ML20210S442

Whipjet Program Final Rept

ML20210S442
Person / Time
Site:	Beaver Valley
Issue date:	01/30/1987
From:	ELECTRIC POWER RESEARCH INSTITUTE, ROBERT L. CLOUD ASSOCIATES, INC., STONE & WEBSTER ENGINEERING CORP.
To:
Shared Package
ML19292G736	List: ML20210S409 ML20210S442
References
NUDOCS 8702170635
Download: ML20210S442 (174)

Text

{{#Wiki_filter:. e. BEAVER VALLEY POWER STATION ,y UNIT NUMBER 2 O .O WHIPJET PROGRAM FINAL REPORT O JANUARY 30, 1987

O l

'O Prepared for Duquesne Light Company by: ,0 stone and Webster Engineering Corporation Robert L. Cloud Associates, Incorporated l Electric Power Research Institute O I 8702170635 870202 ADOCK0500g2 O PDR A

O TABLE OF CONTENTS SECTION PAGE C) 1. Introduction 1-1 2. Summary of Results 2-1 3. Program Scope 3-1 3.1 High energy systems C). within WHIPJET 3-1 3.2 System descriptions 3-6 4. Screening Considerations 4-1 4.1 Anomalous conditions 4-1 4.2 Fluid transi'ents 0 4-4 4.3 Stress corrosion cracking 4-7 5. Crack Propagation Analysis 5-1 5.1 Stress determination 5-1 5.2 Crack growth law 5-9 5.3 Stress intensity factor calculation 5-10 0 5.4 Initial crack size 5-10 5.5 Acceptance criteria for fatigue crack growth 5-11 5.6 Example 5-12 5.7 Fatigue crack growth results 5-16 0 6. Material Property Data 6-1 6.1 Austenitic stainless steel lines 6-1 6.2 Highest stresses and minimum material properties 6-2 7. Leak Detection 0 7-1 7.1 Introduction 7-1 7.2 Leak detection systems (inside containment) 7-2 7.3 Leakage diagnosis 7-3 7.4 Actions 7-4 C) 8. Leak Rate Calculation Summary 8-1 i 9. Crack Stability Calculations 9-1 [For Normal + SSE Loads] l 10. Crack Stability Under 10-1 .O Excessively High Loads O i O

1 D. Table of Contents (continued) APPENDICES A. NUREG-0582 Fluid Transients A-1 B. Stainless Steel Material Properties B-1 ) C. Ferritic Steel Material Properties (DELETED) C-1 D. Welding Procedure D-1 E. Leak Rate Calculation Methodology E-1 ) F. BVPS-2 Leak Rate Curves F-1 G. FLET Verification Calculations G-1 D D D D D e 9 ii

C; TABLES No. Title PAGE G 3.1 Piping Eliminated from WHIPJET Scope by Screening or Other Considerations 3-8 3.2 Final Piping Systems for WHIPJET Analysis 3-9 3.3 WHIPJET Program Piping Data 3-10 0 3.4 WHIPJET Program Pipe Breaks and Hardware 3-11 4.1 PWR Water Hammer 4-12 l-4.2 Reactor Coolant Chemistry Specification 4-13 5.1 Section III Stress Indices 5-18 5.2 Postulated Break in SIS System 5-19 5.3 Break Location Internal Loads 5-19 5.4-Piping Stresses 5-20 5.5 Stress Indices 5-20 5.6 SIS Line Crack Growth Results 5-21 5.7 BVPS-2 Fatigue Crack Growth Results 5-21 6.1 BVPS-2 High Stress Locations for WHIPJET LBB Analysis 6-4 O 7.1 Effect on Containment Parameters vs. The Type of Leak 7-9 8.1 Leak Rate Results for Stainless Steel Lines Inside Containment 8-3 9.1 Circumferential Crack Stability Evaluation for Inside Containment 9-5 9.2 Longitudinal Crack Stability Evaluation for Inside Containment 9-6 10.1 Excessive Load Circumferential Crack Stability for Inside Containment 10-2 10.2 Excessive Load Crack Stability for Longitudinal Lines Inside Containment 10-3 O lii g

l (. FIGURES No. Title PAGE ). 1.1 WHIPJET Program 1-4 3.1 6" SIS Piping to Cold Leg A 3-12 3.2 6" SIS Piping to Cold Leg B 3-13 3 3.3 6" SIS Piping to Cold Leg C 3-14 3.4 High Head Safety Injection to Hot Leg A 3-15 3.5 6" SIS Piping to Hot Leg B 3-16 3.6 6" SIS Piping to Hot Leg C 3-17 3.7 8" Bypass Piping - Cubicle A 3-18 3.8 8" Bypass Piping - Cubicle B 3-19 3.9 8" Bypass Piping - Cubicle C 3-20 3.10 10" Residual Heat' Removal Discharge to Loop B 3-21 3.11 10" Residual Heat Removal Discharge 3 3-22 to Loop C 3.12 12" Residual Heat Removal Suction from Loop A 3-23 ,3 3.13 12" SIS Accumulator Injection Lines - Cubicle A 3-24 3.14 12" SIS Accumulator Injection Lines - Cubicle B 3-25 3 3.15 12" SIS Accumulator Injection Lines - Cubicle C 3-26 3.16 14" Pressurizer Singe Line 3-27 4.1 Corrosion Review 4-15 3 5.1 Decomposition of Temperature Distribution Range 5-22 5.2 Temperature Distribution 5-22 and their 8 5.3 Comparisons of AT, AT2 1 Combination 5-23 O iv

O. FIGURES (continued) No. Title PAGE in the WHIPJET O-5.4 Approximation Used for AT2 Program 5-24 5.5 Comparison of Actual PWR Environment Data with The Austenitic Stainless Steel Crack Growth Law for Various R Values 5-25 0 5.6 Through-wall Stress Distribution 5-26 9.1 DPFAD Diagram for 12-inch Diameter SIS Lines 9-7 O O. O O O O. O. v O

. =. - O-SBCTION 1 INTRODUCTION 0- 'WHIPJET is an engineering program which assures that essential -structures, systems, and equipment are given protection at least 5) equivalent to that conventionally'provided by designing hardware to protect against the dynamic effects of postulated pipe whip and jet impingement. WHIPJET descnstrates that the fluid leakage from a postulated defect at the highest stress location

O concurrent with minimum material properties (in terms of normal plus safe Shutdown Earthquake loads) in a high energy piping line can be detected well before the rupture of the pipe.

Adequate time is assured to both detect and repair the leak and to bring O the plant to a safe, controlled shutdown. The program hec been applied to the Beaver Valley Power Station, Unit 2 (BVPS-2) for balance-of-plant (BOP) piping (i.e. other than the Primary Reactor Coolant Loop). Since WHIPJET is based on the leak-before-break (LBB) approach, using elastic-plastic fracture mechanics for assessing the potential for pipe rupture, WHIPJET is consistent with the procedural recommendations and analytical criteria found in NUREG-1061, Volume 3. [1.1) The basis for assurance of piping design and construction quality is presented in the BVPS-2 Final Safety Analysis Report (FSAR). The WHIPJET program is limited in scope to only those lines which

O have been previously determined to have a threat to essential surrounding safe shutdown equipment through potential damage caused by either pipe whip impacts or jet impingement loadings.

These lines were determined following NUREG-0800 [1.2] Standard ) Review Plan (SRP) Section 3.6.:2, except that arbitrary intermediate breaks were previously eliminated as approved by the Nuclear Regulatory Commission (NRC). Many lines do not require any protection hardware on the basis of detailed target and ) 1-1

O

O hazard analyses. Remaining system lines were then screened to eliminate piping for which crack growth mechanisms (such as corrosion or fatigue) and other failure potential (eg., water hammer) could not be ruled out or there was no economic benefit S-to include this piping in the program. Each remaining line was also analyzed for fatigue crack growth possibilities at the highest stress location for a postulated part-through wall crack 8' of an initial size equal to the American Society of Mechanical Engineers (ASME) Boiler and Pressure Vessel Code, Section XI acceptance standards; this defect was shown not to grow to a size (with margin) which could lead to a through-wall crack during the 8 normal operating life of the piping system. Lines were checked to determine which locations (not limited to 4 normal postulated pipe break locations) were most critical in O ! terms of highest stress and minimum material properties; these locations were then used in the elastic-plastic fracture mechanics calculations for assuring LBB. The pertinent system locations were analyzed to determine the stability of postulated 8 through-wall flaws in the base metal or weld given the local load and stress conditions in the pipe [ dead weight plus thermal plus pressure plus safe Shutdown Earthquake (DW + TH + P + SSE)). Leak rate versus crack size was determined considering normal operating loads (DW + TH + P) and fluid conditions. A limiting leak rate based upon the detectable leakage limit multiplied by a margin was used to determine a leak rate crack length. This leakage size crack length was analyzed for stability using a further margin on crack size and piping loads equal to normal + l

SSE, Crack stability was further assured by evaluating the limiting leakage crack size for excessively high piping loads i

equal to 1.414 (normal + SSE). l If these engineering criteria were satisfied, protection hardware l ( will not be required and need not be installed for the entire line. No relaxation of environmental qualification requirements 1-2 . - ~....

0 was intended with this approach. In summary, WHIPJET addresses an alternative engineering approach for the provision of protection from the mechanistic effects of postulated pipe ) rupture. Figure 1.1 shows the steps required in the WHIPJET program to satisfy the leak-before-break approach for eliminating pipe rupture hardware. 3 REFERENCES 1.1 Report of the U. S. Nuclear Regulatory Commission Piping Review Committeer Evaluation of Potential Pipe Breaks, NUREG-1061, Vol. 3, November 1984. 1.2 Standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants, LWR Edition, NUREG-0800, July 1981. 3 3 O O O O l-3

O Lines SRP 3.6.2 Requiring 4-(Except arbitrary Prutection intermediate breaks) O v Target and Interaction Evaluations e 1r Screening: Industry Experience, e Economics Leak Detectability Margin Material Highest Stress e Properties 3 Locations (Normal +SSE) With Minimum Material Properties 1r Fatigue Crack 9 Growth Evaluation 1r Limiting Detectable Leak Rate G Leak Rate x Margin > Calculations 4 Normal Loads l l 1r Margin on Crack Stability Normal +SSE m m Crack Size Evaluation Loads O ir Stability Check For Excessively 1.414 (No r+SSE) High Loads Loads e FIGURE 1.1 WHIPJET PROGRAM l-4 g

D i SECTION 1

SUMMARY

OF RESULTS D Fracture mechanics technology has advanced to the point that an engineering approach using the concept of leak-before-break in D lieu of postulating double-ended pipe rupture is now possible. The results for such a program (termed WHIPJET) has been successfully applied at BVPS-2. The overall results indicate that pipe rupture hardware is not necessary for lines inside ) containment greater than or equal to 6-in, nominal pipe size for systems that have passed initial screening. The screening process mentioned above included evaluation for the ) potential for stress corrosion cracking, excessive fatigue, water hammer, or other conditions which could result in failure of a pipe. Also because of economic considerations, some lines were not evaluated using the WHIPJET methodology. The main steam s'ystem lines f ell into this category due to the advanced state of construction and insta11ation of the pipe rupture hardware. Some other lines in the steam generator blowdown system and reactor coolant system were deleted from WHIPJET analysis due to D accessibility or projected low fluid leakage. Fatigue crack growth was also evaluated, and none of the lines f ailed the crack growth acceptance criteria (i.e., the extent of crack growth was minimal over the 40-year life of the plant). The evaluation for leak-before-break involved three key processes: leak rate calculations for normal operating loads, crack stability analysis for normal + SSE inertia loads, and an 3 excessive load case where stability is assessed for loads much higher than normal + SSE. Three margins were assessed related to these analyses: (1) a margin on leak detection of 10 inside containment on the minimum detectability limit of 0.5 gpm; this g: 2-1 7)

0-results in a 5 gpm value for assessing stability crack size; (2) a margin on crack size for assessing stability of at least 1.8; and (3) a margin on loads of the square root of two (1.414) for the excessively high load stability check. The intent of all of G these margins was met for the lines analyzed for BVPS-2. The application of the WHIPJET leak-before-break program at BVPS-8 2 has not only eliminated unwanted and unnecessary pipe rupture hardware devices, but much more is now known about the piping and its capabilities.than was the case before WHIPJET was applied. 4 e. O O l e l O O 2-2

O' sacTIon 1 ^ PROGRAM SCOPE -O 3.1 HIGH ENERGY-SYSTEMS WITHIN WHIPJET a O The BVPS-2 pipe rupture evaluation program identifies the high energy lines which have potentially hazardous consequences to adjacent plant equipment which are required to mitigate the effects of the postulated event. The high energy systems are: ASS Auxiliary Steam System BDG Steam Generator Blowdown System l-BRS Boron Recovery System CBS Chemical and Volume Control System DGS Gaseous Drains System FWE Auxiliary Feedwater System FWS-Main Feedwater System ,O GNS Gaseous Nitrogen System MSS Main Steam System RCS Reactor Coolant System RHS Residual Heat Removal System O SIS Safety Injection System These twelve systems were reviewed to establish operating pipe stress levels and then were subjected to break postulation j) criteria (in accordance with NUREG-0000 (3.1]). Further, the resulting pipe whip and jet impingement trajectories were i delineated in order to identify any interactions that could occur with adjacent safety-related structures, systems, and components. lO If-the interactions were judged severe enough to cause ^ significant impairment of the target's safety function, then the resulting damage was evaluated in order to determine whether loss of essential safe plant shutdown function would result. 3) 3-1 .() ,._,_-,.-_,---m--_--...,-.,,,,__o ..-_,,,_,,_,m_,-.w .-.,-----.__-,,__,w._, .-.__,--__-m_,, --._.-----..v.

---

,n

O This review indicated that there are no essential targets for postulated pipe breaks in several systems, and consequently no pipe rupture protection hardware needs to be provided. These systems were therefore eliminated from further consideration by 0-the WHIPJET program. Note that several of the detailed targeting analyses were performed concurrently with the WHIPJET program and in some cases resulted in the corresponding breaks being eliminated based upon targeting. The remaining systems which O have breaks that target essential safety equipment are: BDG Steam Generator Blowdown System O FWS Main Feedwater System MSS Main Steam System RCS Reactor Coolant System RHS Residual Heat Removal System SIS Safety Injection System Within some of these systems, some break's have been evaluated as not requiring hardware based upon detailed target analyses; these systems and affected pipe sizes include: RCS -- 6-in. inside containment RHS -- 10-in. inside containment (only one of two lines) SIS - and 3-in, inside containment 3-and 4-in. outside containment BDG -- 1.5-and 2-in. outside containment The above relevant systems and lines were reviewed against nuclear industry experience with pipe cracking as documented in NUREG-0691 [3.:2] where several occurrences of thermal fatigue induced cracking in main feedwater piping were noted. The I e feedwater system also has potential for water hammer. Based on l l this information and discussions with the NRC staff and its l l consultants, it was decided to exclude FWS from further review within the WHIPJET program. O i l l 3-2 S:

)- C The NUREG-0691 review also noted that a number of cases of vibration-induced fatigue cracking were experienced at socket weld fittings for small lines in particular. Except for the 2-h in. RCS lines, there are no postulated break locations in small lines which target essential shutdown equipment. Accordingly, all 2-in. lines were eliminated from further consideration. ) The material type used for the remaining systems were reviewed for documented toughness properties. Carbon steel was used for i the MSS piping and for some lines in the BDG system. The MSS pipe whip restraints have been fabricated and installation is )' near completion. Therefore, there was no economic benefit to include this system in the WHIPJET review. Consequently, a carbon steel testing program, planned originally to establish toughness properties of actual BVPS-2 materials, was determined not to be cost-effective. Steam Generator Blowdown piping located in the pipe tunnel area was considered too remote and inaccessible to be visually ' nspected for leaks. Accordingly, protective hardware is i prov'ided for the breaks in this area. Additionally, due to the recent concern for erosion / corrosion in ferritic steel systems and the need to prove adequate leak detection capability outside containment (plus the economics for small 3-inch diameter lines), the remaining EDG lines in the cable vault area were eliminated from WHIPJET and protection hardware was provided. 3 The balance of the systems use austenitic stainless steel which was judged to be sufficiently tough for inclusion in the WHIPJET program. Industry data for the stainless steel materials were utilized in the WHIPJET analysis. ) S 3-3 3

O Additionally, initial calculation of leakage rates using the NUREG-1061, Vol. 3 (1JL] safety margins of 10 on leakage and 2 on crack size suggested that some small stainless steel lines would not be applicable for LBB, Therefore, these lines were not 9 included in the WHIPJET program for eliminating pipe rupture hardware. After final target analyses, only one 4-in RCS line was ultimately eliminated from WHIPJET by this low leakage determination. O The piping deleted after screening and consideration of other factors (i.e., economics and leakage) are listed in Table 3.1. The piping systems remaining and the associated hardware are shown in Table 3.2. An indirect benefit due to increased awareness of the piping systems and loads also resulted -- many other pipe break locations and associated hardware have been O' determined to be unnecessary due to detailed targeting and interaction studies. Many of these studies would not have been performed at this level of detail if the WHIPJET program had not prompted them. 9 Also indicated in Table 3.2 are the piping materials and sizes. As shown, the pipe size ranges from 6 to 14-in. nominal pipe size for inside concainment austenitic stainless steel (Types 304 and 316). All field welds were made using the shielded metal arc welding (SMAW) process. A few shop welds utilized submerged arc welding (SAW). Of course, the stainless steel welds in these systems are used in the as-welded condition. O I The Beaver Valley Project uses the following nomenclature for piping: O f 5 9 3-4 g

O Line Number 2RCS-008-020-1 n a a a a

).

Pipe Class Designates BVPS-2 piping- -Line Number g. Piping System Abbreviation --Nominal Piping Size (e.g., 8-in. diameter) .O Table 3.3 presents the WHIPJET Program piping by line number and . lists the piping dimensions, material type, operating conditions, insulation type and thickness, and indicates the plant area code l were the piping is located. CS202, CS204, and CS206 are the three reactor coolant loop cubicles. CS203-is the pressurizer ~ cubicle which contains the surge line. 'O Table 3.4 lists the postulated break locations which would normally require hardware for each of the lines listed in Table 3.3 where, O 2RCS-004-C-C a a a Break Number

O Location Code Break Geometry C = Coolant Loop Cubicle C = Circumferential L = Longitudinal Split

) 4 The hardware associated with these breaks is also identified in i Table 3.4. Figures 3.1 through 3.16 show the location of these breaks, as well as the highest stress location based upon normal O 4 3-5 j) ,..,_v.--.m__,.,___,,,.,,,,_m. ._,,,,,.._......-._...,_,.m,,_.._m...z., _.

O plus SSE loads. The associated hardware (pipe rupture restraints and jet impingement shields) are also identified in the figures. Note that all of the stainless steel lines in the WHIPJET scope are Class 1 lines. In ansessing the highest stress locations, 9 portions of the lines (anchor to anchor) were Class 2, but the highest stress locations were always within the' Class 1 portions. i J 9 3.2 SYSTEM DESCRIPTIONS The following is a description of each WHIPJET system as to its

1. l primary function and its location within safety related areas.

3.2.1 Reactor Coolant System (RCS) i O The RCS transports heated water from the reactor core to the steam generators where heat is transferred to the main steam system. Leak-before-break for the main reactor coolant loop has been demonstrated by Westinghouse; primary loop breaks are O excluded'because of this work and the new limited scope rule ~ change to GDC-4. The RCS piping remaining in the WHIPJET program consists of the pressurizer surge line and the reactor coolant loop bypass lines. The piping is directly connected to the I primary reactor coolant loops and is part of the reactor coolant system pressure bcundary. 3.2.2 Residual Heat Removal System (RHS) O The RHS transfers heat from the reactor coolant system to the primary plant component cooling water system in order to reduce the fluid temperature of the reactor coolant system to the cold 9 shutdown temperature. The high energy portion of the piping is located directly adjacent to the connection to the reactor coolant loop and is part of the reactor coolant system pressure boundary. 3-6 g: l l

) 3.2.3 Safety Iniection System (SIS) The SIS is part of the emergency core cooling system (ECCS). The 3 short term function of the SIS is the prompt delivery of borated water to the reacto'r core following a loss-of-coolant accident. The system is divided into two parts: the low pressure safety injection system piping between the accumulator tanks and the RCS O cold leg, and the high pressure injection system piping between the charging pumps and the primary loop. The high energy portion of the piping is located directly adjacent to the reactor coolant loop and is part of the reactor coolant system pressure boundary. ) REFERENCES 3.1 Standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants, LWR Edition, NUREG-0800, July 1981. 3.2 Investigation and Evaluation of Cracking Incidents in D Piping in Pressurized Water Reactors, NUREG-0691, September 1980. 3.3 Report of the Uz Sz Nuclear Regulatory Commission Piping Review Committee; Evaluation of Potential for Pipe Breaks, NUREG-1061, Vol. 3, November 1984. 3 J e 3-7 3

O TABLE 3.1 PIPING ELIMINATED FROM WHIPJET SCOPE BY SCREENING OR OTHER CONSIDERATIONS g SYSTEM RUPTURE PIPE RETAINED REASON FOR g LOCATION SIZE HARDWARE ELIMINATION (IN) 1 FWS CONTAINMENT 16 5 INDUSTRY EXPERIENCE 4. FWS OUTSIDE 16 18 INDUSTRY EXPERIENCE CONTAINMENT RCS CONTAINMENT 1.5, 2 2 INDUSTRY EXPERIENCE g; i MSS CONTAINMENT 32 9 ECONOMICS MSS OUTSIDE 32 3 ECONOMICS 3 CONTAINMENT l l BDG TUNNEL - 3 5 INACCESSIBILITY OUTSIDE l CONTAINMENT .G. BDG CABLE VAULT - 3 3 LEAK DETECTION & OUTSIDE ECONOMICS CONTAINMENT RCS CONTAINMENT 4 1 LEAKAGE MARGIN < 10 9' TOTAL 46 9 l' O O-3-8 I, - - -.

