ML032810438
| ML032810438 | |
| Person / Time | |
|---|---|
| Site: | Vermont Yankee  |
| Issue date: | 10/01/2003 |
| From: | Devincintis J Entergy Nuclear Vermont Yankee |
| To: | Document Control Desk, Office of Nuclear Reactor Regulation |
| References | |
| BVY 03-87 | |
| Download: ML032810438 (13) | |
Text
Entergy Nuclear Vermont Yankee, LLC E n tely gy Efor m r Entergy Nuclear Operations, Inc.
185 Old Ferry Road Brattleboro, VT 05302-0500 October 1, 2003 BVY 03-87 U.S. Nuclear Regulatory Commission ATTN: Document Control Desk Washington, DC 20555
Subject:
Vermont Yankee Nuclear Power Station License No. DPR-28 (Docket No. 50-271)
Supplement to Third-Interval Inservice Inspection (IS1) Proyram -
Submittal of Relief Request B-5, "Limited Examinations" Vermont Yankee Nuclear Power Station's (VY) third ten-year interval of the Inservice Inspection Program concluded on August 31, 2003. During this interval, the components identified within received less than "essentially 100%" examination. Accordingly, pursuant to IOCFR50.55a(g)(5)(iii), relief is requested on the basis that the required "essentially 100%" coverage examination is impractical due to physical obstructions and limitations imposed by design, geometry and materials of construction of the subject components.
Attachment I identifies the commitments contained within this letter. Attachment 2 contains Relief Request B-5.
If you have any questions on this transmittal, please contact Mr. Thomas B. Silko at (802) 258-4146.
Sincerely, DeVincentis Manager, Licensing Attachments cc: USNRC Region I Administrator USNRC Resident Inspector - VY USNRC Project Manager - VY Vermont Department of Public Service 1647
Docket No. 50-271 BVY 03-87 Attachment 1 Vermont Yankee Nuclear Power Station Supplement to Third-Interval Inservice Inspection (ISI) Program -
Submittal of Relief Request B-5, "Limited Examinations" List of Commitments
SUMMARY
OF VERMONT YANKEE COMMITMENTS BVY NO.: 03-87 The following table identifies commitments made In this document by Vermont Yankee. Any other actions discussed in the submittal represent Intended or planned actions by Vermont Yankee. They are described to the NRC for the NRC's information and are not regulatory commitments. Please notify the Licensing Manager of any questions regarding this document or any associated commitments.
COMMITMENT COMMITTED DATE
_ OR "OUTAGE" NONE I N/A
-t 4
4 4
4 4
4
Docket No. 50-271 BVY 03-87 Attachment 2 Vermont Yankee Nuclear Power Station Supplement to Third-Interval Inservice Inspection (ISI) Program -
Submittal of Relief Request B-5, "Limited Examinations" Relief Request B-5
LICENSEE/UTILITY NAME - Entergy Nuclear Operations, Inc.
PLANT NAME, UNIT - Vermont Yankee 10-YEAR INTERVAL - Third Interval REQUEST FOR RELIEF No. B-5 Proposed Alternative In Accordance with 10 CFR 50.55a~g)(5)(iii)
-- Inservice Inspection Impracticality -
- 1. ASME Code Component(s) Affected Code Classes: 1, 2,
3
References:
Subarticle IWB-2500, Subarticle IWC-2500, GL 88-10, NUREG 0313, Code Case N-460 Examination Categories: B-D, B-F, B-H, B-J, B-K, B-O, C-A, C-C, and C-F-2 Item Numbers: B3.90,'B3.100, B5.10, B8.I0, B9.1 1, B10.I0, B14.10, CIl 0, C3.10, and C5.51
Description:
Volumetric and Surface Examination Coverage Component Numbers: Various, see Table RR-B-5
2. Applicable Code Edition and Addenda
1986 Edition with no Addenda
- 3. Applicable Code Requirements Subarticle IWB-2500 states, in part, "Components shall be examined and tested as specified in Table IWB-2500-1." Table IWB-2500-1 requires a volumetric examination or a surface and volumetric examination be performed on the component based on Category and Item Number. The applicable examination area or volume and method required are as shown below from Table IWB-2500-1:
Examination Item Examination Requirements Examination Method Category Number /Figure Number B-D B3.90 & IWB-2500-7(b) Volumetric B3.100 B-F B5.10 IWB-2500-8(c) Volumetric and Surface B-H B8.10 IWB-2500-13 Surface B-J B9.11 IWB-2500-8(c) Volumetric and Surface B-K B110.10 IWB-2500-15 Surface B-Q B 14.10 IWB-2500-18 Volumetric or Surface BVY 03-87, Attachment 2, Page I of 9
Subarticle IWC-2500 states, in part, "Components shall be examined and pressure tested as specified in Table IWC-2500-1." Table IWC-2500-1 requires a surface examination or a surface and volumetric examination to be performed on the component based on Category and Item Number. The applicable examination area or volume and method required are as shown below from Table IWC-2500-1:
Examination Item Examination Requirements / Examination Method Category Number Figure Number C-A C 1.10 IWC-2500-1(a) Volumetric C-C C3.10 IWC-2500-5(a) and (b) Surface C-F-2 C5.51 IWC-2500-7(a) Surface & Volumetric
4. Reason for Request
Pursuant to 10 CFR 50.55a(g)(5)(iii), relief is requested on the basis that the required "essentially 100%" coverage examination is impractical due to physical obstructions and limitations imposed by design, geometry and materials of construction of the component.
