finally, if the next round of measurements shows that the margin of safety is less than it was in when the last valid round of testing occurred (in 1992 or 1994), the testing intervals must be increased accordingly.

2. Proposed Area To Be Measured Is Too Small Large variations in remaining wall thickness have been observed. Minimum wall thicknesses of as little as 0.603 inches have been reported within the 6x6 inch grids.

In addition, many other thin areas, with thickness measurements as low as 0.618, have been observed from the outside of the drywell. It is therefore entirely unreasonable to assume that the small 6x6 inch areas on top of the sandbed are representative of the over 3 foot thickness, Ex. NC 8 at 40, of the entire sandbed area, simply because around two thirds of the sand bed shell below the 6x6 inch grid was not accessible from the inside. See Ex. NC 10. 7

Furthermore, the spatial scope of the monitoring must be sufficient to allow meaningful comparison with the acceptance criteria that are to be applied to the results. In various submissions AmerGen has laid out how the monitoring was done in the sand bed region in 1992, 1994, and 1996. Initial investigations, carried out before the sand was removed, measured the thickness of the drywell shell in the sand bed region from the inside "at the lowest accessible locations." Ex. NC 5 at

11. However, because the interior concrete floor and curb is over two feet higher than the exterior floor this meant around two thirds of the sand bed area was not tested. To see if the corrosion extended to these areas the reactor operator dug a trench into the floor in bays 17 and 5 and found that the thinning below the floor level in bay 17 was similar to that observed above the floor, but eventually became less severe. Id. This confirmed that much of the area below the interior floor was corroding, showing that this area should not have been omitted from the monitoring regime.

In bays where initial investigations found significant wall thinning, 49 readings were taken within a 6 inch by 6 inch square centered at elevation 11'3". Ex. NC 2 at 5. In other bays, 7 readings were taken along a 7 inch horizontal line at the same elevation. Id. Thus, the initial selection of the points to be monitored periodically was fundamentally flawed because it omitted to establish monitoring of known thin areas below the interior floor level, and failed to even attempt to identify thin areas below the floor level in eight of the ten bays. Measurements conducted from the outside of the drywell shell in 1992 highlighted these deficiencies in the initial investigations. The 1992 measurements demonstrated that there are extensive areas in bays 1 and 13 that are not proposed to be tested, but are already well below 0.736 inches thick. Ex. NC 3. For example in bay 13, nine areas below 0.736" were widely scattered over a large area. Id. at Sheet 26-29. Measurements of nine one to two inch diameter areas showed thicknesses ranging from 0.618 inches to 0.728 inches. Id. at Sheets 26, 28. Figure 13 on Sheet 29, shows the locations. To give an idea of scale, the distance between locations 5 and 7 was "about 30 inches apart." Id. at Sheet 26. For point 7 alone, the area below 0.736 inches was conservatively estimated to be 6 by 6 inches with a thickness of 0.677 inches on average. Id. at Sheet 26. Similarly, the measurements in bay 1 showed eight areas below 0.736 inches, whose thickness ranged from 0.700 to 0.726 inches. Id. at Sheet 11. The thinnest area was at location 7, which was located well below the "bathtub ring" and so cannot be easily monitored from the inside of the drywell. Id. at Sheet 12. These results show that the spatial scope of the proposed monitoring is wholly inadequate to assess whether the drywell shell is meeting the acceptance criteria. Many areas that are thinner than 0.736 inches limit are not proposed to be monitored at all. Even those that have been monitored once are not fully characterized. To fully address all the areas that are below 0.736 inches, AmerGen must devise a systematic approach to identify and measure all such areas. Thereafter, each area must be measured and tracked to enable AmerGen to estimate 8

the worst case corrosion rate and the worst case rate at which the thin areas could expand. Because AmerGen is now proposing to measure at the same locations that it measured in 1992, 1994, and 1996, the scope of the monitoring will remain inadequate, even though the exterior of the sand bed is now accessible, so that the cause of the initial inadequacy no longer exists. Unless AmerGen can devise a way to monitor through the concrete in the interior of the drywell, it appears likely that future monitoring will need to be conducted from the outside of the shell. A second, less difficult problem is that the square grid pattern employed in the previous testing may miss extended areas of thinness that are not square. For instance, if a 5 inch by 30 inch horizontal trough were present in the shell and intersected the measured area, its area would only be estimated as 5 inches by 6 inches because of the area limitation of the measurements. Thus, its area would be estimated as 0.2 square feet, whereas the actual area would be over 1 square foot, in violation of an acceptance criterion. This means that if the testing finds points below 0.736 inches on the outside of the grid, it will underestimate the continuous area that is below 0.736 inches. To avoid this error AmerGen should expand the search area where or when readings at the edge of the grids show readings of less than 0.736 inches. It has failed to propose such a change.

3. The Quality Assurance For The Measurements Is Inadequate Recently, the NRC concluded that the quality of the calibration for the UT measurements taken after 1992 is in question. Transcript of Meeting on June 1, 2006, attached as Citizens' Exhibit NC 4, at 28. Further, NRC said that the 1996 results are anomalous because they show that the drywell shell got dramatically thicker between 1994 and 1996. Id. at 28, 31. AmerGen responded that they had spent a lot of time trying to find the source of the problem, but were unable to explain why the results were so high. Id. at 29. AmerGen also acknowledged that it could not explain the increase between 1994 and 1996, Id. at 31, but would do additional calibration to see if the coatings on the inside and outside of the drywell affected the results. Id. at 29.

The systematically higher wall thicknesses observed in 1996 cannot be explained purely by the presence of the epoxy coating, because the coating was present when the previous two measurements were taken from the inside in 1992 and 1994. One potential explanation for the anomalous 1996 results is the start of coating deterioration. It is known that certain poly-epoxides tend to swell in the presence of humidity and at elevated temperature. It is proposed, as a working hypothesis, that the higher measurements in 1996 may well be due to such swelling, which could not have been calibrated out of the measurements. As a consequence, the actual thickness of the drywell shell in 1996 might well have been lower than the 1994 measurements due to ongoing corrosion, albeit a slower pace than pre-1992. What is clear is that the 1996 results cannot be used to predict future corrosion rates, and that even in the 1992 and 1994 post-coating results are in question. 9

Had AmerGen had an effective quality assurance program in place when the results were taken in 1996, it would have identified any problems with the data close to the time that they were taken. As illustrated by my memo of May 3, 2006, the anomaly in the 1996 results was not difficult to find, provided systematic rather than random error was the focus. Thus, Amergen obviously did not have an adequate quality assurance program in place. AmerGen has recently stated that the same methodology will be used to analyze the 1992, 1994, and 1996, will be used for the new UT results. Ex. NC 2 at 2. This means that AmerGen will continue to fail to identify questionable data in a timely manner, unless it changes its approach to the identification of systematic error. Furthermore, although AmerGen realized at some point that there were questions about the reliability of the thickness data taken after 1992, especially the 1996 results, it has continued to use these data to predict the thickness of the drywell shell during any license renewal period. See M Citizens' Exhibit NC 1 at 19-30. This is wholly unjustifiable. Unless questions about calibration of the results taken after the coating can be answered, the post-coating thickness data provide little knowledge about the actual thickness of the drywell shell, let alone the corrosion rate.

4. Statistical Analysis Of Results Is Flawed

a. Background As the NRC has recognized, uncertainty is the key issue when analyzing the UT results.' Ex. NC 4 at 63-64. In fact, there are a number of uncertainties, all of which need to be taken into account in the design of the monitoring regime. The first is that the UT results themselves are subject to uncertainty. This uncertainty means that the thickness at the time the measurement is taken is uncertain and it also means that the rate of corrosion is uncertain. Adding to the uncertainty in the corrosion rate is that conditions may change over time. For example, coatings may deteriorate, or the volume and composition of the water reaching the corroded area may change.

Since the actual original UT measurements were not available for a detailed statistical analysis, a hypothetical 6x6 inch grid was constructed to illustrate a point to be made here. Figure 1 shows hypothetical UT measurements in a 6x6 grid with 1 inch spacings. The average wall thickness over the grid is 0.81 inches. However, as is often the case in real life, a corrosion trough is depicted parallel to the y-axis with an average depth of 0.68 inches and a maximum depth of 0.55 to 0.60 inches. While this example is not a real life observation, it nevertheless illustrates how averages can be misleading. In this particular case, the corrosion trough exceeds the grid, and one could not tell whether corrosion would become more severe or less severe beyond the boundaries of the grid. Similarly, when Amergen talks about "isolated minimum thickness measurements", one does not know where these were recorded and whether there were others, which exceeded the average wall loss, but 10

may have been above the quoted minimum wall thickness. When the same data shown in Figure 1 are plotted from a different perspective (Figure 2), conclusions may be different, but again it appears that there may be an extensive corrosion phenomenon on one side of the grid. In the treatment of the current thickness, AmerGen has set various acceptance criteria: one for small areas of around 2.5 inches in diameter (0.490 inches), one for areas of less than 12 by 12 inches (0.535 inches), one for the total area where the wall thickness is less than 0.735 inches (one square foot), and one for the mean thickness of the vessel (0.753 inches). In comparing the measured data to the acceptance criteria, the reactor operator actually evaluated the UT results by comparing the means of the 6 by 6 inch grids to 0.535 inches, and comparing each measurement to the small area criterion. Ex. NC 2 at 11.

b. Modeling It is generally accepted that failures do not occur as a result of average corrosion, but are generally occasioned by the weakest spot in the system. As a consequence, one cannot interpret the data by calculating averages and standard deviations.

Figure 3 for instance shows a histogram of the 49 hypothetical UT measurements. It can clearly be seen that in this example a bimodal distribution exists. The first mode, covering small pit depths, is perhaps Gaussian, as is often observed for pit initiation, because the smallest pits, too difficult to count, are rarely included in the analysis. The second mode is represented by a skewed distribution, perhaps a Weibull distribution with very high extreme values. Again this type of distribution is often observed after pitting has progressed for some time. It would clearly be irrational to try and present data of this kind by a Gaussian distribution and disregard the values that are outside "confidence limits". Rather, data of this kind should be analyzed by Extreme Value statistic. It turned out, as shown in Figure 4 that a reasonably straight line is obtained when the pit depths are plotted as a function of the "reduced variate". Only 49 points were available for the correlation. Extrapolation to the virtual 100Ih point results in a pit depth of about 0.77 inch, a remaining wall thickness of about 0.4 inches, or in this hypothetical case, a remaining wall thickness of less than minimum allowable. Because a worst case analysis is necessary for a safety-critical condition, the data must be analyzed by a methodology similar to the one demonstrated in the above procedure. Unfortunately, at present AmerGen appears to take no account of the chance that the true value of the remaining wall thickness at each point could actually be substantially less than indicated. See Ex. NC 3 at Sheet 6. Turning to the corrosion rate, AmerGen attempted to predict corrosion rates based on the '92, '94, and '96 UT measurements. They used the averages for each grid measured in each by over the time period indicated. (This procedure is based on the notion that all pits grow at the same rate, which is quite erroneous since the deepest pits usually grow faster than the smaller ones.) In most instances it turned out that the 92 averages were higher than the 94 averages, while the 96 averages were again 11

higher then the previous two. This is shown in Figure 5. However, a statistical Analysis of Variance (ANOVA), Figure 6, shows that there is no significance to these variations from date to date if the data are amalgamated. However, the differences from location to location are indeed very significant. On the basis of these data Amergen concluded that the corrosion was arrested following the application of the epoxy coating. It is probably correct that on average the corrosion was significantly slowed or even arrested during the four years covered by the measurements. Whether the extreme corrosion rates were also similarly affected remains an open question. Nevertheless, it would be logical to expect that corrosion slowed down following the application of the coating, at least for a period of time. It is, however, stretching credulity to assume that such protection would last in excess of the stated lifetime of the coating, which was specified as 8 to 10 years. Turning to the details of the analysis, the way in which AmerGen calculated corrosion rates was flawed in at least four ways. The calculation of estimated corrosion rates erroneously assumed that the rate would be constant over time, the means of the 49 point grid were used for curve fitting, the most extreme values were often omitted from the calculation of the means, and a ninety five percentile statistic is used as the appropriate level of uncertainty for future predictions. Taking each of these flaws in turn AmerGen first made the erroneous assumption that "if corrosion is continuing, the mean thickness will decrease linearly with time." Citizens' Exhibit NC2 at 6. In fact, if the coating starts to fail, the corrosion rate could increase rapidly in a non-linear fashion. The projected coating life is around eight to ten years, and that life has now been exceeded by around four to six years. Ex. NC 8 at 54. In addition, other conditions could change. Thus, the assumption that the corrosion rate will be constant with time is simply invalid. Second, using the means of the 49 points rather than the individual points to produce the curve fit that is used for future predictions only serves to mask the inherent uncertainty in the data, because the means are less variable than the individual points. See I Ex. NC 1 at 21. The fit statistics from the curve fit program therefore do not fully represent the uncertainty in the fit because the errors are artificially lowered by only feeding in the means, rather than individual measurements. A more appropriate procedure would be to plot all the individual measurements and then do a curve fit and find the predicted errors on the curve fit at an appropriate level of uncertainty. Third, AmerGen does not include the thinnest points in the means it reports, because it treats pits separately in the analysis when the data are not normally distributed. E& Ex. NC 5 at 25. A more recent analysis of upper region results by AmerGen best illustrates the problem. The analysis candidly states "points that were considered pits are.., excluded from the mean." Ex. NC 6 at 15. Such a procedure obviously leads to an underestimation of the mean value and the corrosion rate. Thus, in some cases the mean values that have been plotted and 12

fitted are actually thicker than the mean thicknesses of the areas that were measured. This is obviously a major problem with the analysis because one acceptance criterion is based on the mean values of the grids. Fourth, only the ninety five percentile of extreme values are used for the prediction of the corrosion rate. E Ex. NC 5 at 1. This means that even if the prediction is correct and the 95%ile confidence limit is taken as the worst case corrosion rate, there is a one in twenty chance that the actual corrosion rate will be higher than that calculated. For a safety-critical evaluation, this level of uncertainty is far too high. The statistical procedure must be redesigned to insure that safety margins are met with a substantial degree of certainty. Further flaws have crept into the analysis over time. In 1992 the reactor operator recognized that to estimate a 95%ile of the corrosion rate, at least four data sets are needed. Ex. NC 5 at 1. It further recognized that where only two points were available, the uncertainty in the individual points should be used to plot a straight line. Id. However, more recently AmerGen concluded the corrosion rate was zero based on only three points, one of which it has now recognized as unreliable. AmerGen now intends to confirm this conclusion by taking one more set of measurements before the start of any license extension period. Because at least four reliable sets of measurements are needed AmerGen would continue to have insufficient data to predict the corrosion rate reliably, even if conditions over time had remained constant. In fact, it is highly likely that conditions have changed since 1994, therefore realistic wall thickness measurements must be made as soon as possible to establish the current baseline and margins of safety. In addition, a worst case corrosion rate needs to be established for the current time period. This is obviously impossible based on just one point. I therefore suggest a pragmatic solution. AmerGen should use the corrosion data it gathered previously to estimate a statistically valid worst case corrosion rate based on previous conditions, which were with water and sand present and without a coating. This approach should have some inherent conservatism because the removal of the sand and the coating appeared to slow the corrosion for the period from 1992 to 1994. Thus, even if the coating has now become ineffective, the previous conditions should continue to provide a worst case scenario, provided a statistically valid approach is used.

c. Modeling and Statistical Analysis with Actual Data The calculation sheet (EX NC 3 (DRF 143071)) contains sufficient original data to analyze GPU Nuclear's evaluation of the UT and micrometer pit measurements.

Figure 7 is a schematic of what I understand was done to arrive at a representative remaining wall thickness in the former sandbed region in order to subsequently perform GE type vessel integrity calculations. First: UT measurements were taken from the inside as described earlier. Second: an imprint (or cast) was taken from the 13

outside in order to characterize the roughness of the corroded surface in addition to the UT measurements. The roughness is also characterized as "dimples". The depth of the dimples was measured from the imprint by means of a micrometer. Thus, Figure 7 shows the UT measurement (1) from the inside, which characterizes the remaining wall thickness. Second, the dimple depths were measured (repeatedly) and averaged (2). This average was added to each UT measurement. Third, a characteristic average dimple depth was determined and used as a global average to be used in all areas were imprints were not available, or where such were performed in a reduced fashion. The reason for this procedure is not entirely clear, other than hopefully arriving at a representative average, which could be the basis for the integrity calculations. As can be seen from Figure 7, the first location: if the average dimple depth is added to the UT remaining wall thickness, and then a global average dimple depth is being subtracted from the result, the actual pit depth may be reduced. In the second location the actual average pit depth may be increased by this procedure. However, a more detailed analysis of the some of the available data shows that this procedure performed by GPU Nuclear may show milder corrosion than what actually prevails. Detailed Data Analysis Appendix A of above reference document lists the measurements of impressions taken from Bay # 13, presumably the Bay where corrosion was the roughest. The average of all "dimple" measurements is 0.13 in with a standard deviation of 0.07 in. GPU Nuclear used the same average plus one standard deviation to arrive at the value of 0.2 in for the characterization of the average roughness of the corroded surface. It is not clear why only on standard deviation was added to the average when in fact 2 standard deviations represent a confidence limit of 95%. Hence it is my opinion that 0.27 in should have been used to represent worst case, or 270 mils. If this had been done, for instance, for the UT measurements summarized in Table 1-b (page 11) of referenced document, five of the 8 locations cited would have been below the acceptable criterion of 736 mils, while GPU Nuclear found all eight locations acceptable. It is therefore concluded that the procedure employed by GPU Nuclear is highly arbitrary, since the one vs. two standard deviations has not been explained. Extreme value Statistics Figure 8 shows the extreme value statistical evaluation of the UT Measurements in Bays 1 and 13. It can be seen that worst case penetrations can be predicted to be of the order of 550 to 600 mils, or dangerously close to the criteria for the remaining local wall thickness of 490 mils. Hence, predictions of this nature, which in the case of Bay 13 are reasonably accurate, (R = 0.95), are considerably less optimistic than those of GPU Nuclear. 14

Figure 9 shows a comparison of the measurements in Bay 13 of the UT remaining wall thickness and the dimple depths. The correlations are reasonably good. The prediction for the most severe dimple depth is about 300 mils, or 50% larger than the average used by GPU Nuclear, and more in line with the use of 2 standard deviations. Interestingly, the difference between the UT measured pit depth from the outside and the pit depth arrived at by micrometer measurements using the cast imprint turns out to be 200 mils at the lower pit depths and 300 mils at the higher pit depts. The 300 mil figure results in an average remaining wall thickness over the measured area of 1154 mils minus - 300 mils equals 854 mils, a number which has also been used in integrity calculations aimed at the buckling question. However, as pointed out earlier, this is clearly an average and one does not know how large the area is, which was further reduced by localized pitting. And herein lies the difficulty of what has been done in the past and what Amergen/Exelon proposes to do in the future. 99% of the sandbed region has not been monitored overtime and even the small areas that have been monitored are incompletely characterized. The overall area of the sandbed region is of the order of 300 ft. AmerGen are proposing more measurements at 12 6 inch by 6 inch areas, or a total of 3 ft2 . Thus only 1% of the total area is proposed to be monitored. In those small areas, point UT measurements, as have been done in the past, using a template and positioning the sensor always at the same location give information about the remaining wall thickness at this location (z- direction), but contain no information about the extent of the reduction in wall thickness around the point measurement (x-, y-directions). Hence around 93% of the 0.25 ft 2 area of each template remains unexplored. For these reasons it is urged that Amergen/Exelon consider using more modern UT methods which are capable of scanning large areas and can generate data in all three directions, x, y, and z. Conclusions This brief analysis of the original data presented in 1993 (measured in 1992) depicts a more severe corrosion situation than was extracted by GPU Nuclear on the basis of averages. Hence we think that a much more detailed analysis of the integrity of the remaining wall thickness is warranted and required, the repeated assertions that the coating has arrested any further corrosion not withstanding. Signed Rudolf H. Hausler 15

Figure 1 Hypothetical UT Measusrementsover a 6x6 Inch Grid 1.05 1 0.95

        ~0.9
       .~0.85 0.8 0.75 Z       0.7 0.65 0.6 1            3     4       5   6       7       -      ,i x Axis Figure 2 HypotheticalUT Measurements over a 6x6 Inch grid 1.05 1

0.95 0.9-0.85-Remaining Wall 0.8-thickness (inch) 0.75-0.7-0.65- 05S3 7 0.6 0.55 S3 x-axis 2 3 y-axis 7 17

Figure 3 0.8-0.7-0.6 - [ 0.5-0.4 0.3-0.2 Mean 0.34655 Std Dev 0.14139 Std Error Mean 0.02000 Upper 95% Mean 0.38674 Lower 95% Mean 0.30637 N 50.00000 Sum Weights 50.00000 Test for Normality Shapiro-Wilk W Test W Prob<W 0.943173 0.0284 Test Mean=value= ) Histogram of 49 Hypothetical UT Measurements over a 6x6 inch grid with 1 inch spacing. 18

a. Figure4 I

red variate By pit depth ) 0

             .1                .2             .3              .4              .5      .6 .7 .8 pit depth Unear Fit     I Lnr    Fr                it red variate = -2.4887      + 8.93985 pit depth Summary of Fit RSquare                                     0.954781 RSquare Adj                                 0.953839 Root Mean Square Error                      0.277923 Mean of Response                            0.629364 Observations (or Sum Wgts)                         50 "Analysis of Varifjan Source             DF      Sum of Squares        Mean Square               F Ratio Model                 1           78.284650              78.2845         1013.507 Error             48               3.707584               0.0772          Prob>F C Total            49             81.992234                                <,0001 (arameter     imates Term                    Estimate       Std Error     t Ratio      Prob>ltI Intercept            -2.488701        0.104952      -23.52        <0001 pit depth           8.9398488         0.280806        31.84       <.0001 Extreme Value Statistical Plot of 49 Hypothetical UT Measurements over a 6x6 inch grid. The last point at 0.77 inch pit depth is the most probable pit depth obtained by extrapolation if 100 data point had been measured. It is within the statistical 95% boundaries for the fit.

19

b. Figure5 UT Meassurments at Different Locations and Different Dates 1.1 1.05 SB 9-D 1* -U--SB 11-A

 - 0SB                                                                  11-C top 0.95                                                         SB 11-C bot
                                                                -- SB 13-A 0.9                                                         SB 13 A top
                                                                -SB 13 A bot i   0.85                                                         -SB     15-D 0.8, Sep-91 Jan-93        Jun-94       Oct-95        Mar-97 Date of Measurements 20

c. Figure6 Wall TWiess By Date 1.15 1.10 1.05 1.00 0.95 0.90 0.85 0.80 I

0.75

                                                                                    , II, Simmiwyof FR                                                                        su(   ar of FtR RSquare                                  0.010207                                    RSquare                             0.979337 RSqure At                                -0.03378                                    RSquare A4                          0.969651 Root Mean Square Error                   0.101758                                    Root Mean Square Error              0.017435 Mean of Response                         0.930158                                    Mean of Response                    0.930158 Obsenriatons (or Sum Wgts)                       48                                  Obrtý ion (or Sux wgts)                    48 (Anaysis of Valance                                                                  Analysis of Variance Source          DF      Sum of Squares          Mean Square          F Rato          Source          DF   Sum of Squares       Mean Square     F Ratio Model            2           0.00480510              0.002403        0.2320          Model           15       0.46103987           0.030736 101.1099 Error           45           0.48596232              0.010355        Prob>F          Error           32       0.00972755           0.000304   Prob>F C Total         47           0.47078742              0.010016        0.7939          C Total        47        0.47076742           0.010016   <.0001 (Means for Oneway~nv Level        Number             Mean     SWEror

["Mea Comparisns ) Sep-92 16 0.927588 0.02544 Sep-94 18 0.919394 0.02544 Sep-98 16 0.943494 0.02544 SWtError uses a pooled estimate of error valance [Mean Comparisons) Dlf=-MeanWl-MeanI Sep-90 Sep-92 Sep-94 Sep48 o.oooooo 0.015906 0.024100 Sep-92 .0.01591 0.008194 Sep-94 .0.0241 .0.00819 0.000000 Alpha- 0.05 Comparisons for al ,pairstusingTukey-Kramer HSD q! 2.42362 As(Otf)-LSO Sep-9 Sep- Sep-94 Sep-96 -0.08719 -0.07129 -0.0M309 Sep-92 .0.07129 -0.08719 .0.079 Sep-94 -0.0M30M .0.079 -0.08719 Postve values show pairs of means VWat ae signticany dfrent. Statistical Evaluations of all available UT Measurements performed in 1992, 1994 and 1996 on the drywell liner in the sandbed area 21

Figure 7 Schematic of Evaluation of Pit Depth Measurements and Averaging Procedure External Wall Toward Sandbed Internal Wall Wall thickness used for integrity evaluations: (1) + average of (2) - 200 = T (evaluation) Figure 8 Extreme Value StatisticalEvaluation of PittingMeasurements in Bay I and Bay 13 0.7 0.6 0.5 0o.4 S0.3 0.2 UT Measurements In Bay 13 UT Measurements In Bay I 0.1 -Linear (UT Measurements in Bay 13)

                                                 -    Lnear (UT MeasurementsIn Bay 1) 0+           i   I      I      Ii                                    I
      -1.5   -1    -0.5    0      0.5       1       1.5    2     2.5    3      3.5     4 red variate 22

Figure 9 Comparisonof UT Measuremetsand Micrometer Measuremens In Bay 13 Evaluated by Extreme Value Statistics 0.7 MicrometerPit Depth Measurements 0.6 0. - Linear

         -U-- UT Pit Depth Measurements (MicrometerPitDepth Measurements)

• l

                                                          . /_1   *   *   -

0.50.5 - Linear(UT Pit Depth Measurements) y 0.0918x+ 0.2982 S0.4 2  ;ý D0.3 0.2 y 0.0576x + 0.0986 0.1 0

      -2          -1             0          1          2          3            4 reduced variate 23

Citizen's Exhibit NC 1 I D. Ashley -FW: AuditQ &A (Question Numbers AMP- 141, 210, 356)a -page!l_ Citizen's Exhibit NCI From: <George.Beck@exeloncorp.com> To: <djal @nrc.gov>, <rkm @nrc.gov> Date: 04/05/2006 5:02:53 PM

Subject:

FW: Audit 0 & A (Question Numbers AMP-141, 210,356) Note: As originally transmitted this email was undeliverable to the NRC; it exceeded the size limit. It is being retransmitted without the AMP-210.pdf. This file will be reconstituted and sent In smaller ".pdf's; the first 11 pages are attached. George

                > -Original       Message-
                > From:             Beck, George
                > Sent:             Wednesday, April 05,2006 4:39 PM
                " To: Donnle Ashley (E-mail); 'Roy Mathew (E-mail) ' (E-mail)
               " Cc: Ouaou, Ahmed; Hufnagel Jr, John G; Warfel Sr, Donald B; Polaski, Frederick W
               " 

Subject:

Audit 0 & A (Question Numbers AMP-1 41,210,356)

               >  Donnie/Roy,
               >  Attached are the responses to AMP-21 0 and AMP-356 in an updated version of the reports from the AMP/AMR Audit database. Also Included Is a revised version of AMP-141. These answers have been reviewed and approved by Technical Lead, Don Warfel.
               " Regarding AMP-210, please note:
               " As ponted out In our response to NRC Question AMP-21 0, (8a)(1), "The 0.806" minimum average thickness verbally discussed with the Staff during the AMP audit was recorded In location 19A in 1994.

Additional reviews after the audit noted that lower minimum average thickness values were recorded at the same location in 1991 (0.803") and in September 1992 (0.800"). However, the three values are wthin the tolerance of +/-0.010" discussed with the Staff."

               > Regarding AMP-141, please note:
              > Our response to AMP-141 has been               revised to reflect additional information developed during the ongoing preparation of RAI responses.
              >  Please let John Hufnagel or me know ifyou have any questions.
              >  George
              >     >> <<Pages from AMP-210.pdf>>
              >                                   > > <<AMP-141.pdf>>
             >> <<*.MP-356.pdf>>

Q*******O, ********h*********hhI**~*4************O************&*******Q*tt This e-mail and any of Is attachments may contain Exelon Corporation proprietary Information, which Is privileged, confidential, or subject to copyright belonging to the Exelon Corporation family of Companies. This e-mall Is Intended solely for the use of the Individual or entity to which it Is addressed. ifyou are not the Intended recipient of this e-mail, you are hereby notified that any dissemination, distribution,

ID. Ashley - FW: AuditQ& A (Question Numbers AMP;141, 210, 356) - Page 2: copying, or action taken In relation to the contents of and attachments to this e-mail Is strictly prohibited and may be unlawful. If you have received this e-mail In error, please notify the sender immediately and permanently delete the original and any copy of this e-mail and any printout. Thank You.

             *******.4t¢ iI~tntttttt~eet
                           *****************Ot   ****O *  *O*   *O **....O......

CC: <ahmed.ouaou@exeloncorp.com>, <john.hufnagel@exeloncorp.com>,

             <donalcl.warfel@exeloncorp.com>, <fred.polaski@exeloncorp.com=.

I c.)ýq pGW)OOOOI.TMP Page 1

• I c:\temrAGWlOOOOl .T'AP Pane 1 Mail Envelope Properties (44343066.C5F: 19 : 7263)

Subject:

FW: Audit Q & A (Question Numbers AMP-141,210,356) Creation Date: 04105/2006 5:01:46 PM From: <George.Beck@exeloncorp.com> Created By: George.Beck@exeloncorp.com Recipients nrc.gov OWGWPO01.HQGWDO01 DJA1 (D. Ashley) nrc.gov TWGWPOO1.HQGWDOOI RKM (Roy Mathew) exeloncorp.com fred.polaski CC donald.warfel CC john.hufnagel CC ahmed.ouaou CC Post Office Route OWGWPOO1.HQGWDOO1 nrc.gov TWGWPOO1.HQGWDOO1 nrc.gov exeloncorp.com Files Size Date & Time MESSAGE 2679 05 April, 2006 5:01:46 PM TEXT.htm 5457 Pages from AMP-210.pdf 64593 AMP-141.pdf 47353 AMP-$56.pdf 71556 Mime.822 262768 Option:s Expiration Date: None Priority: Standard Reply Requested: No Returm Notification: None Concealed

Subject:

No ' Security: Standard 3

ENRCInformationRequestFor Item No Date Received: Source AMP-210 1/2412006 AMP Audit Topic: Status: Open IWE Document

References:

B.1.27 NRCRepresentative Morante, Rich AmerGen (Took Issue): Hufnagel, Joh Question Pages 25 through 31 of the PBD present a discussion of the OCGS operating experience. (8a)The following statements related to drywell corrosion In the sand bed region need further explanation and clarification: As a result of the presence of water In the sand bed region, extensive UT thickness measurements (about 1000) of the drywell shell were taken to determine Ifdegradation was occurring. These measurements corresponded t6 known water leaks and Indicated that wall thinning had occurred in this region. Please explain the underlined statement. Were water leaks limited to only a portion of the circumference? Was wall thinning found only Inthese areas? After sand removal, the concrete surface below the sand was found to be unfinished with Improper provisions for water drainage. Corrective actions taken In this region during 1992 included; (1) cleaning of loose rust from the drywell shell, followed by application of epoxy coating and (2) removing the loose debris from the concrete floor followed by rebuilding and reshaping the floor with epoxy to allc'w drainage of any water that may leak Into the region. UT measurements taken from the outside after cleaning verified loss of material projections that had been made based on measurements taken from the Inside of the drywell. There were, however, some areas thinner than projected; but In all cases engineering analysis determined that the drywell shell thickness satisfied ASME code requirements. Please describe the concrete surface below the sand that is discussed In paragraph above. Please provide the following information: (1) Identify the minimum recorded thickness Inthe sand bed region from the outside Inspection. and the minimum recorded thickness In the sand bed region from the inside Inspections. Is this consistent with previous Information provided verbally? (.806 minimum) (2) What was the projected thickness based on measurements taken from the Inside? (3) Describe the engineering analysis that determined satisfaction of ASME code requirements and Identify the minimum required thickness value. Is this consistent with previous Information provided verbally? (.733 minimum) (4) Is the minimum required thickness based on stress or buckling criteria? (5) Reconcilo and compare the thickness measurements provided in (1) and (3) above with the .736 minimum corroded thickness that was used In the NUREG-1540 analysis of the degraded Oyster

INRC Information Request Form Creek sand bed region. Evaluation of UT measurements taken from Inside the drywell, in the in the former sand bed region, in 1992, 1994, and 1996 confirmed that corrosion Is mitigated. It Is therefore concluded that corrosion in the sand bed region has been arrested and no further loss of material is expected. Monitoring of the coating in accordance with the Protective Coating Monitoring and Maintenance Program, will continue to ensure that the containment drywell shell maintains its intended function during the period of extended operation. NUREG-1540, published in April 1996, Includes the following statements related to corrosion of the Oyster Creek sand bed region: (page vii) However, to assure that these measures are effective, the licensee Is required to perform periodic UT measurements, and (page 2) As assurance that the corrosion rate Is slower than the rate obtained from previous measurements, GPU is committed to make UT measurements periodically. Please reconcile the aging management commitment (one-time UT inspection and monitoring of the condition of the coating) with the apparent requirementicommitment documented in NUREG-1540. (8b)The following statement related to drywell corrosion above the sand bed region needs further explanation and clarification: Corrective action for these regions Involved providing a corrosion allowance by demonstrating, through anaysis, that the original drywell design pressure was conservative. Amendment 165 to the Oyster Creek Technical Specifications reduced the drywell design pressure from 62 psig to 44 psig. The new des.ign pressure coupled with measures to prevent water Intrusion into the gap between the drywell shell and the concrete will allow the upper portion of the drywell to meet ASME code requirements. Please describe the measures to prevent water Intrusion into the gap between the drywell shell and the concrete that will allow the upper portion of the drywell to meet ASME code requirements". Are these measures to prevent water Intrusion credited for LR? If not, how will ASME code requirements be met during the extended period of operation? (8c)The following statements related to torus degradation need further explanation and clarification: Inspection performed In2002 found the coating to be in good condition in the vapor area of the Torus and vent header, and in fair condition in Immersion. Coating deficiencies in Immersion include blistering, random and mechanical damage. Blistering occurs primarily in the shell Invert but was also noted on the upper shell near the water line. The fractured blisters were repaired to reestablish the protective coating barrier. This Is another example of objective evidence that the Oyster Creek ASME Section XI, Subsection IWE aging management program can identify degradation and Implement corrective actions to prevent the loss of the contalnment's Intended function. While blistering Is considered a deficiency, it Is significant only when It Is fractured and exposes the base metal to corrosion attack. The majority of the blisters remain Intact and continues to protect the base metal; consequently the corrosion rates are low. Qualitative assessment of the Identified pits Indicate that the measured pit depths (50 mils max) are significantly less than the criteria established

INRCInformnationRequest Form m in Specification SP-1302-52-120 (141- 261 mils, depending on diameter of the pit and spacing between pits). Please confirm or clarify (1) that only the fractured blisters found in this Inspection were repaired; (2) pits were idontified where the blisters were fractured; (3) pit depths were measured and found to 50 mils max;, (4) the inspection Specification SP-1302-52-120 Includes pit-depth acceptance criteria for rapid evaluation of observed pitting; (5) the minimum pit depth of concern Is 141 mils (.141) and pits as deep as 261 mils (.261) may be acceptable. Please also provide the following Information: nominal design, as-built, and minimum measured thickness of the tows; minimum thickness required to meet ASME code acceptance criteria; the technical basis for the pitting acceptance criteria Include In Specification SP-1 302-52-120 Assigned To: Ouaou, Ahmed

Response

(8a) Questicn: Please explain the underlined statement. Were water leaks limited to only a porticn of the circumference? Was wall thinning only in these area?

Response

This statemont was not meant to indicate that water leaks were limited to only a portion of the circumference. The statement is meant to reflect the fact that water leakage was observed coming out of certain sand bed region drains and those locations were suspect of wall thinning. No. Wall thinning was not limited to the areas where water leakage from the drains was observed. Wall thinning occurred in all areas of the sand bed region based on UT measurements and visual inspection of the area conducted after the sand was removed in 1992. However the degree of wall thinning varied from location to location. For example 60% of the measured locations in the sand bed region (bays 1, 3, 5, 7, 9, and 15) indicate that the average measured drywell shell thickness Is nearly the same as the design nominal thickness and that these locations experienced negligible wall thinning; whereas bay 19A experienced approximately 30% reduction Inwall thickness. Question: Please discuss the concrete surface below the sand that is discussed in paragraph above.

Response

The concreto surface below the sand was Intended to be shaped to promote flow toward each of the five sand bed drains. However once the sand was removed It was discovered that the floor was not properly finished and shaped as required to permit proper drainage.' There were low points, craters, and rough surfaces that could allow moisture to pool instead of flowing smoothly toward the drains. These concrete surfaces were refurbished to fill low areas, smooth rough surfaces, and coat these surfaces with epoxy coating to promote Improved drainage. The drywell shell at juncture of the concrete flocr was sealed with an elastomer to prevent water Intrusion Into the embedded drywell shell. Question: Please provide the following Information:

                        *NRC InformationRequestFor (1) Identify the minimum recorded thickness In the sand bed region from the outside Inspection, and the minimum recorded thickness in the sand bed region from the inside inspections. Is this consistent with previous Information provided verbally? (.806 minimum)

(2) What was the projected thickness based on measurements taken from the inside? (3) Describe- the engineering analysis that determined satisfaction of ASME code requirements and identify the minimum required thickness value. Is this consistent with previous Information provided verbally? (.733 minimum) (4) Is the minimum required thickness based on stress or buckling criteria? (5) Reconcile and compare the thickness measurements provided in (1) and (3) above with the .736 minimum coarroded thickness that was used in the NUREG-1 540 analysis of the degraded Oyster Creek sand bed region.

Response

1. The minimum recorded thickness In the sand bed region from outside Inspection Is 0.618 Inches.

The minimum recorded thickness In the sand bed region from Inside Inspections is 0.603. These minimum recorded thicknesses are Isolated local measurement and represent a single point UT measurement. The 0.806 Inches thickness provided to the Staff verbally is an average minimum general thickness calculated based on 49 UT measurements taken In an area that Is approximately 6"x 6". Thus the two local isolated minimum recorded thicknesses cannot be compared directly to the general thickness of 0.806". the 0.806" minimum average thickness verbally discussed with the Staff during the AMP audit was recorded in location 19A In 1994. Additional reviews after the audit noted that lower minimum average thickness values were recorded at the same location in 1991 (0.803") and in September 1992 (0.800"). However, the three values are within the tolerance of +1-0.010" discussed with the Staff.

2. The minimum projected thickness depends on whether the trended data Is before or after 1992 as demonstrated by corrosion trends provided in response to NRC Question #AMP-356. For license renewal, using corrosion rate trends after 1992 Is appropriate because of corrosion mitigating measures such as removal of the sand and coating of the shell. Then, using corrosion rate trends based on IE92, 1994, and 1996 UT data; and the minimum average thickness measured In 1992 (0.800"), the minimum projected average thickness through 2009 and beyond remains approximately 0.800 inches. The projected minimum thickness during and through the period of extended operation will be reevwluated after UT Inspections that will be conducted prior to entering the period of extended operation, and after the periodic UT inspection every 10 years thereafter.

3.The engineering analysis that demonstrated compliance to ASME code requirements was performed in two parts, Stress and Stability Analysis with Sand, and Stress and Stability Analyses without Sand. The analyses are documented in GE Reports Index No. 9-1, 9-2, 9-3, and 9-4, were transmitted to the NRC Staff in December 1990 and In 1991 respectively. Index No. 9-3 and 9-4, were revised later to correct errors Identified during an internal audit and were resubmitted to the Staff in Janvary 1992 (see attachment I & 2). The analyses are briefly described below. The drywell shell thickness In the sand bed region Is based on Stability Analysis without Sand. As

FNýlnomdnRequest described Indetail In attachment I &2, the analysis Is based on a 36-degree section model that takes advantage of symmetry of the drywell with 10 vents. The model Includes the drywell shell from the base of the sand bed region to the top of elliptical head and the vent and vent header. The torus is not included Inthis model because the bellows provide a very flexible connection, which does not allow significant structural Interaction between the drywell and the tows. The analysis conservatively assumed that the shell thickness In the entire sand bed region has been reduced uniformly to a thickness of 0.736 Inches. As discussed with the Staff during the AMP audit, the basic approach used In the buckling evaluation follows the methodology outlined InASME Code Case N-284 revision 0 that was reconciled later with revision I o" the Code Case. Following the procedure of this Code Case, the allowable compressive stress is evaluated In three steps. In the first step, a theoretical buckling stress is determined, and secondly modified using appropriate capacity and plasticity reduction factors. In the final step, the allowable compressive stress Is obtained by dividing the buckling stress calculated Inthe second step by a safety factor of 2.0 for Design and Level A &B service conditions and 1.67 Level C service conditions. Using the approach described above, the analysis shows that for the most severe design basis load combinations, the limits of ASME Section III, Subsection NE 3213.10 are fully met. For additional details refer to Attachment I & 2. As described above, the buckling analysis was performed assuming a uniform general thickness of the sand bed region of 0.736 inches. However the UT measurements identified isolated, localized areas where the drywell shell thickness Is less than 0.736 Inches. Acceptance for these areas was based on engineering calculation C-1302-187-5320-024. The calculation uses a Local Wall Acceptance Criteria". This criterion can be applied to small areas (less than 12" by 12"), which are less than 0.736" thick so long as the small 12" by 12" area Is at least 0.536" thick. However the calculation does not provide additional criteria as to the acceptable distance between multiple small areas. For example, the minimum required linear distances between a 12" by 12" area thinner than 0.736" but thicker than 0.536" and another 12" by 12" area thinner than 0.736" but thicker than 0.536" were not provided. The actual data for two bays (13 and 1) shows that there are more than one 12" by 12" areas thinner than 0.736" 5ut thicker than 0.536". Also the actual data for two bays shows that there are more than one 2 %'"diameter areas thinner than 0.736" but thicker than 0.490". Acceptance is based on the following evaluation. The effect of these very local wall thickness areas on the buckling of the shell requires some discussion of the buckling mechanism Ina shell of revolution under an applied axial and lateral pressure load. To begin the discussion we will describe the buckling of a simply supported cylindrical shell under the Influence of lateral pressure and axial load. As described in chapter 11 of the Theory of Elastic Stability, Second Edition, by Timoshenko and Gere, thin cylindrical shells buckle Inlobes In both the

Smation RequestFor axial and circumferential directions. These lobes are defined as half wave lengths of sinusoidal functions. The functions are governed by the radius, thickness and length of the cylinder. Ifwe look at a specific thin walled cylindrical shell both the length and radius would be essentially constants and If the thickness was changed locally the change would have to be significant and continuous over a majority of the lobe so that the compressive stress Inthe lobe would exceed the critical buckling stress under the applied loads, thereby causing the shell to buckle locally. This approach can be easily extrapolated to any shell of revolution that would experience both an axial load and lateral pressure as Inthe case of the drywell. This local lobe buckling is demonstrated in The GE Letter Report "Sandbed Local Thinning and Raising the Fixity Height Analysis" where a 12 x 12 square inch section of the drywell sand bed region Is reduced by 200 mils and a local buckle occurred in the finite element elgenvalue extraction analysis of the drywell. Therefore, to Influence the buckling of a shell the very local areas of reduced thickness would have to be contiguous and of the same thickness. This Is also consistent with Code Case 284 In Section -1700 which Indicates that the average stress values Inthe shell should be used for calculating the buckling stress. Therefore, an acceptable distance between areas of reduced thickness is not required for an acceptable buckling analysis except that the area of reduced thickness Is small enough not to Influence a buckling lobe of the shell. The very local areas of thickness are dispersed over a wide area with varying thickness and as such will have a negligible effect on the buckling response of the drywell. In addition, these very local wall areas are centered about the vents, which significantly stiffen the shell. This stiffening effect limits the shell buckling to a point Inthe shell sand bed region which is located at the midpoint between two vents. The acceptance criteria for the thickness of 0.49 inches confined to an area less than 2,4 inches in diameter experiencing primary membrane + bending stresses Is based on ASME B&PV Code, Section III, Subsection NE, Class MC Components, Paragraphs NE-3213.2 Gross Structural Discontinuity, NE-3213.10 Local Primary Membrane Stress, NE-3332.1 Openings not Requiring Reinforcement, NE-3332.2 Required Area of Reinforcement and NE-3335.1 Reinforcement of Multiple Openings. The use of Paragraph NE-3332.1 Is limited by the requirements of Paragraphs NE-3213.2 and NIE-3213.10. In particular NE-3213.10 limits the meridional distance between openings without reinforcement to 2.5 x (square root of Rt). Also Paragraph NE-3335.1 only applies to openings In shells that are closer than two times their average diameter. The Implications of these paragraphs are that shell failures at these locations from primary stresses produced by pressure cannot occur provided openings Inshells have sufficient reinforcement. The current design pressure of 44 pslg for drywell requires a thickness of 0.479 Inches in the sand bed region of the. drywell. A review of all the UT data presented InAppendix D of the calculation Indicates that all thicknesses in the drywell sand bed region exceed the required pressure thickness by a substantial nargin. Therefore, the requirements for pressure reinforcement specified Inthe previous paragraph are not required for the very local wall thickness evaluation presented in Revision 0 of Calculation C-1302-187-5320-024. Reviewing the stability analyses provided In both the GE Report 9-4 and the GE Letter Report Sand bed Local Thinning and Raising the Fixity Height Analysis and recognizing that the plate elements In the sand bed region of the model are 3"x 3" it is clear that the circumferential buckling lobes for the

rNRC Information Request Form! drywell are substantially larger than the 2 1a Inch diameter very local wall areas. This combined with the local reinforcement surrounding these local areas Indicates that these areas will have no Impact on the buckling margins Inthe shell. It Is also clear from the GE Letter Report that a uniform reduction in thickness of 27% to 0.536" over a one square foot area would only create a 9.5% reduction In the load factor and theoretical buckling stress for the whole drywell resulting Inthe largest reduction possible. In addition, to the reported result for the 27% reduction Inwall thickness, a second buckling analysis was performed for a wall thickness reduction of 13.5% over a one square foot area which only reduce:l the load factor and theoretical buckling stress by 3.5% for the whole drywell resulting in the largest reduction possible. To bring these results into perspective a review of the NDE reports indicate there are 20 UT measured areas in the whole sand bed region that have thicknesses less than the 0.736 inch thickness used In GE Report 9-4 which cover a conservative total area of 0.68 square feet of the drywell surface with an average thickness of 0.703" or a 4.5% reduction in wall thickness. Therefore, to effectively change the buckling margins on the drywell shell in the sand bed region a reduced thickness would have to cover approximately one square foot of shell area at a location Inthe shell that Is most susceptible to buckling with a reduction in thickness greater than 25%. This leads to the conclusion that the buckling of the shell Is unaffected by the distance between the very local wall thicknesses, Infact these local areas could be contiguous provided their total area did not exceed one square foot and their average thickness was greater than the thickness analyzed in the GE Letter Report and provided the methodology of Code Case N284 was employed to determine the allowable buckling load for the drywell. Furthermore, all of these very local wall areas are centered about the vents, which significantly stiffen the shell. This stiffing effect limits the shell buckling to a point Inthe shell sand bed region', which is located at the midpoint between two vents. The mlnimumn thickness of 0.733" Is not correct. The correct minimum thickness is 0.736".

4. The minimum required thickness for the sand bed region Is controlled by buckling.

5. We cannot reconcile the difference between the current (lowest measured) of 0.736" In NUREG-1540 and tha minimum measured thickness of 0.806 inches we discussed with the Staff. Perhaps the value in NUREG-1540 should be labeled minimum required by the Code, as documented In several comr.spondences with the Staff, Instead of lowest measured. In a letter dated September 15, 1995, GPU provided the Staff a table that lists sand bed region thicknesses. The table Indicates that nominal thickness Is 1.154". the minimum measured thickness In 1994 is 0.806", and the minimum thickness required by Code is 0.736". These thicknesses are consistent with those discussed with the Staff during the AMP/AMR audit.

Question: NUREG-1540, published in April 1996, Includes the following statements related to corrosion of the Oyster Creek sand bed region: (page vii) However, to assure that these measures are effective, the licensee is required to perform periodic UT measurements, and (page 2) As assurance that the corrosion rate Is slower than the rate obtained from previous measurements, GPU Is committed to make UT measurements periodically. Please reconcile the aging management commitment (one-time UT Inspection and monitoring of the condition of the coating) with the apparent requirementcommitment documented In NUREG-1540.Please reconcile the aging management commitment (one-time UT Inspection and monitoring of the condition of the coating) with the apparent requirement/commitment documented In NUREG-1540. (io

Ifr IR aion ReII uesI

Response

Our review of NUREG-1540, page 2 Indicates that the statements appear to be based on 1991, or 1993 GPU commitment to perform periodic UT measurements. In fact UT thickness measurements were taken :n the sand bed region from inside the drywell In 1992, and 1994. The trend of the UT measurements Indicates that corrosion has been arrested. As results GPU Informed NRC In a letter dated September 15, 1995 (ref. 2) that UT measurements will be taken one more time, in 1996, and the epoxy coating will be Inspected In 1996 and, as a minimum again in 2000. The UT measurements were taken In 1996, per the commitment, and confirmed corrosion rate trend of 1992 and 1994. The results of 1992, 1994, and 1996 UT measurements were provided to the Staff during the AMPIAMR audits. In response to GPU September 15, 1995 letter, NRC Staff found the proposed changes to sand bed region commitments (I.e. no additional UT measurements after 1996) reasonable and acceptable. This response is documented in November 1, 1995 Safety Evaluation for the Drywell Monitoring Program. For license renewal, Oyster Creek was previously committed to perform One-Time UT inspection of the drywell shell in the sand bed region prior to entering the period of extended operation. However. in response to NRC Question #AMP-141, Oyster Creek revised the commitment to perform UT inspections .eriodically. The Initial Inspection will be conducted prior to entering the period of extended operation and additional Inspections will be conducted every 10 years thereafter. The UT measureme.its will be taken from Inside the drywell at same locations as 1996 UT campaign (8b) Question: Please describe the measures to prevent water intrusion into the gap between the drywell shell and the concrete that will allow the upper portion of the drywell to meet ASME code requirements. Are these measures to prevent water intrusion credited for LR? If not, how will ASME code requirements be met during the extended period of operation?

Response

The measures taken to prevent water intrusion Into the gap between the drywell shell and the concrete that will allow the upper portion of the drywell to maintain the ASME code requirements are,

1. Cleared the former sand bed region drains to Improve the drainage path.

2. Replaced reactor cavity steel trough drain gasket, which was found to be leaking.

3. Applied stainless steel type tape and strippable coating to the reactor cavity during refueling outages to seal Identified cracks in the stainless steel liner.

4. Confirmed that the reactor cavity concrete trough drains are not clogged

5. Monitored former sand bed region drains and reactor cavity concrete trough drains for leakage during refueling outages and plant operation.

Oyster Creek Is committed to Implement these measures during the period of extended operation. (8c) Please confirm or clarify (1) that only the fractured blisters found in this inspection were repaired; (2) pits were Identified where the blisters were fractured; (3) pit depths were measured and found to

lNIRC Information Requestrm! 50 mils max:; (4) the inspection Specification SP-1302-52-120 includes pit-depth acceptance criteria for rapid evaluation of observed pitting; (5) the minimum pit depth of concern is 141 mils (.141) and pits as deep as 261 mils (.261) may be acceptable.

Response

(1) Specification SP-1302-52-120, Specification for Inspection and Localized Repair of the Torus and Vent System Coating, specifies repair req'uirements for coating defects exposing substrate and fractured blisters showing signs of corrosion. The repairs referred to in the inspection report included fractured blisters, as well as any mechanically damaged areas, which have exposed bare metal showing signs of corrosion. Therefore, only fractured blisters would be candidates for repair, not those blisters that remain Intact. The number and location of repairs are tabulated in the final Inspection report prepared by Underwater Construction Corporation. (2) Coating deficiencies In the Immersion region Included blistering with minor mechanical damage. Blistering occurred primarily In the shell Invert but was also noted on the upper shell near the water line. The majority of the blisters were Intact. Intact blisters were examined by removing the blister cap exposing the substrate. Corrosion attack under non-fractured blisters was minimal and was generally limited to surface discoloration. Examination of the substrate revealed slight discoloration and pitting with pit depths less than 0.001. Several blistered areas included pitting corrosion where the blisters were fractured. The substrate beneath fractured blisters generally exhibited a slightly heavier magnetite oxide layer and minor pitting (less than 0.010") of the substrate. (3) In addition to blistering, random deficiencies that exposed base metal were Identified in the torus immersion region coating (e.g., minor mechanical damage) during the 19R (2002) torus coating inspections. They ranged in size from 1116" to Y" in diameter. Pitting in these areas was qualitatively evaluated and ranged from less than 10 mils to slightly more than 40 mils In a few Isolated cases. Three quanlitative pit depth measurements were taken in several locations In the Immersion area of Bay 1. Pit depths at these sites ranged from 0.008u to 0.042" and were judged to be representative of typical conditions found on the shell. Prior to 2002 Inspection 4 pits greater than 0.040" were Identified. The pits depth are 0.058" (1 pit in 1988), 0.05" (2 pits in 1991), and 0.0685" (1 pit in 1992). The pits were evaluated against the local pit depth acceptance criteria and found to be acceptable. (4) Specification SP-1302-52-120, Specification for Inspection and Localized Repair of the Torus and Vent System Coating, includes the pit-depth acceptance criteria for rapid evaluation of observed pitting. The acceptance criteria are supported by a calculation C-1302-187-E310-038. Locations that do not meet the pit-depth acceptance criteria are characterized based on the size of the area, center to center distance between corroded areas, the maximum pit depth and location In the Torus based on major structural features. These details are sent to Oyster Creek Engineering for evaluation. (5) The acceptance criteria for pit depth Is as follows: -Isolated Pits3 of 0.125" In diameter have an allowed maximum depth of 0.261" anywhere In the shell provided the center to center distance between the subject pit and neighboring isolated pits or areas of pitting corrosion Is greater than 20.0 Inches. This Includes old pits or old areas of pitting corrosion that have been filled and/or re-coated.

JNRCInformation Reques

 -Multiple Pits that can be encompassed by a 2-1/2" diameter circle shall be limited to a maximum pit depth of 0.141" provided the center to center distance between the subject pitted area and neighboring isolated pits or areas of pitting corrosion is greater than 20.0 Inches. This Includes old pits or old areas of pitting corrosion that have been filled and/or recoated.

Question: Please also provide the following Information: nominal design, as-built, and minimum measured thickness of the tows; minimum thickness required to meet ASME code acceptance criteria; the technical basis for the pitting acceptance criteria include In Specification SP-1302-52-120

Response

Submersed area: (a) The nominal Design thickness is 0.385 inches (b) The as-built thickness Is 0.385 Inches (c) The minimum uniform measured thickness Is, 0.343 inches - general shell 0.345 inches - shell - ring girders 0.345 Inches - shell - saddle flange 0.345 inches - shell - tors straps (d) The minimum general thickness required to meet ASME Code Acceptance Is 0.337 inches. Technical basis for pitting acceptance criteria included in Specification SP-1302-52-120 is based on engineering calculation C-1302-187-E310-038. At the time of preparation of calculation C-1 302-187-E310-038 In2002 there were no published methods to calculate acceptance standards for locally thinned areas in ASME Section III or Section VIII Pressure Vessel codes. Therefore, the approach in Code Case N-597 was used as guidance Inassessing locally thinned areas in the Tows. This Is based on the similarity in approaches between Local Thinning Areas described in N597 and Local Primary Stress areas described In Paragraph NE3213.10 of the ASME B&PV Code Section III, particularly small areas of wall thinning which do not exceed 1.0 x (square root of Rt). In addition, the ASME B&P'I Code Section III, Subsection NB, Paragraph NB-3630 allows the analysis of pipe systems In accordance with the Vessel Analysis rules described In Paragraph NB-3200 of the same Subsection as an alternate analysis approach. Therefore, the approach used In N597 for local areas of thinning vias probably developed using the rules for Local Primary Membrane Stress from paragraph NB-3200 In particular Subparagraph 3213.10. The Local Primary Stress Limits In NB-3213.10 are similar to those discussed InSubsection NE, Paragraph NE-3213.10. Since the Code Case had not yet been Invoked Into the Section XI program, the calculation provided a reconciliation of the results obtained from the code case against the ASME Section III code requirements as discussed above. This reconciliation demonstrated that the approach In N597 used on a pressure vessel such as the Torus would be acceptable since the results are conservative compared to the previous work performed In MPR-953 and Lm(a) (defined In N597 Table- 3622-1) £ (Rmintmln)1/2. Currently, the maximum pit depth measured In the Torus Is a 0.0685" ( measured in 1992 in bay 2). It was evaluated as acceptable using the design calculations existing at that time and was not based on

FNRC Information"Request om Calculation C-1302-187-E310-038. This remains the bounding wall thickness in the Tows. The criterion developed In 2002 for local thickness acceptance provides an easier method for evaluating as-found pils. The results were shown to be conservative versus the original ASME Section III and VIII Code requirements for the Tors. The Torus lispection program Is being enhanced per IR 373695 to improve the detail of the acceptance criteria and mkrgin management requirements using the ASME Section III criteria. The approach u,-ed in C-1302-187-E310-038 will be clarified as to how It maintains the code requirements. If Code Case N-597-1 Is required to develop these criteria for future inspections, NRC review and approval will be obtained. It should also be noted that the program has established corrosion rate criteria and continues to periodically monitor to verify they remain bounded. LRCR #: LRA A.5 Commitmentt #: IR#: Approvals: PreparedBy.: Ouaou, Ahmed 4/5/2006 ReviewedBy: Miller, Mark 4/512006 ApprovedBy.: Warfel, Don 4/5/2006 NRCAccepltnce (Date):

INRC InformationRequest FormI Item No DateReceived: Source AMP-356 2/16/2006 AMP Audit Topic: Status: Open IWE Document

References:

NRC Representative Morante, Rich AmerGen (Took Issue): Ouestion IWE AMP Question 4 WE AMP Revised Feb. 17, 2006 R. Morante (AMP-356) (1) Identify the specific locations around the circumference in the former sandbed region where UT thickness readings have been and will be taken from Inside containment. Confirm that all points previously recorded will be included In future Inspections. (2) Describe, the grid pattern at each location (meridional length, circumferential length, grid point spacing, total number of point readings), and graphically locate each grid pattern within the former sandbed region. (3) For eac& grid location, submit a graph of remaining thickness versus time, using the UT readings since the initiation of the program (both prior to and following removal of the sand and application of the external coating). (4) Clearly describe the methodology and acceptance criteria that Is applied to each grid of point thickness readings, including both global (entire array) evaluation and local (subregion of array) evaluation. Assigned To: Ouaou, Ahmed

Response

Response:

1. The circumference of the drywell Is divided Into 10 bays, designated as Bays 1, 3, 5, 7, 9, 11,13, 15, 17, and 19. UT thickness readings have been taken in each bay at one or more locations. The specific locations around the circumference In the former sand bed region where UT thickness reading have been taken from Inside containment are Bay 1D, 3D, 5D, 7D, 9A, 9D, 1lA, IIC, 13A, 13C, 13D, 15A, 15D, 17A, 17D, 17/19 Frame, 19A, 19B, and 19C. For each location, UT measurements were taken centered at elevation 11 '-3". These represent the locations where UT measurements were taken in 1992, 1994, and 1996.

V<

[NR'CInformationRequest Form[ In addition UT measurements were taken one time inside 2 trenches excavated in drywell floor concrete. The purpose of these UT measurements is to determine the extent of corrosion in the lower portions of the sand bed region prior to removing the sand and making accessible for visual Inspection. Future UT thickness measurements will be taken at the same locations as those inspected In 1996 in accordance with Oyster Creek commitment documented InNRC Question #AMP-209.

2. For locations where the Initial Investigations found significant wall thinning (9D, 11A, 11 C, 13A, 13D, 15D, 17A, 17D, 17119 Frame, 19A, 19B, and 19C) the grid pattern consists of 7 x 7 grid centered at elevation 1 '-3 (meridian) and centered at the centerline of the tested location within each bay, which consists of 6"x 6" square template. The grid spacing is 1"on center. There are 49 point readings. For graphical location of the grid, refer to attachment 1.

For locations where the Initial Investigations found no significant wall thinning (ID, 3D, 5D, 7D, 9A, 13C, and 15A) the grid pattern consists of I x 7 grid centered at elevation 11"-3" (meridian) on V" centers. There are 7 point readings. For graphical location of the grid, refer to attachment 1.

3. A graph representing the remaining thickness versus time using UT reading since the initiation of the program (both prior to and following removal of the sand and application of the external coating) for location GD, 11A, 11C, 13A, 13D,15D,17A,17D,17119, 19A, 19B, and 19C is included in the attached graph. Other locations (i.e. ID, 3D, 5D, 7D, 9A, 13C, and 15A) are not included because wall thinning Is not significant and the trend line will be essentially a straight line.

4. The methodology and acceptance criteria that is applied to each grid of point thickness readings, Including both global (entire array) evaluation and local (subregion of array) is described in engineering specification IS-328227-004 and In calculation No. C-1 302-187-5300-011. These documents wvere submitted to the NRC In a letter dated November 26, 1990 and provided to the Staff during the AMPIAMR audit. A brief summary of the methodology and acceptance criteria Is described below.

The initial lo.zations where corrosion loss was most severe In 1986 and 1987 were selected for repeat Inspection oier time to measure corrosion rate. For location where the Initial investigations found significant wall thinning UT inspection consists of 49 individual UT data points equally spaced over a 6"x 6" area. Each new set of 49 values was then tested for normal distribution. The mean values of each grid were then compared to the required minimum uniform thickness criteria of 0.736. In addition each Individual reading Is compared to the local minimum required criteria of 0.49. The basis for the required minimum uniform thickness criteria and the local minimum required criteria Is provided In response to NRC Question #AMP-210. A decrease Inthe mean value over time Is representative of corrosion. Ifcorrosion does not exist, the mean value will not vary with time except for random variations In the UT measurements.

INRC Information Request Fornl If corrosion is continuing, the mean thickness will decrease linearly with time. Therefore the curve fit of the data is tested to determine if linear regression is appropriate, In which case the corrosion rate Is equal to the slope of the line. If a slope exists, then upper and lower 95% confidence intervals of the curve fit are calculated. The lower 95% confidence interval Is then projected into the future arid compared to the required minimum uniform thickness criteria of 0.736. A similar process is applied to the thinnest individual reading in each grid. The curve fit of the data is tested to determine if linear regression Is appropriate. If a slope exists, then the lower 95% confidence nterval Is then projected into the future and compared to the required minimum local thickness criteria of .49. LRCR #: LRA 4.5 Commitment #: IR#I: Approvals: PreparedBy." Ouaou, Ahmed 4/412006 Reviewed By: Getz, Stu 415/2006 ApprovedBy- Warfel, Don 415/2006 NRCAcceptancc (Date):

Ir Ir- r[ *rL r r ... I rI yr r Ir r .. r r Dr r. Oyster Ceek Drywell Vessel Corrosion Rate Trending Program Average Measured Thicknesses 7o~ F.b.I P4 P 81 A@7 "71 JL44 095487 .~4 $*" .164 AM" 1"41 olw 079.7 I I I - 1.115 3D 1 1.10 A.

                                                                                                    .1?

3D 1. 1.012 7D ...... 7.1? 1.02, - .s 1.13L I A I : _. 1. 7 171 VA !01.7, -.-- 1 1.151 1 . 7 ,712 7.71 1. _o. ... 7.7 Ilf 1.3 I ,A 0.11 7. 7.*& 0 .*1 0 S.AM 0. 0."1 1e 0.81t 7.1 0.e4 0.4 GAIN 0 _ GO S 6.111 .gs 0.. 7 1 0.7 0.0? D$ 0.. 0.am 0.A 0. 7534 Go @tax 1.04 7.7 1 1.0 1.07 1.0 1 17 7.9 0.52 7.77 A..,,,070

                ,                                                                                               7           0.                                  1.017         .074         7.0"      OA              .04

7. 78.0 710 7. 0.7876 74 C

75. 0. S1A 0.a 0.7 &W,93 on

                                                                                                                       ..                ..              7     1.94          1.74          1A     .            .A 1           W      7.001                7.7                  7 TIA                                                 1                                    0..0737 1321:           .34 I     7 1.3    .2.i        79. 0.1.
                                                                                                                                                         "1 i   7.1
                                                                                                                                                                 .1         " 7.722

1. 71

• 11411 7DOtt"$ 1 ,Oi O.1t , O.14 81 015"' 9 am Ut to"1S 4 a1.l1 I a1. 0. .

                                                                                                                                                                 .87        "'1.1I
                                                                                                                                                                               -82                    .

OL 1I. Sl iA 04 8011 ,,11 ,,1 11 OA IA . H :

___.. ___m ___. L- 6=m No m--i k w ~ =a Lhmoi faw Sandbed Bay 9 Location D 1~.10 1.05 C 0.95 0, 901 :L . CY IýP: TIMEL Slope Beat Est Date bogfoadl twavilu IcT~kne" ~ iami~fr Required Thkkaua Aveag Snc, ft 4.12 C32 09U1192 1.1Wt D"-" IelMay47 Aug47 S*1p47 JsIJE Get-83 JUUJl Sqp43 Feb-90 Apr-90 M-941 1*1.-*l Ks,.9 ay4 sqp4j Sep4 e4 90 1,0715 M9 LI~ ~6 L217 IJ1 0.26 1,47 01Z4 14000 1=0J &GfLG 1.008

II

• r+i, y VL-- r- (r,- rý"-

                                                                                                   -~       -

rP--,

                                                                                                                           -

rPP--

                                                                                                                                  ~-       -

rp,-

                                                                                                                                               ~.-

wpl-, U rl-U~ (9---- spl-U or, U ff-U-

                                                                                                                       %

08 r 0), tk 141, Ilk,

                                                                                                               ý&        109 TIME Based on Cailcuiation C1302*t1875300-021
                             *t Est pos.             Date              Average Since 194i                              Original Nominal Tiunejig                              Unformiakeufred Thtmdis
                                                                                                                                                              ?Alnlduxmr
                  -.. 17       .031              5O10/9           0.8251                                        114 Dates        e46     Feb47. Apr47      May487 Aug-.7   Sep417*   UI-88 Oct4*8.. Jun4f9        Sep-9. Feb40 Apr-90           Mar-41  May,,1  Nov-911 May42     Sep-O2   Se11004  Sepo6 11A                        0.9187   0.90464 0.92209  0.0052   0.913 0.0882      0.381i      0.8916    0.8808 0.87D4        0.3446   0.844 0.1326     04    0.8252     0.82     0.83,
    . ....

1,~~ L~ & I I _ 6 _L__ ~m 6 N a

                                                                                                                                                                       -ilm        - -. b, w
                                                                       ........
                                                                   .....
                                                              $SandbedBay I11 Location C 1,05, o0.95 M
           -       0.9 0.85 0.8            - -7 TIME                         _   _    _   _      _     _       _    _      _   _

Based on C~alculation C413O2-1187-5300-02i Besut Beat Averago Alv.R4mearhkea aIpnlatn"eard~ka Slope Slope Est. E t. Daet Sinco jno re41 O,14TikkwMimu no i~"Tkw LW 11,1gh 199U2 10922 4.0143 -0.0171 0.480.9465 42 06It01J2 6.841 0.9934 1,154"< 7 Dates Doc-SC Feib47 Apr47T May-67TAug-$? Se-T. ~JuI-f ýlOct4 jun-9 Sep489 Feb-90 ~ApT490 Mar491May-91 Nov-91 May42 SeSe Sep4 Sep46G Bottomi SID-4490.95364 0.I1161 0.961 0.8907 0,5168 0*907 0.9703 0.86 013TS 7 6026 0.9563 0.S82 Q.ils01 #5503 *Saa IIC'TOj 1&046 1-1036 1.0191 10454 11.0089 ~1.0$ 1.05 0.9522 0.977 0)981 1,018 0.9643 '1.01 %1t960 0,083n 1.0418

-r r~r~_r"~r r r r r r r r K C, c

  ,Basied nCacuiftloti C-130kw.m-S~ov Slope BvtExt        ~   Date          AveagoShc I                               OriginalNominalIlickldrni         MMijimova Unflotm Requi*e Thickfien
   -0.012 0.0442,           O0911192.                                                                                0.736" Detu"   De-      Fob-hT   Apr47 NMay-7 Aug-8 Sep487 jui-l4 oct-88 .ltun4. Sep489 Feb-90 Apr-90 Mar-9l      May-91 No~v41 IMay42       Sep42 Sop-9 4

tep46 0.91908 a.9063 o.i828 068n3 0*61 0.1531 0.8545 0.8529 0.8486 0.8645 0.8576 0.6218 0,843

bwi 6iiii bo No_ h." bok .w 6 ibfwý -,ll Laiii ____ 1- ==j Sandbed Bay 13 Location C I' U)

0.95 0.8 "qj Cf (0 rN Co o0) C) N u* to :r c,

                     ~~~D.              4D~O~   0       t-00     IV             4) 0)          QV             M       M        CD0           0       0       C)        0    al             ~      Cl      C        0 TIME Base~don.Calculation C-1302-16753004021 Slope Slope         Best Est LOW-     Best Est. Hig Date       Average Since 1992     Average Since 1992        Original Nominal Thickness        Minium UuforxuJequredTlxfkknss
     -0.013-0-0146                   1.06            0~5/01192 1.~0505                 0.91 14                                   -7~36"          0.54 Dateo       lDec-8      Feb-81 Apr487       May487     Aug487 tep-87 Jul-       Oct-88 Jtiti49 $op-89    Feb-90 Apr490 Mar-91 Nlay-91 Iov-B1 klay42           Sep42 S"p4       ep.045 13~C Tp1,0722                                                                                            '1.0438 1.04'7    .089 -t.0546     11.111311.6663

r T~Tr rz~ 1 TU~~7~ Z - --------- r IF ri r~r Saindbed Bayi5 Location D 1i15 0," 0 1.i05 0.95 D f-C (D -CO 0 O ,.L ( -C )0CoI CO oC Ommm0 ) )00 0C )C __ME _ _ _ _ _ - Based Om Catoutatfoti C-32-i6?453GO@421 Soe Best Eft Dato Average Since 1992 OriginialNo,,tiaI 71itknew 0.736" 4.O01 11055 " /0i92 1.0518 Feb471 Aug-87 Sep487 JhA48 Oct-88 Jun-89 Sep489 Feb-00 Apr-09 Mar491 May.9l :14oivl Apr-47 MAay-81 BMy42 SOP42 8ep44 Sep#6 Dates, Dec-6 1.Os: 1.0516 1.0565 1.0596 1.0502 1.0417 1.0652 1.0577 1.053 1.066 I Ass s1.05 1.0"

SandbecIBa~y 17 Location A S1.05 0.9ý CD co 0) C? 0 <) o c, 0 0o 0) 0 C) 0 C) CD) (1) "( 0 AD) TIME fale. Va.H Fail? AspJ7 aiV,.t *44?7 Sa .W i GOsW.lAW5 Sp4 ft.540 Apv. 04 M ay41 m"I 41 Now-it 9.0' UaP." S$004 iTAadwlW J7.4 0,W10 0140812 10,13H

                                                                                %    ,1301 0.5424 Ol2itt  141    ei W13l    l     01148 I7Abp.       0.0"                                    1.1331 if a j0i      tin  13281 4.43W  1.1129 ltill   1.1M 1  .1   1.3110~l 1.141

L I&--" mllýw Lý, 11, L---.ý

• 1 Li ~ ~* I. i

                                                                            -        -

I -j

                                                                                                     -            -           -

A. k

                                                                                                                                                       ~4             -

A.

                                                                                                                                                                           - ~ . I S andbed Bay 17 Location D 0

U Iy TIME Based on Calculation C-1302-87-4UO-021 Soe Be" st .. Date Avorage Since 1992 OriginlNminal hlkuam ~Minknm U~fcwm )Rdqatra TI&Imma 0510119B2 0*8222 1.154" Dateas Deb-8G Psb-6 Apr47T May487 Aug47 Sep-7 Jul48l Oct-88 ain-898eop-89Feb-SOApr40i Mar491FMai,4l Nov~-2144y-2 Sep-9Z fSepl4 $646 17D) 0.92217 1.8907 0.89069 1D.6952*0.8779 0.8622 0.8501111471 0.3358 0.82g 0,82S3 0.8291 01.222 OA6230.8172 Oil 0.641 p

rI r_ r- W- IV r ~r rr ... r Sandbed Bay 17-19

             .1 AI CD 0..95 0.9 r 0.85-0.8 -

C, q, 5~b I

                                                                                               ý V                                                90 4QP                                      9 TIME Sesod on Cslautoto C-1302-181-530M~

Slope Slope Best Eat. Low Oat. Ovis. H7P Avotage #Inc 1992 Awirag Sinc. i"2 Origsi Nmnal Ttnhi~kbvh vit abmma Unform Itetvvolikidbckmwo -0,087 4,01010 019211 V5IO1I92 0.9871 0.96" 1,154-Dat" Ne-90 lF.b-57 Apr-87 Mayr47 Aug-81P Sep-87 JuI.S8 Oct.$# ,tim.*9 S909 Fob4O Apr-"O Mar-91 May-01 NOV-91May402 "I~f UP-34 SiO6. 0,9417 1.0191 1,W68 0ý38*S 0.98 0.974 0,9692 0.9542 17110 1.0038 6*968 .1552 1.01 1.00V o0187 Oa..U O711t 0.92 .5.90187 O&7M 09.114

                          -Jr--                                                                                                                                -r-u -u Bbted do C;ilcultion C-.1302487-3094fl slope    Best Est.         Nht               Avensge Sin"  199              OdginalNomins Thicknless               MinimumUo -mRqh'dTlcnl MV0192          Of.8071                       1.154"

,Datem Dec*6 Feb.87 Apr47 May47 Aug487 SSepT7 Jul4 Oct488 Jun-89 Sep-49 Feb,0:A*r40 Mar41 May-91 Nov-I1 May-42 Sep2. 86p"4 0e0046 19A 0,8834 0.87293 0,85 0.85829 0.8488 0,JW 9M0.8288 0.8254 0.8399 0.8070 068167 018028 0.8032 .0.8091 0.8002 0.806 0.815

No..- Now..~ L. k- ___b Sandbed Bay 19 Location B 0 0.75 TIMIE Ba~sed anCluainC-3217504 Slope Best Fst. D~ate Average Sin'ce 1992 Original Nominal Tblckne~s M~In~iinwi n~t~infReq~aittd lThikness 0.0099 01330 OeSJ@I192 O.E337' Dts Dec-6 Feb#7' #Apr-7 Nigy47 Aug-97 Sep4S7 Jusf- Oci-*4 Jun489 Sep-0 Feb.tO Ap- Mar491 MIay91 ?frv-91 May-92 &Sp-01 Svp-04 sc"69 19B kW63mn 0.89221 0i76 *.M4 *OSS65 0,1256 044549 0.812 W0.8 0.95; 0*44" 4.463 0.9472 0.8j% 0.9,24 0.617

v r IP- -.-- ~----------~------ -V U ~ ~ -3 N w U ~U U - F _----- -~ u -r Sandbed Bay 19 Location C (I~ U

                'Based on Caclto       46-1750-2 Soe Best Est.       Date                  Average Since 1992             Oulgzingi Nornilnvl Thickness         'Mialln~is Upnfu Retuiratd hiclume
                         -005 0.8117           O5101192              0.820                        1,154" Dates        Dec486 Feb.87      Apr487 MAay47    ~Aug87'     Sp 4
                                                                                   ? Jucl-88 Oc$8 Jun-49 Sep-,9 170-            Apr40, Mar-t1     May-41  Nov4,1  May-92    Sep-2 SeI-94 Sep96 19C                                     0490651, 0.88816 0.88831 0.0735 0.8563      0.845   0.8447      0.8305 0.8261   0.8428    0.8232 0.8223   0.8319   0.8192     0.82 0.848 0

JNRC Information RequestFom Item No Date Received: Source AMP-141 10/612005 AMP Audit Topic: Status: Open IWE Document

References:

B.1.27 NRCReprestative Morante, Rich AmerGen (Took Issue): Hufnagel, Joh ouestion AMP B.1.27 IWE

a. Visual inspection of the coatings In the former sandbed region of the drywell Is currently conducted under the applicant's protective coatings monitoring and maintenance program; only this AMP is credited for managing loss of material due to corrosion for license renewal. Visual Inspection of the containment shell conducted in accordance with the requirements of IWE is typically credited to manage loss of material due to corrosion.