O TABLE 3.2 FINAL PIPING SYSTEMS FOR WHIPJET ANALYSIS O PIPING PIPE MATERIAL BREAKS HARDWARE SYSTEM -SIZE 0 (IN) PRR JIS (1) (2) SIS 6 SA 376, 20 7 5 TYPE 316 O RCS 8 SA 376, 6 6 0 TYPE 304 Q-RHS 10 SA 376, 1 0 (3) 0 TYPE 316 RHS 12 SA 376, 1 1 0 TYPE 316 O SIS 12 SA 376, 28 10 3 TYPE 316 O RCS 14 SA 376, 13 8 0 TYPE 304 TOTAL 69 32 8 O TOTAL 40 0 NOTES: i l) (1) PRR = Pipe rupture restraint (2) JIS = Jet impingement shield (3) Break requires SIS restraint () 3-9

1 b' i i TROLE 3.3 4 MHIPJET PROGRAM PIPIN8 DATR i ^ LINE NUMBER 00 WALL MRTERIAL TEMP PRESS INSULATION

• THK LOCATION l

~ 1 2SI5-006-012-1 6.625 0.718 58376 TYPE 316 545 2327 REFLECTIVE 3.5 CS206 l 25I5-006-269-1 6.625 0.718 58376 TYPE 316 545' 2327 REFLECTIVE 3.5 CS206 25I5-006-015-1 6.625 0.718 SR376 TYPE 316 545 2327 REFLECTIVE 3.5 CS204 25I5-006-270-1 6.625 0.718 SR376' TYPE 316 545 2327 REFLECTIVE 3.5 CS206 I 2515-006-016-1 6.625 0.718 5R376 TYPE 316 545 2327 REFLECTIVE 3.5 CS202 2515-006-271-1 6.625 0.718 SR376 TYPE 316 545 2327 REFLECTIVE 3.5 CS206 2S15-006-026-1 6.625 0.718 SR376 TYPE 316 545 2327 REFLECTIVE 3.5 CS206 25I5-006-268-1 6.625 0.718 58376 TYPE 316 545 2327 REFLECTIVE 3.5 CS206 25I5-006-024-1 6.625 0.718 SR376 TYPE 316 613 2327 REFLECTIVE 3.5 CS204 25I5-006-266-1 6.625 0.718 SR376 TYPE 316 545 2327 REFLECTIVE 3.5 CS206 2 SIS-OO6-025-1 6.625 0.718 SR376 TYPE 316 613 2327 REFLECTIVE 3.5 CS202 2SI5-006-267-1 6.625 0.718 SR376 TYPE 316 545 2327 REFLECTIVE 3.5 CS206 ^ 2RCS-OOS-020-1 8.625 0.906 SR376 TYPE 304 613 2327 REFLECTIVE ' 3. 3 CS206 2RCS-OO8-021-1 8.625 0.906 SR376 TYPE 304 545 2327 REFLECTIVE 3.3 CS206 ? 2RCS-OOO-040-1 8.625 0.906 SR376 TYPE 304 613 2327 REFLECTIVE 3.3 CS204 2RCS-OOS-041-1 8.625 0.906 SR376 TYPE 304 545 2327 REFLECTIVE 3.3 CS204 j Y 2RCS-DOS-060-1 8.625 0.906 SR376 TYPE 304 613 2327 REFLECTIVE-3.3 CS202 H 2RCS-OOS-061-1 8.625 0.906 SR376 TYPE 304 545 2327 REFLECTIVE 3.3 CS202 o j 2RHS-010-023-1 10.750 1.125 SR376 TYPE 316 140 640 REFLECTIVE 3.7 CS202 2RHS-010-024-1 10.750 1.125 SR376 TYPE 316 140 640 REFLECTIVE 3.7 CS202 2RHS-012-001-1 12.750 1.312 SR376 TYPE 316 613 2327 REFLECTIVE 3.7 CS206 2RHS-012-056-1 12.750 1.312 5R376 TYPE 316 613 2327 REFLECTIVE 3.7 CS206 2S15-012-289-1 12.750 1.312 5R376 TYPE 316 545 2327 REFLECTIVE 3.7 CS206 l 25I5-012-067-1 12.750 1.312 SR376 TYPE 316 545 2327 REFLECTIVE 3.7 CS206 2 SIS-012-066-1 12.750 1.312 SR376 TYPE 316 545 2327 REFLECTIVE 3.7 CS206 4 2 SIS-012-250-2 12.750 0.375 SR376 TYPE 304 105 688 NONE O.0 CS102 I 2 SIS-012-288-1 12.750 1.312 58376 TYPE 316' 545 2327 REFLECTIVE 3.7 CS204 j 2 SIS-012-071-1 12.750 1.312 SR376 TYPE 316 545 2327 REFLECTIVE 3.7 CS204 2 SIS-012-070-1 12.750 1.312 SR376 TYPE 316 545 2327 REFLECTIVE 3.7 CS204 l 2 SIS-012-252-2 12.750 0.375 SR376 TYPE 304 105 688 NOPE 0.0 CS105 1 25I5-012-287-1 12.750 1.312 SR376 TYPE 316 545 2327 REFLECTIVE 3.7 CS202 1 2 SIS-012-069-1 12.750 1.312 58376 TYPE 316 545 2327 REFLECTIVE 3.7 CS202 2 SIS-012-068-1 12.750 1.312 58376 TYPE 316 545 2327 REFLECTIVE 3.7 CS202 2 SIS-012-251-2 12.750 0.375 SR376 TYPE 301 105 688 NONE O.0 CS107 2RCS-014-004-1 14.000 1.406 SR376 TYPE 304 656 2327 REFLECTIVE 3.7 CS.~9. 3 NOTE: e Reflective insulation has both an inner and outer-stainless steel layer I t 4 O e e e e e e e e-g

O O O O O '**" '^ O O O O O C UNIPJET PROGRAM PIPE BREAKS AND HARDWARE LINE Hun 8ER BREAK nun 8E2 TYPE MELS HARDuRRE LINE MUMBER BREAK HUnDER TYPE MELD. HARDWARE (Figure 3.1 o-anch SIS to Cold Leg A) (Fi g.are 3.13 12-inch SIS to Cold Leg A) 2515-006-012-1 25t S-091-C-C TE Fu 25IS-PRR-850 2515-012-289-1 2SI S-Oh3-C-L EA Sn 25I54RR-806 2 SIS-151-C-C EA Fu 2SI S-PRR-850 2 SIS-OOS-C-L EA Sn 2 SIS-PRR-806 25IS-152-C-L EA FM 2 SIS-JI5-850 2 SIS-007-C-L EA SM 2 SIS 4 RR-806 25IS-153-C-C EA Fu 2 SIS-PRR-851 2 SIS-OO9-C-L EA SW 25154RR-806 25IS-154-C-L EA Fu 2SI S-JI 5-851 2SI5-010-C-C EA Su 2SI 54RR-806 2 SIS-155-C-C ER FM 2 SIS-PRR-851 2SI S-013-C-L EA FM 2SI S-PRR-806 2515-156-C-L EA FM 2 SIS-JI S-851 2SI S-011-C-L EA Su 25I5-PRR-806 2 SIS-158-C-L EA Fu 25I5-J15-851 2SI S-012-C-C ER FM 2SI54RR-8%,804 2SI 5-181-C-L EA FM 2515-JI S-851 2515-012-067-1 NO ASSOCI ATED HARDunRE OSIS-CD6-269-1 NO ASSOCIATED HARDWARE 25I5-012-066-2 NO ASSOCIATED HAROMARE 25I5-012-250-2 NO A5SOCI ATED HARDWARE < Figure 3.2 o-ar.ch SIS to Cold Leg B) 2515-OO6-01*s-1 2 SIS-162-C-L TE Fu 2515-JI5-860 (Fi y e 3.14 12-inch SIS to Cold Leg B) 25IS-164-C-L EA FM 2 SIS-JI 5-860 25I:>-012-288-1 2515-016-C-C TE Fu 2SI5-JIS-803 25IS-166-C-L EA FM 25 t S-JI 5-860 2SI S-017-C-C EA Su 25I5-JIS-803 2 SIS-166-C-L EA Fu 2 SIS-JIS-860 2SI5-018-C-L ER Su 2 SIS-JIS-805 25IS-183-C-L EA FM 2 SIS-JIS-860 2SI5-020-C-L EA Su 2SI 5-JIS-805 2 SIS-aXa6-270- 1 NO ASSOCIATED HAPDMARE 2SI S-022-C-L EA SM 25I5-JIS-805 2SI5-024-C-L EA 5u 2SI S-PRR-8 L5,818 (Figure 3.3 6-ir.ch SIS to Cold Leg C) 2SI S-026-C-L EH Bn 25IS-PRR-817 2515-006-010-1 2 SIS-177-C-C TE Fu 2SI S-PRR-871,JI 5-871 2SI5-028-C-L EA BN 2SI 5-PRA-816 2 SIS-179-C-C EA FM 2 SIS-PRR-872 25I5-012-071-1 HO ASSOCI ATED HARDuRRE 2515-006-271-1 NO ASSOCIATED HARDWARE 25I5-012-070-2 NO ASSOCIATED HARDWARE 2515-012-252-2 HO ASSOCI ATED HAROuRFI (Figure 3A o-anch SI5 to Hot Leg A) 25IS-OO6-026-1 NO ASSOCIATED HARDWARE (Figure 3.15 12-inch SIS to Cold Leg C) 251S-006-266-1 NO ASSOCIATED HARDWARE 25I5-012-287-1 2SI5-032-C-L EA SM 2SI5-JIS-800 2SI S-034-C-L EA Su 2 SIS-JIS-8CG (Figure 3.5 o-anch SIS to Hot Leg 8) 2SI5-035-C-C EA Su 2SI5-PRR-824 25I5-006-024-1 2 SIS-118-C-C TE FM 2 SIS-PRR-860 2515-036-C-L ER SM 2 SIS-JIS-800 25IS-119-C-C EA Fu 2 SIS 4RR-861 2SI5-037-C-C EA Fu 25I54RR-827 2515-006-260-1 NO ASSOCIATED HARDuRRE 2SI S-038-C-L EA Fu 2SI S-PRR-827 2515-039-C-C EA Bn 2 SIS-PRR-824 (Fi gure 3.6 o-anch SIS to Hot Leg C) 2SI5-040-C-L ER Bn 25IS-PRR-824 2515-006-005-1 2515-147-C-C TE Fu 2515-PRR-870 2SI5-041-C-C EA Su 2515-PRR-826 25IS-148-C-L EA FM 2 SIS-JIS-8?O 2 SIS-042-C-L EA Su 2SI S-PRR-827 25I5-006-267-1 HO ASSOCIATED HAkDWARE 2 SIS-043-C-C EA SW 2 SIS-PRR-825 2 SIS-044-C-L ER Su 2SI54RR-827 (Fiqure 3.7 8-snch Loop A 8 pass) 25I5-012-069-1 NO ASSOCIATED HARDuRRE 9 2RCS-OOS-020-1 2RCS-OO4-C-C TE Fu 2RCS-PRR-818 2515-012-068-2 HO ASSOCIATED HARDuGPE 2RCS-OOS-021-1 2RCS-OO1-C-C TE fu 2RCS-PRR-853 2515-012-251-2 NO ASSOCIATED H4RDWARE (Figure 3.8 O-anch Loop B bypass) (Fi gure 3.16 14-inch Surge Lane to Hot Leg C) ORCS-OOS-040-1 2RCS-DOS-C-C TE Fu 2RCS-PkR-820 2RCS-014-084-1 2RCS-241-C-C TE fu 2RCS-PRR-814,815 2RCS-OO8-041-1 2RCS-DO5-C-C TE FM 2RCS-PRR-852 2RC5-242-C-C EA Su 2RC54RR-814 2RCS-243-C-L EA Su 2RC5-PRR-816 2kC5-244-C-C EA SM 2RC5-PRR-825 (Figure 3.9 6-anch Loop C 8 pews) 9 2RCS-OO8-060-1 2RCS-012-C-C TE FM 2RCS-PRR-828 2RCS-245-C-L EA Su 2RCS-PRR-816 2RCS-OO8-061-1 2RCS-GO9-C-C TE Fu 2RCS-PRR-851 2RC5-246-C-C EA BN 2RC54 RR-814,815 2PCS-247-C-L ER Sn 2RCS-PRR-815,816 (Fiquee 3.10 10-inch RHS Discharge, Cubacle B) 2RC5-248-C-C EA SN 2RC5-PRR-825,814 2RHS-010-023-1 2RHS-OO A-C -C TE FM 2 SIS-PRR-817 2RCS-249-C-L EA Bn 2RCS4RR-816 2RCS-250-C-C EA SM 2RC54RR-825,815 (Figure 3.11 10-ie.ch RHS Dis 4herge, C.4biel C) 2RCS-252-C-C ER 8n 2RC5-PRR-815,814,826 2RHS-010-024-1 2RH5-OO2-C-C TE Fu 2 SIS-PRR-826 2RCS-254-C-C EA FM 2RC5-PRR-815,825,810 2RCS-256-C-C TE Fu 2RCS-PRR-807,324 (Figure 3.12 12-ie.ch RH5 Suction from Not Leg A) 2RHS-012-OO3-1 2RMS-OO3-C-C TE FM 2RHS-PRR-801 2RMS-012-056-1 NO ASSOCIATED HARDWARE TE = T eres a nal End Break EA = Exceeds the Stress end/or Fatigue Factor Threshold for Br==k CODE: F u = A Fi el d F ebra c at ed peld i s nade at This Postulated Pipe Rupture Location Su = A Shop Fabricated peld i s nede at This Postulated Pipe Ruptura Location BM = 8ase hetal. There as eso eld at Tha s Postulated Pipe Rupture Locataon

C' e 6'- CB S e se # 2 SIS-006-263-1 e 2 SIS-FRR851 2 SIS-JIS851 ch 2 SIS-151-C-C L. 2 SIS-158 C-L I{ 2 SIS-PRR850) 2 SIS-152-C-L ESIS-I SS-C-C E

2. SIS-091-C-C c-TE LE 2 SIS-156-C L 2 RCS-275 1 CE 2 SIS-153-C-C (COLD LEG) 3ha CL EL 732-2 Sis-alS850 i5 2515-154 C-t e

2S15-006-12-1 Figure 3.1 6" SIS PIPING TO COLD LEG A CUBlCLE A WEST LOOP e e 3-12

_ _.=. 2 SIS-JIS860 2 SIS-183-C-L 2 SIS-PSSP 208X CL EL 738'-O' L-A / l -{L_S_Sl ~ If0 2 SIS-lG 8-C-L U L-EA 2 SIS-16 6-C-L N 6 4 -E 2 SIS-164 -C-L u b '2 SIS-006-270-1 2 SIS-006-15-lQ 2 SIS-162-C-L G 2RCS - 27 S-6-1 CL EL 732'-3k! (COLD LEG) SCI SC2 A 6" SIS PIPING TO COLD L EG B ngure 3.2 CUBICLE B SOUTH LOOP -Y

l 6" CBS

l CL EL 74 l'- 5" 2 SI S-179-C- C '.h 2 SIS-17 7-C- C '7 2SI S-0 06-271-1 E

2 SIS-PRR872 2 SIS-PHit871 3 - 1 SCI 2RCS-275-9 -l SC2 C L EL 732'-3,I ~ j (COLD LEG) / Figure 3.3 6' SIS PIPING TO COLD LEG C CUBICLE C EAST LOOP O e e e e e e e e e e

1 6 2 0 6 0 0 S is 2 8 6 C A '6 GEL T v OH O T NO I T v CE JN I YTEF v A S DAEH H v G I R H E C U 4 D R E 3 O R 1 v H e "3 8 C r 6 N X ug A "6 2 i -6 F 0 0 S 1 5 2 o Y l 1

O 06 I B 8. 4 R 2 1 C 0 6 R 8 P C 6 R 8 0 R S I P 0 S l S 2 l I s S S 9 I iS S 2 2 2 C -C B G 9 E L l I -S T 9 I O S H 2 O 9 T E 1 1 G 8 N 8 I C P I '6 P PO S O I L /"6 S HT l U O 4 3 S f '2 9 G 3 5 2 B 0 E 7 3 s - L 'L E / S E e L C T rC RO L l uI 2H C G gB i U G F C / I CS Y 1 I i 'i I

2 SIS-PRR 870 ' s' 2 SIS-OO G 1 4 2 SIS-148-C-C -E N 251S-0 06-267-I u h 0 2 SIS-J IS870 Sci Jc2 2RCS- 029 1 HCL EL 732'-3l[ y (HOT LEG) -E 2 SIS -14 7-C- C I Figure 3.6 6' SIS PIPING TO H0T LEG G CUBICLE C EAST LOOP

eCL STEAM GENERATOR j 2RCS-SG2fA T \\ 2 RCS-OO 8-20 2 RC S-O O4-C-C a 1 CT C L EL 738'-7 CL REACTOR COOLANT PUMP 2RCS-031-2-1 OV 590 2 RCS-P2 t A % CROSSOVER) u LEG y 2RCS-PRR818 -2 RHS-Ol2-1-1 C L EL 722'-OI3, 2 RCS-029-I-l L2RCS-PRR853 i (HOT LEG) 16 ) 25'S 6-26-' - l Mov 89 } l i C-T E 2 RCS-008 1 rCL REACTOR 2 RCS-001-C-C 2RCS-27S-3-1 (COLD LEG) .i 2RCS -PEV 21 CL EL 732-3k, Figure 3.7 B BYPASS PIPING - CUBICLE A ~ WEST LOOP o e e e e e-e e e e e

4 i l [CL REACTOR /CL STEAM GENERATOR 2RCS-REV21 4 2RCS-SG218 15' CL EL 732-3Tr> o.- 2nsitai -' . - xl ' d l MOV 592 I4 CL EL 73B'-7[ O j MOV 586 8 2RCS-OO8-40-1 2 RCS-DOS-C-C l C-TE 5 C-T l 2RCS-OO8-CG 2RCS-PRR820 2 RCS-275-6-1 2RCS-OOB 1 (COLD LEG) I E MOV 593 i 2RCS-031-35-1 (CROSSOVER LEG) 2RCS-PRR852 l 'y CL EL 722 Ok' I ' CL RC PUMP y 2RCS-P218 n ure 3.8 8" BYPASS PIPING-CUBICLE B a SOUTH LOOP l

eCL RC PUMP 2RCS-P2tC 2 RCS-OI2-C-C 7 j C-HOV 595 2RCS-OO8-60-1 \\ 2 RCS-P RR828-\\; 2RCS-275-9-1 I (COLD LEG) gCL REACTOR [ - 2RCS-OO8-61-1 2RCS REV2I 2 RCS-OO9-C-C j C 738'-7T '2RCS PRR851 4 CL EL l Y 732'-315" 5 l 2RCS-031-8-1 i g CL EL 722' 0 13,, j NOV594 (CROSSOVER LfG) l I 2 R C S-029-7-1 (" CL STEAM GENERATOR 2 RCS-Ol4-84-1 (CL PRESSURIZER ~ 2RCS PRE 21 Figure 3.9 8' BYPASS PIPING - CUBICLE C l EAST LOOP o e e e e e e e e

l O O 1 O 1 7 2 1 0 S B I O S P 2 OOL OT H O C S I ) D L A A I 0 C V 2 S O 7 M O E V R 1 O 2 3 C M 2 T S A E 0 H 1 0 L O A S U i' H D I R S 2 ER 0 O 1 0 1 3 e O rug i F O O vU 1 lllll1Illll 1 ll l1 ll1< 1

o e 1 e 9 6 2 1 0 S t S C e 2 P 7 OOL O 1 T 4 e 2 H 0 CS 1 0 ID S L E H AV z T-R O e c 2 M ) E C R S#i TAE 8 , s, 2 H 0 L e 2 7 C A V S U O D I M S E R e 0 1 1 1 3 e erug s i F e e m i i! i! ii l l l l !I

111, 1

O O O C -C 1 '6 3 k 6 O I 3 5 '0 O 1

• -)

22 -S 1-2 G l 7 O R 3E O A H 9 - L 27L 2 0 SE P - LT H O SEO RL O H 2C L CRL B M 2C( 1 O 0 R 7 F O 1 I C V 1 O 2 ' M NO lO 8 O I l -S 2 B T H 0 R C R 7 R U S P 2 V O O S 6 L M H A 1 R 0 V 1 2 2 2 O 7 M lO E L R SE D y H T RL AE 2C H LAUD P I O S O O A E L 2 R 0 T 7 S 2 E V 1 W O H 2 O A 1 1 0 7 3 V e I C O r SW H ug i F O 2 CS O lli i! ll!