Relief is requested from performing a complete coverage examination of the entire volume or area required. Entire volume or area required is defined by ASME Section Xl Code Case N-460 titled "Alternative Examination Coverage for Class 1 and Class 2 Welds,Section XI, Division 1." Code Case N460 states in part, "...when the entire examination volume or area cannot be examined.. .a reduction in examination coverage.. .may be accepted provided the reduction in coverage for that weld is less than 10%."
The NRC through, Information Notice 98-42 titled "Implementation of 10 CFR 50.55a(g) Inservice Inspection Requirements," termed the reduction in coverage of less than 10% to be "essentially 100 percent." Information Notice 98-42 states, in part, "The NRC has adopted and further refined the definition of 'essentially 100 percent' to mean 'greater than 90 percent'...has been applied to all examinations of welds or other areas required by ASME Section XI."
Relief is requested from performing an examination of "essentially 100%" of the required volume or area as applicable for the identified components in Table RR-B-5.
Note: Category B-A reactor pressure vessel shell welds (Item Nos. BI.l 1, Bl.12 and BI .30) are not listed in Table RR-B-5. A relief request was submitted for coverage limitations of this augmented examination of the reactor pressure vessel shell welds in a letter from Vermont Yankee to USNRC',
dated January 23, 1997. The NRC issued an SER 2 , dated February 18, 1999, related to that submittal.
- 5. Proposed Alternative The statistical basis provided below demonstrates that a reduction in coverage does not affect the ability to confidently detect potential degradation mechanisms. With an earlier design coupled with VYNPC Letter to USNRC, BVY 97-1S, "Augmented Examination of the Reactor Pressure Vessel Shell Welds," dated 2 January 23, 1997.
USNRC Letter to VYNPC, NVY 99-16, "Augmented Examination of the Reactor Pressure Vessel Shell Welds at Vermont Yankee Nuclear Power Station (Tac No. M99389)," dated February 18, 1999.
BVY 03-87, Attachment 2, Page 2 of 9
the examinations complete to the extent practical and results evidencing no unacceptable flaws present, the underlying objectives have been met. Additionally, a VT-2 examination performed on the subject components during system pressure test per examination category B-P each refueling outage and category C-H each period provides additional assurance that the structural integrity of the subject components is maintained.
Basis for Use Vermont Yankee obtained a Construction Permit on July 7, 1967. Vermont Yankee's piping systems and associated components were designed and fabricated before the examination requirements of ASME Section Xl were formalized and published. Since this plant was not specifically designed to meet the requirements of ASME Section XI, literal compliance is not feasible or practical within the limits of the current plant design.
Physical obstructions imposed by design, geometry and materials of construction are typical of vessel appurtenances, biological shield wall, insulation support rings, structural and component support members, adjacent component weldments in close proximity, unique component configurations (valves and pumps), and dissimilar metal weldments.
