L-2002-239, Proposed License Amendments Request for Additional Information Response on Risk Informed One Time Increase in Integrated Leak Rate Test Surveillance Interval

Proposed License Amendments Request for Additional Information Response on Risk Informed One Time Increase in Integrated Leak Rate Test Surveillance Interval

ML023520056
Person / Time
Site:	Saint Lucie
Issue date:	12/13/2002
From:	Jernigan D Florida Power & Light Co
To:	Document Control Desk, Office of Nuclear Reactor Regulation
References
L-2002-239
Download: ML023520056 (28)

Text

& Florida Power & Light Company, 6501 South Ocean Drive, Jensen Beach, FL 34957 N PL F:PL December 13, 2002 L-2002-239 10 CFR 50.90 U. S. Nuclear Regulatory Commission Attn: Document Control Desk Washington, DC 20555 RE: St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 Proposed License Amendments Request for Additional Information Response on Risk-Informed One Time Increase in Integrated Leak Rate Test Surveillance Interval Pursuant to 10 CFR 50.90 on August 15, 2002, Florida Power & Light Company (FPL) submitted requested to amend Facility Operating Licenses DPR-67 and NPF-16 for St.

Lucie Units 1 and 2. The proposed amendments revise Unit 1 and Unit 2 Technical Specifications Section 6.8.4.h, Containment Leakage Rate Testing Program, to allow a one time 5-year extension to the current 10-year test interval for the containment integrated leak rate test (ILRT). St Lucie has implemented the 10 CFR 50, Appendix J, Option B performance-based containment leak rate test program.

The proposed changes were submitted on a risk-informed basis as described in Regulatory Guide (RG) 1.174, An Approach for Using ProbabilisticRisk Assessment in Risk-Informed Decisions on Plant-Specific Changes to the Licensing Basis. The proposed changes to extend the ILRT surveillance interval are justified based on a combination of risk informed analysis and assessment of the containment structural condition utilizing ILRT historical results and containment inspection programs. The risk aspects of the justification have been prepared by the Combustion Engineering Owners Group (CEOG) and are presented in a joint applications report (JAR), WCAP-15691, Joint Applications Report for Containment IntegratedLeak Rate Test Interval Extension, Revision 2, June 2002. Revision 2 of WCAP-15691 was submitted to the NRC for review by CEOG letter CEOG-02-125 dated June 14, 2002. A brief description and history of St. Lucie Unit 1 and Unit 2 ILRT testing results and the containment inspection program are discussed in the CEOG report with a more detailed description provided in this submittal.

During a conference call with the NRC on October 29, 2002, the NRC staff requested FPL to provide additional information in support of the NRC review of the proposed license amendments.

an FPL Group company

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Page 2 requested by the NRC on October 29, Attachment I provides the additional information 2002 for containment inspectable area. Attachment 2 provides the additional risk sensitivity. Attachment 3 provides information on the latent containment corrosion probabilistic risk assessment (PSA).

additional information on the quality of the St. Lucie submitted by FPL letter L-2002-143 The no significant hazard evaluation that was remains valid and unchanged.

of the proposed amendments is being In accordance with 10 CFR 50.91 (b)(1), a copy forwarded to the State Designee for the State of Florida.

requested by January 31, 2003 to Approval of these proposed license amendments is (SL2-14). Please issue the support the spring St. Lucie Unit 2 refueling outage and to be implemented within 60 amendments to be effective on the date of issuance Madden at 772-467-7155 if there are days of receipt by FPL. Please contact George any additio I questions about this submittal.

Ver uly ours,s D aid EE. Jernigan Vice Presi-i St. Lucie Plant DEJ/GRM Attachments cc: Mr. William A. Passetti, Florida Department of Health

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Page 3 STATE OF FLORIDA )

ss.

COUNTY OF ST. LUCIE )

Donald E. Jernigan being first duly sworn, deposes and says:

That he is Vice President, St. Lucie Plant, for the Nuclear Division of Florida Power &

Light Company, the Licensee herein; That he has executed the foregoing docu ent; that the statements made in this document are true and correct to the best his k wiedge, information, and belief, and that he is authorized to execute t on ehalf of said Licensee.

DontIdE A gan STATE OF FLORIDA COUNTY OF ST LUCIE Sworn to and subscribed before me this ( day of "c,-. t _, 2002 by Donald E. Jernigan, who is personally known to me.

Name of Notaryl 'Pic - State of Florida Leslie 3.Whitwell MY COMMISSION # DD020212 EXSIR$

May 1Z 2005 BONDEDTHRU TROYFAaNINSURA4NCE&

INC (Print, type or stamp Commissioned Name of Notary Public)

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 1 Page 1 ATTACHMENT 1 Background Information and Relative Inspectable Area of Containment

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 1 Page 2 ATTACHMENT I

Background

requested information During telephone conference on October 29, 2002 the NRC regarding the following:

area is large

1. In order to determine whether the leakage from the uninspectable area of the surface enough to be detected by ILRT, provide the total accessible percentage of the metal containment to be examined (surface area should be a entire surface area and not just what is accessible).