2S1S-012-250-2 MOV86SA %-CL SAF INJ l3 ^ ACC TANK -E 2 SIS-Ol3-C - L 2 SIS-Ol2-66-2 2 SIS -TK21A 12 -E 2 SIS-Ol2-C-C CL EL 704'- 8" -E 2 SIS-Ol l-C-L 12" C8 8 10 ,SC2 -E 2 SIS-OIO-C-C CL EL742'-O* 2 SIS-Ol2-289-i E 2 SIS-OO9-C-L h 2 SIS-PRR806 2 SIS-PRR804 3 (j 2 SIS-007-C- L y ] = L. 2 SIS-OOS-C - L 2 sis-Oiz-67-i .i 2 SIS-OO3-C-t Figure 3.13 12" SIS ACC INJECTION LINES CUBICLE A WEST LOOP 2RCS-275-3-1 EL 732'- 3[ 4 l l O e e e e e e e e e e

O O O O O O O O o o e CL SAFETY INJECTION ACCUMULATOR TANK 1 2 C 8 8 n L fA 2 SIS-028-C-L 2 SIS -T K 218 2 SIS-PRR817 2 SIS-Ol 2-288-1 uu U ( y 2 SIS-PRR818 f 2SI S-PRR816 2 SIS-PRR815 2 SIS-026-C -L 2 SIS-024-C -L QN2 SIS-JIS805 4 -) 2 RH3-oio -oz3-1 2SI S-Ol 2 1 ? Y 2 SIS-Ol2-2S2-2 L-EA 2 SIS- 022-C - L Mov8658 2 SIS- 020-C-L gg 2 SIS- 018 - C-L s 2 17 . SIS-JIS80' C-EA 2 SIS-Ol7-C-C 2 RC S-275 1 Ch (COLD LEG) 2S15-012-7 0-2 E 2 SIS-Ol6-C-C Figure 3.14 SIS ACCUMULATOR INJECTION LINES CUBICLE B SOUTH LOOP

o 9 41 43'N 2 SIS 041-C-C C-E'A CE 2 SIS-043 -C-C,, ,_2 SIS-Ol 2 1 42 44 2 SIS-044 C-L 2 SIS-042 C-L L-EA L-E 12"C88 IVSb c P_ g2 SIS-PRR827 3, 4, 2 SIS-PRR82S e 2 SIS-PRR824 Uh 7 C-A 2 SIS-039 C-C 2 SIS,JIS800 4 40 2 SIS-040-C-L L-E t 2RHS-Ol0 24-1 n 7 j/E C-EA 2 SIS-037-C- C f 2 SIS-038-C -L L EA 35 j C-EA 2 SIS-035-C-C V 12" C 8 8 36 L-E 2 SIS- 036-C - L-Sq -2S1S-012-287-1 4 hv L-E 2 SIS-03 4-C - L NSC2 c 2 SIS-Ol2-68-2 p 3 ~ .E 2 SIS-032-C - L g l k SAFETY EJECTION l ACCUMULATOR TANK 2 SIS.-TK 2lC 2S1S 012-251-2 Figure 3.15 12" SIS ACCUMULATOR INJECTION LINES CUBICLE C EAST LOOP O 9

CL RC PUMP 2RCS.OO6-61-1 MOV$87X l 2RCS OC6-GO-8, L j ) 2 RCS-275 9-I (COLD LEG kg l n-CL REACTOR l 2RCS REv21 l 2 RCS-029-7-1 MOV 594 / CL EL 732-3,3-iy g3 y (HOT LEG) 2RCS-ONS l

• CL EL 722*

i tj 4 rCL PRESSU#1ZER (CROSSOVER G) 41 2RCS-PRE 21 2RCS-241-C-C -T 2RCS-242-C-C -T 2RCS-256 C C l 2RCS-243C-L 2 RCS-PRR607 44 2RCS-24+C-C 1 2RCS Ol4-064-l 25-mis 2RCS-245-C-L 245 m7 2RCS-PRR810 l 2 RCS-246-C-C l 2RCS-PRR824 4 I 233CS-247-C-L 'I 2RCS-PRR625 gd, 2RCS-246-C-C 5 2RCS-249-C-L ZINS-PRRel4 C-2 RCS-254-C-C 5 l 2RCS-250-C-C ,j0 2RCS-252 C-C ,E ~ Figure 3.16 14 PRESSURIZER SURGE LINE

O. SBCTION 1 SCREENING CONSIDERATIONS ~ 9 Four systems,(Feedwater, Reactor Coolant, Residual Heat Removal, and Safety Injection) were evaluated through a screening process O. designed to eliminate from consideration any piping which exhibits the potential for o Anomalous conditions: excessive vibration, flow stratification, failure of pipe fittings, failure of - equipment supports, and uneven mixing o Fluid transients (water hammer) o Stress corrosion cracking The overall results were summarized in Section 3, and as indicated, the FWS lines were screened out of the WHIPJET scope. The following sub sections discuss each of these items in detail for the three systems (RCS, RHS, and SIS) remaining in the WHIPJET program. O' 4.1 ANOMALOUS CONDITIONS h 4.1.1 Flo_w Stratification O Flow stratification is a thermal fluid phenomenon identified in IE Notice 84-87. It potentially occurs in large diameter fluid systems where the following circumstances exist: (1) the piping i system is long and horizontally oriented, (2) the pipe is filled I with hot (or cold) fluid flowing very slowly, and (3) mech colder (or hotter) fluid is introduced at some point upstream of the horizontal run at a rate by which mixing does not occur. O i [ 4-1 e i

) This phenomenon was evaluated for those systems within the final scope of WHIPJET systems and was concluded to have no potential for application. The safety injection system is a predominantly cooler, small diameter, high flow system. It is appropriate to note that, while the feedwater system is not in the final WHIPJET program, this system was reviewed in detail for specific potential for flow stratification and concluded not to have the sufficient conditions to be a concern. The potential for backflow from the steam generators is minimized by the main feedwater isolation valve (FIV) and check valve in each line. Each system within the WHIPJET program was reviewed for its susceptibility to fluid flow atratification. It is concluded that no potential for flow stratification exists. Previous plant experience has shown acceptably high flow rates at primary loop branch connections. The Safety Injection System is predominantly smaller piping which experience cooler, high volume fluid flow rates. g 4.1.2 Induced System Vibration Positive displacement pumps are potentially a source of vibration and fatigue failure. BVPS-2 systems remaining in the WHIPJET program have no positive displacement pumps in service with the exception of the hydrostatic test pump. Vibration from this pump is insignificant because of its low flow and infrequent use. g General system vibration is addressed in the BVPS-2 startup l testing program. 4.1.3 Pipe Fittings Consideration for LBB Analysis g A review was conducted to determine the presence of any cast materials for systems evaluated in WHIPJET. All fittings were wrought stainless steel, SA403 WP316. Other than for the large O 4-2 g

O-bore primary coolant piping, no cast material is used for i fittings within the WHIPJET program high energy systems. Typical PWR piping system fittings include elbows, tees, and branch connections. 9 The most often expressed concern regarding wrought fittings is for weld cracking in elbows. Elbows may be either seamless or 8' welded. Seamless elbows are generally forged, but may also be made by bending seamless pipe. Welded elbows are usutlly made from plate materials which are either rolled and welded into pipe l sections before being bent into elbows or formed into halves of 8 l elbows which are then welded together by longitudinal welds. Service experience with elbows has been excellent except for a few instances of cracks caused by a defective longitudinal elbow i weld. O All of the BVPS-2 lines remaining in the WHIPJET program have only forged seamless fittings. Radiographic, ultrasonic, or liquid penetrant examinations were performed by the manufacturer O in accordance with ASME Section III, NB-2550 for Class 1 seamless materials, thus assuring material soundness. Satisfactory experience combined with the type of elbows and NDE checks, reduce the probability of any longitudinal cracking of the 8 elbows. Also, the material properties for the fittings are i comparable to those in the wrought pipe itself. Additionally, each line of piping was evaluated to determine limiting stress locations. The limiting locations where WHIPJET is being applied correspond to those determined to be the highest stress points as dictated by the ASME Code and standard piping i analysis procedures. These procedures consider the entire piping 9 i system including pipe fittings. O i f. 4-3

'O 4.1.4 Equipment Supports Evaluation The potential for the failure of equipment supports resulting in 5) the subsequent failure of attached high energy piping exists if the supports have not been seismically qualified. Since ASME Section III piping systems terminate at ASME Section III Seismic Category I components, seismic qualification is assured. O 4.1.5 Uneven Mixing Thermal fatigue due to water temperature fluctuations caused by O slow, continuous mixing of water flows with different temperatures, has been reported to occur for certain BWR plants. BVPS-2 WHIPJET systems were reviewed and concluded to be free of causative factors leading to this phenomenon. i 4.2 FLUID TRANSIENTS O 4.2.1 Introduction BVPS-2 evaluated the potential for flow transient events in accordance with applicable Stone and Webster Engineering }) Corporation (SWEC) and NRC guidelines. A review by fluid system was conducted. Applicable flow transients were identified and the transient description and design information used in the pipe stress evaluation. Potential water hammer sources considered for g the design of BVPS-2 piping systems were based on industry experience and the concerns presented in various NUREG's, including specifically NUREG-0582 and NUREG-0927 [4.1, 4.5]. O Water hammer caused by steam voids, condensation, flashing and thermal mixing are controlled through system design, operating procedures, and operator training. O 4-4 O \\ -

O Since NUREG-0582 was issued, additional NUREG's that address the issue of water hammer were published [1,2 through 1J.]. These documents recognize that water hammer events are inevitable and 8 I provide guidance to eliminate or reduce the frequency and effects of these events. They also acknowledge that water hammer is not as significant a safety issue as previously anticipated since the resulting damage from water hammer events has been limited to piping and equipment supports. Table 4.1 is reproduced from Table 3-2 of NUREG-0927 and provides specific PWR system water I hammer causes and preventive measures. The detailed discussion below addresses the BVPS-2 provisions for minimizing steam / water hammer effects and, where applicable, addresses specific details associated with that guidance for minimizing water hammer. Specific flow transients for BVPS-2 were identified subsequent to 1 O a comprehensive review of NRC Guidelines and plant experience. I Guidance was obtained from NUREG-0582 (1,1], which is a comprehensive summary of water hammer and fluid system transients. Appendix A contains a working summary of and 9 l comments on fluid transients evaluated against NUREG-0582, and as appropriate, included in the Plant's design basis for pipe stress analysis. 4.2.2. Provisions for Minimizing Water Hammer Effects Systems within Westinghouse scope of supply are not in general I susceptible to water hammer. The reactor coolant, and residual heat removal systems have been specifically designed to preclude water hammer. Operating experience at other plants with Westinghouse systems have verified this design approach. O j Westinghouse has conducted a number of investigations into the causes and consequences of water hammer events. The results of these investigations have been reported to Westinghouse operating l plant customers and have been reflected in design interface i 4-5 g

.O requirements to the BOP designer for plants under construction, to assure that water hammer events initiated in the secondary systems do not compromise the performance of the Westinghouse-supplied safety-related systems and components. ~ In general, the approaches taken, individually or in combination, to address water hammer concern were to prevent or minimize water 'O hammer effects through system design features and operating procedures. Potential water hammer sources to be considered were based on industry experience and the concerns presented in various NUREG's. The following discusses in more detail the jo potential water hammer sources, if any, that were considered in the design of the subject systems and the actions taken to minimize and prevent water hammer effects. -O 4.2.2.1 Reactor Coolant System (RCS) i There is a very low potential for water hammer in the subcooled water solid portions of RCS since these portions of RCS are , designed to preclude void formation. Relief valve discharge loads associated with the pressurizer have been specifically identified and analyzed for BVPS-2. (NUREG-0927) i O 4.2.2.2 Safety Injection System (SIS) l As discussed below, it is considered unlikely that water hammer uld occur in the Safety Injection System. The low temperature O SIS lines, which are normally water solid, have a very low probability of steam void formation. Proper initial fill and venting ensures that low and high head safety injection system O piping remains filled. In addition, the head of water provided by the RWST provides a continuous mechanism for ensuring that the l low head safety injection system lines remain full. l O 4~6 O

. _ ~ T O For the SIS lines which are part of the Residual Heat Removal System return flow path, operating procedures for RHS minimize the potential for water hammer in these lines. For the SIS lines which are part of the Reactor Coolant Pressure Boundary to the O' first isolation valve, there is a very low potential for water hammer as indicated in the RCS discussion. O 4.2.2.3 Residual Heat Removal System (RHS) A portion of the RHS piping is high energy because it is normally l pressurized by the RCS or SIS during normal plant operating O conditions. When RHS is operating (i.e., the short operational period), valve closure times and operating procedures minimize the potential for water hammer. Proper fill and venting will l initially ensure that air does not become trapped in any part of O the RHS during start-up. Additionally, just prior to RHS l initiation, the RBS will be cross-connected with the Chemical and Volume Control System (CHS). This action utilizes the pressure head in the CHS to collapse any voids (should they remain) prior 7 O to opening the RHS suction valves from the RCS. 1 l When the RHS system is not operating, the normally pressurized . portions of the system are water solid and are either at a low 8 temperature or subcooled. RHS voiding, therefore, has been addressed by a combination of operator training and startup procedures which provide for the complete filling and venting of the system before operation. (NUREG-0927) O 4.3 STRESS CORROSION CRACKING The pertinent lines which were determined to have essential targets and passed the material and fatigue screening criteria were reviewed further to determine if they were susceptible to stress corrosion cracking (SCC). The review has concluded that 4-7

based on specified chemistry, cleanliness, fabrication, and operating controls, and successful operating experience, SCC is not expected to occur. Figure 4.1 presents an overview of the O corrosion-related review conducted for BVPS-2. The WHIPJET program lines in the primary side environment are in the RCS, RHS, and SIS branch connections to the reactor coolant I loop piping. Type 304 and Type 316 austenitic stainless steel piping materials are used for these lines. These materials when used in other than the solution annealed condition are susceptible to SCC when exposed to three simultaneous conditions: high tensile stress, high temperature, and a corrosive environment. Controls were used in the manufacture and fabrication of this D piping to minimize the material susceptibility to SCC by limiting both cold work effects, (i.e., strain hardening) and sensitization effects (i.e. chromium-carbide precipitation in the grain boundaries). Cold working of the material was controlled J by requiring the original manufacturing process of the pipe to include solution annealing and by also requiring subsequent forming operations such as pipe bending to be done with bend radii greater than or equal to five pipe diameters for cold bending and by re-solution annealing all hot bent piping. Piping sizes greater than 2-inch nominal pipe size included in the WHIPJET program were hot bent. Hence, the subject stainless steel piping was furnished in the solution annealed condition g prior to welding. Welding-induced sensitization was controlled by limiting the weld interpass temperature and weld heat input. The weld interpass temperature was limited to 350 degrees F maximum, and the weld heat input was limited to 50 kJ/ inch 3 maximum. These controls were evaluated to the requirements of ASTM A708 and were judged to be effective in minimizing weld-induced sensitization. In summary, the measures taken to O 4-8

O~ minimize material susceptibility to SCC by controlling the material condition followed the recommendations of Regulatory Guide 1.44 and help to prevent ~ SCC from occurring. S' Since welding does result in a limited degree of material - sensitization-further controls were used to prevent SCC from occurring. Of the three simultaneous conditions required for . SCC, preventing a corrosive environment was used at BVPS-2. The primary controls are provided through adoption of proven water quality standards and pipe inside diameter (ID) and outside 8 diameter (OD) cleanliness requirements. The reactor coolant chemistry is described in the FSAR, Table 5.2-5 which is included in this report as Table 4.2. This chemistry control has proven to be effective in preventing SCC due to the control of oxygen to 7 O' less than 0.1 ppm when operating at temperatures above 180 F and by controlling chlorides to less than 0.15 ppm at all times. The oxygen control is accomplished by using hydrazine to control the initial oxygen content to less than 0.1 ppm when operating above O 180 F and by a hydrogen overpressure during normal plant operation. Chloride control is accomplished by using strict purity requirements for the procurement of reactor coolant chemical additives. Monitoring of the reactor coolant chemistry O assures a corrosive environment will not be formed. Pipe ID and OD cleanliness is controlled in accordance with the i recommendations of Regulatory Guides 1.37, 1.3 8,and 1.3 9. Swipe i O testing of the pipe OD for contaminants provides assurance against externally initiated SCC. In addition, the use of l thermal insulation in accordance with the recommendations of Regulatory Guide 1.36 helps assure against OD initiated SCC. O' Industry experience with these materials has been reviewed in NUREG-0679, NUREG-0691, and NUREG-1061 [4.7 to 41]. SCC has 2 neither been reported nor is it expected under the specified ,p 4-9 .%,w-e--m-c w-w -y w ww -,---w,- ,,r,-eww-ew,----www w-e.-vw--~w =-.w,%. -n+-.-- - - -+ - - -

) operating conditions. Additionally, erosion / corrosion problems as evidenced at Surry-2 are not a problem at BVPS-2 due to the use of stainless steel materials and the geometry of the ) connecting fittings. Based on the strict fluid chemistry, cleanliness, fabrication, and operating controls, and successful plant operating ) experience, the subject lines are judged to be not susceptible to stress corrosion cracking. Therefore, these lines pass this WHIPJET program screening criteria. ) REFERENCES 4.1 Water Hammer in Nuclears Plants, NUREG-0582, July 1979. 4.2 Evaluation of Water Hammer in Nuclear Power Plants, NUREG-0927, March 1984. D 4.3 Compilation of Data Concerning Known and Suspected Water Hammer Events in Nuclear Power Plants, NUREG/CR-2059, May 1982. D 4.4 Evaluation of Water Hammer Events in Light Water Reactor Plants, NUREG/CR-2781, July 1982. 4.5 Prevention and Mitigation of Steam Generator Water g Water Hammer Events in PWR Plants, NUREG-0918, November 1982. 4.6 Requlatory Analysis for USI A-1, ' Water Hammer', NUREG-g 0993, March 1984. 4.7 Pipe Cracking Experience in Light Water Reactors, NUREG-0679, August 1980. 4-10

O 4.8 Investigation and Evaluation g Cracking Incidents in Piping in Pressurized Water Reactors, NUREG-0691, September 1980. 9: 4.9 Report of the U.- E Nuclear Regulatory Commission Piping Review Committee; Evaluation of Potential for Pipe Breaks, NUREG-1061, Vol. 3, November 1984. O^ O / O O 0 O l i O 9: 4-11 y l

TABLE 4.1 ); ~ PWR WATER HAMMER 5YSTEM PRlleutY CAUSES rREVENT!VE MEASURES (*) 0F WATEA HAfstER. DE5IGN PLANT OPEllATION )* Feed-Feedwater Cortrol FCV Oesign Yer1-w ter Valve (FCV),Over-fication (3.6) sitfag & Instati1Ity Unknown and Operator Operating Procedures Error Induced Steam (3.12), operator Sebble Collapse Training (3.11) ) Main Steam Hammer (Valw Include Valve Steam Closure) Closure Loads in Pipe Support and Component Design Basis (3.9) ReIf af Valve' Include Relief ) Ofscharge Valve Dischalge Loads in Pipe Sup-port and Components Design 8 asis (3.10) Steam Water Entrain-Operating Procecures sent, Unknown (3.12), Operator Trafn- ) fng (3.11) Reactor Relfef Valve Include Retief Valve Coolant Ofsenarge Olsenarge Loads in (Pres-Pipe Support and surizer) Components Design 8 asis (3.10) } RHR Vofding Venting (3.37 Operating Procedures (3.12), Operator Training (3.11) ECC5 Vofding Venting (3.3). Operating Procedures Void Detection (3.1) (3.12), Operator Training (3.11) ) 'CVC5 Staan 8ubble Col-Operating Procedures lapse or Vibration (3.12), Operator Training (3.11) Essen-Vofding venting (3.1), Filling Essentfal Cool-tial Filling Essential ing Water (3.4), Oper-Cooling Cooling Water ating Procedures (3.12), Water (3.4), Analysis Operator Training (3.4) (3.11) 5 team Line Voiding 8TP A58 10-2 8TP A58 10-2 Provisions Gener-Followed by 5 teas Provisions (3.13): (3.13): Testing. Keep-l ator SuDele Collapse Top Ofscharge. Ing Line Full. Auto-Short Line matic AFW IntTiation 1 Lengths, External Header (84W Only) (*)f!cfers to section of this report providing details of oreventive measures. 4-12 J