As a minimum, all components received the required examination(s) to the extent practical with regard to the limited or lack of access available. The examinations conducted confirmed satisfactory results evidencing no unacceptable flaws present, even though "essentially 100%" coverage was not attained. Vermont Yankee has concluded that if any active degradation mechanisms were to exist in the subject welds, those degradations would have been identified in the examinations performed. The basis for this conclusion derives from the statistical approach put forth in Appendix A of this request.
The statistical approach concludes that even for large reductions in coverage, the reduction in degradation detection confidence is insignificant.
For surface examinations, Vermont Yankee calculated the coverage percentage based on the area that was examined within the required coverage area divided by the required surface area to be inspected.
For volumetric (invariably ultrasonic) examinations, VY elected to use the following method to calculate coverage. The required examination volume was calculated. The examination was performed in accordance with an approved ultrasonic procedure that met the governing Code requirements. The approved procedure requires a number of angles and a number of beam directions for each angle. For each angle/beam direction combination the volume interrogated by that beam was calculated (within the required coverage volume). Then that value was divided by the required examination volume to determine a percentage of coverage for each angle/beam-direction combination. Then those required angle/beam-direction coverage percentages were averaged to determine an overall composite coverage. For example, prior to invoking Appendix VIII, ASME Section V, Article 4 required 0°, 450, and 600 search units for examining vessel welds from the OD of the vessel. The 450 and 600 search units are each required to be scanned in four orthogonal directions. Therefore, a total of nine angle/beam-directions are required and a coverage percentage is calculated for each of those nine angle/beam-direction combinations. Then those nine values are averaged to determine the overall composite coverage. (Note: Since Appendix VIII was invoked for vessel welds, the required number of angle/beam-direction combinations now depends on the qualified procedure, and thus the calculation would be different.) There are many other ways to calculate volumetric coverage, and it may be that using other calculation methods, more ultrasonic exams may have fallen above or below the 90% coverage threshold. The calculation method is of relative unimportance however, when one considers the statistical relevance (see Appendix A, "Statistical Significance of Limited Coverage").
BVY 03-87, Attachment 2, Page 3 of 9
Usually, a change in the NDE procedure will not make a dramatic change in the amount of coverage.
Also, typically, the original NDE procedure is optimized to perform the test with the greatest amount of confidence in the results. When a change is made to the procedure to achieve greater coverage, usually compromises in the NDE technique must be made. Further, since the implementation of Appendix VIII, it is not permissible to change ultrasonic techniques in order to attain more coverage because the UT technique would then not be qualified. It is more important to be confident in the results that are obtained, rather than to compromise the technique to obtain a little more coverage. As can be seen in Appendix A, "Statistical Significance of Limited Coverage," unless the coverage increase is quite dramatic, the impact is not really significant.
For the most part, Vermont Yankee did not select alternative welds when coverage was limited on the scheduled weld. A sample plan implies a certain amount of random choice in the selection of welds for examination - unless there are more conservative ways to select the sample, such as selecting high stress points or welds where industry experience indicates that damage mechanisms are more likely. This is why for Category C-F-2, terminal end welds are singled out; they are more typically high stressed. The reason for interferences is usually independent of the flaw mechanism. However, there may be cases where this is not true. For example, valve-to-pipe welds and pump-to-pipe weld geometries may inhibit coverage. But, these welds may actually have higher stresses because of their configurations. In these cases, if alternative welds were selected, the sample of higher stressed welds in the population would be diluted. If alternative welds are chosen, the selection randomness decreases. Flaw mechanisms associated with test limitations may be missed. It may be better to accept the limited coverage than to select alternative welds.
There is Code precedent for allowing limited coverage due to inaccessibility. ASME Section Xl allows certain Class I and Class 2 welds to be exempt based on the criteria that they are inaccessible.
Paragraphs IWB-1220(c), IWB-1220(d), and IWC-1223 exempt welds that are inaccessible due to control rod drive penetrations, because they are encased in concrete, are buried underground, or are encapsulated by guard pipe. The Code recognizes that examination of these welds is not possible and, therefore, that a Relief Request would not be necessary. The same logic applies to portions of welds that are inaccessible and where examination of those portions of welds is not possible.
The statistical study in Appendix A, "Statistical Significance of Limited Coverage," makes the interesting point that the Code and the industry implicitly accept a certain number of flaws in a population of welds by accepting one or more of the following: I) a sample inspection plan, 2) a reduction of up to I0% in coverage, and/or 3) a probability of detection of any one flaw of less than 100% (80% in the examples).