The applicant is requested to provide its technical basis for not also crediting Its IWE program for managing Icss of material due to corrosion in the former sandbed region of the drywell. B. During discussions with the applicant's staff on 10/04105 about augmented inspection conducted under IWE, the applicant presented tabulated Inspection results obtained from the mid 1980s to the present, to rmonitor the remaining drywell wall thickness In the cylindrical and spherical regions where significant corrosion of the outside surface was previously detected. The applicant Is requested to provide (1) a copy of these tabulated Inspection results, (2) a list of the nominal design thicknesses In each region of the drywell, (3) a list of the minimum required thicknesses In each region of the drywell, and (4) a list of the projected remaining wall thicknesses In each region of the drywell In the year 2029. AMP B.1.27 IWE Question on Remaining Wall Thickness In the Former Sandbed Region of the Drywell

c. During discussions with the applicant's staff on 10/05/05, the applicant described the history and resolution of corrosion In the sandbed region. After discovery, thickness measurements were taken from 1986 tirough 1992, to monitor the progression of viall loss. Remedial actions were completed In early 1993. At that time, the remaining wall thickness exceeded the minimum required thickness. The applicant concluded that it had completely corrected the conditions which led to the corrosion, and terminated Its program to monitor the remaining wall thickness. At that time, the remaining years of operation was expected to be no more than 16 years (end of the current license term).

INRC Information Request Formn The applicant's aging management commitment for license renewals is limited to periodic inspection of the coating that was applied to the exterior surface of the drywell as part of the remedial actions. The applicant has not made a license renewal commitment to measure wall thickness In the sandbed region In order to confirm the effectiveness of the remedial actions taken. Assigned To: Ouaou, Ahmed

Response

a) Visual Inspection of the containment drywell shell, conducted Inaccordance with ASME Section XI, Subsection IWE, Is credited for aging management of accessible areas of the containment drywell shell. Typically this Inspection Is for Internal surfaces of the drywell. The exterior surfaces of the drywell shell In the sand bed region for Mark I containment Is considered inaccessible by ASME Section XI, Subsection IWE, thus visual Inspection is not possible for a typical Mark I containment Including Oyster Creek before the sand was removed from the sand bed region in 1992. After removal of lhe sand, an epoxy coating was applied to the exterior surfaces of the drywell shell In the sand bed region. The region was made accessible during refueling outages for periodic Inspection of the coating. Subsequently Oyster Creek performed periodic visual Inspection of the coating in accordance with an NRC current licensing basis commitment. This commitment was Implemented prior to implementation of ASME Section XI, Subsection IWE. As a result inspection of the coating was conducted In accordance with the Protective Coating Monitoring and Maintenance Program. Our evaluation of this aging management program concluded the program is adequate to manage aging of the drywell shell in the sand bed region during the period of extended operation consistent with the current licensing basis commitment, and that inclusion of the coating inspection under IWE Is not required. However we are amending this position and will commit to monitor the protective coating in the exterior surfaces of the drywell in the sand bed region Inaccordance with the requirements of ASME Section XI, Subsection IWE during the period of extended operation. For details related to implementation of this commitment, refer to the response to NRC AMP Question #188. b) A tabulation of ultrasonic testing (UT) thickness measurement results in monitored areas of the drywell spherical region above the sand bed region and in the cylindrical region is Included in ASME Section Xl,.Subsection IWE Program Basis Document (PBD-AMP-B.1.27) Notebook. The tabulation contains information requested by the Staff and Is available for review during AMP audit. The tabulation is also provided in Table -1, and Table-2 below. c) In December 1992, with approval from the NRC a protective epoxy coating was applied to the outside surface of the drywell shell Inthe sand bed region to prevent additional corrosion in that area. UT thickness measurements taken In 1992, and In 1994, in the sand bed region from inside the drywell confirmed that the corrosion in the sand bed region has been arrested. Periodic inspection of the coating Indicates that the coating in that region is performing satisfactorily with no signs of deterioration such as blisters, flakes, or discoloration, etc. Additional UT measurements, taken in 1996 from Inside the drywell in the sand bed region showed no ongoing corrosion and provided objective evidence that corrosion has been arrested.

INRC Information Request Form As a result of these UT measurements and the observed condition of the coating, we concluded that corrosion has been arrested and monitoring of the protective coating alone, without additional UT measurements, will adequately manage loss of material Inthe drywell shell in the sand bed region. However to provide additional assurance that the protective coating Is providing adequate protection to ensure drywell Integrity, Oyster Creek will perform periodic confirmatory UT inspections of the drywell shell Inthe sand bed region. The Initial UT measurements will be taken prior to entering the period of extended operation and then every 10 years thereafter. The UT measurements will be taken from Inside the drywell at the same locations where the UT rMeasurements were taken In 1996. This revises the license renewal commitment communicated to the NRC In a letter from C. N. Swenson Site Vice President, Oyster Creek Generating Station to U. S. Nuclear Regulatory Commission, "Additional Commitments Associated with Application for renewed Operating Ucense - Oyster Creek Generating Station", dated 12/9/2005. This letter commits to one-time Inspection to be conducted prior to entering the period of extended operation. The revised commitment will be to conduct UT measurements on a frequency of 10 years, with the first Inspection to occur prior to entering the period of extended operation. This response was revised to incorporate additional commitments on UT examinations for the sand bed region discussed with NRC Audit team on 1/26/2006. This response was revised to reference response to NRC Question #AMP-188 and RAI 4.7.2-1(d). AMO 4/1/2006. The response was revised to add Table-I, and Table-2, and delete reference to RAI 4.7.2-1(d) AMO 4/512006. LRCR #: 229 LRA A.5 Commitazent #: 27 IR#: Approvals: PreparedBy.: Ouaou, Ahmed 4/5/2006 ReviewedBy., Getz, Stu 4/5/2006 ApprovedBy.: Warfel, Don 41512006 NRCAcceptencc (Date):

ii K r r7 E r7-r r77r --r -7~ r r r r r r r ii Table-I. UT Thickness measurements ror the Upper Region of the Drywell Shell

                                                    -   Average   Measured Thickness ,,zS Inches Monitored   Location  Minimum                         A Elevation           Required,,                                                                                       Projected Lower ThIckness,   1987 1988 7 1989 8

1990 1991 1 1992 1 1993 1994 1 1996 2000 ... 2004 9P19 Confidence Thickness In 2029 Thinches$ 7 Elevation 0.541" 50' 2" Bay 5- 0.743 0.742 0.747 0.741 0.748 0.74,1 0.743 No Ongoing D12 0.745 0.745 0.747 Corrosion S - 0.746 0.748 Bay 5- 5H 0.761 0.755 0.758 0.754 0.757 0.754 0.756 0.7384 0.761 0.758 0.758

                                                              - 0.760                                       -

Bay 5- 5L 0.706 0.703 0.703 0.702 0.705 0.706 0.701 No Ongoing 0.703 0.705 0.707 Corrosion 0.706 Bay 13- 0.762 0.760 0.765 0.759 0.766 0.762 0.758 No Ongoing 31H 0.779 0.758 0.763 Corrosion 0.765 Bay 13- 0.687 0.689 0.685 0.683 0.690 0.682 0.693 No Ongoing 31L 0.684 0.678 0.688 Corrosion 0.688 - Bay 15- 0.758 0.762 0.767 0.758 0.760 0.758 0.757 0.738 23H 0.764 0.762 0.763 0.765 Bay 15- 0.726 0.726 0.726 0.728 0.724 0.729 0.727 23L No Ongoing 0.728 0.729 0.724 Corrosion 0.725 Elevation 0.541" 51' 10"

W-- r F r U-ýz rp-- r r r7.1-- - V7 7ý --r -7r r- U F r Table-1. UT Thickness measurements for the Upper Region of the Drywell Shell

                                                                        -Average Measnrcdl    Thickness "*', Inches Monitored     Location     Minimum Elevation                                                                                                                                    Projected Lower Required         -.
                                                                                                                                    -        -    95% Confidence Thickness,      1987      1988     1989      1990    1991     1992    1993       1994     1996      2000     2004   Thickness In 2029 Inches -

Bay 13- 0.716 0.715 0.717 0.714 0.715 0.715 0.713 No Ongoing 32H 0.715 0.717 Corrosion

                                                                                .0.719 Bay 13-                                                 0.686    0.683   0.683   1           0.680    0.684     0.679   0.687  No Ongoing 32L                                                              0.683   0.676                                                 Corrosion 0.682                -,_

Elevation 0.518," 60'10" 0.693 10.11 0.692 0.689 0.689 No Ongoing 122 1 1 Corrosion Elevation 0.452" 87'5" Bay 9-20 0.619 0.622 0.619 0.620 0.614 0.629 0.613 0.613 0.604 0.612 0.604. 0.620 0.612 0.614 Bay 13- 0.643 0.641 0.645 0.643 0.635 0.641 0.640 0.636 0.635 I 0.640 No Ongoing 28 0.642 - 0.629 0.637 Corrosion Bay 15- 0.638 0.636 0.638 0.642 0.628 0.631 0.633 0.632 0.628 10.630 0.615 31 10.636 1 0.627 0.630 Notes: I. The average thickness is based on 49 Ultrasonic Testing (UT) measurements performed at each location

2. Multiple inspections were performed in the years 1988, 1990, 1991, and 1992.

3. The 1993 elevation 60" 10" Bay 5-22 inspection was performed on January 6, 1993. All other locations were inspected in December 1992.

4. Accuracy of Ultrasonic Testing Equipment is plus or minus 0.010 inches.

5. Reference S-000243-002.

w- r F r_- r-77 r7- r r r rrr r r or_- r r Table -1. UT Thickness measurements for the Upper Region of the Drywell Shell Conclusion-Summary of Corrosion Rates of UT measurements taken through year 2004

• There is no ongoing corrosion at two elevations (51" 10" and 60' 10")
• Based on statistical analysis, one location at elevation 50' 2" is undergoing aminor corrosion rate of 0.0003 inches per year,
• Based on statistical analysis, two locations at elevation 87' 5"are undergoing minor corrosion rates of 0.0005 and 0.00075 inches per year

r .. r Ir

                                  **r         7             7                  r                         r          r-.         r.         r-         r,              r         r          V       ir Table -2 UT Thickness measurements for the Sand Bed Region of the Drywell Shell Location Sub         Dec    Feb       Apr   May       Aug     Sep      Jul      Oct      Jun        Sep      Feb      Apr ay                                                                                                                               Mar       May     Nov       May       ep        Sep     Sep ocation 1986    1987     1987   1987     1987     1987    1988      1988     1989       1989     1990     1990       1991       1991    1991      1992      992       1994    1996 ID                                                                                  1.11.x                                                                                            1.101  1.151A D                                                   .....-.-                      t                                                                           -                  1.154   1.181 D                                                                                 1.17                                                                        ....                 1.16q    1.1 D                                                                                 1.131    -       .--.-                            -.                                              1.135   1.13 A                                                                                 1.15.

D 1.07

                                                                                            -

1.12 1.15 1.021 1.0.4 1.0 1.021 1.02, 0.99" 1.00 0 1.0c 1.004 0.994 1.00 SIA 0.911 0.9y. 0.r 0.90! 0.912 0.88- 0.881 0.8 0.881 0.87c 0.844 0.84ý 0.83 0.844 o.82 0.8M 0.83 IC Bottom 0.91 0.9M 0.916 0.901 0.891 0.8 0.891 0.87C 0.86! 0.851 0.868 0.85 0.884 0.85, 0.851 0.88 Top 1.04( 1.10 1.07E 1.045 1.00f 1.016 1.005 0.952 0.97A 0.984 1.014 0.9 1.01q 0.971 0.981 1.04 13A 0.919 -.... 0.90! 0.88 00.88 0.86 2 0.58 0.851 0.85? 0.84 0.868 0.851 0.828 0.843 13C ottom - --

                                                                                                        -        -       -           0--

0.90 0901 0. 0.931 0.0 0.894 0.933

          -  op 1.07      1.0     1.041       1.08a    1.051    1.037   1.054 13D                                                                                 0.96    8                              0.93-                                             1.001-  0.95     0.0 ISA                                                                                 1.121          1             -

1.11 1.127 1SD 1.08E 1.05- t06" 1.061 1.05 1.05 1.06C.- 1.0 41 1.068 1.054 1.051 1.068 17A Bottom 0.99C 0.95 0.986 0.95! 0.95ý 0.951 0.93! 0.944 0.93 0.948 0.941 0.93M 0.99A Top 0.99S 1.131 1.13C 1.131 1.121 1.124 1.131 1.121 21.121.124 1.121 1.12 1.144 17D 0.922 0.89! 0.891 0.89.9 0.871 0.861 0.851 0.841 0.83E 0.829 0.82- 0.821 0.8 2 0.823 0.811 0.811 0.841 17/19 ýop 0.98 1.011 1.131 0.99C 0.980 0.974

            ýottom                                                                                                                             0.961   0.951       0.972   0.97     0.96A    0.96A 1.004   0.99         0.951    1.01C    1.001      0.98A     0.98"   0.971       0.994   0.98     0.97!    0.991 19A                            0.884           0.87.. 0.85   0.85     0.8.4     0. 3    0.8a         0.821   0.84C     0.80     0 0.81        08. 0.80.       0.801   0.80     0.80     0.81 19B               -.-                          0.89       0.8    0.88Z    0.864     0.85    0.82        0.841    0.81     0.837       0.851     0.84    0.841      0.84     0.84C    0.829    0.83 19C                                            0.901      0.888  0.884    0.873     0.854   0.84-       0.844    0.831     0.621      0.844     0.821   0.82A      0.83W    0.81     0.82     0.84

Citizen's Exhibit NC2 D. Ashley - Questions to go over tomorrow... Page I11 [I7*Ihley- Questios to go oer tomorrw... Pag 1i Citizen's Exhibit NC2 From: <john.hufnagel@exeloncorp.com> To: <djal @nrc.gov>, <rkm@nrc.gov> Date: 04)24/2006 6:17:58 PM

Subject:

Questions to go over tomorrow... Roy and Donnie, These attached questions are those from the database that we currently have statused as Open, but which have responses that should allow closure. Although in the closed status, AMP-071 and AMP-204 were also included because they were updated to reference additional information provided In AMP-072. Also, we did not send AMP-358, which Is the item on Fatigue Analysis. We plan on sending that to you tomorrow. Hope to talk with you tomorrow PM.

              - John.
               <<AMP-071.pdf>> <<AMP-072.pdf>> <<AMP-141.pdf>> <<AMP-204.pdf>> <<AMP-209.pdf>>
             <<AMP-210.pdf>> <<AMP-264.pdf>> <<AMP-356.pdf>> <<AMP-357.pdf>> <<AMP-359.pdf>>
             <<AMP-360.pdf>> <<AMP-361.pdf>> <<AMP-362.pdf>> <<AMR-164.pdf>> <<AMR-167.pdf>>
             <<AMR-355.pdf>>

This e-mail and any of Its attachments may contain Exelon Corporation proprietary information, which Is privileged, confidential, or subject to copyright belonging to the Exelon Corporation family of Companies. This e-mail Is Intended solely for the use of the Individual or entity to which It Is addressed. If you are not the Intended recipient of this e-mail, you are hereby notified that any dissemination, distribution, copying, or action taken In relation to the contents of and attachments to this e-mail Is strictly prohibited and may be unlawful. If you have received this e-mail In error, please notify the sender Immediately and permanently delete tie original and any copy of this e-mail and any printout. Thank You. CC: <donald.wartel@ exeloncorp.com>, .fred.polaski@ exebncorp.com> p(C4Py

INRCInformation Request rm lIe::: No DateReceived: Source AMP-357 2/16/2006 AMP Audit Topic: Status: Open IWE Document

References:

NRCRepresentative Morante, Rich AmerGen (Took Issue): Ouestion (1) When a new set of point thickness readings Is taken Is the former sandbed region, prior to entering the LR period, what will be the quantitative acceptance criteria for concluding that corrosion has or has not occurred since the last Inspection in 1996. (2) If additional corrosion is detected in the upcoming inspection, describe in detail the augmented Inspections and other steps that will be taken to evaluate the extent of the corrosion, and describe the approach to ensuring the continued structural adequacy of the containment. Assigned To: Ouaou, Ahmed

Response

(1).The new set of UT measurements for the former sand bed region will be analyzed using the same methodology used to analyze the 1992, 1994, and 1996 UT data. The results will then be compared to the 1992, 1994,1996 UT results to confirm the previous no corrosion trend. Because of surface roughness of the exterior of the drywell shell, experience has shown that UT measurements can vary significantly unless the UT Instrument Is positioned on the exact point as the previous measurements. Thus acceptance criteria will be based on the standard deviation of the previous data (+/-11 mils) and Instrument accuracy of (+1-10 mils) for a total of 21 mils. Deviation from this value will be considered unexpected and requires corrective actions described in item (2) below. (2). If additional corrosion Is identified that exceeds acceptance criteria described above, Oyster Creek will Initiate corrective actions that Include one or all of the following, depending on the extent of Identified corrosion.

a. Perform Edditional UT measurements to confirm the readings

b. Notify NRC within 48 hours of confirmation of the identified condition

c. Conduct inspection of the coatings in the sand bed region in areas where the additional corrosion was detected.

d. Perform engineering evaluation to assess the extent of the condition and to determine If additional inspections are required to assure drywell Integrity.

e. Perform operability determination and justification for continued operation until next scheduled

INRC Information Request Fr inspection. These actions will be completed before restarting from an outage LRCR #: 293 LRA A.5 Commiitment #: IR#: Approvals: PreparedBy.: Ouaou, Ahmed 4/1/2006 ReviewedBy: Muggleston, Kevin 4/3/2006 ApprovedBy.: Warfel, Don 413/2006 NRCAcceptwnce (Date):

INRC InformationRequest Form[ Item No Date Received: Source AMP-356 2/1612006 AMP Audit Topic: Statuss: Open IWE Document

References:

NRC Representative Morante, Rich AmerGen (Took Issue): Onettlon IWE AMP Question 4 IWE AMP Revised Feb. 17, 2006 R. Morante (AMP-356) (1) Identify the specific locations around the circumference in the former sandbed region where UT thickness readings have been and will be taken from Inside containment. Confirm that all points previously recorded will be included in future Inspections. (2) Describe the grid pattern at each location (meridional length, circumferential length, grid point spacing, total number of point readings), and graphically locate each grid pattern within the former sandbed region. (3) For each grid location, submit a graph of remaining thickness versus time, using the UT readings since the Initiation of the program (both prior to and following removal of the sand and application of the extemal coating). (4) Clearly describe the methodology and acceptance criteria that is applied to each grid of point thickness readings, including both global (entire array) evaluation and local (subregion of array) evaluation. Assigned To: Ouaou, Ahmed

Response

Response:

1. The circumference of the drywell is divided Into 10 bays, designated as Bays 1, 3, 5, 7, 9. 11,13, 15, 17, and 19. UT thickness readings have been taken in each bay at one or more locations. The specific locations around the circumference In the former sand bed region where UT thickness reading have been taken from Inside containment are Bay 1D, 3D, 5D, 7D, 9A, 9D, I1A, 1 C, 13A.,

13C, 13D, 15A, 15D, 17A, 17D, 17/19 Frame, 19A, 19B, and 19C. For each location, UT measurements were taken centered at elevation 1 V-3". These represent the locations where UT measurements were taken in 1992, 1994, and 1996.

INRC Information Request Form l In addition UT measurements were taken one time Inside 2 trenches excavated in drywell floor concrete. The purpose of these UT measurements Is to determine the extent of corrosion In the lower portions of the sand bed region prior to removing the sand and making accessible for visual inspection. Future UT thickness measurements will be taken at the same locations as those inspected in 1996 in accordance with Oyster Creek commitment documented in NRC Question #AMP-209.

2. For locations where the initial Investigations found significant wall thinning (9D, I1A, IIC, 13A, 13D, 15D, 17A, 17D, 17/19 Frame, 19A, 19B, and 19C) the grid pattern consists of 7 x 7 grid centered at elevation 11'-3 (meridian) and centered at the centerline of the tested location within each bay, which consists of 6"x 6" square template. The grid spacing is 1" on center. There are 49 point readings. For graphical location of the grid, refer to attachment 1.

For locations where the initial investigations found no significant wall thinning (ID, 3D, 5D, 7D, 9A. 13C, and 15A) the grid pattern consists of I x 7 grid centered at elevation 1 1'-3" (meridian) on 1 centers. There are 7 point readings. For graphical location of the grid, refer to attachment 1.

3. A graph representing the remaining thickness versus time using UT reading since the initiation of the program (both prior to and following removal of the sand and application of the external coating) for location 9D, I1A, 11C, 13A, 13D,15D,17A,17D,17/19, 19A, 19B, and 19C is Included in the attached graph. Other locations (i.e. ID, 3D, 5D, 7D, 9A, 13C, and 15A) are not Included because wall thinning Is not significant and the trend line will be essentially a straight line.

4. The methodology and acceptance criteria that Is applied to each grid of point thickness readings, Including both global (entire array) evaluation and local (subregion of array) Is described in engineering specification IS-328227-004 and in calculation No. C-1302-187-5300-011. These documents were submitted to the NRC in a letter dated November 26, 1990 and provided to the Staff during the AMPIAMR audit. A brief summary of the methodology and acceptance criteria is described below.

The initial Io=ations where corrosion loss was most severe in 1986 and 1987 were selected for rep-eat inspection over time to measure corrosion rate. For location.where the Initial Investigations found significant wall thinning UT inspection consists of 49 Individual UT data points equally spaced over a 6"x 6" area. Each new set of 49 values was then tested for normal distribution. The mean values of each grid were then compared to the required minimum uniform thickness criteria of 0.736. In addition each Individual reading is compared to the local minimum required criteria of 0.49. The basis for the required minimum uniform thickness criteria and the local minimum required criteria is provided In response to NRC Question #AMP-210. A decrease in the mean value over time Is representative of corrosion. If corrosion does not exisl, the mean value will not vary with time except for random variations In the UT measurements.

N'RC Information Request Form If corrosion is continuing, the mean thickness will decrease linearly with time. Therefore the curve fit of the data is tested to determine if linear regression is appropriate, Inwhich case the corrosion rate Is equal to the slope of the line. Ifa slope exists, then upper and lower 95% confidence Intervals of the curve fit are calculated. The lower 95% confidence Interval is then projected Into the future and compared to the required minimum uniform thickness criteria of 0.736. A similar process is applied to the thinnest Individual reading In each grid. The curve fit of the data is tested to determine Iflinear regression Is appropriate. If a slope exists, then the lower 95% confidence interval Is then projected Into the future and compared to the required minimum local thickness criteria of .49. LRCR #: LRA A.5 Commitment #: IR#: Approvals: PreparedBy: Ouaou, Ahmed 414/2006 ReriewedBy: Getz, Stu 4/5/2006 ApprovedBy,: Warfel, Don 415/2006 NRCAcceptance (Date): 6

INRCInforinationRequest Form Item No DateReceived: Source AMP-210 1/2412006 AMP Audit Topic: Status: Open IWE Document

References:

B.1.27 NRCRepresentative Morante, Rich AmerGen (Took Issue): Hufnagel, Joh OuestTon Pages 25 through 31 of the PBD present a discussion of the OCGS operating experience. (8a)The following statements related to drywell corrosion In the sand bed region need further explanation and clarification: As a result of the presence of water in the sand bed region, extensive UT thickness measurements (about 1000) of the drywell shell were taken to determine if degradation was occurring. These measurements corresponded to known water leaks and Indicated that wall thinning had occurred in this region. Please explain the underlined statement. Were water leaks limited to only a portion of the circumference? Was wall thinning found only In these areas? After sand removal, the concrete surface below the sand was found to be unfinished with Improper provisions for water drainage. Corrective actions taken in this region during 1992 included; (1) cleaning of b*ose rust from the drywell shell, followed by application of epoxy coating and (2) removing the loose debris from the concrete floor followed by rebuilding and reshaping the floor with epoxy to allow drainage of any water that may leak into the region. UT measurements taken from the outside after cleaning verified loss of material projections that had been made based on measurements taken from the inside of the dryweli. There were, however, some areas thinner than projected; but in all cases engineering analysis determined that the drywell shell thickness satisfied ASME code requirements. Please describe the concrete surface below the sand that Is discussed in paragraph above. Please provide the following Information: (1) Identify the minimum recorded thickness In the sand bed region from the outside Inspection, and the minimum recorded thickness In the sand bed region from the Inside Inspections. Is this consistent with previous Information provided verbally? (.806 minimum) (2) What was the projected thickness based on measurements taken from the inside? (3) Describe the engineering analysis that determined satisfaction of ASME code requirements and Identify the minimum required thickness value. Is this consistent with previous information provided verbally? (.733 minimum) (4) Is the minimum required thickness based on stress or buckling criteria? (5) Reconcile and compare the thickness measurements provided in (1) and (3) above with the .736 minimum corroded thickness that was used in the NUREG-1 540 analysis of the degraded Oyster

                                                                                                       -7

INRC Information Request Form Creek sand bed region. Evaluation of UT measurements taken from inside the drywell, in the in the former sand bed region, in 1992, 1994, and 1996 confirmed that corrosion is mitigated. it Is therefore concluded that corrosion In the sand bed region has been arrested and no further loss of material is expected. Monitoring of the coating In accordance with the Protective Coating Monitoring and Maintenance Program, will continue to ensure that the containment drywell shell maintains its Intended function during the period of extended operation. NUREG-1540, published in April 1996, includes the following statements related to corrosion of the Oyster Creek sand bed region: (page vii) However, to assure that these measures are effective, the licensee is required to perform periodic UT measurements. and (page 2) As assurance that the corrosion rate is slower than the rate obtained from previous measurements, GPU is committed to make UT measurements periodically. Please reconcile the aging management commitment (one-time UT Inspection and monitoring of the condition of the coating) with the apparent requirement/commitment documented in NUREG-1 540. (8b)The following statement related to drywell corrosion above the sand bed region needs further explanation and clarification: Corrective action for these regions Involved providing a corrosion allowance by demonstrating, through analysis, that the original drywell design pressure was conservative. Amendment 165 to the Oyster Creek Technical Specifications reduced the drywell design pressure from 62 psig to 44 psig. The new design pressure coupled with measures to prevent water Intrusion Into the gap between -he drywell shell and the concrete will allow the upper portion of the drywell to meet ASME code requirements. Please describe the measures to prevent water Intrusion into the gap between the drywell shell and the concrete that will allow the upper portion of the drywell to meet ASME code requirements". Are these measures to prevent water intrusion credited for LR? If not, how will ASME code requirements be met during the extended period of operation? (8c)The following statements related to torus degradation need further explanation and clarification: Inspection performed in 2002 found the coating to be in good condition In the vapor area of the Torus and vent header, and In fair condition In Immersion. Coating deficiencies In Immersion include blistering, random and mechanical damage. Blistering occurs primarily In the shell Invert but was also noted on the upper shell near the water line. The fractured blisters were repaired to reestablish the protective coating barrier. This Is another example of objective evidence that the Oyster Creek ASME Section XI, Subsection IWE aging management program can Identify degradation and Implement corrective actions to prevent the loss of the containment's Intended function. While blistering is considered a deficiency, It Is significant only when It is fractured and exposes the base metal to corrosion attack. The majority of the blisters remain Intact and continues to protect the base metal; consequently the corrosion rates are low. Qualitative assessment of the Identified pits indicate that the measured pit depths (50 mils max) are significantly less than the criteria established

INRC Information Requlest ro-I In Specification SP-1302-52-120 (141- 261 mils, depending on diameter of the pit and spacing between pits). Please confirm or clarify (1) that only the fractured blisters found in this inspection were repaired; (2) pits were Identified where the blisters were fractured; (3) pit depths were measured and found to 50 mils max; (4) the inspection Specification SP-1302-52-120 Includes pit-depth acceptance criteria for rapid evaluation of observed pitting; (5) the minimum pit depth of concern Is 141 mils (.141) and pits as deep as 261 mils (.261) may be acceptable. Please also provide the following Information: nominal design, as-built, and minimum measured thickness of the torus; minimum thickness required to meet ASME code acceptance criteria; the technical basis for the pitting acceptance criteria Include in Specification SP-1 302-52-120 Assigned To: Ouaou, Ahmed

Response

(8a) Question: Please explain the underlined statement. Were water leaks limited to only a portion of the circumference? Was wall thinning only in these area?

Response

This statement was not meant to Indicate that water leaks were limited to only a portion of the circumference. The statement is meant to reflect the fact that water leakage was observed coming out of certain sand bed region drains and those locations were suspect of wall thinning. No. Wall thinning was not limited to the areas where water leakage from the drains was observed. Wall thinning occurred in all areas of the sand bed region based on UT measurements and visual Inspection of the area conducted after the sand was removed in 1992. However the degree of wall thinning varied from location to location. For example 60% of the measured locations in the sand bed region (bays 1, 3, 5, 7, 9, and 15) Indicate that the average measured drywell shell thickness Is nearly the same as the design nominal thickness and that these locations experienced negligible wall thinning; whereas bay 19A experienced approximately 30% reduction in wall thickness. Question: Please discuss the concrete surface below the sand that Is discussed In paragraph above.

Response

The concrete surface below the sand was Intended to be shaped to promote flow toward each of the five sand bed drains. However once the sand was removed it was discovered that the floor was not properly finished and shaped as required to permit proper drainage. There were low points, craters, and rough surfaces that could allow moisture to pool Instead of flowing smoothly toward the drains. These concrete surfaces were refurbished to fill low areas, smooth rough surfaces, and coat these surfaces with epoxy coating to promote improved drainage. The drywell shell at juncture of the concrete floor was sealed with an elastomer to prevent water intrusion Into the embedded drywell shell. Question: Please provide the following Information: 1

INRC InformationRequest For, I (1) Identify the minimum recorded thickness In the sand bed region from the outside Inspection, and the minimum recorded thickness in the sand bed region from the inside inspections. Is this consistent with previous information provided verbally? (.806 minimum) (2) What was the projected thickness based on measurements taken from the inside? (3) Describe the engineering analysis that determined satisfaction of ASME code requirements and Identify the minimum required thickness value. Is this consistent with previous information provided verbally? (.733 minimum) (4) Is the minimum required thickness based on stress or buckling criteria? (5) Reconcile and compare the thickness measurements provided in (1) and (3) above with the .736 minimum corroded thickness that was used In the NUREG-1540 analysis of the degraded Oyster Creek sand bed region.

Response

1. The minimum recorded thickness in the sand bed region from outside inspection Is 0.618 inches.

The minimum recorded thickness In the sand bed region from inside inspections is 0.603. These minimum recorded thicknesses are Isolated local measurement and represent a single point UT measurement. The 0.806 Inches thickness provided to the Staff verbally is an average minimum general thickness calculated based on 49 UT measurements taken In an area that is approximately 6"x 6". Thus the two local Isolated minimum recorded thicknesses cannot be compared directly to the general thickness of 0.806". The 0.806" minimum average thickness verbally discussed with the Staff during the AMP audit was recorded in location 19A in 1994. Additional reviews after the audit noted that lower minimum average thickness values were recorded at the same location in 1991 (0.803") and in September 1992 (0.800"). However, the three values are within the tolerance of +1-0.010" discussed with the Staff.

2. The minimum projected thickness depends on whether the trended data is before or after 1992 as demonstrated by corrosion trends provided In response to NRC Question #AMP-356. For license renewal, using corrosion rate trends after 1992 is appropriate because of corrosion mitigating measures such as removal of the sand and coating of the shell. Then, using corrosion rate trends based on 1992, 1994, and 1996 UT data; and the minimum average thickness measured in 1992 (0.800"), the minimum projected average thickness through 2009 and beyond remains approximalely 0.800 Inches. The projected minimum thickness during and through the period of extended operation will be reevaluated after UT Inspections that will be conducted prior to entering the period of extended operation, and after the periodic UT Inspection every 10 years thereafter.

3.The engineering analysis that demonstrated compliance to ASME code requirements was performed In two parts, Stress and Stability Analysis with Sand, and Stress and Stability Analyses without Sand. The analyses are documented in GE Reports Index No. 9-1, 9-2, 9-3, and 9-4, were transmitted to the NRC Staff In December 1990 and In 1991 respectively. Index No. 9-3 and 9-4, were revised later to correct errors Identified during an internal audit and were resubmitted to the Staff In January 1992 (see attachment I & 2). The analyses are briefly described below. The drywell shell thickness In the sand bed region is based on Stability Analysis without Sand. As I0

NRC Information Request FormI described in detail in attachment I & 2, the analysis Is based on a 36-degree section model that takes advantage of symmetry of the drywell with 10 vents. The model includes the drywell shell from the base of the sand bed region to the top of elliptical head and the vent and vent header. The torus is not Included in this model because the bellows provide a very flexible connection, which does not allow significant structural interaction between the drywell and the tows. The analysis conservatively assumed that the shell thickness In the entire sand bed region has been reduced uniformly to a thickness of 0.736 Inches. As discussed with the Staff during the AMP audit, the basic approach used in the buckling evaluation follows the methodology outlined in ASME Code Case N-284 revision 0 that was reconciled later with revision I of the Code Case. Following the procedure of this Code Case, the allowable compressive stress is evaluated In three steps. In the first step, a theoretical buckling stress is determined, and secondly modified using appropriate capacity and plasticity reduction factors. In the final step, the allowable compressive stress is obtained by dividing the buckling stress calculated in the second step by a safety factor of 2.0 for Design and Level A & B service conditions and 1.67 Level C service conditions. Using the approach described above, the analysis shows that for the most severe design basis load combinations, the limits of ASME Section III, Subsection NE 3213.10 are fully met. For additional details refer to Attachment 1 & 2. As described above, the buckling analysis was performed assuming a uniform general thickness of the sand bed region of 0.736 Inches. However the UT measurements identified isolated, localized areas where the drywell shell thickness is less than 0.736 Inches. Acceptance for these areas was based on engineering calculation C-1302-187-5320-024. The calculation uses a Local Wall Acceptance Criteria'. This criterion can be applied to small areas (less than 12N by 12"), which are less than 0.736" thick so long as the small 12" by 12" area is at least 0.536" thick. However the calculation does not provide additional criteria as to the acceptable distance between multiple small areas. For example, the minimum required linear distances between a 12" by 12' area thinner than 0.736" but thicker than 0.536' and another 12" by 12" area thinner than 0.736" but thicker than 0.536" were not provided. The actual data for two bays (13 and 1) shows that there are more than one 12" by 12" areas thinner than 0.736" but thicker than 0.536". Also the actual data for two bays shows that there are more than one 2 Y'/"diameter areas thinner than 0.736" but thicker than 0.490". Acceptance Is based on the following evaluation. The effect of these very local wall thickness areas on the buckling of the shell requires some discussion of the buckling mechanism in a shell of revolution under an applied axial and lateral pressure load. To begin the discussion we will describe the buckling of a simply supported cylindrical shell under the influence of lateral pressure and axial load. As described In chapter 11 of the Theory of Elastic Stability, Second Edition, by Timoshenko and Gere, thin cylindrical shells buckle In lobes in both the

NRC Information Request Form. axial and circumferential directions. These lobes are defined as half wave lengths of sinusoidal functions. The functions are governed by the radius, thickness and length of the cylinder. If we look at a specific thin walled cylindrical shell both the length and radius would be essentially constants and if the thickness was changed locally the change would have to be significant and continuous over a majority of the lobe so that the compressive stress in the lobe would exceed the critical buckling stress under the applied loads, thereby causing the shell to buckle locally. This approach can be easily extrapolated to any shell of revolution that would experience both an axial load and lateral pressure as in the case of the drywell. This local lobe buckling is demonstrated In The GE Letter Report aSandbed Local Thinning and Raising the Fixity Height Analysis' where a 12 x 12 square Inch section of the drywell sand bed regloh is reduced by 200 mils and a local buckle occurred In the finite element elgenvalue extraction analysis of the drywell. Therefore, to Influence the budding of a shell the very local areas of reduced thickness would have to be contiguous and of the same thickness. This is also consistent with Code Case 284 in Section -1700 which indicates that the average stress values in the shell should be used for calculating the buckling stress. Therefore, an acceptable distance between areas of reduced thickness Is not required for an acceptable buckling analysis except that the area of reduced thickness is small enough not to influence a buckling lobe of the shell. The very local areas of thickness are dispersed over a wide area with varying thickness and as such will have a negligible effect on the buckling response of the drywell. In addition, these very local wall areas are centered about the vents, which significantly stiffen the shell. This stiffening effect limits the shell buckling to a point In the shell sand bed region which is located at the midpoint between two vents. The acceptance criteria for the thickness of 0.49 Inches Confined to an area less than 2Y2 inches in diameter experiencing primary membrane + bending stresses Is based on ASME B&PV Code, Section III, Subsection NE, Class MC Components, Paragraphs NE-3213.2 Gross Structural Discontinuity, NE-3213.10 Local Primary Membrane Stress, NE-3332.1 Openings not Requiring Reinforcement, NE-3332.2 Required Area of Reinforcement and NE-3335.1 Reinforcement of Multiple Openings. The use of Paragraph NE-3332.1 Is limited by the requirements of Paragraphs NE-3213.2 and NE-3213.10. In particular NE-3213.10 limits the meridional distance between openings without reinforcement to 2.5 x (square root of Rt) . Also Paragraph NE-3335.1 only applies to openings In shells that are closer than two times their average diameter. The Implications of these paragraphs are that shell failures at these locations from primary stresses produced by pressure cannot occur provided openings In shells have sufficient reinforcement. The current design pressure of 44 psig for drywell requires a thickness of 0.479 inches In the sand bed region of the dryweil. A review of all the UT data presented In Appendix D of the calculation indicates that all thicknesses In the drywell sand bed region exceed the required pressure thickness by a substantial margin. Therefore, the requirements for pressure reinforcement specified In the previous paragraph are not required for the very local wall thickness evaluation presented in Revision 0 of Calculation C-1302-187-5320-024. Reviewing the stability analyses provided In both the GE Report 9-4 and the GE Letter Report Sand bed Local Thinning and Raising the Fixity Height Analysis and recognizing that the plate elements in the sand bed region of the model are 3"x 3"It Is dear that the circumferential buckling lobes for the

INRC InformationRequest FormI drywell are substantially larger than the 2 % inch diameter very local wall areas. This combined with the local reinforcement surrounding these local areas indicates that these areas will have no Impact on the buckling margins in the shell. It is also clear from the GE Letter Report that a uniform reduction in thickness of 27% to 0.536" over a one square foot area would only create a 9.5% reduction in the load factor and theoretical buckling stress for the whole drywell resulting in the largest reduction possible. In addition, to the reported result for the 27% reduction In wall thickness, a second buckling analysis was performed for a wall thickness reduction of 13.5% over a one square foot area which only reduced the load factor and theoretical buckling stress by 3.5% for the whole drywell resulting in the largest reduction possible. To bring these results into perspective a review of the NDE reports Indicate there are 20 UT measured areas In the whole sand bed region that have thicknesses less than the 0.736 Inch thickness used In GE Report 9-4 which cover a conservative total area of 0.6E square feet of the drywell surface with an average thickness of 0.703" or a 4.5% reduction In wall thickness. Therefore, to effectively change the buckling margins on the drywell shell In the sand bed region a reduced thickness would have to cover approximately one square foot of shell area at a location In the shell that is most susceptible to buckling with a reduction in thickness greater than 25%. This leads to the conclusion that the buckling of the shell is unaffected by the distance between the very local wall thicknesses, in fact these local areas could be contiguous provided their total area did not exceed one square foot and their average thickness was greater than the thickness analyzed In the GE Letter Report and provided the methodology of Code Case N284 was employed to determine the allowable buckling load for the drywell. Furthermore, all of these very local wall areas are centered about the vents, which significantly stiffen the shell. This stiffing effect limits the shell buckling to a point in the shell sand bed region, which is located at the midpoint between two vents. The minimum thickness of 0.733" is not correct. The correct minimum thickness is 0.736".

4. The minimum required thickness for the sand bed region is controlled by buckling.

5. We cannot reconcile the difference between the current (lowest measured) of 0.736" in NUREG-1540 and the minimum measured thickness of 0.806 inches we discussed with the Staff. Perhaps; the value in NUREG-1 540 should be labeled minimum required by the Code, as documented In several correspondences with the Staff, Instead of lowest measured. In a letter dated September 15, 1995, GPU provided the Staff a table that lists sand bed region thicknesses. The table Indicates that nominal thickness Is 1.154". the minimum measured thickness in 1994 Is 0.806", and the minimum thickness required by Code is 0.736". These thicknesses are consistentwith those discussed with the Staff during the AMP/AMR audit.

Question: NUREG-1540, published in April 1996, Includes the following statements related to corrosion of the Oyster Creek sand bed region: (page vii) However, to assure that these measures are effective, the licensee is required to perform periodic UT measurements, and (page 2) As assurance that the corrosion rate is slower than the rate obtained from previous measurements, GPU is committed to make UT measurements periodically. Please reconcile the aging management commitment (one-time UT Inspection and monitoring of the condition of the coating) with the apparent requirement/commitment documented In NUREG-1540.Please reconcile the aging management commitment (one-time UT Inspection and monitoring of the condition of the coating) with the apparent requirementlcommitment documented in NUREG-1540.

INRCnformation Request rm

Response

Our review of NUREG-1540, page 2 Indicates that the statements appear to be based on 1991, or 1993 GPU commitment to perform periodic UT measurements. In fact UT thickness measurements were taken in the sand bed region from Inside the drywell in 1992, and 1994. The trend of the UT measurements indicates that corrosion has been arrested. As results GPU Informed NRC in a letter dated September 15, 1995 (ref. 2) that UT measurements will be taken one more time, in 1996, and the epoxy coating will be Inspected in 1996 and, as a minimum again In 2000. The UT measurements were taken In 1996, per the commitment, and confirmed corrosion rate trend of 1992 and 1994. The results of 1992, 1994, and 1996 UT measurements were provided to the Staff during the AMP/AMR audits. In response to GPU September 15, 1995 letter, NRC Staff found the proposed changes to sand bed region commitments (i.e. no additional UT measurements after 1996) reasonable and acceptable. This response Is documented in November 1, 1995 Safety Evaluation for the Drywell Monitoring Program. For license renewal, Oyster Creek was previously committed to perform One-Time UT Inspection of the drywell shell in the sand bed region prior to entering the period of extended operation. However, In response to NRC Question #AMP-141, Oyster Creek revised the commitment to perform UT Inspections periodically. The initial inspection will be conducted prior to entering the period of extended operation and additional Inspections will be conducted every 10 years thereafter. The UT measurements will be taken from inside the drywell at same locations as 1996 UT campaign (Bb) Question: Please describe the measures to prevent water intrusion Into the gap between the drywell shell and the concrete that will allow the upper portion of the drywell to meet ASME code requirements. Are these measures to prevent water Intrusion credited for LR? If not, how will ASME code requirements be met during the extended period of operation?

Response

The measures taken to prevent water Intrusion Into the gap between the drywell shell and the concrete that will allow the upper portion of the drywell to maintain the ASME code requirements are,

1. Cleared the former sand bed region drains to Improve the drainage path.

2. Replaced reactor cavity steel trough drain gasket, which was found to be leaking.

3. Applied stainless steel type tape and strippable coating to the reactor cavity during refueling outages to Eeal Identified cracks In the stainless steel liner.

4. Confirmed that the reactor cavity concrete trough drains are not clogged

5. Monitored former sand bed region drains and reactor cavity concrete trough drains for leakage during refueling outages and plant operation.

Oyster Creek Is committed to Implement these measures during the period of extended operation. (8c) Please confirm or clarify (1) that only the fractured blisters found In this inspection were repaired; (2) pits were Identified where the blisters were fractured; (3) pit depths were measured and found to

                                                                                                           /4-

INRClInformnation Request Form l 50 mils max; (4) the Inspection Specification SP-1302-52-120 includes pit-depth acceptance criteria for rapid evaluation of observed pitting; (5) the minimum pit depth of concern is 141 mils (.141) and pits as deep as 261 mils (.261) may be acceptable.

Response

(1) Specification SP-1302-52-120, Specification for Inspection and Localized Repair of the Torus and Vent System Coating, specifies repair requirements for coating defects exposing substrate and fractured blisters showing signs of corrosion. The repairs referred to in the Inspection report Incduced fractured blisters, as well as any mechanically damaged areas, which have exposed bare metal showing sig.is of corrosion. Therefore, only fractured blisters would be candidates for repair, not those blisters that remain Intact. The number and location of repairs are tabulated in the final inspection report prepared by Underwater Construction Corporation. (2) Coating deficiencies in the Immersion region Included blistering with minor mechanical damage. Blistering occurred primarily in the shell Invert but was also noted on the upper shell near the water line. The majority of the blisters were Intact. Intact blisters were examined by removing the bliste-cap exposing the substrate. Corrosion attack under non-fractured blisters was minimal and was generally limited to surface discoloration. Examination of the substrate revealed slight discoloration and pitting vith pit depths less than 0.001. Several blistered areas included pitting corrosion where the blisters were fractured. The substrate beneath fractured blisters generally exhibited a slightly heavier magnetite oxide layer and minor pitting (less than 0.010") of the substrate. (3) In addition to blistering, random deficiencies that exposed base metal were identified in the torus immersion region coating (e.g., minor mechanical damage) during the 19R (2002) torus coating Inspections. They ranged in size from 1116" to W" in diameter. Pitting in these areas was qualitatively evaluated and ranged from less than 10 mils to slightly more than 40 mils in a few isolated cases. Three quantitative pit depth measurements were taken in several locations in the immersion area of Bay 1. Pit depths at these sites ranged from 0.008" to 0.042" and were judged to be representative of typical conditions found on the shell. Prior to 2002 Inspection 4 pits greater than 0.040" were Identified. The pits depth are 0.058" (1 pit in 1988), 0.05" (2 pits in 1991), and 0.0685" (1 pit In 1992). The pits were evaluated against the local pit depth acceptance criteria and found to be acceptable. (4) Specification SP-1302-52-120, Specification for Inspection and Localized Repair of the Torus and Vent System Coating, Includes the pit-depth acceptance criteria for rapid evaluation of observed pitting. The acceptance criteria are supported by a calculation C,1302-187-E310-038. Locations that do not meet the pit-depth acceptance criteria are characterized based on the size of the area, center to center distance between corroded areas, the maximum pit depth and location in the Torus based on major structural features. These details are sent to Oyster Creek Engineering for evaluation. (5) The acceptance criteria for pit depth Is as follows: -Isolated Pits of 0.125" in diameter have an allowed maximum depth of 0.261" anywhere in the shall provided the center to center distance between the subject pit and neighboring Isolated pits or areas of pitting corrosion Is greater than 20.0 inches. This includes old pits or old areas of pitting corrosion that have been filled andlor re-coated. is"

JNRC Information Reuest ForI

 -Multiple Pits that can be encompassed by a 2-112" diameter circle shall be limited to a maximum pit depth of 0.141" provided the center to center distance between the subject pitted area and neighboring isolated pits or areas of pitting corrosion is greater than 20.0 inches. This Includes old pits or old areas of pitting corrosion that have been filled and/or recoated.

Question: Please also provide the following Information: nominal design, as-built, and minimum measured thickness of the tows; minimum thickness required to meet ASME code acceptance criteria; the technical basis for the pitting acceptance criteria Include In Specification SP-1302-52-120

Response

Submersed area: (a) The nominal Design thickness Is 0.385 inches (b) The as-built thickness Is 0.385 Inches (c) The minimum uniform measured thickness Is, 0.343 inches - general shell 0.345 inches - shell - ring girders 0.345 Inches - shell - saddle flange 0.345 Inches - shell - torus straps (d) The minimum general thickness required to meet ASME Code Acceptance is 0.337 inches. Technical basis for pitting acceptance criteria included in Specification SP-1302-52-120 is based on engineering calculation C-1302-187-E310-038. At the time of preparation of calculation C-1302-187-E310-038 in 2002 there were no published methods to calculate acceptance standards for locally thinned areas InASME Section III or Section VIII Pressure Vessel codes. Therefore, the approach In Code Case N-597 was used as guidance in assessing locally thinned areas in the Tors. This is based on the similarity in approaches between Local Thinning Areas described in N597 and Local Primary Stress areas described In Paragraph NE3213.10 of the ASME B&PV Code Section III, particularly small areas of wall thinning which do not exceed 1.0 x (square root of Rt). In addition, the ASME B&PV Code Section III, Subsection NB, Paragraph NB-3630 allows the analysis of pipe systems Inaccordance with the Vessel Analysis rules described in Paragraph NB-3200 of the same Subsection as an alternate analysis approach. Therefore, the approach used In N597 for local areas of thinning was probably developed using the rules for Local Primary Membrane Stress from paragraph NB-3200 Inparticular Subparagraph 3213.10. The Local Primary Stress Limits In NB-3213.10 are similar to those discussed in Subsection NE, Paragraph NE-3213.10. Since the Code Case had not yet been invoked In to the Section XI program, the calculation provided a reconciliation of the results obtained from the code case against the ASME Section III code requirements as discussed above. This reconciliation demonstrated that the approach in N597 used on a pressure vessel such as the Torus would be acceptable since the results are conservative compared to the previous work performed in MPR-953 and Lm(a) (defined in N597 Table- 3622-1) £ (Rmlntmin)12. Currently, the maximum pit depth measured Inthe Tors is a 0.0685" ( measured In 1992 in bay 2). It was evaluated as acceptable using the design calculations existing at that time and was not based on

FNR C Information Request Form Calculation C-1302-187-E310-038. This remains the bounding wall thickness Inthe Torus. The criterion developed in 2002 for local thickness acceptance provides an easier method for evaluating as-found pits. The results were shown to be conservative versus the original ASME Section III and VIII Code requirements for the Torus. The Torus inspection program is being enhanced per IR373695 to Improve the detail of the acceptance criteria and margin management requirements using the ASME Section III criteria. The approach used In C-1302-187-E310-038 will be clarified as to how It maintains the code requirements. IfCode Case N-597-1 is required to develop these criteria for future Inspections, NRC review and approval will be obtained. It should also be noted that the program has established corrosion rate criteria and continues to periodically monitor to verify they remain bounded. Supplemental Information - 0411912006. This supplements response to Item 8a(1) above. The lowest recorded reading was 0.603 in December 1992. A review of the previous readings for the period 1990 thru 1992 and two subsequent readings taken in September 1994 and 1996 show this point should not be considered valid. The average reading for this point taken in 1994 and 1996 vwas 0.888 Inches. Point 14 in location 17D was the next lowest value of 0.646 Inches recorded during the 1994 outage. A review of readings, at this same point, taken during the period from 1990 through 1992 and subsequent reading taken in 1996 are conistent with this value. Thus the minimum recorded thickness in the sand bed region from inside Inspections is 0.646 inches, instead of 0.603 inches. For additional information on torus coating refer to AMP-072. LRCR #: LRA A.5 Commitment 9: IR#: Approvals: PreparedBy: Ouaou, Ahmed 4/20/2006 ReviewedBy. Miller, Mark 4/20/2006 ApprovcdBy$: Warfel, Don 4/20/2006 NRCAcceptance (Date):

Citizen's Exhibit NC3 Citizen's Exhibit NC3 - RMNuclear Calculation Sheet The purpose of this calculation is to evaluate the UT thickness measurements taken in the sandbed region during the 14R outage in support of O.C drywell corrosion mitigation project. These measurements were taken from the outside of the shell. Access to the sandbed region was achieved by cutting ten holes completely through the shield wall from the torus room. 2.0

SUMMARY

OF RESULTS: This calculation demonstrates that the UT thickness measurements for all bays meet the minimum uniform and local required thicknesses. The evaluation was performed by evaluating the UT measurements for each bay and dispositioning them relative to the uniform thickness of 0.736 inch used in GE structural analysis reports. Additional acceptance criteria was developed to address measurements below 0.736 inch. The results are summarized in Table 1. UT measurements for bays 3, 5, 7, 9, and 19 were all above the 0.736 inches and therefore acceptable. UT measurements for bays 11, 15, and 17 were all above 0.736 inches except for one measurement for each bay. After further evaluation of these three measurements including an examination of adjacent areas, it was determined that they were acceptable as shown on Table 1. UT measurements for bays 1 and 13 were evaluated using detailed criteria described in this calculation and the results are summarized in Table 1 below: OCLR00000363

0fllNuclear Calculation Sheet 2.0

SUMMARY

OF RESULTS ( Continued ): Summary of UT Evaluations Table (1) 4g

                                               '~~:*........   .                            '               .:*g Bay it/1/oc. I            0.       fM f              0.46                 0.2                    0.751?          Acceptable Bay IS/ Loc. 9            0.72r"                     0.33                 0.20("                 0.859           Acceptable Bay 17/ L.w 9             0.7120"                    0.3511               oo0"V                  o,87l           Accctable Bay I/ Loc. I             0.720r                     0218,                0.20(r                 0.739"          Accepable Bay I/ Loc. 2             0,7l6*                     0.143'               0.200                  0.659"          Arceptabl¢ Bay I/ Loc. 3             0.705"                     0.347"               0.2W0                  0.M52           Acceptable Day I/Loc. 5             _70.*o                      0313"                o~zoo                  0.82-           ,A-,eptib Bay I/ Loc*.Ba+*S 7 to:.7     VOW700                    0.7.66               0.2V0
                                                                          *oo; .o-ot.*.,         0.7660          Acceptable om.                        o z ...... C                                e.+)m,  _-

Bay I/Loc. ,1, 0..714" o2r o02o ..... 726' Acceptable Bay 13/ InOCC907 Day I/ Lo. 12 o0.724, o3o0-W020 0.2 Vo o. , Acpal Accptable

]Say 13/ Loc.21           0.6726"                       M0211             0.2V0                  0.73T           Acceptable
.Bay13/ Loc. 2r           0.6729-                    0.360"               0.2w0          [       082             Acceptable BayI/ Loc 21 ..         0.7260                     0*;.211.             0.2vo                  0.737.          Acceptable Bay Bay 13/  Lo.1 W3/Loc. 75         000 0.71V"                     0.266 0.207"               0.20T?

0.200" 0.7410 0.735- Acceptable

                                                                                                                 .Acceptable Bay    /'*,o. .. 6  .  .         -.. ..              0.'_ ....            0.20                     .756          Acceptabl Bay Bay 13/ Lot.

Loc. 1S7 0.672' 0.605' 0.51l.21 0.24" C.200, 0.723' 0.751. Acceptable 0.200" 0.7W9 Acceptable Bay 13/ LoC. 9 0.720" 0.7118 Bay 13/ Lot. 75 0.613" 0.25?" 0.200' 0.75'" Acceptable Bay 13/Loc. 10 0.7208 0.2" 0.200" 0.739' Acceptable 3Lt-a .1'028 ,ioc....... A,,,bl OCLR00000364

i JI]Nuolear Calculation Sheet Subject Colo No. Rev. No. Sheet No. O.Q Drywell Ext. Ut Evaluation C-1302-167-5320-04 hnSaded- 0. 3 Originator to Reviewed by Date MARK YEKMA 01/12/93 S. C. Tuminelli 0

3.0 REFERENCES

3.1 Drywell sandbed region pictures (see Appendix C ). 3.2 An ASHE Section VIII Evaluation of the Oyster Creek Drywell for Without Sand Case Performed by GE - Part 1 Stress Analysis, Revision 0 dated February, 1991 Report 9-3. 3.3 An ASME Section VIII Evaluation of the Oyster Creek Drywell for Without Sand Case Performed by GE - Part 2 Stability Analysis, Revision 2 dated November, 1992 Report 9-4. 3.4 ASME Section III Subsection NE Class MC Components 1989. 3.5 GE letter report " Sandbed Local Thinning and Raising the Fixity Height Analysis ( Line Items 1 and 2 In Contract

         -     PC-0391407 )" dated December 11, 1992.

3.6 GPUN Memo 5320-93-020 From K. Whitmore to J. C. Flynn "Inspection of Drywell Sand Bed Region and Access Hole", Dated January28, *1093. 4.0 ASSUMPTIONS AND BASIC DATA: 4.1 Raw UT measurements are summarized for each bay in the body of calculation. 4.2 Observations of-the outside surface of the drywell shell indicate a rough surface with varying peaks and valleys. In order to characterize an average roughness representing the depth difference of peaks and valleys, two impressions were made at the two lowest UT measurements for bay 13 using Epoxy putty . Appendix A presents the calculation of the depth of surface roughness using the drywell shell impressions taken in the roughest bay. Two locations in bay 13 were selected since it is the roughest bay. Approximately 40 locations within the two impressions were measured for depth and the average plus one standard deviation was calculated. A value of 0.200 inch was used in this calculation as a conservative depth of uniform dimples for the entire outside surface of the drywell in the sandbed region . OCLR00000365

[ Nuclear U*11 Calculation Sheet Subject Caic No. Rev. No. Sheet No. p.c Drvwell EXt. Ut Evaluation inSanQb* g-1302-187-5320T.024 0. 4 Originator Date Reviewed by Date MARK YEKTA 01/12/93 S. C Tummindefi

5.0 CALCULATION

ACCEPTANCE CRITERIA - GENERAL WALL: The acceptance criteria used to evaluate the measured drywell thickness is based upon GE reports 9-3 and 9-4 (Ref. 3.2 & 3.3) as well as other GE studies (Ref. 3.5) plus visual observations of the drywell surface ( Ref. 3.6 and Appendix C

         ). The GE reports used an assumed uniform thickness of 0.736 inches in the sandbed area. This area is defined to be from the bottom to top of the sandbed, i.e., El. 8'-111" to El.

12'-311 and extending circumferentially one full bay. Therefore, if all the UT measurements for thickness in one bay are greater than 0.736 inches the bay is evaluated to be acceptable. In bays where measurements are below 0.736 inches, more detailed evaluation is performed. This detailed evaluation is based, in part, on visual observations of the shell surface plus a knowledge of the inspection process.' The first part of this evaluation is to arrive at a meaningful value for shell thickness for use in the structural assessment. This meaningful value is referred to as the thickness ;for evaluation. It is computed by accounting for the depth of the spot where the thickness measurement is taken considering the roughness of the shell surface. The surface of the shell has been characterized as being "dimpled" as in the surface of a golf ball where the dimples are about one half inch in diameter ( Appendix C ). Also, the surface contains some depressions 12 to 18 inches in diameter not closer than 12 inches apart, edge to edge (Ref. 3.6). Appendix A presents-the calculation of the depth of surface roughness using the drywell shell impressions taken in the roughest bay. Two locations in bay 13 were selected since it is the roughest bay. Approximately 40 locations within the two impressions were measured for depth and the average plus one standard deviation was calculated to be at 0.186 inches. A value of 0.200 inch was used in this calculation as a conservative depth of uniform dimples for the entire outside surface of the drywell in the sandbed region . OCLROO000366

V0JU1Nuclear Calculation Sheet

• 5.0 CALCULATION:

ACCEPTANCE CRITERIA - GENERAL WALL: (Continued) The inspection focused on the thinnest portion of the drywell, even if it was very local, i.e., the inspection did not attempt to define a shell thickness suitable for structural evaluation. Observations indicate that some inspected spots are very deep. They are much deeper than the normal dimples found, and very local, not more than 1 to 2 inches in diameter. (Typically these observations were made after the spot was surface prepped for UT measurement. This results in a wide dimple to accommodate the meter and slightly deeper than originally found by 0.030 to 0.100 inches). The depth of these areas was measured and averaged with respect to the top of local areas as shown in Appendix A. These depths are referred to herein as the AVG micrometer measurements. The thickness for evaluation is then computed from the above information as: T (evaluation) - UT (measurement) + AVG (micrometer)

                                 . 0.200 inches where:

T (evaluation) = thickness for evaluation UT (measurement) thickness measurement at the area

                               . (location) -..

AVG (micrometer) average depth of the area relative to its immediate surroundings 0.200 inch = a conservative value of depth of typical dimple on the shell surface. After this calculation, if the thickness for analysis is greater than 0.736 inches; the area is evaluated to be acceptable. 0CLR00000367

[AiIju ocear Calculation Sheet Subjet Calc No. Rev. No. Sheet No. .Q.Q Drvwell Ext. Ut Evaluation inSandbed4 C-1302-187-5390-024 0 6 Originator ate Reviewed by Date MARK YEKTA 01/12/93 S. C Tumnminem C

5.0 CALCULATION

ACCEPTANCE CRITERIA,- LOCAL WALL: If the thickness for evaluation is less than 0.736 inches, then the use of specific GE studies is employed (Ref. 3.5). These studies contain analyses of the drywell using the pie slice finite element model, reducing the thickness by 0.200 inches in an area 12 x 12 inches in the sandbed region, tapering to original thickness over an additional 12 inches, located to result in the largest reduction possible. This location is selected at the point of maximum deflection of the eigenvector shape associated with the lowest buckling load. The theoretical buckling load was reduced by 9.5% from 6.41 to 5.56. Also, the surrounding areas of thickness greater than 0.736 inches is also used to adjust the actual buckling values appropriately. Details are provided in the body of the calculation. ACCEPTANCE CRITERIA - VERY LOCAL WALL (24 Inches In DIAMETER): All UT measurements below 0.736.inches have been determined to be in isolated locations less than 21 inches in diameter. The acceptance criteria for these measurements confined to an area less than 2J inches in diameter is based on the ASKE Section III Subsection NE Class MC.Components paragraph NE 3332.1 and NE 3335.1 titled "OPENING NOT REQUIRING REINFORCEMENT AND REINFORCEMENT OF MULTIPLE OPENINGS". These Code provisions allow holes up to 21 inches in diameter in Class MC vessels without requiring reinforcement. Therefore, thinned areas less than 2J inches in diameter need not be provided with reinforcement and are considered local. Per NE 3213.10 the stresses in these regions are classified as local primary membrane stresses which are limited to an allowable value of 1.5 Sm. Local areas not exceeding 24 inches in diameter have no impact on the buckling margins. Using the 1.5 Sm criteria given above, the required minimum thickness in these areas is: T ( required ) = ( 2/3 ) * ( 0.736 ) = 0.490 inches Where 2/3 is Sm/1.5Sm and is the ratio of the allowable stresses. UT thickness measurements for all ten bays are above 0.490 inches. 0CLR00000368

PflNuolear Calculation Sheet Subje c Noo Rev. No. Sheet No. O.c nrvwell rxt. Ut Eygiuation in Sandbe4 C30-187-5320-024- 0 7 Originator Date Reviewed by Date MARX YKTA 01/12/93 S. C Tumminefli 0'

5.0 CALCULATION

UT EVALUATION: BAY # 1: The outside surface of this bay is rough and full of dimples similar to the outside surface of golf ball. This observation is made bythe inspector who located the thinnest areas for the UT examination. This inspection focused on the thinnest areas of the drywell, even if it was very local, i.e., the inspection did not attempt to define a shell thickness suitable for structural evaluation. The shell appears to be relatively uniform in thickness except for a band of corrosion which looks like a "bathtub" ring, located 15 to 20 inches below the vent pipe reinforcement plate, i.e, weld line as shown in Figure 1. ( Figure 1 and others like figures presented in this calculation are NOT TO SCALE). The bathtub ring is 12 to 18 inches wide and about 30 inches long located in the center of the bay. Beyond the bathtub ring on both sides, the shell appears to beuniform in thickness at a conservative value of 0.800 inches. Above the bathtub ring the shell exhibits no corrosion since the original lead primer on the vent pipe/reinforcement plate is intact. Measurements 14 !and 15 confirm that the thickness above the bathtub ring is at 1.154 inches starting at elevation 11'-00". Below the bathtub ring the shell is uniform in thickness where no abrupt changes in thicknesses are present. Thickness measurements below the bathtub ring are all above 0.800 inches except location 7 which is very local area. Therefore, a conservative mean thickness of 0.800 inches is estimated to represent the evaluation thickness for this bay. Given a uniform thickness of 0.800 inches, the buckling margin for the refueling load condition can be recalculated based on the GE report 9-4 (Ref. 3.3). The theoretical buckling strength from report 9-4 (ANSYS Load Factor) is a square function of plate thicknesses. Therefore, a new buckling capacity for the controlling refueling load combination is calculated to be at 13% above the ASME factor of safety of 2 as shown in Appendix B. OCLROO000369

wri2]Nuclear Calculation Sheet Subject CaIC No. Rev. No. Sheet No. OC Drywall Ext., Ut Evaluation 6nande C-1302-187-5320-0947 0. a Originator Date Reviewed by Date MARK 'YEKTA 01/12/93 S. C. Tummieli C 5

• 0 CALCULATION:

UT EVALUATION: BAY # I ( Continued): Locations 1, 2, 3, 4, 5, 10, 11, 12, 13, 20, and 21 are confined to the bathtub ring as shown in Figure 1. An average value of these measurements is an evaluation thickness for this band as follows; Location Evaluation Thickness 1 0.738" 2 0.659" 3 0.852" 4 0.760" 5 0.823" 10 0.839" 11 0.726" 12 0.825" 13 0.792" 20 0.965" 21 0.737" Average = 0.792" An average evaluation thickness of 0.792 inches for the bathtub ring may raise concern given that the bathtub ring is noticeable and that the difference between its average evaluation thickness (0.792 inches) and the average thickness taken for the entire region (0.800 inches) is only 0. 008 inches. This results from the fact that average micrometer readings were generally not taken for the remainder of the shell since each reading was greater than 0.736 inches. In reality, the remainder of the shell is much thicker than 0.800 inches. The appropriate evaluation thickness can not be quantified since no micrometer readings were taken. The individual measured thicknesses must also be evaluated for structural compliance. Table 1-a identifies 23 locations of UT measurements that were selected to represent the thinnest areas, except locations 14 and 15, based on visual examination. These locations are a deliberate attempt to produce a minimum measurement. Locations 14 and 15 were selected to confirm that no corrosion had taken place in the area above the bathtub ring. OCLROO000370

0 ZINuclear

5.0 CALCULATION

UT EVALUATION: BAY # I ( Continued): Eight locations shown in Table 1-a (1, 2, 3, 5, 7, 11, 12, and 21) have measurements below 0.736 inches. Observations indicate that these locations were very deep and not more than 1 to 2 inches in diameter. The depth of each of these areas relative to its immediate surroundings was measured at 8 locations around the spot and the average is shown in Table 1-a. Using the general wall thickness acceptance criteria described earlier, the evaluation thickness for all measurements below 0.736 inches were found to be above 0.736 inches except for two locations, 2 and 11, as shown in Table 1-b. Locations 2 and 11 are in the bathtub ring and are about 4 inches apart. This area is characterized as a local area 4 x 4 inches located at about 15 to 20 inches below the vent pipe reinforcement plate with an average thickness of 0.692 inches. This thickness of 0.692 inches is 0.108 inches reduction from the conservative estimate of 0.800 inches evaluation thickness for the entire bay. In order to quantify the effect of this local region and to address structural compliance, the GE study on local effects is used (Ref. 3.5). This study contains an analysis of the drywell shell using the pie slice finite element model, reducing the thickness by 0.200 inches (from 0.736 to 0.536 inches) in an area 12 x 12 inches in the sandbed region located to result in the largest reduction possible. This location is selected at the point of maximum deflection of the eigenvector shape associated with the lowest buckling load. The theoretical buckling load was reduced by 9.5%. The 4 x 4 inches local region is not at the point of maximum deflection. The area of 4 x 4 inches is only 11% of the 12 x 12 inches area used in the analysis. Therefore, this small 4 x 4 inches area has a negligible effect on the buckling capacity of the structure. In summary, using a conservative estimate of 0.800 inches for evaluation thickness for the entire bay and the presence of a bathtub ring with an evaluation thickness of 0.792 inches plus the acceptance of a local area of 4 x 4 inches based on the GE study, it is concluded that the bay is acceptable. OCLR00000371

0!Jj1Nuolear Calculation Sheet

5.0 CALCULATION

UT EVALUATION: BAY # 1 (Continued): Bay # 1 UT Data

                           .Table- I-a
                                       .~~~..:.