O (Frte BVPS-2 FSAB) TABLE 4.2 REACTOR COOLANT CHEMISTRY SPECIFICATION Electrical conductivity Determined by the concentration of boric acid and alkali present. Expected range is <1 to 40 pMhos/cm g at 258C. Solution pH Determined by the concentration of boric acid and alkali present. Expected values rangg between 4.2 (high boric acid concentration) and g 10.5 (low boric acid concentration) at 258C. Oxygeni18 0.005 ppm, maximum chloridecas 0.15 ppm, maximum e Fluoride (28 0.15 ppe, maximum Hydrogents: 25-50 cc (STP)/kg H O 2 Suspended solids (48 1.0 ppm, maximum 'O ~ pH control agent (Li70H) si o,7 2.2 ppm as Li Boric acid Variable from 0-4000 ppm as B Silicaas 0.2 ppm, maximum 8 Aluminumes 0.05 ppm, maximum Calcium 888 0.05 ppm, maximum Magnesiums 0.05 ppm, maximum O NOTES: 1. Oxygen concentration must be controlled to less than 0.1 ppm in the reactor coolant at temperatures above 180*F by scavenging with O hydrazine or by maintaining the proper hydrogen concentration. During power operation with the specified hydrogen concentration maintained in the coolant, the residual oxygen concentration must not exceed 0.005 ppm. 2. Halogen concentrations must be maintained below the specified values at all times regardless of system temperature. 4-13 9

l ) 's (From BVPS-2 FSAR) l TABLE 4.2 (Conti d ) NOTES: (Cont) i 3. Hydrogen must be maintained in the reactor coolant for all plant operations with nuclear power above 1 MWt. The normal operating 2) range should be 30-40 cc/kg H o. a 4. Solids concentration determined by filtration through filter having 0.45 micron pore size. 5. The specified limits for lithium hydroxide must be established for prestart-up testing prior to heatup beyond 150'5e During cold hydrostatic testing and hot functional testing, in the absence of 2) boric acid, the reactor coolant limits for lithium hydroxide must be maintained to provide inhibition of hr. logen stress corrosion cracking. Upon plant restart, the lithiura hydroxide limits should be established at 180'F. 6. These limits are included as recommended standards for monitoring coolant purity. Establishing coolant purity within the limits shown ) for the species is judged desirable with the current data base to minimize fuel clad crud deposition which affects the corrosion resistance and heat transfer <of the clad. f:D r e O 4-14 ~s O G

O, AUSTENITIC STEEL o RCS o SIS o RBS g' u HYDROXIDES O ABSENT u OXYGEN O ABSENT 1r SULFUR O (reduced forms) ABSENT O v HALOGENS o Chlorides o Fluorides O ABSENT v O TEMPERATURE o Threshold < 180 F O YES sr PASS O Figure 4.1 CORROSION REVIEW 4-15 g

O SBCTION 1 CRACK PROPAGATION ANALYSIS

O A fatigue crack growth analysis is performed at piping locations which have the highest stress (worst-case) based on normal and O

SSE loads. All break locations which could result in pipe whip and jet impingement targeting of essential safe shutdown j equipment are bounded by the these high stress locations. Not all break locations are explicitly analyzed; instead, at least ~O one break location for each pipe size is analyzed. The fatigue crack propagation analysis assumes that a material flaw of a size which exceeds the acceptance criteria of ASME Section XI is present at this piping location. This postulated flaw is then ) analytically subjected to the internal piping loads which occur at each specified location described above. Next, the flaw is analytically examined in order to determine the maximum potential for growth, during the 40-year plant life, into a through-wall g crack of a size whic would, at some further point, present risk of unstable growth resulting in a complete pipe severence. The process for performing the crack growth analysis involves determination of the through-wall stress distributions, a crack O growth. law, and a postulated defect size. These items will each be discussed separately. O 5.1 STRESS DETERMINATION 5.1.1 Thermal Transient Development 'O The WHIPJET Program utilizes ASME Class 1 piping analyses. Thermal transient analysis of ASME Class 1 piping is performed in accordance with the ASME Section III requirements for pipe stress O analysis where the system design specification defines all o 5-1

O thermal transients associated with anticipated plant operating conditions. The resulting piping loads are used for the WHIPJET review of crack growth. The ASME Class 1 piping systems O associated with the WHIPJET crack propagation analysis are the reactor coolant system, the safety injection system, and the residual heat removal system. To facilitate the analysis, the piping system is subdivided into appropriate thermal zones. The temperature conditions j experienced by the piping in each zone'is detailed in the Stress Analysis Data Package which provides input to the pipe stress i analysis. Where appropriate, the transients for each piping zone are reviewed and combined into more generalized transient envelopes. The rates of temperature change are conservatively estimated as stepwise changes for this evaluation. O: 5.1.2 Pipe Stress Evaluation The state of stress at the piping locations selected for WHIPJET O analysis is a combination of piping loads (i.e. forces and moments) and internal thermal induced stresses. The method of analysis used to develop these loads is the same analytical approach which satisfies the ASME III Section NB 3600 Code O: requirements for Class 1 piping analysis. Therefore, for those piping segments which are Class 1, the pipe stress data provided as input to the WHIPJET piping review come directly from the final SWEC pipe stress calculation results. The piping loads are evaluated using a SWEC computer program NUPIPE, which uses a finite element stiffness method of analysis. In the analysis, the effects of individual static and dynamic loading conditions are input and a set of forces and moments are developed for each piping location. Thermal expansion (including anchor displacement), piping deadweight, and the seismic inertia (including seismic induced anchor movements) are the loading conditions analyzed. 5-2 g - - -, _ - _ - =_-_

( D A11 of the piping thermal load cases are examined in order to determine the dominant thermal case which represents the most likely thermal state of the piping. The forces and moments of this representative thermal expansion load case, the corresponding forces and moments for the deadweight load case and the system pressure are used to determine the typical steady-state stress condition. J For ASME Section III NB 3600 thermal transient analysis a SWEC computer program "BTLOAD" which uses a one-dimensional finite difference method of calculating piping through-wall temperature ) variations is used. The resulting linear ( ATy) and nonlinear ( A T ) portions of the temperature differential predicted to 2 occur in the pipe wall resulting from a given change in internal fluid temperature due to a thermal transient is computed. The O ATy term produces the linear portion of the through wall stress compressive on one side, tensile on the other, depending on the direction of the temperature change. The nonlinear AT term 2 O produces the inside skin stress which is seen to become negligible a short distance from the inside piping surface and would contribute minimally to the overall through-wall stress in the crack growth evaluation. This is discussed in greater detail next. O In addition to these terms, the gross discontinuity stress (related to TA and T ) would be considered at any structural and B () material discontinuities such as branch connections and dissimilar metal interfaces. HTLOAD addresses these discontinuities using a conservative one dimensional analysis which predicts a peak value for the temperature differential O existing between the two dissimilar materials or geometries. The overall effect of the gross discontinuity stress is included in terms for simplicity. At the high stress the AT1 and AT2 contribution to through-wall locations of interest, the T -TB g O stress is significantly less than the AT1 and AT2 eff*Ct-O 5-3

r O 5.1.3 Throuch-wall Stress Distribution Approximation The pipe through-wall stress distribution needed to grow a defect O by fatigue is characterized by a tensile stress at the inner radius and a lower tensile or even a compressive stress at the outer radius. The through-wall stresses used in this program are approximated from the NUPIPE piping analysis results for the e appropriate thermal transients. A linear stress distribution, assumed for simplicity, is broken down into membrane and bending stress components following the approach used in Appendix A of Section XI of the ASME Boiler and Pressure Vessel Code. O The determination of the through-wall stress distribution involves two Section III stress analysis approximations. The first item is the use of "K" indices for peak stress O concentration, and the second is the breakdown of the true thermal distribution into a linear through-wall bending moment and a nonlinear portion which has a zero average value of stress and bending moment but account for an increased inner wall stress addition (which falls off very rapidly). 5.1.3.1 "K" Stress Indices O The "K" indices are essentially stress concentrations of the inner pipe wall used for the internal pressure, moment loading, and thermal loading components of the Equation 11 (NB 3600) peak stresses. The breakdown and definition of terms are as follows: P D o o 1) Pressure: K1 C1 ------ (P ) 2t g O Do Mi e 2) Moment: K C 2 2 (M ) 2I y 5-4 9

O 1 3) Thermals K3 E a AT1 (DELT ) 2(1 - v) 1 4) Thermal: K T T 3 C Eab aA- "b B G 3 (T -T ) g B 1 Ea AT2 5) Thermals (DELT ) (1 ~ VI 2 where: Po = internal pressure }) Do = outside pipe diameter t = pipe thickness Mi = bending moment I = moment of inertia

c) v = Poisson's ratio E

= Young's modulus a = thermal coefficient of expansion a = thermal expansion coefficient for material A S O ab = thermal expansion coefficient for material B AT1 = linear temperature distribution AT2 = nonlinear temperature distribution

O TA = cange of average temperature on side A of discontinuity TB = range of average temperature on side B of discontinuity The "C" indices are secondary stress indices which account for

() more global stress discontinuities. The "K" indices are stress l scaling factors which account for local geometric discontinuities at welds or transition welds in Class 1 piping. The factors which make up the indices are a function of reducer / transition O length, pipe diameter and weld condition as shown in Table 5.1, O 5-5

0: which is Table NB-3683.2-1 from the ASME Code, Winter 1972 Addenda. When the weld is flush with the base metal, the "K" indices are less than the "as-welded" case. None of the analyzed O high stress locations were associated with either curved pipe, branch connections or butt welding tees. Therefore, no " global" geometric effects to the "K" indices are experienced beyond the inside pipe walls. O When the geometry of the weld or base metal transition zone is changed (e.g. a postulated crack 10% through the wall) the underlying assumptions which validate the "K" stress indices are O altered radically. Foremost of the assumptions pertain to the fundamental characteristic of a peak stress that is very local in nature and affects only a small volume of material. In addition, the crack propagation (fracture mechanics) methodology converts the global piping stresses to local, inside-wall stress intensity factor values at the crack tip, thereby accomplishing the intent of the surface-effect "K" stress indices. Therefore, the "K" indices do not provide a meaningful measure of actual stress conditions for the case involving postulated part-through-wall cracks. The "C" indices are maintained as described in Section III of the ASME Code. O In summary, the reasons for minimizing the effect that "K" indices have on Equation II peak stresses are: 1. "K" indices are " surface effect" stress intensifications. g WHIPJET postulates fatigue cracks which begin at greater than 10% through the wall. 2. The local geometric discontinuity resulting in any of g l the prescribed "K" values is not significant once a crack has formed and penetrated an appreciable distance l through the weld or base metal. O 5-6 e

) 3. The fatigue crack propagation (fracture mechanics) method used in the WHIPJET program intensifies and localizes stresses at the postulated fatigue crack tip. Nonlinear Stress Effect 5.1.3.2 AT2 As indicated earlier, the actual through-wall non-linear D temperature gradient is modeled in Equation 11 calculations by two separate thermal parameters. One is a temperature difference between the outside surface and the inside surface based on an equivalent linear distribution and is termed DELT. The second 1 ) is a temperature difference between the surface and the equivalent linear distribution. Figures 5.1 and 5.2 give a graphical representation of this discussion. O The LT2 (DELT ) stress component is a non-linear portion which 2 has a zero average value and a zero first moment with respect to the mid-thickness. If integrated over the entire wall thickness, the net effect to stress would be zero. Its effect is seen most ) dramatically on the inside wall surface and falls off sharply as the outer wall is approached. is computed as follows; it is the maximum The value of DELT2 3 of: 0.5 DELT 1. T -T o 1 3 0.5 DELT1 2. Ti-T 3. 0 3 where: T = Average temperature distribution e Ti = Inside surface temperature To = Outside surface temperature O 5-7

O is a fraction of the ATy (DELT ) value On the inside wall, DELT2 1 and is an additive stress. Near the mid-thickness region, the DELT2 stress, while remaining a small fraction of DELT, becomes 1 O a stress which counteracts the DELTy stress. Near the outside wall, the DELTy stress is compressive and DELT2 makes it less compressive. Neither stress greatly adds to crack growth at the outside wall. O For example, assume the inside peak DELTy stress is 10,000 psi. The corresponding DELT2 stress is typically 2,000 psi (see Figure 5.3). The total inside thermal stress is 12,000 psi. At the midpoint, the DELTy stress is 0 while the DELT2 stress is compressive by a value of approximately 2,000 psi. At the 0.60t (t = wall thickness) point (the maximum crack growth allowed in the WHIPJET program), the total stress is -1000 psi. O The analysis used in the WHIPJET program assumes that the DELTy stress contributes to the total through-wall stress as indicated in the ASME Code and shown in Figure 5.2. However, the DELT2 g-stress is an additive tensile stress in the first 0.25t only and is, in fact, a stress which tends to close fatigue cracks beyond that point and up to about 0.75t. A value equal to 50% of the l inside wall peak stress (see Figure 5.4), taken up to 0.25t DELT2 as an additive tensile stress, is conservatively assumed. Beyond this point (i.e. f rom 0.25t to 0.60t), a value of [10% x DELT ] is used for the stress distribution. Again, because the 2 fatigue cracks are postulated at a part-through-wall depth g. greater than 0.10t, the crack tip does not experience the skin stress value. In other words, WHIPJET effect of the DELT2 assumes that the DELT2 stress adds a tensile stress, which would tend to grow the fatigue crack, across the entire pipe wall. g. These approximations are very conservative, but were used for all of the high stress locations analyzed. O I 5-8 g

D 5.2 CRACK GROWTH LAW The fatigue crack growth analysis of an austenitic weld or base I metal defect is based on compiled experimental crack growth data. The Battelle/EPRI data base management system, EDEAC [5.1] provides this information. For the present study, the crack growth correlation of James and Jones [5.2] was used. This work I was based on the EDEAC data and was recently presented at the ASME B&PVC Section XI Working Group on Flaw Evaluation. The crack growth relation for 304, 316, and weld metal is given by: D da/dn = (F) (C) (S) ( AK)3.3 where: F is a frequency factor C is a temperature factor; and S is an R-ratio factor determined as: S = 1.0 for R<0 or S = 1.0 + 1.8R for 0 < R < 0.79 or S = -43.35 + 57.97R for 0.79 < R < 1.0 3 This da/dn relationship describes incremental crack growth,

where, AK is the difference between the maximum and minimum values of the stress intensity factor (K

-Kmin) and R is the 3 max The temperature factor, C, is set as a ratio Kmin / Kmax. constant corresponding to 600 F -- 2.0E-20; units for da/dn are in./ cycle and for A K units are psi-sqr(in.). The da/dn least squares fit is dependent on many factors, and environment is not directly included. To account for a water environment with a low (<50 ppb) oxygen content, the value for F was taken as 2.00 at 550 F, whereas in air at temperatures less than 600 F it would be D 1.00. For comparison purposes PWR environment data [5.3] were p 5-9

O. compared with this crack growth law as shown in Figure 5.5. The comparison shows good agreement when F = 2.0 as used in the WHIPJET calculations for approximate R ratios of 0.3 to 0.7 for the high stress locations. 5.3 STRESS INTENSITY FACTOR CALCULATION O A two pronged approach is taken in calculating the linear elastic l l stress intensity factors. Values of K and Kmin are computed max l to determine AK for use in the crack growth law described in O Section 5.2. These values are approximated using cylindrical i pipe methods (using both tensile and bending stresses) which promote the growth and extension of an interior semi-elliptic surface crack. The through-wall stress distributions for a given 4 set of transients are used to determine Ky using the methods of i Raju and Newman [5.4]. The Raju and Newman analyses determines the stress-intensity factor using influence coefficients derived from three-dimensional finite element elastic stress analysis for O a wide range of semi-elliptical surface cracks on the inside of cylindrical vessels. Most importantly, these solutions can be superimposed to obtain stress-intensity factor solutions for pressure and thermal-related loads. For the WHIPJET program, the O Raju and Newman stress intensity factor solutions were used to correlate and predict fatigue crack propagation rates under various loading conditions. O, 5.4 INITIAL CRACK SIZE Section XI of the ASME Code, Table IWB 3514-1 was used to obtain the initial (allowable) surface crack size. The conservative assumption is that the crack aspect ratio (2c/a) = 6 remains constant as the crack grows radially through the wall is used. For example, for a defect in an austenitic stainless steel pipe g 5-10 0

D with an aspect ratio of 2c/a = 6, the normalized crack depth (a/t) is 0.110. Here the radial depth is a, the circumferential half length is c, and the wall thickness is t. 5.5 ACCEPTANCE CRITERIA FOR FATIGUE CRACK GROWTH D For each fatigue crack propagation analysis performed on a part-through-wall defect, certain allowable acceptance criteria must be satisfied. The first criteria are for the maximum allowable defect depth and the second criteria are concerned with the maximum defect length. After each transient cycle, the crack grows and the remaining pipe wall thickness (ligament) is compared to the crack-induced 3 plastic zone size. The plastic zone size (rc )is determined from the following relationship: 1 AK 3 ( ----- )2 r c-8 II 2 yc where: 3 B= 6 for plain strain, K -Kmin' AK = max oyc = cyclic yield stress (approximated for stainless steel as twice the monotonic yield strength at operating 3 temperature). If the plastic zone size is greater than the remaining wall g ligament, the fatigue crack propagation computer calculation terminates and a warning statement is issued. As previously stated, this calculation is performed after each cycle. In addition to the plastic zone size criterion, the fatigue crack e depth must not grow beyond 60% through the wall. To verify G 5-11

t O compliance with this requirement, the 40-year life cycle crack growth for each part-through-wall crack was checked to ensure it has not grown more than 60% through the pipe wall. This .8 requirement was always met, as well as the plastic zone size check. There are 2 criteria to be satisfied with regard to the crack O length. The length of the fatigue crack must be less than both of the following: i 1. Critical size through-wall crack (1.414x(Normal +SSE) ] 9 2. Critical size through-wall crack (Normal +SSE) 2 4 Again, each defect length is compared to these two crack criteria atter calculating the 40-year crack growth. For calculations in this program, all of these criteria were satisfied. O. 5.6 EXAMPLE 4 In order to show the methodology used to determine the linear through-wall stress distribution in a BVPS-2 Class 1 line, the l 12-inch SIS system will be used as an example of the methodology. i y l 5.6.1 Stress Analysis Requirements First, a location is identified within the WHIPJET program as requiring a fatigue crack growth analysis based upon the highest stress within that particular system. The following I documentation is gathered to identify the applicable stress information: O l l 5-12 e i l

D 1. A high stress location table which includes the p applicable node, pipeline, stress calculation number, and usage f actor (see Table 5.2).

0 j

2. The internal' loads at the break location resulting from deadweight, thermal, and SSE loads (see Table 5.3). O 3. The contributions to the NB 3600 stresses at the locations, and the stress indices (see Tables 5.4 and 5.5). 5.6.2 Load Ranges Node 301 has four load cases which contribute to cumulative fatigue usage during the 40-year plant life. The transient pairs .O come from the five separate transients listed below: 1. Loss of Flow in Idle Loop (576 F to 498 F - Load Case 32) g 2a. Initiation of Cooldown (Part 1) with RHR flow through SIS Nozzle (Load Case 33). iO 2b. Initiation of Cooldown (Part 2) with RHR flow through SIS Nozzle (350 F to 70F to 350F at 66F/ minute - Load Case 34). O 3. Inadvertent RC System Depressurization (Load Case 35). 4. Loss of Load (Load Case 37). iO 5.6.3 Equation 11 Stresses and Through-wall Stress Distribution The stress distribution is a maximum at the inside wall and o decreases linearly to a minimum at the outside wall. The inside O 5-13

O wall peak stress is unique for each load range and is calculated using the actual NUPIPE Equation 11 peak stress component values (see Table 5.4) for the pressure and moment stresses. For conservatism, the highest system pressure and moment stresses were then used for each transient. In each case, the Class 1 pressure and moment peak stresses were determined with "K" stress indices and without the "C" stress indices as described earlier. stress is a skin Also, as discussed earlier, because the DELT2 effect, a fractional value of the DELT2 stress was used in calculating the inside wall peak stress. The DELT1 and DELT2 stresses were determined directly from the HTLOAD calculation, O effects were enveloped without the "K" indices. The T -TB A terms. The steady-state condition within the DELT1 and DELT2 stress was determined from the internal loads shown in Table 5.3 and the pressure loading. For this example, the total inside wall stress (S ) for the load i i case 33-35 is: O Si=PO+My + DELT1 + DELT2 Si = 13,706 + 12,424 + 3667 + 1892 Si = 31,689 psi (tension) O These values are calculated as described in Section 5.1.3.1. This stress is a summation of the individual and/or system-high peak stress components for load case 33-35. All "C" and "K" stress indices (see Table 5.5) are included in the pressure and moment terms and the full DELT2 value is used. By contrast, the WHIPJET program total inside wall stress for load case 33-35 is: Si=PO+MI + DELT1 + 0.5(DELT ) O 2 Si = 13,706 + 12,424 + 3667 + 946 Si = 30,743 psi (tension) O 5-14 0

) This stress is a realistic and conservative estimate of the actual nodal stresses for this particular transient condition as required for a fracture mechanics crack growth analysis. A ) similar technique was used for each load pair. The stress distribution through the pipe is modeled to decrease linearly with a discontinuity at 0.25t (t = wall thickness). The ) 0.25t point corresponds to the location where DELT2 changes from a 50% to a 10% maximum value; recall that this approximation for DELT is conservative (see Figure 5.4). The linear decrease from 2 30,743 psi to about 29,000 psi is due to the linear decrease in ) the DELT1 stress component. These stresses and the value at the 0.60t point (which is the maximum allowable crack depth) can be seen in Figure 5.6. ) The final stress manipulation concerns the bending and membrane (tensile) components as suggested in Section XI of the ASME Code, Appendix A. This approach has been shown to be conservative [5.5]. To find the bending component, the stress at the 0.50t 3 I, point is needed. This stress is then subtracted from the inside wall stress determined above. For the SIS line, 2 separate bending' stresses are required. One for the 50% DELT2 stress and cne for the 10% DELT2 stress. When the 50% DELT2 stress is used, 3 the stress at 0.50t is 27,076 psi (tension):

1. e., the 26,316 psi shown in Figure 5.6 at 0.50t plus 760 psi (the difference between 10% and 50% DELT ).