This is not surprising when one considers that the ISI program is spread over ten years. If no flaws were acceptable, I 00% of the safety related components in the plant would have to be inspected every day.
The impact that a reduction of coverage would have is not significant when one takes the above implicit assumptions into account.
Compliance with the proposed alternatives described above will provide an adequate level of quality and safety for examination of the affected welds, and will not adversely impact the health and safety of the public.
- 6. Duration of Proposed Alternative Relief is requested for the third ten-year interval of the Inservice Inspection Program for Vermont Yankee, which began September 1, 1993 and concluded August 31, 2003.
BVY 03-87, Attachment 2, Page 4 of 9
TABLE RR-B-5 COMPONENTS WITH LESS THAN "ESSENTIALLY ]00%" COVERAGE Section XI Component Exam Category System Component Condition Limiting Coverage &
& & Description Cvrg Item No. Number Percent B-D RPV Vessel-to-Nozzle Nozzle, radius blend, & weld configuration UT B3.90 N IA (Recirculation) 50.4%
B-D RPV Vessel-to-Nozzle Nozzle, radius blend, & weld configuration UT B3.90 NI B (Recirculation) 50.4%
B-D RPV Nozzle-to-Vessel Nozzle, radius blend, & weld configuration UT B3.90 N2A (Recirculation) 51.6%
B-D RPV Nozzle-to-Vessel Nozzle, radius blend, & weld configuration UT B3.90 N2B (Recirculation) 51.6%
B-D RPV Nozzle-to-Vessel Nozzle, radius blend, & weld configuration UT B3.90 N2C (Recirculation) 51.6%
B-D RPV Nozzle-to-Vessel Nozzle, radius blend, & weld configuration UT B3.90 N2D (Recirculation) 51.6%
B-D RPV Nozzle-to-Vessel Nozzle, radius blend, & weld configuration UT B3.90 N2E (Recirculation) 51.6%
B-D RPV Nozzle-to-Vessel Nozzle, radius blend, & weld configuration UT B3.90 N2F (Recirculation) 51.6%
B-D RPV Nozzle-to-Vessel Nozzle, radius blend, & weld configuration UT B3.90 N2G (Recirculation) 51.6%
B-D RPV Nozzle-to-Vessel Nozzle, radius blend, & weld configuration UT B3.90 N2H (Recirculation) 51.6%
B-D RPV Nozzle-to-Vessel Nozzle, radius blend, & weld configuration UT B3.90 N2J (Recirculation) 51.6%
B-D RPV Nozzle-to-Vessel Nozzle, radius blend, & weld configuration UT B3.90 N2K (Recirculation) 51.6%
B-D RPV Vessel-to-Nozzle Nozzle, radius blend, & weld configuration UT B3.90 N3A (Main Steam) 51.6%
B-D RPV Vessel-to-Nozzle Nozzle, radius blend, & weld configuration UT B3.90 N3B (Main Steam) 51.6%
B-D RPV Vessel-to-Nozzle Nozzle, radius blend, & weld configuration UT B3.90 N3C (Main Steam) 51.6%
B-D RPV Vessel-to-Nozzle Nozzle, radius blend, & weld configuration UT B3.90 N3D (Main Steam) 51.6%
B-D RPV Nozzle-to-Vessel Nozzle, radius blend, & weld configuration UT B3.90 N4A (Feed Water) 63.3%
B-D RPV Nozzle-to-Vessel Nozzle, radius blend, & weld configuration UT B3.90 N4B (Feed Water) 63.3%
B-D RPV Nozzle-to-Vessel Nozzle, radius blend, & weld configuration UT B3.90 N4C (Feed Water) 63.3%
B-D RPV Nozzle-to-Vessel Nozzle, radius blend, & weld configuration UT B3.90 N4D (Feed Water) 63.3%
B-D RPV Nozzle-to-Vessel Nozzle, radius blend, & weld configuration UT B3.90 N5A (Core Spray) 67.1%
B-D RPV Nozzle-to-Vessel Nozzle, radius blend, & weld configuration UT B3.90 N5B (Core Spray) __67.1%
BVY 03-87, Attachment 2, Page 5 of 9
TABLE RR-B-5 (Continued)
Section Xl Component Exam Category System Component Condition Limiting Coverage &