Anna,

2. Inspections of some reinforced and steel containment buildings (e.g., North from the Brunswick, D.C. Cook, and Oyster Creek) have indicated degradation containments.

uninspectable (embedded) side of the steel shell and liner of primary Units 1 and 2 containments are part The major uninspectable areas of the St. Lucie both sides areas on of the steel shell embedded in the basemat and the inaccessible to age related of the cylinder and dome. Address how potential leakage, due areas, is factored into the risk assessment in degradation from these uninspectable support of the requested ILRT interval extension from 10 to 15 years.

submittals, were

3. Issues with St. Lucie PSA quality, related to support of risk based that originally identified in relation to a separate risk based ISI submittal. Provided to it may not be necessary the aging analysis demonstrates acceptable results, submittal.

associate a response to these issues with the ILRT interval extension by one of However, the staff will require that the PSA quality questions be addressed these two submittals.

deals with the To be more specific, given the examples provided, the aging issue a leak path would potential for a corrosion mechanism to progress to the point at which to be based on the be created in the containment vessel. This in turn is expected that this failure accessibility of the containment surface to inspection and the likelihood could go undetected.

is addressed in The first question is addressed in this attachment. The second question 3 has been Attachment 2. The response to issues related to PSA quality in Question recently completed and is included as Attachment 3.

St. Lucie Containment Design leakage steel shell The containment vessel, including all its penetrations, is a low to confine the designed to withstand a postulated design basis accident (DBA) and of the reactor radioactive materials that could be released by accidental loss of integrity circular cylinder coolant pressure boundary. The containment vessel is a right (approximately 1 inch thick) and (approximately 2 inches thick) with hemispherical dome

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 1 Page 3 thick). It houses the reactor vessel, the ellipsoidal bottom (approximately 2 inches steam generators, the pressurizer and reactor coolant system piping and pumps, the connections of the reactor coolant the pressurizer quench tank, and other branch The containment vessel penetrations system including the safety injection tanks. a personnel air lock, and escape air include a construction hatch, a maintenance hatch, containment vessel is also equipped lock and various sized penetration nozzles. The a circular crane girder with a crane with a dome inspection walkway, access ladder and concrete shield building encloses rail attached to the shell of the vessel. The reinforced the containment vessel.

of the containment vessel An annular space is provided between the walls and domes operations and in-service and the shield building in order to permit construction during a loss of coolant accident inspection, and to filter any leakage from containment (LOCA) to minimize dose consequences.

with a net free volume The containment vessel is an independent freestanding structure is rigidly supported at its of approximately 2.5E6 cubic feet. The containment vessel base was placed after base near the elevation of its bottom spring line. The concrete and post weld heat the cylindrical shell and the ellipsoidal bottom were constructed are supported on a treated. Both the shield building and the containment vessel placed underneath and common foundation mat. With the exception of the concrete ties between the near the knuckles at the sides of the vessel, there are no structural slab. Therefore, there containment vessel and the shield building above the foundation vessel is virtually unlimited freedom for differential movement between the containment23 feet.

elevation and the shield building above the top of the concrete base at shell bottom, after the vessel was post Concrete floor fill is placed above the ellipsoidal weld heat treated, to anchor the vessel.

thickness of 1.92 The cylindrical portion of the steel containment shell has a minimum plates are welded inches on an inside radius of 70 feet. The polar crane girder support platform miscellaneous to the shell at approximately six feet on center. Except for some restraint or seismic framing and some minor seismic restraints, no major floor framing girder, a heating and supports are attached to the shell. Immediately below the crane five feet wide by five feet ventilating duct for the containment ring header, approximately supported at 30 deep and running the entire containment circumference, is structurally containment shell is places and attached to the shell by means of welded clips. The cranes.

also used to support temporary construction loads from the pedestal of four inches The 1.92-inch minimum shell plate thickness increases to a minimum dome hemispherical adjacent to all penetrations and openings. The inside radius of the portion of the is 70 feet with a dome plate 0.96 inches thick connected to the cylindrical containment spray shell at the tangent line by means of a full penetration weld. The piping is attached to the dome by means of welded clips as are the dome inspection external missiles by walkway and platforms. The containment vessel is protected from

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 1 Page 4 shield walls and other containment the shield building. The primary and secondary internal structures provide protection from internal missiles.

contact with the concrete on both The lower portion of the steel containment vessel is in small area at the upper interface sides below the 23-foot elevation level except for a acts to protect steel in contact where an expansion material is used. Since concrete in the lower section of the steel with it, there is little likelihood of corrosion occurring the areas of the containment vessel. During inspection of the expansion material, affected by corrosion. This most vessel at this interface were determined to be the area in the original FPL submittal L area has been evaluated on both units (see discussion with the St. Lucie ISI-IWE 2002-143) and continues to be inspected in accordance performed under the site inspection program plan and the evaluation requirements corrective action program.

Inspectable Area is exposed to permit visual Approximately 80 percent of the steel containment vessel inspection include the area inspection. The 20 percent that is inaccessible for visual transfer tube. The relative beneath the concrete floor and a small area around the fuel drawings, St. Lucie Plant surface areas are approximated using St. Lucie Plant UFSARs, and CRC Standard Mathematical Tables.

Accessible Area square feet Hemispherical dome area = 2nRh = 2n(70 feet)(70 feet) = 30,788 ethafoam at the concrete The cylinder conservatively includes the area protected by of cylinder and goes 3 feet metal interface (4-foot depth which starts 1 foot from bottom into ellipsoidal bottom head).

feet Cylinder area = 2nRh = 2-n(70 feet)(127+3 feet) = 57,177 square The total accessible area is 87,965 square feet.