Therefore, the bending stress is: 2 D Sb = 30,743 - 27,076 Sb = 3667 psi 9 The remaining tensile stress is taken to be the membrane stress (S,). In this case, the membrane stress is the stress at 0.50t: 27,076 psi. A similar argument is used for the case for a 10% DELT2 stress value (i.'e. beyond 0.25t); the corresponding Sb and D S values are 3667 psi and 26,319 psi, respectively. m D 5-15

O 5.6.4 Crack Growth The next step is to take the linear stress distributions for each load range, apply the change in stress to the postulated part-through crack, grow the crack for that load case, and analyze the next transient. After all cycles during the 40-year plant life are analyzed, this process determines the total fatigue crack O growth. In the case of the SIS line, the four load cases were analyzed in a random combination manner (using a random number i generation procedure) during the 40-year plant life. The change in stress is the difference between the peak inside wall stress e described above and the steady-state stress conditions existing in the pipe wall prior to the transient. The steady-state stress is found by combining the loads f rom Table 5.3 in accordance with the ASME Code, Section III, thereby maintaining conservatism. Results for the SIS sample calculation are shown in Table 5.7. For this sample case with a relatively high usage factor, no significant crack growth occurs during the 40-year plant life. g None of the acceptance criteria described in Section 5.5 are violated. O 5.7 FATIGUE CRACK GROWTH RESULTS The results for all lines analyzed for fatigue crack growth are listed in Table 5.8. All lines pass the crack growth acceptance g criteria. Results by Westinghouse for a generic surge line (not BVPS-2 specific) indicate that fatigue crack growth is limited to less than 40% of the pipe wall (5.6]; the Westinghouse analysis was performed in much more detail. The WHIPJET results for the g BVPS-2 surge line are within 15% of the Westinghouse results. O 5-16 g

D REFERENCES 3 5.1 Data Base for Environmental Crack Model Development, maintained for EPRI by Battelle Columbus Laboratories.

5. 2. L.A. James and D.P.

Jones, " Fatigue-Crack Growth 3 Correlations for Austenitic Stainless Steels in Air", Predictive Capabilities in Environmentally Assisted Cracking, edited by Ravi Rungta, PVP-Vol.99, ASME Winter Annual Meeting Special Publication, November ~) 1985. 5.3 W. Bamford, American Society of Mechanical Engineers Journal of Pressure Vessel and Technology, New York, J New York, February 1979, p. 73. 5.4

Raju, I.S.

and Newman, J.C., Stress Intensity Factor Influence Coefficients for Internal and External J Surface Cracks in Cylindrical Vessels", Aspects of Fracture Mechanics in Pressures Vessels and Piping. PVP-58, American Society of Mechanical Engineers, New York, New York, July 1982. D 5.5 J. M. Bloom and W. A. Van Der Sluys, " Determination of Stress Intensity Factors for Gradient Stress Fields," Journal of Pressure Vessel Technoloov, Vol. 99, No. 3, 3 August 1977, pp. 477 - 484. 5.6 S. A. Swamy et al., " Leak-Before-Break of PWR Auxiliary Piping," EPRI Research Project 1757-55, in press. D 6 'O 5-17

O TAB E 5.1 SECTIm III SIRESS INDICES TABE O TAst.E N8 3683.21 STRESS INDICES FOR USE WITH EQUATIONS 9.10. AND 11 of N8 3650 0 Internet Moment Thermal l.osding(8 3 1.oeding Pressure Con ~nocent 8, C, K, B, C, K, C, K, C,' Straig7t pioe. remote from welds or other dis:entinuities 0.5 1.0 1.0('I 1.0 1.0 1.0 1.0 1.0 O Girth butt v. eld between svaigit oice or bermeen pipe and butt welding componentsI8I (al flush 0.5 1.0 1.18I 1.0 1.0 1.1 - 1.0 1.1 0.5 I (b) u w eided t>3/1G" 0.5 1.1 1.2('I 1.0 1.0 1.8 1.0 1.7 0.5 (c) es welded t<3/16* 0.5 1.1 1.2('I 1.0 1.4 2.5 1.0 1.7 0.5 Girth fillet eld to socket weld fittings slipen fis7ges, or socketwiding g' fiseges(88 0.5 2.0 3.0 1.0 1.5 2.0 1.8 3.0 1.0

• Lqtngitudinal butt welds in straight pipe 'I I

til flush 0.5 1.0 1.1(83 1.0 1.0 1.1 1.0 1.1 tbl as welded t>3/16" 0.5 1.1 1.2( ' I - 1.0 1.2 1.3 1.0 1.2 (c) as welded t<3/16" 0.5 1.4 2.5('I 1.0 1.2 1.3 1.0 1.2 Taoe ed transition joints per N3 3SSI.2 anc Fig. h B.4233-1('! 0.5 1.4 1.5 1.0 1.2 1.8 1.0 1.5 1.0 0 Br'an:5 con ectio s cer 'JB.3643I 1.Q 2.0 1.7 ('I ('I ('I 1.8 1.7 1.0 Curved cIpe or butt-we: ding elbo. s per 24 r *I I ANSI B16.9. ANSI 816.38 1.0 1.0('I (*I I*I 1.0 1.0 1.0 0.5 l or MSS SP48( el 2(AV) Buttmelding-toes per ANSI 816.9 or MSS SP48(I 1.0 1.5 4.0 ('I I'I 1.0 1.0 1.0 0.5 g] But.v.elding reducers per ANSI 816.9 or MSS SP48(3 1.0 1.5 2.0 1.0 1.3 1.0 1.0 1.0 0.5 O l e O t 5-18 l

'O TABLE 5.2 POSTULATED BREAK IN THE SIS SYSTEM O BREAK PIPELINE NODE TYPE OF REASON FOR NUMBER NO. BREAK BREAK 2 SIS-016-C-C 2 SIS-012-288-1 301 Circum-Terminal ferential End 30 STRESS' USAGE FACTOR MAX NOR. OPER. CONDITION CALCULATION T(F) P (PSI) 12241-NP(B)-X70B 0.7739 636 2327 O 0 TABLE 5.3 i BREAK LOCATION INTERNAL LOADS O REFERENCE NUPIPE RUN NUMBER: 4140 DATE: 2/10/83 !O NUPIPE INTERNAL LOADS AT BREAK POINT (1) LOAD CASE NO. F F F M M M x y g x y z !O DEADWEIGHT 2 2807 579 38 1031 -4231 -2405 THERMAL 27 -737 3706 4866 -38393 -22826 38127 SSE 10 1894 1109 1676 2847 10799 90456 'O (1) Forces are in Ibf, Moments are in ft-lbf 'O 5-19

O TABLE 5.4 O PIPING STRESSES Equation.11 (Peak Stress, component) O 2 P M DELT DELT TOTAL (P i). (Psi) (* Ff (* Ff (psi) (33-35) 1700 6955 51.6 24.8 21,332 I (34-37) 13706 12424 22.1 11.4 31,689 (33-34) 0 3314 31.4 13.1 10,701 (32-33) 12103 5004 12.7 5.3 20,095 e: O TABLE 5.5 0-STRESS INDICES K K NODE C C C K 1 2 3 i 2 3 301 1.10 1.00 1.00 1.20 1.80 1.70 O i 5-20 0

O TABLE 5.6 O SIS LINE CRACK GROWTH RESULTS INITIAL FINAL INITIAL FINAL O a/t RATIO a/t RATIO CRACK DEPTH CRACK DEPTH 0.11 0.123 0.143" 0.171"

O

O TABLE 5.7 BVPS-2 FATIGUE CRACK GROWTH RESULTS O

SYSTEM LINE SIZE a/t FINAL SURFACE (INCHES) INITIAL FINAL CRACK LENGTH O SIS 6 0.117 0.1295 -0.56 RCS 8 0.114 0.1267 0.69 !O RHS 10 0.111 0.1120 0.76 RHS 12 0.110 0.1107 0.87 SIS 12 0.110 0.1224 0.96 O RCS 14 0.109 0.1495 1.26 P 5-21

O e L Outside I o S"d \\ e---. T (y) ---* e--- Tm Vi2 y f+ p 7 M 12 -- - - - - - - - - - - - + Midthick ss,,, Inside

• Tg ---,

V /2 --* +- Surface Figure 5.1 DECOMPOSITION OF TEMPERATURE DISTRIBUTION RANGE g 9 k.aTz I i i O l O l l Tecnerature Distribution A T, < ATg 6 1 e Figure 5.2 TEMPERA'IURE DISIRIBUTION 5-22 9

o-o o CONTRIBUTIONOFDELIIANDDELI210THERMALSTRESS 12,50 \\ 8 10,00' Combined ATt and AT2 Stresses 7,5g. s 5,00 " o AT2 2.50 Stress "N._ Contribution %, ~~' 0,00 -2,50

o

~ I h ATt Stress -5,00 Contrib. -7,50 " o ~ e, -10,00~ ~I'),00 20.00 40,00 60,00 80,00 100,00 0 WALLRADIUS(x) o O Comparisons of AT, ATz, and Their Combination Figure 5.3 1

O 5-23

o e. e l#11PJETDELTAT2STRESSAPPR0XIMATION 100,00-80,00 S 60,00 T % of f 40,00 % W EP E Approximation l l 20,00 - 0.a -20,00 Actual AT # -4g,00 2 j -60,00- 'Y,00 20.00 40,'00 60,00 80,00 ~ PIPERADIUS(%) i e-e e. ) Figure 5.4 Approximation Used for AT2 in the WIIPE Program ~ i 5-24

O O 10 100 ' r I I I I l ll1 I I I 1.4 /l ~ key t=o.2 R=o.7 ) O ~ C D 4 10 2 m o-R O FotGEo 304 O E a g roncEo sisu o 10.. T=550 F (288 C) h } P=2000 psi (14MPa) I alc ,e "i" d-10-3 U W 5 5 " = F C S (AK)3.3 [ g g dn g 5 F 5 F=2. =o 'O 10-5 g ,y y g e o 10" O Bamford, ASME JPVT Feb. 1979 p. 73

o.6 I

I i 1/ I l l/ I I I I I Ill O 10 100 aK (KSI M ) O Comparison of Actual PWR Environment Data Figure 5.5 with the Austenitic Stainless Steel Crack Growth Law for Various R Values O !O 5-25

O 8 IHROUQl-WALL {IRESS DISTRIBUTION FOR 12-INCH SIS LIN 36.N e g 32.00-I ~- e S g,o.25e S 28.00 S 's I 24.00" N N l e .00 20.00 40.00 60.00 80.00 100.00 WALL RADillS (x) e-e I e Figure 5.6 'Ihrough-wall Stress Distribution 5-26

LO gacTIou 1 MATERIAL PROPERTY DATA ,O i 6.1 AUSTENITIC STAINLESS STEEL LINES + Tensile true stress-strain and J-resistance curve data for the applicable BVPS-2 stainless steel material (Type 316, 304, and welds) have been collected. The stainless steel material data for both base (Type 316 and 304) and welds (both '&hielded metal 0 arc and submerged arc) were gathered from numerous engineering soarces (see Appendix B). O. The stainless steel base metal information, derived from previously published data, was required for analytical purposes. The WHIPJET program required stress-strain and J-R curve i properties. These data, along with the piping certified material test reports (CMTR), show that the BVPS-2 BOP high energy lines O are made of high grade material, extremely resistant to unstable tearing. The characteristics of the true stress-strain diagram and the J-R curves were used forsleak rate calculations and crack stability analyses. Appendix B also shows how the industry data O were evaluated to derive appropriate stress-strain and J-R curve properties. The stainless steel shielded metal arc weld (SMAW) and submerged 9 arc weld (SAW) metal information was also derived from existing data. Again, the properties were required to perform leak rate and crack stability analyses for BOP high energy piping welds. O Appendix D discusses the weld procedures and conditioning of materials used for actual BVPS-2 welding. Some SAW shop welds were employed in the RCS 14-in. line, the SIS 12-in. lines, and the RHS 10-in and 12-in. lines. All other welds (shop O O 6-1

~ O-and field) were SMAW; ' the shop and field welding parameters were essentially identical with the same range of heat inputs and the same filler materials. G 6.2 HIGHEST STRESSES AND MINIMUM MATERIAL PROPERTIES NUREG-1061, Vol. 3 [fiL] indicates that within a particular line only the location of highest stress corresponding to the minimum material properties needs to be analyzed for LBB. This same philosophy was expressed by the NRC staff during an August 27, O 1986 meeting on this program. Therefore, the highest stress locations have been evaluated in terms of the highest bending moment for normal plus SSE loads (combined absolutely and using the SRSS): (M 2+Mz )1/2, n'ormal moments were determined from y the algebraic combination of deadweight and thermal components. The axial loads (including pressure) which would open a circumferential crack (F ) are small and result in small stresses x in comparison to the bending moment effect. In considering the highest stress location coupled with the minimum material properties, the piping node points were not separated by weld types or base metal. Only the highest stress location throughout the line (anchor to anchor) was determined, and this location for each pertinent pipe size and system will be analyzed for all material conditions which exist in that line (base metal, SMAW, and/or SAW). For consideration of longitudinal breaks in a pipe size in which the pressure is a constant, all results will be identical. Therefore, each pipe size and system will be checked for longitud 4,nal break leakage and stability. All stresses and data input are "as-constructed" based upon the g '- overall plant reconciliation program performed by SWEC. st l 6-2 g;

'O REFERENCES -O 6.1 Report g the L L Nuclear Regulatory Commission Piping Review Committee; Evaluation g Potential for P_ipe Breaks, NUREG 1061, Vol. 3, November 1984. 'O -O O .O O

O

.O -O 6-3 .O

-~ TROLE 6.1 BVPS-2 HIGH STRESS LOCRTIONS FOR NHIPJET L88 RNRLYSIS LINE LORDING FORCES (1bs) NOMENTS (ft-lbs) CONDITIONS Fx Fy Fz Mx My Mz SIS 6-INCH OER0 HEIGHT -1067 -10 -5 10 -43 135 THERNAL -16 -843 80 1135 721 11223 SSE 169 172 283 732 2720 1610 4 RCS 8-INCH OERDEIGHT 1115 -1%7 114 285 52 4344 THERNAL -2128 -763 -884 -3884 -4314 11098 SSE 1250 1735 3% 1219 854 1761 RHS 10-INCH DERONEIGHT 7 -2949 9 -7 -179 -4015 THERNAL 48 -2318 -80 2162 714 -31215 mL SSE 2566 3338 2017 5194 4998 7888 RHS 12-INCH OER0 HEIGHT -77 4500 318 -6440 -1555 6306 THERNRL -1319 -2369 -1679 15713 361 -17014 SSE 5210 3639 2077 7147 20757 29201 i SIS 12-INCH OEROWEIGHT 2007 579 38 1031 -4231 -2405 THERNAL -737 3706 4866 -38393 -22826-30127 SSE 1894 1109 1676 2847 10799 9045 RCS 14-INCH OER0 HEIGHT 105 2135 59 -2237 -1729 10360 THERNAL 11592 3479 -14015 2951 161404 47651 SSE 1791 1241 1529 1525 12814 8969 I l O e e e e e e e e e e

lO SBCTION 1 l l LEAK DETECTION lO l f

7.1 INTRODUCTION

Cne of the key parameters needed for leak-before-break (LBB) is the lowest detectable leakage for inside containment. This parameter is then multiplied by a margin for use in the LBB (WHIPJET) analyses. For breaks inside containment, the NRC staff O position is that a value of 10 must be applied for the leak rate margin. This margin on leak detection is hichiv conservative, but meets NUREG-1061, Volume 3 (131] and the NRC broad scope rule change to GDC-4 (out for public comment). O The following discussion focuses on BVPS-2 leak detection systems, leak detection capabilities and operator actions. Values for the limiting detectable leakage are determined. O A diverse number of containment parameters are monitored to detect leakage inside containment. The WHIPJET systems within the containment are RCS, RHS, and SIS. These piping systems are g all within the reactor coolant system pressure boundary as defined in Section 50.2 of 10CFR50, and are identified in this section as RCS pressure boundary. O The BVPS-2 technical specifications define the categories of leakage to be: o 1. Identified Leakage -- Identified reactor coolant system pressure boundary leakage is leakage into closed systems, such as pump seal or valve packing leaks that are captured and conducted to a collecting tank, or O leakage into the containment atmosphere from sources o 7-1 1 ______-_________-_-___-_-__--_____________-___-_N

O 'that are both specifically located and known not to interfere with.the operation of leakage detection system or not to be pressure boundary leakage, or leakage of 9 reactor coolant through a steam generator to the secondary system. 2. Unidentified Leakage -- Unidentified reactor coolant system pressure boundary leakage is all reactor coolant system pressure boundary leakage which is not identified. O 3. Pressure Boundary Leakage -- Pressure boundary leakage is leakage (except steam generator tube leakage) through a nonisolable fault in a reactor coolant system component body, pipe wall, or vessel wall. The technical specifications provide limits for each of these reactor coolant system pressure boundary leakage categories. 9: The leak detection systems are employed to monitor leakages so that any abnormal leakage can be diagnosed well before the technical specification limits are reached. RCS inventory balances are also performed to determine the unidentified leakage from the RCS. Containment entry may be required to determine the source and the quantity of leakage. G: 7.2 LEAK DETECTION SYSTEMS (INSIDE CONTAINMENT) The leak detection systems employed within containment include the following: e 1. Containment sump pump flow and level monitoring, 2. Containment atmosphere particulate and gaseous 3: radioactivity monitoring, and 7-2 S'

~O 3. Containment pressure, temperature and humidity monitoring. .O A brief description of these systems, which are provided to meet the intent of Regulatory Guide 1.45, follows: 1. Containment Sump Pump Flow and Level Monitoring -- This z) system employs a programmable controller which periodically measures leak rate to the containment sump by determining flow rate for the sump pumps and m nitoring the pump operating times. This programmable O controller alarms if there is an increase in flow of one (1) gpm over normal. In addition, sump flow rates are determined and logged on a shift basis by dividing the difference between the current shift and the previous O shift totalizer readings, in gallons, by the elapsed time between reading. Operators in the control room also monitor sump pump running frequency and changes in O the water level of the containment sump to check for any abnormal operation. 2. Containment Airborne Radiation Monitoring -- This system !O monitors both gas and perticulate radioactivity. Indicators and alarms are located in the main control room.

O 3.

Containment Pressure, Temperature and Humidity Monitoring -- Containment pressure, temperature and l humidity values are indicated and recorded in the main i control room. O i 7.3 LEAKAGE DIAGNOSIS !O The leak detection systems are monitored frequently. The containment sump pump discharge is monitored on a shift basis; at lO 7-3 l

o! least one reading every 12 hours is required by technical specifications. The containment atmosphere gaseous and particulate radioactivity is also monitored at least at this 9 frequency. Plant operator logs are reviewed for trending. If the containment sump pump flow shows a step increase of 50% between shift readings or an increase of 50% over a seven day period, an abnormal leakage procedure is followed. If the containment sump pump flow alarms indicating an increase of one (1) gpm over normal, the same procedure is followed. The abnormal leakage procedure calls for an evaluation of the containment parameters to determine as accurately as possible the source and magnitude of the leak. In addition, the procedure calls for an RCS inventory balance which is a controlled test performed to determine the magnitude of RCS unidentified leakage. Operating experience from other plants indicates that the avetage long-term unidentified leakage from the RCS is between 0.1 and 0.3 gpm. Recent data for BVPS-1 indicates an RCS unidentified, leakage rate of 0.32 gpm inside containment. Based upon the above discussion (plus the operator actions described in the next section), the detectable leak rate limit is l 0.5 gpm. This value when multiplied by the NRC margin of 10 gives 5 gpm as the level for crack size determination under normal operating loads. Section 8 on leak rate calculations uses the 5 gpm leak rate level for breaks inside containment. 7.4 ACTIONS g Actions depend on the evaluation of the containment parameters. Actions are based on Table 7.1 and the calculated value for RCS unidentified leakage. e 7-4 e l

O 7.4.1 Analyze / Evaluate Containment Parameters The following actions are performed whenever the conditions of O Section 7.3 are exceeded. 1. Analyze information from control room instrumentation, including a comparison of charging and letdown flow O rates, to determine the nature of the leak. 2. Upon determination,-notify the operating supervisor. 3. Evaluate all information available from control room instrumentation as well as locally mounted instruments to identify the leak and determine its location. O 7.4.2 Perform An RCS Inventory Balance An RCS inventory balance' is performed when the conditions described in Section 7.3 are exceeded as specified by the O abnormal operating procedure. It is a controlled test to determine the magnitude of the unidentified RCS leakage. The accuracy is approximately 10.2 gpm.