1 0.720 0.218 2 0.716 0.143 3 0.705 0.347 4 0.760 --- 5 0.710 0.313 6 0.760 --- 7 0.700 0.266 8 0.805 --- 9 0.805 --- 10 0.839 --- 11 0.714 0.212 12 0.724 0.301 13 0.792 14 1.147 15 1.156 16 0.796 __--- 17 0.860 --- is 0.917 --- 19 0.890 --- 20 0.965 --- 21 0.726 0.211 22 0.852 -- 23 0.850 --- I 0CLR00000372

EM~ufluoear CAlculation Sheet

5.0 CALCULATION

UT VýUATION: BAY # 1:(continued)

SUMMARY

OF Measurements BELOW 0.7 Table I-b 2 0.720 0.218 0.200 0.7380 Acceptable 2 0.716' 0.143" 0.200" 0.659" Acceptable 3 0.705" 0.347? 0.200 0.52" Acc=ptable s 0.710" 0.313' 0.200W 0.823' Acceptable 7 0.700 0.266" 0,200' 0.7660 Acceptable 11 0.714" 0.212' 0.200W , 0.726" Acceptable 12 0.724" 0-301' 0.00" 0.825' Acceptable 21 0.726" 0.211" 0.200" 0.73?' Acceptable OCLROO000373

01Nuclear Calculation Sheet BAY #1 DATA NOTES:

1. All 'Location" measurements from Intersection of the DW shell and vent collar fillet welds.

2. Pit depts are average of four readings taken at 0/45°190W/135 within 1"band surrounding ground spots. Only measured where remaining wall thk.

was below 0.736m. 14 a 15 DW SHELL 9* .19 18 7

                              "'    6 16 23,:.k17 FIGURE (1)

OCLROO000374

0 Nuolear 5#0 CALCULATIONt UT EVALUATION: BAY # 3: The outside surface of this bay is rough, similar to bay one, full of dimples comparable to the outside surface of golf ball. This observation is made by the inspector who located the thinnest areas for the UT examination. The shell appears to be relatively uniform in thickness except for a bathtub ring 8 to 10 inches wide approximately 6 inches below the vent header reinforcement plate. The upper portion of the shell beyond the band exhibits no corrosion where the original red lead primer is still intact. Eight locations were selected to represent the thinnest areas based on the visual observations of the shell surface (Fig. 3). These locations are a deliberate attempt to produce a minimum measurement. Table 3 shows measurements taken to measure the thicknesses of the drywell shell using a D-meter. The results indicate that all of the areas have thickness greater than the 0.736 inches. Given the UT measurements, a conservative mean evaluation thickness of 0.850 inches is estimated for this bay and therefore, it is concluded that the bay is acceptable. Day # 3 UT Data Table 3 1 0.795 2 1.000 --- 3 0.857 --- 4 0.898 --- 5 0.823 --- 6 I 0.968 --- 7 0.826 --- 8 0.780 --- OCLR00000375

[Al 9Nuclear Calculation Sheet FIGURE (3) 0CLR00000376

0f lNuclear ar Calculation Sheet Subject Wc No. Rev. No. Sheet No. O.C Drvwe11 Ext. Ut Evaluation in Sndbed C-1302-187-5320-024 0 15 Originator Date Reviewed by Date MARK YEKTA 01/12/93 S. C Tumminelli

5.0 CALCULATION

UT EVALUATION: BAY # 5: The outside surface of this bay is rough and very similar to bay 3 except that the local areas are clustered at the junction of bays 3 and 5, at about 30 inches above the floor. The shell surface is full of dimples comparable to the outside surface of golf ball. This observation is made by the inspector who located the thinnest areas for the UT examination. The shell appears to be relatively uniform in thickness. Eight locations were selected to represent the thinnest areas based on the visual observations of the shell surface (see Fig. 5). These locations are a deliberate attempt to produce a minimum measurement. Table 5 shows readings taken to measure the thicknesses of the drywell shell using a D-meter. The results indicate that all of the areas have thickness greater than the 0.736 inches. Given the UT measurements, a conservative mean evaluation thickness of 0.950 inches is estimated for this bay and therefore, it is concluded that the bay is acceptable. Bay # 5 UT Data Table 5 OCLROO000377

t0r rNNuclear Calculation Sheet FIGURE (5) OCLROO000378

S N uclear Calculation Sheet Subject Cale No. R. No. Sheet No. OC DrvwP1l Vxt. Ut Evaluation inSndbed C-1302-187-5320-024T,--0 17 Originator Date Reviewed by Date MARK YEXTA 01/12/93 S. C TummineHi 0

5.0 CALCULATION

UT EVALUATION: BAY L 7: The observation of the drywell surface for this bay showed uniform dimples in the corroded area, but they are shallow compared to those in bay 1. The bathtub ring seen in the other bays, was not very prominent in this bay. This observation is made by the inspector who located the thinnest areas for the UT examination. The shell appears to be relatively uniform in thickness. Seven locations were selected to represent the thinnest areas based on the visual observations of the shell surface (Fig. 7). These locations are a deliberate attempt to produce a minimum measurement. Table 7 shows readings taken to measure the thicknesses of the drywell shell using a D-meter. The results indicate that all of the areas have thickness greater than the 0.736 inches. Given the UT measurements, a conservative mean evaluation thickness of 1.00 inches is estimated for this bay and therefore, it is concluded that the bay is acceptable. Bay 1-7-ET Data OCLROO000379

[jZig ue ear Calculation Sheet FIGURE (7) OCLROO000380

PlUIJ~uclear Calculation Sheet 5.*0 CAALCULAION: UT EVALUATION: BAY j 9: The observation of the drywell shell for this bay was very similar to bay 7 except that the bathtub ring was more evident in this bay. The shell appears to be relatively uniform in thickness except for a bathtub ring 6 to 9 inches wide approximately 6 to 8 inches below the vent header reinforcement plate. The upper portion of the shell beyond the band exhibits no corrosion where the original red lead primer is still intact. Eight locations were selected to represent the thinnest areas based on the visual observations of the shell surface (Fig. 9). These locations are a deliberate attempt to produce a minimum measurement. Table 9 shows readings taken to measure the thicknesses of the drywell shell using a D-meter. The results indicate that all of the areas have thickness greater than the 0.736 inches. Given the UT measurements, a conservative mean evaluation thickness of 0.900 inches is estimated for this bay and therefore, it is concluded that the bay is acceptable. Bay # 9 UT Data Table 9 f ........ 1 0.960 --- 2 0.940 --- 3 0.994 --- 4 1.020 --- 5 0.985 --- 6 0.820 --- 7 0.825 8 0.791 --- 9 0.832 --- 10 0.980 --- OCLRO0000381

9fNuclear Calculation Sheet FIGURE (9) OCLR00000382

Ri Nuclear Calculation Sheet Subject Calc No. R . No. Sheet No. O.C Drvw=11 Ext. Vt Evaluation LWn Snbdc-1302-187-5320-024 0 21 Originator Date Reviewed by Date MARK YEKTA 01/12/93 S. C Tumminelli 0

5.0 CALCULATION

VT EVALUATION: BAY # 11: The outside surface of this bay is rough, similar to bay 1, full of uniform dimples comparable to the outside surface of a golf ball. The shell appears to be relatively uniform in thickness except for local areas at the upper right corner of Figure 11, located at about 10 to 12 inches below the vent pipe reinforcement plate. Eight locations were selected to represent the thinnest areas based on the visual observations of the shell surface (Fig. 11). These locations are a deliberate attempt to produce a minimum measurement. Table 11-a shows readings taken to measure the thicknesses of the drywell shell using a D-meter. The results indicate that all of the areas have thickness greater than the 0.736 inches, except one location. Location 1 as shown in Table 11-a, has a reading below 0.736 inches. Observations indicate that this location was very deep and not more than 1 to 2 inches in diameter. The depth of area relative to its immediate surroundings was measured at 8 locations around the spot and the average is shown in Table 11-a. Using the general wall thickness acceptance criteria described earlier, the evaluation thickness for location 1 was found to be above 0.736 inches as shown in Table 11-b. Given the UT measurements, a conservative mean evaluation thickness of 0.790 inches is estimated for this bay and therefore, it is concluded that the bay is acceptable. OCLROO000383

LdJIJNuclear Calculption Sheet

5.0 CALCULATION

UT EVALUATION: BAY # It (Continued): Day # 11 UT Data Table 11-a 1 0.705 0.246 2 0.770 --- 3 0.832 4 0.755 --- 5 0.831 --- 6 0.800 7 0.831 8 0,815 --- Summary of Measurements Below 0.736 Inches Table 11-b OCLROO000384

[!JrIiNuclear Calculation Sheet FIGURE ( 11 ) OCLR00000385

A IINucear Calculation Sheet Subject Caic No. Rev. No. Sheet No. O.C Drvwell Ext. Ut Evaluation CnSnb

                                                  -1302-187-5320-O4                    0       24 Originator                       rate          Reviewed by                                Date MARK YEKTA                   01112/93          S. C Tunmmiuneli                           0 S.0 CALCULATION:

UT EVALUATION: DAY # 13: The outside surface of this bay is rough and full of dimples similar to bay 1 as shown in Appendix C. This observation is made by the inspector who located the thinnest areas in deep valleys thereby biasing the remaining wall measurements to the conservative side. This inspection focused on the thinnest areas, even if very local, i.e., the inspection did not attempt to define a shell thickness suitable for structural evaluation. The variation in shell thickness is greater in this bay than in the other bays. The bathtub ring below the vent pipe reinforcement plate was less prominent than was seen in other bays. The corroded areas are about 12 to 18 inches in diameter and are at 12 inches apart, located in the middle of the sandbed. Beyond the corroded areas on both sides, the shell appears to be uniform in thickness at a conservative value of 0.800 inches. Near the vent pipe and reinforcement plate the shell exhibits no corrosion since the original lead primer on the vent pipe/reinforcement plate is intact. Measurement 20 confirms that the thickness above the bathtub ring is at 1.154 inches. Below the bathtub ring the shell appears to be fairly uniform in thickness where no abrupt changes in thickness are present. Thickness measurements below the bathtub ring are all 0.800 inches or better. Therefore, a conservative mean thickness of 0.800 inches is estimated to represent the evaluation thickness for this bay. Given a uniform thickness of 0.800 inches, the buckling margin for the refueling load condition is recalculated based on the GE report 9-4 (Ref. 3.3). The theoretical buckling strength from report 9-4 (ANSYS Load Factor) is a square function of plate thicknesses. Therefore, a new buckling capacity for the controlling refueling load combination is calculated to be at 13% above the ASME factor of safety of 2 as shown in Appendix B. OCLROO000386

P-1 U Nuclear Calculation Sheet Subject Caic No. Rev. No. Sheet No. O.C DrvwM11 Ext. Uyt rvaluation in__Sandbed. C-1302-167-5320-024 0 2.5 Originator Date Reviewed by Date MARKYEKTA 01/12/93 S. C Tummineni 5.0 CALCULATIONt UT EVALUATION: DAY # 13 f-Continued ): Locations 5, 6, 7, 8, 10, 11, 14, and 15 are confined to the bathtub ring as shown in Figure 13. An average value of these measurements is an evaluation thickness for this band as follows; Location Evaluation Thickness 5 0.735" 6 0.756" 7 0.675" 8 0.796" 10 0.7391" 11 0.741" 12 0.885" 14 0.868", 15 0.756" 16 0.829" Average = 0.778" The inspector suspected that some of the above locations in the bathtub ring were over ground. Subsequent locations with suffix A, e.g. 5A, 6A, were located close to the spots in question and were ground carefully to remove the minimum amount of metal but adequate enough for UT examination as shown in Table 13-a. The results indicate that all subsequent measurements were above 0.736 inches. The average micrometer measurements taken for these locations confirm the depth measurements at these locations. In spite of the fact that the original measurements were taken at heavily ground locations they are the ones used in the evaluation. The individual measurements must also be evaluated for structural compliance. Table 13-a identifies 20 locations of UT measurements that were selected to represent the thinnest areas, except location 20, based on visual examination. These locations are a deliberate attempt to produce a minimum measurement. Location 20 was selected to confirm that no corrosion had taken place in the area above the bathtub ring. 0CLR00000387

t!J a Nuclear Calculation Sheet Subjct CaIc No. Rev. No. Sheet No. O.C D3:3=1l Ext. Ut 39valu tioan inSndbed c-1302-187-5320-0241 0 26 Originator Date Reviewed by Date MARK YETA 01/12/93 S. C Tumminelli 0

5.0 CALCULATION

UT EVALUATION: BAY 1 13 ( Continued ): Nine locations shown in Table 13-a (1, 2, 5, 6, 7, 8, 10, 11, and 15) have measurements below 0.736 inches. Observations indicate that these locations were very deep, overly ground, and not more than 1 to 2 inches in diameter. The depth of each of these areas relative to its immediate surroundings was measured at 8 locations around the spot and the average is shown in Table 13-a. Using the general wall thickness acceptance criteria described earlier, the evaluation thickness for all measurements below 0.736 inches were found to be above 0.736 inches except for two locations, 5 and 7, as shown in Table 13-b. In addition, subsequent measurements close to the locations identified above, were taken and they were all above 0.736 inches. Locations 5 and 7 are in the bathtub ring and are about 30 inches apart. These locations are characterized as local areas located at about 15 to 20 inches below the vent pipe reinforcement plate with an evaluation thicknesses of 0.735 inches and 0.677 inches. The location 5 is near to location 14 for an average value of 0.801 inches and therefore acceptable. Location 7 could conservatively exist over an area of 6 x 6 inches for a thickness of 0.677 inches. This thickness of 0.677 inches is a full 0.123 inches reduction from the conservative estimate of 0.800 inches evaluation thickness for the entire bay. In order to quantify the effect of this local region and to address structural compliance, the GE study on local effects is used (Ref. 3.5). This study contains an analysis of the drywell shell using the pie slice finite element model, reducing the thickness by 0.200 inches (from 0.736 to 0.536 inches) in an area 12 x 12 inches in the sandbed region located to result in the largest reduction possible. This location is selected at the point of maximum deflection of the eigenvector shape associated with the lowest buckling load. The theoretical buckling load was reduced by 9.5%. The 6 x 6 inch local region is not at the point of maximum deflection. The area of 6 x 6 inches is only 25% of the 12 x 12 inches area used in the analysis. Therefore, this small 6 x 6 inch area has a negligible effect on the buckling capacity of the structure. OCLROO000388

0.INuclear Calculation Sheet 5,0 CALCULATION: UT EVALUATION: BAY # 13 ( Continued ): In summary, using a conservative estimate of 0.800 inches for evaluation thickness for the entire bay and the presence of a bathtub ring with a evaluation thickness of 0.778 inches plus the acceptance of a local area of 6 x 6 inches based on the GE study, it is concluded that the bay is acceptable. Bay # 13 UT Data Table 13-a S . -U....'Z...~t.

                 *.          . .        .    .    . ...........
                                                            ............

1/lA 0.672/0.890 0.351 2/2A 0.722/0.943 0.360 3 0.941 --- 4 0.915 --- s/SA 0.718/0.851 0.217 6/6A 0.655/0.976 0.301 7/7A 0.618/0.752 0.257 8/8A 0.718/0.900 0.278 9 0.924 --- 10/10A 0.728/0.810 0.211 11/1hA 0.685/0.854 0.256 12 0.885 --- 13 0.932 14 0.868 --- 15/15A 0.683/0.859 0.273 16 0.829 --- 17 0.807 --- 18 0.825 --- 19 0.912 --- 20 1.170 --- OCLROO000389

0I3jiNuclear Calculation Sheet 5 *0 CALCULRTION: UT EVALUATION: BAY # 13 C Continued ): Summary of Measurements Below 0.736 Inches Table 13-b 1 0.67r 0.351' O~w U230 Acceptablc 2 0.722" 0.360' 0.200W 0.882' Acceptable S 0.71W 0.217' 0.200" 0.735' Acceptable 6 0.655' 0.301' 0.200" 0.756' Acceptable 7 0.618" 0.25" 0.200w 0.675' Acceptable 8 0.718" 0.278' 0.200' 0.796' Acceptable 10 0.728' 0.2110 0.200V 0.739' Acceptable 11 0.685" 0.256r 0.200' 0.741V Acceptable 15 0.683- 0.273a 0.200' 0.7S6. Acceptable OCLRO0000390

P1 d Nuclear Calculation Sheet BAY #13 DATA NOTES:

1. All measurements from Intersection of the DW shell (butt) end vent collar (fillet) welds.

2. Spots with suffix (e.g. IA or 2A) were located close to the spots in question and were ground carefully to remove minimum amount of metal but adequate enough for UT.

3. Pit depths are average of four readings taken at 0/450/90"/135*

within V distance around ground spot. Taken only where remaining wall showed below 0.736".

                                                  *20 2.1
                                                    .

17 0 DW 160 7715 .14 4.0

                                                          .-   SHELL 013 8         ;6    5 12,           ,11              010                so FIGURE ( 13 )

OCLR00000391

0IId Nuclear Calculation Sheet Subject Calc No. Rev. No. Sheet No. O.C Dnr3=ll Ext. ut Rvalu~t~on in Sandbedl C-1302-187-S320-024 0 30 Originator Date Reviewed by Date MARK YEKTA 01/12/93 S. C TbmmineW C

5.0 CALCULATION

UT EVALUATION: BAY # IS: The outside surface of this bay is rough, similar to bay 1, full of uniform dimples comparable to the outside surface of golf ball (Appendix C ). The bathtub ring seen in the other bays, was not very prominent in this bay. This observation is made by the inspector who located the thinnest areas for the UT examination. The upper portion of the shell beyond the ring exhibits no corrosion where the original red lead primer is still intact. The shell appears to be relatively uniform in thickness. Eleven locations were selected to represent the thinnest areas based on the visual observations of the shell surface (Fig. 15). These locations are a deliberate attempt to produce a minimum measurement. Table 15-a shows readings taken to measure the thicknesses of the drywell shell using a D-meter. The results indicate that all of the areas have thickness greater than the 0.736 inches, except one location. Location 9 as shown in Table 15-a, has a reading below 0.736 inches. Observations indicate that this location was very deep and not more than 1 to 2 inches in diameter. The depth of area relative to its immediate surrounding was measured at 8.1ocations around the spot and the average is shown in Table 15-a. Using the general wall thickness acceptance criteria described earlier, the evaluation thickness for location 9 was found to be above 0.736 inches as shown in Table 15-b. Given the UT measurements, a conservative mean evaluation thickness of 0.800 inches is estimated for this bay and therefore, it is concluded that the bay is acceptable. OCLR00000392

0AI2JANuclear

5.0 CALCULATION

UT EVALUATION: BAY #L15 Bay #15 UT Data Table l5-a 1 0.786 --- 2 0.829 3 0.932 --- 4 0.795 -- 5 0.850 6 0.794 7 0.808 --- 8 0.770 --- 9 0.722 0.337 10 0.860 --- 11 0.825 _I --- Summary of Measurements Below 0.736 Inches Table 15-b OCLR00000393

V1Li#JNuclear -CalculationSheet BAY #15 DATA NOTES:

1. All measurements from Intersection of the DW shell and vent collar (fillet) welds.

2. Pit depths are average of four readings taken at 0/451907/135° within 1 distance around ground spots. Taken only when remaining wall thickness shown below 0.736.

6 a 1.. S DW S 2I SHELL 11 10 7 4 3 "9 is a FIGURE ( 15 ) OCLROO000394

IAl UINuclear Calculation Sheet Subject Ca*c No. Sheet No. Q_.C Drvwell Ext. Ot Evaluation -InanddC-3-873202 0- 33 Originator Date Reviewed by Datc MARK YEKTA 01/12/93 S. C Tumminelli

5.0 CALCULATION

UT EVLUATION: BAY # 17: The outside surface of this bay is rough, similar to bay 1, full of uniform dimples comparable to the outside surface of golf ball. The shell appears to be relatively uniform in thickness except for a band 8 to 10 inches wide approximately 6 inches below the vent header reinforcement plate. The upper portion of the shell beyond the band exhibits no corrosion where the original red lead primer is still intact. Eleven locations were selected to represent the thinnest areas based on the visual observations of the shell surface (Fig. 17). These locations are a deliberate attempt to produce a minimum measurement. Table 17-a shows readings taken to measure the thicknesses of the drywell shell using a D-meter. The results indicate that all of the areas have thickness greater than the 0.736 inches, except one location. Location 9 as shown in Table 17-a, has a reading below 0.736 inches. Observations indicate that this location is very deep and not more than 1 to 2 inches in diameter. The depth of area relative to its immediate surroundings was measured at 8 locations around the spot and the average is shown in Table 17-a. Using the general wall thickness acceptance criteria described earlier, the evaluation thickness for location 9 was found to be above 0.736 inches as shown in Table 17-b. Given the UT measurements, a conservative mean evaluation thickness of 0.900 inches is estimated for this bay and therefore, it is concluded that the bay is acceptable. OCLR00000395

PIL1NIuclear Calculation Sheet

5.0 CALCULATION

UT EVALUATION: BAY # 17 (Continued): Bay #17 UT Data Table 17-a 1 0.916-- 2 1.150 --- 3 0.898 4 0.951 --- 5 0.913 --- 6 0.992 7 0.970 --- 8 0.990 --- 9 0.720 0.351 10 0.830 --- 11 0.770 --- Summary m T J ofl Measurements m Below 0.736 i Inches-Table 17-b

                                                     .. . .. ...
                                                             .:

9 0.7200 03510 O.20 0,871- Acceptable OCLR00000396

MENuclear Calculation Sheet BAY #17 DATA NO0TES:

1. All ruazuruniont frorn In Hatinfla 01 ths DW (butt) shill Aa YOMi Collar (11111l) WWIde 2.Pit depthivane everaga of four readingi talcan at QWaIrM"a~ within 1'1 distance round gwuund 45 spot&. Taken only when rormn.Inbi wAll I hicdknse was bIIOW 0.736.

                                                  .2 DW II 10                       SHELL 7                                      *1 4

lie L FIGURE ( 17 ) OCLROO000397

0IAUI~uclear Calculation Sheet

5.0 CALCULATION

UT EVALUATION: BAY # 19s The outside surface of this bay is rough and very similar to bay 17. Locations 1 through 7 as shown in Table 19, were ground carefully to minimize loss of good metal. The shell surface is full of dimples comparable to the outside surface of a golf ball. This observation is made by the inspector who located the thinnest areas for the UT examination. The shell appears to be relatively uniform in thickness. Ten locations were selected to represent the thinnest areas based on the visual observations of the shell surface (Fig. 19). These locations are a deliberate attempt to produce a minimum measurement. Table 19 shows readings taken to measure the thicknesses of the drywell shell using a D-meter. The results indicate that all of the areas have thickness greater than the 0.736 inches. Given the UT measurements, a conservative mean evaluation thickness of 0.850 inches is estimated for this bay and therefore, it is concluded that the bay is acceptable. BaY #19 UT Data Table 19

                       ...................              ~..:YK 2                      0.924--

3 0..950... 6 0

• 860 ---

7 0.969 8 0.753 9 0.776 10 0.790 OCLR00000398

I'di UJIMuclear Calculation Sbeet Subject Calc No. Rev. No. Sheet No. OC Dnry=el Ext, Ut Evaluation i Snded. C-1302-187-5320-024I 0 2-37 L5 Originator ate Reviewed by Date MARK YEKTA 01/12/93 S. C. Tumminelli 04/16/93 BAY #19 DATA NOTES: 1- All measurmmnla from lItersecoiup of thq OW RlNIl (butt) and Gent cllr 11110el) waldi. J:WD DW aSHELL C~S I It .3ýmt~ P 6F

5. 5e o m ~ 06

                                                                                         ý. ih

• M eaA

                                                                                        ;...                 t%~ ar
                                                      . . ..
                                                                           ~ C ott~
                                   &ptDJl dhpg61T    t~d.U               %,IA 6d          +cci    c z:O       I:

otc

• A

                     %   su oor 4 W 45~av a 4                                                         P        I V; FIGURE ( 19 )

OCLROO000399

0~Nuclear Calculation Sheet APPENDIX A

SUMMARY

OF MEASUREMENTS OF IMPRESSIONS TAKEN FROM BAY #13 OCLROO000400

0~Nuclear Calculation Sheet The purpose of this appendix is to characterize the depth of typical uniform dimples on the shell surface. This depth is used in acceptance criteria to quantify the evaluation thickness for an area where the micrometer readings are available. Two locations in bay 13 were selected since bay 13 is the roughest bay. Impressions of drywell shell surface using DMR_503 Epoxy Replication Putty manufactured by Dyna Mold Inc were made. These impressions were about 10 inches in diameter and about 1 inch thick. The UT locations 7 and 10 in bay 13 were identified in each of these impression as the reference points. This is a positive impression of the drywell shell surface. The depth of the typical dimples were measured as follows; READING DEPTH # 10 DEPTH # 7 (Location) (inches) (inches) 1 0.150 0.075 2 0.000 0.110 3 0.200 0.135 4 0.140 0.200 5 0.150 0.000 6 0.040 0.000 7 0.150 0.170 8 0.010 0.205 9 0.134 10 0.145 0.145 11 0.118 0.064 12 0.105 0.200 13 0.125 0.045 14 0.200 0.180 15 0.135 0.105 16 0.100 17 0. 175 0.035 18 0.175 0.015 19 0.155 0.190 20 0.175 0.055 21 0.175 0.305 22 0.135 OCLROO000401

SIWINuclear Calculation Sheet Subject Caic No. Rev. No. Sheet No. O.C Drywell .Ext. Ut Evaluatonn Sandbed C-1302-187-5320-024 0 4 Originator Date Reviewed by Date MARK YEKTA 01/12/93 S, C Tumminelli C Location # 10: Mean Value = 0.131 Standard Deviation = 0.055 Mean Value + One S.D = 0.186 Location # 7: Mean Value = 0.118 Standard Deviation = 0.082 Mean Value + One S.D = 0.200 Therefore, a value of 0.200 inches was used as the depth of uniform dimples for the entire outside surface of the drywell in the sandbed region. OCLROO000402

0INuclear Calculation Sheet APPENDIX B BUCKLING CAPACITY EVALUATION FOR VARYING UNIFORM THICKNESS OCLROO000403

r Lr rr7 r- r W 0?Nuclear Calculation Sheet CALCULATION OF BUCKLING MARGIN - REFUELING CASE, NO SAND - GE OYCR1S&T - UNIFORM THICKNESS t= 0.736 Inch LOAD ITEM PARAMETER UNITS VALUE FACTOR

               *** DRYWELL GEOMETRY AND MATERIALS 1         Sphere Radius, R                                   (in.)    420 2         Sphere Thickness, t                                (in.)  0.736 3         Material Yield Strength, Sy                        (ksi)     38 4         Material Modolus of Elasticity, E                  (ksi)  29600 5         Factor of Safety, FS                                           2
               *** BUCKLING ANALYSIS RESULTS 6         Theoretical Elastic Instability Stress, Ste        (ksi) 46.590   6.140
               *** STRESS ANALYSIS RESULTS 7         Applied Meridional Compressive Stress, Sm          (ksi)  7.588   5.588 8         Applied Circumferential Tensile Stress, Sc         (ksi)  4.510   3.300 AAA CAPACITY REDUCTION FACTOR CALCULATION 9         Capacity    Reduction   Factor,     ALPHAI                0.207 10        Circumferential Stress Equivalent Pressure, Peq    (psi) 15.806 11        'X' Parameter, X- (Peq/8E) (d/t)A2                        0.087 12        Delta C (From Figure - )                                  0.072 13        Modified Capacity Reduction Factor, ALPHAi,mod            0.326 14        Reduced Elastic Instability Stress, Se             (ksi) 15.182   2.001
               *** PLASTICITY REDUCTION FACTOR CALCULATION 0     15        Yield Stress Ratio, DELTA=Se/Sy                           0.400 r-    16        Plasticity Reduction Factor, NUi                          1.000 0     17        Inelastic Instability Stress, Si - NUi x Se        (ksi) 15.182   2.001 0C)
               *** ALLOWABLE COMPRESSIVE STRESS CALCULATION C)   18        Allowable Compressive Stress, Sall     SI/FS       (ksi)  7.591    1.000 o    19        Compressive Stress Margin, M=(Sall/Sm -1) x 100%   M%       0.0 4ýý-

ir - r r -- r r Ut r r Wr 0~Nuclear Calculation Sheet CALCULATION OF BUCKLING MARGIN - REFUELING CASE, NO SAND GE OCRFST01 - UNIFORM THICKNESS t-0.776 Inch LOAD ITEM PARAMETER UNITS VALUE FACTOR

                  *** DRYWELL GEOMETRY AND MATERIALS 1         Sphere Radius, R                                     (in.)    420 2         Sphere Thickness, t                                  (in.)  0.776 3         Material Yield Strength, Sy                          (ksi)     38 4         Material Modolus of Elasticity, E                    (ksi)  29600 5         Factor of Safety, FS                                             2
                  *** BUCKLING ANALYSIS RESULTS 6         Theoretical Elastic Instability Stress, Ste          (ksi) 49.357   6.857
                  *** STRESS ANALYSIS RESULTS 7         Applied meridional Compressive Stress, Sm            (ksi)  7.198   5.588 8         Applied Circumferential Tensile Stress, Sc           (ksi)  4.248   3.300
                  *** CAPACITY REDUCTION FACTOR CALCULATION 9         Capacity Reduction Factor, ALPHAI                           0.207 10        Circumferential Stress Equivalent Pressure, Peq      (psi) 15.697 11        'X' Parameter, X- (Peq/8E) (d/t)A2                          0.078 12        Delta C (From Figure - )                                    0.066 13        Modified Capacity Reduction Factor, ALPHA,J,mod             0.316 14        Reduced Elastic Instability Stress, Se               (ksi) 15.583   2.165
                  *** PLASTICITY REDUCTION FACTOR CALCULATION 0        15        Yield Stress Ratio, DELTA=Se/Sy                             0.410 0        16        Plasticity Reduction Factor, NUi                            1.000 17        Inelastic Instability Stress, Si - NUi x Se          (ksi) 15.583   2.165 4--

0 ALLOWABLE COMPRESSIVE STRESS CALCULATION 0 0 18 Allowable Compressive Stress, Sall = SI/FS (ksi) 7.792 1.082 0o3 19 Compressive Stress Margin, M=(Sall/Sm -1) x 100%. 8.2

I-r u- r- r77r-7 r--" r r-77. r7-- r- r7r>r r r r r r 0flflNuclear Calculation Sheet CALCULATION OF BUCKLING MARGIN - REFUELING CASE, NO SAND GPUN EVALUATION FOR UNIFORM THICKNESS t-0.800 Inch USING THICKNESS RATIO LOAD ITEM PARAMETER UNITS VALUE FACTOR

                  *** DRYWELL GEOMETRY AND MATERIALS 1         Sphere Radius, R                                            (in.)     420 2         Sphere Thickness, t                                         (in.)  0.800 3         Material Yield Strength, Sy                                 (ksi)      38 4         Material Modolus of Elasticity, E                           (ksi)  29600 5         Factor of Safety, FS                                                    2
                  *** BUCKLING ANALYSIS RESULTS 6         Theoretical Elastic Instability Stress, Ste                 (ksi) 50.884   7.288 6.857 * (0.800/0.776)A2 = 7.288
                  *** STRESS ANALYSIS RESULTS 7         Applied meridional Compressive Stress, Sm                   (ksi)  6.982   5.588 8         Applied Circumferential Tensile Stress, Sc                  (ksi)  4.120   3.300
                  *** CAPACITY REDUCTION FACTOR CALCULATION 9         Capacity Reduction Factor, ALPHAI                                  0.207 10       Circumferential Stress Equivalent Pressure, Peq             (psi) 15.697 11       'X' Parameter, X- (Peq/8E) (d/t)A2                                 0.073 12       Delta C (From Figure - )                                           0.063 13       Modified Capacity Reduction Factor, ALPHA,i,mod                    0.311 14       Reduced Elastic Instability Stress, Se                      (ksi) 15.824    2.266
                  *** PLASTICITY REDUCTION FACTOR CALCULATION 0         15       Yield Stress Ratio, DELTA-Se/Sy                                    0.416 0I--      16       Plasticity Reduction Factor, NUi                                   1.000 17        Inelastic Instability Stress, Si - NUi x Se                (ksi) 15.824    2.266 0

0 0 ALLOWABLE COMPRESSIVE STRESS CALCULATION 0 18 Allowable Compressive Stress, Sall - SI/FS (ksi) 7.912 1.133 0 19 Compressive Stress Margin, M=(Sall/Sm -1) x 100% 13.3 0 0)

Ur U r _ C7 C7- rý r- r, r r r- U r- r r r-RMIIuclear Calculation Sheet CALCULATION OF BUCKLING MARGIN - REFUELING CASE, NO SAND GPUN EVALUATION FOR UNIFORM THICKNESS t-0.850 Inch USING THICKNESS RATIO LOAD ITEM PARAMETER UNITS VALUE FACTOR

               *** DRYWELL GEOMETRY AND MATERIALS 1         Sphere Radius, R                                            (in.)         420 2         Sphere Thickness, t                                         (in.)       0.850 3         Material Yield Strength, Sy                                 (ksi)          38 4         Material Modolus of Elasticity, E                           (ksi)       29600 5         Factor of Safety, FS                                                         2
               *** BUCKLING ANALYSIS RESULTS 6         Theoretical Elastic Instability Stress, Ste                 (ksi)      54.063     8.227 6.857 * (0.800/0.776)A2 = 7.288
               *** STRESS ANALYSIS RESULTS 7         Applied meridional Compressive Stress, Sm                   (ksi)       6.571     5.588 8         Applied Circumferential Tensile Stress, Sc                  (ksi)       3.878     3.300
               *** CAPACITY REDUCTION FACTOR CALCULATION 9         Capacity Reduction Factor, ALPHAI                                       0.207 10        Circumferential Stress Equivalent Pressure, Peq             (psi)      15.697 11        'X' Parameter, X- (Peq/8E) (d/t)A2                                      0.065 12        Delta C (From Figure - )                                                0.057 13        Modified Capacity Reduction Factor, ALPHAi,mod                          0.300 14        Reduced Elastic Instability Stress, Se                      (ksi)      16.257     2.474
               *** PLASTICITY REDUCTION FACTOR CALCULATION

.00 15 Yield Stress Ratio, DELTA-Se/Sy 0.428 16 Plasticity Reduction Factor, NUI 1.000 17 Inelastic Instability Stress, Si - NUi x Se (ksi) 16.257 2.474 J"- 0C) 04 ALLOWABLE COMPRESSIVE STRESS CALCULATION 18 Allowable Compressive Stress, Sall - SI/FS (ksi) 8.128 1.237 0 0- 19 Compressive Stress Margin, M=(Sall/Sm -1)x 100% 23.7

0I-ilNuolear Calculation Sheet APPENDIX C PICTURES SHOWING CONDITION OF THE DRYWELL IN THE SANDBED REGION OCLROO000408

Ld W jNucIear Calculation Sheet Subject O.nnrwSanEt.dbPeauain C-1302-187-5320-024 Cd~c No. iRev. INo 0 Sheet 47 NO. of 54 Originator t Dat Reviewed by Dat MARK YF*TKA 01/12/93 S. C. TumminclW 04/11/93 I Sand Bed Region - Typical condition found on inita: entry. Corrosion product on drywell vessel OCLROO000409

9'diJIuclear Calculation Sheet qtilhjecl COic NO. Rev. No. SheerI No. ut Evaluation Cn~d~

                                                               -1102-187-5320-074     1      0,         AS4 of     5 n r flrvwoll Et Date griinator

[ARK YVI.cKA IDate 01/12/93 1Reviewed S.byC. Tumminelli 01/16193_ Bay #13 - DNW shell showing plug. The plug is located in the middle of the worst cor-roded area of tne shell The plug showed no sign of corrosion. Bay #13 - DAN shell showed less prominent "Tub Ring" than what was seen in other I OCLROO00041 0

0JNuclear wrt7*+ M

• S, k ~

Bay #1 - Looking at the worst corroded area on shell near jent tube collar/ring. The ground spots seen here correspond to UT spot 20!21 '2.'3 Bay #13 - Lower Mid portion of the DAN shell showing UT spot 5.6 and 10. This close up photo shows the roughness of the corroded surface and how each UT spot has been picked up in the deep valleys thereby biasing the remaining wall readings to the con-servative sitIP OCLROO000411

r C7. r -7 I IF 1r lr- Ir ir Ur 7 r- - U r r C

t. " a o.4 zC 0

0 a) rt ml A"8 0 0 U n 0 I~)

                                                                                                                                      '0 w

n S. ra H C N).:* 0 o Bay #13 - Looking towards Bay#11 - Upper right cornero f DMW shell. Note (1)- Grinding depth on UT spot #1 & 2, (M - A part o- '-S Bat; Tub Ring" as delineated by marking and Q location~s of UT spots 3,4,13 & 17. The photo on right (although blurred by flash o reflection) shows 1/8" projection of plug. C, 0 0 C' -Ii 0 '0 31 IS

P0U1Nuclear Calculation Sheet Bay #!5 Looking towards Bay#17 which has been closed with foam for coating work

n Bay #17. Note the typical surface of the DW.shell --!!Id Innrli7ePd corroded spot A

                                                                              -    '
                                                                       .   -

Bay #13 - Looking toward Bay #15 - Lower left corner showing UT spot #7,12 & 16. This close up has captured the peaks and valleys of the corroded shell in vivid detail. Later NDE insoectinn revealid ne;th between peaks and valleys in the 0.25" - 0.40" OCLROO000413

Ii r- r- r r- r I- E77 U U_ r,- r-. _ r" r7V r r- r- W C Z

                                                                                                                                    >          *2j z

0 w p.'

                                                                                                                                       'S 0

0 I

                                                                                                                                    *0 Ca a         =5.

0 S. 0 0 a. (A 0 .

                                                                                                                                           '3
                                                                                                                                           -J "Ii 0
                                                                                                                                           ".3
                                                                                                                                           '3
                                                                                                                                           '3 0

0 0 0 Bay #15 - Note the original lead primer on vent tube OD Bay-,115 Looking toward Bay #13 showing portions of I) X

c D/W shell and concrete floor, after removal of loose debris 0 0 sturface. The "Tub Ring" was less prominent on the shell in J1z 00 / sand / rust. The concrete floor in this bay is one of the '39 0

this bay except a portion in lower left corner. Also note 0C)~ presence of lead primer on vent collar/ring plate. better ones. However - Note C) no drainage channel and 0* 0 O cratered holes near shell corner. '0 Ca J1 I..

EfilflNualear Calculation Sheet Subject Carc No. Rev. No. Sheet No. 1Vu1natinn in Sandbecd C-1302-187-5320-024 0 53 of 54 Originaior Date Revicwed by Date MARK YFKT*A 01/12!/93 S. C. Tumminclli 04/16/93 or, Wie Bay #13 - Looking toward Bay #11 - Lower right corner of D/W shell showing UT spots 9, 10. 18 & 19 Note the location of these spots - all are located in the valleys of the cor-roded surface This photo also shows the condition of the concrete floor. It appears Bay #13 - Looking toward Bay #15 - This photo captures the concrete floor condition and a oortion of lower shell corroded surface in very great detail. The floor in this area OCLROO000415

0ýNuclear Calculation Sheet Finished floor, vessel with two top coats - caulking material applied. Drain after floor has been refurbished OCLROO000416

Citizen's Exhibit NC4 izen's Exhibit NC4 Official Transcript of Proceedings Cit NUCLEAR REGULATORY COMMISSION Title: Oyster Creek Generating Station License Renewal Docket Number: (05000219) Location: Rockville, Maryland Date: Thursday, June 1,2006 Work Order No.: NRC-1087 Pages 1-129 It NEAL R. GROSS AND CO., INC Court Reporters and Transcribers 1323 Rhode Island Avenue, N.W. Washington, D.C. 20005 (202) 2344433 .

I 1 1 UNITED STATES OF AMERICA 2 NUCLEAR REGULATORY COMMISSION 3 ++++ + 4 CATEGORY 1 PUBLIC MEETING 5 BETWEEN 6 U.S. Nuclear Regulatory Commission 7 AND 8 AmerGen Energy, LLC, 9 Applicant for Oyster Creek Generating Station 10 License Renewal 11 12 THURSDAY, 13 JUNE 1, 2006 14 +++++ 15 The meeting was convened in the 16 Commissioners' Conference Room in One White Flint 17 North, 11555 Rockville Pike, Rockville, Maryland, at 18 9:20 a.m., Donnie Ashley, Presiding Official. 19 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202)

    . o 234-4433          WASHINGTON, D.C. 20005-3701           (202) 234-4433

2 1 =RC PERSONNEL PRESENT: 2 DONNIE ASHLEY 3 HANS ASHAR 4 FRANK GILLESPIE 5 REBECCA KARAS 6 P.T. KUO 7 LOUISE LUND 8 AMERGEN AND EXELON PERSONNEL PRESENT: 9 MICHAEL GALLAGHER 10 JOHN HUFNAGEL 11 AHMED OUAOU 12 FRED POLASKI 13 HOWIE RAY 14 PETER TAMBURRO 15 DONALD WARFEL 16 ALSO PRESENT: 17 KYOTO TANABE, Japan NRC 18 19 20 21 C-O-N-T-E-N-T-S 22 AGENDA ITEM PAGE 23 INTRODUCTIONS 3 24 USE OF ASME CODE SECTION 3 SECTION NE-3213.10 25 FOR LOCALIZED CORROSION AREAS 24 26 VALIDATION OF UT MbEASUREMENTS AND 27 BUCKLING ANALYSIS 34 28 USE OF ASME CODE CASE 284-1 45 29 ULTRASONIC TESTING ISSUES 52 30 INSPECTION INCREMENTS WITH UT COMMITMENT 69 31 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433

3 1 P-R-O-C-E-E-D-I-N-G-S 2 (9:23 a.m.) 3 MR. ASHLEY: Okay, I'm going to go ahead 4 and get started now. The other two participants can 5 just call in when they can. 6 This is a public meeting between the NRC 7 and AmerGen who is the applicant for Oyster Creek 8 license renewal. 9 It's a category one meeting. We will 10 conduct the meeting according to the agenda. At the 11 end of the meeting we will give those people on the 12 phone line and also the folks that are here at 13 headquarters and opportunity to make comments or ask 14 questions of the staff. 15 This meeting is being transcribed, and as 16 a result, if When you make your statements or you make 17 your presentations, please state your name and who you 18 represent go that the recorder can pick that up for 19 you. 20 Rather than introducing everybody in the 21 room, probably have maybe 25 or 30 people here, I just 22 want to introduce the participants here today. 23 And we'll start with our folks, and then 24 we'll give it to you, Mr. Gallagher. 25 Dr. Kuo. 26 MR. KUO: P.T. Kuo, division of license 27: renewal. 28 MS. LUND: I'm Louise Lund, a branch chief 29 in the division of license renewal. 30 MR. ASHLEY: My name is Donnie Ashley. I Im 31 the project manager for Oyster Creek license renewal 32 project. 33 MR. GILLESPIE: Frank Gillespie, director, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W." (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

4 1 division license renewal. 2 MR. ASHAR: Hans Ashar, NRC. 3 MS. KARAS: Becky Karas. I'm the chief of 4 the GSI and civil engineering branch in the division 5 of engineering. 6 MR. GALLAGHER: Okay, Frank. Can you hear 7 me? 8 I'm Mike Gallagher. I'm the vice 9 president of license renewal for AmerGen and Exelon. 10 And I'll turn it over to our team to introduce 11 themselves. 12 MR. TAMBURRO: I'm Peter Tamburro. I'm 13 senior mechanical engineer at Oyster Creek. 14 MR. OUAOU: My name is Ahmed Ouaou. I 'm a 15 civil structural engineer at Oyster Creek. 16 MR. RAY: My name is Howie Ray, and I'm at 17 Oyster Creek, the new manager. 18 MR. POLASKI: For Polaski, Exelon's license 19 renewal manager. 20 MR. WARFEL: Don Warfel, the technical lead 21 for the Oyster Creek project. 22 MR. HUFNAGEL: John Hufnager, the licensing 23 lead for the Oyster Creek project. 24 MR. ASHLEY: Thank you very much. We' ll go 25 ahead and get started with the agenda. 26 We have a very focused agenda today. But 27 first of all, before I get started into the agenda, we 28 really appreciate having the opportunity to meet here 29 at this commissioners' conference room. It's not 30 often we get such nice facilities to meet in. 31 This particular meeting is the-first of 32 two meetings that will be conducted. The next meeting 33 is tentatively scheduled for June the 22nd, and I NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

S 1 think we're going to try to do that in the afternoon 2 so you folks won't have to either drive down early on 3 95 or the night before. 4 I'd like to welcome everyone again. We 5 have participants here from the State of New Jersey, 6 from Region 1, and Kyoto Tanabe from the Japanese NRC, 7 NISA. 8 And of course the people that are on the 9 phone line with us. 10 We're going to talk about two concerns 11 with you, and they're going to be very focused, and 12 we're not looking for answers from you. 13 John Hufnagel and I have done this sort of 14 thing many times since the license application was 15 received in July of 2005. Since that time we've had 16 three onsite audits. We've had regional inspection. 17 And I believe that Roy Matthew, the audit team leader, 18 is here. And he is still working on the audit report 19 and the inspection report. 20 The next step in this process as we go 21 through is collecting all the information that we have 22 garnered over these months in preparing the safety and 23 evaluation report. 24 Part of that has involved the request for 25 additional information. To date, we've processed 26 about 128 questions, plus or minus a few. 27 Normally, that's the reason I mentioned 28 John Hufnagel, we usually do these requests for 29 additional information on the phone. And the way that 30 we do that is, we have a discussion of what our 31 concerns are, and make sure that you understand what 32 our concerns are. 33 When we do the phone calls and meetings NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

6 6 1 for the request for additional information, we're not 2 looking for the answers at this time. Just that you 3 understand where we're coming from. 4 In addition, there are several hundred 5 questions that are in the Q&A database, that Matthew 6 and his team and your team put together during the 7 audits that are publicly available in the Adams 8 (phonetic). 9 The next thing that we do after we have 10 the meetings is, we're going to look for your 11 responses. And then eventually we're going to process 12 the safety evaluation report, and hopefully there 13 won't be any open items. Right now there are probably 14 some open items there that we still have to follow up 15 on. This meeting is going to address some of those. 16 Because the thing that we have to make -17 sure of is that we have reasonable assurance that all 18 of your assumptions and all of your calculations and 19 all of the programs that you've put in place will be 20 valid for the period of the extended operation. 21 So with that, I'm going to turn it over to 22 Frank Gillespie, and we' Ill go ahead and get started on 23 the discussions. 24 MR. GILLESPIE: Okay, thank you, Donnie. 25 The way.we've organized this is, I'm going 26 to present kind of a bulletized issue, and kind of a 27 broader context. And on those issues where we think 28 we really need to give you some detail on what our 29 issue is, then I'm going to turn to Hans. And we've 30 already coordinated between Hans and I. He knows 31 which issues. And we've got some notes developed on 32 it. And as we go through the meeting, we'll do our 33 best, then potentially, to.put out in a rapid way some NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON. D.C. 200054701 (202) 234-4433

6W ~ 1 meeting minutes to share the notes.7 2 And one of our concerns is, you've been an 3 applicant who has been very responsive to our requests 4 for RAle. You have always made it on time. And on 5 the particular issue we're going to address today, you 6 almost overwhelmed us with information. 7 And so Hans in the last two weeks, and we 8 actually had Noel Dudley helping Donnie and Hans, to 9 try to take this large volume of information and say, 10 okay, here are our original questions; map that 11 information into the original questions and say, 12 what's the residual? 13 And the residual were such - and this is 14 an interesting comment, and I have to thank Mitzi, 15 who's from our general counsel, who put this in 16 perspective when we were kind of going through this

    *17  for me, she said, gee, this discussion is far more 18  focussed and detailed than the way we traditionally 19  write RAls.

20 And from that I said, really, we need to

21. sit face to face, because they are very very specific 22 things that we need blanks filled in, and they are all 23 very technically oriented. And we probably wouldn't 24 do justice in kind of randomizing, because you'Id focus 25 on answering the RAI, as opposed to maybe 26 understanding the underlying concern it causes us to 27 write it.

28 And so that's when I proposed this set of 29 meetings. And so therefore I think today is your 30 opportunity to pummel us with, questions to ensure 31 complete understanding of the-Rhls. 32 It is not our opportunity to request from 33 you an answer to them, because 3Ireally think you need NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

I 8 1 time to take - it's very detailed information, and 2 it's the kind of thing you need engineers back at the 3 plant, I think, probably to mull over and look at. 4 So I do thank you for your responsiveness, 5 and you overwhelmed us with information. We've gone 6 through it all. And it really has narrowed it down. 7 So now I'm going to go through the 8 bullets, and on specific ones, you're going to see me 9 turning to Hans. But I'Im going to go through each 10 bullet and ask you, do you-have any questions. 11 If you don't, then I'm turning to Hans, 12 and he's going to go through the specific details. 13 And some of them are more straightforward, and we can 14 go through even quickly. 15 As Donnie said, we narrowed our focus down 16 to two concerns, and both of these concerns with I 17 -want to call uncertainty. So we are not making a 18 judgement as to adequacy at this time. or in anyway 19 absolute. I'm going to suggest that much of what 20 we're talking about deals with the uncertainly in the 21 information, because of some voids in the information 22 that have to come in. 23 so we have dry wall corrosion 24 uncertainties. And then we have some ultrasonic 25 testing issues. And there are two subsets to the 26 ultrasonic testing issues. One is testing in the 27 upper portion, which is really a pressure retention 28 question, and then there is some questions on the 29 lower portion. 30 And if that doesn'It come out clear as I go 31 through the bullets, you need to ask us about that. 32 Because there are two different kinds of points in 33 there, and we're trying to leave this meeting with no NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

9 1 confusion on any part. Because these are really fine 2 tuning now, kind of issues that we're in. 3 With that, let me - 4 MR. GALLAGHER: And Frank, just so you 5 know, we have Pete and Ahmed and Howie, and they are 6 like our experts on this issue. So they provided a 7 lot of that information that you're talking about that 8 we provided. So if we really have any detailed 9 questions, these will be the three individuals - 10 MR. GILLESPIE: But again, I want you 11 asking us questions today, as opposed to putting you 12 in the awkward position of thinking you need to 13 respond. 14 And so as Donnie said, there's no reason 15 to respond today. 16 MR. GALLAGHER: Right, and we really 17 appreciate that, Frank, to make sure we really 18 understand the issues. 19 MR. GILLESPIE: So let me start off. 20 Dry wall corrosion uncertainties. There 21 were assumptions in the 1991 GE report. Within these 22 assumptions there's uncertainties in the simulations 23 of degradation calculations in the associated 24 ,analyses. This is not a your action, this is an our 25 action. I just want to let you know that as part of 26 our review, we may be doing an independent calculation 27 or something to reinforce the assumptions in the basic 28 calculation itself. 29 So that's not an action for you; that's 30 really an action for the staff. And you might say 31 it's by way of almost what we do in other areas of 32 thermal hydraulics and other things, where we' ll do a 33 confirmatory calculation. NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

10 1 But so you know it' s happening, it ' s going 2 on, we may be coming back to you for data in support 3 of that calculation later. 4 Uncertainties in ultrasonic testing 5 results: There are two key issues here that Hans is 6 going to go over in a little more detail. 7 One, there's a report that's referenced 8 that has a disclaimer in it. And the disclaimer says 9 something like, no one at GE or AmerGen can be held 10 responsible for the accuracy of this report. 11 It sounds like a boilerplate disclaimer. 12 But nonetheless, it's kind of - again, we're really 13 fine tuning, so I'm being very specific here. The 14 disclaimer raises issues of, well, do you believe the 15 report you referenced. 16 The second piece, .which is probably more 17 important with this, and now I'm going to turn to H 18 ans on this, is what I'm going to call the evaluation 19 of the grid data itself. 20 And so I've broken this one down into two 21 things: the disclaimer, which I'd like to 22 administratively have you - ask you to please look at 23 it and deal with it. Does it still stand? 24 And now I'm going to turn to Hans on the 25 grid data evaluation. 26 Hans. 27 MR. GALLAGHER: Yes, okay, and Frank, is 28 this in the lower portion? Because you broke it up 29 into upper .and lower? 30 MR. GILLESPIE: In general. 31 MR. ASHAR: In general. 32 Let me narrate what I have written, 33 because this is transcribed; I cannot be informed NEAL R.GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE, N.W. (202) 234-4433 WASHINGTON, D.C. 20005-701 (202)234-4433

1 about it. 2 So attachment I-A of the GPU letter dated 3 November 26th, 1990, makes a statistical evaluation of 4 the UT measurements data taken up to 1990. 5 On the cover page of the report, GPU 6 Nuclear Corporation states in their disclaimer: The 7 work is conducted by individuals for use by GPU. 8 Neither GPU nor the authors of the report warrant that 9 the report is complete or accurate. 10 In view of this disclaimer, the applicant 11 is requested to provide a detailed description of the 12 way the UT measurements data, whether taken as part of 13 the 6X6 grid, or isolated readings, were evaluated and 14 used in performing the analysis. 15 Do you understand? 16 MR. GALLAGHER: I don't.

17. MR. GILLESPIE: We're open for questions.

18 MR. ASHAR: Regarding the clarity of the 19 question. 20 MR. GALLAGHER: Okay. What was the report 21 name again? 22 MR. ASHAR: It is attachment 1-A to the GPU 23 letter dated November 26th, 1990, which is a 24 statistical inference of the UT data. 25 MR. GALLAGHER: Okay. And your question is 26 related to the disclaimer itself. 27 MR. ASHAR: Disclaimer, which is in 1-A has 28 been- used, or if something different is used, what 29 kind of confidence level has been used. Because that 30 particular report talks about the mean and confidence 31 level.. But whether it is used effectively all the 32 time, we have no idea. 33 Because we looked at the report. We NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

12 1 didn't interpret report. 2 MR. ASHAR: Let me see if I can - this is 3 the kind of dialogue we need to have. Because we' re 4 down in the details, really, now. Okay? 5 MR. GALLAGHER: Yeah, this is pretty 6 focused. 7 MR. GILLESPIE: And Hans and I have spent 8 a lot of time together in the last week or so. So let 9 me say it and then ask him if I've said it right. 10 Basically in the measurements that were 11 taken it's a scanned area with a grid array of 12 measurements. And it's not clear whether that's a 95 13 percent confidence interval, is it a median value? 14 It's not clear how those area level calculations were 15 used. And much of the information that we used in the 16 1991 report, and in fact in the graphs that you sent . 17. in in your RAIs literally takes the result of this 18: calculation. 19 But there is really no description of what 20 - how these data points were combined. 21 MR. GALLAGHER: This would be to determine 22 the average thickness for the - 23 MR. GILLESPIE: Thickness data, yes, and 24 the projection of that thickness data as applicable to 25 the liner. 26 MR. GALLAGHER: Now I thought we had 27 provided a description of that in. one of our Q&As. 28 Ahmed or Pete, do you guys recall that? 29 MR. OUAOU: My name is Ahmed Ouaou. I'm 30 with Exelon Oyster Creek. 31 We did - this question came up, we did 32 provide response to the question, what type of 33 statistical analysis did you do. NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

13 1 It's also in the RAI, and it's part of the 2 - that's why we submit that report. 3 MR. GILLESPIE: So is there anything beyond 4 that that we provided? 5 MR. OUAOU: Well, my main question is, 6 have you used that particular report? Or you use 7 something different? 8 MR. TAMBURRO: My name is Pete Tamburro. 9 The attachment one to that letter, is there a document 10 number. 11 MR. ASHAR: Yeah, I think you sent to me. 12 It came to us. 13 MR. GALLAGHER: This is in your 14 application. 15 MR. OUAOU: Again, this is Ahmed. We just 16 -don't recall what Attachment 1-A is. And what that 17 is. 18 MR. ASHAR: I think title is statistical 19 inference. 20 MR. GALLAGHER: That calculation was 21 submitted previously. It's part of the original 22 approach that was developed to calculate the average 23 thickness in thinned areas, submitted back in 1991. 24 But that's the calculation we use, and 25 Pete can talk about that. 26 MR. TAMBURRO: This is Pete Tamburro. 27 The words that you reference about no 28 claims made by the author, that sounds like a 29 technical data report which describes methodology. 30 It's not intended to be a calculation. 31 So-I think we owe it to you to go do the 32 research and see what the intent of that report was. 33 I believe we later did calculations which normally NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) I g 234-4433 WASHINGTON. D.C. 20005-3701 (202) i . 234-4433

14 1 treated the data. 2 MR. ASHAR: This is what we would like to 3 know, that is the question, basically. 4 MR. GALLAGHER: So specifically about that 5 report, how we arrive at our statistical analysis. 6 MR. ASHAR: And what actually you used. 7 MR. GALLAGHER: And what we used, okay. 8 And I just want to make sure, because I think we 9 provided a lot of that. So I want to make sure we 10 don't just provide the same information, and we're 11 missing something. 12 So if it's just that we can make sure we 13 sharpen our response 14 MR. GILLESPIE: Yes, we're trying to be 15 very - this is really a very incremental meeting. 16 We're really trying to deal with the piece that we 17 don't feel that we have. 18 And right now if this is the grid, and you 19 take a six by six measurement - 20 MR. ASHAR: There's 49 probes in it. 21 MR. GILLESPIE: There's 49 probes. I 22 think, Hans, a fundamental question was, but you come 23 up with a single point that is than used in the next 24 level of calculation. We're not pushing the next 25 level of calculation; what we're doing is saying, how 26 was that point come up with? Was it a 95 percent? 27 There's a number of ways that are actually all valid 28 to do it. Was it the median of the 49 measurements? 29 Was it a 95 percent confidence level? How were those 30 49 points combined to get to the one point which was 31 than used at the next calculational level. 32 .. And by the way if there is anything that 33 you want to actually respond to in writing like, we NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE, N.W. (202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433

1-5 1 really didn't understand that. We think we answered 2 it in response this, this and this, following this 3 meeting, feel free to send us that. 4 MR. GALLAGHER: Okay. 5 MR. GILLESPIE: That'Is quite - because what 6 I really want to do is, this is a starting point, to 7 get clarity in every one of these details. Because I 8 think we are down in the details. I will fully 9 concede we are really fine tuning it. 10 MR. GALLAGHER: And that's what I was 11 getting at in how that 49 point array, how the 12 statistical analysis is done. I think we've provided 13 that answer. We can look at it. 14 And then about the disclaimer, we can 15 specifically talk about that. Because I think like

16. Pete said, the intent on that was, the data was taken 17 in the field,-and that was validated. And that data 18 was used in this analysis.

19 .And it's just saying, we didn't take the 20 data. All we did was do a statistical analysis. 21 MR. GILLESPIE: So just make it clear on 22 how that report was used, and then we're kind of okay 23 there. 24 And if you write us a letter and it says, 25 in reply, in reply RAI this, we think we've answered 26 this specific question, that would allow us to reply 27 back, no, here's the specifics of what is missing in 28 that.. And that's a perfectly - I mean that's all part 29 of the process. 30 Believe me, you flooded us with so much 31 information, could we have missed something? Yes. 32 And that's okay. 33 MR. GALLAGHER: We wouldn't have talked NEAL R GOSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-43 WASHINGTON. D.C. 20005-3701 (202) 234-4433

16 1 about that disclaimer. 2 Pete, did you have something? 3 MR. TAMBURRO: Yes, I just wanted to make 4 sure I understood a point you made. You would like a 5 description of how we started with the 49 points and 6 came up with a representative value for those 49 7 points? 8 MR. GILLESPIE: I believe that's the point 9 that Hans and I - Hans, why don't you. 10 MR. ASHAR: Let me explain. 11 I think some of the readings that you have 12 taken are based on the grid. Some of the measurements 13 you might have taken isolated away from the grid, or 14 may not have used the grid. I'm not sure what was 15 done where. 16 But that doesn't matter. The important: 17 thing is how you really used this data in coming out 18 with. the final thicknesses at those points, that is 19 important. 20 MR. GALLAGHER: For the grids. 21 MR. ASHAR: For the grids, yes. 22 MR. GILLESPIE: So what we're seeing is a 23 layering in this calculational process, where you 24 start with raw data and then you do one thing to that, 25 and then you do the next thing. 26 And we're down at the real fine tuned 27 question here at the bottom. And it's that detail 28 that we're not sure that we have. 29 Now it might have been submitted -in 1990; 30 I give you that. Could you repackage it and get it 31 back to us? 32 It may be easier for you to do that than 33 for us to do it again. NEAL F. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

17 1 MR. ASHAR: I might say that you might have 2 even provided some description as a result of the 3 audit. 4 MR. GALLAGHER: Yes, that's what I was just 5 referring to. 6 MR. ASHAR: I understand. I did not have 7 a chance to see everything that they have acquired and 8 have responded to. 9 MR. GALLAGHER: Oh, okay. 10 MR. ASHAR: Not all. I am aware of it, 12 most of them, the basic things. But I did not see 12 anything related to this one. But if it is there, 13 Just give an answer. 14 MR. GALLAGHER: That would be helpful, 15 because we can pinpoint it, and then we can go from 16 there. 17 MR. ASHAR: But to me, it looked like at 18 least in 1990 it appears that this particular. report 19 was used, and to what extent it was used is not quite 20 clear. How does it relate to what you did, and 21 responded to as a part of the AMP questions, I don't 22 know. 23 MR. GALLAGHER: Right, okay. 24 Now, Pete, -Ahmed, you guys okay with 25 understanding that, Howie? 26 MR. MUAOU: I understand the question. 27 This is Ahmed again with Exelon. Part of that 28 response was provided in the RAI and in the questions. 29 So we'll go back and take a look specifically and look 30 at that concern. 31 But we' re not providing a response to you. 32 MR. GILLESPIE: No, no, again, it's 33 perfectly acceptable for you to say, go back, go back NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433

18 1 to New Jersey and say, you know what, we understand 2 your concern, and we think we addressed this is in 3 these RAIs, and the RAIs we have completed reviewed, 4 in fact with multiple people. And if it makes sense 5 that we're looking at how these 49 points - and there 6 may be points that weren't grid points, that weren't 7 49 points, that were individual measurements, or maybe 8 smaller samples. It's not clear that they were all 9 uniformly 49 points, they were all uniform grids. 10 That level of detail was not necessarily seen. 11 And I will say that we are trying to get 12 you this information as we're putting our draft SE 13 together so we can get these issues closed out. 14 MR. OUAOU: Again, this is Ahmed again, 15 the reason I was kind of, I guess, thinking a little 16 bit, it's such a straightforward question, we can 17 answer that today. 18 MR. GILLESPIE: Again, my promise was, we 19 have a second meeting scheduled, and we' d really like 20 to get it in writing before that meeting so we can 21 have a substantive meeting. 22 It was important for me, because of the 23 detailed nature of our concerns, to get them to you 24 and make sure you understood them. And I didn't say 25 they were hard to answer. So just because we have a 26 concern doesn't mean it's difficult to answer. What 27 we wanted to do was get this kind of detail to you so 28 you could answer it, and that was the important 29 aspect. 30 MR. GALLAGHER: I -think we understand 31 that. 32 USE OF ASME CODE SECTION 3 SECTION NE-3213.10 FOR 33 LOCALIZED CORROSION AREAS NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 ('2M) 234-4433

19 1 MR. GILLESPIE: The next point was use of 2 ASME Code Section III, Section NE-3213.10 - now you 3 know why I say we'll publish the meeting minutes to 4 this early - was used for localized corrosion areas. 5 And this is a comment that is also going to come up 6 later. 7 And by the way, this is all dealing with 8 the 1991 GE report. So it's not that you used it; 9 it's that it was used in the 1991 GE report. In 10 general that code was written for and applicable to 11 new containment shells. And the methodology for the 12 buckling calculation, it's not clear its applicability 13 to a shell that's actually older and has corrosion. 14 And I 'm going to get Hans to amplify that, 15 but in my simplistic terms - I get to be the non-16 engineering interpreter, and. he gets to put the 17 details on it. - if things corrode in a manner that's 18 pitting or discontinuous, and you have a shape that is 19 much different than the discontinuity from two 20 different sized-plates. 21 And so this code was specifically 22 developed for one purpose. That doesn' t mean it' s 23 wrong to use it for this purpose; what it means is, 24 the transition to using it for this purpose wasn't 25 included in the 1991 GE report. 26 Now with that I'm going to turn it over to 27 Hans. 28 MR. ASHAR: Let me just narrate the way I 29 formulated the question. 30 For the localized thin areas, the 31 applicant is using the provision of Section 3213. 10 of 32 the subsection NE of Section III of the ASME Code. 33 This provision, though not directly applicable to the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

20 1 randomly thin areas caused by corrosion, if used with 2 care and adequate conservatism, may provide some idea 3 about the primary stress levels at the junction of the 4 thin and thick areas. The applicant is requested to 5 provide a summary of the process used and to address 6 this issue. 7 MR. GALLAGHER: In this particular 8 analysis, I note that that particular question was 9 looked at earlier when the analysis was originally 10 reviewed and approved. 11 And I think, did we have discussion about 12 that in the Q&As? 13 MR. OUAOU: This is Ahmed Ouaou again with 14 Exelon. There was a discussion in the Q&As on the 15 issue - on the concern. :We spent a lot of time with 16 the audit team talking about the calculation in 17 particular, and it was reviewed by the audit team. 18 This. same question that you have a concern 19 was asked - again, I have to go back a little bit, 20 because. I spent a lot of time looking at the history 21 on this - the same exact question came back in '91, 22 and we -,there was a formal report that was generated 23 and submitted to address the question. It was done by 24 Teledyne; it's not the GE report. It's in response to 25 an RAI. 26 Our understanding is that after review of 27 the calculation at the site that it appeared the 28 approach was reasonable that it should not be a 29 concern from a stress concentration perspective. 30 MR. GALLAGHER:. And this review was done 31 when the 1991 account was generated? 32 MR. OUAOU. That's correct. 33 MR. GALLAGHER: So what we were looking at, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) _ o 234-4433

21 1 we didn' t see how there was any aging management 2 related effect on the differences between the way it 3 was evaluated before and when it was evaluated now. 4 MR. ASHAR: Let me again restate. 5 The question is, this particular provision 6 in the ASME code is not written for the localized 7 corroded areas. It has been used here between thin 8 and thicker parts to justify the use of, you know, in 9 a particular way. 10 Now I can see that there is no other way 11 you can do that except to use this type of provisions. 12 But I want to understand what kind of conservatism you 13 have used. 14 Because there are a number of items 15 related to this provision that are in the ASME code. 16 For example, for primary membrane stress there is one 17 particular areas where you go up to the square root of 18 RT; for the secondary stress, you go to the 2.5 square 19 root RT, and figure it out as to, now, I want to make 20 sure that you have considered representation of the 21 thin areas in this particular process. 22 MR. GALLAGHER: I think that's helpful, 23 Hans. Because we didn't identify any other specific 24 method to use, .other than use this. And then there 25 were some I guess checks that was done for, like, one 26 thing Frank mentioned was about the plate changes; 27 that was one check. Another check was done as far as 28 a one-by-one depression, a one-by-one-foot depression 29 in the shell; and then another would be a fairly 30 localized 2-1/2 inch depression. 31 So they were kind of checks that said, 32 they didn' t look like there was any significant impact 33 on the analysis. NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

22 1 The roughness looked like to us it'd be 2 more of a - maybe it related to a fatigue concern, 3 which really isn't an issue for the drywell. 4 So that's why that kind of a review, I 5 think, was done in the 1991 review and analysis, and 6 the staff had accepted that at that point. 7 And I don't think there's any methods 8 that's changed since. 9 MR. ASHAR: This is for the license renewal 10 we are talking about. So I understand that the staff 11 will issue a report based on certain things. 12 But we are looking at this in more depth. 13 And we want to understand the mechanism before we go 14 and say, hey, this is the reasonable assurance that 15 something would happen. 16 So you might see this as duplicative or 17: something, in your mind, but for us that information 18 is necessary to make that reasonableness estimate. 19 So even if you might have done something, 20 you might have responded to this type of question in 21 past, in 1992, 1993, I think we would like you to tell 22 us more about it. If you done it during the audit 23 team, please let us know about it. We can go and 24 check it out in the AMP's responses. There is nothing 25 - but I just want for you to understand that you 26 understand the question. 27 MR. GILLESPIE: Yeah, this is not to say 28 you haven't done it before, and it's part of 29 everything that happened from the mid-'80s through 30 "91. It wasn't reviewed by the staff for the purposes 31 of the current existing license. But this is a 32 question as part of the renewal review. 33 And if we're requesting you to repackage NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 2344433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

23 1 something and send it in as part of this, then that's 2 our request, because we're a second time dealing with 3 what I said was the uncertainties. We're not saying 4 anything negative about the GE calculation. What 5 we're saying is, we have a set of reviews reviewing 6 this for another 20 years on your license beyond, and 7 so this is an aging issue. 8 I mean we met with ACRS on this yesterday. 9 We're not saying you didn't do it 20 years ago. What 10 we're saying is, it's not really readily available to 11 the staff to be able to include it in their more 12 global judgment on the liner today. 13 So if pulling it out of your records and 14 getting the Teledyne report, if that's easy - I didn't 15 say we.were asking anything that was hard. I said we 16 were going. to try to give you our specific concerns.. 17  : MR. OUAOU: The Teledyne report -was not in 18 QA. But we can .provide the Teledyne report and 19 several correspondences to that address the question. 20 MR. GILLESPIE: That would be appreciated. 21 Remember our goal here is to answer the questions. 22 This is a bit collaborative in nature. 23 The other thing I have to ask your 24 forbearance in part of our idea of trying to stay on 25 a certain schedule is that things get done in 26 parallel. And the audit team is in the process of 27 writing a report, and the last I heard they were on 28 page 700. And they have to look at it in an 29 integrative way also. And that is one input to the 30 SE. 31 But that's input eventually to Hans. 32 Because Hans is the guy who on the line has to really 33 make the safety judgment on behalf of the agency. NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202)234-4433 WASHINGTON. D.C. 20005-701 (202) 234-4433

24 1 1 So I'm asking for your assistance. If 2 it's a bit of repackaging, or a resubmittal, this is 3 what's going to get the job done. 4 MR. GALLAGHER: Okay, so if we describe 5 this analysis, and what - how we think it was put 6 together to conservatively address some of these 7 issues, then we could do that and talk about the 8 Teledyne work. And I guess, I want to make sure, 9 Hans, do you have any other methodology that we should 10 be looking at? 11 MR. ASHAR: Well, yes, I think I'd refer to 12 one report which Sandia developed for big area of 13 containments. But I don't know to the extent to this 14 particular aspect, it addresses that area. 15 What it does is, it models certain 16 enclosures:and certain degradation in containments of 17 various types. It's a Mark I, Mark II, all 18 containments have been considered in those. 19 MR. GALLAGHER: Okay, that report is 20 available? We hadn't found that report, had we? 21 MR. ASHAR: I know. I'll try to get it. 22 MR. GALLAGHER: Can we get that today? 23 Because that would be real important to us. 24 2R. GILLESPIE: Yes, if we get the ML 25 number, since we're adjourning at lunch. 26 MR. ASHAR: Yes, we'll put it in ADAMS. 27 MR. GALLAGHER: That would be helpful, 28 because we can review that report. 29 MR. GILLESPIE: And it may.be as easy as 30 saying, here's what we're done. Here's this other 31 report that's a little newer.` And here's why we're 32 consistent with it, and why this makes sense. 33 But that's your judgment to do. We're NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

25 1 trying to give you our concern, and Hans is trying to 2 give you at least one reference that's available to 3 kind of the NRC sponsor which is kind of a benchmark. 4 And again, we're dealing with the 5 uncertainty of the information at a very fine level, 6 SO. 7 MR. OUAOU: Again, this is Ahmed Ouaou. 8 I just want to ask a question. 9 Do I understand you to say that that 10 report, the Sandia report, has a benchmark we should 11 be measuring against? 12 MR. ASHAR: I don't think so. The reason 13 I don't recommend that is because it is meant only for 14 internal reference. 15 MR. OUAOU: This is information? 16 MR. ASHAR: To the extent you can use iti 17 It is not something that is endorsed for use for 18 anybody. 19 MR. OUAOU: Do you know of any other 20 methodology that would take surface corrosion areas 21 that you're concerned with? 22 MR. ASHAR: No, I'm not aware of any. 23 MR. OUAOU: You're not aware of any? 24 Okay, thank you. 25 MR. GALLAGHER: And we had looked at it, 26 that stress ride issue looked like it was more of a 27 fatigue issue, and the containment fatigue really 28 isn't a concern. 29 MR. ASHAR: Well, containment for the ease 30 of concern in the area of events, right. 31 MR. GALLAGHER: Right. 32 MR. ASHAR: But away from there, you don't 33 have that concern. NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W. WASHINGTON. D.C. 20005-3701 (202) 234-4433

26 1 MR. GALLAGHER: That's correct. 2 MR. GILLESPIE: But again, this is not 3 saying you don't have the information on site. It's 4 only saying we don't have it in a form which we can 5 identify that it specifically addresses this question. 6 And so if you can help put that 7 information in a form that specifically addresses the 8 question - this is why I didn't want to get - this is 9 why I said, let's have a meeting, versus writing RAIs 10 where we didn't - have a total misinterpretation of 11 the RAIs. 12 MR. GALLAGHER: Yes, right. 13 MR. GILLESPIE: So again if you get back to 14 the site, and you want to email us, because emails are 15 on the record, and we try to keep everything on the 16 record, to get further amplification, that's 17 perfectly. And you know if .- you have thoughts when you 18 go-back, just say - you know. 19 MR. GALLAGHER: Okay, that's helpful, thank 20 you. 21 MR. GILLESPIE: We finished with - this was 22 really the assumptions in the 1991 GE report section. 23 And so there were really two bullets that we had in 24 summary. And that was, the first was the 25 uncertainties in ultrasonic testing results, and this 26 was the grid thing we talked about, and then the next 27 one. 28 And the first one was just for you to know 29 that we're going to do something of an independent 30 nature to verify the calculation. And that's not an 31 action on you, that's an action on us. 32 VALIDATION OF UT MEASUREMENTS AND BUCKLING ANALYSIS 33 The next major topic - and major doesn't NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 200043701 (202) 234-4433

27 1 2 mean important; major just means it's the next heading on my notes - is validation of UT measurements and I 3 buckling analysis. 4 In this I have three principal notes, and 5 let me just go through them. And the first note is, 6 UT results indicating increase in shell thickness. 7 A* -tnere was this, anomalous point. 8 And the anomalous point raises questions 9 that are probably unanswerable. So let me say in 10 retrospect, looking back, the answer to the specific 11 question might be unanswerable going back, but the 12 actions to be taken in the future might be very 13 doable. And that's questions on the accuracy of 14 measurements, the appropriateness of calibration, the 15 :.one point was significantly above the curve. 16

• So with that, let me turn that one over to 17 Hans, so he can go into details of that concern.

18 MR. ASHAR: Okay, I 'm going to narrate that 19 -again. 20 In the sand pocket region of a drywell 21 shell, the most susceptible base are incorporated into 22 assembly. However, there are a number of issues that 23 need to be addressed to ensure that the readings are 24 taken at the vulnerable locations and techniques used 25 are reliable. 26 I'm talking about the technique right now 27 first, and then I'm going to talk about the other 28 points. That will come with discussion of the other 29 bullets. 30 Review of table two indicates that the UT 31 measurements taken from inside the drywell after 1992 32 shows a general increase in the measurement taken from 33 inside the metal thickness. In some cases it NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. WASHINGTON, D.C. 20005-3701 (202) 234-4433

28 1 increases as much as 50 mils in a two-year time frame. 2 MR. GALLAGHER: What was that number? 3 MR. ASHAR: Fifty mils. 4 MR. GALLAGHER: Fifteen? 5 MR. ASHAR: Fifty, 5-0. 6 MR. GALLAGHER: Oh, 50. 7 MR. ASHAR: Fifty mils within a time frame 8 of two inspections, 1994 and 1996, I think. 9 In general it appears that the UT 10 measurements taken after 1992 requires proper 11 calibration considering the coatings on both sides of 12 the drywell shell. 13 The applicant is requested to address this 14 issue. 15 MR. GILLESPIE: Now, again, as I said, you 16 can't go back and fix what is.

17. MR. ASHAR: Well, Frank, I don't agree. I 18 think if the tests done outside on an epoxy-coated and 19 galvanized inside, and you've calibrated that, the 20 :readings taken earlier can be reduced to this.