Inaccessible Area of an oblate spheroid less The lower ellipsoidal head may be approximated by one-half included above.

the surface area of a cylinder of the height of the ethafoam Spheroid = 2ia 2 + 7(b 1F n(1+ E1- c) = 2n;(70) + n[(35)2/0.866)] In(l+ 0.866/1- 0.866) 2 2 a = major semiaxis = 70 feet b = minor semiaxis = 35 feet 2 2 c = eccentricity = (a2 - b )1" / a = 0.866 Spheroid/2 = area of lower elliptical shell = 21,246 square feet

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 1 Page 5 3

Ethafoam area = 2nRh = 2n(70)( ) = 1319 square feet Total inaccessible area is 19,927 square feet Percent accessible = [87,965/(87,965 + 19,927)] x 100 = 81.5%

This approximation is the same for both St. Lucie Unit 1 and Unit 2.

Industry Corrosion Events a through-wall Two corrosion events have occurred in the industry that have resulted in condition for the metal liner of reinforced concrete containments. There are no reported incidents in which the thicker free standing type steel containment vessel has exhibited a through-wall condition. The events pertaining to the metal liner are summarized below.

the On September 22, 1999, North Anna Unit 2 experienced through-wall corrosion of metal liner. The corrosion appeared to have initiated from a piece of lumber imbedded in the concrete behind the liner plate.

and On April 27, 1999, inspection at Brunswick 2 discovered two through-wall holes pitting in the drywell shell. The through-wall condition was believed to have originated from the coated (visible) side.

It should be noted that neither of these events is specifically applicable to the free standing containment design of St. Lucie Units I and 2. Unlike the previously considered containment structure, the St. Lucie metal vessel is surrounded by a shield building with an annular space between them which permits a general visual inspection of the containment exterior surface. The containment metal is substantially thicker (2 is inches thick as opposed to 1/4 inch thick) and the outer surface, which is also coated, normal operating conditions, the in a dry protected air space. In addition, under containment exterior surface is inherently warmer than its surrounding environment a viable preventing condensation on the surface and thus minimizing the potential for corrosion mechanism.

would Even presuming that a corrosion induced flaw were to breach the containment it The only not initially allow a path of sufficient size to produce a large early release.

of this conceivable corrosion mechanism which could lead to a through wall condition which results in a small leading edge type of containment is a pitting type corrosion be contained by the shield producing a small orifice. Leakage of this magnitude would building and filtered by the shield building ventilation system.

factor Therefore, it should be considered that the potential for large early release (LERF) is dependent upon containment pressurization propagating a small pre-existing

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment I Page 6 through-wall flaw or near minimum wall flaw produced by the aforementioned corrosion mechanism.

St. Lucie Inspection Program The St. Lucie containment vessel is examined in accordance with the requirements of ASME Code Section XI, Subsection IWE, the plant protective coatings program, and Technical Specifications. These inspection processes are described as follows.

The Containment Inservice Inspection Program at St. Lucie Units 1 and 2 is described in detail in ISI/IWE-PSL-1/2-PROGRAM, Metal Containment Inservice Inspection Program, which provides the rules and requirements. The specific areas and in components scheduled for inspection in accordance with the program are provided ISI/IWE-PSL-1 -PLAN, ASME Section X1, Subsection IWE Containment Building Metal Containment Inservice Inspection Plan for St Lucie Unit 1, and ISI/IWE-PSL-2-PLAN, ASME Section XI, Subsection IWE Containment Building Metal Containment Inservice of Inspection Plan for St Lucie Unit 2. The program requirements include inspection containment surfaces, pressure retaining welds, bolting, seals, gaskets, and moisture barriers using visual, surface, and volumetric techniques as required. Examinations that detect flaws or evidence of degradation shall be documented through the condition report process and evaluated in accordance with the requirements IWE-3000.

Personnel performing NDE are qualified and certified in accordance with IWA-2300 of the 1992 Edition with 1992 Addenda of ASME Section XI and implemented by CSI-QI 9.1, Qualification and Certification of Nondestructive ExaminationPersonnel.

The IWE program performs inspection of the entire accessible interior surface of the containment in each of 3 periods within a 10-year surveillance interval. The 100%

2001.

general surface area inspection for the first period on Unit 1 was completed April The 100% general surface area inspection was completed for the first period in April 2000 and second period in November 2001 on Unit 2. One-third of the moisture is barriers at the concrete floor to vessel interface on both sides of containment inspected during each period. To date, 2/3 of the Unit 1 moisture barrier and 1/3 of the Unit 2 moisture barrier have been inspected. Unit 2 will be 2/3 complete following the upcoming spring 2003 refueling outage (SL2-14).

Inspection results indicate that no significant corrosion effects have been experienced on the containment vessels. At the moisture barrier interface, there have been small areas of surface corrosion and minor pitting detected. However, it does not represent an issue considering the available design margin. A more detailed description was provided in the original proposed license amendments.

During activities that require repair of the containment vessel coatings, ASME Section XI, Subsection IWE, requires visual exams to assess the condition of the vessel metal surface for evidence of flaking, blistering, peeling, discoloration, and other signs of distress. Prior to any repair, NDE personnel perform an inspection to assess the

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 1 Page 7 condition of the base material. Following completion of coating repairs, a final inspection is performed by NDE personnel to determine acceptability of the final condition and to act as a reference for future inspections. Using the most recent Unit 1 outage as an example, approximately 25 of these inspections were performed. There has been no indication of containment vessel metal degradation on either unit resulting from these types of inspections.

The Protective Coatings Program at St. Lucie requires that a walkdown of the containment interior be performed each refueling outage by the FPL coatings specialist and engineering personnel to inspect any existing areas of non-qualified coatings and to determine any other areas in need of repair. Personnel familiar with the ASTM coatings standards, in accordance with plant procedures inspect the accessible exterior containment surface. Portions of the upper exterior containment vessel surface are not accessible for inspection due the unavailability of sufficient installed ladders or platforms and so the containment external surface above the floor is not inspected each outage.