& & Description Cvrg Item No. Number Percent B-D RPV Nozzle-to-Vessel Nozzle, radius blend, & weld configuration UT B3.90 N6A (Head Spray) 52.5%
B-D RPV Nozzle-to-Vessel Nozzle, radius blend, & weld configuration UT B3.90 N6B (Head Spray) 52.5%
B-D RPV Vessel-to-Nozzle Nozzle, radius blend, & weld configuration UT B3.90 N7 (Head Vent) 49.1%
B-D RPV Vessel-to-Nozzle Nozzle, radius blend, & weld configuration UT B3.90 N8A (Jet Pump Instr.) 65.9%
B-D RPV Vessel-to-Nozzle Nozzle, radius blend, & weld configuration UT B3.90 N8B (Jet Pump Instr.) 65.9%
B-D RPV Vessel-to-Nozzle Vessel, radius blend, & weld configuration UT B3.90 N9 (CRD - Capped) 61.6%
B-D RPV Nozzle-to-Vessel Vessel, radius blend, & weld configuration UT B3.90 N1O (SLC) and support skirt 42.8%
B-D RPV Nozzle Inner Radius Support skirt UT B3.I00 NlO-IR (SLC 88.3%
B-F RPV Nozzle-to-Safe-end Flange bolting & surface depression UT B5.10 N6A-SE (Head Spray) 58%
B-F RPV Nozzle-to-Safe-end Flange bolting UT B5.10 N6B-SE (Head Spray) 89.9%
B-H RPV Bottom Head-to- OD of weld inaccessible due to biological PT B8.10 Support Skirt Support Skirt shield wall and insulation 50%
B-J Core Spray Valve-to-Pipe Valve configuration (one-sided exam) UT B9. 11 CS4B-MF5 38.6%
B-J RHR Pipe-to-Valve Valve configuration (one-sided exam) UT B9.11 RH28-12 82.5%
B-J RHR Valve-to-Elbow Valve configuration (one-sided exam) & UT B9.11 RH29- 10 branch connection obstruction 82.1%
B-J RHR Valve-to-Pipe Valve configuration (one-sided exam) & OD UT B9.11 RH32-8 surface configuration 73.8%
B-K Recirc 8 Lug Welds Pipe clamp on bottom of shear lugs PT B10.10 RR-89,90 77.8%
B-O CRD CRD Housing-to- Two I" diameter support rods UT B14.10 02-27HF Flange 85.2%
B-O CRD CRD Housing-to- Two I" diameter support rods UT B14.10 02-27SH Flange 85.2%
B-O CRD CRD Housing-to- Two I" diameter support rods UT B14.10 26-03HF Flange 85.2%
B-O CRD CRD Housing-to- Two I" diameter support rods UT B14.10 34-39HF Flange 85.2%
C-A RHR RHR Heat Exch. Flange geometry & 12 welded attachments UT C 1.10 A-HTEX 10-4 Shell-to-Flange 80.2%
C-C RHR RHR Heat Exch. Limited access between shell and floor MT C3.10 A-RHR-CC-4 Welded Support 80.1%
C-F-2 Condensate Pipe-to-Valve Valve configuration UT C5.51 CT27-S30 84.5%
BVY 03-87, Attachment 2, Page 6 of 9
TABLE RR-B-5 (Continued)
Section XI Component Exam Category System Component Condition Limiting Coverage Coverage
& & Description Cvrg Item No. Number Percent C-F-2 Feedwater Valve-to-Pipe Valve configuration (one-sided exam) UT C5.5I FW-17-S5 79.8%
C-F-2 RHR Valve-to-Pipe Weld OD profile configuration (limited UT C5.5 1 RH3D-S206 circumferential scan) 79.0%
C-F-2 RCIC Pipe-to-Valve Valve configuration UT Augmented RC3-S15 (0.280" thick) 80.5%
NUREG RWCU Pipe-to-Flange Flange configuration UT 0313 CU54-16 (Class 3) 57.3%
BVY 03-87, Attachment 2, Page 7 of 9
Appendix A Statistical Sienificance of Limited Coverage Certain Code Categories require a sample inspection. A typical population of Category B-J or C-F welds is several hundred welds. The sample size for B-J welds was 25% (prior to the implementation of Code Case N-560) and the sample size for C-F is 7.5%. The historical reasoning for choosing these sample sizes is not well known, however, it is presumed that if the numbers are based on statistical arguments, the percentages are not extremely rigid.