21 It's possible to do it too to the existing 22 - but I don't know what you want to do. 23 MR. GILLESPIE: What you're saying is, if 24 they did some calibration samples, that had the proper 25 codings on either side, there may be the data 26 available in their records to go back and - 27 MR. ASHAR: Yes, compare what they have 28 done earlier with or without coatings, you know' that 29 kind of thing. 30 MR. OUAOU: Again, this is Ahmed with 31 Exelon. I was surprised, too, that those points wereq 32 as high as they are. We expected some variation 33 because of surface roughness, of the shell itself. NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON. D.C. 20005-3701 (202)

• o 234-4433

29 1 Although we use a template, and we use a 2 probe. If you just happen to move the probe just a 3 little bit you would get a different reading. 4 But in that particular '92, it appears 5 that one set of readings were consistently higher than 6 the rest. And we spent a lot of time trying to find 7 the cause that caused that, and talked to Rich Morante 8 at the site there during the audit review. And 9 frankly, what they came up with I'm not sure that's 10 satisfactory. 11 We just couldn't. Qualified people were 12 doing the testing, same methodology that was used 13 before. We haven't looked at the potential, because 14 there is a grease where you do UT measurements, 15 potentially, that might not have been removed. We 16 looked at all that, but really couldn't come up with 17 a specific answer why those values were higher. 18 MR. TAMBURRO: Going forward, the potential 19 items that.we've looked at, we're going to reduce or 20 eliminate them. For example the grease will be 21 removed prior to the inspections. We will do 22 calibrations, both on the external coating and the 23 internal coating, to get an understanding of how they 24 affect the measurements. 25 So we intend on reducing all those 26 potential variants out of the future inspections. 27 MR. GILLESPIE: The importance of this 28 issue may be one, the narrow technical issue itself. 29 And it's a good response. Didn't really need it, but 30 it was a good response. 31 But it does contribute'to the general 32 thought we have which we' ll get to later when we talk 33 about some of the commitments you made already on the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

II 30 1 level of uncertainty. And these things just 2 contribute to the level of uncertainty of the 3 measurement. 4 And I would use - maybe it's not 5 invalidating the measurement, but it's the uncertainty 6 involved with any individual measurement, and the 7 trend. 8 And uncertainty, like I said in the 9 beginning, is kind of what we're trying to reduce or 10 understand through all of these points. 11 And so I think the idea that you can only 12 make these measurements just so certain, and just - 13 let's just keep it - there's only so much you can do 14 with these kind of UT measurements. 15 But: this seemed to be a very large 16 uncertainty, in.fact much in exceedance of some of the 6 17 things you've actually measured in other areas. 18 relative to thickness changes. 19 So as-long as you understand our concern, 20 this - minimize the contribution of these to 21 uncertainty, and if you can't do anything about the 22 past one, you can't; you did this examination then. 23 But this contributes to some of the 24 thoughts we have relative to the 10-year commitment 25 that we'll talk about later. 26 MR. ASHAR: I feel that you will have to do 27 some kind of a comparative testing in order to, at 28 least for the future readings that you take is going 29 to influence that. 30 If that was the cause, because of a 31 coatings on two sides, this thing we have normal 32 readings that showed more thickness than the other 33 thicknesses, then I think it is something that you NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 2U-4.4= WASHINGTON, D.C. 20005-3701 =2o)234-4433

31 1 ought to look into it and come to grips with it. 2 MR. GALLAGHER: How many points did you 3 have a concern with, Hans, do you remember? 4 MR. ASHAR: Well, I just think in general. 5 Just like Ahmed said before, in general you can see 6 when you look at the readings that they are increasing 7 in 1996 compared to 1994. 1992 and 1994 are almost 8 same; they are not changing too much in general. But 9 there are a few places where it is about 30 mils 10 higher or 50 mils higher, like that, you know. 11 So there's an anomaly here, and that has 12 to be resolved. 13 MR. 'OUAOU: If I may just add, that we're 14 benchmarking other people doing the UT measurements in 15 the past, but that. was before 1996. For instance, GE 16 - GPO brought in GE to do some UT measurements. And 17 I don' t believe the methodology has changed in the way 18 we did it. 19 Whatever, we couldn't explain it. We 20 couldn't explain why. these particular points were that 21 much different than the previous UT measurements. 22 MR. GILLESPIE: Just keep the word 23 uncertainty in mind, and let's move on. We live with 24 uncertainty; we're not asking for absolutes. 25 The next item is sensitivity studies for 26 localized corroded areas. And I'm going to turn this 27 one over to Hans, because my notes are that we 28 basically have - we've only reviewed the results in 29 the application on these reference sensitivity 30 studies. And that we really weren't provided with an 31 expansion of what was - how was the sensitivity study 32 done, how were uncertainties considered in it; that 33 there's kind of an absence of detail at the next level NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

32 1 down. 2 And because we didn't get the detail, I 3 can't give you a specific question on what's missing. 4 So let me ask Hans on that one, because I think he's 5 at the same loss I am on that one. 6 MR. ASHAR: Yes, that is true. 7 I think I did point out about 8 sensitiveness, that they have to be correct enough 9 that we have confidence that the metal thickness is 10 what it's measuring. That is all I can say at this 11 time. 12 MR. GALLAGHER: What sensitivity studies 13 are we talking about? 14 MR. ASHAR: What we are talking about, as 15 I explained earlier, that you take a plate, similar 16 plate, and take the UT measurements outside, without 17 any coatings --inside. And then you take the 18 measurements with zinc oxide, whatever coating you 19 have applied inside, and outside epoxy coating, and 20 see - the measurements - and see if there is any - I 21 mean you have to take enough sample to make sure that 22 you have got confidence in what you are doing, even 23 for the tests. This is what we are thinking. 24 It's up to you. 25 MR. OUAOU: This is Ahmed with Exelon. 26 Inside, we don't have a coated - 27 MR. TAMBURRO: No, we have a protective 28 grease. They're supposed to clean off that in the 29 grid area, clean it off and then do the - 30 MR. GILLESPIE: Then I think you're exactly 31 where I think Hans is at, is, there was no evidence in 32 the submission, I think it talked about doing the 33 representativeness, but there was no description that NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

33 1 would say, okay, we do it with the grease steam 2 blasted off and we do it actually under conditions 3 that are in containment where they may wipe it down 4 with acetone or something else to get it clean, just 5 to get a handle on the uncertainty involved in the 6 measurement itself. 7 Remember, all the topics we're talking 8 about now are really uncertainties involved in the UT 9 measurements. And we're trying to get an 10 understanding of, how do you think about them, and 11 what have you don' t to make sure you have a handle on 12 the uncertainties. 13 And in this case, it wasn't really - the 14 description, you've already said more here than we've 15 had in the application dealing with the uncertainty. 16 We have an organic grease; we clean it off. So what 17 we're looking for is some understanding so Hans can 18 say, you know what, that's a pretty credible way to 19 understand the uncertainty involved with the 20 measurement technique being applied. 21 MR. TAMBURRO: This is Pete Tamburro again. 22 So when you said sensitivity, you're really talking 23 about sensitivity testing of how we do our ultrasonic 24 tests. 25 MR. GILLESPIE: You might say, what you're 26 doing to assure yourselves that you've got a handle 27 that the reading coming out -'and I know that every 28 utility has a program that does this kind of 29 qualification thing. It just wasn't described in 30 there. And the sensitivity here is a large component 31 very much of interest, and that information just 32 wasn't there. 33 On the other hand, we didn't ask you for NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE, N.W. (202) 234-4433 WASHINGTON, D.C. 2000543701 (202) 234-4433

34 1 it. 2 MR. GALLAGHER: Okay, thank you. 3 MR. GILLESPIE: So it wasn't your fault it 4 wasn't there; we didn't ask you for it. So we're 5 asking this. That's why we're putting it on the table 6 right now. 7 And again, I think you probably have a 8 program there. 9 MR. OUAOU: No, this was not in the QA. 10 In the QA we said that we're going to take the UT 11 measurements through the epoxy coating on the outside, 12 because it was qualified previously; and we're going 13 to use the most up-to-date techniques to do that. 14 MR. GILLESPIE: What is the most up-to-date-15 techniques for Oyster Creek? 16 -MR. GALLAGHER: So we' ll give you a 17: description.: 18 1 MR. GILLESPIE: Again, I didn't say any of 19 this was hard; I just said we don't feel we have it. 20 MR. GALLAGHER: Right. 21 USE OF ASME CODE CASE 284-1 22 MR. GILLESPIE: One last one - any 23 questions? One more under UT measurements, that is 24 going to be the use of ASME Code Case 284-1. And I 25 want to temper this a little bit, because there is 26 already a 284-2, as best I've been told, out. Neither 4 27 one have been endorsed by the NRC, but not being 28 endorsed by the NRC actually does not invalidate them. 29 But it does put a burden on you into 30 having to convince us on the applicability. And these 31 deal with buckling of the shell. 32 And the validation of the underlying 33 assumptions, you can't depend on ASME, because you're NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

35 1 using it, and they haven't - we haven't really looked 2 at it on their behalf yet. And that's kind of how I 3 understand the issue. 4 But now I want to turn to Hans for the 5 details on this one. 6 MR. ASHAR: Yes, this Code Case has been 7 within the agency for a number of years now, since it 8 was, the first one was proposed by Dr. Miller, who had 9 done the testing, and committed to all the results. 10 Now we did not endorse it during review of 11 reactors for the buckling analysis, 284-1. We did 12 take a Branch Position during that time. And in 13 addition to what they have done in 284-1, we require 14 them to do more in the bifurcation analysis, and 15 reduce the plasticity index, and those kind of stuff. 16 284-2, which ASME still is struggling 17 with, has a number of changes made in this area, and 18 that is - and with the typographical corrections that 19 they are making right now; put into the equations. 20 Because that makes a lot of difference in the research 21 you have. 22 So I think that looks to be something 23 acceptable you might accept in the future. Until now 24 there is uncertainty regarding the use of 284-1. 25 Now, if it is used only the way I saw it 26 being used is in one particular provision that is 27 quoted in response to the TLAA is that you have 28 assumed that the stresses are uniform along the 29 thickness of the metal. 30 Now in the case of a localized corroded 31 area, that may not be the case. Because when you 32 start from a corroded area to an uncorroded area, you 33 lose metal thickness. But it might have a lower NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

36 1 strength than the strength than you go up above at the 2 end of the plate. 3 Okay, if it's conservative, that's fine, 4 use it. But I believe it may not be conservative. 5 Because there will be a decreasing strength as you go 6 near the corroded area. And it might show you as the 7 metal thickness, but the strength may be different. 8 MR. GILLESPIE: Basically you've got that 9 oxide layer on the outside. And we're not saying it's 10 right or wrong. As I said, endorsement or not 11 endorsement doesn't affect the applicability, but it 12 puts the burden on you, because we have not accepted 13 it in this application to give us the explanation of 14 why you still think it remains conservative enough. 15 .. And this -is in addition to the RAI. It's 16 kind of the next level of detail down on that RAI. 17  : :_ MR. OUAOU: And again, Ahmed with Exelon, .18 we did, spent a lot of time at the site review on this 19 particular item. And the calculations that were based 20 on 284-1 were reviewed. And the conclusion is that 21 the impact of 284 for what we're using it for is not 22 significant. 23 There are a number of questions that deal 24 with those provided in response to these questions, as 25 well as the previous discussion, back in '91 or 26 whatever, that came up when this was used. 27 But one of the things - 28 MR. GALLAGHER: Just one question I have 29 here, isn't this really the same issue as item two, 30 the '91 GE document. 31 MR. ASHAR: Well, they have different 32 implications. One thing is about the area considered 33 for discontinuity analysis, and one is about the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W. WASHINGTON, D.C. 20005-3701 (202) 234-4433

37 1 buckling analysis itself. So those are two different 2 aspects there. 3 MR. GILLESPIE: In principle you could say 4 it's ASME code and (garbled) code. In principle and 5 philosophically even. One is dealing with some of the 6 assumptions in the GE calculation, and this one is 7 really dealing - 8 MR. ASHAR: The buckling analysis. 9 MR. GILLESPIE: -- with the buckling 10 analysis. 11 MR. GALLAGHER: The other one was related 12 to the buckling analysis also, right? 13 MR. ASHAR: Well, not necessarily. 14 MR. OUAOU: The difference with 284 is, 15 that':s what's actually again the capacity factor. 16 MR. ASHAR: Capacity factor. That is where 17 - 18 MR.. OUAOU: -- factors that you use to: 19 correct your allowable stress to come up with a stress 20 at the end; .. 21 MR. GALLAGHER: So did we provide a 22 description of the use of the Code Case 284? 23 MR. OUAOU: It was not in the RAI; it was 24 in questions. Yes. 25 MR. GILLESPIE: So again, we have two 26 processes going on. And the audit guys are still 27 writing their piece up. 28 But if you feel you've answered it, but 29 you need to understand Hans' specific concern is still 30 lingering in his parallel collection process is the 31 application of this code. 32 And we put this under UT measurement and 33 buckling analysis, because it's how you take the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

38 1 measurement itself, as I understand it, and then 2 incorporate that into the calculation, which is a 3 little different than the translation we talked about 4 in the calculation, the other ASME code piece. 5 What I'm trying to do is, we'll get it on 6 the table here. The audit process is going on in 7 parallel. And if you feel, if you can point to the 8 Q&A that it's answered, and just for convenience, 9 you'll be helping us out, for Hans. We're not trying 10 to make you recreate a whole new report if you've 11 already given us the information. 12 We'll internally check with the audit team 13 on the Q&As on this, but if you want to hold our feet 14 to the fire, because we've already asked it to you, 15 and email it in, that would be appreciated too, and 16 we'll make :sure we get the point covered. 17 But you need to know right now in the 18 overall evaluation, this is right now kind of an 19 unanswered issue. 20 MR. ASHAR: The main thing is that in the 21 response that you provided to the TLAA you say you 22 made use of a particular provision 1700, which is - 23 allows you to use it as the same test level throughout 24 the thickness. Now the point that I am trying to 25 make, it may not be true. So there might be a 26 possible distribution of the strength, and you might 27 have a different output from that pint. The analysis 28 is based on this type of assumption. 29 So I want to make sure that you are doing 30 the right thing. 31 MR. GALLAGHER:-Is there a different code 32 case or assumption we should be using? 33 MR. ASHAR: No, I think it would be - NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-701 (202) 234-4433

39 1 because this is very specific to the characterization 2 of the various containments. You have to make a 3 certain judgment as to how the strength near the core 4 area would be as compared to away from the core area, 5 and make a - if you have done the average strength 6 analysis, it will not be conservative, and you might 7 have to pull your neutral axis up, and it might change 8 the character of you compressive stresses. That's 9 what I'm thinking about. 10 MR. GILLESPIE: Mike, it's plant specific, 11 and ASME, as I understand it, doesn't really have a 12 lot of code cases that go out to 60 year lives, and 13 deal with longer term corrosion issues, and the 14 specific effects, and how they may modify codes that 15 were actually there for design codes. 16 And:.so we have to look to you to now 17 explain the application. And we're not saying the 18 application is wrong; we're only saying, you need to 19 explain this piece to us on the application. 20 . So we're not telling you to do it 21 different. We're only saying, again, this contributes 22 to the uncertainty of the application of it. And if 23 you've answered this in the RAI database and you can 24 point that out to us, and we'll check internally, 25 that's fine. 26 But this as of this morning is kind of an 27 uncertainty in the engineering case. 28 MR. GALLAGHER: Okay, you guys have any 29 other questions related to that. 30 MR. OUAOU: No, understood. 31 MR. GALLAGHER: Good, because when he 32 starts going into moving the axes on compressive 33 stresses, that's why he has to sit here. NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W. (202) 234-4433 WASHINGTON, D.C. 205-3701 (202) 234-4433

40 1 MR. GILLESPIE: Okay, next topic, and in 2 fact, the last topic, and again, we're trying to be as 3 fine tuned and as crisp as we can, because if we sent 4 you some general RAI to try to get where I hope we're 5 getting at this meeting, it would not have the 6 specifics that just transpired right here in it, so 7 that we can nail this thing down. 8 ULTRASONIC TESTING ISSUES 9 Ultrasonic testing issues: And now we're 10 shifting not to the technique, and not to the 1981 11 report, but sample size and sample locations. 12 And again, we have - I'm going to say - 13 three areas of clarification that are needed. 14 And this one is junctions between plates 15 of different thicknesses. - The generalization that I'm 16 understanding is, the reason for which points are. 17 being selected where..: And now we're really talking 18 about the upper parts, and the representativeness, bad 19 word, how.representative: the points you're using are 20 to the whole, which if it was demonstrated 20 years 21 ago, it's not clear that there has been a 22 redemonstration, as we're trying to add yet another 20 23 years on to the license. 24 And so with that, let me turn this one 25 again over to Hans for some detail. 26 MR. ASHAR: I'm going to go through three 27 areas here, okay. The cylindrical portion of the 28 sample, size and the spherical portion of the sample 29 size and the sanbed area. 30 The samples taken at this time in the 31 upper portion of the cylindrical portion it is taken 32 I think at one elevation of 87 foot 5 inch. Represent 33 a cylindrical portion of a drywell, and then it is our NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

41 1 suggestion at least for the future UT results to add 2 one more elevation for taking the samples, which is 3 71.6 inches. 4 And what the significance of that 5 particular elevation is that is where the lower 6 thickness meets the knuckle (phonetic) area. And the 7 question here is that if the water even in a small 8 conduit is passing through there, it is going to 9 stagnate in the area there, because the ledges form in 10 that area on the upside. And that is where the water 11 is going to accumulate, or it might be absorbed into 12 the insulation itself, wouldn't know what would 13 happen. 14 But that is a sensitive area which could

15. be subject to more corrosion than the straight portion 16 of the cylindrical area.

171 So*.our suggestion for the future is to

18. have -you include that. area near the junction of the -

'19 to get a confidence that you are good enough, your 20 sample size, enough locations taken. 21 MR. GALLAGHER: So this is elevation 71.6? 22 MR. ASHAR: 71.6, that is the suggestion. 23 You might not have platform there, you might have do 24 something else. So you may change a little bit here 25 and there. But the point is that the dissimilar 26 thickness, wherever you go to the joint between the 27 courses, you know, thickness courses. 28 MR. GALLAGHER: Just where the knuckle is. 29 1R. ASHAR: Just before the knuckle. 30 MR. TAMBURRO: This is Pete Tamburro. 31 So are you asking to take a representative 32 sample of one plate, and then the weld, and then 33 another plate? NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433

42 1 MR. ASHAR: Yeah, I think if you use the 2 6X6 grate in the grated area you can cover the whole, 3 including weld and everything, in one grid. 4 MR. GILLESPIE: Remember, the underlying 5 question is, because we're not telling you what to do. 6 What Hans has done is very nicely given you a specific 7 example of where he feels the physical configuration 8 forms an area which could be conducive to higher 9 corrosion rates than potentially your sample that 10 you're taking. 11 So the real question is the 12 representativeness of your current sample as we go 13 forward for even another 20 years. And it's not that 14 we're asking you to do this everytime; what we're 15 asking.you toido is reinforce the assertion that your 16 current sample is in fact representative. But we've 17 noted that you haven' t been looking at this area which 18 by.physical configuration could be picked out as maybe 19 a high corrosion area. 20 -. So .it's kind of the validation of what 21 you're doing, and so it - I guess what we're asking 22 is, remove this uncertainty in your sampling process 23 somehow. And the only way we can think to do it is to 24 pick a high corrosion area that's not being sampled 25 and ensure it actually is - continues to be enveloped 26 if you would by the current area. 27 MR. GALLAGHER: And I guess I 'm - Imean we 28 actually did some exploratory on the knuckle area, 29 didn't we, Pete? 30 MR. TAMBURRO: Yes, we did. 31 MR. GALLAGHER: In the past. So we have - 32 MR. GILLESPIE: Sometimes you only have to 33 document what you did if you did a good job. NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4M3 WASHINGTON. D.C. 20005-3701 (202) 234-4433

43 1 MR. GALLAGHER: Yeah, this drywell has been 2 very thoroughly looked at over the years. So the 3 chances are, we have that data, and I think we talked 4 about that earlier. 5 MR. ASHAR: Yes, I would have relied on the 6 thousand UTs you have done before. But because of the 7 continuing water - 8 MR. GILLESPIE: Yeah, there' s an operating 9 history there that gives us a concern in operations. 10 And again, we're updating - you know, in always 11 sampling a measurement, what you're really trying to 12 do is bring the applicability of that calculation up 13 to date, and that's really -- and the only way to do 14 that sometimes is a positive measurement. 15 -s And so yes, you might have done it 15 16 years ago* but there's been an operating history and 17 an experience base since then which has affected the 18 environment .in that gap. 19  :- And so it's your option. You can either 20 explain why 15 years ago applies to today, given all 21 of that operating history, or positive knowledge on 22 both of our parts, versus arguing words and pencil 23 notes, well, take a measurement. 24 MR. GALLAGHER: No, I think it's a good 25 idea. 26 MR. GILLESPIE: You know what I mean? It 27 eliminates all the bias, and to a degree the 28 uncertainty, and gives you a new point to project. 29 Because you're asking for a license for an additional 30 20 years. 31 And so we had confidence in that assertion 32 for the remaining portion of your current license, 33 which is 2009, and now we need something a little more NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 2 *3443

44 1 to project that past 2009 for an additional half a 2 life. 3 MR. GALLAGHER: Yeah, I think that's a good 4 idea. I just wanted to make sure you knew we'd looked 5 at that. 6 MR. GILLESPIE: I didn't know you had 7 looked at that area, because it wasn't part of the 8 application. 9 But again, you have to understand our 10 concern. It's not just isolated to that area; it's 11 that area combined with operating history subsequent 12 to those measurements being taken. 13 Again, to revalidate the trends, 14 revalidate the calculations. So we're kind of looking 15 at a revalidation process given the operating history. 16 MR. GALLAGHER: Okay, so any other 17 questions on that, guys? 18 MR. ASHAR: A similar request in the area 19 where the thickness of I think .622 missed the 1.548 20 thick area. There is the area likely to be - there is 21 some accumulation of water if anything is going on. 22 Similar to the cylindrical portion. The junction of 23 the thickness change. 24 MR. OUAOU: This is aside that region - 25 MR. ASHAR: No, above the same region. 26 MR. OUAOU: So from 1.154 to .77. 27 MR. ASHAR: Exactly. 28 MR. OUAOU: Okay. 29 MR. GILLESPIE: So the plate above the - 30 MR. OUAOU: Right. 31 MR. GILLESPIE: That joint right there. 32 MR. OUAOU: Yeah. 33 MR. OUAOU: This is Ahmed again. One NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

45 1 thing I'm not totally sure on yet, I understand the 2 differences in the thicknesses. But typically you 3 grind that so you wouldn't have that. We have to go 4 back to confirm that. So I just wanted to mention, 5 typically you wouldn't leave a discontinuity like that 6 going from one thickness to the other without grinding 7 it. 8 MR. ASHAR: Well, if they use a groove weld 9 to weld those two courses, I think you are going to 10 have a ledge. There won't be a transition there. 11 MR. GILLESPIE: Again, the big question is, 12 the representativeness of the current sampling program 13 for areas that when another engineer looks at it says 14 you could have a ledge there. 15 Again, we are not here to give you the 16 answer; we're giving you our concerns. And there akre 17 two ways you can do it, and there are a combination of 18 two ways you can explain it. 19 . 4 MR. OUAOU: The only thing I may add is 20 that when the investigative work was going on to come 21 up with the very 1,000 UT measurements to find the 22 thin areas, we didn't stay away - I don't think we 23 stayed away from the areas where we transitioned from 24 one plate to the other, especially when you do that 25 from the outside. 26 You move the template along the elevation 27 to see where you have a corrosion, and you don't 28 specifically say I'm going to exclude this area 29 because it's not - 30 MR. GILLESPIE: Again, that was for the 31 life of the current license. And really what you're 32 asking for in renewal space in your application, 33 fundamentally, is to take that projection and now move NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

46 1 it forward now almost 17 years or 20 years to the end 2 of the license, and you're asking the NRC to make 3 another 20-year judgment. 4 We're fundamentally remaking the 20-year 5 judgment we made before. And so it's the same 6 technical issues, are still the same technical issues, 7 and again, it's your choice. But what we're looking 8 for is the least uncertainty on the measurement of 9 making this projection forward 20 years that's also 10 rational. And it's your judgment. 11 So you understand, we still have this 12 uncertainty. We're not negating the finding from 13 1991, but you're asking us to take that and now move 14 it forward, and all Hans is saying is, actually no new 15 .. positive measure which now we're not arguing

16. calculations or philosophy, there is no.new positive 17 measure in :this area of potential. We're not saying 18 it is a high area, but there is a potential, normal 19 industry practices grind down welds and make them 20 smooth. We're also not disagreeing with that.

21 But it seems that you need to understand 22 our concern is, in just looking at the physical 23 arrangement, this is an area of potentially higher 24 corrosion, and we're asking why is your current sample 25 set still representative of that area? 26 If the explanation is, we were 1,000 27 percent sure that this was ground down, and that there 28 are no crevices or anything in that grinding that 29 could catch water, that's one way of doing it. 30 There are two approaches to everything. 31 MR. GALLAGHER: And I think what you're 32 saying, Frank, is that some of these areas that helps 33 to narrow the uncertainty. So I think we - NEAL R. GROSS COURT REPORTERS A* TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202)234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433

47 1 MR. GILLESPIE: Remember, we're trying to 2 be as clear as possible to you. 3 MR. ASHAR: In the pocket region of the 4 drywell shell, the most susceptible bays are 5 incorporated in the sampling, the present sampling is 6 fine. 7 However, there are a number of issues that 8 need to be addressed to ensure that readings are taken 9 at whatever locations, and techniques used are 10 reliable. 11 It is not clear if the junction between 12 the 1.154 inch plate and the .676 inch plate, which I 13 think I had explained to Ahmed when I was there in 14 audit on April 28th. 15 . That area - we do have a concern in that 16 area. Because you took out the sand: from the sand .17 pocket area, before you put the ceiling in the

18. junction:between the steel and the concrete, quite a 19 bit. of amount, of water might have seeped through in 20 those areas, and might have caused corrosion in those 21 areas.

22 And the way we are writing is, we'd prefer 23 that you try to find out some technique to measure the 24 thicknesses in those areas and alleviate any doubt 25 about there is no corrosion. Or if there is 26 corrosion, then you know about it, how much it is. Or 27 justify why this area should not be included in the 28 sand pocket areas. 29 You understand what junction I'm talking 30 about? 31 MR. GILLESPIE: Let me give you a little 32 more amplification on this, because Oyster Creek is 33 not alone on this. We had an ACRS meeting yesterday NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433 wwD * - FwF*

48 1 where this area between the concrete steel concrete 2 sandwich, in there, were addressed. 3 And as best I understand it right now, 4 there really is likely - while we have some research 5 going on in this area, and a research letter from Oak 6 Ridge available, this is not an area where I think - 7 and I think we recognize this - that there is a lot of 8 commercial activity. It's accurately being able to 9 measure through concrete, through steel, and into 10 concrete in that environment, without chipping 11 concrete out, and we have to be, at ACRS, talking to 12 a licensee' that actually chipped concrete out 13 yesterday. 14 But some of the discussion went on, with

.15   ACRS. And again Hans'        second comment was, provide us 16    at least with a rationale that is coherent and makes 17   sense.-FAnd some of the points that ACRS raised in.

-.18 challenging: the staff on our interim staff guidance,: 19 where we:;had to kind of make a rationale for such - 20 aspects: of the fact that the inside containment 21 temperature is like 130 degrees. And therefore it's 22 going to drive moisture out. The lack of oxygen in 23 the area. 24 Once it's been sealed, the initial 25 oxidation is going to consume the available free 26 oxygen, and therefore, there is some severe 27 limitations on corrosion. 28 These are the kinds of things we discussed 29 with ACRS in a broad sense of applicability of how 30 we'd see an applicant trying to address the rationale 31 portion, of this, -if chipping up the concrete was 32 really not rational. 33 That explanation I don't believe was in NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W. (202) 2443 WASHINGTON, D.C. 20005-3701 (202) 234-4433

49 1 the RAIs or the application. You have to make it for 2 us. I know what we did as staff to help support our 3 interim staff guidance on why it wasn't more demanding 4 if you would in this area. And these are the kinds of 5 things that were going through our mind. And I want 6 you go away understanding that that's the same thing 7 that was on the record at the ACRS meeting was the 8 kind of rationale the staff had in mind as to why this 9 actually should be. And looking at the chemistry of 10 it, an area of fairly low concern. 11 But you have to tell us why for your plant 12 it's a fairly low concern with your operating 13 history. And so then there's timing elements about 14 when the seal went on, when various leakages might 15 have occurred, when water could. have accumulated, 16 groundwater levels, and the ACRS asked about, what 1?. about concrete, it's .porous, it contains water. Then 18: an ACRS member said, yeah, but there is no oxygen-.

• 19 left. .

20 So it's. that rationale, or advanced 21 measurement techniques that might or might not be 22 available. That's not my area; I don't know. As you 23 know we have an Oakridge report, and I think we've 24 already supplied you with our ADAMS number. 25 MR. ASHAR: ADAMS number. 26 MR. GALLAGHER: Do we have that report? 27 MR. GILLESPIE: But let me be careful, I do 28 have to be rational, we're not asking you to be in 29 advance of the state of the art of applicable 30 commercial techniques, and again, in RAIs I couldn' t 31 say that, but we're trying to keep this in context. 32 But we do need a signed understanding from you, in 33 your words, as to why this should be a low susceptible NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 23"M43 WASHINGTON, D.C. 20005-701 (202) 234-4433

50 1 area. 2 And that's your choice on how to do that, 3 and we've supplied you with the one letter report, and 4 since it's a letter report on a NUREG, that tells you 5 right there, it's very advanced information. 6 And so you have to digest the information. 7 Just understand our concern, and the rationale we're 8 looking for. 9 Does that make sense? 10 MR. OUAOU: I understand. This is Ahmed 11 with Exelon. We understand the question. 12 We did not provide all that detail you're 13 talking about in the application. We specifically 14 used a NUREG-1001 as a basis why that area is not 15 susceptible to accelerated corrosion. -16 Basically the idea is, if it's embedded in

*17    the concrete,. you don't have an adverse environment, 18    chlorides :and sulfates, you should not - you know; you 19    have an alkaline environment that is not conducive to 20    corrosion of the shell.

21 And all those items that you mentioned 22 contribute to why that area is not - 23 MR. GALLAGHER: So we did not provide 24 that. 25 MR. OUAOU: We have. 26 MR. GALLAGHER: Where is that, in the 27 application? 28 MR. OUAOU: It's in the application; it's 29 in the questions, Q&A. We did not provide, we did not 30 state that it's totally sealed; the oxygen is limited. 31 We didn't get into that detail. 32 MR. GILLESPIE: Yes, and again we didn't 33 ask for it. Again, we're at a level of detail, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., KW. (202) 234-4433 WASHINGTON, D.C. 20006-3701 (202) 234-4433

51 1 because you did supply us a lot of information in the 2 RAIs, we're really now fine tuning and focusing in on 3 these real specifics. 4 MR. ASHAR: We did mention about the 5 inaccessible areas, and we did provide certain 6 guidance to if the concrete is like this or that. 7 Then you may not have to do much in that area. Butý- 8 Oyster Creek is a little different animal here, 9 because it has a history of contaminated water going 10 into the sanbed area. It might have seeped through in 11 the area with the thinnest part of the steel is there. 12 Though it is bearing on concrete, still, it is very 13 thin. And if it is rusting, there are problems with 14 it, and with the analysis, too. 15 :MR. GALLAGHER: And as far as the 16 techniques for looking at this, we had looked into 17 that, and-we hadn:'t really found anything. 18 Did you see anything, Hans? 19 MR. ASHAR: Yeah, in this Oak Ridge report 20 that Frank talked about does have three separate 21 matters came in. Each will have a different 22 applicability. I don't know which is more suitable. 23 I cannot recommend to you that. 24 But there is a potential for use of one of 25 those methods. We requested Research to have Oak 26 Ridge National Laboratory conduct a study. They are 27 state of the art kind of report. They give a contract 28 with three separate independent people to develop some 29 kind of techniques to have the metal thickness results 30 being given when the metal is embedded in concrete on 31 both sides; that was the main purpose of it. 32 So there is some applicable review. 33 MR. GALLAGHER: And we have that Oak Ridge NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W. (202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433

52 1 report? 2 MR. OUAOU: We have the - John you have 3 the - yeah, right. 4 MR. GILLESPIE: But again, I really went 5 out of my way to try to keep that in perspective. 6 MR. GALLAGHER: Right, so we'll take a look 7 at it. 8 MR. GILLESPIE: Sometimes people say, the 9 NRC asked me a question, and that's telling them to do 10 something. I'm not. I'm asking, just reevaluate the 11 data. I'm not insisting on people to do the 12 impossible. But it's the rationale and the details 13 underlying it. You didn't give it to us; we didn't 14 ask for it. And that ' s why we ' re here saying, this is 15 that little piece that's missing under here. 16... And as we told ACRS yesterday, although we 17 kind of.have a generic position, our generic position 18 is really applicable,ýto facilities!that have had no 19 history at all of leakage. And then you step off from 20 that, and when we reviewed Brown' s Ferry, they were a 21 little different. You were a little different. Your 22 operating histories are slightly different. 23 And so the generic applicability strictly 24 of the new reg what we're saying is, there is some 25 customization you have to do specific to your 26 operating history. 27 MR. GALLAGHER: Okay. 28 MR. GILLESPIE: Ready for the next one? 29 MR. GALLAGHER: Yes. 30 MR. GILLESPIE: Okay. 31 INSPECTION INCREMENTS WITH UT COMMITMENT 32 Sanbed region inspection increments 33 associated with UT commitment in letter dated April NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

53 1 4th, 2006, page 3, item two. 2 This - actually XIm going to get your 3 commitment - don't throw it out. It was a good 4 commitment. Let me try to articulate this one. 5 And our thinking is, we're trying to be 6 very consistent with the previous thinking back in the 7 '80s. And also with concepts that we kind of have in 8 the maintenance rule and other things. And the idea 9 is, the intent here is to bring all of this technical 10 information that was developed in the early '90s 11 forward, and essentially revalidate for today. 12 I do understand, in the press clippings, 13 although I don't think you've written it to us, that 14 you were going to do some measurements in 2006. 15 .1read.that in the paper. But we probably 16 would - it would be ,beneficial- to have that on the 17 record. And I assume that's your commitment actually 18 to do it, that that"s -the one you're going to do prior 19 to entering.the period. 20 MR. GALLAGHER: !That's correct. 21 MR. GILLESPIE: Well, I'm making that leap 22 of faith assumption. So the measurement you are going 23 to do prior to entering the period is really the first 24 measurement that's been done since 1996. and there's 25 been a significant amount of history since 1996. 26 And I'm going to simplify this down to my 27 kind of thinking. It takes two points to have a line 28 in order to have a slope. And there's been some 29 operating history between "96 and now that one point 30 validates to some degree I'm going to say current 31 thickness for the last 15 years. 32 But then you're asking in your commitment 33 to jump to not do anything for 10 years, okay. Now NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

54 1 I 'm going to invoke the concept that we have in the 2 maintenance rule, which is kind bf more of an OR gate 3 (phonetic) if you would for any measurement which 4 says, if we do the measurement in '06, and we see some 5 level of degradation which is inconsistent with what 6 you would have predicted, then you're going to do 7 something. 8 Then if I go to the maintenance rule, it 9 says, I 'm going to increase my surveillance frequency. 10 And then if you increase your surveillance 11 frequency and see with the second measurement that 12 it's stable, then you decrease your surveillance 13 frequency. 14 What we'd ask is it's - there is no 15 criteria for-what happens,.. what you're going to find 16 in '06. It still leads in our.: mind to a degree of 17 uncertainty. And we'd like to.'ask consideration in 18 terms of what~s the basis for 10 years? If you say 19 you're going to do something in '06, and if that's 20 part of some criteria, then.we're going to do 21 something within four years after that again. 22 Now you are really consistent with our 23 previous judgments from last time, in which you 24 committed to do several I think measurements I think 25 in a row at four-year intervals. 26 But then if you come -up with a second 27 measurement, and it's better, then there should be an 28 opportunity to extend it past that. 29 So what we're suggesting is, in our minds, 30 we're looking for some sense of commitment to what 31 happens, what's your criteria if you find something, 32 thinner, thicker. What happens if this measurement 33 comes out like the '96 measurement, and comes out as NEAL R.GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON. D.C. 200054701 (202) 234-4433

55 1 ii growing more? 2 Then there is a calibration issue I hope. 3 And so what we're looking for is I'm going 4 to say a bit of a more disciplined reliability 5 approach to the sampling plan maybe as opposed to the ik-6 rigidity of 10 years. 7 And there's a sense on our part right now 8 that given our current knowledge base, and the 9 uncertainties in operating history, the uncertainties 10 in the '96 measurement itself, which may not be - you 11 might not be able to do anything - it's 10 years ago. 12 I'm just being realistic. 13 The coatings are getting older. Yet you 14 aren't going to do the inspections. We're not i15 questioning your inspection regimes,. your commitment, 16 that's very good, to do 100 percent in 30 years of 17 commitments. But it'is getting older, so there is 2 18 these degree of uncertainties that more progressive 19 sampling - the broad RAI-. would be, what is the 20 justification for 10 years? 21 Because 10 years is independent of what 22 you find in '06? 23 MR. GALLAGHER: I think one of the things 24 we tried to do, Frank, was, if you look at all those 25 commitments, they're kind of like an integrated 26 package, you know what I mean? Because the agent-27 management program is an integrated package on that. 28 And I guess what we were trying to say and 29 maybe it didn't come across, we take the readings 30 before the end of the period and we did have some - 31 and our expectation is that the corrosion has been 32 arrested, and has been arrested Since 1996. So our 33 expectation is, we would have similar measurements. NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202)234-4433

56 1 And then we said, we had the criterion, 2 and I think it was plus or minub 21 mils? And it was 3 based on -Ue uncertaint-ies o meaSurin-g and equipment. 4 And then if we were outside of that we 5 would notify the NRC within 48 hours. And we made a 6 commitment to that effect. And that we would have 7 specific actions. 8 And those specific actions relate to doing 9 the projection, increase the frequency of the testing, 10 and things like that. 11 We didn't put the decision tree in there,j 12 but that's our intent. db 13 MR. GILLESPIE: Okay, and on this aspect - 14 as I said, don't throw out the commitment. The 15 commitment, it was a very good. commitment. 16 Our question really is the decision tree, 17 and we've had this same discussion -actually with Nine

12 Mile Island on could you give us the acceptance 19 criteria. ..

20 Because while you can assume that 21 everything will be correct, as the regulator, we 22 cannot assume everything will be correct. 23 And so it's a decision tree that affects 24 inspection frequency. You're reporting to us, all of 25 that was fine. What we're doing is, saying that the 26 specific commitment that says, we're going to do a 27 measurement before we hit the period, and then, 28 really, reading it word for word literal, the next 29 measure is at 10 years. 30 We're absent that decision logic that you 31 have internally that would make perfect sense. And so 32 on the frequency thing, we're asking, could you give 33 us a relook at that in 10 years, and either NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (M0)234-4433 WASHINGTON, D.C. 20005-3701 (2M2)234-44,33

57 1 rationalize why 10 years as an absolute is okay, or 2 provide the commitment of what your decision tree is, 3 relative to frequency of remeasuring versus which 4 goals. 5 Again, you're assuming it won't. And 6 we're regulators, so we have to assume it will. And 7 we need to address both sides. 8 And quite honestly, I think, in the 9 public's view, they need to have a certain assurance 10 that if this becomes a commitment, or whatever, within 11 the license itself as we reissue it, then it becomes 12 real solid, it's inspectable, and what I'm saying it 13 has all the bells and whistles on it that go with the 14 regulatory process. 15 So we'd ask you to relook at the 10-year, 1i and you've just described an internal logic that is 17 not visible to people on the outside who read the 18 literal words of that commitment. 19 So the request is, could you look at the 20 commitment on the 10 years. Because we' re reading it 21 like an absolute. Yeah, you report to us, you'll do 22 all those things, but gee, they never sai. tney.'d go 23 in and remeasure. 24 MR. HUFNAGEL: Frank, this is John Hufnagel 25 from Exelon. Just a clarification. Because when I 26 was listening to you, I believe you may have said that 27 even if we went in and found essentially the same 28 result with the ultrasound testing, the 10-year 29 frequency may not be enough. 30 So I think what Mike described was if we 31 would go in, we would find some degradation, we would 32 consider corrective actions including things such as 33 more frequent inspections. NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

58 1 MR. GILLESPIE: Okay, now we get to the 2 uncertainty issue on that. Akd that's why I can't 3 give you a specific answer. That's why I said it kind 4 of nebulously. 5 The uncertainty issue is, if you go in and 6 you do the measurements, and let me say you have the 7 same issues that you had in '96 that were kind of 8 inexplicable but why it grew, then 10 years is 9 probably too much. 10 And so what I'm dealing with, and I can't 11 do it for you, I'm dealing with, there is an operating 12 history there. There are these uncertainties that in 13 fact you may be within - if it's an asymmetrical 21 14 mil objective, then you still have the same regulatory 15 question, well, it grew again. They don'thave to do 16 anything. - -* 17 MR. GALLAGHER: Right, :and we would take 18 -corrective action. So I guess maybe related to the 19 question John just asked, so. I guess. the corollary 20 would maybe be, if we were-within that plus or minus 21 21 mils, is 10 years okay? 22 MR. GILLESPIE: There is no absolute on 10 23 years. Okay? That was what was in your application. 24 There is an uncertainty connected with these 25 measurements. There is a specific uncertainty 26 demonstrated in measurements at Oyster Creek 27 specifically over time. 28 If you really are trying to bring that 29 forward, you have to make the judgment, is once at the 30 beginning of the period, and doing a second one at 31 four years, and then not doing any more for 16 years, 32 is the right answer. 33 Because remember, what you're trying to do NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON. DC. 20005-3701 (202)234-4433

59 1 is take this calculation and all this body of 2 information from the '80s and "90s and reapply it to 3 a new 20-year period. And if you're going to do two 4 measurements, should that second commitment for all of 5 these questions actually be way out there at 10 years? 6 Or if you're going to do two measurements anyway, 7 should it be at four years or six years? Because 8 that's giving us assurance on the projection of all 9 this body of data forward. 10 And by the 10th year it's not really 11 contributing to the projection doing forward. 12 And now I'm going to make a leap of faith 13 to a new topic - 14 MR. GALLAGHER: Before you go there, Frank? 15 MR. GILLESPIE: It'll make sense though if 16 you let me do it. 17 MR. GALLAGHER: All right. 18 MR. GILLESPIE.: Because it' 11 make sense to 19 why I just said what I'm saying. In the interim staff 20 guidance there is an event aspect to it, which says, 21 if you ever see water, you have to go do a 22 measurement. 23 And so it's not mutually exclusive. And 24 so if you're committed to two measurements on a 25 frequency that allows us to translate this body of 26 information forward for most of the period, we would 27 still ask you, you have not committed to the ISG 28 rvlative to that event aspect to it, which says, if 29 you see water, you have to measure again. 30 And so I'm .saying th-9-Mar ent thing 31 is kind of an integral case. And if you're really 32 good and you never get a leak again, you've still only 33 done two measurements, but you're adding to the NEAt R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

60 1 principle of moving the body of knowledge forward. 2 But if you ever see Water again, you are 3 committed to a third measurement. 4 And so it's a package. I'm agreeing with 5 you; it's a package. And that's not in the 6 commitment. And it's kind of the event based aspect 7 65 7hat-ISG which then says, you need to redo your 8 rate calculation and project it forward it forward if 9 you see moisture. 10 And that's the package I wanted to get 11 out, because it's not like I'm - we've kind of done 12 something thinking on this, and what are we really 13 trying to achieve relative to the staff's approval of 14 your application, and we're trying to approve is that 15 projection forward for the next 20 years. 16 We're not actually trying to specifically 17 find the thin spot at any given:year; we' re trying to 18 have enough comfort if.. you would or faith. And it's 19 faith in that calculation we're trying to get, not 20 just a random measurement at a 10-year point of a 21 vessel. 22 So it depends on how you look at - what is 23 your objective of doing those measurements. If the 24 objective is a random point in time, to make everyone 25 feel comfortable with something you've already 26 approved, the first thing is, get the piece already 27 approved. 28 MR. GALLAGHER: We'll definitely look at 29 that, Frank, because again, I think that was our 30 intent, outside this region, we would change the 31 frequency. 32 But one thing I just want to clarify 33 because even sometimes we fall into this trap, and we NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W. (202) 234-4433 WASHINGTON, D.C. 2006-4701 (202) 234-4433

61 I

*1      talk about the individual components of the aging 2    management program.

3 Like people would say, hey, the last 4 management we took was '96, and that's a long time 5 ago. You know we have the advantage at Oyster Creek 6 where that area is accessible now, because we made 7 these modifications. So we've had eyes on on the 8 coding ever since then, that the coding is put on in 9 '92. 10 So that was our look, ongoing look, to 11 make sure that corrosion was arrested, and was gone. 12 Except so we see that as a real good advantage for us, 13 because we have that area to be accessible. 14 So when you look at the package of UTs and 15 visuals, it's a pretty good one. 16 MR. GILLESPIE: But that's why I just - and 17 Hans, you can jump:in, because I might. say something 18 wrong here. But you notice coatings wasn't on our 19 list, and you're answer and-your commitments in that 20 does reinforce what you just said. 21 So again it reinforces if the codings are 22 being expected reasonably vigorously at one time 100 23 percent, and then it'll go to 30 each outage, that's 24 confirming the underlying assumption that moisture 25 isn't present and therefore corrosion doesn't occur, 26 which makes the usefulness of a 10-year out 27 measurement potentially less useful than one that 28 29 might be in more like a four-year duration that allows us to do what we did in the '80s, to say, okay, you've I 30 got enough information to project this forward, and 31 now depend on' your commitment on the coatings 32 examination. 33 And so at least and I'm going to ask Hans, NEAL FL GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

62 1 because I - and so there is a thought process there 2 that is different than just picking 10 years because 3 it's in the middle. 4 Hans. 5 MR. ASHAR: Yeah, I think programmatically, 6 I think the way you have committed to coating 7 inspections, if during the inspection of coatings, you 8 see seepage of water that you have seen earlier in 9 2004, 2006 time frames, then there is always a 10 question as to what is going on. 11 -And that's why what Frank is trying to 12 explain is that you've got to have a program based on 13 what you find rather than straightforward to 10 years. 14 And I think Frank did describe it very 15 vividly, but I'm trying to simplify it. That's what 16 we are looking at here. Programmatically. . 17: MR. GALLAGHER: And I think that was our

. 18    intent, but we can clarify that.

19 MR. GILLESPIE: So:- the summary is, could 20 you relook at the purpose of the 10 years, and is the 21 10 years really serving the purpose of bringing this 22 data point forward so that we can make the same 23 decision now for the next 20 years we made before for 24 the last 20 years. 25 26 27 questions? MR. GALLAGHER: MR. OUAOU: You guys have some Well, the only thing I really 1 28 want to add - this is Ahmed with Exelon - is the UT 29 measurements we're using in the sanbed region is to 30 confirm that in fact corrosion is not undergone, which 31 is stated, it's arrested. 32 But you've got to remember, on a forty-I 33 year basis, we're still doing UT measurements on the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-443 WASHINGTON, D.C. 200054701 (202) 2344433

63 1 upper region of the drywell, which is not coated, and 2 it really should bound the other areas. 3 MR. GILLESPIE: Again, you're making my 4 case why 10 years may be a random point that is just 5 out there that was picked because it's in the middle, 6 as opposed to being a point that in a real early part 7 of a period contributes to reinforcing the fact that 8 the body of knowledge in the inspection techniques for 9 both how you apply that corrosion rate you're finding 10 at the top which is uncoated, and how you look at the 11 coatings, is doing. 12 All the reasons you're giving me are 13 reasons why you want to reinforce your technical bases 14 early as opposed to late. That' s all I'm saying. I'm

15 just asking you to think about it.

16 MR. OUAOU:. The only .thing I want to point .17 out is, the basis for .the 10 years we -used for -was 18 certainly not random. It 's based on the ISI interval. 19 MR. GILLESPIE: Okay, the ISI period is. 20 also 10 years. 21 MR. OUAOU: That was the basis for it. 22 MR. GILLESPIE: We've actually had some 23 discussions with people that the whole ASME code 24 issue, which is not yours, is given us great pain in 25 aging management as you know with relief, because the 26 code is written to cycles, et cetera, et cetera, that 27 are really based on a 40-year life. 28 And so we're, again, that may be the code, 29 but that is not - I'm trying to say-, it could be a 30 technical rationale, other than it's convenient with 31 the code for doing it. And because of your answer and 32 commitments in the coatings, because of the 33 reinforcing measurements at the top in the uncoated NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE. N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

64 1 areas, we're looking for as much definitive 2 information early in the period that there will be 3 success during the period as we can, relative to 4 projections. 5 And again, the kicker in here is, we'd be 6 looking at the event part of the ISG which then 7 applies to future - because you got a 16-year period 8 I just suggested in there. But the ISG would say, 9 if water shows up, you're doing UTs again. I mean 10 that's what the ISG says. 11 But if you're real good and you never have 12 a leak, because of the inspections and the projections 13 and that, and then the validity of your projections 14 are doubly reinforced early in the period. 15 So I 'm asking you to look at the rationale 16 for the 10 years, and what I'm suggesting is, in light 17 of how we thought about the maintenance rule when we 18 were writing that, and what we were doing and what was 19 happening in the maintenance area; I 'm applying those 20 same principles. 21 Remove the uncertainty early, and that 22 allows you to have a justification and a rationale for 23 the extension, and why a more minimal surveillance 24 program is unsatisfactory. 25 MR. GALLAGHER: And Frank, do you have any 26 thought in mind for what an early interval would be? 27 MR. GILLESPIE: No. If you can rationalize 28 10 years as being early in providing the moving 29 forward into the entire period of that - 30 MR. ASHAR: We'll look at it. 31 MR. GILLESPIE: -- we'll look at it. But 32 the uncertainties involved - and I will admit, this is 33 - there is some subjectivity to this. I mean this is NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234"4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

65 1 not an algorithm that we can put into a spreadsheet 2 and do a calculation on. 3 But there are a number of uncertainties, 4 the residual ones we went over today, which we're 5 looking for clarity in. And I think what you as an 6 applicant are trying to do is reduce or minimize those 7 uncertainties to the degree possible for the maximum 8 period of operation. 9 And what we're suggesting is, 10 years 10 leaves a great deal of uncertainty in our minds 11 relative to the sample selections, and projecting this 12 vast body of data and this calculation forward. 13 Again, we're dealing with taking a vast 14 amount of information which was reviewed now almost

15- 15, 16 years ago,. it.was probably developed close to 16 18 years ago, and.bringing that forward for a new 20-

.17 year period. - 18 And yes, that was satisfactory for the 19 last 20 years of the license, but now we're making a 20 new finding that it's satisfactory for yet even 21 another 20 years. 22 MR. GALLAGHER: Okay, I think we understand 23 it. Okay. 24 MR. GILLESPIE: With that, I've got one 25 other issue, and this - to close out containment, and 26 to let people know that what we've talked about here 27 is only a small piece of the whole. 28 And this has no action .for you. But 29 actually in looking at the whole thing, we were really 30 trying to look holistically as a staff at the entire 31 containment structure. And we did note that your last 32 appendix J integrated leak .rate test was in 2000, 33 which means your next one is due in 2010, which is NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

66 1 really close to the beginning of the period. 2 And while that is not a design test, there 3 are other things going on as part of our body of rules 4 that do affect the integral look that we take at 5 things like the containment shell. 6 And so I didn't want people to think that 7 we only looked at what we talked about at this 8 meeting, which I'm hoping was very focused and quite 9 narrow to our residual concerns. 10 But that we do see that kind of under the 11 rules you have to pick a date, 2008 or 2010 plus or 12 minus a year under 10 years, and that's probably 13 either one within six months of the renewal period. 14 So there are other things going on to give 15 us increased assurance of the. operability of the 16 shell. And.because.this :isn't just a meeting between 17 you and us,?:I want people - and this is an example of 18 other things that we're considering.. So we're not 19 just narrow-people. We're not just looking at the 10 20 years and asking about that. We actually found some 21 really satisfactory things, and just in compliance 22 with the regular body of rules that was going on. 23 And with that, I 'm down to my topic called 24 general discussion, but Im' about worn out. 25 GENERAL DISCUSSION 26 I would ask you and then 1'11 ask - or 27 perhaps I should ask Hans if he has anything else he 28 would like? 29 MR. ASHAR: No, I don't think I have 30 anything more than what you described, no. 31 MR. GILLESPIE: I would like to ask you as 32 an applicant - I mean we're trying to be real crisp 33 here, because we want to get on with the job. NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234433 WASHINGTON, D.C. 200053701 (202) 234-4433

67 1 MR. GALLAGHER: John, do you think we have 2 succinctly what the issues are we need to respond back 3 on would be? 4 MR. HUFNAGEL: I have a lot of notes, Mike. 5 It would take me more than a couple of minutes to go 6 through these notes. So I'm not sure I can go and 7 summarize all that right now. But I think between us 8 I'm sure we have enough notes and understanding. 9 MR. GILLESPIE: And we're going to do our 10 best, by the way, John, to get what Hans was reading. 11 We went through a lot of effort to try to really 12 narrow this down. But we do have the audit process 13 you know kind of going on in parallel. And we t I1 try 14 to get these meeting notes out in a timely way for us, 15 and timely for us, given our secretarial situation, 16 can be long. 17 . But .in this case we're going to push this 18 to ki-nd of the front of the list, and try to maybe - 19 I need to get.these notes'.out in a.public forum. 20 And again, if there is a follow on email 21 needed to clarify the issue, that's fine. 22 The other question that came up, because 23 normally we would have probably followed this meeting 24 with a formal set of RAIs. When I saw what Hans had 25 written in coordination with the bullets we wanted to 26 covered, the RAIs are really embedded in his detailed 27 words. And these, I think, are more - are better 28 words than we generally send in kind of our 29 whitewashed versions of RAIs that require phone calls 30 for clarity on. I 31 So we will try to get the meeting notice, 32 the meeting minutes out, with basically Hans' comments 33 and the bullets, as quickly as possible. NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433

68 1 Now, if we do that, then my intention 2 would be not to issue a separate document of RAIs in 3 prep for your opportunity to come back and talk to us. 4 And the other thing is, we'd like to ask 5 that you send something in in writing before that 6 meeting so that we can really be kind of at the end of 7 the road at that meeting. And the question I have of 8 you is, when we were setting this up, we scheduled it 9 so we could talk to you, and scheduled the next 10 meeting so you could talk to us. But that is actually 11 your option. 12 If we don't need a meeting, and you'd 13 rather answer these in writing, I would just ask that 14 you get back to us in a timely enough way so that we 15 can cancel .the meeting at least a week before. 16 And it's really your option, but we were 17 trying to set this whole thing up to make sure that we 18 had alln:the vehicles for communications. And since 19 we have a 10-day. noticing period for public meetings, 20 and it takes a couple of extra days - it really takes 21 about 15 days to do it - then we had to in a positive 22 way set up both meetings at once just to have a 23 process put in place. 24 But it's your application and it's your 25 answers and it's your choice. So right now we do have 26 it scheduled. Donnie was going to put a notice out, 27 but I would ask for the other people in the public who 28 might want to participate, a timely notification of 29 them is an obligation we have. 30 And so let me leave that to you and not 31 even ask you to answer that 4gestion today. But you 32 can get back to us on how you want to do th at. 33 MR. GALLAGHER: I think what we' re going to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

69 1 do, Frank, is we'll meet and go over what issues we 2 think we have. Maybe John and Donnie can communicate 3 to ensure that these are the things we' re going to be 4 providing in a written format, and we would want to 5 get that to you a few days before the 22nd, and 6 whether we meet or not, we can determine that at a 7 later date. And talk with Donnie about that. 8 MR. GILLESPIE: I'll leave that to Donnie 9 and John, then, 10 MR. GALLAGHER: And then so that would be 11 our - the things that we talked about providing, we 12 would provide that in writing; that's what you're 13 looking for. 14 MR. GILLESPIE: Hans and I are going to try 15 to get everything that we have in writing out as part .16. of the meeting minutes with Donnie. I think it's

.17  going to be a more fruitful meeting if everyone, all 18   the participants, has it in black and white. And then

19. you leave that-with either a markup or a nonmarkup, 20 and everyone knows where we stand on these issues.

21 Because I think we 've really narrowed some things down 22 here. 23 MR. GALLAGHER: That's what I was going to 24 say. Like the issues, like the thank you for getting 25 clear with us on what these issues are. Because I 26 think they are very pinpointed, and I think that will 27 help us really see what information you need to close 28 the issues. 29 Like you said, we provided a ton of 30 information, and we have it down to just a handful 31 right now to really get you just the information you 32 need. 33 MR. GILLESPIE: By the way, we're not NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

70 1 looking for another ton. We're really trying to see - 2 if these answers end up coming in 57 pages long, then 3 we've miscommunicated what we think our residual 4 concern. 5 MR. GALLAGHER: Okay. 6 MR. GILLESPIE: So really, as you're doing 7 it, keep it in perspective. And if that requires 8 calling Donnie, say, you know what, Frank said he 9 didn't expect the Encyclopedia Britannica for every 10 question. We think these concerns are very focused. 11 MR. GALLAGHER: Right, right. 12 MR. ASHLEY: In addition - this is Donnie 13 Ashley - in addition, John, to your notes and the rest 14 of our notes, we're going to try to get a quick 15 turnaround on the transcript so that you can have that. 16 available to you as well. And we'll have that in 17 ADAMS just as quickly as we can. 18  : MR. GILLESPIE: Final part of this meeting

• 19 I'll turn over to Donnie, and that's I believe t 20 requests from any members of the public, or anyone 21 else, to ask questions -

22 MR. GALLAGHER: Wait, Frank, did you have 23 a question? 24 MR. HUFNAGEL: Just a brief, if I may - 25 John Hufnagel here - just a brief comment that it goes 26 without saying, but I I'1be working with Donnie to try 27 to coordinate the next three weeks such that as he's 28 working on pulling. together the notes from this 29 meeting, and we're working on providing the 30 information as we understand it, that there will 31 hopefully be a brief period where we can check what 32 we've done against the meeting notes prior to us 33 sending it in. NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. WASHINGTON. D.C. 20005-3701 (202)f 234-4433 (202) 234-4433 q v

71 1 So we'll obviously need to coordinate to 2 do that. 3 MR. GILLESPIE: That's why we're going to 4 do everything we can to get these notes out pretty 5 quickly for everybody whose participated in listening 6 in on the meeting. 7 MR. HUFNAGEL: Thank you. 8 MR. GALLAGHER: Thanks. 9 MR. GILLESPIE: And Donnie, now I think 10 it's time to ask - 11 MR. ASHLEY: I would like to continue on 12 because we only have the phone for a short period of 13 time, and I don't want to lose the people that are on 14 the bridge.

.15                    Can I go ahead, Frank?

16 MR. GILLESPIE:.Go ahead.. 17 MR. ASHLEY: We've got a little bit of A18 housekeeping for the purposes of the transcript that -19 I need to;.take care of. I need to verify the spelling 20 of your names for the people who are on the telephone. 21 bridge. And in particular order, Ron Zak with the New 22 Jersey DEP, would you spell your name for me, please? 23 MR. ZAK: Z-a-k. 24 MR. ASHLEY: Tom Quintenz from Oyster 25 Creek? 26 MR. QUINTENZ: Q-u-i-n-t-e-n-z. 27 MR. ASHLEY: Thank you. 28 Nick Clunn with the Astbury Park Press, 29 would you spell your name please for the reporter? 30 MR. CLUNN: C-i-u-n-n. 31 MR. ASHLEY: Thank you. 32 Mr. Webster? 33 MR. WEBSTER: Richard, R-i-c-h-a-r-d NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W. (202)234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433

72 1 Webster W-e-b-s-t-e-r. 2 MR. ASHLEY: And your organization, sir? 3 MR. WEBSTER: Directors Environmental Law. 4 MR. ASHLEY: Thank you. 5 Mr. Brown, Jeff Brown? 6 MR. BROWN: B-r-o-w-n. 7 MR. ASHLEY: And your organization, Mr. 8 Brown? 9 MR. BROWN: Is G-r-a-m-m-e-n. 10 MR. ASHLEY: Thank you. 11 Ms. Gotsch? 12 MS. GOTSCH: G-o-t-s-c-h, same 13 organization. 14 MR. ASHLEY: Thank you. 15- Mr. Atherton? 16 MR. ATHERTON: A-t-h-e-r-t-o-n. 17 MR. ASHLEY: And you represent? 18 MR. ATHERTON: I'm working with Jersey h 19 Shore Nuclear Watch. 20 MR. ASHLEY: Thank you, sir.. 21 Ms. Gbur. 22 MS. GBUR: G-b-u-r, Jersey Shore Nuclear 23 Watch. 24 MR. ASHLEY: Thank you. 25 Mr. Warren? 26 MR. WARREN: W-a-r-r-e-n, and I'm also with 27 Jersey Shore Nuclear Watch. 28 MR. ASHLEY: Thank you very much. 29 Is there anyone that came on the line that 30 I didn't mention your name? 31 MR. LAIRD: Name is Jim L-a-i-r-d, Exelon. 32 MR. ASHLEY: Thank you, Mr. Laird. 33 MR. PINNEY: My name is Richard Pinney. P NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 2344433 WASHINGTON, D.C. 200054701 (202) 234-443

73 1 as in Paul -i-n-n-e-y, New Jersey DED. 2 MR. ASHLEY: Anyone else that we didn't 3 recognize? 4 In the interest of having an opportunity 5 for the people that are on the phone bridge, is there 6 anyone who would like to ask the staff a question, or 7 to make a statement at this time? 8 MR. ATHERTON: My name is Atherton. I have 9 a technical background in technical and nuclear 10 engineering. And the first complaint I have is, half 11 the conversation I heard was inaudible. And I didn t 12 know whether it was bad technology in the electronics 13 that you have for transmitting this, or some other 14 cause. And I did phone the public affairs office to 15 complain about that, and I was hoping you got the 16 message. 17 But toward the end of the conversation you .18 - were slightly more audible. So I missed out on a -lot. 19 And I do have a couple of questions I'd like to ask or 20 get clarification for. Is that possible? 21 MR. ASHLEY: Go ahead, Mr. Atherton. 22 MR. ATHERTON: I'm going to back up to the 23 specifics concerning the issue of uncertainty and 24 sensitivity analysis and the like. 25 The basic question would be, is there the 26 potential, since I didn't catch all the information 27 that was taking place back and forth, is the potential 28 for harm to the shell or the liner significant enough 29 with the uncertainties involved so that it would be 30 better not to use uncertainty as a sole means of 31 analyzing the situation, but to approach it from the 32 worst case analysis perspective; and if so, why? 33 MR. GILLESPIE: Yeah, this is Frank NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

74 1 Gillespie. 2 MR. ATHERTON: You're barely audible. I 3 heard the Frank. 4 MR. GILLESPIE: This is probably because 5 we're using 20-year-old technology for our phone 6 system here. 7 .MR. ATHERTON: And how did you spell your a last name, sir? 9 MR. GILLESPIE: Gillespie, G-i-l-l-e-s-p-i-10 e. 11 MR. ATHERTON: Okay. 12 MR. GILLESPIE: The context of this meeting 13 was very incremental, in addition to a lot of 14 information that we've already gotten in the request

• 15 for additional information.

16 And in some ways, if you - have you read 17 all the additional information:that's been sent in to

18 us that's been made available?.

19 MR. ATHERTON: Unfortunately I haven't had 20 the opportunity to do that yet. I just received a 21 disk a couple of days ago, and I haven't had the 22 opportunity to go through that yet. 23 The general question concerned, I doubt 24 the information that 3 'm seeking is going to be on the 25 disk, because I'm questioning whether you should use 26 uncertainty analysis versus worst case analysis. 27 MR. GILLESPIE: Well, to some degree, I 28 think you' 11 find in the applicant' s information, and 29 this is a little beyond the narrow scope of this 30 meeting,. but in general, in the applicant's 31 .information, there are discussions about measurements 32 taken in the'upper portion of this light bulb shell, 33 which is uncoated, which presents a - any application NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 2000D-3701 (202) 234-443

75 1 it would make a case, it presents a case that is far 2 less conservative than the bottom section of the shell 3 which is uncoated. 4 And so there are some assumptions on rates 5 where projections are made where exactly what you're 6 saying I think has been taken into consideration. 7 Now what could be up for discussion is 8 different people's view of what worst case is. And 9 you have to go through the material and give me a 10 specific, but it's really kind of a blend of, we 11 basically have an estimate line, and the estimate 12 comes from various data sources that get combined to 13 make the estimate. 14 And we're trying to have the highest 15 possible confidence in the estimate and the calculated 16 projections. And-.the projections have been made; the 17 measurements have been made. - And that 's why the focus 18 of a lot of -the --discussion. here :was the residual 19 questions on the part of the .staff-to ensure that we 20 understand the -uncertainties. involved in that 21 projection. 22 But that projection involves some 23 assumptions on corrosion rates which some people would 24 say in their minds is worst case of the situation in 25 the environment of the facility. 26 So I think both in different viewers, 27 different readers' views, have probably been done, and 28 we're wrestling with that total decision right now. 29 So it's not uncertainty .is everything or 30 nothing; it just happens to be our residual concern. 31 (Telephone operator voice interrupts) 32 MR. ASHLEY: Mr. Atherton, are you still 33 with us? NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

76 1 MR. GILLESPIE: Anyway, for whoever was 2 listening. 3 MR. ASHLEY: Just a second, Frank? Is 4 anyone still on the line? 5 (Loud telephone noise) 6 MR. ASHLEY: They cut us off. 7 MR. GILLESPIE: We had inadequate safety 8 margin in our bridge. 9 (Technical interruption) 10 MR. ASHLEY: We'll try to pick up Mr. 11 Atherton as he comes back on. 12 Did anyone else have a comment so we can 13 continue on? 14 MR. ATHERTON: Hello. 15 MR. ASHLEY: Yes.  : 16 MR. ATHERTON:. This is Peter Atherton... I 17 don't want what: happened. But- I suddenly got 18 disconnected during Mr. Gillespie.'s part. 19 MR. ASHLEY: We did to. We're glad to have

• 20 you back again.

21 MR. GILLESPIE: Go ahead. 22 MR. ATHERTON: Well, Mr. Gillespie was 23 talking about the use of a version of the worst case 24 analysis for a bottom uncoated part of the containment 25 structure or the shell. 26 MR. GILLESPIE: The bottom part - 27 MR. ATHERTON: And that's where I lost you. 28 MR. GILLESPIE: The bottom part - and this 29 is difficult, because what we've got is a staff here 30 that's gone through literally thousands of pages of i 31 documentation to come down to these residual comments. 32 And in going through that there are i 33 estimates made with corrosion rates that are believed NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4M WASHINGTON, D.C. 20005-3701 (202) 234-4433

77 1 by the applicant - and this is a finding we're trying 2 to make - is believed by the applicant to be 3 reasonably conservative in nature. 4 And there is a coating on the bottom 5 portion of this light bulb fixture containment, and 6 they have measurements from the top part of the 7 containment, which is uncoated, but in a similar 8 environment on the inaccessible side. 9 And I believe the applicant has made some 10 projections using this, and then making the case that 11 the coating really provides this uncoated area 12 measurements are in essence a worst case in their 13 projection. 14 And therefore we've looked at that as a 15: staff, and all their information. And this 16 information was. really:'focusing on the uncertainties 17 that were connected to. that projection.. 18:. It's not that-.we' re. making judgments on 19 the uncertainties, but we're trying to make sure that. 20 we have the soundest possible number and a good 21 understanding of what could be viewed by some as a 22 worst case projection. 23 Now others could view this projection and 24 the numbers used as not being the worst case, and so 25 I'm very hesitant to use the word, worst case. 26 It's a projection that I think is 27 generally believed, actually representing a 28 measurement in an environment that is more harsh than 29 the environment it's being applied on to a carbon 30 steel piece of metal. 31 And that's what' s in the application. And 32 so this meeting is trying to deal with making sure 33 that when we make whatever judgment we need to make, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

78 I 1 that we understand what the pluses and minuses 2 connected with that are. 3 And so the staff has actually read the 4 application, and so we had that part done, and we 5 really weren't questioning the rate. We were 6 questioning the uncertainties around it to make sure 7 we could make an appropriate finding. 8 MR. ASHLEY: Thanks a lot, Frank. 9 Mr. Atherton, you still with us? 10 MR. ATHERTON: Yes, can anybody hear me? 11 MR. ASHLEY: Yes, sir. 12 MR. ATHERTON: I'm having connection 13 problems. 14 Let me back up just a little bit farther. 15 On a very general or holistic view of the containment 16 structure, the plant was approved originally to last 17 40 years. That essentially meant back in those days, 18 the '60s and .'70s!,, that the major components of the 19 plant would not fail for a total of 40 years. 20 We're seeing the drywell apparently 21 degrade prematurely which was not anticipated 40 years 22 ago. 23 The projecting that type of discovery into 24, the future for 20 more years, how are we to know as 25 members of the public that you're going to have 20 26 good years left on the material that was supposed to 27 last 40 years and hasn't? 28 MR. ASHLEY: Who's speaking? 29 MR. ATHERTON: My name is-Atherton. 30 MR. ASHLEY: Okay, go ahead. 31 MR. GILLESPIE: Well, that's exactly the 32 finding we're being asked to make as part of the 33 license renewal. NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

79 1 And first I would refute your assertion 2 that every component in the plant was designed to last 3 40 years. 4 In the basic underlying premise of 5 operation is a large number of surveillances, tests 6 and inspections. And the intent is that the structure 7 and the license be safe for the term of the license, 8 and that includes special tests and analysis, which 9 would detect, prior to violating or causing a safety 10 issue, the degradation of components. 11 And what we're really talking about is 12 taking that same principle and pushing it forward 13 another 20 years. In fact, many of the components in 14 the plant have seen a less severe environment than 15 they were projected in their original design. 16 And it's :that baseline and moving it 17 forward;i that we.'re doing. with renewal, which is why 18 there are extra ,commitments in the overall renewal 19 effort to .extra special tests and analysis. 20 The intention is not to say it will last 21 20 years; that's an economic issue. It's to say that 22 the licensee has processes and procedures in place 23 that we can inspect and that they can follow that will 24 detect and remediate anything that would cross a 25 safety margin. 26 And that's a different statement than 27 saying, we're saying it will last 20 years. In fact 28 if they would do a test and do a projection in 29 accordance with our interim staff guidance for the 30 renewal period, see water, and do an event test and 31 find out they were approaching minimum wall thickness, 32 they have to do an operability analysis under the 33 current requirements, which also project forward. And NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 200054701 (202) 234-4433

80 1 they have a decision to make to either repair or shut 2 down. 3 And instances of this we have in other 4 cases in pressurized thermal shock where we're 5 evaluating licenses for 20 years where the pressurized 6 thermal shock analysis for other licensees will not 7 make it to 20 years. But there is a requirement in 8 the rules that if you don't make it you shut down, or 9 you can replace your vessel. 10 And so it's not saying everything will 11 last the period of the license; it's saying the plant 12 will operate safely for the period of the license, and 13 we have reasonable assurance of that. 14 MR. ASHLEY: Thanks, Frank, I appreciate 15 that. 16 Mr. Brown or Ms. Gotsch, do you have a 17 question or: comment? 18 MsMs-Gubr;, are you on- the line? Did you 19 have a. question or comment? 20 MS.. GUBR.: I have a question. In the 1996 21 inspection report -- 22 MR. GILLESPIE: The 1996 inspection report? 23 All the -- that's actually beyond the scope of this 24 meeting, and our general counsel is here. And I 25 understand that that is tied up in the litigation 26 issues right now. 27 All of the NRC's information that we have 28 from 1996 in the NRC inspection reports are public 29 information. The licensee's information, which the 30 NRC at this time does not and has not possessed, is 31 actually tied up in the litigation right now, and 32 we're really not in a position to comment on that. 33 MR. ASHLEY: Go ahead, Mr. Webster. NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

81 1 MR. WEBSTER: Okay, great. 2 With regard to the drywell liner and the 3 UT measurements, I guess I'm somewhat surprised that 4 the licensee had already known that the '96 results 5 weren' t good, but nonetheless based predictions 6 forward on those '96 results. It seems to me, though, 7 that the QAQC for those results should have identified 8 the level a long time ago, so I'd just like a 9 clarification of why the rejecting wasn't treated 10 closer to the time. 11 MR. GILLESPIE: This is Frank Gillespie 12 with the NRC. And since this is really an opportunity 13 for people to ask the NRC for clarification on what we 14 said, I will answer from the NRC's perspective that 15 right now. the: people sitting in this room were 16 generally not involved in the details of what happened 17 in 1996. -  : 18 But in looking at that anomaly, I think it 19 would be unfair. to say :that that was - I forget what 20 your word. was -, but IIll .. use an irrelevant 21 measurement. It was a measurement, as we heard from 22 the licensee at this meeting, they looked into it and 23 examined it. They saw it as anomalous. But there was 24 really no reason probably at the time to either 25 exclude it or not include it. 26 MR. WEBSTER: There were three measurements 27 taken, and that '96 result was one of those three. If 28 you take that '96 result out of the'analysis the 29 uncertainties become huge. 30 MR.- GILLESPIE: And what I'm going to 31 suggest is, that's the exact question that staff has 32 just asked the licensee on uncertainties. 33 MR. WEBSTER: Absolutely, that's why I -- NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W. (202) 2344433 WASHINGTON, D.C. 20005-3701 (202)B 234-4433

                                                                              ....

82 1 MR. GILLESPIE: And so I'm just saying, I'm 2 not in a position, and I 'm not trying to put anyone in 3 a position to defend what was done over 10 years ago. 4 But because of the anomalous look at the results, 5 we're really focusing on removing that uncertainty 6 that we specifically pointed out as we project 7 forward. 8 So we fundamentally have just asked the 9 licensee to respond to that question. And we've, by 10 design at this meeting, asked the licensee not to feel 11 obligated to respond today to the staff's concerns. 12 So I guess we're in agreement. 13 One of our concerns you heard from Hans 14 Ashar and I were on the calibration techniques. And 15 I think the licensee-responded,. they recognize that 16 there.are certain::coatings and stuff that they have to 17 reallybe very, careful of when they're doing these, 18 and so-we have:.to see what they answer. 19 You're asking for the answer we've asked 20 for, and it'.s just not the right time for the answer 21 yet. 22 MR. WEBSTER: Now the second issue that I 23 think also relates to the questions you're asking is 24 about how the actual raw measurements get 25 incorporated. One of our concerns is that the 26 uncertainties in these measurements become hidden in 27 the way they're presented, because you take the 28 measurements, get an average and put into one 29 measurement, which is then put on a scatter graph. 30 And then. when you look at the scatter graph you don't 31 actually see the underlying uncertainty. All you see 32 is some scatter of averages, which is much less than 33 the actual scatter and the underlying results. NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202)234-4433

83 1 Now one of the concerns I have, and we've 2 reviewed these documents from the licensee, and it 3 seems that they're editing the data, that they omit. 4 They actually omit an outlier from the analysis. And 5 again I think this is another way where the 6 uncertainty is made to appear lower than it really is. 7 MR. GILLESPIE: Let me try to answer that. 8 Now this is going to be dangerous. Because I was an 9 engineer 35 years ago, but I'm going to - Hans has 10 been training me for three weeks, Hans Ashar, who is 11 our expert. So let me take a shot at the answer. 12 One, you have to understand, we've 13 basically asked the same question that we need to have 14 a good understanding about how that lower level 15 combination of numbers .was done. 16 That-was & ;concern we had, and that's a. 17 question we asked. -" 18 Two,.. you also: have to differentiate; 19 there 's two phenomena- of- interest here. One is 20 pressure during an accident, and the other is 21 buckling. 22 And the interest in the buckling sense, 23 which is really the sandbed region interest, is 24 buckling down at the lower level, is one of general 25 area corrosion, a very broad degradation, and not one 26 of pitting. 27 In fact in any structural member you can 28 actually drill holes in it, and you do not 29 significantly reduce its structural strength. 30 And so knowing that principle I would not 31 want to draw a conclusion on information we don't 32 know. And that's why we've asked for information on 33 how they've done the statistical combination; what was NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433

84 1 their basis for whatever, throwing out outliers, in a 2 95 percent confidence interval. 3 But for the purposes of buckling, a 4 localized thinning spot is not a principal concern. 5 MR. WEBSTER: Well, I told you, I 6 understand that. But my point is that if you permit 7 that as part of the uncertainty analysis, then you 8 tend to regard the measurement -- 9 MR. GILLESPIE: Again, I don't know how 10 they've been included or how they've been admitted, or 11 has it followed standard practice. We've asked that 12 question, and I hope within the next month we'll have 13 a little more amplifying information, and I could give 14 you a more satisfactory answer. 15 We're sharing the same concern. 16 MR. WEBSTER: Absolutely. I understand. 17 I'm very pleased to see that we do share the same 18 question.  ; . 19 My present issue - 20 MR..-ASHLEY: Mr. Webster, this is Donnie 21 Ashley. You said you had two. 22 Hold the third one, and let me get 23 through, make sure we can touch base with everyone. 24 If we have time we'll come back to you. I have some 25 uncertainty about all four here. 26 Let me leave this - 27 MR. GILLESPIE: Just in case we get cut off 28 from everybody, email Donnie Ashley and we will get 29 back to you by email on any questions that we don't 30 get to, because our phone system doesn't seem to be 31 working as good as I'd like it to. 32 MR. ASHLEY: Thanks, Frank. 33 Mr. Clunn from the Astbury Park Press, do NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W. (202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433

85 1 you have any questions or comments? Nick Clunn? I 2 guess we lost him a few minutes ago. 3 Ronzak (phonetic) or Ridgepenny (phonetic) 4 with New Jersey DEP, any questions or comments from 5 you? 6 MR. PINNEY: No, we have no questions 7 here. 8 MR. ASHLEY: Okay. Dennis Zannoni, would 9 you like to come down to the podium? I would like for 10 you to go ahead so they can hear your comments as 11 well. 12 Mr. Warner, if you'd wait just one second. 13 MR. ZANNONI: Dennis Zannoni, Z-a-n-n-o-n-14 I. 15 I-'d Z also like to thank the Nuclear 16 Regulatory Commission for:having this meeting. I 17 think it's obviously necessary. 18 So having the next meeting if it's 19 conductedin -the afternoon would also help me, since 20 I have to drive up, since we're facing a very 21 substantial budget deficit in New Jersey as you 22 probably heard. 23 First, I want to mention that - and this 24 is mostly for Frank's edification, because he is 25 coming to our office I guess within two weeks with 26 some of his staff, to give you a little bit of the 27 flavor, of what we're going to talk about, and it does 28 relate to what we're covering here, and that is, there 29 is a little bit of confusion on the ruling made by 30 ASLB and its staff's attorneys, and it's mostly a 31 question directed at the attorney, that we would like 32 the NRC to clear up the fact that we are not a party 33 or involved with the contention on the liner or NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W. (202) 234-4433 WASHINGTON. D.P. 20005-3701 (202) 234-4433

86 1 drywell shell in any way. 2 And I guess ASLB made that clear, but some 3 kind of communication has come down the path, and it's 4 affecting our ability to do work, that we're somehow 5 tied up with that. 6 It would be nice if you could clarify that 7 here today, but I know you're not. 8 We're going to pick that up when we talk 9 too, because it is affecting what we're doing. We go 10 to meetings, and people are confused about what our 11 role is. 12 We do have three appeals to the 13 Commission, but they have nothing to do with the 14 liner. 15  ;- And.-we have a good reason for that,

16. because we have, our own staff that have made their own 17, conclusions, and I have to tell you, quite frankly, I 18 was at a meeting here discussing the same drywell line

• 19 issue when the company was going for a conversion from 20 the full-term operating license to the - or from the 21 provisional operating license to the full term 22 operating license, and it was only at the insistence 23 of New Jersey that they took very aggressive 24 protective corrective actions. I don't know if even 25 anybody here at the AmerGen table was here. But 26 removing the sand and all of that was very, very 27 positive, and we view that in a way that we thought at 28 the time was good for until April, 2009.