Inspections of the upper exterior surfaces of both containments have been performed during previous outages. Inspection of the upper section of the exterior side of the containment vessel identified no degraded areas and no potential means by which corrosion would be promoted such as moisture sources or equipment interface. Those areas identified by inspection which do not meet acceptance criteria are evaluated and scheduled for repair as necessary. Following repairs, containment vessel coatings are re-examined upon completion by certified NDE examiners and the as-left condition documented. This allows identification of any potential for containment vessel degradation. As previously stated, there have been no indications of significant degradation of the containment vessel base metal.

General visual inspections of both sides of the accessible containment vessel surface and the shield building are performed as required by Technical Specifications in accordance with Quality Instruction QI 10-PR/PSL-5, Technical Specification Surveillance Inspection of Reactor Building. Results of these inspections have not revealed any additional conditions to that already noted other than minor concrete spalling of the shield building.

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 2 Page 1 Attachment 2 Liner Corrosion Risk Assessment

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 2 Page 2 Liner Corrosion Request:

Inspections of some reinforced and steel containments (e.g., North Anna, Brunswick, D.C. Cook, and Oyster Creek) have indicated degradation from the uninspectable (embedded) side of the steel shell and liner of primary containments. The major uninspectable areas of the St. Lucie Units I and 2 containments are part of the steel shell embedded in the basemat and the inaccessible areas on both sides of the cylinder and dome. Address how potential leakage, due to age related degradation from these uninspectable areas, are factored into the risk assessment in support of the requested ILRT interval extension from 10 to 15 years.

Response

The following approach was used to assess the change in large early release factor (LERF) as a result of undetected containment vessel corrosion. Previously evaluated intact sequences were evaluated against the likelihood of an undetected through-wall corrosion event and the result used to establish the potential increase in LERF. The following are issues factored into the analysis:

"* Differences between the concrete encased containment lower head and the exposed containment cylinder and dome.

"* Historical probability of corrosion producing a through-wall flaw without prior detection.

"* Aging impact on failure probability.

"* Leakage dependency on containment pressure.

"* Probability that visual inspections will be effective in detection.

It should be emphasized that this approach to estimating the additional potential for LERF due to corrosion is a conservative bounding exercise for the type of containment structure utilized for St. Lucie Units 1 and 2. Based on the lack of failure history, mechanical, and environmental differences which resist or preclude previously identified failure mechanisms, and the considerably greater time for discovery that exists due the order of magnitude thicker steel structure a more detailed analysis is likely to conclude that the risk component was not of significant magnitude to warrant augmenting existing risk parameters.

Assumptions

1. A half failure is conservatively assumed for the concrete concealed lower containment vessel due to the lack of any identified failures for this part of the structure. (See Table 1, Step 1.)

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 2 Page 3

2. The success data was limited to 5.5 years to reflect the years since September 1996 when 10 CFR 50.55a started requiring visual inspection. Additional time was not utilized to limit the aging impact of the corrosion issue even though inspection had been required and performed under other programs prior to this date and there is no evidence that liner corrosion issues were identified. (See Table 1, Step 1.)

3. The potential for corrosion induced flaw likelihood is assumed to double every 5.5 years. This is based on reasonable statistical judgement and is consistent with prior analysis accepted by the NRC. It is included to address the likelihood of corrosion as the liner ages. Sensitivity studies are included that addresses doubling this rate every 10 years and every 2 years. (See Table 1, Steps 2 and 3, and Tables 5 and 6.)

4. The likelihood of a breach in the steel vessel due to corrosion produced localized wall thinning or flaw is a function of the containment pressure. It should be considered that in the case of a free standing steel containment vessel with a nominal 2-inch thickness that significant deterioration would be detected. However, it may be conservatively considered that the potential exists for a localized corrosion mechanism to produce a small (less than LERF) through-wall or near through-wall flaw prior to detection. Failure at lower pressures is proportionately unlikely.

However, at the point of postulated containment failure, a through-wall breach would be statistically certain. Probability values were assumed as 0.1% at 20 psia and 100% at 110 psia (based on IPE level 2 containment failure pressure) with intermediate failure probabilities determined through logarithmic interpolation. Credit for the shield building is taken only in that a leak of magnitude that remains less than LERF is contained and does not contribute to the risk parameter. Sensitivity studies are included that decrease and increase the probability of containment failure at 20 psia anchor point by a factor of 10. (See Table 4 for sensitivity studies)

5. The probability of minimum wall flaw resulting in containment breach in the concrete enclosed lower containment vessel is considered to be 10 times less likely than the exposed containment cylinder and dome regions. (See Table 1, Step 4.)

6. A five percent visual inspection detection failure likelihood, given the flaw, is visible and a total detection failure likelihood of 10% is used. (See Table 1, Step 5.)

Sensitivity studies are included that evaluate total detection failure likelihood of 5 percent and 15 percent. (See Table 4 for sensitivity studies.)

7. The total CDF is included in determining the potential for large early releases. This approach avoids a detailed analysis of containment failure timing and operator recovery actions. The CDFs from internal events of 2.99E-5/Yr for Unit 1 and 2.44 E-5/Yr for Unit 2 are based on a conservative model. Although CDF from external events has not been calculated using a more detailed and realistic approach, it is estimated to be less than that of the internal events.