O l

It is noted that in addition to being performed when these conditions are exceeded this 4m.ventory balance is performed every 72 hours as required by the technical specifications. The j) following equations summarize the procedure: Uncorrected RCS leakage rate = Leakage rate based on Volume Control Tank level change lO i Corrected RCS leakage rate = (Uncorrected RCS leakage rate) + (Leakage rate correction ( for temperature change) lg + (Leakage rate correction for pressurizer level change) 7-5 o

O Next, the unidentified reactor coolant system leakage is ~ determined by subtracting the leakages that are known to go to collecting tanks as follows: O Unidentified RCS leakage rate = (Corrected RCS leakage rate) - (Leakage rate to pressurizer relief tank) - (Leakage rate to primary drains transfer tanks) 7.4.3 Containment Entry / Inspections Containment entry / inspection is required by operating procedures when unidentified RCS leakage is greater than 0.5 gpm. For leakage 10.5 gpm, action is decided by the Nuclear Shift Supervisor (NSS). If the leakage source is not found on the first inspection, O further inspection will be performed with emphasis on the WHIPJET piping lines and systems. When containment inspection is called for, the source of leakage O. is determined by visual inspection and measured by a graduated cylinder to determine the leakage rate. When the source is specifically located, the leakage is then considered identified. O' The following technical specification limits apply: Pressure Boundary Leakage f rom a RCS Component Body, Pipe Wall, or Vessel Wall - None g RCS Unidentified Leakage i 1 gpm RCS Identified Leakage i 10 gpm 7-6 g

O >7.4.4 Limitina Conditions for Operation Depending on the location and magnitude of the leak (0.5 to 1.0 O gp,),. appropriate repair actions will be taken or the plant will be shutdown. If the leakage is identified as a cracked pipe in the RCS pressure boundary (on the main coolant side of the second check valve; i.e., Class 1 piping) the technical specifications O dictate immediate shutdown and repair. If the leakage is a cracked pipe in the non-pressure boundary RCS (Class 2), 'that system will be isolated (if possible), or it will be evaluated and appropriate repair action will be taken. g In accordance with technical specifications, if the calculated unidentified leakage is greater than 1 gpm or identified leakage is greater than 10 gpm, the leakage must be reduced to acceptable O limits within 4 hours or be at least in HOT STANDBY within the next 6 hours and in COLD SHUTDOWN within the following 30 hours. In accordance with technical specification, if there is any g pressure boundary leakage from a RCS component body, pipe wall or vessel wall, the plant muet be in at least HOT STANDBY within 6 hours and in COLD SHUTDOWN within the following 30 hours. O Evaluations will consider the source and magnitude of the leak, rates of change of detection variables and if shutdown is required. The evaluations will be used to determine shutdown O rates and condition. Action taken will be z.ecorded in a written log. After shutdown, corrective action wil] be taken before operation O is resumed. O 0 7-7

O REFERENCES O 7.1 Report o_{ the L L Nuclear Regulatory Commission Piping Review Committee; Evaluation p_f Potential for Pipe Breaks, NUREG-1061, Vol. 3, November 1984. O i l 9 O' O O i O O O 7-8 g

D TABLE 7.1 EFFECT ON CONTAINMENT PARAMETERS VS. THE TYPE OF LEAK p Containment RCS Secondary Component Auxiliary Parameter Leak Leak Cooling Cooling (Feedwater Water Water D or Steam) Leak Leak Air Particulate Increase No No No and Radioactive Change Change Change D Gas Monitors Containment Increase Increase No No Humidity / Dew-Change Change 3 point Temper-aturee Containment Increase Increase Increase Increase O Sump Flow rate Containment Increase Chromates E Sump Chemistry Activity . in Sample Analysis RCS Water Net No No No O Inventory Loss Change Change Change 0 D 7-9 D

Cl secTIon 1 L LEAK RATE CALCULATION

SUMMARY

O' The WHIPJET program.used the EPRI PICEP (Elpe grack Evaluation Program) [giL] computer code to analyze pipe leakage. The PICEP 8: i . code calculates the crack-opening area, the limit load critical crack length, and flow rate for through-wall cracks in pipes. The cracks can be oriented either axially (longitudinally) or circumferential1y. Crack opening area was calculated using O elastic-plastic fracture mechanics methods similar to those used later in the WHIPJET program for determining crack stability; leak rate calculations employ two-phase flow based on homogeneous i non-equilibrium critical flow theory. The extensive validation-results for PICEP are presented in Appendix E. The accuracy of PICEP is judged to be 1254. 1 Each applicable BVPS-2 line in WHIPJET (see Table 6.1) was S.' ' analyzed using the PICEP computer program to determine leak rate { as'a function of crack size. The forces and moments used to calculate leak rates were combined algebraically (. ); for the 1 leak rate determinations, normal operating loads were considered f to be the combination of deadweight, thermal, and pressure. All of the piping thermal load cases were examined in order to determine the dominant thermal case which represents the most likely thermal state of the piping during normal steady-state 9 operation. Only those loads which resulted in an opening of the crack were used (i.e., torsional moments were ignored). The PICEP code also required a stress-strain relationship to be used for determining the crack-opening area needed to calculate leak rates. Results obtained by Combustion Engined. ring [1d[] indicate that the crack-opening area in the case of circumferential welds is governed better by the stress-strain 8-1 g

O - properties of the base metal. Therefore, all leak rate calculations used base metal properties for estimating crack-opening areas; this was also true for the longitudinal welds O. since the size of a crack extends appreciably outside the small . width of the weld.itself. Additionally, since lower bound stress-strain properties result in larger crack-opening areas (and subsequently higher leak rate estimates), best fit stress- ) strain properties for the base metal were used for conservatism. (see Appendices B and C). Results f rom the leak rate calculations - are presented in Appendix ) F as a series of leak rate versus crack size plots. Based upon the limiting detectable leak rate of 0.5 gpm inside containment plus the leak rate margin of 10 (required by the NRC staff), a

g leak rate of 5 gym was used to determine a detectable leak rate (with margin) crack size for crack stability analysis.

Therefore, the leak rate results in Appendix F were analyzed to determine the crack size corresponding to 5 gpm for normal i

O perating loads.

These results are tabul,ated in Table 8.1 fcr the stainless steel lines inside containment. The next sections l of this report will then use these results for evaluating crack i stability. lO ~ REFERENCES I o 8.1 D. Norris et al., PICEP: Pipe Crack Evaluation Program, EPRI NP-3596-SR (Revision 1), in press. 8.2 D. M. Norris, private communication regarding O unpublished Combustion Engineering work for EPRI. O O 8-2

O TABLE 8.1 LEAK RATE RESULTS FOR STAINLESS STEEL LINES INSIDE CONTAINMENT g LINE SIZE 5 GPM CRACK SIZE (1) (inches diameter) (inches) O CIRCUMFERENTIAL BREAKS SIS 6 4.26 RCS 8 5.67 g RHS 10 7.54 RHS 12 8.36 g SIS 12 6.55 RCS 14 4.88 g LONGITUDINAL BREAKS SIS 6 3.61 O' RCS 8 4.35 RHS 10 8.68 O RBS 12 5.33 SIS 12 5.12 S RCS 14 5.53 NOTE: (1) Crack size corresponding to a 5 gpm leak rate under normal operating loads 8 8-3

D SECTION 9 CRACK STABILITY CALCULATIONS [FOR NORMAL + SSE LOADS] D The methodology to evaluate the stability of through-wall cracks requires knowledge of the applied loads, a leak rate crack size, D and the material properties. The piping axial loads and bending moments were obtained from the NUPIPE stress calculations. The dead weight, thermal, and pressure loads and moments were combined algebraically to define normal operating conditions D (exactly the same as for the leak rate calculations in Section 8). These loads and moments were then combined absolutely with the SSE inertia loads and moments to define the appropriate conditions for assessing crack stability. The appropriate leak rate crack sizes for inside containment were developed in Section 7. For inside containment piping, a leak rate crack size for 5 gpm (which includes a multiplicative margin of 10) was used as required by the NRC staff. These crack sizes were evaluated for stability by applying a further margin on crack size of at least 1.8 as required by the NRC staf f; i.e., 10% less than the value of 2.0 specified in NUREG-1061, Volume 3 [9.1]. The material properties used for crack stability analyses were all lower bound as shown in Appendix B. The NFC staff requested D two sets of calculations to be performed for circumferential welds: first, lower bound weld metal stress-strain and J-R curve properties were used, and second, lower bound base metal stress-strain and J-R curve properties were evaluated and compared to g the weld metal results. The more conservative of the two sets of calculations was used to assess stability. Other combinations of stress-strain and J-R curve properties for base-weld cases could have been used leading to more conservative calculations, but the D g 9-1

O 4

minimum of the two cases run was deemed reasonable considering the large margins already employed in the analyses.

O For the circumferential break locations, crack stability was assessed using the EPRI computer code FLET (Ekaw Evaluation for Tearing Instability). The FLET code is based on elastic-plastic fracture mechanics, and the verification results for this code' gy i - are discussed in Appendix G. The deformation plasticity failure assessment diagram (DPFAD) method for evaluating crack stability was chosen particularly since it allows for crack growth which is not the case for the FLET J-T analysis. FLET utilizes four g 4 separate analytical approaches including the two most common, J DPFAD and J-T. Required input into FLET includes the material stress-strain properties (in terms of Ramberg-Osgood curve fit parameters) and the experimental fracture resistance curve (J-R). gr r = (J / J) / is plotted 8 In the DPFAD approach, the quantity K to define the DPFAD failure curve using the versus Sr " #/UL i material stress-strain and J-R values. J' is the elastic J, J is the total J (which includes combined tension and bending), o is GF is the limit the stress for tensile and bending load, and on load stress defined using the ultimate tensile strength or flow f stress _ as indicated in Reference 9.2. Since none of the DPFAD I curves extend into the fully plastic regime at instability GF l (including crack growth), either stress value gives the same result. Next, assessment points using applied J values are plotted on the G same K and S curve to determine whether initiation or r r instability has occurred. The assessment point values, K ' and r l S', both account for combined tension and bending loads. r Instability assessment points lying inside the DPFAD curve are gf safe, and those lying outside the curve indicate an unsafe, unstable situation for the particular size crack and applied loads. Figure 9.1 illustrates the DPFAD curve for a specific case identified in Table 9.1. 1 OI 9-2

C)- E The circumferential crack results are listed in Table 9.1 for the inside containment lines. The columns in the tables identified IO as " margin on crack size" correspond to the minimum value of 1.8. When the DPFAD margin (for normal + SSE loads) is greater than or equal to unity, _ the crack is stable. It should also be noted that the FLET DPFAD calculations for stability in the 6-in. SIS -() lines involve crack sizes which extend slightly beyond 180 1 degrees around the pipe, which is in the extrapolation regime for FLET applicability. It is felt, due to the conservative use of lower bound material properties, that the 6-in. SIS lines easily .3: pass the intent of the margin on crack size. ( All DPFAD calculations were also checked using the J-T method; } the instability DPFAD stress agrees with 10% of the J-T

O instability stress.

For longitudinal cracks, only pressure loads are important in assessing crack opening area and stability. As shown in Table !O 9.2, the margin on crack size was evaluated by ratioing the critical crack size calculated using the empirical results of Eiber (gel) to the 5 gpm leak rate crack size. Note that these resultant margin numbers are significantly greater than the limit !O of 1.8. If a stability analysis had been performed these numbers would only decrease slightly, especially since failure would occur in base metal for these long cracks which extend well l outside any welds. 'O REFERENCES lO 9.1 Report gi the 1hf2 Nuclear Requiatory Commission Pipinq Review Committeer Evaluation gf Potential for Pipe Breaks, NUREG-1061, Volume 3, November 1984. O 1 9-3

O 9.2 J.M. Bloom and S.N. Malik, " Procedure for the-Assessment of the Intensity of Nuclear Pressure Vessels and Piping Containing Defects," EPRI NP-2431, June 1982. 9.3 R. J. Eiber et al., " Investigation of the Initiation and Extent of Ductile Pipe Rupture," Battelle Columbus O Laboratories Report BMI-1908, June 1971. O O O I O' O O f 1 O' 1 9-4 e.

'C: TABLE 9.1 CIRCUMFERENTIAL CRACK STABILITY EVALUATION FOR INSIDE CONTAINMENT MARGIN ON DPFAD MARGIN FOR LINE SIZE CRACK SIZE NORMAL + SSE (inches) (1) BASE SMAW SAW ,O, =. SIS 6 1.8 1.06 1.60 (2) O RCS 8 2.0 1.00 1.48 (2) RHS 10 2.0 1.25 1.79 >2.0 .O RHS 12 2.0 1.08 1.54 1.45 SIS 12 2.0 1.59 2.31 2.02 .; O RCS 14 2.0 1.43 2.10 1.76 O i 'O O i NOTES: (1) Ratio of DPFAD stability crack size analysis (for Normal + ,0 SSE loads) to the 5 gpm crack size (Normal loads only); minimum value is 1.8 corresponding to DPFAD ratio 11.00. (2) No SAW analysis required O 9-5

---

a y

9-m --_,-w-__ --,m-- --w -eep-p -pw.y, -., _.,_., - -__9 __,y _.ym, .y--y-e y-p9mw -p-- --,u'--w7w - mN 7=wNw

O TABLE 9.2 LONGITUDINAL CRACK STABILITY EVALUATION FOR INSIDE CONTAINMENT O LINE PIPE SIZE RATIO OF CRITICAL SIZE (INCHES) TO 5 GPM LEAK RATE SIZE (1) O SIS 6 3.61 RCS 8 3.76 O RHS 10 9.35 i RHS 12 4.33 SIS 12 4.54 i RCS 14 4.44 e i O O NOTES: (1) Critical size is based upon empirical failure results [M] for pressure loads only 8 9-6

6 o. 2 SIS-12-289-1N0DE449 DPFAD (N0 M L + SSE) CASE IYPE 316 o 1,90 g 0.90" 0,89" unst die aesion o 9.70" 8,69" -o S. 0,50 " Stable k gJ., Region Loading Line g, O / 1,30 " 1 I Assessinent Point 0.29 \\ = o e,18 8-{g 8,88 1,68 2,40 EP IO l I 'O Figure 9.1 DPFAD Diagram for 12-inch Diameter SIS Line .O I 9-7 10 i 1 lO l

O SBCTION 11 CRACK STABILITY UNDER EXCESSIVELY HIGH LOADS O Crack stability for excessively high loads was the final fracture analysis effort performed on the high energy piping systems. Assuming a detectable, but stable, leakage size crack has occurred in the pipe, it was shown-that, if an SSE occurs prior to detection of the leak, this through-wall crack was stable under loads that are significantly greater than SSE. The e' excessively high loads that were applied are 1.414 x (normal + SSE). The stability analysis followed the procedures described in e Section 9 except that no extra margin was placed upon the leakage size crack. The required margin was applied to the loads. Results for the circumferential breaks are listed in Table 10.1. Since all of the DPFAD margins are significantly greater than g. I 1.00, all of the high stress break locations meet the requirements for stability under high loads. Again, as for the l other stability analyses in Section 9, all DPFAD results were in the elastic-plastic regime. For the longitudinal breaks, the approach used in Section 9 was again used except that stability was based upon loads equal to l 1.414 x pressure. The results in Table 10.2 show a significant margin on crack size above 1.00 for these high loads, and therefore, stability is verified. O I e l 10-1 9

^ f TABLE 10.1 EXCESSIVE LOAD ~CIRCUMFERENTIAL CRACK STABILITY FOR INSIDE CONTAINMENT O DPFAD MARGIN FOR 1.414 [ NORMAL +SSE] LINE PIPE SIZE (INCHES) BASE SMAW SAW SIS 6 1.57 2.43 (1)

O RCS 8

1.94 2.95 (1) i RHS 10 2.48 3.68 3.22 10 j RHS 12 2.04 3.05 2.63 SIS 12 2.19 3.31 2.85 l0 RCS 14 1.49 2.27 1.94 f 'O 1

O IO i

4 NOTES: {O (1) No SAW analysis required j

O 10-2 i

n._,--- ,n,-, ~

O TABLE 10.2 O EXCESSIVE LOAD CRACK STABILITY FOR LONGITUDINAL LINES INSIDE CONTAINMENT e LINE PIPE SIZE RATIO OF CRITICAL SIZE (INCHES) TO 5 GPM LEAK RATE SIZE (1) e SIS 6 2.52 RCS 8 2.63 RHS 10 6.56 RHS 12 3.02 g SIS 12 3.15 RCS 14 3.07 l l e e 1 NOTES: e (1) Critical size is based upon empirical failure results for (1.414 X pressure) loads 10-3 e

J ) l t ) \\ APPENDIX & 3 NUREG-0582 FLIIID TRANSIENTS ,] 3 D D e J

O APPENDIX & NUREG-0582 FLUID TRANSIENTS .O This appendix is a list of fluid transients compiled from a review of NUREG-0582 as applied to BVPS-2 The tabulation of transients was extracted'directly from project design documentation and includes systems beyond the NHIPJET scope. ~O Also, when appropriate, references to operation and amintenance procedures (OM - # ) are provided as indications of operational controls intended to mitigate certain fluid transient postulations. .O lO LO l l l O O 1 !O 1 !O l fO ^-1

e BVPS-2 FLOW TRANSIENT APPLICABILITY 9 Title / Description Applicability A. Pump Start with t Inadvertently Voided Discharge ID Lines A1. RHS Not Applicable - To establish a controlled cooldown the system is vented and temperature and pressure between RCS and RHS must be within certain criteria before RHS operation (See Operating Manual II 4 OM-2.10.4.A) A2. HHSI NA - Discharge lines are normally filled and operating as the chemical and volume control system. System is filled and vented in accordance O-with OM-2.7.4.W. A3. LHSI NA - System is normally filled and vented in accordance with OH-2.11.4.H.RWST minimum water level is'above the elevation of the cold leg injection. nozzles, keeping lines full of water. A4. SWS' Applicable-Enveloped by transient of restart of pumps. (See Section F, Transient No. F2). Vacuum breakers were installed to reduce water hammer from restart of pumps. Voids would form as a result of -amp stop and SWS water drain back to ultimate heat g en:k. Before normal start of a service water pump (See Operating Manual OH-2.30.4.A), the system is filled and vented in accordance with OM-2.30.4.F. A5. CCP NA - The CCP System is filled and vented in accordance with OH-2.15.4.L. RHS coolers and RHS g-pump seal water coolers are the only two coolers not normally in operation. The portion of the CCP system to the RHS coolers is kept full since only the inlet motor operated isolation valve is closed during normal operation. I A6. FWE NA - Voids are not expected due to filling and venting procedures after ISI tests. t i I A7. RSS Applicable - RSS discharge lines to spray rings are normally empty. Enveloped by system fill of empty l lines (see Section B, Transient No. B6). Analyzed qy in EMD Fluid Transient Calc. No. 12241 - NP(B)- 316-FA and -260-FIA. t 0 A-2

3 Title / Description Applicability A8. QSS Applicable - QSS discharge lines to spray rings are 3 normally empty. Enveloped by system fill of empty lines (see Section B, Transient No. B7). Analyzed in EMD Fluid Transient Calc. No. 12241-NP(B) - 273-FA and -319-FA. A9.RCS Not Applicable - System is filled and vented in accordance with OM-2.6.4.E or a steam bubble is 3 formed in the pressurizer before a Reactor Coolant Pump is started in accordance with0M.2.6.4.A. B. Expected Flow Discharge into

)

Initially Empty Lines Bl. RCS Salety and Applicable - Analyzed in EMD Fluid Transient Calc. Relief Valve Nos. 12241-NP(B)-328-FA and -517-FA Discharge O B2. RHS Suction Applicable - Analyzed in EMD Fluid Transient Calc. Relief Valve No. 12241-NP(B)-312-FA and -298-FA. Discharge (2RHS*RV721A,B) C) B3. HHSI NA - Lines normally filled and operating as CVCS. B4. LHSI - SIS NA - Lines normally filled with water from the RWST Lines from Voids are not expected due to filling and venting RWST to RCS procedures. (See OM-2.11.4.H). Cold and C) Hot Legs B5. LHSI - Cross Applicable - Analyzed in EMD Fluid Transient Calc. over from RSS No. 12241-NP(B)-287-FA for Longterm Safety In C) jection from the Containment Sump B6. RSS - Normal Applicable - Analyzed in EMD Fluid Transient Calc. Pump Start in Nos. 12241-NP(B)-316-FA and -260-FIA ,3 to Empty Lines and Full Flow Pump Test B7. QSS Normal Applicable - faalyzed in EMD Fluid Transient Calc. Pump Start Nos. 12241-NP(B)-273-FA and 319-FA. f) ~ into Empty Lines O A-3