29 So our position right now, and Ron is 30 online, and he's our expert actually on the drywell 31 shell - he keeps telling me to call it a shell, not a 32 liner - is right now positive. And the rigor that I 33 see addressed here for that one issue, I wonder if NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (=0) 234-4433 WASHINGTON, D.C. 20005-3701 234-433 (MO)

87 1 that's going to spill into many, many other issues 2 that we feel an equal amount of rigor is needed. 3 Because you guys are going into some depth 4 here that we are going to talk about again to see if 5 it applies in maybe some other areas that could 6 benefit from that, more so than the liner. 7 Anyway, that said, we also need to have 8 some kind of - we don't know when the commission is 9 bound. If not, I understand it's not to make 10 decisions on the appeals that we submitted. Again, it 11 has bearing, because the more they wait, the less we 12 can interact with NRC staff on those specific issues. 13 And if they made a decision one way or the 14 other, then we could get on with it. So we'll 15 probably-submit that in writing, but I'm just giving

• 16 you .a flavor of some of the topics that we' re going to
• 17 :talk.about.
• 18 Now specific to this meeting. Frank, you 19 said earlier :in the meeting you said you may - the NRC 20 may recalculate something. And then later you said 21 they will recalculate something.

22 I just need to know, you are going to 23 recalculate something. What are you going to 24 recalculate? 25 MR. GILLESPIE: Our intention right now is 26 to do a comparative calculation to the GE calculation 27 of 1991. 28 MR. ZANNONI: The one with the disclaimer? 29 MR. GILLESPIE: The one with - well, that 30 was a piece of it. That report fed into the data that 31 went into that calculation, and our intention would be 32 to do kind of a comparative calculation. 33 Ours doesn't need to be as rigorous as NEAL e. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202)234-4433 WASHINGTON, D.C. 20006-701 (202) 234-4433

88 1 1 theirs, because we're doing it as a confirmatory 2 measure, not as a decision tool on their part. So 3 we're likely going to do that to get a perspective 4 ourselves on the conservatisms that have been assumed 5 in that calculation. 6 And so it's just an independent look. And 7 we do this in thermal hydraulics. We do it in a 8 seismic area. We do it in a lot of different areas 9 occasionally. 10 The other piece is, we have six more Mark-11 ls coming in, and so for the renewal group, we're kind 12 of setting a precedent. Because all of those same 13 questions exist on all of those same containments. 14 And so part of this calculation will be 0 15

• giving us knowledge to a specific operating history
• 16 and a specific calculation that GE did..

17 MR. ZANNONI: Is it going to be done in

18 house or contracted?

19 "* MR. GILLESPIE: Part of this meeting is not 20 discussing how the NRC will do this piece of the 21 review. 22 MR. ZANNONI: I'll ask it at some point in 23 the future. It tells what kind of depth you're going 24 to do which is pretty - if it's in house it's one 25 thing - 26 MR. GILLESPIE: Well, we're going to have 27 outside experts helping us. And any report that's 28 done will be public. 29 MR. ZANNONI: You mentioned, I guess for my 30 own information and information concerning New Jersey, 31 are there other - the rigor that you - the depth and 32 the rigor that I send that you're requesting from 33 AmerGen for Oyster Creek, have there been other plants NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE, N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202)

                                                                      *   .

234-4433 _w

1

       'I     that have similar drywells gone through similar rigor?

89 2 Or is this something new that you are going to ask 3 plants to take a closer look at that have already 4 gotten license renewal? 5 MR. GILLESPIE: There 's two questions. The 6 answer is yes, everyone else is going through a 7 virtually similar process. But everyone has different 8 operating histories. 9 I ' 11 give you a specific one. We 're going 10 to ACRS, Nine Mile Point. Nine Mile Point has an 11 operating history with no visual leakage. They also 12 have welds around their seals. And so seals, for 13 example, at bellows, are not an issue. 14 They have actual electronic alarm systems

• 15 on their drains. They actually have a float alarm on 16 - there is a ledge in there 'that goes under the seal.
• 17 And they put bore scopes up there and looked in with 18 :TV cameras and saw dust.

          ..

19 And so it's a form of:rigor, but it's a 20 different operating situation, and a; slightly 21 different design. So I would suggest that in essence 22 all the licensees with this kind of containment are 23 going through the same process, and the same level of 24 detail, and trying to be just as certain about their 25 projections, and the projections being used there, 26 they're taking them from the torus at the water level 27 where they do UTs, and it's a very aggressive area. 28 MR. ZANNONI: Plants that have already been 29 approved? 30 MR. GILLESPIE: No, this one is in house. 31 Brown's Ferry we did a similar rigorous review. And 32 they had some unknown leakages, and they committed to 33 an inspection regime. And theirs was the 10-year kind NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W. (202)234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

90 1 of one. And that's through, and that license has been 2 issued. 3 Brunswick does not have a shell; it has a 4 liner. And the design differende there is, the 5 structural elements, the concrete, is not the steel; 6 the steel is basically a seal. 7 And so the answer is yes. Now the 8 difference here is, the visibility of Oyster Creek is 9 different than the others. And so a lot of what we do 10 with these other facilities is closer - you know what 11 I mean -- it's not quite as visible. 12 So every one is going through, you could 13 say, an equal type of review, customized to their 14 operating history, the operating conditions and the 15 past events. 16 - MR. ZANNONI: I know Donnie is going to cut 17 me off. But just one last ,comment :for the public 18 that's listening, I know Peter Atherton did mention .19 about confidence that the public, is looking for, not 20 only in this issue but all of license renewal. 21 I'll just throw out, and I always mention 22 this, that in addition to AmerGen's huge workload to 23 meet all the requirements - they got the NRC looking 24 at it - we also as a state have a group of about 15 to 25 20 professionals, I already mentioned that we have a 26 very sound expert in structural stuff on staff who has 27 worked with Oyster Creek for awhile. And this hearing 28 if anything comes out of it, hopefully it will be 29 positive. 30 So the net result here, and I don't want 31 anybqdy .to miss this, and it's too bad the press 32 wasn't here, is that this is getting a lot of eyes and 33 a lot of attention. So that has to give the public NEAL R.GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON. D.C. 200054-701 (202) 234-4433

91 1 some sense of, they're not alone in this process. 2 So that's why I exist just to put it 3 bluntly, so thanks. 4 MR. ASHLEY: Thank you, Mr. Zannoni. 5 MR. ASHAR: This is Hans ASHAR, NRC. 6 Let me say that for the general analysis 7 purpose, the applicant has taken an approach where 8 they are taking an average, but in addition to that, 9 they also do the discontinuity analysis for the thin 10 areas. Thin areas are where there are small sparks 11 which might have been missed in averaging they might 12 have counted as thin areas, but they have taken a 13 number of places which are thin, and they have 14 analyzed separately to understand the discontinuity 15 stresses and their ability to withstand the loads 16 they're supposed to withstand. 17 MR. GILLESPIE: Okay, this is Frank 18 Gillespie. !Let me amplify a little more. Because now 19 I'm-going to take what Hans just said and say, that's 20 also part of the actual sample of the smaller area 21 that's scanned. 22 This is a very, very, very large vessel, 23 and the representative nature of the sample that was 24 earlier worked on with literally thousands of 25 measurement points by the applicant to ensure that 26 even those areas that are scanned, and the 49 points 27 that are averaged, are the right areas to be scanned. 28 And that's why we did ask an additional 29 question here to reconfirm right now the 30 representative nature of those areas, exactly so you 31 couldn't get a substantial elongation or a major flaw. 32 So there's two things. One is the 49 33 points, which is a smaller, very small area, and the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) v . 2344433 v .

                                                                             !

92 1 other is the location of those small areas through the 2 vessel itself. 3 And if you would look at the much earlier 4 data of all the thousands of points that were done and 5 reviewed by the NRC, it's that rbpresentative nature 6 that actually covers your large perforation kind of 7 question. It's not the 49 measurement points which 8 were averaged, for maybe a 6 by 6 inch kind of area. 9 MR. ASHLEY: Thanks, Frank. 10 In closing I appreciate everyone's 11 participation. I appreciate -- I'm sorry, we're going 12 to be out of time, and the phone is going to shut you 13 off in about two minutes. 14 But we do appreciate everyone ' s coming out 15 to participate in this meeting. :And again, if you 16 need additional information, or if you have questions, 17 send me email. Mycemail address.is on the website. 18 And once again, thanks to everyone, and 19 we'll- adjourn rat°this point. 20 MR. GILLESPIE: Thank you. 21 (Whereupon at 11:58 p.m. the 22 proceeding the above-entitled matter went off the 23 record to return on the record at 11:58 a.m.) 24 MR. GUNTER: That's all right. This is 25 Paul Gunter, G-u-n-t-e-r. 26 I'm with Nuclear Information Resource 27 Service. 28 There's a whole lot of questions, and I'm 29 sorry that Richard Wester wasn't able to complete, but 30 we'll go ahead and supplement the record by email. 31 And I guess that could be incorporated into the 32 transcript as well? Can we have email questions 33 incorporated into the transcript? NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON. D.C. 2000"-3701 (202) 234-4433

  *   °

93 1 MR. ASHLEY: I don't think we can have 2 email questions in the transcript. But we can include 3 it in the summary. We'll put it in the meeting 4 summary. 5 iki.GUNTER: Okay, that's fine, that's fair 6 enough. 7 MR. GILLESPIE: And our meeting summaries 8 are all put on our website. 9 MR. GUNTER: You know for the sake of time 10 I'm just going to ask one question here, and it gels 11 back earlier to a comment that Frank made with regard 12 to the 1990 GE report, and the assumptions that went 13 into the corrosion and degradation. 14 I thought I heard you say that the NRC has 15 - they've identified a degradation uncertainties 16 within that GE report. Was that correct? Was I 17 correct in hearing that? 18 .. ..And I: think that was the basis of your. 19 going back and doing the recalculations; right? 20 So. I'm asking first of all for 21 clarification on what you've identified in the GE 22 report that raised degradation uncertainties. And if 23 you could identify those for us right now. 24 MR. GILLESPIE: Okay, I'm not sure how much 25 detail Hans is in a position to go into. It was an 26 accumulation fo fundamentally the underlying 27 assumptions that went into it. And they appear to be 28 conservative, but one of the only ways to test the 29 overall conservatism of the assumptions is just to do 30 a calculation with an independent person making an 31 independent view of it. 32 But Hans, you did that? 33 MR. ASHAR: Yes, if you heard us on the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON. D.C. 20005-3701 r2021 234-4433

94 1 first or second questions that we had for the 2 applicant, you might have heard that we requested the 3 applicant to at least clarify as to what has been said 4 in their statistical inference report that is attached 5 to the GE teport by the way they interpreted the 6 measurements, and how they statistically put together, 7 both that particular report findings were used, or 8 some other metrics were used. That was our question 9 to them before, and I'm looking for those answers. 10 MR. GUNTER: Right. So it's not so much 11 that you're questioning the degradation mechanism 12 itself? 13 MR. ASHAR: No. 14 MR. GUNTER: So one of our concerns is 15 that, for example, :,-I think it's been referenced here 16 a number. of times that there was - in order for the 17 - sandbed wegion. to -be - for the UT to resume at the. 18 sandbed, .there was the event trigger for the presence 19 of water. 20 .:But it's always been our concern that - 21 there was I believe a '95 exemption that provides for 22 a 12-gallon-per-minute leak rate, and that constitutes 23 what we believe to be a significant event. 24 So during the refueling outages, there is 25 this "95 exemption that provides, to reiterate, 12 26 gallons a minute leak rate. 27 So it's been a question for us why we've 28 not seen this reevaluation with UT at the sandbed, and 29 more particularly for the embedded region, so I think 30 it 's been raised here this morning that there needs to 31 be a closer look at a number of areas for the 32 reevaluation with UT. Crevice corrosion should be one 33 of these areas, we believe. And I don't know what NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 2344433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

95 1 level of confidence we have on the seals around - 2 between the steel liner and the concrete. But I think 3 that it's reasonable that we shouldn' t be relying upon 4 - that these seals are necessarily going to be high 5 confidence seals. 6 So as you are looking at the ledges that 7 were raised here today, we would strongly advise that 8 the UTs be resumed at the levels below the sandbed 9 region. 10 Hans, do you think that that is a 11 reasonable request? 12 MR. ASHAR: Well, because this area is not 13 accessible from any side, there is a state of the art, 14 which is not being used by so many people. And we 15 recommended its use if they can do that.

 -  16     S:,              So we *are trying to understand from them 17   -what .they,'are going to do to gain the confidence that

• 18 *that.*area'is being considered in a sample size.
• 19 - * *MR. .GILLESPIE: I'd also like to say, we.

20 have a broader level of operating experience than just 21 Oyster Creek. And so we do have some sense, and a 22 generic idea of - there are some licensees who 23 actually went in and chipped concrete up and did some 24 measurements. Not all of them. They did that in a 25 response to the generic letter in 1987 we put out. 26 The other element is, we do kind of have 27 an understanding of the environment. But we need the 28 applicant to tell us what that environment is, and why 29 it's okay. 30 They're going in and looking at the 31 coatings in those areas. Basically you've committed 32 to verifying those 100 percent, and a third, as you've 33 been doing each time, for the three bays each time, or NFAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

96 1 something. 2 I don't know the details of that, and when 3 the little person goes in this gap and does this 4 inspection, whether they can eyeball the seals or not. 5 MR. GUNTER: Again, I've not seen a 6 commitment to the seals. 7 MR. GILLESPIE: Okay. I'm going to leave 8 it to Hans, as the expert, to say whether we need a 9 commitment to that. 10 The other thing is, at least in the prints 11 I saw, when we looked at the drain arrangement without 12 the sand, it looked like the low points is where the 13 drains were located in the sandbed area. 14 So there are some actual physical 15 .limitations on the accumulation it appears of water,.

• . 16 that actually could accumulate by those seals.

17 - We're asking the licensee to come in and 18 put all7of these things together in this integral 19 discussion of this area that is sandwiched with 20 concrete. 21 It's more than just the chemistry that I 22 mentioned we talked to ACRS about. And so that's on 23 their plate to explain it. 24 It may not be everything that someone else 25 may want, but we're charged with making an adequate 26 protection or reasonable assurance finding, and we do 27 have like I said other operating experience from other 28 plants, so we're not totally isolated here. 29 Yes. 30 MR. ZANNONI: I think someone in the room 31 knows the answer to this question, but is water an 32 intrusion on this vessel part of license renewal 33 space? I was told it wasn't. I mean it could leak, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (202) 234-4Mq WASHINGTON, D.C. 20005-3701 (202) 23"-M*

97 1 it could flow, but it doesn't have a basis in license 2 renewal space. 3 MR. GILLESPIE: Let me say it this way. 4 MR. ZANNONI: I was told that it did. 5 M. GILLESPIE: The component is large, the 6 component corrodes, and the component has a safety 7 function. 8 That means the component is part of 9 license renewal and has to be addressed. In fact that 10 means it has to have an aging management program. 11 And if the water is allowed then the aging 12 management program has to be such that it ensures the 13 14 component's safety function will not be compromised with the water there. fi 15 ."  : And so the water leakage is not part.:of

• 16
• renewal.

17 But the environment, which is a high 18 corrosive environment that the water creates, is part E

• "* 19 of license renewal. And so that's really why we 're 2.0 talking. Because part of the general solution for 21 most licensees - and I'll get off Oyster Creek now -

22 most licensees are using is a combination fo coatings 23 - we just did Monticello with ACRS - they have a 24 primer coating on the external surface. So it's a 25 combination of coatings, leak control and leakage ( 26 monitoring. 27 Both leak control and leakage monitoring, 28 which put their seals in scope, because they said, 29 okay, part of our aging management program for this 30 environment is the seals, and we're not going to have 31 leakage in the seals, so we'll have highly reliable - 32 and so no. 33 But certainly the absence of water makes NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (M0)234-4433, WASHINGTON, D.C. 20005-3701 (202) 234-4433

98 1 aging management far easier. 2 MR. ZANNONI: That's a helpful 3 clarification. 4 MR. GILLESPIE: Thank you, Mr. Zannoni. 5 Let me not make thio mistake again. Is 6 there anyone else who has a question or a comment in 7 the room? 8 Mr. Recorder, you can turn it off. 9 10 (Whereupon at 12:08 p.m. the proceeding in the above-11 entitled matter was adjourned) J J I NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. (2.02)234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433

Citizen's Exhibit NC5 Citizen's Exhibit NC5 TiNuclear NCalculation Sheet Subjct STATISTICAL ANALYSIS OF DRYWELL Cuic No. Rev. No. Sheet No. THICKNESS PATA TERU NOVEMBER 1991 C-1302-187-300-019 1, 0 A I of 39 OdgInatarzw ll ets flevle=e4b 12012091 Date

            ./   _:P. Moore, dr.                                                      I--'l 6/                            Verification: V-1302                    Rev. 6 1        TIROBLEX STATEMENT The basic purpose of this calculation is to update the thickness measurement analyses documented in References 3.7, 3.8,      3.11,   3.12,    3.13 and 3.14 by incorporating the measurements taken in November 1991.               Since no measurements were taken at the 87'-5" elevation in November 1991 due to high temperatures at that elevation, the results for the May 1991 measurements at 87'-5" (Ref. 3.14) are included for completeness.

Specific objectives of this calculation are: (1) Determine the mean thickness at each of the monitored locations. (2) Statistically analyze the thickness measurements to determine the corrosion rate at each of the monitored locations. 2.0

SUMMARY

OF RESULTS The results of the calculation are summarized in the following tables. The terms used are defined below. (1) Best Estimate Corrosion Rate 0 With three or more data points, this is the slope of the regression line. For only two data points, this is the slope of the steepest line which can be drawn within the +/- one-sigma tolerance interval about the mean. (2) 95% Upper Bound Corrosion Rate The corrosion rate for which we have 95% confidence that it is not being exceeded. At least four data sets are required to make a meaningful estimate of this value. OCLROO000174

9, Calc. No. C-1302-187-5300-019 Rev. 0 Page 2 of 39 (3) Best Estimate Mean Thickness When the regression is statistically significant (F-Ratio is 1.0 or greater), this is the predicted value +/- standard error from the regression for the date of the last measurement. o When the regression is not statistically significant (F-Ratio less than 1.0), this is the grand mean of all the data +/- standard error. (4) Measured Mean Thickness The mean +/- standard error of the valid data points from the most recent set of measurements. (5) F-Ratio o An F-Ratio less than 1.0 occurs when the amount of corrosion which has occurred since the initial measurement is less than the random variations in the measurements and/or fewer than four measurements have been taken. In these cases, the computed corrosion rate does not necessarily reflect the actual corrosion rate, A and it may be zero. However, the confidence interval about the computed corrosion rate does accurately reflect the range within which the actual corrosion rate lies at the specified confidence level. 0 An F-Ratio of 1.0 or greater occurs when the amount of corrosion which has occurred since the initial measurement is significant compared to the random variations, and four or more measurements have been taken. In these cases, the computed corrosion rate more accurately reflects the actual corrosion rate, and there is a very low probability that the actual corrosion rate is zero. The higher the *-Ratio, the lower the uncertainty in the corrosion rate. o Whereas an F-Ratio of 1;0 or greater provides confidence in the historical corrosion rate, the F-Ratio should be 4 to 5 if the corrosion rate is to be used to predict the thickness confidence inin the future. the predicted To have a the thickness, highratio degree of should be at least 8 or 9. OCLROO000175

Calc. No. C-1302-187-5300-019 Rev. 0 Page 3 of 39 The number of data sets used in the analysis. (7) Years The time span between the first and last of the analyzed data sets. 0CLROO000176

                                                             ...-.        rr..-        -   --             -         *w        -i--                    -r.....
                                                                                                                                                          ---- --. F Calc. No. 1302-187-5300-010 Rev. 0 Page 4 of 39 2.1 land Red bgton Nil Data Thru Nownw IMt Bay &Area                    CoIoTlSn Nita, mpy                                      Mesa Th1ckaMls                             F--Retoli    Me)     Data $pon, YRSM7 Bestdlul)             j      5 m(2              Best Et.3) r                         Masurd (4) 0*               11.3+/-,2.2                         -15.8        994.4,38                       992.4W ,10.2                     4.0         7         2.9 11A               182 k1.7                             21.2      837.3,8.0                      832.0 A.4                       234         13         4.5 11C Top          *24.7 4.                            .33,2        959.5   13.0                    843U    24.1                    6.8        12         4.5 11C Bet

• 17.9 A2,5 22.4 948.9 8U5.3,4.5 10.3 12 4.5 13A 17.8 4.1 -264 839.9 *7.1 849.8 ,7.7 3.4 a 2.9 130 Top + 1.2 13M7 1053.8 A.2 1047.8 213.3 4 1.6 130 Rot + 2.8 AD,0 - 901.5, 2.9 899.56 7.8 4 1.8 15D - 4.3,2.2 8.6 10154.3 *2.4 1041.7 010.1 0.7 8 2to 17A Top 1.8, 1.8 -5 1129.1 A1.6 1122.8 +/-89 0.2 a 2.9 17A Bat
• 9,8,2.8 - 13.1 931.5 4.9 032.8 9.7 1.9 8 2.9 170 -19.7 1.8 .229 810.4 5.2 822.2 9.2 28.0 13 4.7 17110 Top
• 11,8,3.4
• 18.2 981.4,A.9 A541. *4.8 2.0 8 2.8 17M19 Bot
• 14.7 3.3 . 212 974.5 65.7 871.1 ,54 3.3 B 2.8 I1A
• 16.3,1.8 . 18.1 793.8 4.7 903.2 ,8.9 22.0 13 4.7 198
• 11.5 +/-.1
• 15.4 833.8 ,6.9 848.3 ,10.0 5.A 12 4.5 19C
• 17.0 2.1 . 20.8 813.2 +/-5. 822.3 108 13.0 12 4.5 0

0-C0 0) C0 0) C0 -4

a- It

     -if    -mint            -rz                                    rr           r-          r                                                          r r               U       u    rI OCac. No. 1302-187-5300-010 Rev. 0 Page 5 of 39 L.2 Elevation BW-2" Using Dti Thru November 1991 Bsy & Area                         Cerwisn Rate,                                  e    a      Mesn Thiknem MheF.Ratio()                                N(s)   cate Spon, 510-12                 .3.5    1.1                               5.4       7423 *2.4                          748.1 *2.0                    2.0          11        4.0 50 >Mail

• 1.7. Z* 8.0 759.0 *1.2 70.4 *1.8 <0.1 5 1.8 515 <Men *4.0 *65.6 173 707.8 *3.3 7078 *5.9 <0.1 6 1.8 13131 >Moon
• U ,6.1 *18.3 707.0 3.1 787A *1.5 <0.1 5 1.8 13131 sMoen .4.0 - 22,0 833, 4.5 89.1 *9.8 <0.1 5 1,0 1523 >Mean ,3.0 2.5 2. 704,1 *1.8 78.2 *I.0 0.1 5 1.8 15123 .:9Moen 3.8*2.2 - 1.5 735.7 *13 739.3 *4.2 U 5 1.8 2.3 Sevetien $1" Using D" Thrm November 1la1 BSy 5 Area Co oe Rate, r y Mean Thickness, Mis F.Relio(S) N(9) Date Span,

                                                       ....                                                                                                                   YRS Bant E"t4)                   95% (2)              Best ESL (3)                       Meosured (4) 3132 >705                1. *,2.1                         I    4.3       718. +/-1.1                         720,0 *0,0                     <0.1          4        1.5 13132 .095             .13* 1.2                               *6.2          92.8 *0.8                       882.1 *45                       0.1           4        1.5 ZA
          .eveton        87-5" using Data Thru May 181 Bey & Ar                            Correasien Rate, py                                        MWan Thickness, Mils                       F-Ratlo(SW    N(OS   Data Span, YRS(7)

Best Est*.ll 7[ X Beo ESt. 3 Maisured (4) J0 2 _ __0.7_* 3.5 514.3 *1.5 12.2 01. 1.4 7 3.5 1328 .1.3*1.2 .5OW 1634.1*2.7 6289 *3.4 0.95 7 3.5 1j .2.. 2 *A .5.4 1 o34.82.1 628.8 *1. 0.5 7 3.5

Caic. No. C-1302-187-5300-019 Rev. 0 Page 6 of 39 2.5 Evaluation of Individual Measurements Exceeding 99%/99% Tolerance Interval The following data points fell outside the 99%/99% tolerance interval and thus are statistically different from the mean thickness. sayJI Bev jAms ~POInt MilS [ Div. Sga 5 51 D.1 2 9 889 -60.9 .2. 15 61 23 28 850 .104.3 W.7 is 51 23 27 83 -11&.3 4.1 13 52 32 23 801 -101.8 -2.9 13 62 32 28 583 -139.6 4.0 13 B6 28 32 549 .7.7 4t to 85 31 34 559 .9.3 4.2 Evaluation of the data for each of these points indicate that none of them is corroding more rapidly than the overall grid. NOTE: Since no data was taken at the 86' elevation in November 1991, the results of the analyses of the May 1991 data at this elevation are listed above. (Ref. 3.14) OCLROO000179

Calc. No. C-1302-187-5300-019 Rev. 0 Page 7 of 39 2.6 Mean Thickness of All Points in the Grid The following table lists the mean thickness sigma for all the valid points in each 6"x6" grid. E--v Date anld 90 Sand Bed 11M91 687 *11.1 11A Sand Bad 1119 832.6 *8.4 liC Sand Bed 11191 902.5 *13.0 13A Sand Bed 11191 W48.6*7.7 130 Sand ead11101 06U

• 12.0 15D Sand Bad 11191 1041J *10.1 17A Sand Bed 11191 1014.1 k14.0 17D Sand Bed 11891 8222 A2 17119 Frame 11191 083.9 *39 10A Sand Bed 11191 9032 *08 19B Sand Bed 11191 145.3 &10.0

           ?8M   Sand Bed             11191  822.3 *108 6    61,         0.1 2    11191  749L *3.3 5    51          5        1191   742.4 4.3 13    51"         31       11161  742.8 *SA 15    51.         23       11191  754.3 *4.0 13    57          32       11191  7026 AO 80

Calc. No. C-1302-187-5300-019 Rev. 0 Page 8 of 39

3.0 REFERENCES

3.1 GPUN Safety Evaluation SE-000243-002, Rev. 0, "Drywell Steel Shell Plate Thickness Reduction at the Base Sand Cushion Entrenchment Region" 3.2 GPUN TDR 854, Rev. 0, "Drywell Corrosion Assessment" 3.3 GPUN TDR 851, Rev. 0, "Assessment of Oyster Creek Drywell Shell" 3.4 GPUN Installation Specification IS-328227-004, Rev. 3, "Functional Requirements for Drywell_ Containment Vessel Thickness Examination" 3.5 Applied Regression Analysis, 2nd Edition, N.R. Draper & H. Smith, John Wiley & Sons, 1981 3.6 Statistical Concepts and Methods, G.K. Bhattacharyya & R.A. Johnson, John Wiley & sons, 1977 3.7 GPUN Calculation C-1302-187-5300-005, Rev. 0, "Statistical Analysis of Drywell Thickness Data Thru 12-31-88" 3.8 GPUN TDR 948, Rev. 1, "Statistical Analysis of Drywell Thickness Data" 3.9 Experimental Statistics, Mary Gibbons Natrella, John Wiley & Sons, 1966 Reprint. (National Bureau of Standards Handbook 91) 3.10 Fundamental Concepts in the Design of Experiments, Charles C. Hicks, Saunders College Publishing, Fort Worth, 1982 3.11 GPUN Calculation C-1302-187-5300-008, Rev. 0, "Statistical Analysis of Drywell Thickness Data thru 2-8-90" 3.12 GPUN Calculation C-1302-187-5300-011, Rev. 1, "Statistical Analysis of Drywell Thickness Data Thru 4-24-90" 3.13 GPUN Calculation C-1302-187-5300-015, Rev. 0, "Statistical Analysis of Drywell Thickness Data Thru March 1991" 3.14 GPUN Calculation C-1302-187-5300-017, Rev. 0, "Statistical Analysis of Drywell Thickness Data Thru May 1991" 0 OCLROO000181

Calc. No. C-1302-187-5300-019 Rev. 0 Page 9 of 39 . 4.0 ASSUMPTIONS & BASIC DATA 4.1 Background The design of the carbon steel drywell includes a sand bed which is located around the outside circumference between elevations 8"-11-1/4" and 121-3n. Leakage was observed from the sand bed drains during the 1980, 1983 and 1986 refueling outages indicating that water had intruded into the annular region between the drywell shell and the concrete shield wall. The drywell shell was inspected in 1986 during the 1OR outage to determine if corrosion was occurring. The inspection methods, results and conclusions are documented in Ref. 3.1, 3.2, and 3.3. As a result of these inspections it was concluded that a long term monitoring program would be established. This program includes repetitive Ultrasonic Thickness (UT) measurements in the sand bed region at a nominal elevation of 11'-3' in bays 11A, 11C, 17D, 19A, 19B, and 19C. The continued presence of water in the sand bed raised concerns of potential corrosion at higher elevations. Therefore, UT measurements were taken at the 50'-2" and 87"-5" elevations in November 1987 during the I1R outage. As a result of these

• inspections, repetitive measurements in Bay 5 at elevation 50"-2" and in Bays 9, 13 and 15 at the 87'-5" elevation were added to the long term monitoring program to confirm that corrosion is not occurring at these higher elevations.

A cathodic protection system was installed in selected regions of the sand bed during the 12R outage to minimize corrosion of the drywell. The cathodic protection system was placed in service on January 31, 1989. The long term monitoring program was also expanded during the 12R outage to include measurements in the sand bed region'of Bays 1D, 3D, 5D, 7D, 9A, 13A, 13C, 13D, 15A, 15D and 17A which are not covered by the cathodic protection system. It also includes measurements in the sand bed region between Bays 17 and 19 which is covered by the cathodic protection system, but does not have a reference electrode to monitor its effectiveness in this region. The high corrosion rate computed for Bay 13A in the sand bed region through February 1990 (Ref. 3.11) raised concerns about the corrosion rate in the sand bed region of Bay 13D. Therefore, the monitoring of this location using a 6"x6" grid was added to the long term monitoring program. In addition, a 2-inch core sample was removed in March 1990 from a location adjacent to the 6"x6" monitored grid in Bay 13A. OCLROO000182

Caic. No. C-1302-187-5300-019 Rev. 0 Page 10 of 39 , Measurements taken in Bay 5 Area D-12 at elevation 50'-2" through March 1990 indicated that corrosion is occurring at his location. Therefore, survey measurements were taken to determine the thinnest locations at elevation 50'-21'. As a result, three new locations were added to the long term monitoring program (Bay 5 Area 5, Bay 13 Area 31, and Bay 15 Area 23). The indication of ongoing corrosion at elevation 501-2 1 raised concerns about potential corrosion of the plates immediately above which have a smaller nominal thickness. Therefore, survey measurements were taken in April 1990 at the 511-1011 elevation in all bays to determine the thinnest locations. As a result of this survey, one new location was added to the long term monitoring plan at elevation 511-10"1 (Bay 13 Area 32). Some measurements in the long term monitoring program are to be taken at each outage of opportunity, while others are taken during each refueling outage. The functional requirements for these inspections are documented in Ref . 3.4. The purpose of the UT measurements is to determine the corrosion rate and monitor it over time, and to monitor the effectiveness of the cathodic protection system. OCLROO0001 83

Cale. No. C-1302-187-5300-019 Rev. 0 Page 11 of 39 4.2 Selection of Areas to be Monitored A program was initiated during the 11R outage to characterize the corrosion and to determine its extent. The details of this inspection program are documented in Ref. 3.3. The greatest corrosion was found via UT measurements in the sand bed region at the lowest accessible locations. Where thinning was detected, additional measurements were made in a cross pattern at the thinnest section to determine the extent in the vertical and horizontal directions. Having found the thinnest locations, measurements were made over a 6"x6" grid. To determine the vertical profile of the thinning, a trench was excavated into the floor in Bay 17 and Bay 5. Bay 17 was selbcted since the extent of thinning at the floor level was greatest in that area. It was determined that the thinning below the top of the curb was no more severe than above the curb, and became less severe at the lower portions of the sand cushion. Bay 5 was excavated to determine if the thinning line was lower than the floor level in areas where no thinning was detected above the floor. There were no significant indications of thinning in Bay 5. It was on the basis of these findings that the 6"x6" grids in Bays 11A, 11C, 17D, 19A, 19B and 19C were selected as representative A locations for longer term monitoring. The initial measurements at 4 these locations were taken in December 1986 without a template or markings to identify the location of each measurement. Subsequently, the location of the 6"x6" grids were permanently marked on the drywell shell and a template is used in conjunction with these markings to locate the UT probe for successive measurements. Analyses have shown that including the non-template data in the data base creates a significant variability in the thickness data. Therefore, to minimize the effects of probe location, only those data sets taken with the template are included in the analyses. The presence of water in the sand bed also raised concern of potential corrosion at higher elevations. Therefore, UT measurements were taken at the 50'-2" and 87'-5" elevations in 1987 during the 11M outage. The measurements were taken in a band on L6-inch centers at all accessible regions at these elevations. Where these measurements indicated potential corrosion, the measurements spacing was reduced to 1-inch on centers. If these additional readings indicated potential corrosion, measurements were taken on a 6"x6" grid using the template. It was on the basis of these inspections that the 6"x6" grids in Bay 5 at elevation 50"-2" and in bays 9, 13 and 15 at the 87'-5" elevation were selected as representative locations for long term monitoring. OCLROO0001 84

Rev.0 e.Calc. No. C-1302-187-5300-019 Page 12 of 39 A cathodic protection system was installed in the sand bed region of Bays 11A, 11C, 17D, 19A, 19B, 19C, and at the frame between Bays 17 and 19 during the 12R outage. The system was placed in service on January 31, 1989. The long term monitoring program was expanded as follows during the 12R outage: (1) Measurements on 6"x6" grids in the sand bed region of Bays 9D, 13A, 15D and 17A. The basis for selecting these locations is that are but they not wereincluded originally considered in the system being for cathodic protection installed. (2) Measurements on 1-inch centers along a 6-inch horizontal strip in the sand bed region of Bays ID, 3D, 5D, 7D, 9A, 13C, and 15A. These locations were selected on the basis that they are representative of regions which have experienced nominal corrosion and are not within the scope of the cathodic protection system. (3) A 6"x6" grid in the curb cutout between Bays 17 and 19. The purpose of these measurements is to monitor corrosion in this region which is covered by the cathodic protection system but does not have a reference electrode to monitor its performance. The long term monitoring program was expanded in March 1990 as follows: (1) Measurements in the sand bed region of Bay 13D: This location was added due to the high indicated corrosion rate in the sand bed region of Bay 13A. The measurements taken in March 1990 were taken on a 1"x6" grid. All subsequent measurements are to be taken on a 6"x6" grid. (2) Measurements on 6"x6" elevation 50'-27: Bay grids 5 Area at5, the Bay following 13 Area 31,locations and Bay at 15 Area 2/3. These locations were added due to the indication of ongoing corrosion at elevation 50'-2", Bay 5 Area D-1. The long term monitoring program was expanded in April 1990 by adding Bay 13 Area 32 at elevation 51"-10". This location was added due to the indication of ongoing corrosion at elevation 50'-2" and the fact that the nominal plate thickness at elevation 51'-10" is less than at elevation 50"-2". OCLROO000185

Calc. No. C-1302-187-5300-019 Rev. 0 Page 13 of 39 4.3 UT Measurements The UT measurements within the scope of the long term monitoring program are performed in accordance with Ref. 3.4. This involves taking UT measurements using a template with 49 holes laid out on a 6"x6" grid with 1" between centers on both axes. The center row is used in those bays where only 7 measurements are made along a 6-inch horizontal strip. The first set of measurements were made in December 1986 without the use of a template. Ref. 3.4 specifies that for all subsequent readings, QA shall verify that locations of UT measurements performed are within +/- 1/4" of the location of the 1986 UT measurements. It also specifies that all subsequent measurements are to be within + 1/8" of the designated locations. OCLROO000186

Calc. No. C-1302-187-5300-019 Rev. 0 Page 14 of 39 4.4 Data at Plug Locations Seven core samples, each approximately two inches in diameter were removed from the drywell vessel shell. These samples were evaluated in Ref. 3.2. Five of these samples were removed within the 6"x6" grids for Bays 11A, 17D, 19A, 19C and Bay 5 at elevation 50'-2". These locations were repaired by welding a plug in each hole. Since these plugs are not representative of the drywell shell, UT measurements at these locations on the 6"x6" grid must be dropped from each data set. The following specific grid points have been deleted: Bay Area Points 11A 23, 24, 30, 31 17D 15, 16, 22, 23 19A 24, 25, 31, 32 19C 20, 26, 27, 33, 5 EL 50"-2" 13, 20, 25, 26, 27, 28, 33, 34, 35 The core sample removed in the sand bed region of Bay 13A was not within the monitored 6"x6h grid. OCLRO0000187

Calc. No. C-1302-187-5300-019 Rev. 0 Page 15 of 39 0 4.5 Bases for Statistical Analysis of 6"1x6"' Grid Data 4.5.1 Assumptions The statistical evaluation of the UT measurement data to determine the corrosion rate at each location is based on the following assumptions: (1) Characterization of the scattering of data over each 6"x6" grid is such that the thickness measurements are normally distributed. If the data are not normally distributed, the grid is subdivided into normally distributed subdivisions. (2) Once the distribution of data is found to be normal, the mean value of the thickness is the appropriate representation of the average condition. (3) A decrease in the mean value of the thickness with time is representative of the corrosion occurring within the 6"x6" grid. (4) If corrosion has ceased, the mean value of the thickness will not vary with time except for random errors in the UT measurements. (5) If corrosion is continuing at a constant rate, the mean thickness will decrease linearly with time. In this case, linear regression analysis can be used to fit the mean thickness values for a given zone to a straight line as a function of time. The corrosion rate is equal to the slope of the line. The validity of these assumptions is assured by: (a) Using more than 30 data points per 6"x6" grid (b) Testing the data for normality at each 6"x6" grid location. (c) Testing the regression equation as an appropriate model to describe the corrosion rate. These tests are discussed in the following section. In cases where one or more of these assumptions proves to be invalid, non-parametric analytical techniques can be used to evaluate the data. OCLROO000188

Calc. No. C-1302-187-5300-019 Rev. 0 Page 16 of 39 . 4.5.2 Statistical Approach The following steps are performed to test and evaluate the UT measurement data for those locations where 6"x6" grid data has been taken at least three times: (1) Edit each 49-point data set by setting all invalid points to "missing." Invalid points are those which are declared invalid by the UT operator or are at a plug location. (The computer programs used in the following steps ignore all "missing" thickness data points.) (2) Perform a Univariate Analysis of each 49 point data set to ensure that the assumption of normality is valid. (3) Calculate the mean thickness and variance of each 49 point data set. (4) Perform an Analysis of Variance (ANOVA) F-test to determine if there is a significant difference between the means of the data sets. (5) Using the mean thickness values for each 6"x6" grid, perform linear regression analysis over time at each location.

• (a) Perform F-test for significance of regression at the 5%

level of significance. The result of this test indicates whether or not the regression model is more appropriate than the mean model. In other words, it tests to see if the variation due to corrosion is statistically significant compared to the random variations. (b) Calculate the ratio of the observed F value to the critical F value at 5% level of significance. For data sets where the Residual Degress of Freedom in ANOVA is 4 to 9, this F-Ratio should be at least 8 for the regression to be considered "reliable" as opposed to simply "significant." (See paragraph 4.10.2) (c) Calculate the coefficient of determination (R2 ) to assess how well the regression model explains the percentage of total error and thus how useful the regression line will be as a predictor. (d) Determine if the residual values for the regression equations are normally distributed. 0L)ULUUUUU1tW

Calc. No. C-1302-187-5300-019 Rev. 0 Page 17 of 39 (e) Calculate the y-intercept, the slope and their respective standard errors, The y-intercept represents the fitted mean thickness at time zero, the slope represents the corrosion rate, and the standard errors represent the uncertainty or random error of these two parameters. Calculate the upper bound of the 95% one-sided confidence interval about the computed slope to provide an estimate of the maximum probable corrosion rate at 95% confidence. This is explained in greater detail in paragraph 4.10.2. (f) When the corrosion rate is not statistically significant compared to random variations in the mean thickness, the slope and confidence interval slope computed in the regression analysis still provides an estimate of the corrosion rate which could be masked by the random variations. This is explained in greater detail in paragraph 4.10.1. (6) Use the chi-square goodness of fit test results to determine if low thickness measurements are significant pits. If the measurement deviates from the mean thickness by three standard deviations, it is considered to be a significant pit. HE I OCLROO000190

rCaIc. Rev. 0 No. C-1302-187-5300-019 Page 18 of 39 4.6 Analysis of Two 6"x6" Grid Data Sets Regression analysis is inappropriate when data is available at only two points in time. However, the Analysis of Variance F-test can be used to determine if the means of the two data sets are statistically different. 4.6.1 Assumptions This analysis is based upon the following assumptions: (1) The data in each data set is normally distributed. (2) The variances of the two data sets are equal. 4.6.2 Statistical Approach The evaluation takes place in three steps: (1) Perform a univariate test of each data set to ensure that the assumption of normality is valid. (2) Perform an F-test at 5% level of significance of the two data sets being compared to ensure that the assumption of equal

*variances              is valid.

(3) Perform an Analysis of Variance F-test at the 5% level of significance to determine if the means of the two data sets are statistically different. A conclusion that the means are not statistically different is interpreted to mean that significant corrosion did not occur over the time period represented by the data. However, if equality of the means is rejected, this implies that the difference is statistically significant and could be due to corrosion. The range of potential corrosion rates is estimated by computing the slope of the steepest line which can be drawn within the +1 sigma confidence interval about the mean thickness for the duration between the two measurements. 0 OCLROOO000191

Calc. No. C-1302-187-5300-019 Rev. 0 Page 19 of 39 4.7 Analysis of Single 6"x6" Grid Data Set In those cases where a 6"x6" data set is taken at a given location for the first time during the current outage, the only other data to which they can be compared are the UT survey measurements taken at an earlier time. For the most part, these are single point measurements which were taken in the vicinity of the 49-point data set, but not at the exact location. Therefore, rigorous statistical analysis of these single data sets is impossible. However, by making certain assumptions, they can be compared with the previous data points. If more extensive data is available at the location of the 49-point data set, the Analysis of Variance F-test can be used to compare the means of the two data sets as described in paragraph 4.5. When additional measurements are made at these exact locations during future outages, more rigorous statistical analyses can be employed. 4.7.1 Assumptions The comparison of a single 49-point data sets with previous data from the same vicinity is based on the following assumptions: (1) Characterization of the scattering of data over the 6"x6" grid is such that the thickness measurements are normally distributed. (2) Once the distribution of data for the 6"x6" grid is found to be normal, then the mean value of the thickness is the appropriate representation of the average condition. (3) The prior data is representative of the condition at this location at the earlier date. 4.7.2 Statistical Approach The evaluation takes place in four steps: (1) Perform a univariate analysis of each data set to ensure that the assumption of normality is valid. (2) Calculate the mean and the standard error of the mean of the 49-point data set. (3) Determine the two-tailed t value from a t distribution table at levels of significance of 0.05 for n-i degrees of freedom. (4) Use the t value and the standard error of the mean to calculate the 95% confidence interval about the mean of the 49-point data set. OCLROO000192

Calc. No. C-1302-187-5300-019 Rev. 0 Page 20 of 39 (5) Compare the prior data point(s) with these confidence intervals about the mean of the 49-point data sets. If the prior data falls within the 95% confidence intervals, it provides some assurance that significant corrosion has not occurred in this region in the period of time covered by the data. If the prior data falls above the upper 95% confidence limit, it could mean either of two things: (1) significant corrosion has occurred over the time period covered by the data, or (2) the prior data point was not representative of the condition of the location of the 49-point data set in 1986. There is no way to differentiate between the two. If the prior data falls below the lower 95% confidence limit, it means that it is not representative of the condition at this location at the earlier date. OCLROO000193

Calc. No. C-1302-187-5300-019 Rev. 0 Page 21 of 39 4.8 Analysis of Single 7-Point Data Set In those cases where a 7-point data set is taken at a given location for the first time during the current outage, the only other data to which they can be compared are the UT survey measurements taken at an earlier time to identify the thinnest regions of the drywell shell in the sand bed region. For the most part, these are single point measurements which were taken in the vicinity of the 7-point data sets, but not at the exact locations. However, by making certain assumptions, they can be compared with the previous data points. If more extensive data is available at the location of the 7-point data set, the Analysis of Variance F-test can be used to compare the means of the two data sets as described in paragraph 4.5. When additional measurements are made at these exact locations during future outages, more rigorous statistical analyses can be employed. 4.8.1 Assumptions The comparison of a single 7-point data sets with previous data from the same vicinity is based on the following assumptions: (1) The corrosion in the region of each 7-point data set is normally distributed. (2) The prior data is representative of the condition at this location at the earlier date. The validity of these assumptions cannot be verified. 4.8.2 Statistical Approach Perform the Analysis of Variance and F-test If the prior data falls within the 95% confidence interval, it provides some assurance that significant corrosion has not occurred in this region in the period of time covered by the data. If the prior data falls above the upper 95% confidence interval, it could mean either of two things: (1) significant corrosion has occurred over the time period covered by the data, or (2) the prior data point was not representative of the condition of the location of the 7-point data set in 1986. There is no way to differentiate between the two. If the prior data falls below the lower 99% confidence limit, it means that it is not representative of the condition at this location at the earlier date. In this case, the corrosion rate will be interpreted to be "Indeterminable". OCLROO0001 94

Wý Calc. No. C-1302-187-5300-019 Rev. 0 Page 22 of 39 4.9 Evaluation of Drywell Mean Thickness This section defines the methods used to evaluate the drywell thickness at each location within the scope of the long term monitoring program. 4.9.1 Evaluation of Mean Thickness Using Regression Analysis The following procedure is used to evaluate the drywell mean thickness at those locations where regression analysis has been deemed to be significant (F-Ratio is 1.0 or greater). (1) The best estimate of the mean thickness at these locations is the point on the regression line corresponding to the time when the most recent set of measurements was taken. In the SAS Regression Analysis output (App. 6.2), this is the last value in the column labeled "PREDICT VALUE". (2) The best estimate of the standard error of the mean thickness is the standard error of the predicted value used above. In the SAS Regression Analysis output, this is the last value in the column labeled "STD ERR PREDICT". (3) The two-sided 95% confidence interval about the mean thickness is- equal to the mean thickness plus or minus t times the jestimated standard error of the mean. This is the interval 6for which we have 95% confidence that the true mean thickness will fall within. The value of t is obtained from a t distribution table for d:tails at n-2 degrees of freedom and 0.05 level of significance, where n is the number of sets of measurements used in the regression analysis. The degrees of freedom is equal to n-2 because two parameters (the y-intercept and the slope) are calculated in the regression analysis with n mean thicknesses as input. (4) The one-sided 95% lower limit of the mean thickness is equal to the estimated mean thickness minus t times the estimated standard error of the mean. This is the mean thickness for which we have 95% confidence that the true mean thickness does not fall below. In this case, the value of t is obtained from a t distribution table for p= tail at n-2 degrees of freedom and 0.05 level of significance. 4.9.2 Evaluation of Mean Thickness Using Mean Model The following procedure is used to evaluate the drywell mean thickness at those locations where the regression analysis is not significant (F-Ratio is less than 1.0). This method is consistent with that used to evaluate the mean thickness using the regression model. (1) Calculate the mean of each set of UT thickness measurements. OCLRO00001 95

Calc. No. C-1302-187-5300-019 Rev. 0 Page 23 of 39 (2) Sum the means of the sets and divide by the number of sets to calculate the grand mean. This is the best estimate of the mean thickness. In the SAS Regression Analysis output, this is the value labelled "DEP MEAN". (3) Using the means of the sets from (1) as input, calculate the standard error about the mean. This is the best estimate of the standard error of the mean thickness. (4) The two-sided 95% confidence interval about the mean thickness is equal to the mean thickness plus or minus t times the estimated standard error of the mean. This is the interval for which we have 95% confidence that the true mean thickness will fall within. The value of t is obtained from a t distribution table for g_ tails at n-1 degrees of freedom and 0.05 level of significance. (5) The one-sided 95% lower limit of the mean thickness is equal to the estimated mean thickness minus t times the estimated standard error of the mean. This is the mean thickness for which we have 95% confidence that the true mean thickness does not fall below. In this case, the value of t is obtained from a t distribution table for one tail at n-1 degrees of freedom and 0.05 level of significance. 4.9.3 Evaluation of Mean Thickness Using Single Data Set The following procedure is used to evaluate the drywell thickness at those locations where only one set of measurements is available. (1) Calculate the mean of the set of UT thickness measurements. This is the best estimate of the mean thickness. (2) Calculate the standard error of the mean for the set of UT measurements. This is the best estimate of the standard error of the mean thickness. Confidence intervals about the mean thickness cannot be calculated with only one data set available. OCLROO000196

Calc. No. C-1302-187-5300-019 Rev. 0 Page 24 of 39 4.10 Evaluation of Drywell Corrosion Rate 4.10.1 Regression Not Significant If the ratio of the observed F value to the critical F value is less than 1 for the F-test for the significance of regression, it indicates that the regression is not significant at the 5% level of significance. In other words, the variation in mean thickness with time can be explained solely by the random variations in the measurements. This means that the corrosion rate is not statistically significant compared to the random variations. The possibility does exist that the variability in the data may be masking an actual corrosion rate. Although the regression is not the result of the .regression analysis can be used to estimate the potentially masked corrosion rate. We can also state with 95% confidence that the corrosion rate does exceed the upper bound of the 95% one-sided confidence interval of the slope computed in the regression analysis. The 95% upper bound is equal to the computed slope plus the one-sided t-table value times the standard error of the slope. The value of t is determined for n-2 degrees of freedom. 4.10.2 Regression Significant e Ifthe ratio of the observed F value to the critical F value is or greater, it indicates that the regression model is more 1 appropriate than the mean model at the 5% level of significance. In other words, the variation in mean thickness with time cannot, be explained solely by the random variations in the measurements. This means that the corrosion rate is significant compared to the random variations. Although a ratio of 1 or greater indicates that regression is significant, it does not mean that the slope of the regression line is an accurate prediction of the corrosion rate. The ratio should be at least 4 or 5 to consider the slope to be a useful predictor of the corrosion rate (Ref. 3.5, pp. 93, 129-133). A ratio of 4 or 5 means that the variation from the mean due to regression is approximately twice the standard deviation of the residuals of the regression. To have a high degree of confidence in the predicted corrosion rate, the ratio should be at least 8 or 9 (Ref. 3.5, pp. 129-133). The upper bound of the 95% one-sided confidence interval about the computed slope is an estimate of the maximum probable corrosion rate at 95% confidence. The 95% upper bound is equal to the computed slope plus the one-sided t-table value times the standard error of the slope. The value of t is determined for n-2 degrees of freedom. OCLROO0001 97

Calc. No. C-1302-187-5300-019 Rev. 0 Page 25 of 39 5.0 CALCULA2TIONS 5.1 6"x6" Grids in Sand Bed Region 5.1.1 Bay 9D 12/19/88 to 11/02/91 In the analysis of data thru May 1991, these data sets did not meet the acceptance criteria for either regression or difference between means. Examination of the analysis revealed two reasons for this: (1) the mean value of the 6/26/89 data set fell about 30 mils above the regression line, and (2) there was a pit at point 15 which deviated from the mean thickness by more than 3-sigma. The data was reanalyzed without the 6/26/89 data set and without point 15. The regression of these data sets met the acceptance criteria and the regression was statistically significant. Eight 49-point data sets were available for the period through November 1991. With the November 1991 data, the data sets meet the acceptance criteria for regression with the 6/26/89 data set and point 15. However, the regression accounts for only 61% of the variability in the data. The regression without the 6/26/89 data set accounts for 85% of the variability in the data. The regression without the 6/26/89 data and without point 15 accounts for 84% of the variability in the data. The deviation of point 15

• from the mean thickness is 2.89-sigma for the November 1991 data, and thus is close to the 3-sigma Value for a deep pit. The regression without the 6/26/89 data is much stronger than the regression with it, and the-regressions with and without point 15 are essentially identical. Therefore, to provide continuity with the prior analyses, the reported regression results are without the 6/26/89 data and without point 15. The regression of these data sets meet the acceptance criteria and is statistically significant.

5.1.2 Bay 11A: 4/29/87 to 11/02/91 The regression of thirteen data sets for this period meets the acceptance criteria and is statistically significant. 0

                                        ýM OCLRUOO000198

Calc. No. C-1302-187-5300-019 Rev. 0 Page 26 of 39 5.1.3 Bay 11C: 5/1/87 to 11/02/91 Twelve 49-point data sets were available for this period. Prior analysis have shown that there has been minimal corrosion in the top 3 rows of the 6" x 6" grid with more extensive corrosion in the bottom 4 rows. Therefore, these subsets are analyzed separately. Top 3 Rows The regression of these data sets meets the acceptance criteria and is statistically significant. Bottom 4 Rows The regression of these data sets meets the acceptance criteria and is statistically significant. 5.1.4 Bay 13A: 12/17/88 to 11/02/91 The regression of nine data sets for this period meets the acceptance criteria and is statistically significant. OCLROO0001 99

Calc. No. C-1302-187-5300-019 Rev. 0 Page 27 of 39 5.1.5 Bay 13D: 3/28/90 to 11/02/91 One 7-point data set and four 49-point data sets were available for this period. Prior evaluation showed that the 7-point data set of 3/28/90 and the 49-point data set of 4/25/90 were normally distributed. However, there was a line of demarcation separating a zone of minimal corrosion at the top from a corroded zone at the bottom. Thus, it was concluded that corrosion has occurred at this location. The 49-point data set of 2/23/91 contains an invalid measurement at point #47. Therefore, this was input as a "missing" value to exclude it from the analyses. The data sets have a line of demarcation separating the upper and lower zones. Therefore, the grid was divided into two zones consisting of the following points: Top Zone Bottom Zone 1 - 16 17 - 18 19 - 22 23 - 26 27 - 28 29 - 49 Top Zone This zone consists of 22 points. (1) The data are normally distributed. (2) The regression is not statistically significant and has a positive slope. (3) Analysis of variance shows no significant difference between the means. Thus, there is no indication of statistically significant corrosion during this period. Bottom Zone This zone consists of 27 points. (1) The data are normally distributed except for the 4/25/90 data which is skewed to the thin side. (2) The regression is not statistically significant and has a positive slope. (3) Analysis of variance shows no significant difference between the means. Thus, there is no indication of statistically significant corrosion during this period. OCLROO000200

Cala. No. C-1302-187-5300-019 Rev. 0 Page 28 of 39 5.1.6 Bay 15D: 12/17/88 to 11/02/91 Eight 49-point data sets were available for this period. (1) The regression is not statistically significant. (2) The data are normally distributed. (3) The Analysis of Variance shows that there is no significant difference in the means at 95% confidence. Thus, there is no indication of statistically significant corrosion during this time period. 5.1.7 Bay 17A: 12/17/88 to 11/02/91 Eight 49-point data sets were available for this period. Prior analyses have shown a lack of normality due to minimal corrosion in the top 3 rows and more extensive corrosion in the bottom 4 rows. Therefore, these subsets are analyzed separately. Top 3 Rows (1) the regression is not statistically significant. (2) the data are normally distributed. (3) The Analysis of Variance shows that there is no significant difference in the means at 95% confidence. Thus, there is no indication of statistically significant corrosion during this time period. Bottom 4 Rows The regression of these data sets meets the acceptance criteria and is statistically significant. 5.1.8 Bay 17D: 2/17/87 to 11/02/91 The regression of thirteen data sets for this period meets the acceptance criteria and is statistically significant. OCLROO000201

Calc. No. C-1302-187-5300-019 Rev. 0 Page 29 of 39 5.1.9 Bay 17/19 Frame Cutout: 12/30/88 to 11/02/91 Eight 49-point data sets were available for this period. Prior analyses have shown a lack of normality due to more extensive loss of thickness in the top 3 rows than in the bottom 4 rows. Therefore, these subsets are analyzed separately. Top 3 Rows (1) Based on the Univariate Analysis, seven of the eight subsets are normally distributed. The other one (February 1990) contains two high readings and one low reading which cause the maldistribution. (2) The Analysis of Variance shows that there is a significant difference amongst the means at 95% confidence. This indicates that there is significant ongoing corrosion. (3) The regression of the eight data sets meets the acceptance criteria and is statistically significant. Bottom 4 rows (1) Based on the Univariate Analysis, three of the subsets are normally distributed. The other two (February 1990 and April 1990) each contain a data point which deviates significantly from the mean and causes the maldistribution. (2) The regression of the eight data sets meets the acceptance criteria and is statistically significant. 5.1.10 Bay 19A: 2/17/87 to 11/02/91 Thirteen 49-point data sets were available for this period. Since a plug lies within this region, four of the points were voided in each data set. The regression of these data sets meets the acceptance criteria and is statistically significant. 5.1.11 Bay 19B: 5/1/87 to 11/02/91 The regression of twelve data sets for this period meets the acceptance criteria and is statistically significant. 0 OCLRO0000202

Calc. No. C-1302-187-5300-019 Rev. 0 Page 30 of 39 5.1.12 Bay 19C: 5/1/87 to 11/02/91 Twelve 49-point data sets were available for this period. Since a plug lies within this region, four of the points were voided in each data set. The regression of these data sets meets the acceptance criteria and is statistically significant. 0

Calc. No. C-1302-187-5300-019 Rev. 0 Page 31 of 39 5.2 6" x.6" Grids at Elevation 50'-2" 5.2.1 Bay 5 Area D-12: 11/1/87 to 11/02/91 Ten 49-point data sets were available for this period. Since a plug lies within this region, seven of the points were voided in each data set. The initial analysis of these data sets indicated that they are not normally distributed. The following adjustments were made to the data: (1) Point 9 is a significant pit. Therefore, it was dropped from the overall analysis and is evaluated separately. (2) Points 13 and 25 are extremely variable and are located adjacent to the plug which removed from this grid. They were dropped from the analysis. (3) Point 43 in the 11/01/87 data set is much less than any succeeding measurement. Therefore, this data point was dropped from the analysis. (4) Point 29 in the 9/13/89 data is much greater than the preceding or succeeding measurements. Therefore, this data point was dropped from the analysis. (5) Points 1 and 37 in the 4/25/90 data set are much greater than the preceding or succeeding measurements. Therefore, these two data points were dropped from the analysis. (6) Points 3 and 36 in the 11/02/91 data set are much greater than the preceding or succeeding measurements. Therefore, these two data points were dropped from the analysis. With these adjustments, the Univariate Analyses indicate that all of the data sets are normally distributed at the L1% level of significance. The regression of these data sets meets the acceptance criteria and is statistically significant. OCLROO000204

Calc. No. C-1302-187-5300-019 Rev. 0 Page 32 of 39 Pit at Point 9 Analyses show that the high reading of 746 mils in July 1988 for the pit at point 9 is an outlier and must be dropped to obtain a meaningful least squares fit. Dropping this point, the mean thickness of the remaining points is 694.6 11.9 mils, and the standard deviation of the measurements is +/-6.1 mils. The best estimate of the corrosion rate is -3.6 +/-1.2 mils per year with an R2=52%. It is concluded that the corrosion rate in the pit is essentially the same as the overall grid. 5.2.2 Bay 5 Area 5: 3/31/90 to 11/02/91 Three 49-point data sets were available for this period. The data are not normally distributed due to a large corroded patch near the center of the grid and several smaller patches on the periphery. The data was split into two subsets consisting of points whose mean value is less than or equal to the grand mean, and those greater than the grand mean. Points With Mean Less than Grand Mean (1) The regression is not statistically significant. (2) These 15-point subsets are normally distributed. (3) Analysis of variance shows that there is not a significant difference between the means of the subsets. Thus, there is no indication of statistically significant corrosion during this period. Points with Mean Greater than Grand Mean (1) The regression is not statistically significant. (2) These 34-point subsets are normally distributed. (3) Analysis of variance shows that there is a statistically significant difference between some of the means. However, the differences do not correlate with time and are not attributed to corrosion. (4) Thus, there is no indication of statistically significant corrosion during this period. UU;LKUUUUU/LUO

Calc. No. C-1302-187-5300-019 Rev. 0 Page 33 of 39 5.2.3 Bay 13 Area 31: 3/31/90 to 11/2/91 Five 49-point data sets were available for this period. The data are not normally distributed. This is due to a large corroded patch at the left edge of the grid. The data was split into two subsets consisting of those points whose mean value is less than or equal to the grand mean, and those greater than the grand mean. Points with Mean Less than Grand Mean (1) The regression is not statistically significant. (2) These 14-point subsets are normally distributed. (3) Analysis of Variance shows that there is not a significant difference between the means of the subsets. Thus, there is no indication of statistically significant corrosion during this period. Points with Mean Greater than Grand Mean These 35-point subsets are not normally distributed. This is due to two points with low readings in March 1990, two points. with high readings in April 1990, two points with low readings in February 1991, and one point with a low reading in November 1991. When these seven points are deleted, the subsets are normally distributed. These subsets with the outliers deleted are evaluated below. (1) The regression is not statistically significant. (2) Analysis of variance shows that the second data set (April 1990) is statistically different from the other four. The November 1991 subset is greater than all except the April 1990 subset. Thus, there is no indication of statistically significant corrosion during this period. U)ULI<UUUUUL&UO

Calc. No. C-1302-187-5300-019 Rev. 0 Page 34 of 39 5.2.4 Bay 15 Area 23: 3/31/90 to 11/2/91 Five 49-point data sets were available for this period. The data are not normally distributed. This is due to a large corroded patch near the center of the grid and a significant pit at point 26. There are also some random readings over 780 mils which are outliers. Also, the measurement of 638 mils at point 27 in November 1991 is 118 mils less than the lowest prior measurement. This point is adjacent to the pit at point 26 and was therefore deleted. The data was split into two subsets: (1) Points whose mean value is less than or equal to the grand mean. The pit at point 26 was excluded. (2) Points whose mean value is greater than the grand mean. Readings greater than 780 mils were set to "missing." Points with Mean Less than Grand Mean (1) The regression is not statistically significant and has a positive slope. (2) The 16-point subsets are normally distributed. (3) Analysis of Variance shows that there is not a significant difference between the means of the subsets. (4) There is no indication of statistically significant corrosion during this period. Points with Mean Greater than Grand Mean (1) The regression is not statistically significant and has a positive slope. (2) The subsets are all normally distributed except for the 2/23/91 data which has two measurements (points 22 and

29) which are significantly higher than prior or subsequent measurements.

(3) Analysis of Variance indicates that there is a significant difference between some of the means. However, this is not indicative of corrosion since the later means exceed the earlier means. (4) There is no indication of statistically significant corrosion during this period. Pit at Point 26 OCLROO000207

Calc. No. C-1302-187-5300-019 Rev. 0 Page 35 of 39 The five readings are normally distributed. The best estimate of the corrosion rate is +1.2 +/-3.0 mils per year. There is no indication of significant corrosion during this period. Point 27 had a low measurement of 638 mils in November 1991. All prior measurements fell between 756 and 763 mils. Since this point is adjacent to point 26, it is concluded that the November 1991 measurement is really the pit analyzed above. OCLROO000208

Calc. No. C-1302-187-5300-019 Rev. 0 Page 36 of 39 5.3 6" x 6"1 Grids at Elevation 51"-10" 5.3.1 Bay 13 Area 32: 4/26/90 to 11/02/91 Four 49-point data sets were available for this period. The data are not normally distributed. This is due to a "T" shaped corrosion patch along the right edge and across the center. Examination of the Normal Probability Plot from the Univariate Analysis reveals the following distinct populations: (1) Four pits at points 20, 23, 25 and 28. (2) A group of 13 to 14 readings less than 705 mils. (3) A group of 31 to 32 readings greater than 705 mils. (4) Two outliers with values of 732 mils (Point 34 on 4/26/90) and 736 mils (point 33 on 2/23/91). (5) The 5/23/91 value at point 11 (660 mils) was 39 to 45 mils less than the other three values. If this point were included in the analysis, it would have a major impact on the calculated mean corrosion rate. For subsets (2) and (3) above, all points consistently fell in the same group except for points 1, 5, 7, 14, and 49. For each of these points, three measurements fell in one subset and one measurement fell in the other subset. Subsets (2) and (3) were used to analyze the corrosion rate. Points Less than 705 Mils (1) The regression is not statistically significant. (2) These subsets are normally distributed. (3) Analysis of Variance shows that there is not a significant difference between the means of the subsets. (4) There is no indication of statistically significant corrosion during this period. Points Greater than 705 Mils (1) The subsets are normally distributed except for one exceptionally high reading in February 1991. (2) Analysis of Variance shows that there is a significant difference between the mean of the November 1991 subset OCLROO000209

Calc. No. C-1302-187-5300-019 Rev. 0 Page 37 of 39 and the means of the other three subsets. However, the mean of the November 1991 subset exceeds the others and thus is not indicative of corrosion. (3) The regression is not statistically significant. (4) Thus, there is no indication of significant corrosion during this period. Pits at Points 20, 23, 25 and 28 The measurement at these locations are listed below. 2a 23 25 28 4/26/90 628 594 622 558 2/23/91 626 594 621 558 5/23/91 626 592 620 555 11/2/91 630 601 626 563 The standard deviation of November 1991 data for the points less than 705 mils is 14.3 mils. With 14 data points, the 99%/99% one-sided lower bound is 682-4.5 (14.3) = 617 mils. Thus points 23 and 28 are significant pits. However, the difference between readings is minimal, so there is no indication of significant corrosion in these pits. Low Reading at Point 11 There have been several cases where there have been significant differences between readings at a given location. It usually occurs in grids with significant pitting such that a pit is observed one time but not another time. The large difference at point 11 is attributed to this. OCLROO00021 0

Calc. No. C-1302-187-5300-019 Rev. 0 Page 38 of 39 5.4 6" x 6" Grids at 87"-5" Elevation No measurements were taken at the 87'-5" Elevation during the November 1991 due to the high temperature. Therefore, the May 1991 evaluation of corrosion rates is given below to provide complete documentation of the latest analyses. 5.4.1 Bay 9 Area 20: 11/6/87 to 5/23/91 The regression of the seven 49-point data sets for this period meets the acceptance criteria and is statistically significant. 5.4.2 Bay 13 Area 28: 11/10/87 to 5/23/91 Seven 49-point data sets were available for this period. The data sets are not normally distributed. Examination of the data shows that this is due to the seven thinnest points: 1, 2, 22, 25, 26, 36 and 48. Analysis of Data Without 7 Thinnest Points (1) The data are normally distributed. (2) The regression is not statistically significant. (3) Analysis of variance indicate that there is a significant difference between the means of the February and May 1991 data sets and the means of the other five data sets. This could be caused by actual corrosion, random variations in the data, or a slight bias in the measurements. More data is required to determine the true cause. "I_ _ _ _ _ _ _ _ _ _ OCLROO000211

Calc. No. C-1302-187-5300-019 Rev. 0 Page 39 of 39 5.4.3 Bay 15 Area 31: 11/10/87 to 5/23/91 Seven 49-point data sets were available for this period. (1) The data sets are normally distributed at 95% confidence except for the July 1988 data which is normally distributed at 99%. (2) The regression is not statistically significant. (3) Analysis of variance shows that there is a significant difference between the means of the February and May 1991 data sets and the means of the prior data sets. This could be caused by actual corrosion, random variations in the data, or a slight bias in the measurements. More data is required to determine the true cause. (4) The pit at point 34 is behaving like the rest of the grid. The current reading of 556 mils at point 34 is 9 mils below the mean for this point. The current mean reading for the grid is 8 mils below the grand mean. T OCLROO000212

Citizen's Exhibit NC6 Citizen's Exhibit NC6 APECO EnergyAintish Energy Company CALCULATION COVER SHEET (Ref. EP-006)

Subject:

Statistical Analysis of Drywell Vessel Calculation No. Rev o. System Nos. I Sheet Thickness Data Through September 2000 C-1302-187-E3 10-0377 187 1 of 36

1. Is this calculation within the scope of the GPUN Operational Quality Assurance X Yes O No Plan? (If YES, a verification is required unless the calculation is a non-substantive revision.)

2. Does this calculation contain assumptions I design inputs that require confirmation? 0l Yes X ho (If YES, provide CAP or appropriate configuration control number(s)) (e.g., ECD.

PFU. MD. PCR. etc.)

3. Does this calculation require revision to any existing documents? (If yes, provide 0 Yes X No CAP or appropriate configuration control numberis))

4. Is this calculation performed as a design basis calculation? (If YES, identify design CE Yes X No basis parameters.) (See Section 3.3)

Parameter: Referenced Calculations and Safety Evaluations (See Section 4.3.1.3) Rev. No. Safety Evaluation SE-000243-002, "Drywell Steel Shell Plate Thickness Reduction at the 15 Base Sand Cushion Entrenchment Region."

2) GPUN calculation C-1302-187-5300-005, Rev.0, "Statistical Analysis of Drywell 6, Thickness Data Thru 12-31-88"

3) GPUN Calculation C-1302-187-5300-028, Rev.0. "OCDW Statistical Analysis of 0 Drywell Thickness Data Thru September 1994"

4) GPUN Calculation C-1302-187-5300-028, Rev.0. "Statistical Analysis of Drywell 0 Thickness Data Thru September 1996" Comments: Rev. I updates the Coversheet Referenced Calculation Section with three additional references. In addition reference 3.22 on page II was corrected from C-1302-187-5300-028 to C-1302-187-5300-030. These changes correct or update references and editorial and do not affect the calculation, the conclusions or results.

Therefore the verification is unaffected. APPROVALS I Originator Peter Tamburro Date 1/2,230/0 Verification Engineer/Reviewer Steve Leshnoff ,< 3 IDate 1-1200 Section Manager Tom Date . / I Other Verification Engineer/Reviewer Date i. Other Verification Engineer/Reviewer Date AG5870 (02/00) OCLROO000694

AmerGen CALCULATION VERIFICATION CHECKLIST (Ref. EP-006)

Subject:

Statistical Analysis of Drywell Vessel Calculation No. Rev. No. System Nos. Sheet: Thickness Data Through September 2000 C-1302-187-E310-037 0 187 3 of 36 Place an "X" in the applicable box (Yes, No, NIA) for each item. A 'NO" response may indicate that the design or verification is incomplete and may require a CAP to be assigned by the responsible Section Manager. The Section Manager shall review each "NO" response to determine if the "NO" response requires further investigation. A "N/A" (Not Applicable) response does not require any further action by the Verification Engineer. The Verification Summary (Exhibit 7A) may be used to outline the Verification Engineer's work or to document comments that are deemed appropriate by the Verification Engineer. Review Check ITEMS Design Compliance Yes No N/A

1. Design Input and Data - Were the inputs correctly selected, referenced El 0 0 (latest revision) and incorporated into the calculation?

2. Assumptions - Are assumptions necessary to perform the calculation '0 0 adequately described and reasonable?

3. Regulatory Requirements - Are the applicable codes and standards and I L-regulatory requirements, including Issue and addenda, properly identified and their requirements met?

4. Construction and Operating Experience - Has applicable construction and El f]19 operating experience been considered?

5. Interfaces - Have the design interface requirements been satisfied? 0 9

6. Methods -Is the appropriate calculation method used? Nl LI 0I

7. Outr - Is the output reasonable compared to the inputs? El LI 0

8. Acceptance Criteria - Are the acceptance criteria incorporated in the LI 00l calculation sufficient to allow verification that the design requirements have been satisfactorily accomplished?

   '9. Radiation Exposure - Has the calculation properly considered radiation                    0        LI        I-i exposure to the public and plant personnel?

Comments: Use Additional Sheets if Necessary AG5830 (1199) OCLR00000697

Amer~en CALCULATION SHEET Preparer: Pete Tamburro 2/13/01

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E3I0-037 0 187 4 of 36 Through September 2000

1. Purpose The purpose of this calculation is to update the Drywell Thickness Analyses documented in reference 3.7, 3.8, and 3.11 through 3.22 by incorporating measurements taken in September 2000 (see Appendix 10).

Results of this calculation will be used in to update reference 3.1 Specific objectives of this calculation are:

1) Determine the September 2000 mean thickness at each monitored location

2) Statistically analyze the thickness measurements to determine if a corrosion rate exists at each location,

3) If corrosion rate exists, provide a conservative projection to 2009 and 2029.