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 2 Page 4 Table 1-Liner Corrosion Base Case Step Description Containment Cylinder and Containment Basemat 20%

SDome80%

Historical Liner Flaw Likelihood Events:2 Events: 0 Failure data: Containment location (Brunswick & North Anna Assume half a failure specific 2)

Success data: Based on 70 steel-lined containments and 5.5 years since the 10 21(70*5.5)=5.2E-3 0.5/(70"5.5)=1.3E-3 CFR 50.55a requirement for periodic visual inspection of containment surfaces Aged Adjusted Liner Flaw Likelihood Year Failure Rate Year Failure Rate 2 During 15-year interval, assumed failure 1 2.1E-3 1 5.OE-4 rate doubles every 5 years (14.9% avg 5-10 5.2E-3 avg 5-10 1.45E-3 increase per year). The average for fifth to tenth year was set to the historical 15 1,4E-2 15 4.OE-3 failure rate (See Table 5 for an example) 15-year avg 6.45E-3 15-year avg 1.8E-3 3 Increase in Flaw Likelihood Between 3 and 15 years 8.97% 2.5%

Uses aged adjusted liner flaw likelihood (Step 2), assuming failure rate doubles every 5 years. (See Tables 5 and 6) 4 Likelihood of Breach in Containment Pressure Likelihood Pressure Likelihood of given Liner Flaw (psia) of Breach (psia) Breach The upper end pressure is consistent with St. Lucie Probabilistic Risk 20 0.1% 20 0.01%

Assessment (PSA) Level 2 analysis.

0.1% is assumed lower end. 58.7 (Design) 1.95% 58.7 (Design) 0.195%

Intermediate failure likelihoods are determined through logarithmic interpolation. The basemat is assumed 110 100% 110 10%

to be 1/10 of the cylinder/dome analysis 5 Visual Inspection Detection Failure 10% 100%

Likelihood 5% failure to identify that Cannot be visually inspected.

the flaw is not visible (not through-cylinder but could be detected by ILRT).

All events detected through visual inspection.

5% visible failure detection is a conservative assumption.

6 Likelihood of Non-detected 0.017% 0.0048%

Containment Leakage.

(Steps 3*4*5) 8.97%*1.95%*10% 2.5%*0.195%*100%

St. Lucie Units I and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 2 Page 5 The total likelihood of the corrosion-induced, non-detected containment leakage is the sum of Step 6 for the containment cylinder and dome and the containment basemat.

Total Likelihood of Non-Detected Containment Leakage = 0.017% + 0.0048% =

0.022%

The increase in LERF associated with the liner corrosion issue is estimated as:

Increase in LERF (ILRT 3 to 15 years) = 0.022%

• 2.99E-5 per year = 6.6E-9 per year for Unit 1.

Increase in LERF (ILRT 3 to 15 years) = 0.022%

• 2.44E-5 per year = 5.4E-9 per year for Unit 2.

Change in Risk The risk of extending the ILRT from 3 in 10 years to 1 in 15 years is evaluated by considering the following elements.

1. The risk associated with the failure of the containment due to a pre-existing containment breach at the time of core damage (Class 3 events).

2. The risk associated with liner corrosion that could result in an increased likelihood that containment over-pressurization events become LERF events.

3. The likelihood that improved visual inspections (frequency and quality) will be effective in discovering liner flaws that could lead to LERF.

These elements are presented in detail in the following discussion.

Pre-Existing Containment Breach The original submittal addressed Item 1. The submittal calculated the increase risk using a new CEOG methodology and a previously NRC-approved methodology. Table 2a and Table 2b summarize the risk increase associated with extending the Type A test from 3 in 10 years to I in 15 years for St. Lucie Unit 1 and St. Lucie Unit 2, respectively.

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 2 Page 6 Table 2a Risk Increase Associated with Pre-Existing Containment Breach, St. Lucie Unit I Percentage "Method LERFIncrease Person-REMlyr Increase in

'Increase Person-REMlyr CEOG Method 1.4E-8 10.26 1.14%

NRC Approved 9.4E-8 1.13 0.18%

Method I _ 1__

Table 2b-Risk Increase Associated with Pre-Existing Containment Breach, St. Lucie Unit 2 SMethod LERF Increase' Person-REM/yr .Percentage

-I lncrease 'Increase in

______________Person-REM/yr CEOG Method 1.OE-8 7.41 0.71%

NRC Approved 7.7E-8 0.93 0.11%

Method Liner Corrosion Table 3a and Table 3b summarize the risk increase with liner corrosion included for St.

Lucie Unit 1 and St. Lucie Unit 2, respectively.

Table 3a-Risk Increase Including Liner Corrosion Impact for St Lucie Unit 1

Method... :LERF Increase*, -Person-REM/yr Percentage

" Increase Ilncrease in

_____________ Person-REM/yr CEOG Method 1.4E-8 10.26 1.14%

CEOG Method with 2.1E-8 10.51 1.17%

Liner Corrosion NRC-Approved 9.4E-8 1.13 0.18%

Method NRC-Approved 1.1E-7 1.18 0.19%

Method with Liner Corrosion

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 2 Page 7 Table 3b-Risk Increase Including Liner Corrosion Impact for St Lucie Unit 2 Method LERF Increase -Person-REMlyr 'Percentage Increase increase in

______________ ; Person-REMlyr CEOG Method 9.6E-9 7.41 0.71%

CEOG Method with 1.5E-8 7.61 0.73%

Liner Corrosion NRC-Approved 7.7E-8 0.93 0.107%

Method NRC-Approved 8.25E-8 0.96 0.111%

Method with Liner Corrosion Visual Inspections The original submittal did not fully address the benefit of the Subsection IWE visual inspections. Visual inspections following the 1996 change in the ASME Code are believed to be more effective in detecting flaws. In addition, the flaws that are of concern for LERF are considerably larger than are those of concern for successfully passing the ILRT. Integrated leakage rate test failures have occurred even though visual inspections have been performed. However, the recorded ILRT flaw sizes for these failed tests are much smaller than that for LERF. Therefore, it is likely that future inspections would be effective in detecting the larger flaws associated with a LERF.