O Title / Description Applicability B8. MSS Safety Applicable - Analyzed in EMD Fluid Transient Calc. GP Relief Valves Nos. 12241-NP(B)-464-FD and -278-FC. (2 MSS *SV101-105A, B, C) and Atmospheric Dump Valves (2SVS*PCV101A,B,C) Opening e .B9. MSS Steam Sup Applicable - Initially empty lines at ambient ply to Aux. temperature are filled with main steam. Feedwater Pump Condensate pots remove any condensate formed. Turbine Analyzed in EMD Fluid Transient Calc. No. 12241-NP(B)-452-FD. e BIO. CCP Flow into NA - All coolers supplied by CCP system are Empty Coolers normally filled except when down for maintenance. Venting procedures during gradual fill eliminate the possibility of this transient occurring. (See OM-2.15.4.L.) Bil. SWS Flow into Applicable - When the RSS system starts after an Initially Em-accident, service water valves isolating the RSS pty RSS cooler are opened filling the empty lines and Coolers coolers. Analyzed in EMD Fluid Transient Calc No. 12241-NP(B)-381-FA. e-B12. MSS Radiation Applicable - Water slug between monitor and outlet Monitor isolation valve passes through RO into the Operation discharge piping when valve opens. B13. FWE Opening of Applicable - Relief valve opens filling empty RV101 on the line (relieves excess pressure due to overspeed Turbine Pump of the turbine). Analyzed in EMD Fluid Transient Discharge Calc No. 12241-NP(D)-276-FA B14. ASS Flow into Not Applicable - OM-2.27.4.A for the ASS startup Empty Lines includes provisions and cautions to prevent water / gg, steam hammer or thermal shock. C. Valve Opening, Closing, and Instability I' C1. RHR - Closing / NA - Suction and discharge motor operated Opening of isolation valves have an opening / closing time Suction or of 120 sec. No water hammer will occur due to Discharge valve closing or opening. Also, since the dCVs fail Isolation open and a recire loop is provided around the pumps Valves. a flow path for the pumps would always exist after gp failure of any one of the control valves. No water hammer would occur. 0 A-4

C) ~ Title / Description Applicability .O C2. HHSI - Closing / NA - Discharge motor-operated isolation valve has Opening of Dis-an opening / closing time of 10 sec. The BV-2 HHSI charge Isola-System operates similarly to the BV-1 HHSI System tion Valves which has not experienced any water hammer events. Also, a flow path would always exist after a ., 3 - failure of the control valves. No water hammer ( should occur. C3. LHSI - Opening / NA Discharge motor-Closing of Dis-operated isolation valves charge Isola-have an opening / closing () tion Valves time of 12.2 sec. No water hammer should occur. C4. FWS - Closure Applicable - Analyzed in EMD Fluid Transient Calc. of Flow Con-No. 12241-NP(B)-258-FA. trol Valves O C5. FWS - Instability Not Applicable - BV-2 Flow Control Valves were of Flow Control modified to eliminate flow instability observed at Valves BV-1. Modifications to increase compatibility with Feedpumps included a reduction in trim size and the replacement of plug-type trim with cylindrical .O. trim. C6. FWE Closure Not Applicable - Modulation of control valves of Flow Con-causing pressure waves not expected due to rol Valves reaction time for hand control valves. Control valve closing not considered to produced sign-

C) ificant dynamic loads.

C7. MSS Steam Sup-Not Applicable - The steam hammer event of ply to Aux, closing the isolation valves is analyzed in EMD Feedwater Pump Fluid Transient Calc. No. 12241-NP(B)-452-FD. Turbine Isola-t () tion Valve Closure [ C8. MSS Closure of Applicable - Analyzed in EMD Fluid Transient Calc l Turbine Stop No. 12241-NP(B)-274-FB. Valves j) C9. MSS Opening of Applicable - Analyzed in EMD Fluid Transient Calc Turbine Bypass No. 12241-NP(B)-142-FIB. Valve C10. MSS Closure of NA - MSIVs close in 4 seconds. Loads developed j) MSIVs would be enveloped by those from the closing of the l turbine stop valves (150 milliseconds). l 9 i A-5 i ~~ ,.e .x 1,., - - - -. - -,. -., - -.. - -

Cl Title / Description Applicability 9 C11. RSS Isolation NA - Discharge motor-operated isolation valves Valve Closure have an opening / closing time of 60 sec. (2RSS*MOV156A-D) C12. QSS Isolation NA - Discharge motor-operated isolation valves Valve Closure have an opening / closing time of 60 sec. qp (2QSS*MOV101A,B) C13. ASS Closure of Not Applicable - Air operated isolation valves have Air-Operated an opening / closing time of 15 sec. Isolation Valves O C14. BDG Closure of Not Applicable - 3" Air-operated isolation valves Air-Operated have an opening / closing time of 10 sec. Isolation Valves D. Check Valve Closing and gp. Delayed Opening ~ D1. RSS Check NA - If only 2 of 4 RSS pumps start, RSS water will Valve Closure also fill the risers up to the spray headers in During Minimum the safety trains that are not operating. No water G Safeguards slug is expected due to the displacement of air in Operation the riser. This transient would be enveloped by the startup transient in Section B, No. B6. 7 I D2. QSS Check NA - Same as above_for RSS except that only 1 of l Valve Closure 2-QSS pump starts, dD j During Minimum Safeguards Operation l D3. FWS Main Feed Applicable - Check valve must maintain its l water Check structural integrity only. Analyzed in EMD Fluid GD: Valve Slam Fluid Transient Calc No. 12241 -NP(B)-293-FA. D4. MSS Check Applicable - Check valve closure due to postulated Valve Closure MSS line break. This transient is enveloped by in the Steam steam fill of the empty line (Section B, Transient GD Supply Lines No. B9) l to the Aux. f Feed Pump Turbine l DS. RHS - Check Not Applicable - Pump discharge paths are Valve Closing separated and redundant. Check valves are located dD-and Delayed in horizontal sections of piping. Opening A-6

)-

Title / Description Applicability D D6. HHSI Not Applicable - The BV-2 HHSI System operates similarly to the BV-1 HHSI System which has not experienced any water hammer events. D7. LHSI Not Applicable - Same as NHSI

)

E. Water Entrainment in Steam Lines El. MSS Main Steam NA - Entrained water in main steam line is limited Lines by the collection and draining, via the Steam Drains System (SDS), of any condensate formed in ) the lines. E2. MSS Water En-NA - Same as above for main steam linea trainment in Turbine Bypass Piping O E3. MSS Water En-NA - Same as above for main steam lines. trainment in l Steam Lines to l the Aux. Feed-l water Pump () Turbine E4; ASS Stemalines Not Applicable - Entrained water in " auxiliary steamline is limited since condensate is collected in main steamlines (aux. steam source) and steam traps are located at various points in the system !) I to remove condesate. I l F. Transient Cavita-tion (Column Separation) (). F1. CCP Water NA - BVPS-2 has a surge tank which keeps the system l Column Separa-filled (no drain back after pump stop.) l tion Effects j Following Pump Stopping C) F2. SWS Water Applicable - Analyzed in EMD Fluid Transient Calc Column Separa-No. 12241-NP(B)-173. Vacuum breakers were sized tion Effects at and located to elimiate water hammer for LOOP. Pump Discharge Following Pump Stop and Restart ) \\

O g_7

'O Title / Description Applicability F3. SWS Water Column Not Applicable - Check valves added to eliminate ~9 Separation Effects water column separation. at Main Steam Valve House Cool-ing Coils Follow-ing Pump Stop and gp Restart F4'. RSS Water NA - Pumps have been tested with inadequate NPSH Colum Separa-for 13 minutes without failure. The RSS pump tion Effects and safety trains are separated and redundant. Following Pump A flow transient analysis for column separation 43 Stopping or due to drain back to cont. sump after pump stop and Inadequate subsequent restart is not required since a single NPSH failure (pump stop) in one train would not affect the others. (p I G. Steam Bubble Col-lapse Due to Rapid Condensation Gl. HHSI Collapse NA Cold water injection into the RCS from safety of Steam Bub-injection could cause rapid condensation of steam bles Formed as with resulting water hammer back through safety dDl a Result of injection, lines. Loads are not considered large Local or enough to be of concern (pg. A-21, NUREG-0582). System Also, the BV-2 HHSI System operates similarly to Depressurization the BV-1 HHSI System which has not experienced any water hammer events. GD G2. LHSI Collapse NA Same as HHSI above. of Steam Bub-bles due to l Local or System l Depressuration 4D L i G3. Safety Accum-NA Same as HHSI above. Safety Injection Startup l ulator Collapse Test Procedure (OM-2.11.4.A) details requirements of Steam Bubble to preclude steam pocket formation during test. Due to Local or Also, see OST 2.11.15 " Safety Injection Accumulator System check Valve Test." GD' j Depressuration G4. RHR Collapse NA - Steam bubble formation of Steam between check and closed Bubble isolation valves due to higher temperature in the II RCS avoided by startup procedures. See OM-2.10.4.A. lI A-8

[) Title / Description Applicability [)- G5. FWS Slug In-NA.- Unlikely due to inverted 'J' tube design of pact Due to feedring in steam generators on BVPS-2. Rapid Conden-sation in Steam

)

Generator G6. BDG - Collapse Not Applicable - Steam bubble formation is not of Steam Bubble expected. Feedwater is injected into the steam Generator blowdown water downstream of the containment isolation valves to subcool and prevent slugging or flashing in the blowdown lines. m/ G7. CCP Steam Bub-NA - Unlikely since system and components are ble Formation filled and vented before being placed into and Collapse operation. (OM-2.15.4.L) in Heat Exchangers O G8. SWS Steam NA - Same as CCP above. (OM-2.30.4.F) Bubble Forma-tion and Col-lapse in Heat Exchangers O .H. Pump Start and Postulated Seizure with Full Lines Hl. HHS1 NA - Pump start /stop negligible due to adequate C) venting and priming procedures before start See OM-2.7.4.W. Pump seizure is not analyzed as a transient mode on BVPS-2 since motor overload protection is provided which should trip the motor as rotational resistance increases prior to complete seizure. Also rotational momentum of the (J pump and motor will prevent an instantaneous pressure spike as a result of pump seizure.

50. LHSI NA - Pump start /stop should not produce large dynamic loads due to adequate venting and priming procedures before start.

(See OM-2.11.4.H) O, H3. RHS NA - Same as HHSI above (See OM-2.10.4.A). H4. RSS NA - Pump start /stop are analyzed as part of transient no. B6 (pump start and fill of empty lines). Postulated pump seizure is considered a ) single failure on BVPS-2. Since the RSS pump suction and discharge lines are separated and redundant, no flow transient analysis is required. H5. QSS NA - Same as RSS above. O A-9 - ~.

O - O Title / Description Applicability 9 H6. CCP NA - Same as HHSI above i (See OM 2.15.4.L) H7. SWS NA - Same as HHSI above (See OM-2.30.4.F) 9 H8. FWE NA - Same as HHSI above H9. RCS NA - Pump start /stop should not produce significant loads due to system design and startup procedure (OM-2.6.4.A). Reactor Coolant Pump seisure is considered an accident condition. BV-2 is designed 9 to mitigate its effects. O O O O G. O i 9 A-10

o. 0 APPENDIX 3. STAINLESS STEEL MATERIAL PROPERTIES (This appendix is under a separate cover g and treated as proprietary) O .O O lO {~ O O l lo O-B-1

O APPElIDIX C FERRITIC STEEL MATERIAL PROPERTIES O' (This appendix is no longer applicable and is deleted) O O l 9: O' O i l l O l l l C-1 0:

O 0 .O O APPENDIX.p, .O NELDING PROCEDURE O O O

O

!O O

O. AEEM D WELDING PROCEDURE D.1 INTRODUCTION The terminal end butt welds considered under the BVPS-2 WHIPJET program are welded in the field using the. Gas Tungsten Arc Welding (GTAW), and Shielded Metal Arc Welding (SMAW), processes. Welds at intermediate points in the piping systems are made in e either an off-site pipe fabrication shop or in the field at BVPS-2. The discussion that follows is based on approved field welding procedures.- D.2 FRACTURE TOUGHNESS g Welding variables which may influence fracture toughness of weld metal and weld heat affected zones are specified in ASME Boiler and Pressure Vessel Code, Section IX. The heat input range of SMAW welds studied in this report was 15 to 50 KJ/in. For the Submerged Arc Welding (SAW) process, the heat input range G. 4 j studied was 70 to 100 KJ/in. The heat inputs used for SMAW welds were also checked by corrosion testing of weld test coupons to ensure that excessive sensitization of the base metal would be avoided. 9 D.3 BASE METAL PREPARATION 1 l The ends of the pipe, valves, or fittings to be joined by welding are prepared for welding by machining, flame cutting (except i stainless steel), grinding, sawing, plasma arc cutting, or shearing. The dimensions and configurations are in accordance G with the standard drawings of Stone and Webster's specification 2BVS-920. The method used to prepare the base metal leaves the weld preparation with reasonably smooth surfaces. The surfaces for welding are free of scale, rust, oil, grease, and other harmful foreign material. Welding is not performed on wet surfaces. If welding is not started immediately after cleaning, the weld joint is suitably wrapped to prevent contamination. End prep configuration for standard butt welds, as in the WHIPJET lines, uses a J-bevel. Under certain circumstances, end prepping consists of a V-bevel. 9 D-1 g.

O If piping component ends are counterbored, such counterboring does not result in a finished wall thickness after welding less than the minimum design thickness. Where necessary, weld metal of the appropriate analysis is deposited on the piping component O to correct minimum wall or, in the case of attachment welds to achieve a suitable fit-up. D.4 MATERIAL ALIGNMENT O The forces applied during alignment are limited to amounts that will not deform the piping or components, weld joint, or end prep. Parts that are joined by welding are fitted, aligned, and retained in position during the welding operation by using bars, jacks, clamps, tack welds, temporary attachments, or mechanical alignment clamps. O Localized heating of stainless steel piping is permitted provided the temperature does not exceed 350 F. Hasting of stainless steel above 800 F, the lower limit for sensitization, is not permitted. 'O D.5 FIT-UP Fit-up tolerances are in accordance with the standard sketches referenced on the technique sheet, and the following information as it applies f or ASME, Class 1, 2, and 3 and ANSI B31.1, f or .O Class 4 systems. The acceptable mismatch for ASME III piping is 1/32" per side and for ANSI B31.1 piping is 1/16" per side. D.6 PREHEATING I) Preheating of stainless steel is as follows: Preheating is not required when the base metal temperature is above 60 F. When the base metal temperature of the materials is below l i 60 F, preheating is performed by uniformly heating circumferential1y to a temperature of 60 F minimum (warm z) to the touch). D.7 WELDING GASSES The shielding and purging gas for the GTAW welding process is O welding grade Argon. Gas flow rates are within the range specified on the applicable technique sheet. Shielding gas is used to form an ionized gas (plasma) for metal arc transfer, to shield the molten base / weld metal puddle and to cool the tungsten electrode. O D-2 O

O D.8 WELD FILLER METAL All weld material is purchased in accordance with ASME Section III requirements and the Field Piping Specification. Delta O ferrite determination is required in accordance with Regulatory Guide 1.31 for stainless steel weld filler metal. D.9 WELDING TECHNIQUE e The filler metal sizes shown on the technique sheets are the only sizes used. Typical bead sequence and number of layers shown on the individual welding procedure technique sheets are for illustrative purposes only. The welding technique sheets are applicable to all position welding unless otherwise stated. All vertical welds using the GTAW and SMAW process are performed in G the upward direction. Each bead deposited is thoroughly cleaned of all oxidation, slag or flux using a descaling tool and/or wire brush. Crater cracks, porosity and undercutting are removed by grinding before G[ depositing the next bead of welding. D.10 INTERPASS TEMPERATURE The interpass temperature, specified in the welding procedure technique sheet, is checked on the piping adjacent (1" max.) to the welding groove, using " Tempi 1 sticks", a surf ace pyrometer, or an approved equal. After a bead of welding has been deposited around the complete circumference of the joint, it is cooled to below the maximum interpass temperature before starting the next bead of welding. Generally, by the time the deposited bead is thoroughly cleaned and inspected, the temperature will be below the interpass temperature. To control the interpass temperature of stainless steel welds, water quenching with clean, lint-f ree rags soaked with demineralized water may be used to bring the weld below 350 F 8 before continuing to weld. D.11 WELD FINISHING All weld finishing must be completed before NDT tests are 8 performed. Local grinding with appropriate wheels for the type of base materials involved are used where necessary to achieve the desired surface finish. The preparation of welds for ISI is in accordance with 2BVS-920. Weld edges are to merge smoothly with the base metal. 9 D-3 g J

O D.12 STRESS RELIEVING Post weld heat treatment of stainless steel is not permitted. O D.13 CONTROL OF STAINLESS STEEL WELDING Fabrications of welded austenitic stainless steel Classes 1, 2, and 3 comply with the requirements of ASME Section III and O Section IX as supplemented by the BVPS-2 position on NRC Reg. Guide 1.31 and Reg. Guide 1.44 contained in 2BVS-920. A typical BVPS-2 stainless steel butt weld is performed in the following manner. For stainless steel piping, a TIG root and hot pass is put down first. This is done with an ER308 insert for the root pass and bare filler wire for the hot pass. This first segment O is 1/8" to 3/16" thick with nominally two layers from three beads. After this, the balance of the weld is made from E308 covered electrodes using the shielded metal arc welding process. For each weld, a count of the number of weld wires and electrodes required to make the weld is maintained. The heat input, directly related to the size of weld bead and an indication of O sensitization of the base metal, is nominally 20,000 - 50,000 Joules / inch. End prepping usually consists of a J-bevel and no post weld heat treatment is performed (as just indicated). Certain shop welds included in the WHIPJET program were made with [. the above practices except that Type 316 weld filler metal was F'r O used. Also, the submerged arc welding, SAW, process was used to

.t" complete some of these shop welds.

In these instances, Type 316 i9. weld filler metal with CD-Ni Alloy flux was used. The heat input i.; i range for shop welding processes is nominally 12,000 to 50,000 f.i T Joules /in.