This calculation does not evaluate the sand bed region. The corrosion in the sand bed region was eradicated in 1992 by removing sand. The external side of the Drywell Vessel in these regions was then coated. Follow-up inspections after 1992 (including September 2000) shows that the coating is good condition. Therefore thickness measurements of the sandbed region are not required. This calculation does not use the same software that was used in earlier calculations. Previous calculations utilized the GPUN main frame computer and the "SAS" main frame software. The Oyster Creek Plant has been sold to AmerGen in the year 2000. The GPUN Main Frame will not be available to Ame rGen after the' year 2002. Also the "SAS" software is mainframe based and difficult to learn and maintain. An alternative PC based software, "MATHCAD", has been chosen to perform this calculation. Although software has been changed the overall methodology, with minor exceptions, is the same as in previous calculation. The minor exceptions are the statistical tests which determine whether the data is normally distributed. Also, since the GPUN Maine Frame Computer stored all program data, this calculation documents all data sets since the beginning of the program for each inspection location above the sandbed elevation. OCLR00000698

Aer&AO eAPreparer: Pete Tamburro 2/13/01 CALCULATION SHEET

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E310-037 0 187 4 of 36 Through September 2000

1. Purpose The purpose of this calculation is to update the Drywell Thickness Analyses documented in reference 3.7, 3.8, and 3.11 through 3.22 by incorporating measurements taken in September 2000 (see Appendix 10).

Results of this calculation will be used in to update reference 3.1 Specific objectives of this calculation are:

1) Determine the September 2000 mean thickness at each monitored location

2) Statistically analyze the thickness measurements to determine if a corrosion rate exists at each location,

3) If corrosion rate exists, provide a conservative projection to 2009 and 2029.

This calculation does not evaluate the sand bed region. The corrosion in the sand bed region was eradicated in 1992 by removing sand. The external side of the Drywell Vessel in these regions was then coated. Follow-up inspections after 1992 (including September 2000) shows that the coating is good condition. Therefore thickness measurements of the sandbed region are not required. This calculation does not use the same software that was used in earlier calculations. Previous calculations utilized the GPUN main frame computer and the "SAS" main frame software. The Oyster Creek Plant has been sold to AmerGen in the year 2000. The GPUN Main Frame will not be available to AmerGen after the year 2002. Also the "SAS" software is mainframe based and difficult to learn arid maintain. An alternative PC based software, "MATHCAD", has been chosen to perform this calculation. Although software has been changed the overall methodology, with minor exceptions, is the same as in previous calculation. The minor exceptions are the statistical tests which determine whether the data is normally distributed. Also, since the GPUN Maine Frame Computer stored all program data, this calculation documents all data sets since the beginning of the program for each inspection location above the sandbed elevation. OCLR00000698

AmerGen -CALCULATION SHEET Preparer: Pete Tamburro 2/13/01 mi

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E310-037 0 187 5 of 36 Through September 2000 2.0 Summary of Results 2.1 Elevation 50' 2" Through September 2000. Bay Area/ Sept. 2000 No of F-ratio Mean of all Corrosion Projected Min Location Mean +/- Insp. Inspections Rate Lower 95% Required Standard +1- Standard (mils! Confidence Thickness Error (mils) Error (mils) year) Thickness in (mils) 2009 (mils) 5 D12 741.2+/- 1.8 11 0.13 744.8 +/-0.8 NA' NA 541 5 5 low 706.1+/- 6.6 10 0.06 704.7 +/-0.6 NA NA 541 5 5 hi 754.1+/-1.9 10 1.23 NA 0.6 742.8 541 13 31 low 682.0+/-6.9 10 0.02 685.4+1- 1.2 NA NA 541 13 31 hi 762.4+/- 2.0 10 0.06 764.0 +1-1.8 NA NA 541 15 23 low 729.4 +/-3.6 10 0.51 726,5 +/- 0.6 NA NA 541 15 23 hi 757.8+/- 1.1 10 0.61 761.7+/- 1.0 NA NA 541 Since February 1990, ten or more inspections have been performed on each of the four locations at this elevation. Three of these four locations are not experiencing corrosion. A portion of the fourth location (Bay 5, area 5) may be experiencing a minor corrosion rate of approximately 0.6 mils per year. This corrosion rate is very small. Projection based on this corrosion rate using the 95% lower confidence interval shows that it will not corrode to less than the minimum required thickness by the year 2009 or 2029. There is substantial margin, even when considering plant life extension (see the plot below). Bay 5 Area 5 Corrosion Projection Individual 9 Upper 95% inspection II IAt confidence interval means 750 Projected mean Upper 95% confidence interval coo I. 651 6Wr Minimum Required thickness 550 at this elevation SI1I , I I , I 1 920 1990 2000 2010 2020 030

                                                                                                                  *13, j.9sp5e,0    Yc re!   ~   Y pre"'cY'~preWYr P         =2.029 OCLR00000699

AmerGen CALCULATION SHEET Preparer: Pete Tamburro 2/13/01 m

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-£310-037 0 187 6 of 36 Through September 2000 Analysis of individual points within these four locations shows no ongoing corrosion except for two points. Bay 5 Location D12, point 9 may be experiencing a corrosion rate of 1.3 mils per year. Bay 5 Location 23, point 26 may be experiencing a corrosion rate of 1.5 mils per year. These corrosion rates are very small. Projection based on these corrosion rates using the 95% lower confidence interval shows that these points will not corrode to less than the minimum required thickness by the years 2009. or 2029. There is substantial margin, even when considering plant life extension (see the plot below). Bay 5 Area 23 Point 26 Corrosion Projection Individual I I I I mInspection 6" means - Upper 95% confidence interval 6001 4- Projected mean Mi 26 XXX:- Upper 95% itcurwe confidence interval Uppit 500 I-Thm l-ocal51 400 Minimum Required thickness at this elevation v i i  ! 1990 2000 2010 2020 2030 min(Dates)-2 Dafts'year p'.diý",Yeat Vron 2030 These results are not unexpected given the minor rates, the accuracy of UT technology, and the repeatability of data collection methodology. For example, in Bay 5 location 5 the Standard Error for September 2000 is 1.9 mils. This Standard Error is conistent for data from past inspections and for this location. Therefore it would take a substantial amount of time for a rate of 0.6 mils per year to be observed. It is therefore concluded that the program has finally performed enough inspections over a long enough time frame to observe these minor corrosion rates. OCLROO000700

Aenmer ALC LATIN SEETPreparer: CALCULATION SHEET Pete Tamburro 2/13/01

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E310-037 0 187 7 of 36 Through September 2000 2.2 Elevation 51' 10" through September 2000. Bay Area/ Sept. 2000 No of F-ratio Mean of all Corrosi Projected Min Location Mean +/- Insp. Inspections on Rate Lower 95% Required Standard +/-Standard (mils/ Confidence Thickness Error (mils) Error (mils) year) Thickness in (mils) _2009 (mils) 13 32 ow 678.8 +/-52 9 0.30 681.8 +/- 0.9 NA NA 541 13 32 hi 715.2+/-0.8 9 0.17 716.0 +/-0.6 NA NA 541 Since April 1990, nine inspections have been performed on one location at this elevation. The data indicates that this location is not experiencing corrosion (see the pot below). Bay 13 Area 32 Thickness Individual Inspection means 760 Grand mean of 33 7~u thickest points z 9 eneurcd Grand mean if all

                                                              *113 points 0700 Grand mean of 15 thinnest points
                                    =M00460 I3990      I99M       394       1996      399        2000 Dtus                        we oDwes)+ I Analysis of local individual points within this location shows no ongoing corrosion.

r 'U, OCLROO00701

ArmerGen mee Aen ALCLATIN SEETPreparer: t Pete Tamburro 2/13/01

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E310-037 0 187 8 of 36 Through September 2000 1 1 1 2.3 Elevation 60' 10" through September 2000. Bay Area/ Sept. 2000 No of F-ratio Mean of all Corrosi Projected Min Location Mean +/- Insp. Inspections on Rate Lower 95% Required Standard +/-Standard (mils/ Confidence Thickness Error (mils) Error (mils) year) Thickness in (mils) 2009 (mils) 1 5-22 688.5 +/-4.2 14 1 0.04 696.5 +/- 5.0 NA NA 518 Since December 1992, only four inspections have been performed on this one location. The data indicates that this location is'not experiencing corrosion. Analysis of local individual points within this location shows that there -may be ongoing corrosion at one point. Bay I location 5-22, point 48 may be experiencing a corrosion rate of 4.5 mils per year. This calculated rate, which is greater than all other calculated rates may be due to the limited amount of inspections. The methodology and analysis results in greater rates and confidence levels With less inspection information. As shown for other locations, which have at least 9 inspections, the observed rates are less. Never-the-less projection based on this corrosion rates using the 95% lower confidence interval, which is significantly more conservative than at other locations show that this point will not corrode to less than the minimum required thickness by the year 2009. Additional inspections are required to successfully project to 2029. Bay I Area 5-22 Point 48 Corrosion Projection Individual Inspection means Minimum Required thickness at this elevation 1995 2000 2005 2010 min(Date)- 2 DatesWYepedkY predict 2010 OCLROO000702

AmerGen CALCULATION SHEET Preparer: Pete Tamburro 2/13/01 - I

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E310-037 0 187 9 of 36 Through September 2000 2.4 Elevation 87' 5" through September 2000. Bay Area/ Sept. 2000 No of F-ratio Mean of all Corrosi Projected Min Location Mean +/- Insp. Inspections on Rate Lower 95% Required Standard +/- Standard (mils/ Confidence Thickness Error (mils) Error (mils) year) Thickness in (mils) 2009 (mils) 9 20 603.8 +/-2.1 12 1.84 NA 1.2 596.7 452 13 28 634.9 +/-1.9 12 0.62 638.8 +1- 1.3 NA NA 452 15 31 627.5 +/-2.0 12 1.05 633.1 +1- 1A 0.75 620.5 452 Since November 1987, twelve inspections have been performed on each of the three locations at this elevation. Two of the three locations may be experiencing corrosion. Bay 9, area 20 is experiencing minor a corrosion rate of approximately 1.2 mils per year. Bay 15 area 31 may be experiencing a corrosion rate of 0.75 mils per year. The F-ratio for this second location is 1.05, which is on the threshold as to whether or not a rate exists. These corrosion rates are very small.. Projections based on these corrosion rates using the 95% lower confidence interval shows that they will not corrode to less than the mniimum required thickness by the years 2009 or 2029. There is substantial margin, even when considering life extension. Bay 9 Area 20 Corrosion Projection Bay 15 Area 31 Corrosion Projection l .... a - a a-I-

                                                                                          "5-I-
                          'S. r I-                                                                         h... SSR
               ~  0                                                                         4 a                                                                    a
                                                                                        -        a a

a I I " I/ I I- ~ a. - - ". =N via onS. Analysis of local individual points within these three locations show no ongoing corrosion with the possible exception of point 25 in bay 13 area 28 which has an F-ratio of 0.96. Again this value is on the threshold as to whether or not a rate exists. This point may be experiencing a corrosion rate of 3.0 mils per year. Projection based on these corrosion rates using the 95% lower confidence interval shows that this point will not corrode to less than the minimum required thickness by the years 2009 or 2029. There is substantial margin, even when considering plant life extension. OCLROO000703

AmerGen Aenmer ALCLATIN SEETPreparer: CALCULATION SHEET oouo Pete Tamburro 2/13/01

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E310-037 0 187 10 of 36 Through September 2000 1 Bay 13 Area 28 Point 25 Corrosion Projection 91 23 an. Swig 2050 2020 12030 a.h,(D~ia)-2 D~1a.VafFdWap,~d1a am3a These results are not unexpected given the minor rates, the accuracy of UT technology, and the repeatability of data collection methodology. For example, in Bay 9 location 20 the Standard Error for September 2000 is 2.1 mils. This Standard Error is consistent for data from past inspections and for this location. Therefore it would take a substantial amount of time for a rate of 1.2 mils per year to be observed. It is therefore concluded that the program has finally performed enough inspections over a long enough time frame to observe these minor corrosion rates. I OCLROOOO7Od4-

Am erG en CPreparer: CALCULATION SHEET Pete Tamburro 4/19/01

Subject:

Calculation No. Rev. o. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-8310-037 1 ,.I 187 11 of 36 Through September 2000 641-3.1 References 3.1 GPUN Safety Evaluation SE-000243-002, Rev. 14 "Drywell Steel Shell Plate Thickness Reduction at the Base Sand Cushion Entrenchment Region." 3.2 GPUN TDR 854, Rev. 0 "Drywell Corrosion Assessment" 3.3 GPUN TDR 851, Rev. 0 "Assessment of Oyster Creek Drywell Shell" 3.4 GPUN Installation Specification, IS-328227-004, Rev XX, "Functional Requirements for Drywell Containment Vessel Thickness Examination" 3.5 Applied Regression Analysis, 2 nd Edition, N. R. Draper & H. Smith, John Wiley and Sons 1981 3.6 Statistical Concepts and Methods, G.K. Bhattacharyya & R.A. Johnson, John Wiley and Sons 1977 3.7 GPUN calculation C-1302-187-5300-005, Rev.0, "Statistical Analysis of Drywell Thickness Data Thru 12-31-88" 3.8 GPUN TDR 948, Rev. I "Statistical Analysis of Drywell Thickness Data" 3.9 Experimental Statistics, Mary Gobbons Natrella, John Wiley & Sons, 1966 Reprint (National Bureau of Standards Handbook 91) 3.10 Fundamental Concepts in the Design of Experiments, Charles C Hicks, Saunders College Publishing, Fort Worth, 1982 3.11 GPUN Calculation C-1302-187-5300-008, Rev.0, "Statistical Analysis of Drywell Thickness Data Thru 2-8-90" 3.12 GPUN Calculation C-1302-187-5300-01 1, Rev.1, "Statistical Analysis of Drywell Thickness Data Thru 4-24-90" 3.13 GPUN Calculation C-1302-187-5300-015, Rev.0, "Statistical Analysis of Drywell Thickness Data Thru March 1991" 3.14 GPUN Calculation C-1302-187-5300-017, Rev.0, "Statistical Analysis of Drywell Thickness Data Thru May 1991" 3.15 GPUN Calculation C-1302-187-5300-019, Rev.0, "Statistical Analysis of Drywell Thickness Data Thru November 1991" 3.16 GPUN Calculation C-1302-187-5300-020, Rev.0, "OCDW Projected Thickness Data Thru 11/02/91" 3.17 GPUN Calculation C-1302-187-5300-021, Rev.0, "Statistical Analysis of Drywell Thickness Data Thru May 1992" 3.18 GPUN Calculation C-1302-187-5300-022, Rev.0, "OCDW Projected Thickness Data Thru 5/31/92T 3.19 GPUN Calculation C- 1302-187-5300-025, Rev.0, "Statistical Analysis of Drywell Thickness Data Thru December 1992" 3.20 GPUN Calculation C-1302-187-5300-024, Rev.0, "OCDW Projected Thickness Data Thru 12/8/92" 3.21 GPUN Calculation C-1302-187-5300-028, Rev.0, "OCDW Statistical Analysis of Dryweil Thickness Data Thru September 1994" 3.22 GPUN Calculation C-1302-187-5300-030, Rev.0, "Statistical Analysis of Drywell Thickness Data Thru September 1996" I 3.23 Practical Statistics - "Mathcad Software Version 7.0 Reference Library, Published by Mathsoft, Inc. Cambridge OCLROO000705

 -AmerGen                .A en mer                   ALCUATIN SEETPreparer.                           Pete Tamburro 2/13/01

• CALCULATION SHEET

,

Subject:

Statistical Analysis of Drywell Vessel Thickness Data Calculation No. C-1302-187-E3 10-037 Rev'. No. 0 System Nos. 187 Sheet 12 of 36 Through September 2000 4.0 Assumptions The statistical evaluation of the UT measurement data to determine the corrosion rate at each location is based on the following assumptions: 4.1 Characterization of the scattering of the data over each 6" by 6" grid is such that the thickness measurements are normally distributed. If the data is not normally distributed the grid is subdivide into normally distributed subdivisions. 4.2 Once the distribution of data is found to be close to normal, the mean value of the data points is the appropriate representation of the average condition. 4.3 A decrease in the mean value of the thickness over time is representative of the corrosion. 4.4 If corrosion does not exist, the mean value of the thickness will not vary with time except for random variations in the UT measurements 4.5 If corrosion is continuing at a constant rate, the mean thickness will decrease linearly with time. In this case, linear regression analysis can be used to fit the mean thickness values for a given zone to a straight line as a function of time. The corrosion rate is equal to the slope of the line. 0 OCLROO000706

Preparer: Pete Tamburro 2/13/01 IAIeEr~eE CALCULATION SHEET O

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E3310-037 0 187 13 of 36 Through September 2000 5.0 Design Inputs: 5.1 The minimum required thickness for the three elevations in which the data was collected in September 2000 are documented below (reference 3.1). 5.2 Seven core sample approximately 2" in diAmeter were removed from thb drywell vessel shell for analysis (reference 3.1). In these locations replacement plugs were installed. Five of these removed cores are in grid locations that are part of the monitoring program. Of these, 4 were in sandbed region, which are no longer monitored (reference 3.1). The remaining core was removed from the grid at elevation 50'2" bay 5 area D 12. The replacement plug is located over data points 13, 20, 25, 26, 27, 28, 33, 34, and 35. Therefore the UT data from these points are not included in the calculation. 5.3 Historical data sets were collected from previous calculations (references 3.7, and 3.11 through 3.22) OCLRO0000707

Amer~~en Aenmer ALCLATIN SEETPreparer: o. Pete Tamburro 2/13/01 O

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E310-037 0 187 14 of 36 Through September 2000 6.0 OVERALL APPROACH AND METHODOLOGY: 6.1 Definitions 6.1.1 A Normal Distribution has the following properties

                    - Characterized by a bell shaped curve centered on the mean.
                    - A value of that quantity is just as likely to lie above the mean as below it
                    - A value of that quantity is less likely to occur the farther it is from the mean
                    - Values to one side of the mean are of the same probability as values at the same distance on the other side of the mean 6.1.2 Mean thickness is the mean of valid points, which are normally, distributed from the most recent UT measurements at a location.

6.1.3 Variance is the mean of the square of the difference between each data point value and the mean of the population. 6.1.4 Standard Deviation is the square root of the variance. 6.1.5 Standard Error is the standard deviation divided by the square root of the number of data points. Used to measure the dispersion in the distribution. 6.1.6 Skewness measures the relative positions of the mean, medium and mode of a distribution. In general when the skewness is close to zero, the mean, medium and mode are centered on the distribution. The closer skewness is to zero the more symmetrical the distribution. Normal distributions have skewness, which approach zero. 6.1.7 Kurtosis measures the heaviness of a distribution tails. A normal distribution has a kurtosis, which approaches zero. 6.1.8 Linear Regression is a linear relationship between two variables. A line with a slope and an intercept with the vertical axis can characterize the linear relationship. In this case the linear relationship is between time (which is the independent variable) and corrosion (which is the dependent variable). 6.1.9 F-Ratio - An F-Ratio less than 1.0 occurs .whenthe amount of corrosion which has occurred since the initial measurement is less than the random variations in the measurements or fewer than four measurements have been taken. If the F ratio is less than 1.0, the computed corrosion rate does not reflect the actual corrosion rate but rather is provided to as a conservative projection (reference 2.22). OCLROO000708

Ameru~en A mner ~ CACULAION CALCULATION SHEET HEETPreparer: o, o Pete Tambwrro 2/13/01

Subject:

Calculation No. Rev. No. System 1Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E310-037 0 187 15 of 36 Through September 2000 An 'F" ratio of 1 or less indicates that the data trend is best explained by the grand mean of the data and the trend has no slope. The variability in the data is within the distribution profile for the data, which is normally distributed. Therefore a grand mean (0grand actual) is best estimate of the thickness of the location. An F-Ratio of greater than 1.0 occurs when the amount of corrosion that has occurred since the initial measurement is significant compared to the random variations, and four or more measurements have been taken. In these cases the computed corrosion rate more accurately reflects the actual corrosion rate, and there is a very low probability that the actual corrosion rate is zero. The greater the F-Ratio the lower the uncertainty in the corrosion rate (reference 2.22). Where the F-Ratio of 1.0 or greater provides confidence in the historical corrosion rate, the F-Ratio should be 4 to 5 if the corrosion rate is to be used to predict the thickness in the future. To have a high degree of confidence in the predicted thickness, the ratio should be at least 8 or 9 (reference 3.22). 6.1.10 Grand mean - when the F-Ratio test is less than 1.0 and/or the slope is positive this is the grand mean of all data. 6.1.11 Corrosion Rate - With three or more data sets and the F-Ratio test greater than 1.0 this is the slope of the regression line. 6.1.12 Upper and Lower 95% Confidence Interval - The upper and lower corrosion rate range for which there is 95% confidence that the actual rate lies within. 6.2 The UT measurements within scope of this monitoring program are performed in accordance with ref. 3.4. This specification involves taking UT measurements using a template with 49 holes laid out on a 6",by 6" gr.d with I" between centers on both axes. The first sets of measurements were made in 1987. All subsequent measurements are made in the same location within 1/8" (reference 3.4). 6.3 Each 49 point data set is evaluated for missing data. Invalid points are those that are declared invalid by the UT operator or are at plug locations. 6.3 Past calculations were reviewed to ensure that points that were considered pits are accurately trended and excluded from the calculation of the mean. 6.4 September 2000 data that are not normally distributed were compared to previous calculations to determine if past data was also not normally distributed. In such cases the new data is divided into subsets with the same points as in past calculations. OCLROO000709

Preparer: Pete Tamburro 2/13/01 AmerGen. CA LCULATION SHEET

Subject:

Statistical Analysis of Drywell Vessel Thickness Data Through September 2000 6.5 Methodology 6.6.1 Test Matrix To demonstrate the methodology a 49 member array will be generated using the Mathcad "rnorm" function. This function returns an array with a probability density which is normally distributed, where the size of the array (No DataCells), the target mean (i. ), and the target standard input deviation input) are input. The following will build a matrix of 49 points No DataCells := 49 i := 0.. No DataCclls,- I count :=7 The array "Cells" is generated by Mathcad with the target mean (V input) and standard deviation '. input) P input :=.775 Cr input  :=20 Cells*:= morm(No DataCells, P input, C input)

  "Cells" is shown as a 7 by 7 matrix 766 761 766 756 741 776 773 786 819 791 795' 792 793 788 754 776 760 789 771 762 761 Show    matriiCells, 7 ) = 765 786 770 777 800 761 775 797 793 717 732 779 763, 751 777 790 781 775 760 767 762 772 795 779 785 790 775 781 The above test matrix will be used in sections 6.5.2 through 6.5.8 6.5.2 Mean and Standard Deviation The actual mean and standard deviation are calculated for the matrix "Cells" by the Mathcad functions "mean" and "Stdev".

Therefore for the matrix generated In section 6.5.1 P actual mean(Cells) a actual :=Stdev(Cells) 77 4 pctual .10 o actual = 18.258 Inspection shows that the actual mean and standard deviations are not the same as the target mean and target standard deviation which were input. This Is expected since the "morm" function returns an array with a probability density which is normally distributed. OCLROO00071 0

Preparer: Pete Tarnburro 2/13101 Amer~enCALCULATION SHEET

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E3310-037 0 187 17 of 36 Through September 2000 6.6.3 Standard Error The Standard Error is calculated using the following equation (reference 3.23). For the matrix generated in section 6.5.1 Standar actual Standard error "- Standard error = 2.57 8 No DataCells 6.5.4 Skewness Skewness is calculated using the following equation (reference 3.23). For the matrix generated In section 6.5.1 (NO DataCells) "2(Cells actual) 3

  *KCWrIlSS (No DataCells    1)'No Date i-2)-c actual) 3                       Skewness = 0.354 A skewness value close to zero is indicative of a normal distribution (reference 3.22 and 3.23 6.6 Kurtosis Kurtosis Is calculated using the following equation (reference 3.23).

For the matrix geherated in section 6.5.1 No DataCells kNo DataCells ÷) e 1 actual) Kurtosis := (fNO DataCells 1)J(NODataCells-2)-(No ataCells ' cu) (No Data~ells 2' -(No DataCells -3) Kurtosis = 0.262 A Kurtosis value close to zero is indicative of a normal distribution (reference 3.23) OCLROO000711

 'Ameri~en ACALCULATION                                                            SHEET Preparer: Pete Tainburro 2/13/01

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E310-037 0 187 18 of 36 Through September 2000 6.5.6 Normal Probability Plot An alternative method to determine whether a sample distribution approaches a normal distrib is by anormal probability plkteference 3.22 and 3.23). In a normal plot, each data value is plot against what its value would be if it actually came from a normal distribTdihexpected normal values, callechormal scoreý and can be estimated by first calculating the rank scores of the sort data. The Mathcad function "sorts" sorts the "Cells" array j :=0.. last(Cells) srt :=sort(Cells) Then each data point is ranked. The array "rank" captures these rankings r:=j1 rank.:=- 7-srt-srt Each rank is proportioned into the "p" array. Then based on the portion an estimate for data point. TheVan der Waerden'sformula is used rank, PJ :=rows(Cells) +-I-The normal scores are the correspondirgh percentile points from the standard normal distribution: X:= I N Score.:=rootcnormx)-. p)x If a sample is normally distributed, the points of the "Normal Plot" will seem to form a nearly straight line. The plot below shows the "Normal Plot" for the matrix generated in section 6.5.1 3x 2 x-0- hi sr< hi2 -X 720 740 700 780 800 820 OCLROO000712

Preparer: Pete Tamburro 2/13/01 Amer~enCALCULATION SHEET

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-I302-187-E3310-037 0 187 19 of 36 Through September 2000 .. 1 1 6.5.7 Upper and Lower Confidence Values The Upper and Lower confidence values are calculated based on .05 degree of confidence a" (reference 3.23). a :=.05 Ta: I-1 Ta = 2.011 Therefore for the matrix generated in section 6.1 o actual a Ta Lower 95%ConI :P actual.- Lower 95%Con = 767.726 4No DataCells Cactual actual Upper 95%Con :=P atua + Tct Upper 95%Cffi= 778.094 FNO DataCells These values represent a range on the calculated mean in which there Is 95% confidence. In other words, if the 49 data points were collected 100 times the calculated mean in 95 of those 100 times would be within this range. 6.6.8 Graphical Representation Below is the distribution of the "Cells" matrix generated in section 6.5.1 sorted in one half standard deviation Increments (bins) within a range from minus 3 standard deviations to plus 3 standard deviations. 77 0 0 Bins:= Make bins (,P actual, 0 actual) 3 4 6 Distribution := hist(Bins,Cells) 13 Distribution = The mid points of the Bins are calculated

                                                                                                  !8 A

k:=0.. 11 (Binsk t- Binskl ) 2" 3l M idpoints k - =

• The Mathcad function pnorm calculates the normal distribution curve based on a given mean and standard w deviation. The actual mean and standard deviation generated in section 6.5.2 are input. The resulting plot will provide a representation of the normally distribution corresponding-the the actual mean and standard deviation.

OCLROO000713

Am erGen CALCULATION SHEET Preparer: Pete Tamburro 2/13/01

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E3310-037 0 187 20 of 36 Through September 2000 normal curveo := pnorm (Binsvp actual' 0 actual) normal curvek := pnorm(Binsk-iIP actual,' actual)- pnorm(Binsk, I actualo actual) The normal curve is simply a proportion, which is multiplied by the number of Cells" (49) normal curve :=No DataCells-normal curve The following schematic shows: the actual distribution of the samples (the bars), the normal curve (solid line) based on the actual mean (, actual ) and standard deviation *( actual). the kurtosis (Kurtosis), the skewness (Skewness). the number of data points (No DataCells), and the the lower and upper 95% confidence values Lower95 Con, Upper 95%Con). i actual = 772.91 o actual 18.047 Standard error = 2.578 Skewness = 0354 Kurtosis = 0.262 No DataCells = 49 Distribution normal curve 5-720 740 760 780 800 820 840 Midpoints.MkIpoints Lower95%Con = 767.726 Upper 95VoCon 778.094 OCLROO000714

Preparer: Pete Tamburro2/13101 CALCULATION SHEET

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E310-037 0 187 21 of 36 Through September 2000 6.5.9 The "F" Test for Linear Regression In order to determine whether the historical data for each location is indicative of corrosion, the means collected at each location over time are tested using the "F Test for Linear Regression as follows (reference 3.22 and 3.23). 6.6.9.1 "F" Test Results Indicative of No Corrosion For purposes of demonstration, five 49 point matrixes with the same Input mean are generated. This will illustrate the case in which the means are indicative of a location which is not corroding, bw ~d :=0.. 4 P~ in~utd :=775 rinputd :=20 Cellsd :=mormtNo DataCells, P inut 1p p actuald :=mean(Cellsd) o actuald :=Stdev 1Cellsd\ d

• The five means, standard deviations, and simulated dates are shown below Dates.:=

769.638 18.813 1993+- 775.647 19.4 365 I actual = 771.334 o actual = 23.726 1994- 243+ 14 779.326 19.422 365 9 773.555 .18.793 19 6.i 2431- 16 365 1997+ 356 365 19+105 19991-5365 The following function simply returns the number of means 'No-of means) which will be used later Noofmeas':=rows 11actual) Noof means = 5 OCLROO000715

Preparer:Pete Tamburro 2113/01 CALCULATION SHEET IV

Subject:

Calqulation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E310-037 0 187 21 of 36 Through September 2000 6.5.9 The "F" Test for Linear Regression In order to determine whether the historical data for each location is indicative of corrosion, the means collected at each location over time are tested using the "F" Test for Linear Regression as follows (reference 3.22 and 3.23). 6.5.9.1 "F" Test Results Indicative of No Corrosion For purposes of demonstration, five 49 point matrixes with the same Input mean are generated. L This will illustrate the case in which the means are indicative of a location which is not corroding. d:=0.. 4 P~ input d := 775 Cellsd := morro(No a inputd :=20 (N DataCells, P inputda d inputd\di~t P~ actual d :=Me(CeIlsd) ° actual :Stdev'Cells d d The five means, standard deviations, and simulated dates are shown below Dates.:= 769.638 *18.813I 1993t- 6 775.647 19.4 365 t actual = 771.334 o actual = 23.726 1994t 243+- 14 779.326 19.422 365 773.555 18.793, 1996. 243 16 365 1997- 356 365 1999+ 105 365 The following function simply returns the number of means tNo of means) which will be used later L No-of means' : rows 19 actual) Noof means = 5 OCLROO000715

AmerGen CALCULATION SHEET Preparer. Pete Tamburro 2/13/01 - I.

Subject:

Statistical Analysis of Drywell Vessel Thickness Data Through September 2000

• Calculation No.

C-1302-187-13310-037 Rev. No. 0 The curve fit equation in which the date IDates) is the independent variable and the measured System Nos. 187 I Sheet 22 of 36 mean thickness of the location ( au) is the dependent variable is then defined as the function "yhat'. This function make use of Mathcad function" intercept" which returns the intercept value of the "Best Fit" curve fit and the Mathcad function "slope" which returns the slope value of the "Best Fit" curve fit. yhat(x, y) := intercept(x, y) + slope(x, y).x The Sum of Squared Error (SSE) Is calculated as follows (reference 3.23) Iast(Dates) SSE:= E (1P actualI yiat (Dates, P actual ))2 i SSE= 44.202 i=0 The Sum of Squared Residuals (SSR) is then calculated as follows (reference 3.23) last( Dates) i y=0 at (Dates, p atl) Mean(Pactual)) 2 SSR= 13.158 Degrees of freedom associated with the sum of squares for residual error is calculated (reference 3.23). DegreeFre ss ::No-of meas - 2 The degrees of freedom for the sum of squares due to regression Is calculated (reference 3.22 and 3.23). DegreeFree reg :=No-ofmen Dividing a sum of squares by its degrees of freedom provides the variance estimate (reference 3.22 and 3.23). MSE:= SSE DegreeFree ss MSE = 14.734 OCLROO000716

AmerG en

           =a CPreparer:

CALCULATION SHEET Pete Tamburro 2/13/01

 *   

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E310-037 0 187 23 of 36 Through September 2000 An estimate of the error standard deviation which is also called the standard error of estimate is calculated (reference 3.23). Standard error :=4M Standad error 3.838 MSR, the population error variance is calculated (reference 3.23) MSR:= SSR DegreeFree rg MSR = 2.632 The MSE is the variance estimate for the mean model. Similarly, the MSR is an estimate for the variance that is explained by the regression model. The ratio of regression variance (MSR) to mean L variance MSE, gives measure of the regression relationship. S Factauil := MSR M* MSE For 95% confidence level the "F critical is calculated as follows (reference 3.22 and 3.23) a0.05 F citcal:= qF( - a,DegreeFree reg, DegreeFree ss F critical =.9.013 The "F"ratio for 95% confidence is calculated: F MAU -actaul F ratio = 0.02 F critical An F-Ratio less than 1.0 occurs when the amount of corrosion which has occurred since the initial measurement is less than the random variations In the measurements or fewer than four measurements have been taken. Ifthe F ratio is less than 1.0, the computer corrosion rate does not reflect the actual corrosion rate but rather is provided as a conservative projection(reference 3.22) Aft "F"ratio of I or less indicates that the data trend is best explained by the grand mean of the data and the the trend has no slope. The variability in the data is within the distribution profile for the data which is normally distributed. Therefore a grand mean pgrand actua) Is best estimate of the thickness of the location. An F-Ratio of 1.0 Is greater occurs when the amount of corrosion which has occurred since the initial measurement is significant compared to the random variations, and four or more measurements have been taken. In these cases the computed corrosion rate more accurately reflects the actual corrosion rate, and there is a very low probability that the actual corrosion rate is zero. The greater the F-Ratio the lower the uncertainty in the corrosion rate (reference 3.22) Where the F-Ratio of 1.0 or greater provides confidence in the historical corrosion rate, the F-Ratio should be 4 to 5 if the corrosion rate is to be used to predict the thickness in the future. To have a high degree of confidence in the predicted thickness, the ratio should be at least 8 or 9 (reference 3.22 calculation). OCLROO000717

AmerGen CALCULATION SHEET Preparer. Pete Tamburro 2115/01

Subject:

Calculation No. Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-F.310-037 Through September 2000 The following shows the results in a graph ptgrand actual. :=mean(p actualj I I I I I Ii iI i i . i i i 820 Individual Inspection Noo0-

                                                      /
                                                       'I means P~ actual      780-x0(                         /5e ttgrand actual 7601-points 740-I          I          I         I          I       I   I IAu -                                                I       I   l 1992 1993        1994      1995       1996       1997   1998 1999   2000 Daws IIIII OCLROO000718

AmerGen CALCULATION SHEET Preparer: Pete Tamburro 2/13/01 inum I

Subject:

Statistical Analysis of Drywell Vessel Thickness Data Through September 2000 6.5.9.2 "F" Test Results Indicative of Corrosion Calculation No. C-1302-187-E3 10-037 Rev. No. 0 System Nos. 187 I Sheet 25 of 36 To illustrate the case In which the location is corroding the five, 49 point matrixes will now be generated with input means which are descending over time. d :=0..4 P input d :=775- (d-4) Sinput d := 20 CeIsd : rorni(No ataCelIs' P inlput 0oinu d f ~d) P actuald :=xmean(Ceilsd) Sactuald := Stdev (Cellsý d di Dates. 779.579 19.489 775.201 17.654 1993+ 6 P actual = 769.326 365 o actual = 19.735 766.983 19.979 1994- 243+ 14. 365 762.322 20.121 199 6 + 243- 16 365 19971- 356 365 1999+ 105 365 Total means :=rows(p actul) Total means = 5 The curve fit equation is then defined for the function "yhat" yhat(x, y) := intercept(x, y) + slope(y, y).x The Sum of Squared Error is calculated last( Dates) SSE:= ftactual, - yhat (Dates, i actual) )' SSE= 0.818 i=0 OCLROO000719

AmerGen CALCULATION SHEET Preparer: Pete Tamburro 2/13/01 - F

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E310-037 0 187 26 of 36 Through September 2000 The Sum of Squared Residuals is then calculated last( Dates) SSR:= (yhat (Dates, p actual). mean(11 actual))' SSR= 184.164 i=0 Degrees of freedom associated with the sum of squares for residual error. DegreeFree =Total means- 2 The degrees of freedom for the sum of squares due to regression, DegreeFre regTotal MSE SSE DegreelFree MSE = 0.273 Standard error := M,* Standard error = 0.522 SSR MSR:=- DegreeFree rg MSR = 36.833 Fa MSR MSE a :=0.05 Fcriticai:=qF~I-aDegreeFereegDegreeFree N 55) F critical " 9.013 The "F" ratio for 95% confidence is calculated: F actaul F ratio= 14.983 F critical The "F" ratio is greater than 1.0, therefore the regression model holds for the data. The curve fit for the five means is best.explainedby a curve fit with a slope. OCLRO0000720

Am erGen CALCULATION SHEET Preparer: Pete Tamburro 2/13/01

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E310-037 0 187 27 of 36 Through September 2000 1 6.9.3 Unear Regression with 95% Confidence Bounds Using data generated in section 6.9.2 the curve fit for linear regression is calculated by the Mathcad functions "slope" and "intercept". Ins :=slope (Dates, A actual) Yb :=intercept (Dates, Aactual) m= -2.702 Y b = 6.1655i01 The predicted curve is calculated over time where 'year predict" is time (independent variable), and "Thick predict" is thickness (dependent variable). Remainingpi life:= 13 f:= 0.. Remaining pllife- I Year predictf 1985+ f.2 Thick predict := m s'year predict + Y b The 95% Confidence ("l., t") curves are calculated as follows (reference 3.3) at:=0.05 Thick actualmean := mean( Dates) 2 sum :=z (Datesd- mean(Dates)) d upperr:= Thick predictf 2 oal- 1 Iq(- predict, Thick culen2 2 r (d*1) sum lowerf :=Thick predictf" [le (fa predictf Thc 1culený

           +.[t!kI - 2 Total     a      2 '-Standard 1          err'1       (d+- )                    sum                j OCLR00000721

_____________________________________U Preparer: Pete Tamburro 2/13/01

       ,me                                                   CALCULATION SHEET 0     

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E310-037 0 187 27 of 36 Through Septemriber 2000 6.9.3 Unear Regression with 95% Confidence Bounds Using data generated in section 6.9.2 the curve fit for linear regression is calculated by the Mathcad functions "slope" and "intercept". Ins :=slope (Dates, p actual) Y b :=intercept (Dates, P actual) ms = -2.702 Yb = 6"16-h0I3 The predicted curve is calculated over time where 'year predict" is time (independent variable), and "Thick predict" is thickness (dependent variable). Remainingpl life:= 13 f:= 0.. Remaining Pl._life- I year predictf,:= 1985t-f.2 S Thickpredict :`ms'Year predict i.Y b The 95% Confidence ("1.a t") curves are calculated as follows (reference 3.3)

   '*a              t := 0.05 Thick actualmean := mean(Dates) 2 sum :=ZE (Datesd - mean(Dates))

d upperr: Thick predict. (year predict- Thick actualmean2 Sa t 2.tndrmeans -2j.Staard (d 1-) + sum

                  + qtIJ I- -Total                                 J                                              mean 1 lowerr: Thick predictf.-
     *                       - e,Total mes qt          fl *n2-dl 2.Standard error] + ( d + 1)

(YI e 7 -rie TMick u+- s um actualmean 2 J I OCLROO000721

AmerGen CALCULATION SHEET Preparer: Pete Tamburro 2/13/01 m

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C;-1302-187-E310-037 0 187 28 of 36 Through September 2000 Therefore the following is a plot of the curve fit of the data generated in section 6.9.2 and the Upper and Lower 95% confidence Intervals. The Upper and Lower 95% Confidence Intervals are the two curves shown below which bound the data points and the curve fit. Individual Inspection 80o-o means Thick predict 780 upper lower = 11actual 760 0 2000 2005 yearedpredicya redicbycflpredict, D9W OCLROO000722

Preparer: Pete Tamburro 2/13/01 Amereen CALCULATION SHEET ,

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E310-037 0 187 29 of 36 Through September 2000 7.0 Calculation 7.1. Elevation 50' 2" 7.1.1 Bay 5 Area D12, Feb. 1990 through Sept. 2000 Refer to Appendix #1 for the complete calculation. Eleven Inspections have been performed at this location. A plug lies within this location and therefore the nine points over the plug are eliminated from the calculation (see section 5.2). The data collected in Sept. 2000 inspection is normally distributed after the nine points are eliminated. In addition to the nine points the following adjustments have been made over time:

1) Point 9 is a significant pit and is trended separately

2) Points 1, 4 and 37 in the 4/25/90 data set are much greater than the preceding or succeeding measurements. Therefore these three points were dropped from the 4/25/90 data (ref. 2.22).

3) Points 3 and 36 in the 11102/91 data set are much greater than the preceding or succeeding measurements. Therefore these two points were dropped from the 11102/O91 data (ref. 2.22).

The data indicates no ongoing corrosion since Feb 1990 Point 9 Analysis of this point prior to Sept. 2000 (ref. 3.22) indicated that there was a potential corrosion rate. The addition of the Sept. 2000 data now drives the F-ratio for this point to 1.7, which now confirms that a rate exists. The calculated rate is 1.3 milsper year. This corrosion rate is very small. Projection based on this corrosion rate using the 95% lower confidence interval shows that this point will not corrode to less than the minimum required thickness by the years 2009 or 2029. 7.1.2 Bay 5 Area 5, March 1990 through Sept. 2000 Refer to Attachment #2 for the complete calculation. Ten Inspections have been performed at this location. Previous data sets were not normally distributed since there is a large thin area in the center of the grid and several smaller patches on the periphery. Past calculations separated these regions into subsets. The thinner area has 16 points and the thicker area has 32 points. Analysis of past subsets shows that both data sets are normally distributed. The Sept. 2000 data is consistent With past data. In addition, point 17 is significantly thinner than these two areas. Therefore point 17 is trended separately OCLR00000723

Results of the two subsets are described below: Thinner Points This subset is normally distributed. The F-ratio for this subset indicates that there is no ongoing corrosion. Thicker Points This subset is normally distributed. The F-ratio for this subset (1.2) indicates that there is ongoing corrosion at a rate of 0.6 mils per year. This corrosion rate is very small. Projection based on this corrosion rate using the 95% lower confidence interval shows that it will not corrode to less than the minimum required thickness by the year 2009 or 2029. Point 17 The F-ratio for this point indicates no on going corrosion. 7.13 Bay 13 Area 31, March 1990 through Sept. 2000 Refer to Appendix #3 for the complete calculation. Ten Inspections have been performed at this location. Previous data sets have not been normally distributed since there is a thinner area on the left edge of the grid. Past calculations have separated these regions in two subsets. The thinner area has 16 points and the thicker area has 33 points. Analysis of past subsets shows that both data sets are normally distributed. The Sept. 2000 data is consistent with this past data. Results of the two subsets are described below: Thinner Points This subset is normally distributed. The F-ratio for this subset indicates that there is no ongoing corrosion. Thicker Points This subset is normally distributed. The F-ratio for this subset indicates that there is no ongoing corrosion. 7.1.4 Bay 13 Area 23 March 1990 through Sept. 2000 Refer to Appendix 84 for the complete calculation. Ten Inspections have been performed at this location. Previous data sets were not normally __ distributed since there is a large thinner area in the center of the grid. Past calculations separated

• Wthese regions into subsets. The thinner area has 15 points and the thicker area has 32 points.

Analysis of past subsets shows that both data sets are normally distributed. The Sept 2000 data is consistent with this past data. OCLR00000724

AmerGen CALCULATION SHEET Preparer: Pete Tamburro 2/13/01

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-! 302-187-E310-037 0 187 31 of 36 Through September 2000 Also point 26 is significantly thinner than these two areas. Therefore point 26 is trended separately Point 27 in the November 1991 data set is much less than the preceding or succeeding measurements. Therefore this point was dropped from the 11/91 data (ref.2.22). Results of the two subsets are described below: Thinner Points This subset is.normally distributed. The F-ratio for this subset indicates that there is no ongoing corrosion. Thicker Points This subset is normally distributed. The F-ratio for this subset indicates that there is no ongoing corrosion. Point 26 The addition of the Sept. 2000 data now drives the F-ratio to 1.8, which now confirms that a rate does exist. The calculated rate is 1:5 mils per year. This corrosion rate is very small. Projection based on these corrosion rates using the 95% lower confidence interval shows that this point will not corrode to less than the minimum required thickness by the year 2009 or 2029. OCLR00000725

Preparer: Pete Tamburro 2/13/01 ACALCULATION SHEET

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E310-037 0 187 32 of 36 Through September 2000 7.2 Elevation 51' 10" through September 2000. 7.2.1 Bay 13 Area 32 April 1990 through Sept. 2000 Refer to Appendix #5 for the complete calculation. Nine inspections have been performed at this location. Previous data sets were not normally distributed since there is a "T" shaped thinner area along the right side of the grid. Past calculations separated these regions into subsets. The thinner area has 13 points and the thicker area has 32 points. Analysis of past subsets shows that both data sets are normally distributed. The Sept. 2000 data is consistent with this past data. In addition, points 20, 23,25, and 28 are significantly thinner than these two areas. Therefore these points are trended separately Point 11 in the 5/23/91 data set was much less than the preceding or succeeding measurements. Therefore this point was dropped from the 5/22191 data (ref. 2.22). Results of the two subsets are described below: Thinner Points This subset is normally distributed. The F-ratio for this subset indicates that there is no ongoing corrosion. Thicker Points This subset is normally distributed. The F-ratio for this subset indicates that there is no ongoing corrosion. Points 20, 23, 25, and 28 The F-ratio for these points indicates no on going corrosion. II II OCLROO000726

Amer~en CALCULATION SHEET Preparer: Pete Taraburro 2/13/0 1

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E3310-037 0 187 33 of 36 Through September 2000 7.3 Elevation 60' 10" 7.3.1 Bay 1 Area 5-22 April 1990 through Sept. 2000 Refer to Appendix #6 for the complete calculation. Four inspections have been performed at this location. Data collected in all four inspections are normally distributed after point 48 is eliminated. Point 48 is a significant pit and is trended separately The data indicates no ongoing corrosion since Feb 1990 Point 48 Point 48 may be experiencing a corrosion rate of 4.5 mils per year. This relatively greater calculated rate may be due to the limited amount of inspections. The methodology and analysis results in confidence levels with less inspection information. Never-the-less projection based on this corrosion rates using the 95% lower confidence interval, which is significantly more conservative than other locations, show that this point will not corrode to less than the minimum

• required thickness by the year 2009.

Since the amount of data on this pit is limited to 4 inspections, the upper and lower 95 % confidence intervals are very broad and conservative. This results in a projected lower 95% confidence thickness value for the year 2029 which is less than the local minimum required thickness. More inspections should narrow the upper and lower 95 % confidence intervals. Assuming the rate is constant over time, it is expected that future projections should show that this pit will not corrode to less than the minimum required local thickness by the year 2029. OCLROO000727

Preparer: Pete Taniburro 2/13/01 Amer~enCALCULATION SH'EET

Subject:

Calculation No. Rev. No. System Nos. Sheet Statistical Analysis of Drywell Vessel Thickness Data C-1302-187-E310-037 0 187 34 of 36 Through September 2000 7.4 Elevation 87' 5" 7.4.1 Bay 9 Area 20 November'1987 through Sept. 2000 Refer to Appendix #7 for the complete calculation. Twelve inspections have been performed at this location. UT data collected in September 2000 is normally distributed. Point 13 in the May 1992 data set was much less than the preceding or succeeding measurements. Therefore this point was dropped from the May 1992 data (ref. 2.22). Based on a calculated F-ratio of 1.2, this location is experiencing minor a corrosion rate of approximately 1.2 mils per year. This corrosion rate is very small. Projections based on this corrosion. rate using the 95% lower confidence interval shows that they will not corrode to less than the minimum required thickness by the year 2009 or 2029. 7.4.2 Bay 13 Area 28 November 1987 through Sept. 2000 Refer to Appendix #8 for the complete calculation. Twelve inspections have been performed at this location. Previous data sets were not normally distributed. Past calculations separated out points 1, 2, 22, 25, 26, 36, and 48. Analysis of past data sets without these points show that the data is normally distributed. The Sept. 2000 data is consistent with this past data. Analysis of the data indicates no ongoing corrosion since Feb 1990 Points 1, 2,22,25, 26,36, and 48 Analysis of these individual points show no ongoing corrosion with the possible exception of point 25 which has an F-ratio of 0.96. This point may be experiencing a corrosion rate of 3.0 mils per year. Projection based on these corrosion rates using the 95% lower confidence interval shows that these point will not corrode to less than the minimum required thickness by the year 2009 or 2029. 7A.1 Bay 15 Area 31 November 1987 through Sept. 2000 Refer to Appendix #9 for the complete calculation. Twelve inspections have been performed at this location. Previous data sets have not been normally distributed. Past calculations have separated out points 34 and 35. Analysis of past OCLR00000728

AmerGen Calculation Sheet Appendix 9

Subject:

Calc. No. Rev. No. SystemNo. Sheet No. Drywell Corrosion C-4301-187-e310-037 0 187 A9 - of 25 Appendix 9 - Elevation 87' 5" Bay 15, Area 31 Sept. 17, 2000 Data The data shown below was collected on 10/17/2000 (reference NDE data sheet 2000-034-009). page:= U:'V.VD86310.txt Points 49:= showcells(page, 7, 17) 0.639 0M639 0.627 0.631 0.606 0.614 0.631 0.647 0.639 0.637 0.629 0.639 0.637 0.63 0.631 0.65 0.629 0.628 0.64 0.619 0.627 Points 4 9 = 0.638 0.645 0.594 0.627 0.613 0.623 0.63 0.643 0.63 0.632 0.615 0.624 0.564 0.614 0.649 0.619 0.617 0.602 0.616 0.602 0.609 0.643 0.639 0.61 0.636 0.617 0.612 0.64 Cells := convert(Points 49,7) No DataCeils := length(Cells) The pits at point 34 and 35 are removed from the data an will be trended separately (ref. 3.22). Cells := Zero one(Cells,No DataCells, 34) Cells := Zero one(Cells,No DataCells, 35) Cells := deletezero ceils(Cells,No DataCells) No DataCells  := length(Cells) OCLR00000937

U AmerGen Calculation Sheet Appendix 9

Subject:

Cal c. No. Rev. No. System No. Sheet No. Drywell Corrosion C-I1301-187-310-037 0 187 A9 -2 of 25 Mean and Standard Deviation lt actual:= mean(Cells) 1 actual = 627.532 aactual :=Stdev(Cells) " actual = 13.518 Standard Error a actual a Standard error

                       ./No DataCells                                            Standard error  1.972 Skewness (No DataCells).,(Ceds - Pacua)

SkewnessN (No DataCells-

. .. 1)-(NO DataCells - 2).-(13 actual)*

                                     .(Oat)3su                                        Skewness = -0.485 Kurtosis p actual)4 Kurtosis:             NO DataCeils-(No DataCelis+ 1)T (Cells -

Kurtosis = -0.429 (No DataCelI,- I)'(No DataCells- 2).(NO DataCells-" 3).(; actual) (N 3.NO DataCells- 1)2 2 (No D,%Cells- )-(No.v;Cells- 3.) we OCLROO000938

AmerGen Calculation Sheet Appendix 9

Subject:

Calc. No. Rev. No. System No. Sheet No. Drywell Corrosion C-1301-187-e310-037 0 187 AS -3 of 25 Normal Probability Plot In a normal plot, each data value is plotted against what its value would be if it actually came from a normal distribution. The expected normal values, called normal scores, and can be estimated by first calculating the rank scores of the sorted data. j:=0-. last(Ceils) srt :=sort(Cells) Then each data point is ranked. The array rank captures these ranks r :=j + I xsrt-sr ranN. p rows(Cells) + T The normal scores are the corresponding pth percentile points from the standard normal distribution: x :=1 N-Score, :=roofcnorin(x) - (pi),x] OCLR00000939

AmerGen Calculation Sheet Appendix 9

Subject:

Cale. No. Rev. No. System No. Sheet No. Drywell Corrosion C-1301-187-e310-037 0 187 A9 -4 of 25 Upper and Lower Confidence Values The Upper and Lower confidence values are calculated based on .05 degree of confidence 'a" a :=.05 Ta:= -) Ta = 2.011 Lower 95%Con P actual - Ta. 1 actual Lower 95%Con = 623.567

                                          .jFo DataCells Upper 95%'Con = actual+                    actual
                                         ,PN ýDataC~ells                Upper 95%Con = 631.496 These values represent a range on the calculated mean in which there is 95% confidence.

Graphical Representation Distribution of the "Cells" data points are sorted in 1/2 standard deviation Increments (bins) within +/-3 standard deviations M 010 Bins := Make bins(P actual,a actual) 1 3 5 Distribution hist(Bins, Cells) Distribution = 10 The mid points'of the Bins are calculated _F1 - 4 2 k:=0.. II Midp--n--k "- (Bin + Binsk I Midpoints 0 0 The Mathcad function pnorm calculates a portion of normal distribution curve based on a given mean and standard deviation normal curveo :=pnor (Binslp actual' actual) normal curvek,:=pnorm(Binsk÷ ,,i actual,' actual) - pnorm(Binsk,p actual,° actual) normal curve :=No DataCells-normal curve OCLR00000940

AmerGen Calculation Sheet Appendix 9

Subject:

Calc. No. Rev. No. System No. Sheet No. Drywell Corrosion C-1301-187-e310-037 0 187 A9 -5 of 25 Results For Elevation 86' 5" Bay 15, Area 31 Sept. 17, 2000 The following schematic shows: the the distribution of the samples, the normal curve based oni the actual mean and standard deviation, the kurtosis, the skewness, the number of data points, and the the lower and upper 95% confidence values. Below is the Normal Plot for the data. Data Distribution I actual = 627.532

                                                                                             < actual = 13.518 Distribution
          .L                                                                                 Standard error = 1.972 normyI curve Skewness = -0.485 Kurtosis = -0.429 680 Midpoints, Midpoints Lower 95%Con = 623.567                   Upper 95%Con = 631.496 Normal Probability Plot 2

Based on the Normal Probability x Plot, Skewness, x and the Kurtosis x this data is normally NS=ej distributed. xxx XXX

              -2       X I         I          I            I          I 590        600       610        620          630        640       650 OCLR00000941

AmerGen Calculation Sheet Appendix 9

Subject:

Calc. No. Rev. No. System No. Sheet No. Drywell Corrosion C-13014187-0310-037 0 187 A9 -6 of 25 Elevation 86' 5" Bay 15, Area 31 Trend Data from Feb 1990 to Sept 2000 Is retrieved. d :=O For Nov. 10 1987 page : Datesd:= Day year( 1 , 10,1987) MA.D863iN87bc Points 49 showcells(page, 7,17) Data

                              '0.655    0.648  0.639   0.65     0.62   0.627   0.641 0.659    0.643' 0.646   0.64     0.634  0.651   0.641 0.628   0.657  0.673   0.638    0.654  0.629   0.632 Points 49 =   0.65    0.652   0.646   0.638    0.619  0.633   0.634 0.656    0.633  0.637   0.623    0.634  0.568   0.62 0.65     0.63   0.625   0.607    0.625  0.606   0,614 0.649   0.648   0.615   0.649    0.628  0.628   0.647 nnn := convert(Points 49,7)              No DataCells := length(nnn)

The pits at points 34 and 35 are removed from the data and will be trended separately (ref 3.22). Pit 3 4 d :=Get.Pit datann, N DataCells, 34) 35 Pit d :=Get Pit data(nrm,No DataCells, 3 5 ) d dta(J-These points are deleted from the mean calculation nnn :=Zero one( nn, No DataCells, 34) nnn := Zero one(nnn,No DataCells,35) Cells := deletezero celis(nnn, No DataCells) measuredd lmeasuzred d :=rnean(Cells) a measuredd :Stdev(Cells) Stanarderrod PO DataCells OCLROO000942

AmerGen Calculation Sheet Appendix 9

Subject:

Caic. No. Rev. No. System No. Sheet No. Drywell Corrosion C-1301-187-e310-037 0 187 A9 -7 of 25 d :=d t- I For July 20 1988 page:= Datesd := Day year( 7 , 2 0 , 1988) U:A.D8631J881tx Points 49 =showcells(page, 7,17) Data 0.651 0.645 0.633 0.643 0.615 0.626 0.634 0.651 0.642 0.643 0.641 0.651 0.644 0.638 0.627 0.654 0.654 0.633 0.65 0.652 0.634 0.634 Points 49 = 0.644 0.652 0.654 0.635 0.616 0.632 0.652 0.63 0.64 0.622 0.635 0.566 0.623 0.645 0.627 0.619 0.604 0.624 0.605 0.617 0.648 0.646 0.613 0.639 0.622 0.619 0.643 nnn := convert(Points 49,7) No DataCells':= length(nnn) Pit 34d =:Get-Pit data(nnn,No DataCells, 34) Pit *3 5d :';Get-Pitdatafnnn,No

                                                                       -        k=       DataCells, 35)

These points are deleted from the mean calculation nnn := Zero one(nnn, No DataCells, 34) nnn:=Zer (nnn,No DataCells, 35) Cells - deletezero cells(nnn,No DataCells) crmeasureid P measured' := mean(Cells) measuredd := Stdev(Cells) Standard errord : r

                                                                                          .INo
                                                                                             . DataCells OCLROO000943

AmerGen Calculation Sheet Appendix 9

Subject:

Cale. No. Rev. No. System No. Sheet No. Drywell Corrosion C-1301-187-e310-037 0 187 A9 -8 of 25 d :=d+-I For Oct. 8 1988 page:= Datesd :=Dayyea(10,8, 1988) UA.I.D863108B.txt Points 49:= showcells(page, 7,17) Data

                           '0.651   0.645 0.632     0.642 0.618     0.622    0.636 0.655 0.641 0.644       0.638 0.63      0.643    0.637 0.629 0.654 0.645       0.635    0.649  0.649    0.643 Points 49 = 0.651 0.65 0.619          0.636    0.616  0.632    0.636 0.664 0.63 0.635        0.619    0.634  0.562 0.626 0.65 0.646 0.622       0.605    0.63   0.608 0.622 0.654 0.645 0.612      0.642    0.628  0.622 0.643 nnn := convert(Points 49,7)               No DataCells:= length(nnn)

Pit 34d:= Get-Pit data (nnn,No DataCells, 3 4) Pit 35d := Get Pit data(nnn,No DataCells, 3 5) These points are deleted from the mean calculation nnn := Zero one (nnn, No DataCells, 34) nnn :ýZoro one(nnnNo DataCells, 35 ) Cells:= deletezero celis(nnn, No DataCells) a measuredd I measuredd := mean(Cells). ameasuredd~:=StdeV(CellS) Standard errord OCLROO000944

AmerGen Calculation Sheet Appendix 9

Subject:

Calc. No. Rev. No. System No. Sheet No. Drywell Corrosion C-1 30-1 87-e310-037 0 187 A9 -9 of 25 d:=d-I- I For June 26 1989 page:= 0 Datesd := Day year( 6 , 2 6 , 1989) U-1.108631 JURe8Ix Points 49 :=showcells(page,7,17) Data 0.654 0.649 0.636 0.643 0.619 0.629 0.637 0.656 0.644 0.649 0.641 0.651 0.649 0.642 0.63 0.658 0.651 0.641 0.654 0.629 0.636 Points 49 0.647 0.648 0.623 0.637 0.617 0.631 0.636 0.65,5 0.633 0.641 0.623 0.632 0.576 0.619 0.648 0.636 0.624 0.611 0.627 0.61 0.618 0.653 0.667 0.629 0.649 0.627 0.621 0,647 nnn *=convert(POints 49,7) NO DataCells - length(nnn) Pit 3 4 d := Get-Pit data(nnn,No DataCells, 34/ Pit 35d := GetPit data(nnnNo DataCellsI 35) These points are deleted from the mean calculation nnn =Zero One(nnn, No DataCells, 34 ) nnn := Zero one(nnn, No DataCells, 35) Cells:= deletezero cells(nnnNo DataCells) a measuredd 9 measuredd := mean(Cells) a measuredd := Stdev(Cells) Standard errord -INo DataCells OCLR00000945

AmerGen Calculation Sheet Appendix 9

Subject:

Calc. No. Rev. No. System No. Sheet No. Drywell Corrosion C-1301-187-e310-037 0 187 A9 -10 of 25 d :=d- I For March 28 1990 page:= Datesd := Day year(3 , 2 8 ,1990) UUDB8631MarchgOltd Points 49 := showcells(page, 7, 17) Data

                                      *0.653    0.646 0.644   0.644 0.621 0.626 0.635 0.654    0.643  0.647   0.64 0.637 0.649 0.639 0.638    0.658  0.653 0.656 0.676 0.63 0.642 Points 49     0.646     0.65  0.631 0.644 0.633 0.652 0.636 0.682     0.657 0.66    0.634 0.635 0.573 0.624 0.65      0.644 0.636 0.613 0.627 0.61 0.616 0.65 1    0.648 0.614 0.646 0.655 0.622 0.647 nnn   convert(Points   49 ,7)             No DataCells:= length(nnn)

Pit 34d:= Get-Pit data(nnn,fNo DataCells, 34) Pit 35d:=GetPit datatnnn, No DataCells, 3 5 ) These points are deleted from the mean calculation nnn := Zero one (nnn, No DataCells, 34) nnn := Zeroone(nnn,No DataCells, 35) Cells:= deletezero cells(nnn, NO DataCells) measuredd p measuredd :=mean(Cells) a measuredd :=Stdev(Cells) Standard errord er ,No DataCells IOCLR00000946

AmerGen Calculation Sheet Appendix 9

Subject:

Cale. No. Rev. No. System No. Sheet No. Drywell Corrosion C-I301-187-e310-037 0 187 A9 -11 of 25 d :=dt-I For Feb, 23 1991 page:= Datesd :Day year(2,23, 1991) U:UD8631-F91 txt Points 4 9 :=showcells(page,7,16) Data

                              '0.645    0.641  0.629   0.639   0.613 0.624 0.631 0.646   0.637  0.639   0.633   0.629 0.641 0.634 0.62    0.648  0.63    0.629   0.645 0.624 0.626 Points 49 = 0.637      0.637  0.597   0.629   0.611 0.625 0.629 0.644   0.624  0.631   0.615   0.626 0.556 0.615 0.642   0.623  0.616   0.602   0.619 0.601 0.611 0.644   0.624  0.606   0.636   0.619 0.611 0.639 nnn :=convert(Points 49,7)                No DataCells:= length(nnn)

Pit 3 4 d:= Get-Pit data(nnn,No DataCells, 34) 35 Pit'35 :=GetPit data(nnn,NoDataCells, ) These points are deleted from the mean calculation nnn :=Zero one(nnn,No DataCells, 34) nnn :=Zero one(nnn, NO DataCells, 3 5) Cells := deletezero .lls(nnn,No DataCells) t measuredd p measuredd := mean(Cells) o measuredd := Stdev(Cells ) Standarderrord :=

                                                                                           ,Jio D..a. ll DataCells 0CLR00000947

AmerGen Calculation Sheet Appendix 9

Subject:

Caic. No. Rev. No. System No. Sheet No. Drywell Corrosion C-1301-187-e310-037 0 187 AS -12 of 25 d :=d+ I For May 23, 1991 page : Datesd := Day yea5,23, 1991) UX.1D8831M91-cutd Points 4 9 := showcells(page, 7, 0) Data

                                    '0.647    0.641   '0.631    0.636    0.613 0.621   0.63 0.649   0.637    0.641    0.634    0.633 0.641   0.633 0.62    0.649     0.628   0.629    0.624 0.622   0.626 Points 49 = 0.639       0.642     0.595   0.629    0.611 0.625   0.629 0.646   0.624    0.632    0.614    0.627 0.556   0.615

~mhd 0.642 0.623 0.616 0.601 0.621 0.603 0.61 0.643 0.624 0.607 0.636 0.619 0.612 0.613 nnn := convert(Points 49,7) No DataCells:= length(nnn) Pit 34d := Get-Pit data(nnn, No DataCells, 34) Pit 35 d :=GetPit data(nnn'NO DataCells, 35 ) These points are deleted from the mean calculation non :=ZrOone(nnn.No DataCelis, 3 4 ) non := Zero one(nnn, No DataClls, 35) Cells :=deletezero cellslnnn,No DataCells) Fmeasuredd p measuredd := mean(Cells) a measuredd := Stdev(Cells) Standard errord

                                                                                                     ,No DataCells r

OCLR00000948

AmerGen Calculation Sheet Appendix 9

Subject:

Calc. No. Rev. No. System No. Sheet No. Drywell Corrosion C-1301-187-e310-037 0 187 A9 -13 of 25 d :=d+ I For May 30 1992 page Datesd := Day year(5,30, 1992) U:I.AD8631_M92.txt Points 49 :=showcells(page, 7, 1) Data 0.65 0.639 0.627 0.635 0.614 0.621 0.629 0.651 0.635 0.64 0.635 0.608 0.642 0.635 0.622 0.65 0.633 0.631 0.65 0.627 0.634 Points 49 = 0.64 0.643 0.665 0.632 0.616 0.629 0.63 0.651 0.628 0.635 .0.617' 0.629 0.565 0.613 0.645 0.639 0.619 0.604 0.623 0.607 0.611 0.64 0.639 0.607 0.635 0.619 0.611 0.638 nnn := convert(Points4a9, 7 s3 No DataCells :=Iength(nnn) Pit 344 :=ýGet-Pit data(nnn, NO DataCells, 34) Pit 35d:=GePit data(nnn,No DataCells, 3 5 ) These points are deleted from the mean calculation

. nnn :=Zero one. nnn,No DataCells, 34)                nnn := Zero one(nnn, No DataCells, 3 5)

Cells: deletezero ceiis(nnnNo DataCells) Standard ro measuredd Pmeasured d :=mean(Cells) a measured d := Stdev(.Cells) CrNo DataCells OCLROO000949

I" AmerGen

Subject:

Drywell Corrosion Calculation Sheet Calc. No. C-1301-187-e310-037 Rev. No. 0 System No. Appendix 9 Sheet No. 187 A9 .14 of 25 d :=d÷ I For Dec. 5 1992 page := Datesd :=Day Year(12, 5, 1992) U:LAD8631_92.tx Points 49 :showcells(page, 7,16) Data 0.644 0.643 0.632 0.634 0.612 0.622 0.632' 0.649 0.64 0.642 0.629 0.642 0.639 0.633 hi 0.633 0.647 0.632 0.63 0.643 0.614 0.627 Pd Points 4 9 = 0.641 0.647 0.601 0.634 0.612 0.629 0.631 0.648 0.628 0.634 0.623 0.625 0.567 0.615 0.643 0*.621 0.623 0.604 0.62 0.602 0.606 0.645 0.643 0.615 0.638 0.617 0.618 0.643 nnn := convert(Points 49,7) No

                                                       'DataCells:= length(nnn)

Pit 3 4d := Get-Pit data(nnn,No DataCells, 34) 35 Pit 35:= Get-Pit data(nnn, No-DataCells, ) These points are deleted from the mean calculation nnn := Zero one(nnn, No DataCells; 34) nnn :=Zero one(nnn,No DataCeils,35) Cells := deletezero cells(nnnNo DataCells) Smeasured 4 pmeasured := meari(Ceils) (Fmeasuredd :=Std~ev(Cells) Standard errord := , t ao FO"DataCells OCLROO000950

AmerGen Calculation Sheet Appendix 9

Subject:

Calc. No. Rev. No. System No. Sheet No. Drywell Corrosion C-1301-187-e310-037 0 187 A9 -15 of 25 d :=d.t- I For Sept. 14 1994 page: = Datesd :=Day yea(9,14,1994) U:\..%D8631_94.txt Points 49 :=showcells(page,7,16) Data 0.648 0.643 0.614 0.638 .0.614 0.626 0.633' 0.654 0.645 0.643 0.633 0.643 0.645 0.637 0.63 0.653 0.634 0.633 0.646 0.619 0.632 Points 49 . 0.642 0.65 0.6 0.636 0.617 0.63 0.636 0.651 0.635 0.639 0.623 0.629 0.564 0.617 0.656 0.623 0.624 0.609 0.622 0.605 0.611 0.674 0.644 0.616 0.641 0.622 0.617 0.646 nnn convert(Points 49, 7) No DataCells :=length(nnn) Pit 3 4d:= Get-Pit data(nnn, No DataCells, 34) Pit 35  := GetPit data nnn,No DataCells, 35) These points are deleted from the mean calculation nnn :=Zero one(nnn,No DataCells, 34) nnn :=Zero on(nnn, No DataCells' 5 Cells:= deletezero cells(nnn, NO DataCells)

=, omeasuredd a measuredd :=Stdev(Cells) Standard errord 11 leSUre d:= men(Cells)

                                                                                             ,PNo DataCells OCLR00000951

AmerGen Calculation Sheet Appendix 9

Subject:

Calc. No. Rev. No. System No. Sheet No. Drywell Corrosion C-1301-187-e310.037 0 187 AS -16 of 25 d :=di- I For Sept. 9 1996 page : Datesd :=Day year(9,9, 1996) U:I..D8631_96.txt Points 49 := showcells(page, 7, 17) Data 0.647 0.64 0.627 0.636 0.61 0.622 0.6362 0.653 0.639 0.642 0.637 0.629 0.644 0.635 0.623 0.653 0.635 0.634 0.644 0.625 0.629 Points 49 = 0.643 0.643 0.601 0.633 0.615 0.628 0.634 0.651 0.628 6.635 0.618 0.631 0.562 0.521 0.644 0.627 0.622 0.605 0.627 0.608 0.619 0.649 0.646 0.612 0.641 0.624 0.616 0.647 nnn :=convert(Points 49,7) No DataCells := length(nnn) Pit 34d :=GetPit data(nnnNo DataCells, 34) Pit 3 5 d := Get-Pit data(nnn,No DataCells, 35) These points are deleted from the mean calculation nnn := Zero one (nnn, No DataCells, 34) nnn := Zero one(flflniNo DataCe11s, 35) Cells:= deletezero cells(nnn, No DataCells) a measuredd measured p measured4 := mean(Colls) ameasuredd := Stdev(Cells) Standard errord :- a

                                                                                            ,INo DataCells OCLR00000952

AmerGen Calculation Sheet Appendix 9

Subject:

Calc. No. Rev. No. System No. Sheet No. Drywell Corrosion C-I301 -187-e310-037 0 187 A9 -17 of 25 req d :=d- I For Sept. 16 2000 page Datsd:= Day year(9,16,2000) 12 U.I.M8631-00AXt Points 4 9 := showcelis(page, 7,17) Data 0.639 0.639 0.627 0.631 0.606 0.614 0.631 0.647 0.639 0.637 0.629 0.639 0.637 0.63 0.631 0.65 0.629 0.628 0.64 0.619 0.627 Points 49 = 0.638 0.645 0.594 0.627 0.613 0.623 0.63 0.643 0.63 0.632 0.615 0.624 0.564 0.614 0.649 0.619 0.617 0.602 0.616 0.602 0.609 0.643 0.639 0.61 0.636 0.617 0.612 0.64 nnn := convert(Points 4 9 ,7) N1o DataCells :=length(nnn) Pit 34d:= Get-Pit data (nanNo DataCelIs,34) d Pit data (nnnNo DataCells, 3 5) These points are deleted from the mean calculation nn.n :=Zero one(nnn, No DataCells, 34 ) nnn := Zero onc(nnn.No DataCells, 35) Cells := deletezero cells(fln, No Datacels)

                                                                                                 *omeasuredd d    l 11measuredid'rn~ean(Cells)          a mneasuredd :=Stdev(Cells)          Standard errordsro OCLR00000953

AmerGen Calculation Sheet Appendix 9

Subject:

Calc. No. Rev. No. System No. Sheet No. Drywell Corrosion C-1301-187-e310-037 0 187 AS -18 of 25 Below are matrices which contain the date when the data was collected, Mean, Standard Deviation, Standard Error for each date. 1.9880103 1.9890103 1.9894,103 I.99.IO03 Dates = 1.991.10 1.991.10 1.992.1O3 1.997.10 i81" 2.001.103 17.7,77 ME 637.894 14.318 635.702 1.948 13.636 635.936 1.918 13.428 T38.043  ! .908 13.356

                  ýi 1.9-15                              2.117                     14.819 627.681          Standard error =      1.847     a measured =    12.928 P measured  =

627.021 1.892 13.243 631.064 1.953 13.674 630 ] .88 00T3 1-5 8 633.213 2.152 15.064

                  ý31 6-38                               1.862                     13.032 627.532                                1.931                  a T3 ý 1-9 MIIIMý-                               I OCLR00000954

AmerGen Calculation Sheet Appendix 9

Subject:

Calo. No. Rev. No. System No. Sheet No. Drywell Corrosion C-I 301-187-e31 0-037 0 187 A9 -19 of 25 Total means:= rows(it measured) Total means = 12 The F-Ratio is calculated last(Dates) SSE:= Z (1=measuredi- Ybat(DatesP measured)1 ) SSE = 166.324 1=0 last(Dates) SSR:= (yhat(Dates SSR = 86.812 Pmeasured) , - mnkll measured), i=0 DegreeFree s Total means - 2 DegreeFreeg := I MSE: SSE MSR := SSR DegreeFree ss MSE = 16.632 DegreeFree reg MSR = 86.812 StGrand er:=,[-F StGrand . - 4.078 611 L. OCLROO000955

AmerGen Calculation Sheet Appendix 9 0

Subject:

Drywell Corrosion Cal¢. No. C-1301-187-e310-037 Rev. No. 0 System No. 187 Sheet No. A9 -20 of 26 MSR Factaul :MSER a :=0.05 F critical := qF ( - a, DegreeFree reg, DegreeFree .. ) L F actaul SratioF:= 'critica I F ratio = 1.051 Therefore the curve fit of the means seems to have a slope and the grandmean is not an accurate measure of the thickness at this location i:= 0.. Total means I pgrand measuredi mean(9 measured) ogrand measured 0grand measured :zStdev (F measured) GrandStandard error0 : jToaimeans Plot of the grand mean and the actual means over time X 640 hi x X 635

                                           --  -  -  -  o-  ..- - . - . - .. ----------
                                                                                  . .     .    ----X  .  .  .  .-----------   ----------

P' measured XXX X X P&grnd measured - X - 630 625 620 I I I I I I I! 198S 1990 1992 1994 1996 1998 2000 2002 Dates pgrand measuredo = 633.137 GrandStandard error = 1.3S5 OCLR00000956

AmerGen Calculation Sheet Appendix 9

Subject:

Calc. No. Rev. No. System No. Sheet No. Drywell Corrosion C-1301-187-e310-037 0 187 A9 -21 of 25 Therefore the corrosion rate is calculated and compared to the minimum required wall thickness at this elevation ms:=slope (Dates, p measuredj m S = -0.746 Yb :=intercept(Dates,/* measured) Yb = 2.119-103 The 95% Confidence curves are calculated S:=0.05 k: 923 f :=T 0 ik:- c year predictf= 1985 +-f-2 Thickpredict :=m s-year predict+4 Y b 2 Thick actualmean := mean(Dates) sum := . (Datesd - mean(Dates)' upperf := Thick predictf. . 2 (I-

               +qtdI- -,!,Total men I           (year predict, Thick actuialmeai)

2) S1itGmnd er. J +TTI - u Iowerf: Thick predictr c

             +.-qt 1 - -4, Total means- 2 .StGrand2).S~ran   +

rri.11-1 +ye

                                                                        -I         r pedic(         asuam   e)1 OCLR00000957

Amer~en Calculation Sheet Appendix 9

Subject:

Cale. No. Rev. No. System No. Sheet No. Drywall Corrosion C-1301-187-e310-037 0 187 A9 -22 of 25 The minimum required thickness at this elevation is Tmin gen S6 :=452 (Ref. Calc. SE-000243-002) Location Curve Fit Projected to Plant End Of Life I ~ - I - 650 Thick predict upper 6W0 A lower 550 m s = -0.746 L measured Tm*mj-gn 86 1 Sn') 450 I I I I. 1980 1990 2000 2010 2020 2030 Yea predicY~e" red tY predict, Dates~year prdc Therefore the regression model shows that even at the lower 95% confidence band this location will not corrode to below Drywell Vessel Minimum required thickness by the plant end of life. Year predict1 2 = 2.009-103 Thick predict1 2 = 620.515 OCLROO000958

AmerGen Calculation Sheet Appendix 9

Subject:

Calc. No. Rev. No. System No. Sheet No. Drywell Corrosion C-I301-187-e310-037 0 187 A9 -23 of 25 The Following trends are shown for the pits Local Tmin for this elevation in the Drywell Tminlocal 86r :=300 (Ref. Calc. SE-000243-002) B1 : Max(Pit 34) ,nin(Pit 34) R(Pitý35) I I I I I

                                &W01-                                         *

• Xa~4~)~~4 ~bUU)4iUUI~UuH3*UUUR4:UWW~UuiUX

                                                                                     ~t 4 .*

Pit 3 4 sqoo1-Pit 3 .5 Tmunjocal g6 4001-Li 300-I I I I *I 1988 1990 1992 1994 1996 1998 2000 Dates L7 OCLROO000959

AmerGen Calculation Sheet Appendix 9

Subject:

Calc. No. Rev. No. System No. Sheet No. Drywell Corrosion C-1301-187-e310-037 0 187 AS -24 of 26 0 The following addresses possible corrosion in these pits The F-Ratio is calculated for the worse pit last(Dates) SSEpit .= E O (Pit 35 - yhat(Dates, Pit 5. 2 SSE pit = 6.931-1IO3 last(Dates) SSR pit: 0 (yhat(Dates Pit 35)i- mcan(Pit 35) )2 SSR pit = 1.941u10 SSE pit SSR pit MSE pit DegreeFreess StPit err := -'*it MSR pit : DgreeFre. reg MSR pit Fpit actauf

• E pi Fpit ratio 0.564 Fcrtiical Therefore this pit Is not experiencing corrosion m pit := slope(Dates, Pit 35) m pit = -3.526 Y pit:= intercept(Dates, Pit 35) Y pit = 7.634-103 The 95% Confidence curves are calculated Pit curve := m pit.Year predict 1-Y pit Pit actualmean := mean(Dates) sum  :=" (Datesd - mean(Dates)) 2 uppitr:= Pit curvef -

(Year predictr- Pit actualmcan) 2

                + qt   1 -- ,Totalmeans 2)-StPiterr          I '*-

sum qt( Iopitf := Pit curve

                           -a  t]oa en Tt aa tier 1 OCLROO000960

AmerGen Calculation Sheet Appendix 9

Subject:

Calc. No. Rev. No. System No. Sheet No. Drywell Corrosion C-1301-187-e310-037 0 187 A9 -25 of 25 Curve Fit For Pit 35 Projected to Plant End Of Life Pit 3 5 m pit = -3.526 xX t curve uppit lopit Train local 86 1990 2000 2010 2020 2030 Dates,year predict.year piedict.Year predictYear predict Therefore based on regression model the above curve shows that this pit will not corrode to below minimum required thickness by the plant end of life. OCLROO000961

Citizen's Exhibit NC7 Citizen's Exhibit NC7 EDNuclear Calculation Sheet PROBLEM STATEMENT The statistical analyses of drywell thickness data through November 1991 are documented in Reference 3.1. The analyses show that there was statistically significant corrosion in Bays 9D, 11A, 11C, 13A, 17A, 17D, 17/19 Frame Cutout, 19A, 19B, and 19C in the sand bed region and in Bay 5 Area D-1 2 at elevation 50'-2". The corrosion at elevation 51"-10" was not statistically significant, and no measurements were taken at elevation 87'-5" due to high temperatures. The regression analyses in Reference 3.1 provide the best estimates of the linear regression line which defines the drywell thickness in these bays as a function of time. The purpose of this calculation is to use these linear regression equations to predict the following: (1) The best estimate of the date on which the mean thickness at a monitored location will equal the minimum allowable. (2) The latest date for which we have 95% confidence that the mean thickness will not be less than the minimum allowable. (3) Using the earliest date from (2), compute the lower 95%/95% Tolerance Limit for the local minimum thickness on that date. These analyses are performed at all locations which have been determined to have a statistically significant corrosion rate. For the sand bed region, the analyses are performed for the minimum allowable thickness without sand (736 mils). pie I I OCLROO000213

Calc. No. C-1302-187-5300-020 Rev. 0 2 ofS .SPage T 2.1 Sand Bed Region (1) The earliest best estimate date to reach 736 mils is May 1995 for Bay 19A. (2) The earliest 95% confidence level date to reach 736 mils is June 1994 for Bay 19A. (3) It is predicted at the 95%/95% tolerance level that the minimum local thickness in Bay 19C will not be less than 574.1 mils in June 1994. This is the most limiting monitored location in the sand bed. 2.2 Elevation 50'-2" The only location with statistically significant corrosion is Bay5, Area D-1 2. (1) The best estimate to reach 670 mils is May 2019. (2) The 95% confidence level date to reach 670 mils is June 2006. (3) It is predicted at the 95%/95% tolerance level that the pit at point 9 in Bay 5, Area D-1 2 will have a thickness of not less than 615 mils in June 2006. 2.4 Elevation 51"-10" The only monitored location at this elevation does not have statistically significant corrosion. This means that the slope of the regression line is not statistically significant and that projected thicknesses determined by using inverse regression are meaningless. 2.5 Elevation 86' No measurements were taken at this elevation in November 1991 due to high temperatures. OCLROO000214

Calc. No. C-1302-187-5300-020 Rev. 0 Page 3of A

3. REFERENCES 3.1 GPUN Calculation C-1302-187-5300-019, Rev. 0, "Statistical Analysis of Drywell Thickness Thru November 1991."