Impact of Improved Visual Inspections The raw data for both the CEOG method and previously approved NRC method is contained in NUREG-1493. This containment performance data is pre-1994. In 1996, the USNRC endorsed the use of Subsection IWE to ASME Section XI which provided detailed requirements for in service inspection of containment structures. Inspection and attendant requirements for examination, evaluation, repair, and replacement activities of the MC type containment, in accordance with 10CFR 50.55a, involves consideration of potential corrosion areas. Based on a more rigorous, structured inspection process, it should be considered that the detection of flaws after 1996 is more likely than that prior to inception of IWE requirements, and contributes to a more effective coatings program, further reducing the likelihood of corrosion induced failure.

Visual inspection improvements directly reduce the delta LERF increases as calculated in the CEOG method and NRC-approved method. The increased inspection frequency reduces the delta LERF as calculated by both the CEOG and NRC-approved methods.

Table 7 illustrates the benefit of visual inspection improvements on the delta LERF calculations.

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 2 Page 8 If the improved inspections (additional inspection, improved effectiveness, and larger flaw size) were 90% effective in detecting the flaws in the visible regions of the containment (5% for failure to detect and 5% for flaw not detectable [not-through-wall]),

then the increase ILRT LERF frequency could be reduced by 23.5%. See Tables 7a and 7b for additional sensitivity cases. This indicates significant margin exists for the estimated LERF increase.

Sensitivity Studies The following cases were developed to gain an understanding of the sensitivity of this analysis to the various key parameters. For this sensitivity study, the values performed for Unit 2 are used for Unit 1 as the containment liner contribution is the same for both units.

Table 4-Liner Corrosion Sensitivity Cases

.Age',, -.,Containment Visual Inspection Likelihood Flaw LERF.Increase (Step 2) Breach ' & Non-Visual Is LERF

___________ (Step 4) 'Flaws (Step 5) _, ___

Base Case Base Case Base Case Base Case Base Case Doubles every 5 years 1.95/0.19 10% 100% 5E-9 Doubles every 2 Base Base Base 6E-8 years Doubles every 10 Base Base Base 2E-9 years Base Base point 10 Base Base 1E-9 times lower (0.52/0.05)

Base Base point 10 Base Base 2E-8 times higher (7.2/0.72)

Base Base 5% Base 3E-9 Base Base 15% Base 7E-9 Lower Bound Doubles every 10 Base point 10 5% 10% 4E-1 1 years times lower (0.524/0.05)

Upper Bound Double every 2 Base point 10 15% 100% 3E-7 years times higher (7.2/0.72)

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 2 Page 9 Table 5-Flaw Failure Rate as a Function of Time Year Containment. Basemat 0 1.79E-03 5.00E-04 1 2.06E-03 5.74E-04 2 2.36E-03 6.60E-04 3 2.71E-03 7.58E-04 4 3.12E-03 8.71E-04 5 3.58E-03 1.OOE-03 6 4.11E-03 1.15E-03 7 4.72E-03 1.32E-03 8 5.43E-03 1.52E-03 9 6.23E-03 1.74E-03 10 7.16E-03 2.OOE-03 11 8.22E-03 2.30E-03 12 9.45E-03 2.64E-03 13 1.09E-02 3.03E-03 14 1.25E-02 3.48E-03 15 1.43E-02 4.00E-03 15-year average 6.45E-03 1.80E-03 Delta I in 3 to 1 in 15 8.97E-02 2.50E-02 Table 6-Cumulative Failure Probability

- Years ,Containment-," '. ,,...,..Basemat' 1 to 3 0.71% 0.20%

1 to 10 4.15% 1.16%

1 to 15 9.68% 2.70%

A = 9.68% - 0.71% = 8.97% (delta between 1 in 3 years to 1 in 15 years), for containment A = 2.7% - 0.2% = 2.5% (delta between 1 in 3 years to 1 in 15 years), for basemat

St. Lucie Units I and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 2 Page 10 Table 7a-Benefit of Visual Inspection Improvements for St. Lucie Unit 1 Factor Improvement Reduction in NRC - NRC Approved CEOG CEOG Method "Dueto Visual Delta LERF Approved Methodw/Liner Method wlLiner Inspections Method Corrosion Delta LERF Corrosion Delta LERF Considered Delta Considered LERF Delta LERF Pre-1996 Inspection 0% 9.40E-08 9.95E-08 1.40E-08 1.95E-08 Approach (Base Case)

Post-1 996 with Visual 80% 1.88E-08 1.99E-08 2.80E-09 3.89E-09 Inspections Perfectly Accurate Post-1996 with Visual 76.00% 2.26E-08 2.39E-08 3.36E-09 4.67E-09 Inspections 95%

Accurate Post-1996 with Visual 72.00% 2.63E-08 2.78E-08 3.92E-09 5.45E-09 Inspections 95%