To understand the significance of limited coverage, several examples are presented with conservative assumptions. One initial assumption is that each weld is a separate data point. Another assumption is that if it is desired to find one flaw within a sample of welds, then a larger finite number of flaws must exist in the population to have a certain probability (akin to a level of confidence) of finding at least one flaw. This is implicit from the fact that the Code allows a sample inspection plan. For the sake of argument, assume that a level of confidence of 90% is required. (The following examples use the binomial probability distribution method of determining probabilities.)
A further assumption is that the probability of detecting (POD) any one flaw is 80%. This number is chosen based on Appendix Vill and IGSCC detection qualification acceptance criteria. In actuality, the Appendix VIII acceptance criteria vary between I00% for a small number of flaws (5 flaws) down to 70% for a large number of flaws (20 flaws). The effect of an 80% POD is that in order to assure a 90% chance of detecting at least one flaw in a sample, two flaws must exist in the sample such that there is a 96% (1 - 0.2 x 0.2) chance of detecting at least one of them.
For example, a 25% sample of 400 welds is 100 welds. There must be 20 flaws in the total population in order to assure that there is at least a 90% chance of detecting one of the flaws within the sample, using the binomial probability distribution method and allowing for the 80% POD example. Allowing for POD requires that the probability of two or more flaws occurring in the sample be multiplied by the probability of detecting at least one flaw, given that two exist, i.e. 0.96. The actual probability of detecting at least one of 20 flaws in this example is 92.4%.
What is the effect of limited coverage on several of the sample welds? A more conservative question would be -- what is the effect of not even inspecting several of the sample welds? In the above example, the probability of detecting at least one flaw would only drop a small amount to 90.6% (from 92.4%) if ten welds were dropped completely from the 100-weld sample. The reduced sample size would be 22.5%
(90 of 400) rather than 25%.
For a 7.5% sample of 400 welds (or 30 welds), there must be 60 flaws in the total population in order to assure that there is at least a 90% (actually 91.4%) chance of detecting one of the flaws within the sample, allowing for POD. If three of those 30 welds are not even inspected the probability of detecting at least one flaw only drops to 89.1% (from 91.4%). The respective reduced sample size (27 of 400) would be 6.75%.
Even if entire welds are dropped from the sample, the impact is within acceptable limits, as demonstrated in the above examples that the probability of detecting at least one flaw drops only a couple percent when a full 10% of the sample welds are not even inspected.
Actually, a weld examination is not a singular data point. Every scan increment (raster) should be considered a data point for examination of a weld. The portion of weld that is accessible for examination is typically representative of the full weld (the reason for interferences is usually independent of the flaw mechanism). Therefore, the sample of the weld becomes a sample of a larger population within the entire BVY 03-87, Attachment 2, Page 8 of 9
population of welds. The significance of foregoing a number of rasters in one or more welds is even less than foregoing entire welds.
With this viewpoint, even if there is only one weld in the population, a partial coverage examination becomes a sample of that weld. A vessel weld 10 feet long, which requires a 100% examination, may have a total of 240 rasters.
Current Code Case N460 states that a coverage of 90% is acceptable. This would equate to a 90% sample.
In the previous example, for a 90% sample of 240 rasters (216 rasters), there must be a total of six flaws in the total weld length in order to assure that there is at least a 90% (actually 93.4%) chance of one of the rasters detecting at least one flaw within the total weld length (also allowing for POD). If the total coverage drops to 75% (180 rasters), the probability of detecting at least one of those six flaws only drops to 90.3%
(from 93.4%).
The previous example assumes that the six flaws are spread randomly along the length and that each flaw is raster size or smaller. A critical flaw size might be long enough such that several scan rasters would impact that flaw. Also, a flaw might be situated such that it overlapped an area of scan interference. This would effectively raise the probability of detection for limited examinations.
Therefore, even for 100% sample inspection plans, a reduction in coverage of what one may initially think is quite severe, may not be statistically significant.
BVY 03-87. Attachment 2, Page 9 of 9