3.2 "Applied Regression Analysis", 2nd Edition, N.R. Draper & H. Smith, John Wiley & Sons, 1981. 3.3 "Experimental Statistics", (NBS Handbook 91), Mary Gibbons Natrella, John Wiley & Sons, October 1966. I, 'Oi I 0OCLR0000021 5

Calc. No. C-1302-187-5300-0 20 Rev. 0 Page 4 of 4t 19 4. ASSUMPTIONS & BASIC DATA 4.1 Mean Thickness vs Time The mean thickness for each set of measurements at each location is documented in Ref. 3.1. These values are used as input to this calculation. 4.2 Earliest Time To Reach Minimum Allowable Thickness We need to determine the Best Estimate Date and the date at which we have 95% confidence that the mean thickness at a monitored location will not be less than the minimum allowable thickness. To accomplish this, we must do the following: (1) Perform linear regression analysis for each monitored location with statistically significant corrosion, using the data described in 4.1 as input. (2) Project the regression Line and the two-sided 90% confidence interval about the mean forward in time to locate the intersections with the minimum allowable 1/4 thickness. This is depicted in the figure below. 1~

                '4 min;nimum CV/6/OAJ3W4 X1     .        K,              - K    'ý7W rWVI.JJ law*h rqv     sinwims X bY2tokna  mn r*vau, WaNdO 66MMISa V.

OCLROO000216

Calc. No. C-1302-187-5300-020 Rev. 0 Page 5 of4 In this problem, y is the mean thickness in mils and x is time in years. The two-sided confidence interval is calculated at a confidence level of 90%. This means that we have 90% confidence that the true mean lies within this interval. Since the confidence interval is centered about the regression line, it also means that we have 95% confidence that the true mean lies above the lower band (90% within the interval between bands plus 5% above the upper band = 95%).

 *0 represents our best estimate of the time when the true mean thickness will equal 736 mils.

xL represents the time at which we have 95% confidence that the true mean is not less than y0. yo = 736 mils. The equations for the regression line and confidence interval are generally expressed explicitly in terms of y and implicitly in terms of x. If used in this form, the calculation of x2 and x. would be a trial and error process. 0 OCLROO000217

Calc. No. C-1302-187-5300-020 Rev. 0 Page 6 of //f Paragraph 1.7 in Ref. 3.2 rearranges the equations to perform inverse regression and thus solve directly for xL and x.. b0 = (Ye - b0 )/b,

     = y Intercept yo = Slope
     = 736 mils XL                                      (l-g) x)   =      + (ko  -  x           1 ) [[(k0 -3
                                +/- (ts/b            2 /s=] + (1-g)/n]'/2 I

g= t 2 / [b2 / (S 2 /Sx)'/ 2 12 U t=t(v,1-o:/2) U v = (n-2) = degrees of freedom of s2 (1-a/2) (1-0.10/2) - 0.95 S =2 n = number of observations (yt,x1 ) 3x = Ex.,x1 n A SPEAKEZ program (OLINPREDX") was written to solve the equations for xL, ke and x,. This program calls another SPEAKEZ program ("LINREGNI") which performs a linear regression analysis to calculate the values to input to these equations. Listings of these programs are included in Appendix 6.3. U OCLROO000218

Calc. No. C-1302-187-5300-020 Rev. 0 Page 7 of ' 4.3 Mean Thickness In All Bays At Minimum Time xL Having calculated the minimum time xL to reach t*, the next step is to predict the true mean thickness in all bays on this date. A SPEAKEZ program ("(LINPREDY") was written to use the input from 4.1 and 4.2 to predict the true mean thickness on the above date and the 90% confidence interval about this value. A listing of this program is included in Appendix 6.3. 4.4 Local Minimum Thickness The local minimum thickness at 95%/95% tolerance level is estimated as follows: (1) The predicted mean thickness and 90% confidence interval at minimum time xL is calculated for each monitored location in 4.3 with a statistically significant corrosion rate. This represents the mean thickness within the 6"x6" grid. (2) The thickness at individual points witoin the grid was determined to be normally distributed in Ref. 3.1. The standard deviation of the individual point thicknesses was also calculated in Ref. 3.1. If the standard deviation at a given location does not. vary with time, its mean value can be used as the best estimate of the standard deviation at a future date. However, if it varies with time, regression analysis' may be used to predict the standard deviation at a future date. (3) The one-sided 95%/95% lower tolerance* limit for local minimum thickness is equal to the one-sided 95% lower confidence limit for mean thickness (from 4.3) minus 2.092 times the standard deviation (from 4.4(2)), where 2.092 is the factor for 95%/95% one-sided tolerance limit for a sample size of 45 (Ref. 3.6, Table A-7). Most of the grids have about 45 valid data points. OCLROO000219

2 Caic. No. C-1302-187-5300-0 0 Rev. 0 Page 8 of I MJWAiuA' N\

  ~   Ukt ~,ga~D                        #c40,9"S4 L.1mr.*JOI4 5%GICI&VAM      z Adi u
                                                      ?,#,4 .cAwA:&

dimar~~

   ~~44 A4IAUIWUI4
   ~    ,cx~ve* r qS ?./7s7~

761-8 4 e4qpJC6

    '.4-n/f                IS$ AMMA~

CA@ 4~Af bFqrA,* XL' ir OCLROO000220

Calc. No. C-1302-187-5300-020 Rev. 0 Page 9 of 41

5. CALCULATIONS 5.1 Mean Thickness vs Time The mean thickness for each set of measurements at each location from Ref. 3.1 are tabulated as the variable y on the computer output sheets in Appendix 6.1.

5.2 Sand Bed Without Sand Minimum allowable thickness = 736 mils. Bay & Area Best Estimate 95% Date Confidence Date 9D 11/02/2014 7/08/2005 11A 6/01/1997 3/22/1996 11C Top 3 11/16/2000 7/07/1997 11C Bot 4 1/07/1998 4/06/1996 13A 9/22/1997 6/11/1995 17A BOt 4 4/10/2012 4/24/2004 17D 8/08/1995 9/14/1994 17/19 Bot 4 1/02/2008 8/23/2002 17/19 Top 3 3/16/2011 8/2812003 19A 5/16/1995 6/16/1994 19B 7/15/2000 8/17/1997 19C 5/16/1996 1/27/1995 r

$

ki aw OCLR00000221

Calc. No. C-1302-187-5300-020 Rev. 0 Page 10 of 4 Minimum True Mean Thickness in June 1994 Bay & Area Minimum True Mean Thickness @ 95% Confidence on 6/10/1994 9D 936.0 11A 773.7 11C Top 3 851.3 11C Bot 4 776.7 13A 736.7 17A Bot 4 883.8 17D 742.0 17/19 Bot 4 909.3 17/19 Top 3 903.6 19A 736.0 19B 784.4 19C 749.1 U I* OCLROO000222

Calc. No. C-1302-187-5300-020 Rev. 0 Page 11 of 0 Local Minimum Thickness in June 1994 The variation of the standard deviation of individual point measurements about the mean thickness as a function of time was analyzed using the SAS Procedure "PROC REG". The results of the regression are tabulated below, where: N Number of datasets Mean STD Mean Standard Deviation for the N datasets STD Last Standard Deviation on Date of Last Reading (from the regression line B1 Slope of regression line Prob>F Probability that B1=0 Years Years from date of last measurement to 6/10/94 Bay& N Mean Std Std Last B1 Prob>F Area III 9D 6 71.1 78.1 6.4 0.01 11A 13 46.2 51.0 1.9 0.11 11C Top3 12 106.7 102.1 -2.0 0.23 11C Bot4 11 26.4 23.1 -1.6 0.34 13A 9 57.6 57.4 -0.1 0.96 17A Bot4 8 53.4 56.7 2.2 0.19 17D 13 60.7 66.0 2.1 0.02 17/19 8 33.1 33.0 -0.1 0.98 Bot4 17/19 8 23.3 20.4 -2.0 0.54 Top3 .... _ 19A 13 59.2 62.9 1.5 0.05 19B 12 59.4 63.0 1.5 0.07 19C 12 76.7 80.0 1.4 0.43 OCLROO000223

Calc. No. C-1302-187-5300-020 Rev. 0 Page 12 of 11Y 0 For those bays where BE is negative, the standard deviation is assumed to be constant and equal to the Mean Std. For those bays where Bl is positive, the standard deviation on 6/10/94 is computed from the regression. Std on 6/10/94 = Std Last + Bl

• Years Years = 2.6 Bay & Area STD on K*STD 95% Lower 95%/95%

6/10/94 Confidence Lower Limit for Tolerance Projected Limit for Mean Local Minimum Thickness 9D 94.7 198.2 936.0 737.8 11A 55.9 117.0 773.7 656.7 11C Top3 106.7 223.2 851.3 628.1 11C Bot4 26.4 55.2 776.7 721.5 13A 57.6 120.5 736.7 616.2 17A Bot4 62.4 130.6 883.8 753.2 17D 71.5 149.5 742.0 592.5 17/19 Bot4 33.1 69.2 909.3 840.1 17/19 Top3 23.3 48.7 903.6 854.9 19A 66.8 139.8 736.0 596.2 19B 66.9 140.0 784.4 644.4 19C 83.6 175.0 749.1 574.1 OCLR00000224

I I* Calc. No. C-1302-187-5300-020 Rev. 0 Page 13 of I O 5.3 Elevation 50"-2" The only location at this elevation with It statistically is predicted significant corrosion is Bay 5, Area D-1 2. thickness at at the 95% confidence level that the true mean decrease below 670 mils prior to this location will not limiting. 6/30/2006. Thus, Bay 19A in the sand bed is more Local Minimum Thickness in June 2006 The standard deviation has of individual point measurements about the mean thickness a mean value of +/- 12.0 mils. Regression analysis yielded the following: at last measurement +/- 13.9 1STD Slope BE + 1.0 Prob>F 0.11 Years from 11/02/91 to 6/30/2006 = 14.7 STD 0 6/30/2006 = 13.9 + 14.7(1.0) = 28.6 0 t,5L = 95% Lower Confidence Limit for Projected Mean in June 2006 Thickness tq5L = 670 mils ttmi95L = t95L - K(STD)

             .95/95= 2.126 @ n=40 (Ref. 3.3,  Table A-7) tmin95 = 670 -       2.126(28.6)
                     =  609.3 mils I.

I OCLROO000225

Calc. No. C-1302-187-5300-020 Rev. 0 Page 14 of #1S Based on the regression, the best estimate of the current thickness of the grid is 742.7 +/- 2.4 mils. (Ref. 3.1) Based on the regression, the best estimate of the current thickness of the pit at point 9 in the grid is 688.1 +/- 2.6 mils. (Ref. 3.1) Thus, the depth of the pit is (742.7 +/- 2.4] - (688.1 +/- 2.6] = 54.6 +/- 3.5 mils Where the two uncertainties are propagated using the square root of the sum of the squares. Based on the conclusion that the pit is corroding at approximately the same rate as the grid (Ref. 3.1, para. 5.2.1), the thickness at the pit in February 2003 at the 95% confidence level will be: 670 - 54.6 = 615.4 mils. 5.4 Elevation 511-100 The only monitored location at this elevation does not have statistically significant corrosion. This means that the slope of the regression line is not statistically significant and that projected thicknesses determined by using inverse regression are meaningless. 5.5 Elevation 86' No measurements were taken at this elevation in November 1991 due to high temperatures which prevented access. OCLROO000226

PROGRAM: LIIPREDX RAY NUMBER? 9d E NM " DATE LIST data9d Caic. No. C-1302-187-5300-020 NAME OF MEAN TIEWS DATLSEVT =dgd Rev. 0 Page j6of 1

• DAT9D I1D9D 1 12/4/86 1071.51 2 12/19/88 1021.39 3 6/26/89 1054.42 4 9/13/89 1020.43 5 2/8/90 1009.96 6 4/24/90 1002.12 7 2/23/91 992.55 6 5/23/91 1002.53 9 11/2/91 992.4 EWTER NO. OF DESIRED DATLAinzts(4,9)

Er VALUE T OF MINIMUM TJICXXESS (10) 736 D=ES x y SD 9/13/89 1020.43 2/8/90 .405479 1009.96 4/24/90 .610959 1002.12 2/23/91 1.44658 992.55 5/23/91 1.69041 1002.53 11/2/91 2.13699 992.4 yo XL10 - 736 XU90 = Nov 22, 2055 XOHAT - Nov 2, 2014 XLS0 - Jul 8, 2005 T95 - 2.13182 0 - .39709 INVERSE ESTIMATION 18 NOT OF ,UCZ PRACTrICAL ILUE UNLESS WM REGRESSION IS WL DETRMIUNED, THAT 1S, TME SLC*E (31) IS SIGMICaM , WHICH IMPLIES TwA! 0 S LD BE SMAIJER THAN ABOUT 0.20. WHEN 0 IS MUCH LARGER TTAN WS, TEI RESULTS AIR M*AINGLSS. December 18, 1991 9:21 AM 0CLR00000227

PROGRAM: LINPREDX Calc. No. C-1302-187-5300-020 BAY NUMBER? 11a Rev. 0 ENTER NAME OF DATE LIST date1la Page. 0-o NAME OF MEAN THICKNESS DATASET? mella N DATEllA MEllA 0 ** ******** ****** al112/2/86 891.47 2 4/29/87 918.66 3 5/1/87 904.64 4 8/1/87 922.09 5 9/10/87 905.2 6 7/12/88 912.97 7 10/08/88 888.22 8 6/26/89 881.34 9 9/13789 891.56 10 2/8/90 875.34 11 4/24/90 858.04 12 2/23/91 844.58 13 5/23/91 843.98 14 1 /2/91 832.6 ENTER NO. OF DESIRED DATA intsC2,14) ENTER VALUE OF MINIMUM THICKNESS (YO) 736 BAY DATES X Y

      .11A    4/29/87     0           918.66 5/1/87        .00547945 904.64 8/1/87        .257534   922.09 9Z10/87       .367123   905.2 7/12/88     1.20548     912.97 10//88      1.44658     888.22 6/26/89     2.16164     881.34 9/13/89     2.37808     891.56 2/8/90      2.78356     875.34 4/24/90      2.98904     858.04 2/23/91     3.82466     844.58 5/23/91     4.06849     843.98 1 /2/91     4.51507     832.6 YO      = 736 XU90 = Jan 27,       1999 XOHAT = May 29,      1997 XL90 = Mar 19        1996 T95         1.7958t G       = .0285006 INVERSE ESTIMATION IS NOT OF MUCH PRACTICAL VALUE UNLESS THE REGRESSION IS WELL DETERMINED THAT IS, THE SLOPE (B1) IS SIGNIFICANT RHICH IMPLIES THAT G SHOULD BE SMALLER THAN ABOUT 0.20. WHEN G IS MUCH LARGER THAN THIS, THE RESULTS ARE MEANINGLESS.

Arf1 15, 1992

53 '

I-- 0CLR00000228

pIrXwAZ4 LINP22D NAX 9MMW lic Zop 3 Calc. No. C-1302-187-5300-020 IR NAMU OF DATE LIST datello Rev. 0 NAME*F 1MN MIC1IWSS DASM mdlict Page I7 of N DATMITC Y*D11CT 1 5/1/87 1046 2 6/1/87 1108.6 3 9/10/87 1079.12 4 7/12/88 1045.38 5 10/06/88 1008.86 6 5/26/89 1015.78 7 9/13/89 1005 S 2/8/90 978.45 9 4/24/90 974.9 10 3/4/91 981.68 11 5/23/91 1001.79 12 11/2/91 964.3 ZN= NO. OFVRS1=5 DATA Ints(1,12) UM MVATU*IF MNIMM SWIMES (YO) 736 DAY DAME3 x y 110 TOP 5/1/87 0 1046 6/1/87 .252055 1108.6 9/10/87 .361644 1079.12 7/12/88 1.2 1045.38 10/08/88 1.4411 1008.86 6/26/89 2.15616 1015.78 9/13/89 2.3726 1005 2/8/90 2.77808 978.45 4/24/90 2.98356 974.9 3/4/91 3.84384 981.68 5/23/91 4.06301 1001.79 11/2/91 4.50959 964.3 Y0 - 736 Xu9O - Nov 13, 2006 XOHAT- Nov 16, 2000 XLSO w Dec 7, 1997 T95 - 1.81245 C - .Z18019 V HAT S3w z IS 1o0 oF muCH PHACI.CAT. VA.,0 UNLESS TME R3CauSSic Is WEIL DWM*==vD, TM iS, ME* 81021 (E1) zs ZINICAM, WHIXCDMPLISS 5TM aSN ,D IBE W8M MN ABC= 0.20. WMY a is MUC

    ,TRHAM            TNIS, T12S121TS1A.1 1          IMNINLE-S.

December 18, 1991 9:24 AM I I Im* OCLR00000229

P=RGAM, LINPREDX RAY ,UNmMR? 11c bot 4 Calc. No. C-1302-187-5300-020 ESTER SAMEOF DATE LIST datellc Rev. 0 itAME OF MEAN TaICENESS DTSEVT? mdllcb Page 1rof4Y N* DATEiIC 9 16.79 5/1/87 2 953.64 8/1/87 3 9/10/87 915.71 4 7/12/88 906.05 5 20/08/18 897 6 6/26/89, 876.75 7 9/13/89 890.71 8 210/90 869 g 4/24/90 863.29 10 3/4/91 857.54 11 5/23/91 862.64 12 11/2/91 856.3 EIE lNO. OF DESIZXD DATA ints(,12) ESTERVALUE OF MINIMMM '1'3ZC1WES (TO) 736 SAY x y lic NOT 5/1/87 0 916.79 8/1/87 .252055 953.04 9/10/87 .361644 915.71 7/12/88 1.2 906.05 10/08/88 1.4411 897 6/26/89 2.15615 876.75 9/13/89 2.3726 890.71 2/8/90 2.77808 869 4/24/90 2.98356 863.29 3/4/91 3.84384 857.54 5/23/91 4.06301 862.64 11/2/91 4.50959 856.3 TO - 736 1,190 - Dec 10, 2000 XOEAT - Jan 7, 1998 3M90 - Apr 6, 1996 T95 - 1.81246 0 - .0641842 IZVERSE ZSTI.ATION 18 NOT OF WCH PRACTZCAL VALUE UNLESS T RE GRSSION IS WEZL =ERMINED, TEAT IS, TIM SLOPE (BI) IS SIG"*EIAIW, WEMC IMPLIES TEZT G S=MULD BE SYMALX TrWEA*OUT 0.20. WMENC 19 MUCH LARGER MN THIS, MM RESULTS AM KM*Z1=,LSS. Decembet 18, 1991 9:25 Am I. OCLROO000230

PEMG2AM: .IVP3EDX EAY NMMER? 13a MERR NAME OF DATE .ISTdatel3a Calc. No. C-1302-187-5300-020 IE OF PMK *ICKINS6 DATASET? gdl3a Rev. 0 Page of f N DATE13A MD13A 1 11/15/86 919.08 2 12/17/88 905.27 3 6/26/89 882.84 4 9/13/89 882.98 5 2/8/90 859.04 6 3/28190 852.84 7 4/24/90 855.63 8 2/23/91 854.51 9 5/23/91 852.88 10 11/2/91 848.6 IRTE NO. O. DWSMD DATAIR(2,10) EMM VALUE OP* InM= TICEIES3 (TO) 736 DAt x r 13A 12/17/88 0 905.27 6/26/89 .523288 882.84 9/13/89 .739726 882.98 2/8/90 1.14521 859.04 3/28/90 1.27671 852.84 4/24/90 1.35068 855.53 2/23/91 2.1863 854.51 5/23/91 2.43014 852.88 11/2/91 2.87671 648.6 TO -736 XU9 0 Aug 3, 2003 XOEAT= Sep 22, 1997 AL90 - Jan 11, 1995 T95 - 1.89458 G - .195334 ZNVMRSE ESTIMATZI IS NOT OF )MM PRACTICAL VALUE UNLSS Tim XEGPESSIO zS WELL DzTERMINzD, TET IS, THEELO (81) IS 8ZGNxZICA1IT, PWICR nMPIES 2w a SHOULD BE &MALLE TW~l AB=U 0.20. W=R 9 18 K=C ZAR=E MHN ES, =E SULTS AME NZANIGLESS. December 18, 1991 9:2801[ OCLR00000231

PiMM0AM: LINPREDX Shy NUHGM 17a bot 4 I Calc. No. C-1302-187-53 0 0- 020 XM* NAME OF DATE LIST 4atel7a Rev. 0 I=AE OF LMETSICKNES DATA&S Id17ab Page Aoof DiTE17A MD17AB 12/17/88 957.36 2 6126/89 964.5 3 9/13/89 955.18 4 2/8/90 937.5 5 4/24/90 939.61 6 2/23/91 934.71 7 5/23/91 942.39 a 11/2191 932.8 NEwo. OF DsE D cm ints(1,8) MM VAL= CO MN13M1M M*ZOESS (YO) 736 NAY x 17A SM 12/17/88 0 957.36 6/26/89 .523288 964.5 9/13/89 .739726 955.18 2/8/90 1.14521 937.5 4/24/90 1.35068 939.61 2/23/91 2.1863 934.71 5/23/91 2.43014 942.39 11/2/91 2.87671 932.8 70 - 736

1. XU90 xOHAT -

XL90

             - Apr 15, Apr 2041 10, 2012
             - Apr 24, 2004 T95      -   1.94318 O        - .324478 INVERSE ESTDM4ATON IS NOT OP MUCM MRACTICAL VAi=U UESS mTE         EcREScsO       iS WEL  DERNE,       TMAT Is.

TME SLOPE (51) IS SIGMICANT, W*ICH IMIS 7= r SBEULD BE sLLR 2M AZWT C, 0.20. WEN 9 MICH LARGER THANS. m = m NUT A= EJE InIULss. DeCembO: 18, 1991 9:29 *M ml, r 0CLR00000232 6 OCLROO000232

PIDGPMt LINPIM2DX BAY NMMM 174 EUM NA O D= T.LIST date17d Calc. No. C-1302-187-5300-020 NAMEOF IMN MUXCRMEB DATASEMf mel7d Rev. 0 Page j of qf V DATE17D Pt1tD 1 12/2/86 903.51 2 2/17/87 922.16 3 5/1/87 895.07 4 8/1/87 890.69 5 9/10/87 895.28 6 7/12/88 877.93 7 10/08/88 862.22 8 6/26/89 856.84 9 9/13/89 847.13 10 2/8190 833.37 11 4/24/90 826.8 12 2/23/91 825.27 13 5/23/91 829.08 14 11/2/91 822.2 lN2fl NO. OF DESIRED DATA Inta(2,14) IF= VALd=1OPF .NIME'M THICMESS (YO) 736 S7D DAME x y 170 2/17/87 0 922.16 5/1/87 .2 895.07 8/1/87 .452053 890.69 9/10/87 .561544 895.28 7/12/88 1.4 877.93 10/08/88 1.5411 882.22 6/25/89 2.35616 856.84 9/13/89 2.5726 847.13 2/8/90 2.97808 833.37 4/24/90 3.18356 826.8 2/23/91 4.01918 825.27 5/23/91 4.26301 829.08 11/2/91 4.70959 822.2 Alw yo - 736

          - Oct 30, 1996
          -    aug    8,  i995 XL90 Sep 14, S          1994 T95    -    1.79588 a      -    .0255943 1XINVR      ESTIMATION      is NOT OF MUCHPPACICXI. VALUE ONLES THE NEREM SION IS iVEM. DETEMISED, MM                  IS, WE 8=2         (31) I     SZXCNXZXCANT,   WHICH I*NLIES    -A 0 SHOULD BE 8AMR TOMAABOUT 0.20.                 WHEN 6 1, MUCH LARGE     2tHA   THIS, THE RESULTS AR          ME     EL .

December 18, 1991 9:30 AM 0CLR00000233

PROI~DM: LISPB3D BAY 3WMEER 1719b SATE KME (W DATE LIST dat&27lg Calc. No. C-1302-187-5300-020 NAME OF DO=E TZIcME~S DATAMMW ud17l9b Rev. 0 Page JpJof A DATE1719 MD1719B 1 12/30/88 1003.79 2 6/26/89 1019.07 3 9/13/69 1016.57 4 2/8/90 1000.46 5 4/24/90 988.54 6 2/23/91 987 7 5/23/91 982.39 a 11/2/91 971.1 ENER N0. 0F ESSIED DAMA:Lftn(2,8) EERVALUE Op IUNII= TBICJERsS (YO) 736 17BAY DATES x y 1719B 12/30/88 0 1003.79 6/26/89 .487671 1019.07 9/13/89 .70411 1016.57 2/8/90 1.10959 1000.45 4/24/90 1.31507 988.54 2/23/91 2.15068 987 5/23/91 2.39452 982.39 11/2/91 2.8411 971.1 YO -736 XUO - Aug 15, 2021 XCHAT - Jan 2, 2008 x.90 - Aug 23, 2002 T95 - 1.94318 a .189737 WYMME ESTIMATION IS X= OF HUC IACTICAL VALUE UNLESS 'WE =RZSSIC IS MMELDEOERMINED, MMT IS, THE BLOPE (31) is SIGMIFICAET, IMCH ZneLIES TEAT G SBOULD BE SMALLER MWN ADOMT 0.20. WHEN C IS MUCK LANR= TOMA THIS, TME MEEHILS AME HEANXNGZ=S. Deceber 18, 1991 9:31 AM 0CLR00000234

PROGRAH: LIUPREDX BAT NW RM 1719t

          'iR NAM OF DATE LM           datel719                     Calc. No. C-1302-187-5300-020 NAMECF MEAN THICME=S DATA.T? mdl7l9t                          Rev. 0 Page 73of #9
                         *MD171S".

1 12/30/88 981.71 2 6/26/89 998.81 3 9/13/89 992.4 4 2/8/90 986.29 s 4/24/90 970.5 6 2/23/91 974.67 7 5/23/91 969.3 8 11/2/91 954.2 E1TER NO. Or DESIRED DAMMintR(1,8) ZWM VALE OF KNINUM !IICKRESS (TO) 736 DATES I 1719T 12/30/88 0 981.71 6/26/09 .487671 998.81 9/13/89 .70411 992.4 2/8/90 1.10959 986.29 4/24/90 1.31507 970.6 2/23/91 2.15068 974.67 5/23/91 2.39452 969.3 11/2/91 2.8411 954.2 TO -736 ,~0UO - J=m 25, 2038 XOHAT - Mar 16, 2011 XL90 - Aug 28, 2003 T95 " 1.94318 G - .321285 INVERSE ISTD(8TION IS NOT CP MUCH PR*CICAL VALUE UNLESS ' E RESSIOM 1S WELL 'ENNITNED, THAT 19, ESLMOVE (31) 13 SZGNIICANT, WHICH IXMIES M Q SHOULD BE SMAf1LUR THAN As=O 0.20. VJEEN a is MUCH LARGER MR THIS, T*M RESULTS SAR MEANINGLESS. December 18, 1991 9:32 AM 0CLR00000235

P1OGMMM: LINPREDX DAY RUJWIC? 19a ZRT2R NAME OF DATE LIST datel9a Calc. No. C-1302-187-5300-020 Ge1ga Rev. 0 NAME Or ME IHICUES DASET? Pagesý of i N DWET19A M19A 2. 12/2/86 870.22 2 2/17/87 883.64 3 511/87 872.93 4 8/1/87 658.6 5 9/10/87 858.29 6 7/12/88 848.57 7 10/08/88 836.91 8 6/26/89 828.82 9 9/13/89 825.36 10 2/8/90 807.78 11 4/24/90 807.8 12 2/23/91 816.67 13 5/23/91 802.8 14 11/2/91 803.2 uNTER O. oF ,sIzRED DATA. iuts(2,14) ETERM VALUE OF MNIMUM 95TI1EMS (Y0) 736 DAy DATES x T lSA 2/17/87 0 883.64 5/1/87 .2 872.93 8/1/87 .452055 858.6

         /110/87        .561644   858.29 7/12/08        1.4         848.57 10/08/88      1.6411      836.91 6/26/89       2.35616     828.62 9/13/89        2.5726      925.36 2/8/90        2.97808     807.78 4/24/90       3.18356     807.8 2/23/91       4.01918     816.67 5/23/91       4.26301     802.8 11/2/91       4.70959     803.2 TO      - 736 3,190   - Sep    6, 1996 XO1AT ftMay 16, 1995 30L90 - Jun 10, 1994 T95     - 1.79588 a       a .0302474 MNER SEITIWMATo          ZS NOT OF MUCH MECTZCAL VALUE mumESS M       REO.SION iS WELL D=EI=MINED, TEAT iS, THE    LWEN (MI) IS BIMGnMCA          , WHICH MOIieiS 2SAT 0S HOLD HE SMALLER TERN AOUT 0.20.             WHEN

• II MUCH LAVAER "ELS, THE RESULTS AIR NM NLESS.

December 18, 1991 9:33 AK OCLR00000236

PROGRAM: LINPREDX CaJc. No. C-1302-187-5300-020 BAY NUMBER? 19b Rev. 0 ENTER NAME OF DATE LIST datel9b Pageg* of g NAME OF MEAN THICKNES. DATASET? mdl9b N DATE19B ED19B 1 12/2/186 880.84 2 5/1, 87 897.63 3 8/1 787 892.21 4 9/-10/87 887.6 5 7712:788 863.98 6 101/08/88 856.41 7 6126i/89 852.55 8 92Z81 9/1378990 854.94 840.67 10 4/-24/,90 839.1 11 2/-23Z91 852.53 12 5J23191 13 1 /2/91 844.4 846.3 ENTER NO. OF DESIRED DATA ints(2,13) ENTER VALUE OF MINIMUM THICKNESS (Y) 736 BAY DATES X Y 19B 5/1/87 0 897.63 8Z1/87 .252055 892.21 9/10/87 .361644 887.6 7/12/88 1.2 863.98 10/08/88 1."411 856.41 6/26/89 2.15616 852.55 9Z13789 2.3726 854.94 2Z8/90 2.77808 840.67 4Z24/90 2.98356 839.1 2/23/91 3.81918 852.53 5f23Z91 4.06301 844.4 1 Y0

               /2/91 736 4.50959     846.3 XU90       Dec 12, 2005 XOHAT       May 7, 2000 XL90       Jul 26 1997 T95         1.81246 G           .11486 INVERSE ESTIMATION IS NOT OF MUCH PRACTICAL VALUE UNLESS THE REGRESSION IS WELL DETERMINED THAT IS, THE SLOPE (Bl) IS SIGNIFICANT WHICH IMPLIES THAT G SHOULD BE SMALLER THAN ABOT 0.20. WHEN G IS MUCH LARGER THAN THIS, THE RESULTS ARE MEANINGLESS.

Aril 3:155 A15, 1992 S 4(4 I - 0CLR00000237

PROGRAM, LINPREDX BAY NUMBER? 19C II *e . ENTIR NME OF DATE LIST datel9o Calc. No. C-1302-187-5300-020 KUME OF MEAN THICMWSS DATASET? 3=19c Rev. 0 PageA (of W% DATE19C HE19C

      **     * *at 1  12/2/86       866.02 2 5/1/87         900.51 3 8/1/87         888.16 4 9/10/87        888.31 5 7/12/88        873.46 6  10/08/88       856.27 7  6/26/89       845 8  9/13/89       844.7 9  2/8/9O        830.51 10  4/24/90       822.52 Ii  2/23/91       842.8 12  5/23/91       823.22 13  11/2/91       822.3 ENTER NO. OF DESIRE3D DAM ints(2,13)

ENTER VALUE OF MINIMUM THI ESS (YO) 736 DAY DATES X Y 19C 5/1/87 0 900.51 0/1/87 .252055 888.16 9/10/87 .361644 888.31 7/12/88 1.2 873.46 10/08/88 1.4411 656.27 6/26/89 2.15616 845 9/13/89 2.3726 844.7 2/8/9O 2.77808 630.51 4/24/90 2.98356 822.52 2/23/91 3.81918 842.8 5/23/91 4.06301 823.22 11/2/91 4.50959 822.3 YO - 736 XU90 x May 31, 1998 XOHT - May 16, 1996 XL90 - Jan 27, 1995 T95 - 1.81246 G - .0508207 DIVERSE ESTIMAT Is NOT or MUCH PRACTICAL VALUE. UENESS T REOEESSiON Is WELL DEMWR*IMD, TENT is, TM BLOPE (81) IS sIcnn*YPCJT, WHICH IM2LIES TT 0 SHOM) NE SMZALLER THAN ABOT 0.20. WHEN 9 IS M LANCER TOM THIS, THE I*s*STS ARE MEANINGLESS. December 18, 1991 9:36'AM OCLR00000238

PraPM. LINPPXDX BAY Ni*RB1? 5/51/dMU 1M NMM" 0* 1

                          =    ,8L82 date         512              Calc. No. C-1302-187-5300-020 NIAM OF IRAN !BICMUMEB          DATASET? mf5112                Rev. 0 Page ~of  qf DATES112     175112 1   11/021/87    755.05 2   7/12/88      751.07 3   10/08/88     751.51 4   6/26/89      750.9 5   9/13/89      757.27 6   2/8/90       740.83 7   3128/90      743.85 8   4/25/90      746.29 9  2/23/91       741.56 10   5/23/91       745.28 11   11/2/9%       748.1 ENTER NO.       OF DESIMED DM       Ints(),12)

ETRVALUX OF KaMfWI TOMCEESS (TO) 670 SAY DA*MS X I 5/51/512 0 755.05 11/01/87 7/12/88 .69589 751.07 10/08/88 .936986 751.51 6/25/89 1.85205 750.9 9/13/89 1.86849 757.27 2/8/90 2.27397 740.83 3/28/90 2.40548 743.85 4/25/90 2.48219 746.29 2/23/91 3.31507 741.56 5/23/91 3.5589 745.28 11/2/91 4.00548 748.1 To :70 2W9J0 BNov 6, 2114

'U XMIA       Mar 15, 2019 XZ90       J=
              ,T    30, 2006 T95        1.83311 a          .586272 XNMGRE ESTD=AT0lN is NOT OF MUM NPA=CALCX ViLIR UNLESS Ts      120E7415¢* IS wiz DNZMIMED, TWAT IS, MM MMOIE      (21) II    SIGNIPICMN,      WHlCH IMPLIES  OwA 6  BEOULD RE 8M*       ER TM ABOUT 0.20.           WR  9 18 MMC UORGER 2       ETHIS, TIE     RESULTS ARM)M0AMZXELS.

December 18, 1991 9037 Am I OCLR00000239

PRO0GRAM: LXNPPEfl MA NUIMM~? 5/51/412 ENTER NAME CE DA32 LIST dat5e112 Calc. No. C-1302-187-5300-020 NM CF MEAN TICI*mEss DATASET? mf5112 Rev. 0 Page A of 49 DATE5112 MF5112 2 1 11/01/87 755.05 2 7/12/88 751.07 3 10/08/88 751.51 4 6/26/89 750.9 5 9/13/89 757.27 6 2/8/90 740.83 7 3/28/90 743.85 B 4/25/90 746.29 9 4/23/91 741.56 10 5/23/91 745.28 11 11/2/91 748.1 ElEJA NO. OF DESIED DA,33Ln.ta(1,11) ENTER VALUE OFKMNMN THICIEMSS (YO) 673 DATES x 7 0 755.05 5/51/012 11/01/87 7/12/88 .69589 751.07 10/08/88 .935986 751.51 6/26/89 1.65205 750.9 9/13/89 1.88849 757.27 2/8/90 2.27397 740.83 3/28/90 2.40548 743.85 4/25/90 2.48219 746.29 2/23/91 3.31507 741.56 5/23/91 3.5589 745.28 11/2/91 4.00548 748.1 T0 =673 X=90 - Jam 24, 2110 XOBAT - Jan 30, 2018 X,90 NoVv 10, 2005 T95 - 1.83311 0 - .586272 ENVE=R ESTIATION IS NT OF MUC= 1ACAL VALUE USLUSS THE REGRESSION IS WELL DETERMINED, THAT IS, mm PE (E1) is s8*mC.NT, WHIcH m3LxZS T$a G 32 . 3M SMOLLER TMHN JA3OaT 0.20. WHEN 6 IS MUCH LINCER mm THIS, MME NE3ULTS an mmKE GLESs. Decembez 18, 1991 9:38 AM OCLROO000240

PROPJGRAM, LINPREDX In NUMER? 5/51/d12 3ERNAME OF DATE LIST date5112 Calc. No. C-1302-187-5300-020 NAME F) MAIIIC1RSSC DAWSE MZ5112 Rev. 0 Page of qf N DAME5112 MF5112

      **

U, 3 1 2 11/01/87 755.05 7/12/88 .751.07 10/06/08 751.51 4 6/26/89 750.9 5 9/13/89 757.27 6 2/8/90 740.83 7 3/28/90 743.85 8 4/25/90 746.29 9 2/23/91 741.56 10 5/23/91 745.28 u 11/2/91 748.1 Rx= NO. OF DESZIRE DATA Intar(1,11) RER VALUE OF IUNIMU TgICtESS (TO) 597 DAY x 5/51/D12 11/01/87 0 755.05 7/12/88 .69589 751.07 10/08/88 .936986 751.51 6/26/89 1.65205 750.9 9/13/89 1.86849 757.27 2/8/90 2.27397 740.83 3/28/90 2.40548 743.85 0 4/25/90 2/23/91 5/23/91 2.48219 3.31507 3.5589 746.29 741.56 745.28 11/2/91 4.00548 748.1 70 -597 1119O - Mar 22, 2231 XOar - Jan 22, 2046 XL90 - Dec 12, 2021 T95 - 1.83311 a - .586272 ZINVESE ESTIKMT1ON 13 1OT OF MUCH PAOPCAL VALUE tILESS THE RECEESSIO2 18 WELL DE2EEMZqED, TEAT IS, TEE SLOPE (31) is SZCNIFmCTtM, WlCa LPLI" EMT 0 SEOUID BE SMALLER TEAMC U 0.20. W 9 IS M=UC luARGER TA Mrs, Tg RESULTS ARE MEA2!ZNLSS. Deember 18, 1991 9:39 AM OCLR00000241

PROM8M: LZ1NPREDX MA NMWM31? 5/51/412 INME NAMEOF DATE UIST dateS 112 Calc. No. C-1302-187-5300-020 NAME OF M~AN THICKMS8 DA2ASE? uf0112 Rev. 0 Page 5tof ,/* flT5112 N.F5112 11 1 11/01/87 755.05 2 7/12/88 751.07 3 10/08188 751.51 4 6/26189 750.9 7 9/13/89 757.27 6 2/s/90 740.83 7 3/28/90 743.85 6 4/25/90 745.29 9 2/23/91 741.56 10 5/23/91 745.28 11 11/2/91 748.1 lITE wo. OF DzSnED D=T intou(i,1) iNTr VALuE OF jUNIMIM ISICIEsS (TO) 540 DA7,S x F 5/51/D* 2Ia 5/51/812 11/01/87 755.05 7/12/88 .69589 751.07 10/08/88 .936986 751.51 6/26/89 1.65205 750.9 9/13/89 1.86849 757.27 2/8/90 2.27397 740.83 3/28/90 2.40548 743.85 4/25/90 2.48219 746.29 2/23/91 3.31307 741.56 5/23/91 3.5589 745.28 11/2/91 4.00548 748.1 Y0 -540 XU90 - Feb 2, 2322 X0WAT- Oct 7, 2067 XL90 = Jan 3, 2034 T95 = 1.83311 C - .585272 IMVESE ESTIMATYIW is NOTOF HMCEI PRACICAL ' VTA.UE WILS IRE REGRESSIC? rZ WELL DETER~nED, THAT IS, TMi SLOPE (31) lB BIG~nIPII1 VEYCI umLIs T=A a SBOULD BE SMALLER THUM ABC=J 0.20. WNEEG Z9 MUCH LLUME TA TEES ESULTE SAM MEMNONLESS. December 18, 1991 9:40 AM OCLROO000242

PROGRlAM Lfl!PJEDY SAY SUMMER? 9d Calc. No. C-1302-187-5300-020 ENTER NAMEOF DATE U5ST date9d Rev. 0 NIMM OF HEM!NTEXIOOSS DATASZT? uid9d Page 1 of qf

                                                                                                         /

N DATESO MD9D 1 12/4/86 1071.51

                                                                                                             /

2 12/19/88 1021.39 3 6/26/89 1054.42 4 9/13/89, 1020.43 5 2/8/90 1009.96 6 7 8 4/24/90 2/23/91 1002.12 992.55 1002.53 i 5/23/91 9 2.11291 992.4 ENTER NO. OF DESXINED D=T jnta(4,9) 11=1 DA32 OF HINIMU TSICENESS (X0) 6/10/94 I, EAT DATES X 7 t 9D 9/13/89 0 1020.43 2/8/90 4/24/90

                      .405479
                      .610959 1009.96 1002.12                                                           /

2/23/91 1.44658 992.55 5/23/91 1.69041 1002.53 21/2/91 2.13699 992.4 I X0 - 6/10/94 IL,90 YOM - U90 T95 936.012 962.366 948.72 2.23182 P-RATIO - 2.51832 W=AXS R F-RATIO OF 1.0 OR =A7E WROVEDES C00ID CE Zn THE =ME AND UNERCEPT OF THE RIS=ICAL DM, W 7-RATIO SBOULD BE 4 OR 5 IF TEE REGESSICX 2QATIC! 18 TO BE E TO PREDICT FUTWE VALUES. TO haVE A BIXG DEGREE CF ccFImix IN THE lREDICED VALUE, Mix RATIO E w BE AT LEAST 8 OR29. December 19, 1991 I I 4:26 PM S f I J

                                                                                                           /

I OCLROO000243

PROGRAM: LINPREDY Calc. No. C-1302-187-5300-020 BAY NUMBER? 11a Rev. 0 ENTER NAME OF DATE LIST datella PageIj.of q NAME OF MEAN THICKNESS DATASET? mella N DATE11A ME11A 1 1212/86 891.47 2 4/29/87 918.66 3 5/1/87 904.64 4 8Z1/87 922.09 59Z10187 905.2 6 7712/88 912.97 7 10/08/88 888.22 8 6/26/89 881.34 9/913/89 891.56 10 2/8/90 875.34 11 4/24/90 858.04 12 2/23/91 844.58 13 5(23/91 843.98 14 1832.6 ENTER NO. OF DESIRED DATA ints(2 14) ENTER DATE OF MINIMUM THICKNESS tX0) 6/10/94 BAY DATES X Y 11A 4/29/87 0 918.66 5/1/87 .00547945 904.64 8/1/87 .257534 922.09 9101/87 .367123 905.2 7712*88 1.20548 912.97 10/08/88 1.44658 888.22 6/26/89 2.16164 881.34 9/13789 2.37808 891.56 2Z8/90 2. 78356 875.34 4Z24-90 2. 98904 858.04 2/23/91 3.82466 844.58 5/23/91 4.06849 843.98 11/2/91 4.51507 832.6 r XO 6/10/94 YL90 773.58 YOHAT 789.971 YU9o 806. 363 T95 1.79588 rw~F-RATIO = 35.0869 WHEREAS AN F-RATIO OF 1.0 OR GREATER PROVIDES CONFIDENCE IN THE SLOPE AND INTERCEPT OF THE HISTORICAL DATA THE F-RATIO SHOULD BE 4 OR 5 IF THE REGRESSION EQUATI6N IS TO BE USED TO PREDICT FUTURE VALUES. TO HAVE A HIGH DEGREE OF CONFIDENCE IN THE PREDICTED VALUEs THE RATIO SHOULD BE AT LEAST 8 OR 9. 3:57 A15, 1992 Axril 4 4 i* OCLR0000244

PROGflM: LINPREDY BAY NIQE? Ile top 3 suM K= OF DATE LZS" datellc Calc. No. C-1302-187-5300-020 Rev. 0 SAME O MEAN TuMMzESS DMAE?. idI1ct Page 33of 4 N DA711C M)llCT at I 5/1/87 1048 2 8/1/87 1108.6 3 9/10/87 1079.12 t 4 5 7/12/88 10/08/88 1045.38 1008.86

                                                                                                      .ft I

6/28/89 1015.78 7 9/13/89 1005 8 2/8/90 978.45 4/24/90 974.9 10 3/4/91 981.68 11 5/23/91 1001.79 22 11/2/91 964.3 j uSRn NO. OF DSIRED DA71 ints(1,12) IT=ER DATE OF MIND=I. TBICESS (XV) 6/10/94 MAY x Y 110 TOP 5/1/87 0 1046 8/1/87 .252055 1108.5 9/10/87 .361644 1079.12 7/L2/88 1.2 1045.38 10/08/88 1.4411 1000.86 6/26/89 2.15616 1015.78 0 I 9/13/89 2.3726 1005 2/8/90 4/24/SO 2.77808 2.98355 978.45 974.9 I 3/4191 3.84304 981.68 5/23/91 4.06301 1001.79 11/2/91 4.50959 964.3 Xo - 6/10/94 nso anT Yugo g T5 i 8***a*e ***A*** at9*0*8 *.8**2t 851.317 895.162 939.008 1.81246 F-RATIO - 8.47318 UVMMSAN 1-XA=O CF 1.0 OR GREMR PROVIDES C=FFIDESCE 1 IN MEEV.021 AND INMRC E F T'tE EI5 OZC**r DAZA, TIE 1-RA=IO UChDN 4 OR S IF TeE NE;S8I0R EQuATIOR i3s1 I ISE UED TO MREDICT TUTUE VALUES. To RAVE A DSEGREE OW CCwFtuzzc N ME wnZnI'Crm VALUE, Tax RATIO acor BE AT LEAST 8 OR 9. December 19, 1991 I 4:27 PM t OCLR00000245

PROGAM. LXHPPRDY BAYiN12432? 11C bottozM 4 ENITER RAM5 C DATE LIST datalc Calc. No. C-1302-187-5300-020 MM~~ CF MIEANTSICX(NESS Drl!ASEr? udliob Rev. 0 Page 3dof ((1 DAME11C MD21CB 1 5/1/87 916.79 I 8/1/87 953.64 2 9/lO/87 3 9/10/87 915.71 4 7/12/88 906.05 5 10/08/88 897 6 6/26/89 876.75 7 9/13/89 890.71 8 2/8/90 869 A 9 4/24/90 853.29 10 3/4/91 857.54 11 5/23/91 862.64 12 11/2/91 856.3 ENTER NO. OF DESIRED DATA JLnts(1,12) EVTER DATE 01 )UNn! THICKNESS (X0) 6/10/94 I SA DATEzS x y

  &*....*.       .. **...*       *.******     *0**.*

11.C OT 5/1/87

                 $/1/87 9/10/87 7/12/88 10/08/88 0
                                    .252055
                                    .361644 1.2 1.4411 916.79 953.54 915.71 906.05 897
                                                                                                                  /

6/26/89 2.15616 876.75 9/13/89 2/8/90 2.3726 2.77808 890.71 669 f 4/24/90 2.98356 863.29 3/4/91 3.64384 857.54 5/23/91 4.06301 862.64 I I 11/2/91 4.50959 856.3 X0 a 6/10/94 nr90o 776.699

                *"

800.132 T Yugo0 823.566 T95 1.81246 I F-RATIO a 15.5802 WMMRSAN JrE-RATIO CF 1.0 OR 9322TE2 P201132 CWITDENCR IN 122 ELME JRD IM.TERCEPT OF E EISTOEICMT DATA, F-RATIO MEOUXD BE 4 OR 5 IF TE REGMESSICK XQMT= I8 TO BE uSED TO PREICT FUTUIE VALUES. TO UAVE A NICE DEGREZ 0O CON1FIDEN~CE IN THE PREDICTED VALUE, WM RATICBOUL E AT BE3 ZlEA 8 011 9. December 19, 1991 4029 RM OCLR00000246

PROGRAM: LINPREDY Calc. No. C-1302-187-5300-020 BAY NUMBER? 13a all pts Rev. 0 ENTER NAME OF DATE LIST datel3a Page.35of NAME OF MEAN THICKNESS DATASET? mdl3a S N DATE 13A MD13A

      **    *h*******   ******

1 11/15/86 919.08 2 12/17/88 905.27 3 6/26/89 882.84 4 9/13/89 882.98 5 2/8/90 859.04 6 3/28/90 852.84 7 4/24/90 855.63 8 2/23/91 854.51 9 5/23/91 852.88 10 1 /2/91 848.6 ENTER NO. OF DESIRED DATA intsi(2 10) ENTER DATE OF MINIMUM THICKNESS X0) 6/10/94 BAY DATES X Y 13A ALL 12/17/88 0 905.27 6/26/89 .523288 882.84 9/13/89 .739726 882.98 2f8 90 1.14521 859.04 3/2 /90 1.27671 852.84 4/24/90 1.35068 855.63 2/23/91 2.1863 854.51 5/23/91 2.43014 852.88 11/2/91 2.87671 848.6 XO = 6/10/94 YL90 YOHAT YU90 T95 761.348 793.96 826.572 1.89458 F-RATIO = 5.11944 WHEREAS AN F-RATIO OF 1.0 OR GREATER PROVIDES CONFIDENCE IN THE SLOPE AND INTERCEPT OF THE HISTORICAL DATA THE F-RATIO SHOULD BE 4 OR 5 IF THE REGRESSION EQUATI6N IS TO BE USED TO PREDICT FUTURE VALUES. TO HAVE A 'IGH DEGREE OF CONFIDENCE IN THE PREDICTED VALUE, THE RATIO SHOULD BE AT LEAST 8 OR 9. A~ril 15, 1992

59A 0

? OCLR00000247

PROIRAM: LINPREDY EM NUMBER? 17a bottom 4 WR NRM OF DATE LIST datel7a Calc. No. C-1302-187-5300-020 nn MEAN THIONNESS Oa'sAsET? Mdl7ab Rev. 0 Page 3 (eof 49 K DATE1TA SD17AB 6 ******** ****** 1 12/17/88 957.36 2 6/26/89 964.5 3 9/13/89 955.18 4 2/8/90 937.5 5 4/24/90 939.61 6 2/23/91' 934.71 7 5/23/91 942.39 8 11/2/91 932.8 4 METER nO. OF nDzsn MATA ints(1,O) Z DATn OF MINIMNI *!EICTIESS (XO) 6/10/94 BAY DATES x y 17A *OIT 12/17/88 0 937.36 6/26/89 .523288 964.5 9/13/89 .739726 955.18 2/8/90 4/24/90 2/23/91 5/23/91 1.14521 1.35068 2.1863 2.43014 937.5 939.61 934.71 942.39 I & 11/2/91 2.87671 932.8 XO - 6/10/94 Ji YLs0 !OHAT YIugO T95 883.791 906.56 929.329 1.94318 P-PATIO - 3.08187 V=E AN F-RATIO OF 1.0 OR OR.ATER PNOVIDES COBDIECN IN MM SLOPE MM1 DITMRCET OF TSE RIST3ICAT7 DATA, THE F-NATZO SHOULD ZE 4 OR 5 IF THE RECRESSZON 220MON IS TO BE USED TO UNEICT F=MUE VALUES. TO IAVZ A RXGH DEGREE OF OUTDEN0E ZN ut PREDICTED VALUE, W RATIO SULD BE AT' LEAST 8 OR 9.

                                                                                                                 )

December 19, 1991 4:30 PM

                                                                                                            /

I OCLROO000248

P1ROGAN: LfINPREDY RAY NMM=R? 17d ENTER RAMS OF DATE LIST date27d Calc. No. C-1302-187-5300-020 X=3 OF MEAN TRICKNESS VATASET? mel7d Rev. 0 Page W7of '( DAT17D MN17D 2 I 12/2/86 903.51 2 2/17/87 922.16 3 3/1/87 895.07

    £ 8/1/87           890.69 9/10/87          895.28 6 7/12/88          877.93 7 10/08/88         852.22                                                                                   I 8  6/26/89          856.84 g  9/13/89          847.13 10   2)8/90           833.37 11   4/24/90          826.8 12   2/23/91          825.27 13   5/23/91          829.08 14   11/2/91          822.2                                                                         I,)

ENTER NO. or DEsIRD RENER DATE 07 MIrDOZ snah ints(2,14) THI 'I ISS (XO) 6/10/94 I BRAY DAT X 7

 *t*          ***s********

i/l**** /****.* 17D 2/17/87 0 922.16 5/1/87 8/1/87 9/10/67

                          .2
                          .452055
                          .561644 695.07 890.69 895.28 I

7/12/88 1.4 877.93 10/08/88 1.6411 862.22 6/26/89 2.35616 856.84 9/13/89 2/8/90 2.5726 2.97808 847.13 833.37 I 4/24/90 3.18356 826.8 2/23/91 4.01918 825.27 5/23/91 4.26301 829.08 11/2/91 4.70959 822.2 I X0 = 6/10/94 1L90 YCWLf T YU90 195 742.02 758.955 775.891 1.79588 I F-RATIO - 39.0712 WHEREMS AN F-RATIO OF 1.0 OR GREATER PROVIDES CONFIDENCE IN THE MX.01 EDA InMR=CT OF TEHISTO.RICAL DATA.,M 1-RATIO SHOULD BE 4 OR 5 IF 28E REGRESSION 2ZUATION 13 TO BE USED TO PREDICT FUTURE VAL.UES. COFIDENCE IN TEE REDI-so HAVE A RICH DEGERZ OF VALXUE,28E RATIO ==D BR AT i &EAST' 8 OR 9. December 19, 1991 I 4:31 PM i OCL=R00000249

1'IROCdM: LINPPED! RE! NMMERI? 17119 bottom 4 INZER NAME OR DATE LX5T date1719 Calc. No. C-1302-187-5300-020 Rev. 0 NAME OF HEIR THICM04SB DATXSET? Xd1719b Page 5 of 19 DATE1719 MD1719B 1 12/30/68 1003.79 2 6/26/89 1019.07 3 4 5 9/13/89 2/8/90 4/24/90 1016.57 1000.46 988.54 i i 6 2/23/91 987 I 7 5/23/91 982.39 8 11/2/91 971.1 ERg No. cP DEsIRED DAITE intu(I.B) FNDATE OF DInIWJTHICKNESS (XO) 6/10/94

      /AY        DATES          x 17/19 30     12/30/68   0            1003.79 6/26/89 9/13/89
                            . 487671
                            .70411 1019.07 1016.57                                                      1' 2/8/90     1.10959      1000.46 4/24/90    1.31507       988.54 2/23/91    2.15068       987 5/23/91    2.39452       982.39 11/2/91    2.8411        971.1 ZO   -   6/10/94 j
       .90     YO]RAET    Yugo       T95 909.277      936.099   962.222    1.94318 F-R=TIO - 5.27046 WEREAS ARIF-DiTTO OP 1.0 OR G.,ATR PWVXIDZS COMNFIDEC I

IN 'E WLOPE UNDZETE.CEPT O" THE EISTMUCAL DATA, TE I'-RSTIO SEBBLD BE 4 OR 5 IF THE REGRESSION E1UATIMO I8 TO BE USE TO PREDICT FUTURE VaI.ES. To NAVE uA HIm DEGREE CONFIDENCE IN =E PREDI'MED VWLRE, TSE RATIO 6800W 31 AT LEAST 8 OR 9. December 19, 1991 4:33 PM I I OCLROO000250

PRO)GRAM: LINPREDY SAY NW 3042?7/19 top 3 lERM NZME cP DATE LIST date1719 Calc. No. C-1302-187-5300-020 Rev. 0 Page 3' of q NAME CF )MAN TRICEUSE IDATASMT kudlll9t I; DAME1719 =31719T 1 12/30/88 981.71 2 6/26/19 998.61 3 9/13/89 992.4 4 2/8/90 986.29 970.6 I~ 1~ 4/24/90 2/23/91 974.67 7 5/23/91 969.3 8 11/2/91 954.2 I zzR so. or DESIRED DanA inmtis(,) Ezm~ DATE OP )IflZ14W ZSlESS (XO) 6/10/94 Bar z r CAM*** o 981.71 17119 TO 12/30/88 6/26/89 .487671 998.81 9/13/89 .70411 992.4 2/8/9O 1.10959 98,629 4/24/9o 1.31507 970.6 2/23/91 2.15068 974.67 5/23/91 2.39432 969.3 11/2/91 2.8411 954.2 I X0 " 6/10/94 1L%90 YOEAT 1U90 T95 903.606 931.145 958.684 1.94318 1 F-RATIO - 3.1125 WEEW £ AN -RATIO OF 1.0 OR GREAER PROVIDWS CWPIDERCE IN WM SLOPE RIM hNERCIT CF 1-RATIO a20WIm MENISTMICAL DATA, ZM R 4 OR 5 IF WM RWH6250 IQOATI= BE uSED TO PP.IDICT FuTURE VALUES. IS TO TO HAIVEA SIGH DEGRES OF I C=FIINCE IN T= PREDICTED VALU*, TEE RATIO X0w BE AT LEAST a OR 9. December L9, 1991 I 4:32 PH Aw I II i OCLR00000251

  ?JGIMLX: LINPREDY BYMac=D3T? 19a MER MADM     OF DATE LIST datelga                                  Calc. No. C-1302-187-5300-020 NAME OF KMA TIOESS DATASET? melga                                    Rev. 0 Page 4of  -f DA.TE1BA    MEISA

1. 12/2/86 870.22 2 2/17/87 883.64 3 5/1/87 872.93 4 6/1/87 858.6 5 9/10/87 858.29 6 7/12/88 848.57 7 10/08/88 636.91 8 6/26/89 628.82 9/13/89 825.36 10 11 2/8/90 4/24/90 807.78 807.8 I

12 2/23/91 816.67 13 5/23/91 802.8 14 11/2/91 803.2 TZER No. OF DESIRED DATA i*ts(2,14) EIZ DATE o*FI MM THC4[IM6SS (ZO) 6/10/94 4 9AY DATES x T I 19A 2/17/87 0 883.64 5/1/87 .2 872.93 8/1/87 .452055 85836 9/10/87 .561644 858.29 7/12/88 1.4 848.57 10/08/88 1.6411 836.91 6/26/89 2.35616 828.82 9/13/89 2.5726 825.36 2/8/90 2.97808 807.78 4/24/90 3.18356 807.8 2/23/91 4.01918 816.57 5/23/91 4.26301 802.8 11/2191 4.70959 803.2 X0 - 6/10/94 IL90 TORA? Yu9o T95 i 736.008 751.182 766.356 1.79568 F-RATIO a 33.0607 WUERES AN 1-RATIO OF 1.0 OR CREA2ER PROVIDES CC*WIDERCZ IN MME WLOE AIM IN2ERWT OF TIM HI8SVRXCAL DATA, THE F-RTIO BE=TLD uE 4 OR S IF INNE REM SZO ]==NII is TO ME USDTO PREDICT WU'E3 'VALUES, TO EAVE A HUMF DEG=E OF I COFIDECE IN MM PR.EDICTED VALU, TM PATIO SHOVED IE AT LEAST 8 OR 9. i r Dece-ber 19, 4:33 PM 1991 p / I 0CLR00000252

PROGRAM: LINPREDY Calc. No. C-1302-187-5300-020 BAY NUMBER? 19b Rev; 0 ENTER NAME OF DATE LIST datel9b Page W/of NAME OF MEAN THICKNESS DATASET? mdl9b N DATE19B MD19B

   **   *kf********

1 12/2/86 880.84 2 5/1/87 897.63 3 8/1/87 892.21 4 9Z10/87 887.6 5 7/12/88 863.98 6 110/ /88 856.41 7 6/26/89 852.55 8 9/13/89 854.94 9 2Z8/90 840.67 10 4/24/90 839.1 11 2/23/91 852.53 12 5123/91 844.4 13 11/2/91 846.3 ENTER NO. OF DESIRED DATA ints(2,13) ENTER DATE OF MINIMUM THICKNESS (XO) 6/10/94 BAY DATES X Y 19B 5/1/87 0 897.63 8/1187 .252055 892.21 9Z10/87 .361644 887.6 7/12/88 1.2 863.98 10108/88 1.4411 856.41 6/26/89 2.15616 852.55 9.13/89 2.3726 854.94 2Z8/90 2.77808 840.67

          /+24/90       2.98356     839.1 2/23/91        3.81918     852.53 23/Z91    4.06301     844.4 1 /2191        4.50959     846.3 S XO  = 6/10/94 YL90         YOHAT       YU90       T95 783.755      803.851     823.948   1.81246 F-RATIO = 8.70624 WHEREAS AN F-RATIO OF 1.0 OR GREATER PROVIDES CONFIDENCE IN THE SLOPE AND INTERCEPT OF THE HISTORICAL DATA THE F-RATIO SHOULD BE 4 OR 5 IF THE REGRESSION EQUATI6N IS TO BE USED TO PREDICT FUTURE VALUES. TO HAVE A HIGH DEGREE OF CONFIDENCE IN THE PREDICTED VALUE, THE RATIO SHOULD BE AT LEAST 8 OR 9..

April 15, 1992 10:00 AM I. OCLR00000253

PROGRAM: LINPEDY amT M4ER? 19c NAME OF DATE LIST date19c Cale. No. C-1302-187-5300-020 XNTE aCIO U T2ICXMS DATASM m1l9c Rev. 0 Page 4ZOf N DATLE19C HZ19C

   **2 I   12/2/86     866.02 2   5/1/87       900.51 3   8/1/87      888.16 4   9/10/87      888.31 5

6 7 7/12/88 10/08/88 6/26/89 673.46 856.27 845 j 8 9/13/89 844.7 10 9 2/8/90 4/24/90 030.51 822.52 I I 842.8 I 11 2/23/91 12 5/23/91 823.22 13 11/2/91 822.3

0. OF EZShRDAMA ints(2,13) 50.

XZS EmVATE COFl42 TMI4SS (XO) 6/10/94 BAT DATES x y itt 19C 900.51 5/1/87 0 888.16 8/1/87 .252055 888.31 9/10/87 .361644 7/12/88 1.2 873.45 S 10/08/88 6/26/89 9/13/89 2/8/90 1.4411 2.15816 2.3726 2.77808 2.98356 856.27 845 844.7 830.51 822.52

                                                                                                         )

I' 4/24/90 2/23/91 3.81918 842.8 I 5/23/91 4.06301 823.22 11/2/91 4.50959 822.3 XO - 6/10/94 7L90 2T Yugo090 T95

   *******     7**.**      788.i73   1.8****

749.124 768.93 788.735 2.81246 V-PATIO = 19.677 VESPEA8 AN F-MAO OF 1.0 OR G = IOVIES CWTZ21C3 IN MMLOOE BED IN=BCE2? eF WE X2!LRCIL DA~. MM 7-RLTXo SWULD BE 4 OR 5 IF %= REGMMSI ==F0 IS T BE USED TVOPRDICT rUTUIM VEALES. T RAVE A HICH DEGREE or CMMIMDEMcE IN MM PMEDICMD VALUE, 'MM RATIO 8HOULD BE AT IAS= 8 onl 9. December 19, 1991 I 4035 PM I 0CLR00000254

UISTING OF VROCRAM LINPRElU 100.00 PRO3,M 200.02 S PRGRAM LNPREDX Calc. No. C-1302-187-5300-020 300.00 $ POG1W4D BY J.P. HOW1 1/25/89 Rev. 0 400.02 $ REVISED BY J.P. vDRE 2/27/91 Page4-3of 4? 600.00 500.02 $$

REFERENCE:

                   -REVISED BY"APPLIED J.P. HVQm REGRESSION319/91 ANALYSIS",   2ND EDITION               l. Ae c ncc63

(*"x-S 700.00 $ .R. DRAM21 A U. 8N11J 800.00 $ JOHN WILEY A SONS, 1981, PP 47-51 900.00 JOURNAL Ot 1000.00 RLWPAGK 1100.00 TIRE "PROMMM3: LINPREDX" 31200.00 ASSAME("BAY X=UNB2"*,"BAY- ") i 1300.00 ASK('EETR NAM OF V= DAOE LIST,",ENCEFOINH DAMELIST IS ") 1400.02 aB;("U3T OP mm THICK=S DAT&ASET',"BINCYOP. DATES 1") 1M00.02 $ ATASET ON OCDAT 1600.02 N - ZXTS(1,EOELC(DA.TAT)) 1700.02 1800.02 1900.02 TABUATE N,DhTELIST,DATAS.T AK (I

• DCAMZB

0. or men=I oE a DATlIST(O.ECT)

DATA", "SEL, = *) I 2000.02 T - DATASST(SELECT) 2100.00 AZ.K( U VALUE OF MINDI=I THlCKNES3 (Y0)","Y'- ") 2200.00 CAM3 - DAY NfflHR(DATZS) 2300.00 X - tDAYNO-DAYNO(Z))/365 2400.00 lVV -,.ELS(Y) - 2 2300.00 ENECUTE .IJU'REGN1 2600.00 $ T95 - ONE-SIDED T TA= VALUE AT .95 WITH IMF DEG OP PREEDCM I 2700.00 T95 a ARS(*TPROINmWE(.895,3DF)) 2800.00 $ EQUATI*N 1.7.7, PFA 49 2900.00 6 - 5s**2/f3/B T(8B_/sX0)}**2 . 3000.00 3100.00 3200.00 OHRTYKR , (YO-20)/15 X, - (ZOREATR-),RMA)'o/-G) X2 - (TJS*8/R1)*8gTVC((2 T!R-XBAR)*2/62X)÷(1-0)13)/(1-0) 3300.00 $ 22M=M 1.7.6, PAGE 49 3400.00 XSOYR , XOUA-MM+X1+X2, XAEMTR÷X-X2 3500.00 $ XU901o UPP-NER ROMN OF 90% oC.FIDENCE INTERAL ABOUT XD 3600.00 Xg9OTR , .AX(X9mO) 3700.00 $ XL9091M - E BOUND OF 90% CONFIDENCE IEMVA ABOUT X0 3800.00 XLg90E. - XIN(X90ThR) 3900.00 XOHATN0 - DCAUO(i) + INTPAI* (XOATYR*365) 4000.00 4100.00 4200.00 X0HAT2 a NABDATE(X0LA :DA* IN-DAT1*MR.

                 =U9o0o = 03_30(1) + InTPART(XU90YR*385)

XU90 - iMDTE(90J0NOtDA7EIN-DAz*2 E,ABBR,0ME) RABBPm*R) i 4300.00 Z,90NO - DAXNO(1) + INTPRT2(X290CR*3S5) 4400.00 XLg0 - MNERA(XLO*oQ*DATEIN-DATNU E*.,BBRPAR) 4500.00 TABULBATE lYCDATES,X,Y 4600.00 Tin "Y0 4700.00 TYPE "1=90

                            .'20
                           -"."190 I

4800.00 TYPE "XOEAT 90MMOAT 4900.00 TYPE "XL,90 w"XL90 5000.00 TYPE 'T95 f,"295 5100.02 T2172 "C a" I I I OCLR00000255

5200.02 SPACE(2) 5300.02 T "INVERSE ESTIIM=OI0N IS SOT Or MUCE PRAMIC 5400.02 TZE "UMBESS T REGRESSION IS WELL DMESEMIUED, Calc. No. C-1302-187-5300-020 5500.02 TYPE "TEE S*OPE (Bl) IS 8IGNIFICA*T, WICR WXML Rev. 0 5600.03 TYPE "C 8H,0D E SEL T*EROUT 0.20. MW WS Page 44of 5700.03 TXPE -LER MW TMIS, THE RESULTS AM MI4ZING 5800.00 $STHULEAE *OBl",MAISE1,3,EAR 5go0.o0 $T ULATE G,XI,X2 6000.00 SPACE'2) 6100.00 DA%";TIM  ! 6200.00 J3O .AL OFF 6201.00 M J I

                                                                                           'i i

I

LISTING OP PROGRAM LIM]REDO 100 PROGRAM 200$ PROGRAM: LIXPREDYCa .N ). C-1302-187-5300-020 300 $PROGRAXED BY J.P. MOORE 1/27/89 Rev. 0. 400 s RVISED BY J.P. MOORE 40OSl/i Page 4ri if 500 $

REFERENCE:

"APPLIED RECRFSSION WALYSIS", 2=0 ED 0   600 s 700 N.R. DRAPER A H.