Accurate and 5%

Chance of Undetectable Leakage Post-1996 with Visual 60.00% 3.76E-08 3.98E-08 5.60E-09 7.78E-09 Inspections 80%

Accurate and a 5%

Chance of Undetectable Leakage

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 2 Page 11 Table 7b-Benefit of Visual Inspection Improvements for St. Lucie Unit 2 Factor Improvement 'Reduction -NRC NRC Approved ,CEOG CEOG Method due to Visual in Delta Approved Method wlLiner Method Delta w/Liner Inspections LERF Method - . Corrosion LERF Corrosion Delta'LERF Considered Delta Considered LERF Delta LERF Pre-1996 Inspection 0% 7.70E-08 8.25E-08 1.OOE-08 1.55E-08 Approach (Base Case)

Post-1996 with Visual 80% 1.54E-08 1.65E-08 2.OOE-09 3.09E-09 Inspections Perfectly Accurate Post-1996 with Visual 76.00% 1.85E-08 1.98E-08 2.40E-09 3.71E-09 Inspections 95%

Accurate Post-1996 with Visual 72.00% 2.16E-08 2.31E-08 2.80E-09 4.33E-09 Inspections 95%

Accurate and 5%

Chance of Undetectable Leakage Post-1996 with Visual 60.00% 3.08E-08 3.30E-08 4.00E-09 6.18E-09 Inspections 80%

Accurate and a 5%

Chance of Undetectable Leakage It is noted that the CDF used in the current analysis is conservative. In addition, the large early release used is also conservative. Based on a more recent review of the degraded core phenomena modeling, the LERF for the St. Lucie large dry containment is on the order of 0.01. The LERF values used in this license submittal are approximately an order of magnitude higher (see Table 2, Class 3a and 3b of FPL letter L-2002-1431: 0.085, the sum of 0.064 and 0.021 was used). If the CDF of external events is assumed to be as high as that of the internal events, the increase in the risk may be bounded by doubling the calculated values. The total risk increase (including the external events and the conservatism in both the CDF and LERF modeling) associated with the one time extension of the ILRT interval from 3 in 10 years to 1 in 15 years is approximately a factor of 5 lower than that estimated (a factor of 10 reduction from conservatism of LERF and CDF divided by a factor of 2 from the external event assumption).

Conclusion Considering increased frequency of visual inspections and the benefit of improved visual inspections post-1996, the increase in risk is considered to be less than 1 E-7 for 1 L-2002-143, St. Lucie Units I and 2, Docket Nos. 50-335 and 50-389 Proposed License Amendments: Risk Informed One Time Increase in Integrated Leak Rate Test Surveillance Interval

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 2 Page 12 LERF. Changes less than 1 E-7 are considered small per Regulatory Guide 1.174. The one-time extension of the ILRT interval from 3 in 10 years to 1 in 15 years is considered an acceptable risk increase.

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 3 Page 1 Attachment 3 PSA Quality Related Issues

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 3 Page 2 Note: No formal written question was issued. However, the PSA quality in general similar to that raised for RI-ISI was discussed during the conference call on October 29, 2002. The same questions related to the PSA quality for RI-ISI are summarized below. It is concluded that the PSA model used for the St. Lucie ILRT is robust with respect to the ILRT application. The weaknesses stated in the SE are briefly outlined below.

PSA Quality Related Request I Identify the version of the ProbabilisticSafety Assessment (PSA) model that was used for the R/-ILRT application and when it was last updated. Include when, and which version, of your PSA has been peer reviewed by the Combustion Engineering Owner's Group.

Response 1:

The version of the Level 1 model used for input to the RI-ILRT submittal is dated March 2001. The version of the Level 2 update is dated May 2001.

The St. Lucie CEOG peer review was conducted the week of May 20, 2002. The model reviewed by the peer review team was the draft version of a 2002 update. The latest Level 2 update, dated May 2001, validates that the one percent early containment failure assumption used for the LERF calculations is bounding.

PSA Quality Related Request 2:

The staff evaluation (SE) report on the St. Lucie Individual Plant Examination (IPE),

dated July 21, 1997, concluded that the IPE met the intent of GL 88-20. The SE also stated that "the staff identified weaknesses in the front-end, HRA and back-end portions of the IPE which, we believe, limit its future usefulness." The weaknesses stated in the SE are briefly outlined below. Explain how each of the weakness has been removed by modifications to the PRA or otherwise addressedduring the Ri-ISI evaluation.

Some initiating -eventfrequencies appeared low and some initiating event frequencies which relied on generic values should have received a plant-specific analysis.

Response 2:

Data update has been performed since the IPE. The data update included re quantification of the LOCA initiating event (IE) frequencies based on a CEOG technical position paper. Initiating event fault trees were also developed for loss of component cooling water (CCW), loss of intake cooling water (ICW), loss of turbine cooling water (TCW), loss of DC bus, and loss of instrument air. Plant specific data was used for other initiating events where available.

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 3 Page 3 There is no significant difference between the total CDF due to the change of the initiating event frequencies. It is judged that this IE data issue does not have a significant impact on the results and conclusions for the ILRT application.

PSA Quality Related Request 3:

Some pre-initiator human actions appeared in dominant accident sequences, an unexpected and uncommon result. It appears that a more detailed analysis of pre initiatorhuman actions may appropriatelyreduce the human error probabilities (HEPs) for these events, thus reducing the likelihood that excessively conservative HEPs may distort the risk profile.

Response 3:

Screening values have been used in all updates to date. It is judged that the use of unrefined pre-initiator screening values is conservative for the ILRT application, as a more refined HRA may reduce HEP and thus lower the CDF.

PSA Quality Related Request 4:

It was not clear what basis was used to determine which post-initiatorhuman actions were quantified with a time-independent technique and those post-initiatoractions that were quantified with a time-dependent technique. Three post-initiator human actions (initiatingonce-through cooling, manually initiating recirculation actuation components following loss of the automatic signal, and securing the reactor coolant pumps after loss of seal cooling) are relatively short time frame events. Failure to consider time in these events might lead to unrealisticvalues.

Response 4:

No changes to the HRA analysis to address this issue have been implemented for the PSA updates to date. The St. Lucie IPE SER states that "the HEPs for the events modeled as slips were not unreasonable and several of the events modeled in this way still show up as being important. Therefore, there is no reason to believe that the approach necessarily precluded detection of HRA related vulnerabilities."

A sensitivity study was performed on these actions. The results indicate that this HRA issue does not have a significant impact on the results and conclusions for the ILRT application.

PSA Quality Related Request 5:

The time-dependent human actions used likelihood indices at theirdefault values.

Therefore, the resulting human error probabilities may be generic rather than plant specific.

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 3 Page 4 Response 5:

No changes to the HRA to address this issue have been implemented for the PSA updates to date. The St. Lucie IPE SER states that in general, the way in which the SAIC time-dependent method was applied in the IPE did not appear to violate its basic tenets and that resulting HEPs would not be considered unusual. The SER also states that most of the HEP values themselves would not suggest that identification of human action vulnerabilities was precluded.

A sensitivity study was performed using updated HEPs for events previously quantified as time-independent. The methodology used to calculate the revised HEPs addresses plant specific factors.

The process for evaluating individual human interactions breaks down the detection, diagnosis, and decision-making aspects into different failure mechanisms, with causes of failure delineated for each. Eight different potential failure mechanisms are identified:

• Availability of information

• Failure of attention

• Misread/miscommunicate data

• Information misleading

• Skip a step in procedure

• Misinterpret instruction

• Misinterpret decision logic

• Deliberate violation A relatively simple decision tree is used for each of these mechanisms. Each of these decision trees identifies performance shaping factors that could cause the relevant mechanism to lead to failure to initiate the proper action. The analyst selects branch points in the decision trees that correspond to the aspects of the interaction being analyzed (e.g., the number and quality of cues for the operators, the ease of use of the procedures, etc.). For each outcome in the decision trees, there is a nominal probability of failure.

Depending on the failure cause, certain recovery mechanisms may come into play. The potential for recovery may arise as follows:

"* due to self-review by the operator initially responsible for the misdiagnosis or error in decision-making, as additional cues become available or additional procedural steps provide opportunity to review actions that have been taken and the resulting effects on the plant;

"* as a result of review by other crew members who would be in a position to recognize the lack of proper response;

"* by the STA, whose review might identify errors in the response;

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 3 Page 5

"* by the technical support center (TSC) when it is staffed and actively involved in reviewing the situation; and

"* by oncoming crewmembers when there is a shift turnover (when the time window is very long).

Thus, after processing each of the decision trees to arrive at estimates for the basic failure mechanisms, the analyst must identify and characterize the appropriate recovery factors.

There are other considerations besides time that affect the treatment of the non recovery potential. These included the degree to which new or repeated cues and recurring procedural steps would give rise to considering the action that had not been successfully taken.

Another element represents failure to implement the action correctly, given that the decision is made to initiate the action. A basic task analysis is performed to identify the essential steps that must be accomplished to implement a decision. The corresponding failures to perform them properly are noted. These failures are then quantified.

In considering the execution errors, three levels of stress were identified: optimal, moderately high, and extremely high. Optimal stress would apply for actions that are part of a normal response to a reactor trip, and for which the operators would be alert.

Moderately high stress would apply when the operators are responding to unusual events, including multiple failures. Extremely high stress would apply for scenarios in which there is a significant threat, such as the potential that core damage is imminent if the actions are not successful, or when actions must be accomplished under significantly less than optimal conditions.

The execution errors may be subject to review and recovery as well. This is particularly true for actions taken in the control room, where additional observers may be able to identify the need for corrective action. As in the case of the initiation errors, a set of guidelines for considering review and recovery by other crewmembers has been developed.

Based on the discussion above, it can be seen that the revised HEPs used in the sensitivity study takes into account plant specific factors.

It is judged that this HRA issue does not have a significant impact on the results and conclusions for the ILRT application.

PSA Quality Related Request 6:

An additional sensitivity analysis should have been performed regarding the probability of in-vessel recovery since the licensee assumed a very high probability of in-vessel recovery due to ex-vessel cooling.

St. Lucie Units 1 and 2 Docket Nos. 50-335 and 50-389 L-2002-239 Attachment 3 Page 6 Response 6:

It is recognized that there are variations in the probability of in-vessel recovery. For the ILRT applications, other conservatism embedded in the Level 2 model with respect to other dominant early containment failure mechanisms (e.g., direct containment heating, steam explosion, and the vessel acting like a rocket) outweigh this issue. The revised Level 2 analysis incorporating the insights after the IPE submittal was made indicates that the large early containment failure probability, assuming 25% of ex-vessel cooling, is less than 1%.

In conclusion, there is conservatism in the PSA model used for the ILRT applications. In addition, there is significant margin between the calculated risk increase and acceptance criteria. The PSA quality is thus judged to be sufficient and PSA model robust for the ILRT application.