JONJ WILEY A SONS, 1981, PP 28-31, 129-133 800 JOURNAL ON 900 N)EWAE 1000 TYPE "PROGRAM: LINPREDY' 1100 .q31QUME(9JlAY NUMB*ER?","BAY- It) 1200 ASK("EETER RAM OF DATE LIST-"'e NCEFORTH DATELIST is ") 1300 SK(o"nAME OF MEAN TRIC3ESS DAmlSET?",HENEo DAM&DET IS") 1400 $ GET DATASET ON 0MAT 1500 N - ITE(1,NOmS(DATASET)) 1600 TABULATE ,DATELIST,DPTASET 1700 ASK ("EITER NO. OF DESIRED DATA", "SELECT - 1800 DAh=S - DhTElIT(SZW-) 1900 T , VAMASET(SELECV) 2000 ASAM.("£ENT*R DATE OF MINIMUM TEICI*-NE (XO)","XO- ") 2100 D&TUO = DlAY ER(DAMES) 2200 X - (DAXEO-DATNO(1))/365 i 2300 XONO - DAYXUEER(X0) 2400 X0R* - (X0NO - DnMo(1))/365 2500 2600

          =7 - X0E1SCY) - 2 NOMJUTE LI,=GN1                                                                                      4 2700 $ T95 - ONE-SIDED T TABLE VALUE AT .95 WITH IW DEG OP FREEDOM 2800    T95 - ABs(TPRONINVZRl(.95,XD7))

2900 $ EQUATICU 1.7.7, PACE 49 3000 C - T95**2/(1/sGT=(SSO/S)C))**2 3100 EORAT- D0 + z1*XOYR 3200 VAW*OEM - 8SQ*((I/N) + (X01R-XBAR)**2/SUM((X-XDAR)**2)) 3300 SDYORXT aTD ERROR OP ESTIMATE (SEE) OF KEAN IKICZESS 3400 S*OX*AT A,3.(8=A(%VARYIOEA)) 3500 $ UPPER BOND OP 90% CONFZIDENCE INTERVAL ABOUT Y0

                 -U0O 3600    Yg90 "- OBAT        .+T95*SD=OEAT 3700 $L90       - LOWER SOUND 0F 90% CONFIDENCE InREVAL ABOUT Y0 3800    IL,0 =     OM00T - T95*SDTOE.T 3900    TRELATE BAY,DVAES,X,T 4000    TIE ",0      -"XO 4100      AmLATE YLg90,yOHAT,g00,T095 4200 S TABULPM 20,BI,SSQS,XRAR 4300 $ TBULKTR V3RYOHAT,SDYOA=

4400 TYPE "F-RATIO -" 1/9 4500 ZPACE(2) 4600 TIUB "WHEAS AN 1-RATZO OF 1.0 OR GUAM PROVIDES CONFI CLE" 4700 TYPE "II TEE SLOPE AND INTECEPT OP THE HISTOIRICAL DATA, THE" 4500 TYPE "1-RA*IO SHOULD BE 4 OR 512 TEE SRERESSION EQUATION IS TO" 4900 TYE "IM USED TO PREDICT FUTURE V=ES. TO HAVE A 1CHGDEGREE OF" 5000 TYPE "ý IDErNCE IN ME PREDICTED VALUE, THE RATIO S=O BE AT' 5100 TYPE "IMA.T 8 OR 9." 5200 SPAcE(2) 5300 DATE;TIME 5400 3URRAL OFF 5401 MM I OCLROO000257

X.IB!ZNG OF PROGMAX LIMPEEDS 1.0 PROGRAMt 2.0 S PRomtafl4= zY j.p. momR 1/27/89 Calc. No. C-1302-187-5300-020 2.1 S RSFERENCEt "APPLIED IM=RSrIoW ANALYSIS", 2ND ED Rev. 0 2.2 S X.J. DRAPER a R. SMITH Page (of q 2.3 JOHN WILEY a SONS, 1981, PP 28-31 3.0 MSERMEC"BAY =-MER?0",ftA7. -) 4.0 AZBW'ZRTER NAME OP DATE LIS2-,'.DA=Sw 11) 5.0 .BX("EUTE VALME OF Y'"fet="f) 6.0 ASX('UETE XO DAF',t"xO-" 7.0 DAMN - DAYNUNEER(D=XS) 8.0 x - (DUMO-DI*XN(1))/365 8.2 IOMa - DAYNUMBER4XO) 8.4 KOXR - (ZONO - DUMN(1))/365 9.0 MYD- NOELS(X) - 2 10.0 ERSOUTZ LXNREUE 11.0 T99 - A33S(TPR~n3VES( .29,NDF)) 13.0 YOBAT - 20 + ll'IXCY 14.0 VARYCHA2X - BSg' ((1/N) + tE0R-CDAR)"*2I8M( (X-XMR~)l 14.5 s Svy0EAT - Ba= ERRG OF W3TiIOm (MME) or mOm mu=~ 15.0 BEVOBAT - AM8(BS (VARYEATE) 15.5 YU98 - UIPPER SO1MD OF 98% COMIEnCE InaERVAL ABOUT 16.0 YU98 - TONE! + 299*SDYOVAT 16.5 $ L98 - LO!SR XOUD OF 98% CONtFZDECZ INTERVAL lABOU 17.0 11.98 - TOBAT - 9*DER /1 20.0 JQMUMBL CH 21.0 TYPE IPROGRAMs LmE " 22.0 TASUIATE nAyDamE,x,T 22.5 TOPE "XO -'%D 23.0 MRUZATE Yh98,Y08AT,TU98,T99 24.0 S T&EUWM! S0,D+/-BEQISAAR 25.0 $ TERULATE VAEaNIPSDOUA! 26.0 DATE;TnM I) 27.0 JOURNAL CF? 28.0 ENM I i 0CLR000 OCLROO000258

Us1T210 OF PaOGIAM IINFR0W 1.0 PROC3Ii Caic. No. C-1302-187-5300-020 2.0 $ 1IR0GR04D BY J.P. KX19 1/25/89 Rev. 0 3.0 RWUAM('BAY 3Wuraso","BAY -) Page Aof qq 4.0 LSK("E1f=1 NAME OF DATE LIBT-",DATBS" 5.0 MK("ENME VALME OF V',"-" s) 5.5 RSZ("WER VALUJE OF roll," o

                                      ")o 6.0 DATIIO - DAYSUJER(DAM2ES) 6 7.0 X -   (DAkNO-DAYl1O(1))/3   5 6.0 ND? - NOELS(Y) - 2 g.0 EXECUTE LMhREO1 10.0 T25 - ABsC(PROrxVBz=(.95,NDF))

11.0 8 - S=CFSSQ) 2 12.0 0 - TS5**2/(Z1/6QRT(S8Q/WM1))** 13.0 XDEW - (Y0-30)/31 14.0 X2 - X=XM*/LG 15.0 X2 - (¶T95*6/31)*8Q(((J~tCCX0ER)**2/81)Ii1-C2)/N)/l

16. X90 .XOBAT!+x1*2. XMWA+X1-X2 16.5 X1390 nMAX(190) 17.0 fL.90 M=CZNX90) 18.0 JOURNAL ON 20.0 TABlULATE BAY,DA=8,xy 21.0 TABULATE Y0,11190.X0AT,XL90 22.0 T~ABULATE Z0#21#63Q,52M,G,19 23.0 WThABUAT T9S,G,X1,12 24.0 DAiTF;TZD3 25.0 JOURNAL OFF I

26.0 END I I OCLROO000259

LISTING OF PROGRAM LIfqREGN1 1.0 PROGRAM 2.3 $PRO=W: LInE=l Calc. No. C-1302-187-5300-020 2.0 $ THIS PRGRAM P32PORS LINEAR. 33 salO03 Rev. 0 3.0 S 4.0$SBY WERENCE: "-

                                            .R.

D LINE R ssIo, DRAPER &K.MITH 2 EI Page 4.-Cof 4.' 5.0 $ oi vim a BcS, 1981 6.0 $ PROGRAMME BY J.P.MOCSE 1/26/89 7.0 $ INPUTS: 8.0 $ X - INDEPENDENT VARIABLE ARRAY 9.0 $ T - DEPMWENT VARIABLE AMW 10.0 $ EDP - NUMBE OF DE*QU= OF F1REDCH 11.0 $ 12.0 $ UTiPUTS: 13.0 $ 30 - T INTERCEPT OF FI22TD STRAIGHT LIZE 14.0 $ 51 - SLOPE OF F1118V LIKE 15.0 $ AT - I19rICZZD VALUES OF Y 16.0 $ 80 - STD ERROR OF ESTIMATE OF THE ITERCEPT DO 17.0 $ SD01 - LTD ERROR OF ESTIMATE OF TME SLOPE Bl

• e 18.0 19.0 $ CALcLATIONS 20.0 N - NCELS(X) 21.0 84 - 814(3) 22.0 YEAR ,, SUMY/4 23.0 SUMX - SUM(X) 24.0 21U - 81381/(N) 25.0 5UM - SUM(X*T) 26.0 lCCSQ - LM(X**2) 27.0 S= -,,U - SXMX*SAmUW 28.0 S=01 - SUM(X**2) - (suM(X))**2/N 29.0 STY - UM3(Y**2) - (sm4(Y))**2/N 30.0 31 - EcrT/m 31.0 E0 - TRAR-21*33AR

  .. 132.0          8.S3100 - (SXY**2)/M*X 32.5 $   SSQ - MEAN SQUZ3D ERROR (MSZ) 33.0     SS     - ABS(BSY      - BSDIBO)/(N-2) 33.2 $   9 - ROOT MEAN SCUME ERROR CRISE) 33.5     a - SQlT(SSQ) 33.7 $   SDBI - STD ERWOR OF ESTIMATE (SEE)              OF SLOE B1 34.0        m,1 - SQHT(8Q/S=0) 34.5 $   880 - lT          ERROR OP ESTIMATE (SEE) OF IWMMRGPT DO 35.0     88o - S=((8sQ*SuIxSQ)/(N*e))

36.0 YTAT 3 0 # Sl*X 37.0 .3s a AT-Y 38.0 M - 81E( (TAT - AUR)"*2)/SUM((Y - AR)-*2) 39.0 END O I i

  *
                                     ,                                                                         if
                               ~OCLRO0000260

Citizen's Exhibit NC8 4

Technical Functions Citizen's Exhibit NCI Safety/Environmental Determination and 50.59 Review (EP-01 6) T,2na I f c 1ý11 OCNGS .w - iocumen ActiviwTIe Steel Shell Plate Thickness Reduction SDll S ev. No. tocument No. (ifapplicable) Doc. Rev. No. SE No. 000243-002

 'Type of Activity (modification, procedure, test, experiment, or document):
         -  l                     Document q    .Does      this document involve any potential non-nuclear environmental concern?                0 Yes To answer this question, review the Environmental Determination (ED) form. Any YES answer on the El No ED form requires an Environmental Impact Assessment by Environmental Controls, per 1000-ADM-4500.03. If in doubt, consult Environmental Controls or Environmental Licensing for assistance.

If all answers are NO, further environmental review is not required. In any event, continue with Question 2, below.

  ;2. Is this activity/document listed Section I or II of the matrices in Corporate Procedure      [5 Yes    0 No 1000-ADM-1291.01?

If the answer to question 1 is NO, stop here. This procedure is not applicable and no documentation is required. (if this activity/document is listed in Section IV of 1000-ADM-1 291 review on a case-by-case basis to determine applicability.) If the answer is YES, proceed to question 3.

3. Is this a new activity/document or a substantive revision to an activity/document? I3 Yes 0 No (See Exhibit 2, paragraph 3, this procedure for examples of non-substantive changes.)

If the answer to question 3 is NO, stop here and complete the approval section below. This procedure is not applicable and no documentation is required. If the answer Is YES, proceed to answer all remaining questions. These answers become the Safety/Environmental Determination and 50.59 Review.

4. Does this activity/document have the potential to adversely affect nuclear safety. [N Yes 0 No or safe plant operations?

5. Does this activity/document require revision of the system/component description (9 Yes 0 No in the FSAR or otherwise require revision of the Technical Sepcifications or any other part of the SAR?

6. Does the activity/document require revision of any procedural or operating description 0 Yes EINo in the FSAR or otherwise require revision of the Technical Specifications or any other part of the SAR?

7. Are tests or experiments conducted which are not described in the FSAR, the 0 Yes EK No Technical Specifications or any part of the SAR?

IF ANY OF THE ANSWERS TO QUESTIONS 4. 6. 6. OR 7 ARE YES, PREPARE A WRITTEN SAFETY EVALUATION FORM. If the answers to 4, 5, 6, and 7 are NO. this precludes the occurrence of an Unreviewed Safety Question or Technical Specifications change. Provide a written statement In the space provided below (use back of sheet if necessary) to support the determination, and list the documents you checked. NO, because: Documents checked:

8. Are the design criteria as outlined in TMI-1 SDD-TI-000 Div. I or OC-SDD-000 Div. I 0 Yes El No Plant Level Criteria affected by, or do they affect the activity/document?

If YES, indicate how resolved:

  -            *...                          AF~; ... S. ..(r
                                    .".APPROVA                     name andsign)        ._._.__,_

Engineer/Originator h z-x Date Sconnge J. D. Abramovici La I-Responsible Technical

                    )     Revieweris             ,ý f f      /Date                                                    Z-Other Revieweris)                           N                                                            Date OCLROO001306

SAFETY EVALUATION CONTINUATION SHEET Page 40 of 69 SE-000243-002 Rev. No. 11 1.0 PURPOSE The purpose of this safety evaluation is to assess the structural integrity of the Oyster Creek drywell pressure vessel. This revision incorporates data on vessel thickbess, sandbed coating inspections and resulting corrosion rates based on data obtained through September 1994 and assesses the period of time for which vessel structural integrity can be assured. 1.1 Introduction The Oyster Creek drywell pressure vessel is of steel construction. Its original design incorporates a sandbed which is located around the outside circumference between elevations 81114" and 1213". The sand was removed during the 14R outage (December 1992) and the steel surfaces coated. Leakage was observed from the sandbed drains during the early to mid 1980's, indicating that water had intruded into the annular region between the drywell pressure vessel and the concrete shield wall. The presence of water in the sand was confirmed later when a water level (i.e., free water) was discovered during core boring operations to install anodes for cathodic protection (CP). Concerns about the potential for corrosion of the vessel resulted in thickness measurements being taken in the sandbed region in 1986. These measurements indicated that the vessel in the sandbed region-was thinner than the 1.154 inch nominal thickness originally specified by Chicago Bridge & Iron Company (CBI) (Reference 2.3.1). Additional thickness measurements at elevations 50'2" and 8715" were taken in 1987. These measurements also indicated areas where the pressure vessel was thinner than the originally specified. The specified nominal thickness at these elevations is 0.770 inches and 0.640 inches respectively. Since 1987 OPUN has developed and implemented a drywell vessel corrosion monitoring program (Reference 3.1.4.21) in which inspections are conducted at identified corroded locations. Inspections have been periodically performed during refueling outages and outages of opportunity in the former sandbed region, in the spherical region (elevation 50'2" and 51'10"), and in the cylindrical region (elevation 87',5). 1.2 Backaround Discssion Discovering that the drywell pressure vessel thickness was less than originally specified necessitated a number of activities. The purpose of these activities was to establish that the vessel was structurally acceptable to support continued safe operation of Oyster Creek. A summary of the activities undertaken and the resulting conclusions are provided herein. 1.2.1 Vessel Thickness Measurements References 3.1.4.1, 3.1.4.5, 3.1.4.6, 3.1.4.22 and 3.1.4.23 document the non destructive ultrasonic testing examination methods utilized to measure vessel thickness, the locations chosen for thickness measurements, the locations for metallurgical plug samples taken from the drywell vessel and the extensive amount of data taken (in excess of 1,000 individual UT 008/250 July 21, 1995 OCLROO001345

SAFETY EVALUATION CONTINUATION SHEET Page 41 of 69 SE-000243-002 Rev. No. 11 readings). obtaining the thickness measurements over a large portion of the vessel's circumference at four elevations enabled GPUN to establish an ongoing corrosion rate monitoring program and assess the structural integrity of the vessel. As documented in Reference 3.1.4.29 in April of 1991 a supplemental augmented series of inspections were performed on the drywell vessel. Results were that all inspected locations meet code requirements. 1.2.2 Corrosion Assessment References 3.1.4.2 and 3.1.4.3 document the metallurgical evaluations of the two inch plug samples which were removed from the vessel in the sandbed region in December, 1986 and the upper elevation (EL. 50'20) in November, 1987. Reference 3.1.4.24 documents metallurgical evaluation of an additional two inch plug removed in April, 1990. The type of corrosion noted, coupled with an assessment of the vessel construction and operating history, allowed GPUN to establish the probable cause of the corrosion and to conservatively project corrosion rates. GPUN conducts ongoing periodic vessel thickness measurements which statistically monitor and establish corrosion rates. The ongoing measurements are not taken in all the locations where measurements were taken initially in 1986, 1987 and 1990. The initial locations where corrosion/material loss was most severe were selected for the ongoing program. This-reduction of inspection scope was done primarily to reduce the man-rem exposure received when taking drywell measurements. Note that a spot check of locations measured initially was performed during the 12R (October, 1988) outage which confirmed proper selection for ongoing measurements. In March, 1990 an additional check was performed at elevation 50'2". This check consisted of a continuous UT "A* scan In all accessible areas in a one inch band at elevation 50'2". Results confirmed that the existing grid in Bay 5 was among the thinnest at this elevation. As a result of this check, three additional grids at elevation 50'2" were added to the program. Elevation 50'2" is representative of vessel plates originally delivered with a mean nominal thickness of

                      .770 inch and installed between elevation 23'6" to 51'.

In April, 1990 an additional elevation was investigated for corrosion. This elevation at 51'10" is representative of drywell vessel plate originally delivered with mean nominal thicknesses of .722 inch and installed between elevation 51' to 65'. This investigation was performed by continuous UT "A" scan in a one inch band, at elevation 51'10". Results showed only one area which was less than nominal. An inspection grid of this area (Bay 13) was added to the inspection program. 008/250 July 21, 1995 OCLROO001346

SAFETY EVALUATION CONTINUATION SHEET Page 42 of 69 SE-000243-002 Rev. No. 11 Corrosion assessments have been periodically accomplished as summarized herein. The previous bounding corrosion rate projections (discussed in previous versions of this Safety Evaluation and in Ref. 3.1.4.2 and 3.1.4.3) are no longer accurate and are not discussed in this revision of this safety evaluation. 1.2.3 Corrosion Rate Assessment Reference 3.1.4.7, 3.1.4.10, 3.1.4.11 through 3.1.4.14, 3.1.4.25 through 3.1.4.28, 3.1.4.31 through 3.1.4.34, 3.1.4.36, 3.1.4.37, and 3.1.4.40 document the ongoing I statistical analysis of vessel ultrasonic thickness (UT) measurements as they are taken at specific locations over time. The corrosion rate monitoring program involves the establishment of six inch by six inch grid locations on the vessel interior, the use of a template with 49 holes on one inch centers for locating the UT probe, a specified +/- 1/8 inch tolerance on the location of subsequent measurements and taking thickness measurements periodically. This program has enabled GPUN to statistically determine corrosion rates at these grid locations. since the grid locations are in the known areas where corrosion/material loss is most severe, the corrosion rates and projected wall thicknesses are determined over a small fraction of the drywell but conservatively applied uniformly. 1.2.4 Structural Assessment References 3.1.4.17 through 3.1.4.19 provide an overall - analysis of the Oyster Creek drywell pressure vessel structural requirements. The UT readings obtained through September, 1994 and resulting statistical analysis coupled with the GE Nuclear structural analyses and a recently NRC approved license amendment establishing a 44 psig design pressure in place of 62 psig (Reference 3.1.2) provide the structural basis for assuring safe operation of Oyster Creek until end of plant license (April 9, 2009). The corrosion rates, where available, have been used to project material loss. The structural evaluations have been performed assuming minimum uniform thicknesses in the areas of concern. Since corrosion is confined to specific areas, the existing evaluations and resulting vessel thickness requirements are very conservative in that they do not take credit for actual wall thicknesses in excess of the minimum used in the evaluations. In addition, the coating inspection of the former sandbed region insures the corrosion rate at this area remains at zero. 008/250 July 21, 1995 OCLROO001347

SAFETY EVALUATION CONTINUATION SHEET Page 43 of 69 SE-000243-002 Rev. No. 11 1.3 Purpose Summary This safety evaluation will demonstrate that (based on data collected through September, 1994) plant operations can continue until end of license life based on the structural evaluation of the drywell. Action has been taken to eliminate leakage from the reactor cavity region, and for periodic surveillance (Ref. 3.1.4.21) of vessel thickness at intervals that ensure that the wall thickness will not decrease below acceptable levels between inspections. The former sandbed area of the drywell has been cleaned and coated (during 14R Outage) to stop corrosion. The coating is visually inspected to ensure it remains effective. Additionally, the analysis of the UT'data collected during the most recent inspection (September 1994) indicates that for the upper elevations of the drywell, there is no evidence of ongoing corrosion. 2.0 SYSTEMS AFFECTED 2.1 System No. 243, Drywell and Suppression System, particularly the drywell vessel structure. 2.2 Drawing showing original thickness - Chicago Bridge and Iron Co., Contract Drawings 9-0971, Drawing Nos. I through 11. 2.3 Documents which describe the Oyster Creek drywell pressure vessel design. 2.3.1 "Structural Design of the Pressure Suppression Containment Vessel" for JCP&L/Burns and Roe, Inc., Contract No. 9-0971, by CB&I Co., 1965. 3.0 EFFECTS ON SAFETY 3.1 Documents that Describe Safetx Function & Evaluations 3.1.1 OCNGS Unit I Facility Description and Safety Analysis Report 3.1.1.1 Licensing Application, amendment 3, Section V 3.1.1.2 Licensing Application,"Amendment 11, Question 111-18 3.1.1.3 Licensing Application, Amendment 15 3.1.1.4 Licensing Application, Amendment 68 3.1.2 Technical Specification Vocuments 3.1.2.1 Technical Specification and Bases - OCNGS Unit, Appendix A to Facility License DRP-16, JCP&L Docket No. 50-219, Sections 3.5, 4.5, 5.2. 3.1.2.2 Technical Specification Amendment 165. 008/250 July 21, 1995 OCLROO001348

SAFETY EVALUATION CONTINUATION SHEET Page 44 of 69 SE-000243-002 Rev. No. 11 3.1.3 Rerulatory Documents 3.1.3.1 10CRFS0, Appendix A. General Design Criteria for Nuclear Power plants

                                   -  Criterion 2     -  Design basis for Protection against natural phenomena
                                   -  Criterion 4     -  Environmental and Missile Design Bases
                                   -  Criterion 16    -  Containment Design
                                   -  Criterion 50    -  Containment Design Basis 3.1.4    GPUN Technical Data Reports (TDRI,    Calculations and Drawings 3.1.4.1      TDR 851 Assessment of Oyster Creek Drywell Shell.

3.1.4.2 TDR 854 Drywell Sandbed Region Corrosion Assessment. 3.1.4.3 TDR 922 Drywell Upper Elevation, Wall Thinning Evaluation. 3.1.4.4 (This reference has been superseded by References 3.1.4.17 through 3.1.4.19). I 3.1.4.5 Sketch 3E-SK-S-89, Ultrasonic Testing - Drywell Level 50'2" - 87'5" Plan. 3.1.4.6 Sketch 3E-SK-S85, Drywell Data UT Location Plan. 3.1.4.7 TDR 948, Statistical Analysis of Drywell Thickness Data. 3.1.4.8 NRC Letter Docket 50-219, dated October 26, 1988, subject "Oyster Creek Drywell Containment". 3.1.4.9 Primary Containment Design Report, dated 9/11/67, Ralph M. Parson Company. 3.1.4.10 Calc. C-1302-187-5360-006 "Projection of Drywell Mean Thickness thru October, 1992". 3.1.4.11 Calc. C-1302-187-5300-008 "Statistical Analysis of Drywell Thickness Data thru 2/8/90". 3.1.4.12 Calc. C-1302-187-5300-009 Rev. 0 "OC Drywell Projected Thickness". 3.1.4.13 Calc C-1302-187-5300-001 Rev. 0, "Statistical Analysis of Drywell Thickness Data thru 4/14/90". 3.1.4.14 Calc C-1302-187-5300-012 Rev. 0, "OCDW Projected Thickness Using Data thru 4/24/90"-. 008/250-July 21, 1995 OCLROO001349

SAFETY EVALUATION CONTINUATION SHEET Page 45 of 69 SE-000243-002 Rev. No. 11 3.1.4.15 This reference no longer applicable, therefore, is deleted. 3.1.4.16 This reference no longer applicable, therefore, is deleted. 3.1.4.17 "Justification For Use of Section I1I, Subsection NE, Guidance in Evaluating The Oyster Creek Drywell", Technical Report TR-7377-1, dated November 1990, Teledyne Engineering Services. 3.1.4.18 "An ASME Section VIII Evaluation of Oyster Creek Drywell for without Sand Case, Part I, Stress Analysis", dated February 1991, GE Nuclear Energy, San Jose, CA. 3.1.4.19 "An ASME Section VIII Evaluation of the Oyster Creek Drywell for without Sand Case, Part 2, Stability Analysis", Rev. 2, dated November 1992, GE Nuclear Energy, San Jose, CA. 3.1.4.20 This reference no longer applicable, therefore is deleted. 3.1.4.21 GPUN Specification IS-328227-004, Revision 10, "Functional Requirements For Drywell Containment Vessel Thickness Examination". I 3.1.4.22 Sketch 3E-Sk-M-275, Rev. 0, "UT Drywell Level 50'2", March 1990 Readings". 3.1.4.23 Sketch 3E-Sk-M-358, Rev. 0, "UT Drywell Level 51'-10", April 1990 Readings". 3.1.4.24 "Oyster Creek Drywell Corrosion Evaluation", dated June 1990, GE Nuclear Energy, San Jose, CA. 3.1.4.25 Calc C-1302-187-S300-01S, "Statistical Analysis of Drywell Thickness Data Thru 3/3/91". 3.1.4.26 Calc C-1302-187-5300-016, "OCDW Projected Thickness Using Data Thru 3/3/91". 3.1.4.27 Cale C-1302-187-5300-017 "Statistical Analysis of Drywell Thickness Data thru May, 1991". 3.1.4.28 Calc C-1302-187-5300-018, "OCDW Projected Thickness using Data thru May, 1991". 3.1.4.29 GE Report "Final Report - Oyster Creek Drywell Containment Vessel Random UT Project" dated May 8, 1991. 3.1.4o30 IS-402950-001, Rev. 0 Functional Requirements for Augmented Drywell Inspections. 008/250 July 21, 1995 OCLROO001350

SAFETY EVALUATION CONTINUATION SHEET Page 46 of 69 SE-000243-002 Rev. No. 11 3.1.4.31 Calc C-1302-187-5300-19 "Statistical Analysis of Drywell Thickness Data thru November, 1991". 3.1.4.32 Calc C-1302-187-5300-20 "OCDW Projected Thickness Using Data thru November 1991". 3.1.4.33 Calc C-1302-187-5300-021 "Statistical Analysis of Drywell Thickness Data thru May, 1992". 3.1.4.34 Calc C-1302-187-5300-022 "OCDW Projected Thickness Using Data thru May, 1992". 3.1.4.35 Safety Evaluation SE-402950-005 "Removal of Sand from Drywell Sandbed". 3.1.4.36 Calc C-1302-187-5300-025 "Statistical Analysis of Drywell Thickness Data thru December 1992". 3.1.4.37 Calc C-1302-187-5300-024 "OC DW Projected Thickness Using Data thru December, 1992". 3.1.4.38 TDR 1108 Summary Report of Corrective Action Taken form Operating Cycle 12 through 14R Outage. 3.1.4.39 Calc C-1302-187-5300-024 "O.C. Drywell External UT Evaluations" in the Sandbed. 3.1.4.40 Calc C-1302-187-5300-028 - Statistical Analysis of Drywell Thickness Data thru September, 1994. 3.1.4.41 Memo #5514-94-319 - Dated September 30, 1994 - Subjectt Inspection D.W. Sandbed Coating in Bay 11 - O.C. 3.1.4.42 Calc C-1302-243-5320-071 - Rev. 1, "Drywell Thickness Margins." 3.1.4.43 Memo #5340-94-120 - Dated November 9, 1994 -

Subject:

Video Inspection of DW Sandbed Bay #3. 3.1.4.44 Memo #5340-§5-062 - Dated July 12, 1995 -

Subject:

Life Expectancy of Drywell Shell Coating in Former Sandbed O.C. 3.1.5 Industry Codes and Standards Annlicable Codes 3.1.5.1 The ASME Boiler and Pressure Vessel Code and applicable nuclear code cases utilized for the design of the drywell pressure vessel are as listed in References 3.1.4.17 through 3.1.4.19. 008/250 July 21, 1995 OCLROO001351

SAFETY EVALUATION CONTINUATION SHEET Page 47 of 69 SE-000243-002 Rev. No. 11 3.1.5.2 Annlicable Drvwell Shell Plate Material Standards/Specification SA-212 High Tensile Strength Carbon - Silicon Steel Plates for Boilers and other Pressure Vessels. 3.2 Drywell Pressure Vessel Safety Function Drywell Geometry Descrirtion 3.2.1 The drywell, sometimes referred to as the containment vessel or containment structure, houses the reactor vessel, reactor coolant recirculating loops, and other components associated with the reactor system. The structure is a combination of a sphere, cylinder, and 2:1 ellipsoidal dome that resembles an inverted light bulb. The spherical section has an inside diameter of 70'. The cylindrical portion connecting the sphere to the dome has a diameter of 33'. The structure is approximately 99' high. The plate thicknesses vary from a maximum of 2.56" at the transition between the sphere and the cylinder down to a minimum of 0.640" in the cylinder. The dome wall thickness is 1.18". Figure 1 illustrates the drywell structure along with the pertinent dimensions. The top closure, which is 33' in diameter, is made with a double tongue and groove seal which permits periodic checks for leak tightness. Ten vent pipes, six feet six inches in diameter, are equally spaced around the circumference to connect the drywell and vent header to the pressure suppression chamber. The drywell interior is filled with concrete to elevation 10'3" to provide a level floor. Concrete curbs follow the contour of the vessel up to elevation 12'3" with cutouts around the vent lines. On the exterior, the drywell is encapsulated in concrete of varying thickness from the base elevation up to the elevation of the top head. From there, the concrete continues vertically to the level of the top of the spent fuel pool. The base of the drywell is supported on a concrete pedestal conforming to the curvature of the vessel. For erection purposes a structural steel skirt was first provided supporting the vessel. A portion of the steel skirt was left in place which serves as one of the shear rings that prevent rotation of the drywell during an earthquake. The proximity of the biological shield concrete surface to the steel shell varies with elevation. The concrete is in full contact with the shell over the bottom of the sphere at its invert elevation 213" up to elevation 8"11%". At that point, the concrete is stepped back 15 inches radially to form a pocket which continues up to £ elevation 12'3". The pocket was originally filled withi sand which formed a cushion to smooth the transition of the shell plate from a condition of fully clamped 008/250 July 21, 1995 OCLROO001 352

SAFETY EVALUATION CONTINUATION SHEET Page 48 of 69 SE-000243-002 Rev. No. 11 between two concrete masses to a free standing condition. The sand pocket was connected to drains which allowed drainage of any water which might enter the sand. The sand was removed during the 14R outage (December 1992). The sand "springs" helped to ease this transition. GE analysis (Ref. 3.1.4.18 and 3.1.4.19) has shown that the sand is not required so long as vessel thickness in that region is greater than or equal to .736 inches (with margin as stated in 3.3.2.1). Justification for removing sand from the sandbed is covered under a separate Safety Evaluation (Ref. 3.1.4.35). As stated above, the sand was completely removed and the drywell vessel was coated in the sandbed region during the 14R refueling outage (Figure 2). The sand was removed via ten (10) 20" diameter access holes drilled equally spaced through the containment concrete shield wall. Up from elevation 1213" there is a 3" gap between the drywell and the concrete biological shield wall which is filled with foam material that provides no structural support. An upper lateral seismic restraint, attached to the cylindrical portion of the drywell at elevation 82.17 ft., allows for thermal, deadweight, and pressure deflection, but not for lateral movement due to seismic excitation. All penetrations for piping, instrumentation lines, vent ducts, electrical lines, equipment accesses, and personnel entrance have expansion joints and double seals where applicable. The spherical area is described by 10 segments, one at each downcomer, referred to as bays. The bays are odd numbered 1 thru 19 (Figure 3). 008/250 July 21, 1995 OCLRO0001353

SAFETY EVALUATION CONTINUATION SHEET Page 52 of 69 SE-000243-002 Rev. No. 11 3.2.2 Drywell Pressure Vessel Safety Function 3.2e2.1 Functional Design The drywell pressure vessel is one of the major structural components of the Primary Containment System (PCS) discussed in Section 6.2 of the Oyster Creek Nuclear Generating System Update FSAR. The safety function of the Primary Containment System is to accommodate, with a minimum of leakage, the pressures and temperatures resulting from the break of any enclosed process pipe; and, thereby, to limit the release of radioactive fission products to values which will insure offaite does rates well below 10CFR100 guideline limits. 3.2.2.2 Desion criteria The design criteria for the Containment are as follows:

a. To withstand the peak transient pressures (coincident with an earthquake) which could occur due to the postulated break of any pipe inside the drywell.

b. To channel the flows from postulated pipe breaks to the torus.

C. To withstand the force caused by the impingement of the fluid from a break in the largest local pipe or connection, without containment failure.

d. To limit primary containment leakage rate during and following a postulated break In the primary system to substantially less than that which would result in offoite doses approaching the limiting values in 10CFR10O.

e. To include provisions for leak rate tests.

f. To be capable of being flooded following a Design Basis Accident to a height which permits unloading of the core.

3.2.2.3 Drvwell Vessel Desion Pressure and Temperature Parameters The drywell and connecting vent system tubes are designed for 44 psig, internal pressure at 292"F, and an external pressure of 2 psig at 205*F. 008/250 July 21, 1995 OCLROO001357

SAFETY EVALUATION CONTINUATION SHEET Page 53 of 69 SE-000243-002 Rev. No. 11 The design lowest temperature to which the-primary containment vessel is subjected is 30*F. 3.3 Effects of Drvwell Pressure Vessel Thickness Reduction In order to demonstrate that the vessel thickness reduction will not adversely affect the ability of the drywell to perform its safety function, GPUN establishes a conservative corrosion rate, projects vessel thickness, and shows by analysis that allowable stresses are not exceeded for the design basis load conditions. 3.3.1 Results of Corrosion Monitoring Program 3.3.1.1 Monitoring Program Summary Reference 3.1.4.21 defines the drywell corrosion inspection program. This program identifies nine (9) locations for UT inspection. These nine locations were selected for inspection based on extensive drywell thickness investigation performed during the initial corrosion investigation phase (1986 through 1991). These nine (9) locations (exclusive of the former sandbed region) exhibited that worst metal loss and therefore were selected for monitoring wall thickness. Originally, the knowledge of the extent of corrosion was based on a UT inspection plan involving going completely around the inside of the drywell at several locations. Nine six-by-six grids on either side of each vent penetration were used to characterize the situation at the elevation of the sandbed. At each of the upper elevations a belt-line sweep was used with readings taken on as little as one inch centers wherever thickness changed between successive nominal 6" centers. Grids were established in the upper elevations in this way. As experience increased with each data collection campaign, only grids showing evidence of change were retained in the inspection program. Additional assurance regarding the adequacy of this inspection plan was obtained by a completely randomized inspection, involving 59 grIds,that showed that all inspection locations satisfied code requirements. As a minimum, the nine locations above the former sandbed region specified in the program, will be inspected during the 16R refueling outage and every third refueling outage thereafter. This frequency of 008/250 July 21, 1995 OCLROO001358

SAFETY EVALUATION CONTINUATION SHEET Page 54 of 69 SE-000243-002 Rev. No. 11 inspection is considered adequate because most recent data obtained indicates that there is no evident of ongoing corrosion at the upper elevations of the drywell vessel. Reference 3.1.4.21 also covers coating inspection of the drywell shell exterior at the former sandbed region. The corrosion in this area of the drywell vessel was arrested during the 14R refueling outage (December 1992), as the steel surface was coated for corrosion protection. As stated in 3.3.1.7 of this safety evaluation, the coating was inspected during the 15R refueling outage on a sample basis. Results of the inspection were satisfactory with no indications of coating failures. As a minimum, additional inspections of the coating will be conducted during the 16R refueling outage and again during refueling outage 18R. This frequency of inspections is adequate based on results of prior coating inspection and estimated coating life (8-10 years) per reference 3.1.4.44. After the inspection in refueling outage 18R, an assessment will be made, appropriate actions will be taken, and the need for future inspections will be determined to ensure that the drywell integrity is maintained until at least April 2009. The scope of the inspection as set forth in reference 3.1.4.21 of inspecting two bays, is adequate because the environmental conditions and coating application methods were similar for all ten bays when the coating was applied. Also, the two bays selected for inspection are known to be worst leakage areas with most corrosion attack prior to the coating application. In summary, the inspection program (Reference 3.1.4.21) is adequate to assure drywell vessel integrity until at least April 9, 2009 (end of plant license). 3.3.1.2 Corrosion Rates Reference 3.1.4.40 discusses the statistical analysis of the UT data taken over the time period February, 1987 through September, 1994 for the sandbed region grids and November, 1987 through September, 1994 for the upper elevation grids. A new monitored location (#50-22) above the sandbed was added to the program in December of 1992. The corrosion rate was determined by calculating the rate of change of the mean thickness at each measured grid using linear regression 008/250 July 21, 1995 OCLROO001359

SAFETY EVALUATION CONTINUATION SHEET Page 55 of 69 SE-000243-002 Rev. No. 11 analysis. The corrosion rate has previously been expressed as the slope of the regression line +/- the standard error of the slope. Below are the current corrosion status assessments in the most limiting areas for each of the major elevations. The corrosion at the sandbed region was arrested in December, 1992 when the subject surfaces were cleaned and coated. Inspection of the coated surfaces performed in September of 1994 revealed that the coating is performing satisfactory as documented in reference 3.1.4.41. Sandbed Region - Corrosion arrested. Elevation 50'2" - F-Ratio <1.0 Elevation 5110" - F-Ratio <1.0 Elevation 87'5" - F-Ratio <1.0 Elevation 60'-11" - Insufficient Data Evaluation of the September, 1994 inspection data indicates that for Elevations 50'-2", 51"-10", 60'11", and 87'5", there is no evidence of ongoing corrosion. This assessment (Ref. 3.1.4.40) is based on the fact that the statistical regression estimate can not be used to define a corrosion rate because the F-ratio is far too low for reliable use, or that there are fewer than four measurements. (See paragraph 3.3.1.3--Sphere elevation 60'-11") Because the statistical F-test for significance of the regression rate estimate is very low, there is no evidence of ongoing corrosion, only random variation associated with measuring techniques. 3.3.1.3 Projections Projections are determined by performing regression analysis, when appropriate. Sandbed The entire sandbed region of the drywell shell O.D. was coated during the 14R refueling outage (December 1992). This coating was inspected in September 1994. This inspection showed no coating failure or signs of deterioration. Therefore, the corrosion in this region has been arrested and no further corrosion is expected to occur. To ensure that the coating applied will remain effective, visual inspections by direct and/or remote methods will be conducted per reference 3.1.4.21. The coating will again be inspected during refueling outage 16R and again during refueling outage 18R. Should an inspection reveal coating failure, an assessment will 008/250 July 21, 1995 OCLROO001360

SAFETY EVALUATION CONTINUATION SHEET Page 56 of 69 SE-000243-002 Rev. No. 11 be made, appropriate actions will be taken, and the need for additional inspections will be determined to ensure that the drywell integrity is maintained until at least April 2009 (end of License). The coating has an estimated life prediction of 8-10 years, before signs of local deterioration are expected (Reference 3.1.4.44). Currently, a margin of 70 mils exists between the required metal thickness and the actual mean metal thickness at the thinnest location as measured during the 15R outage in September 1994. This margin provides additional assurance for drywell integrity in the unlikely case of coating failure between inspection intervals. Based upon the arrested corrosion, and future monitoring of the coating, it is reasonable to conclude that this region will not become limiting prior to April 2009. Cylinder. Elevation 871-5" As a result of low F-ratio at this elevation, it can be concluded that there is no evidence of ongoing corrosion at this location. The September, 1994 data indicates that the thinnest location at this elevation has a mean thickness of 613 mils. Therefore, a margin of 161 mils exists between actual and minimum mean acceptable thickness. With the 161 mils margin which currently exists, minimum mean acceptable thickness could not be reached by April 2009, unless there was an ongoing corrosion rate of approximately 11 MYP. A corrosion rate of this magnitude would be observable. A corrosion rate of 11 MPY-has not been observed in any location above the sandbed. Additional assurance will be provided by volumetric inspection during the next refueling outage (16R) and at least every third refueling outage thereafter. Sphere, Elevation 50'-2" As a result of low F-ratio at this elevation, it can be concluded that there is no evidence of ongoing corrosion at this location. The September, 1994 data indicates that the thinnest location at this elevation has a mean thickness of 733 mils. Therefore a margin of 192 mils exists between actual and minimum mean acceptable thickness. 008/250 July 21, 1995 OCLROO001361

SAFETY EVALUATION CONTINUATION SHEET Page 57 of 69 SE-000243-002 Rev. No. 11 Although the data on hand does not permit a statistically rigorous calculation of corrosion rate, it is adequate to support a conclusion that this region will not become limiting prior to April 2009, unless there was an ongoing corrosion rate of approximately 13 MPY. A corrosion rate of this magnitude would be observable. A corrosion rate of 13 MPY has not been observed in any location above the sandbed. Additional assurance will be provided by volumetric inspection during the next refueling outage (16R) and at least every third refueling outage thereafter. Sphere. Elevation 51'-10" As a result of low F-ratio at this elevation, it can be concluded that there is no evidence of ongoing corrosion at this location. The September, 1994 data indicates that the thinnest location at this elevation has a mean thickness of 695 mile. Therefore a margin of 177 mile exists between actual and minimum mean acceptable thickness. Although the data on hand does not permit a statistically rigorous calculation of corrosion rate, it is adequate to support a conclusion that this region will not become limiting prior to April 2009. With the 177 mils margin which currently exists, minimum mean acceptable thickness could not be reached by April 2009, unless there was an ongoing corrosion rate approximately 12 MPY. A corrosion rate of this magnitude would be observable. A corrosion rate of 12 MPY has not been observed in any locations above the sandbed. Additional assurance will be provided by volumetric inspection during the next refueling outage (16R) and at least every third refueling outage thereafter. Sphere, Elevation 60'-11" This locatiohnwas added to the Drywell Corrosion monitoring program with the first UT data set taken in December 10 1992 and a second UT data set taken in September 1994. As a result of the limited data at this elevation, a statistical analysis of the corrosion rate, could not be performed. Therefore, a projection based on regression analysis will not be meaningful. The 008/250 July 21, 1995 OCLROO001362

SAFETY EVALUATION CONTINUATION SHEET Page 58 of 69 SE-000243-002 Rev. No. 11 September, 1994 data indicates that the thinnest location at this elevation has a mean thickness of 709 mils. Therefore, a margin of 191 mils exists between actual and minimum mean accepted thickness. Although the data on hand does not permit a statistically rigorous calculation of corrosion rate, it is adequate to support a conclusion that this region will-not become limiting prior to April 2009. With the 191 mils margin which currently exists, minimum mepn acceptable thickness could not be reached by April 2009, unless there was an ongoing rate of approximately 13 MPY. A corrosion rate of this magnitude would be observable. A corrosion rate of this magnitude has not been observed in any locations above the sandbed. Additional assurance will be provided by volumetric inspection during the next refueling outage (16R) and at least every third refueling outage thereafter. 3.3.1.4 Proiected local' Vessel Thicknesses Because mean uniform thickness can consist of local values less than the mean, consideration has been given to the significance of such readings. The number of such readings is: extremely limited and have been evaluated as not structurally significant as follows (Ref. 4.1.4.40) Sandbed The lowest local reading is .770 inches (Ref. 3.1.4.40). The local acceptable thickness for the sandbed region is .49 inches (Section 3.3.2). As mentioned in 3.3.1.3, the sandbed region was coated and no further corrosion is expected in this area, and the .280" margin is more than adequate for the balance of plant life (April 2009). Cylinder, Elevation 87'5" The lowest local reading is .551 inches (Ref. 3.1.4.40). The local acceptable thickness for this elevation is .300 inches (Section 3.3.2). Therefore, a margin of approximately 251 mils exists between actual and local acceptable thickness. If this local area is actually corroding, it would have to corrode at a rate of approximately 17.mile/year to reach the minimum local acceptable thickness by April 2009. A corrosion rate of approximately 17 mile/year has not been observed to date (above the sandbed) and is not considered credible. 008/250 July 21, 1995 OCLROO001363

Citizen's Exhibit NC9 Citizen's Exhibit NC9 Nuclear UTDR No. 1011 Revision No. 0 Budget Technical Data Report Activity No. Page I of I Project:s Department/Section ,&D/Mechanical Systems OYSTER CREEK Revision Date Document Titles EVALUATION OF FEBRUARY 1990 DRYWELL UT EXAMINATION DATA Originator Signature Date Approval(e) Signature Date Atmroval for External Distribution Date Does this TDR include recommendation(s)? _X_.es .._No If yes, TFWR/TR#_ see next Race

• Distribution Abstracts A. Baig Summary and Purpose
• F. P. Barbieri The purpose of this report is to document the pre-D. Bowman liminary evaluation of the February 1990 Drywall UT
• G. R. Capodanno Examination Data as well as document the possible B. D. Elam reasons for why corrosion has not significantly abated.

S. Giacobi L. C. Lanese Results of UT examination data obtained February 9, S. D. Leshnoff 1990 indicated that some locations of the drywell

    .7.Pelicone           vessel may be experiencing corrosion rates greater

• H. Robinson than recently projected.

P. Tamburro Conclusions Although a more detailed review is currently underway (to be documented by revision to References 7.6 and 7.8), this report is intended to document preliminary analysis which determined that the drywell would be serviceable up to the 13R outage. Based on a preliminary analysis of the February, 1990 data, this evaluation projects the most limiting drywell vessel region to be Bay 5 at the 51 foot elevation. The most conservative rates project that this area will not reach minimum thickness until the 13R outage scheduled in January 1991. (For Additional Space Use Side 2) This is a report of work conducted by an individual(s) for use by GPU Nuclear Corporation. Neither GPU Nuclear Corporation nor the authors of the report warrant that the report is complete or accurate. Nothing contained in the report establishes company policy or constitutes a commitment by GPU Nuclear Corporation.

• Abstract Only OCLROO001669

-Abstract Continuation TDR No. 1011 Revision No. 0.,, Recommendations:t

1. SE 000243-002 Rev. 3 needs to be revised to indicate the new corrosion rates and projections.

2. The use of actual material properties (CHTR) should be pursued for the 50'2" elevation.

3. The dryvell design pressure of the drywell should be lowered.

4. Operation of the Cathodic Protection system needs to be verified and corrected as necessary.

S. Means of abating-corrosion at the upper drywell elevations must be evaluated. NOTE: All recommendations are being performed through ongoing activities. la OCLROO001670

TDR 1011 Rev. 0 Page 2 of 18 TAOF Or CONTENTS

1.0 INTRODUCTION

3 1.1 Background Information 3 2.0 METHODS 4 3.0 RESULTS 5 3.1 Results of February 1990 UT Examination 5 3.2 UT Measuring Device 12 3.3 Existing Corrosion Mechanism 12 3.4 Review of Cathodic Protection System Operation Since Installation 14 3.5 Review of Safety Evaluation 000243-003, Rev. 3 15 4.0 EVALUATION 16 4.1 Evaluation Approach 16 4.2 Sand Bed Region 21 4.3 50'-2" Elevation 23 4.4 86 Foot Elevation 24

5.0 CONCLUSION

25 6.0 RECOMMENDATIONS

7.0 REFERENCES

27 OCLROO001671

                                                                                     \

TDR 1011 Rev. 0 Page 3 of 18 1.0 ZT0M 1.1 Bakckaround Information GPUN has established a drywell corrosion abatement and monitoring program. (References 7.1, 7.2, 7.3, 7.4, 7.5, 7.6 and 7.7.) This program includes: the installation and operation of the cathodic protection system in the sand bed region (3/89)1 reduction of water inleakage sources (10-12/88), mechanical agitating and draining water from the sand bed region (10-11/88), monitoring the most limiting areas (ongoing), and continued analysis of the situation (ongoing). The most limiting areas are listed in the table below: UT INSPECTION RIORITY ELEVATION AREA 1 11'-3" Eleven 6" x 6" grids in Bays 9, 11, 13, 15, 17, 19 and frame 17/19 1 50'-2" One 6" x 6" grid above Bays 5 2 87'-5" Three 6" x 6" grid above Bays 11 a 15 2 11'-3" Eight strips (i" x 6" reading I" apart) in Bays 1, 3, 5, 7, 9, 13 Priority 1 areas are inspected at each outage of opportunity but not more frequently than once every three (3) months. Priority 2 areas are inspected in an outage of opportunity if the previous set of data was taken eighteen months (18) or more before the outage. Review of UT data up to October 1988 (References 7.6 and 7.7) indicated that the most limiting area (sand bed bay 17D) would not corrode below the minimum thickness before June of 1992. The installation of cathodic protection and sand bed draining were intended to significantly abate corrosion and allow extension of the projected date. Interim data taken in September 1989 indicated that corrosion rates in the sand bed regions had been reduced. On February 9, 1990 UT examinations were performed on all Priority 1 locations. Results from this data suggests corrosion rates in some areas may be greater than projected in October 1988 and September 1989. This report documents the assumptions, methods, results of the preliminary analysis, and engineering judgement used to evaluate the corrosion rates in each region. I - OCLROO001672

TDR 1011 Rev. 0 Page 4 of 18 2.0 RMQDQLQgy In order to understand the results from the February 1990 data the following were evaluated and reviewed: 2.1 A preliminary review of the data was performed to determine the data's validity and calculate new conservative corrosion rates. 2.2 A review of the UT measuring device was performed, in addition to a review of the physical application of the device in the field. 2.3 A review of GPUN's understanding of the perceived corrosion mechanism was performed. 2.4 A review of the Cathodic Protection System operation since installation was conducted to identify any operational changes which may have affected the corrosion mechanism in the sand bed region. As part of this effort, a meeting was held with a cathodic protection expert, Mr. Ian Munroe of Corrosion Services, who designed the present system at oC. 2.5 A review of the existing Safety Evaluation (Reference 7.7) which justified continued operation through June 1992 was performed to determine if the conclusions of the SE were still valid. 3.0 RESULTS 3.1 Results of February 1990 UT Examination Although the February 1990 UT examination data is not completely understood, the data seems to be valid. To ensure a completely thorough and conservative approach, this data was used in estab-lishing new corrosion rates. 3.1.1 Mean Thickness Values Each priority 1 inspection location consists of an 6" x 6* area. Measurements were made using the template with 49 holes (7 x 7) laid out on a 60 x 60 grid with 1" between centers. A mean of all points in each grid was calculated. This approach is consistent with earlier mean thicknesses calculations as is documented in Reference 7.5. Table 1 presents the calculated mean thickness values derived from February 1990 and October 1988 examinations. OCLROO001673

TDR 1011 Rev. 0 Page 5 of 18 Mean Thickness Mean Thickness Area Bay asof 10/88 n of 29 Difference (milo) (mile) (milo) Protected IIA 908.6 680.4 -28.2 Sand Bed 1IC Top 3 916.6 978.4 - Regions Bottom 4 869.0 - 17D 864.8 839.1 -25.7 19A 837.S 807.8 -30.1 198 856.5 840.7 -15.8 19C 860.9 830.5 -30.4 17/19 Frame 981.7 994.4 - Unprotected 9D 1021.4 1010.0 -11.4 Sand Bed 13A 905.3 859.0 -46.3 Region 15D 1056.0 1057.3 - 17A Top 3 957.4 1120.2 Bottom 4 937.5 50'2" 5 750.0 739.6 -10.4 Elevation Notes After October 1988, Bays 12C and 17A were split into two regions (the top three rows and bottom four rows). This is because these bays showed regions which were corroding at different rates. The February 1990 data show these differences while the October 1988 data presents a mean for the entire grid. OCLROO001674

r r r r r r r r r r r r r r r ii TDR 1011 Rev. 0 Page 7 of 18 TABLE 2 - ESTIMATED CORROSION RATES - SAND BED REGION (1) (2) (3) (4) (5) (6) (7) (8) (9) CORROSION RATE CORROSION RATE CORROSION RATE CORROSION FEB. 1990 REQ. DATE DATE DATE DAY UP TO 10/88 FROM 6/89-2/90 FROM 10/88 - RATE TO THICKNESS MIN. WHICH WHICH WHICH (POST CP & (2/90 PRE-CP & 2/90 THICK. MINIMUM MINIMUM MINIMUM 920 DRAIN) POST H20 DRAIN) ALL DATA THICK IS THICK IS THICK IS (MP,) (NPY) (MP!) (NP?) (MILS) (MILS) RECHED REACHED REACHED (COL. 2) (COL. 3) (COL.4) (3) 12A NOT SIGNIFICANT -5.0 +19.5 -4.1 +/-6.3 -12.4 +/-3.0 880.4+/- 700 5/91 6/97 3/99 t128oltt-22,

                                                              + (3)__(3)                51 .. 8.11-11C TOP 3     INDETERMINABLE                -62.0 +/-3.S                -20.3 +/-15.2        -35.0 +8.5         978.4+       700    1/93         1/94        1/95

(-86,4, r-64.71 ._I _-5_ 11C BOTTOM)%4 INDETERMINABLE -18.3+30.4 -13.4 +10.0 -22.1 +5.3 869.0+/- 700 10-11/90 9/93 11/94 _1-210.21 -42.,ll M-2.. (3) (3) 17D -27.6 +/-6.1 -27.8 +6.6 -17.7_+/-4.3 -24.0 +/-2.4 839.1+/- 700 12/91 4/94 7/94 i-lu41. ()(3) -69.51 J-3025 -8.51(3) 19A -23.7 +4.3 -35.7 +7.0 -20.7 +5.96 -21.8 +1.8 807.8+ 700 5/91 9/92 1/94 1-32.51 1-79 (3)

                                                             ,1)                              .    -25.2_

(3) _-32. -) 19B -29.2 +0.5 -21.6 +11.7 -10.2 +5.6 -19.6 +/-2.1 840.7+ 700 6/91 12/95 7/95

                          -30J.l                     95.51                 (-26.'951}*         1-23.71                   -

1)(3) 19C -25.9 +4.1 -25.3 +8.6 -18 4+/-3.8 -23.9 +1.5 83O.5+/- 700 8/91 2/94 7/94 ___1-35.5-1 Z2.2.5i,-i9it (-26._ (3) 17/19 INDETERMINABLE -13.0 +/-0.9 - -2.8 +/-8.2 994.4+/- 700 2004 - 2000 _ -18.71 (1-26.71 9D INDETERMINABLE -69.0 +/-41.4 -11.1 +28.0 -16.4 +/-7.5 1010.0+/- 700 12/90 2/93 5/98 (-330.";. (-92,*1 (-34.01 I NOTE: 1) RATES IN PARENTHESIS REPRESENT MOST CONSERVATIVE RATES WHICH CAPTURES 95% CERTAINTY.

2) BAY 17D WAS THE MOST LIMITING BAY AFTER OCTOBER 1988 UT RESULTS

3) STATISTICAL REGRESSION MODELING MORE APPROPRIATE THAN MEAN MODEL.

ooI 0 012/071A.1 01 I-- 00 0 0) 4S. al

r r V- r: r,-V r- Vr-z rr r- r~ r-~ r r r r II TDR 1011 Rev. 0 Page 8 of 18 TABLE 2 - ESTIMATED CORROSION RATES - SAND BED REGION (1) (2) (3) (4) (5) (6) (7) (8) (9) CORROSION RATE CORROSION RATE CORROSION RATE CORROSION FEB. 1990 REQ. DATE BAY UP TO 10/88 DATE DATE FROM 6/89-2/90 FROM 10/88 - RATE TO THICKNESS MIN. WHICH WHICH (POST CP & WHICH (2/90 PRE-CP & 2/90 THICK. MINIMUM MINIMUM MINIMUM B20 DRAIN) POST H20 DRAIN) AL DATA (MILS) THICK is THICK is TRICK iS (NMP) REACHED REACHED REACHED (COL. 2) (COL. 3) (COL.4) (3) (3) 13A INDETERMINABLE -41.8 +/-15.4 -39.3 +/-6.0 -16.3 +4.8 859.0+/- 700 2/91 8/92 5/95 ______________ -1 1 i32_______ 1-56,91 I-27.61L. 15D NOT SIGNIFICANT -5.2 +/-3.2 1-25.4) 1

                                                                          --             1.54 +3.4 1

1057.7+/- 1120.2+/- 700 700 2002

                                                                                                                                 -
                                                                                                                                         -             2018 2006 17A TOP 3                INDETERMINABLE    +17.4 +7.6                   -            -10.9     4   1120.2+/-       700      12/95        -             2006

_ -6, 1-23,5 _ (3~) 173 BOTTOM 4 -44.3 +.Ol -18.1 +/-12.3

                                                                                       .,(-54.)       937.5t       700      12/94                      2/94 1-44.Mt NOTES   1) RATES IN PARENTHESIS REPRESENT MOST CONSERVATIV RATES WHICH CAPTURES 95% CERTAINTY.

2) BAY 17D WAS THE MOST LIMITING BAY AFTER OCTOBER 1988 UT RESULTS.

3) STATISTICAL REGRESSION MODELING MORE APPROPRIATE THAN MEAN MODEL.

012/071A.2

r : r [ Ir.. .. i rw

• r ... r ... r r .... r .... ... r : Ir r........ r:...

TDR 1011 Rev. 0 Page 9 of 18 TABLE 3- ESTIMATED CORROSION .RATES - UPPER ELEVATIONS (1) (2) (3) (4) (5) (6) (7) (8) (9) CORROSION RATE CORROSION RATE CORROSION RATE CORROSION FEB. 1990 REQ. DATE DATE DATE BAY UP TO 10/88 BASED ON. BASED ON RATE TO THICKNESS KIN. WHICH WHICH WHICH SECTION 3.3 STRAIGHT AVG. 2/90 THICK. MINIMUM MINIMUM MINIMUM 6/89 - 2/90 ALL DATA (MILS) THICK IS THICK Is THICK IS (MPY) REACHED REACHED REACHED (COL. 2) (COL. 3) (COL.4) 51' - 4.3 +0.03(3) 16 15 -3.6 +2.9 739.6+/- 725 1/91 2/91 6/91 (AM DATAI (-4.51 ___-9. 1 51' (9189 D.LETE. I N/A N/A N/A -5.6 +1.13) 739.6+/- 725 N/A N/A 7/91 _-2 .j ) 51' N/A 16 15 -3.6 +/-2.9 739.6+/- 671 5/94 7/94 6/96 ISISIG CKT81 J-.l (A S -OF 86' 9 NOT SIGNIFICAM USING 6/26/89) 16 N/A (-9.8) 619.1 591 3/91 N/A 1/92 NOTEt 1) RATES IN PARENTHESIS REPRESENT MOST CONSERVATIVE RATES WHICH CAPTURES 95% CERTAINTY.

3) STATISTICAL REGRESSION MODELING MORE APPROPRIATE THAN MEAN MODEL.

0ý 0 012/071A.3 0-0 0) 0

-'4

-4.

TDR 1011 Rev. 0 Page 10 of 18 In addition the NDE/ISI group at Oyster Creek performed an equip-ment functional check on the UT meter (D-meter) and probe used to record the data. Different D-meter and probe combinations were used on various thickness. The results were generally identical with variances of only several thousands on an inch. 3.3 Existino Corrosion Mechanism 3.3.1 Corrosion Mechanism in Sand Bed Recion Per Reference 7.2, the cause of the corrosion in the sand bed region is the result of water trapped in the sand bed. The water which may have leaked into the sand bed during construction and/or outages in 1980, 1983 and 1986 was contaminated with chlorides, sulfates and numerous other metal ions. Per Reference 7.2, a likely corrosion rate (based on plug samples, analysis of inleakage water, laboratory testing, and literature research of related phenomena) is 17 milu/year. However, to ensure conser-vatLem, Reference 7.8 arrived at a conservative rate assuming all material loss observed in 1986 had occurred in the six year period of water intrusion since 1980. The resulting rate -48 MPY was used to justify continued plant operation to June 1992. 3.3.2 Corrosion Mechanism in U=oer Elevation Per reference 7.3 the cause of the corrosion in the upper elevation was the result of the drywell steel exposed to the "firebar" insulation laden with chloride containing water. This was based on analysis of drywell vessel plug samples, analysis of inleakage water, laboratory testing, and literature research of related phenomena. Reference 7.3 concludes that the most conservative corrosion rate (based on plug samples, analysis of inleakage water, laboratory testing, and literature research of related phenomena) is 16 mils per year. 3.4 Review of Cathodic Protection System Operation Since Installation A review was performed on the Drywell Cathodic Protection System (CPS). This review included verification of the electrical installation and system operating parameters. According to the design documentation, the system is configured correctly. Review of system electrical potential data has shown that since the initial draining of water from the sand bed, generally there has been a steady reduction of current as a function of time. OCLROO001678

TDR 1011 Rev. 0 Page 11 of 18 The data indicates that since June of 1989, many of the cathodic protection system probes have experienced zero current. There are several possible reasons for this occurrence.

1) The sand bed could have become uniformly dry, including the sand in contact with the vessel wall. With the sand bed completely dry, the corrosion mechanism and subsequent rate were expected to halt.

2) Only the sand in areas close to and around the CPS probes has completely dried. The remaining sand bed region, including the sand in contact with the vessel wall, is still wet and the corrosion mechanism is still in place. The locally dry sand around the probes may be developing very high resistivity factors which have resulted in low and/or zero currents. Per discussions with Ian Munroe, of Corrosion Services, this is thought to be unlikely because the current density of the system is not high enough for this kind of phenomena.

3) The current provided initially is too low. Per discussing with Ian Munroe, of Corrosion Services, the electrical power supplied to the system may need to be increased. This may be required due to the grade positioning being different than the conceptual layout of grades.

3.S Review of Safety Evaluation 000243-002. Rev. 3 (Reference 7.71 3.5.1 Band. Bed ReaLon The above referenced Safety Evaluation projects Bay 17D in the sand bed region as the most limiting of all monitored locations. Mean thickness was expected to reach the minimum allowable mean thickness of .700 inches by June 1992. 3.5.2 Elevation 50-20 The above referenced Safety Evaluation projects mean thickness on EL. 50'-2" as .730 inch by June 1992 which is above the minimum mean thickness of .725 inches. Note that this value does not take credit for the actual material properties of the steel plate (CMTRs). Minimum allowable thickness using actual stress values from CMTRs is .671 inches (Ref. 7.4). OCLROO001679

TDR 1011 Rev. 0 Page 12 of 18 3.5.3 Elevation 87 Foot The above referenced Safety Evaluation does not project mean thickness on Elevation 86-'50 as no corrosion was ongoing at this elevation. However, the minimum allowable mean thickness at this elevation is .591. Note that this value is derived from actual material properties of the steel (CT[s). The minimum allowable thickness for localized areas at this elevation is .425 inches. 4 0 EVAMTO 4.1 Evaluation Aroach This evaluation documents and illustrates the preliminary approach used to estimate corrosion rates, identify the limiting bay and project the date at which minimum shell thickness is reached. The statistical appropriateness of these analyses is to be verified by revision to Reference 7.5. Reference 7.5 will be updated to provide statistically appropriate corrosion rates. 4.1.1 Sand Bed Region A logical approach based on an understanding of the corrosion phenomena, a vigorous application of statistics, and sound engineering judgement was necessary to develop appropriate conservative corrosion rates. Rates based on data from June 1989 to February 1990 were intended to capture a rate post cathodic protection installation and sand bed draining. These rates may have indicated the most recent changes in corrosion. However, these rates are based on only three observations (6/89, 9/89 and 2/90 data) which generally resulted in statistically inappropriate rates. Corrosion rates based on all data up to February 1990 would capture an overall rate and would statistically be Smore accurate (Table 2, Column 4). However, these rates may not capture possible recent increases in corrosion rates. Therefore, this approach may not be the most conservative. Rates were also calculated based on data from 10/88 to 2/90. Although these rates are based on only four obser-vations, the time period is almost doubled (compared to the 6/88 to 2/90 period). OCLROO001680

TDR 1011 Rev. 0 Page 13 of 18 Table 2 shows which of the rates are based on data which fit the regression model more appropriately than the mean model (indicated by Note #3). (This will be referred to as "statistical appropriateness" throughout this report.) However, the most "statistically appropriate" rate may not be the most conservative. Therefore, to take a consistently conservative approach, the greatest rate must be chosen, unless that value can be discounted (based on sound engineering judgement coupled with an understanding of the corrosion phenomena). The evaluation approach was to find the date in columns 7, 8 and 9 which would occur soonest in time. The rate used in projecting this date was then evaluated to see if it was based on a statistically appropriate curve fit and if the rate could be realistically expected (i.e. :S 60 MPY). If the rate was not realistic and not statistically appropriate, then it would be disregarded and the next date in time in column 7, 8 and 9 would be chosen. The date which occurs soonest in time is Bay 11C (bottom four rows) which projects a 10-11/90 date (in column 7). The corresponding corrosion rate is -18.3 + 30.4 (column 2). This suggests a standard error which is almost twice as much as the rate. As a result of this uncertainty, and the small number of observations, the 95% confidence rate is -210.4 MPY.. This type of corrosion rate is considered unrealistic (see Section 3.3). Therefore, this rate and the projected date based on this rate must be disregarded. For the next, Bay 9D, the column 2 rate is -69 + 41.4 MPY. This results in a 95% confidence rate of -330.0 MPY. This rate is considered unrealistic and is not based on a statistically appropriate model. Again, this rate and the projected date are disregarded. Bays IlA, lic (top 3 rows), 13A, 17D, 19A, 198 and 19C showed similar unrealistic results in column 2. In general, all column 2 results and projected dates (column 7) were not considered reasonable. 4.1.2 poer Elevations Table #3 presents 3 rows for Bay 5 at the 51 foot elevation. The first row presents an overall rate up to October 1988 (column 1), a rate based on section 3.3 (column 2), a rate based on straight line average from June 1989 to February 1990 (column 3), and an overall rate up to February 1990 (column 4). OCLROO001681

TDR 1011 Rev. 0 Page 14 of 18 Since it appears that a significant amount of material was lost from June 1989 to February 1990 (see Table f4) a straight average using mean thicknesses on these two dates was developed. Bay 5 Elevation 51 Mean Thickness Date of UT Mean Thickness 11/1/87 753.8 "7/12/88 750.0 10/8/88 750.2 6/26/89 749.6 9/13/89 7ss.6 2/9/90 739.6 The second row presents a rate with the September 1989 data disregarded. Review of the September 1989 mean thickness value shows an increase over the June 1989 mean thickness (by approximately 6 mils). This increase, coupled with'a resulting overall rate which is based on a curve fit which is not statistically appropriate, prompted an analysis of the data with the September 1989 observation deleted. The resulting rate of -5.6 + 1.6 is based on a curve fit which is statistically appropriate. Regardless, the more conservative of either resulting 95% confidence rate (with or without the September 1989 data) was chosen as the most conservative projection (-9.8 MPY). The third row for the 51 foot elevation presents the same rates as in the first, except a CMTR based minimum mean thickness is applied. Resulting projections are presented in column 7, 8 and 9. 4.2 Sand Bed Recion 4.2.1 Most Limiting Bay In The Sand Bed Reaion The October 1988 Safety Evaluation (Reference 7.11) projected Bay 17D (in the sand bed region) has the most limiting of all monitored locations. Based on a rate of

            -27.6 +/- 6.1 MPY and a 95% confidence conservative rate of
            -41 MPY, mean thickness was projected to reach the minimum allowable mean thickness of 0.700 inch by June 1992.

OCLROO001682

                                                                                \

TDR 1011 Rev. 0 Page 15 of 18 Results from February 1990 data now suggests that a conservative rate of -17.7 +/- 4.3 MPY and a 95% confidence conservative rate of -30.25 MPY can be applied, and that this bay is projected to reach a mean thickness of 700 mile by April of 1994. The February data now indicates that Bay 19A is the most limiting bay of all monitored locations in the sand bed region. Based on a new conservative rate of -20.7 +/- 5.6 MPY and 95% confidence rate of -38.1 MPY, it is projected that this bay may reach a mean thickness of 700 mile by September 1992. The conservative rate is both realistic and is based on-a statistically appropriate curve fit. Note, this rate is based on data recorded from October 1988 through February 1990 (column 4). 4.2.2 Protected Bave Interim data recorded in September 1989 indicated that corrosion rates in the protected sand bed region had generally decreased, yet the February 1990 data indicates that corrosion rates generally increased almost to former levels before cathodic protection installation. A possible explanation for this may be the reduced or zero probe current rates which has occurred since June 1989 (Section 3.4). Up to June 1989 the sand bed region may have been uniformly wet and Cathodic Protection System may have performed its intended purpose by inducing a current throughout the sand bed. Then in June the sand close to and around the probes may have completely dried with the remaining sand (including the sand in contact with the vessel wall) remaining wet. The locally dried sand around the probe may have developed very high resistivity factors resulting in very low and zero currents. The lack of impressed current prevents the cathodic protection system from performing it's function. This may explain the increased corrosion rates observed in February 1990. 4.3 50"-2" Elevation The most limiting bay at the 50 feet elevation is Bay 5. October 1988 data had resulted in a mean thickness of approximately .75 inches. October 1988 data indicated an on-going rate of -4.3 +/- .03 MPY. OCLROO001683

TDR 1011 Rev. 0 Page 16 of 18 February 1990 data indicates a loss of material resulting in a mean thickness of .7396 inches. Although the February 1990 data is not been thoroughly understood an overall rate of -3.6 + 2.9 MPY and a 95% confidence conservative rate of -9.8 MPY has been calculated. Based on this rate, it is projected that this area may reach a mini-mum mean thickness of .725 inches by June 1991. This thickness is based on code allowable stress values for the steel and not CHTR results. The minimum mean thickness at this elevation based on measured stress values (per vendor CXTHs) is .671 inch (Reference 7.7). Use of this minimum (instead of a minimum based on code allowable stress values) and the -9.8 MPY rate allow a projection for serviceability to June 1996. The more conservative rates of 16 and 15 MPY were also considered. The most limiting projection based on these rates (without COTR stress values) resulted in a January 1991 date. Use of CHTR stress values and resulting minimum mean thickness result in a May 1994 date. 4.4 86 Foot Elevation The most limiting bay at the 86 foot elevation is bay 9. June 1989 data indicates that this bay had a mean thickness of .6191 inches. As of June 1989 this bay was considered to be experiencing a rate of O.MPY. UT examination was not performed at this elevation in February 1990. Although it is very likely that this area is continuing to experience rates close to zero HPY, the conservative rate calculat-ed at the 51 foot elevation applied to the June 1989 mean thickness at Bay 9 on the 86 foot elevation projects that this bay may reach the minimum mean thickness of .591 inches by January of 1992. A more conservative rate of 16 mils/year based on the original safety evaluation (Section 3.3) was considered. Projection based on this rate resulted in a March 1991 date. If CHTR stress values are applied to the 51 foot elevation projection, then bay 9 on the 86 foot elevation becomes the most limiting bay with a serviceability date of March 1991. 5 *.0 ON=81UEION 5.1 Based on this evaluation, the sand bed region is no longer the limiting elevation for drywell vessel service. Bay 5 at the 51 foot elevation is now the most limiting. Based on February 1990 mean thickness of .7396 inches and a conservative rate of 16 UPY (Sec. 3.3), this area is projected to reach the minimum mean thickness of .725 inch by January 1991. This projection is based OCLROO001684

TDR 1011 Rev. 0 Page 17 of 18 on a theoretical rate of 16 MPY. The detailed review currently underway may determine a different projection which is based on a statistically derived rate from the data. However, this conservative projection does show that the drywell will be serviceable until January 1991. 5.2 Use of CKTR stress values applied to bay 5 at the 51 foot elevation projects this area to reach the minimum mean thickness of .671 inch by May 1994. 5.3 Although no data was taken in February 1990 at the 86 foot elevation and it is likely that corrosion rates remain at zero MPY, the conservative rate of 16 MPY (See. 3.3) projects bay 9 on the 86 foot elevation to reach the minimum mean thickness by March 1991. 5.4 February 1990 data now indicates that Bay 17D in the sand bed is no longer the most limiting bay. Results from the February 1990 data projects the most limiting bay in the sand bed is 19A. it is con-servatively projected that this area will reach the minimum mean thickness by September 1992. Based on these results in the sand bed region, it is concluded that cathodic protection Is currently producing very limited positive results in abating corrosion in the sand bed region. 6.0 RECOHN 6.1 Safety Evaluation 000243-002 Rev. 3 (Reference 7.6) which projects drywell service life up to June 1992 must be revised to reflect the new rate and a new date of January 1991. This is ongoing. 6.2 The minimum mean thickness at the 50'20 elevation is .725 inches. This value is based on code requirements. It is recommended that GPUN pursue using CMTR results to calculate a reduced minimum mean thickness value of .671 inches. This would result in projected serviceability date (at this elevation only) of June 1996. This is ongoing. 6.3 It is recommended that GPUN pursue lowering the design pressure of the drywell. This would further reduce the minimum mean thickness value in the upper elevation and provide more margin. This is ongoing. 6.4 Current cathodic protection system potential data indicates a postulated mechanism which may be defeating cathodic protection. The proper operation of this system needs to be verified and corrected as necessary. This is ongoing. 6.5 Evaluate methods for abating corrosion in the upper elevations. This is ongoing. OCLROO001685

TDR 1011 Rev. 0 Page 18 of 18 7.1 TDR 851 Assessment of Oyster Creek Drywall Shell. 7.2 TDR 854 Drywall Sand Bed Region Corrosion Assessment. 7.3 TDR 922 Drywall Upper Elevation, Wall Thinning Evaluation. 7.4 TDR 926 OC Drywall Structural Evaluations. 7.5 TDR 948, Statistical Analysis of Drywall Thickness Data. 7.6 Calculation C-1302-187-5360-006 Projection of Drywall Mean Thickness through October, 1992. 7.7 Safety Evaluation SE 000243-002, Rev. 3. 7.8 Safety Evaluation sE 000243-002, Rev. 1. OCLROO001686

Citizen's Exhibit NC 10 jr-

-                                                      Citizen's Exhibit NC1O aBr
                                      ~. AtQMULAR SPCT (GA~&D FITR Btu   ~        FROMA CL QIV" To ILLZ3'-0 WILDS  SEAMAeO-I                 251-Ce" ILLPX~F~

EL 23-Oa rA W.,TIAL C.ROSS SECTION OF DRrWELL q TORUS}}