1. ,'/R O

REFERENCES {j[.y u1 D.1 Determining Fracture Properties of Reactor Vessel 3 Forging Materials, Weldments, and Bolting Material EPRI RTN NP-122, July 1976. gi ).; O D.2 Impact Testing of the HAZ, P.F. Ivens and A.A. van den Bergh, Metal Construction and Brittish Welding Journal, j July 1974. f.4 if. O O D-4 O

O i O O O' APPENDIX K O LEAK RATE CALCULATION METHODOLOGY O l 1 O l l O l l l e 1 l l O l O l

R a APPENDIX E 3 LEAK RATE CALCULATION METHODOLOGY r W This appendix describes the computer program PICEP (Pipe Crack 4 Evaluation Erogram) [E.11 that calculates the crack-opening area. -+ the stable crack length, and flow rate through cracks in pipes. =g The program evaluates the leak-before-break (LBB) hypothesis for appropriate loads, material. crack length, and pipe geometry, _7 3 W 5! E1 CRACK OPENING AREA CALCULATION A Crack opening area for a through-wall circumferential flaw is gg 3 computed by using the crack opening displacement formulas of Kumar and German EE.23 and assuming a rectangular, diamond. or -A elliptical shape. The plastic contribution to the displacement is computed by 35 summing the contributions assuming pure bending and pure tension. y O a conservative LBB procedure that underestimates the displacement from combined tension and bending. Applications cover most of ]g the practical range of material properties, pipe geometries, and y crack lengths, qg Figure E.1a compares measured [E.33 and computed displacement in O a 6.5-inch outside diameter, t = 0,425 inch, type 304 stainless _g steel pipe with a 90-degree circumferential crack subjected to an 5 increasing bending load, Figure E.1b gives the measured stress- -5 strain curve for the material and the analytical fit used for the jj PICEP analysis, The best results for this calculation were found Z by matching the data in the 0.25% strain range. Other crack = O opening displacement comparisons are given by German et al, -i E' [E. 4 ], 7 Evaluation of axial flaws is given in the PICEP report [E.1 ], jg 3 _j O E.2 CRITICAL FLAN LENGTH 9 The critical flaw length is the maximum stable flaw length for a $H E given load. For austenitic stainless steel and its weldments. the critical flaw size may be computed using the equations in j Appendix C to Section XI of the ASME Boiler and Pressure Vessel 3 O Code. These equations have been implemented in the PICEP E 1 computer program, 9 3 The fracture toughness of carbon steel piping requires d consideration of a wider range of fracture mechanisms and the equations of Appendix C may be inappropriate. For these steels. O the critical flaw length is computed by using other methods g proposed by the Section XI Task Group on Piping Flaw Evaluation. ] y O g_1 =

o The validity of these procedures is discussed in ASME technical support documents [E.5, E. 6 ], c O E3 LEAK RATE CALCULATION An analytical model was developed to predict flow rates f rom __d blowdown of initially subcooled or saturated liquid through j cracks based on Henry's homogeneous non-equilibrium critical flow fi 3 I model [E.7] with several modifications made for use in PICEP O saturated water-steam mixtures and superheated steam (E.10), The ~ l [E.8, E. 9 ], The model has been extended to cover the flow of j model assumes the flow to be isenthalpic and homogeneous, but it j accounts for the non-equilibrium " flashing" mass tri.nsfer process between the liquid and vapor phases. Fluid frictien due to -j surface roughnssa of the walls and curved flow paths has been O l incorporated in the model, Flows through both parallel and convergent cracks can be treated, Due to the coraplicated q geometry within the flow path, tne model nues some approximations l and empirical factors which were confirmer1 by comparison against ]l test data, 1 The general features of the discharge of initially subcooled or saturated liquid through a crack are shown in Figure E.2 and the g configuration of a convergent crack in shown in Figure E.3 In the region 0 i L/D i 3. a liquid det surrounded by a vapor ] annulus is formed. For lengths between L/D = 3 and L/D = 12, the C liquid det breaks up into droplets at the surface, and small g g

== bubbles are entrained within the det. It is assumed that no mass or heat transfer takes place between entrance and L/D = 12 Non- ] equilibrium effects are introduced through a parameter, which is a a function of equilibrium quality and flow path L/D. j For given stagnation conditions and crack geometries, the leak g rate and exit pressure are calculated using an iterative search for the exit pressure starting from the saturation pressure j corresponding to the upstream temperature and allowing for friction, gravitational, acceleration and area change pressure 5 dropa. The inertial flow calculation is performed when the critical pressure is lowered to the back pressure without finding g i a solution for the critical mass flux, a A conservative methodology was developed to handle the flow of wet or dry steam through a crack. The same model was used except 7 that the single phase flow equations were not used. Rather. the a flow was assumed to be two-phase from the entrance to crack exit, 3 f= To make the model continuous, a correction factor was applied to adjust the mass flow rate of a saturated mixture to be equal to that of a slightly subcooled liquid, Similarly, a correction S factor was developed to ensure continuity as the steam became superheated. The superheated model was developed by applying thermodynamic principles to an isentropic expansion of the single 5 phase steam. R 2 m S E-2 a a = 2

.O .The code can calculate flow rates through fatigue or IGSCC cracks i cnd has been verified against data from both types. The crack curf ace. roughness and the number of banda account f or the difference in ge metry f the two types of cracks. The guideline O for predicting-leak rates through IGSCCs.when using FICEP was based on obtaining the number of turns that gave the best cgreement for Battelle Phase II test data (see section E 3.1.1), For fatigue cracks, it is assumed that the crack path has no ban ds, E.3.1 Verification with Experimental Data The experimental data for flow through tight cracks under 2-phase flow conditions is scarce. PICEP has been ammessed using several open literature and proprietary data mets. Key verification C**"81"" f 11 "-

O E.3.1.1 Battelle Columbus Laboratories (BCL) Data This program [Eu(1 ) funded by EPRI generated data to confirm critical flow models used to predict leak rates through cracks, l(3 Phase I tests used simulated cracks in which test variablem included L/Dh ratio, stagnation temperature, stagnation pressure, and crack face surface roughness.

The variation in L/DA ratio was provided by using three different specimens. The curface roughness was varied by machining the crack faces to the desired roughness. The crack faces were initially ground to an () cverage surface roughness of 0.039 mila as measured by a surface profilometer, After completing the ' smooth

• tests, the flanges were removed, disassembled and the crack surfaces were roughened by shot blasting to an average roughness of 0,24 mils, The flanges were reassembled and installed on the test vessels, The minimum crack width opening was measured and a new L/Dg

()

computed. For the final met of tests, the crack faces were roughened to 0,4 mils average roughness. and L/Dh was recalculated, based on new crack width measurements. Predicted versus measured flow rates for simulated cracks are shown in Figure E.4, O The Phase II testa used a test specimen (Figure E.5) with a 90% through-wall circumferential intergranular stress corrosion (IGSCC) crack. A portion of the outer pipe surface was removed to expose the crack tip and progressively wider cuts resulted in the variation of the L/D parameter. The maximum crack length was 1,1", O Following the Phase II tests. the pipe containing the IGSCC crack was cut and an exit to entrance area ratio for cracks was coproximated by measuring the slope of crack convergence, A crack wall surface roughness of K = 0,0051 mm was used. The flow path was assumed to contain twenty 45 degree turns, i l O E-3

O Figure E 6 shows that PICEP results are in good agreement with measured leak rates for tests 19 to 81 with crack lengths from 0.38" to 1,1". The Isak rates f or very narrow cracks (testa 7 to 18 where 6 = 0.02 mm) are overpredicted because of high GD likelihood of plugging in the short cracks. E.3.1.2 Duane Arnold Safe-End Crack Plant Data From June 14 1978 to June 17 1978 the Duane Arnold nuclear plant had an unidentified leakage source of 3 epm into the O> primary containment sump [E LL], The leakage was below the u plant's technical specifications of 5 gym and the plant operation continued between June 14 and June 17, 1978 The reactor, however, scrammed on June 17 1978 during performance of weekly control valve tests, Following shutdown of the plant on June 17. plant personnel found the leak to be from a crack in one of the II oight recircu1ation-inlet-nozzle safe ends. These safe ends are coproximately 12 inches in diameter and are used to f acilitate welding of the stainless steel inlet piping to the carbon steel reactor vessel nozzles. The Duane Arnold recirculation inlet configuration is shown in Figure E.7a, The safe ends were ' () manufactured from a nickel-base alloy (Inconel 600). The thermal sleeve shown in Figure E.7a contributed to crevice corrosion in the Inconel 600 The safe ends were examined thoroughly with UT. These examinations clearly shcwed that all eight safe ends were cracked essentially completely around the circumference. The crack g, penetration from the inner surface typically ranges from 50% to 75% of the wall thickness, except'in the leaking safe end where an 80 degree segment had a through-wall crack. The general characteristics of the crack geometry were later confirmed by a destructive examination of the leaking safe end and one of the remaining safe ends with a part-through crack, Figures E.7b and g, E.7c illustrate'the general character of the cracking in the leaking safe end. The PICEP model used a through-the-wall crack with inside crack length of 7.2 inches and the original pipe thickness of 0.65

inches, It was not necessary to model the compound crack.

The gp critical crack length for the PICEP model in 19.9 inches. The following plant input data was used for the PICEP calculations: Pressure: 1050 psia gp Temperature: 550 F Outside Diameter 12" Thickness: 0.65" Crack Surface Roughness: 2 x 10E-04 inches No. of 45 degree Turns: 17 Crack Shape: Elliptical g) Crack Orientation: Circumferential E-4 ED

D Crack Type IGSCC Young's Modulus: 28,9 kai Yield Stress: 42,0 kai Flow Stress: 75,0 kai g Hork Hardening Exponent: 10,0 Hork Hardening Constant: 3.5 Bending Moments 0 (Case 1) Bending Moment: 225.000 in-lbf (Case 2) g Measured Crack Length: 7,2" (inside) Measured Crack Flows 3 spm The bending moment of 225,000 in-lbf was obtained f rom the stress report (personal communication from H. Mehta and S. Ranganath of General Electric, San Jose) by using the SRSS of My and Mz g coments. The torsional moment Mx was not considered as this does not act to open the crack. The resultant bending moment Mr = (My , g2 2) uz provides a varying stress field along the 2 circumference of the pipe. Since the crack was not exactly at the location of maximum bending stress, the problem was bounded g by making two runs. One used no contribution from bending moment cnd the other used maximum contribution from bending moment. The results from these two calculations bound the measured leak flow rate data as shown in Figure E,7d, These two calculations also bound uncertainties in the bending moment due to the attached thermal sleeve and crack plugging and unplugging effects due to ) corrosion products. E.3.1.3 UC, Berkeley Data C. Amos and V. Schrock [E.13] funded by the USNRC obtained cxperimental data for critical flashing flow of initially ) cubcooled water through rectangular slits, The influences of ctagnation pressure, stagnation subcooling, and slit opening dimension on both the critical mass flux and the pressure profile within the slit, were studied parametrically, Stagnation pressures ranged between 4.1 and 16.2 MPa, Subcooling was in the range of zero to 65 C. Length-to-diameter ratios (L/D) of the ) alits were between 83 and 400, with the length in the flow direction fixed at 6,35 cm, (This is a typical value for the thickness of a nuclear reactor primary cooling system pipe), The test matrix consisted of 60 tests. Each of three slit opening dimensions, (0.127 mm, 0,254 mm and 0,381 mm) was tested J ct 5 pressure levels (4,2 MPa, 7.2 MPa. 9,6 MPa. 11.6 MPa, and 15,6 MPa), and with four degrees of subcooling (60, 30, 15 and 3 C), Five of the highest pressure tests were not completed because of leakage problems encountered when operating the system at high pressure. Slit openings were selected to be of a size cuch that leakas e through crack = of that dimension, with a length O corresponding to 50% of the circumference of a reactor primary oystem pipe, would result in a significant loss of cooling water, ) E-5

O' Stagnation pressures and subcoolings tested cover the range-of conditions which occur in Pressurized Mater Reactors (PMRs) and Boiling Mater Reactors (BNRs) under both normal operating conditions. and during postulated accidents. In addition to the two-phase tests, each test section assembly was calibrated using e cold (25 C) water. A stagnation pressure of approximately 2.5 MPa was used for these tests. Data from these runs were used in determining the single-phase friction factor and entry loss coefficients for the slits. -Figure E.8 shows that PICEP results are in very good agreement 9 with measured leak rates. 4 E.3.1.4 Canadian Fatique Crack Ha_ta_ PICEP predictions were compared with the data taken by D. Scott cnd A. Cook (E 14 ] for critical flow of initially subcooled water O I through cracks in thick wall piping. The specimens were all made from-Schedule 100 ASTM 106 Grade B pipe. The pipe diameters ( tested ranged f rom 305 to 610 mm. To make the specimens, a length of pipe approximately twice the pipe diameter was used. An Electric Discharge Ma. chine (EDM) notch was then placed at the O center of the pipe on the inner surface to initiate the crack. The notch was 0.127 mm wide x 5.1 mm long x 2.54 mm deep. In come cases, an additional stress riser was added by welding a flange on. the inside surf ace of the pipe. Preneure cycling was used to grow the crack. The specimen was attached to a high temperature loop which supplied water at about 250 C and 8.4 MPa. O. The water in the loop was kept at a pH of 9.7 with about 40 ppb of oxygen. Strap heaters could supply 4.9 kN of heat and could bring the temperature of the pips up to 290 C. The temperature was controlled between 255 and 265 C. i On completion of leak testing. each crack was examined. This g j cxamination consisted of: ) (a) cutting out the section of pipe containing the crack, typically 0.1 m x 0.3 m l (b) measuring the crack length on the pipe inside diameter g-and outside diameter (c) measuring the crack opening on the inside and outside diameter using a microscope (d) cooling the section in liquid nitrogen and then g cracking it open to expose the leak path i (e) remeasuring the crack length on the pipe inside and outside diameter, and l (f) measuring the surface roughness of the leak path, g. i. l E-6 g- 'l

) Figure E.9 shows that PICEP results are in good agreement with measured leak rate. An important finding of this work concerns the plugging and 3 unplugging phenomena observed for short crack lengths. For larger crack lengths, plugging phenomena are not important because the crack opening is too large for this to occur in the middle section of the crack, It might still occur at the ends of the crack where the opening is tight but the impact on showing LBB is insignificant. This is also illustrated by the crack O opening displacement of the near critical crack size in Figure E.10 tested at Hyle Labs for EPRI to demonstrate LBB [E.15]. This alleviates the concern that a crack might not be detected due to plugging of the crack. l$ENh E 3.1.5 ANL Experimental Data D. Kupperman and T. Claytor (E.16] funded by USNRC took cxperimental data to evaluate and develop improved leak detection systems. The primary focus of the work was on acoustic emission detection of leaks, Leaks from artificial flaws. laboratory-generated IGSCCs and thermal fatique cracks, and field-induced g IGSCCs from reactor piping were examined. Water at temperature l cnd pressure of 500 F and 1100 psia respectively is supplied to a small pressure vessel which is welded to the inner surface of the pipe and vary the crack opening, Calculated and experimental leak rates through an IGSCC and a g thermal fatique crack are shown as a function of applied stress in Figures E.11 and E 12 As expected from' fracture mechanics, the flow rate is in most cases proportional to the applied stress. This linear dependence breaks down for low stresses. ANL speculated that the residual stresses produced when the crack is welded int the piping system may tend t h Id the rack shut O until a certain threshold level of stress is reached. Since the astimate of the initial crack opening area (zero applied stress) was not known, two curves are shown from PICEP calculation. The first assumes no initial crack opening area and the other assumes cn opening that yields a flow rate equal to the zero applied g stress value. The overall comparison and the trends are good. E.3.1.6 CREC Experimental Data M, Cumo et al. (E.17) at CREC, Italy measured the flow of cubcooled water, saturated water and saturated steam through a O tong (1,5 meter) small diameter (s mm) tube. The conditions testa covered a range of pressures (11-24 bars) and subcooling (0 to 47 F). Their data is presented in Figure E 13 together with the PICEP prediction. The agreement is very good and the predictions are generally conservative, O O E-7

l O E.4 PICEP SAMPLE CALCULATIONS Three sample calculations are provided to show the general trends predicted by PICEP, Figure E.14 shows the effect-of crack length cn flow rate with subcooling as a parameter. For a fixed O geometry surface roughness and inlet pressure conditions, the l predicted critical mass flux increases as subcooling increases. l The influence of crack depth-over-hydraulic diameter (L/D) ratio is shown in Figure E.15. The critical mass flux decreases uniformly with increasing fL/D ration. This is consistent with O 7 the assumption that the mixture quality approaches the long tube value in large L/D cracks. The effect of stagnation pressure is shown in Figure E.16 For a fixed geometry, surface roughness and subcooling. the predicted O mass flux increases with stagnation pressure. 9 O O O. 9 O E-8

) REFERENCES E.1 Norris. D,. Okamoto. A., Chexa1. B., Griesbach. T., "PICEP: Pipe Crack Evaluation Program." EPRI Report ) NP-3596-SR. Special Report. August 1984. 1 E.2 Kumar. V.. et al.. " Advances in Elastic-Plastic i Fracture Mechanics." EPRI Report NP-3607 Palo Alto. California. August 1984 E.3 Personal communication from Kuzuo Kishida of IHI to D.M. Norris of EPRI, April 1. 1986. E.4

German, M.D.,

et al., " Elastic-Plastic Fracture Analysis of Flawed Stainless Steel Pipes." EPRI Report NP-2608-LD, September 1982. ) E.5 " Evaluation of Flaws in Austenitic Steel Piping." ASME Code, Section XI. Task Group on Piping Flaw Evaluation. D.M. Norris. Editor. April 1986 To be published in the Journal of Pressure Vessel Technology, ) E, 6 " Evaluation of Flaws in Ferritic Piping." A.

Zahoor, et al..

EPRI Research Project RP1757-51. October 1985. For consideration by the ASME Code. Section XI. Task Group on Piping Flaw Evaluation. ) E.7 Henry. R.E.. "The Two-Phase Critical Discharge of Initially Saturated or Subcooled Liquid," Nuclear Science and Engineering, Volume 41, 1970 E,8 Abdo11ahian, D. and Chexa1. B., " Calculation of Leak ] Rates Through Cracks in Pipes and Tubes." EPRI Report NP-3395 January 1984 E.9 Chexa1. B., Abdo11ahian. D.. and Norris, D.. " Analytical Prediction of Single Phase and Two-Phase Flow Through Cracks in Pipes and Tubes " 22nd ASME-0 AIChE National Heat Transfer Conference. Niagara Falls. August 1984 E 10 Chexa1 B. and Horowitz, J,. "A Critical Flow Model for Leak-Before-Break Applications," submitted for presentation at the SMIRT-8 Conference. Lussanne. 3 Switzerland. August 1987 E 11 Collier. R., et al.. "Two-Phase Flow Through Intergranular Stress Corrosion Cracks and Resulting Acoustic Emission." EPRI Report NP-3590-LD. Final Report April 1984 D S E-9

O References (continued) \\ E 12 " Investigation and Evaluation of Stress-Corrosion Cracking in Piping of Light Natur Reactor Plants. " O NUREG-0531. Pipe Crack Study Group. February 1979. E 13 Amos. C,N, and Schrock. V.. " Critical Discharge of Initially Subcooled Mater Through Slits." NUREG/CR-3475 LBL-16363. September 1983 O E 14 Scott. D,A,. " Leak Rates Through Cracks in Thick Malled Piping." AECL Report. June 1983 E 15 Bausch. H.P., et al,. " Pipe Rupture and Depressurization Experiments. " EPRI Ramsarch Project O 2176-1 January 1986 Draf t Report. E,16 Kupperman. D.. Shack. H.J,. and Claytor. T., " Leak Rate Measurements and Detection Systems." Argonne National Laboratory. CSNI LBB Conference. Monterey. California. September 1-2. 1983. O E.17 Cumo. M, et al.. " Experimental Results on Critical Flow Internal Technical Report. NDAG 2TS4B020 Centro Ricerche Energia Casaccia. October 1982 O l O 0 t O l l O l l E-10 g-

0.05 g,,, g,,,.g. 0.04 O 0.03 3 8 ~ ~ EXPERIMENTAL DATA 0 0.02 O 8 0.01 PICEP CALCULATION O I,,, l, I, I,,,,- g,gg 0 0.2 0.4 0.6 0.8 1 ab/cy 'O 50 l.., l l l

O 40 RAMBERG oSGooD FIT.

Q L 30 E 'O e ~ g 20 LJ IHI DATA g g O 10 l,,,,I,,,,I, I,,,,- g O O.2 0.4 0.6 0.8 1 'O ELONGATION (7.) Figure E.1 (a) Comparison of meawred versus computed crack-opening displacement versus bending stress. The stress is normalized with the 0.2%-offset yield stress of 24.8 kg/mm ; (b) stress-strain data and the Ramberg-Osgood fit used for the 2 lO 2 PICEP calculation: c =24.8 kg/mm, a=1.0, n=5.0, c =0.125%. Data is at 20*C. o o

O

. =__ O; e Ftre fills efeck A mwasssssssomlm_memm_ms4 / chemiae a- , *le e = - g%%%N'*%N%%WN%%%%g N 9. T.anesemistee f sp;re E.2. Two-Phase Flow Through a Long, Narrow Crack 0-O' Length 9 Width 6 ,-l , 2c y c .L gai, piene / e ,d / ,/ A, tarse a vo = 121 s / / l L' / A. Inlet pleae e t ca.e.,_ _ Figure E.3. Geometry of a Convergent Crack O 1 O' l

m l D = J O 6 6 4 4 e e 1e I**e t 6 e - .s 7 r J 'F i a t j 'j n O V # e V l CC m O Y a J O<p l r c_ I e b-./ d <3 e w 4J l 0 kt "3 De l g O X T or 'o,'- E [ r_ u M l f U L -3 L.' g av m m 0, g + -m x I f e = iii D E I i g r f O t x u) Iil j E s k x O< x g L t 1 t t t t I I t t t 1 1 9 f f f 1 I I O o o o o o o o a o e a o (Wd9) 31V8 XV310313IO38d o i

O O .... se <AA A \\ l \\ l i I i k w osce g-ca l l en mene et 9see ac me.me.en.ag l t..... O 4 c,.,e,..e s., sEcTioN A-A g' Figure E.5. Sketch of Crack Details O' l O. O' O' i l 9

O O' O O , _ _o __ o- _O'~~ O ~ O, ~ i 101 ,,,iii. i i i i ni i i i iiin i i i iviii i-/ 2 } N 100 M y-O_ ~ e 9 w I 10-1 (r 'x [4 M X M M 3 M

• 10-2 9P x

x i o W I h I-0 o O m O ig-3 X Tests 1-6 W O' I Tests 19-81 l 1 Q Tests 7-18 [ l 10-410-4 10-3 10-2 10-1 100 iOI l I MEASURED 1.E AK RATE (GPM1 l Figure E.6. BCL Phase II Results l l l

l l i t i I g h .g( I 'I twa a*W4 J rei s j -g 1' 1) ,,5 ~ r n n ? h/ A.' l a; ggy,g( ag--,l.--.i !(-e k w m_.s(yi .I assur m .e g w o k3 _!j h }, I Mwt m em ~ f a-w __w !A .._ sit.. '""'N ~ I wh l l l 1 g k,y' Rtctae., teat t a, pe=1, Lo. ~ ~ ~~ y ? am*A;;'"' s>~~,ww-so m

d Asessed 4I I

~d if y o e:1 i g = Figure E.7a Duane Arnold Recirculation Inlet (Old Configuration) From Drawing No. VPF 2655-96 0 8 8 9 9 e e e g

r 4 0 O ~O S 55 h .O mzn. cue =saue=.. --- m . c=..

o r

Figure E.7b Representation of IGSCC in Duane Arnold Leaking Recirculation-Inlet-Nozzle Safe End O l t lO l l .;o .;O a O

O L w, i STA.~0 sc y 330' \\ 1 308 l / / 18 ~~- \\ g e, S " sN 3 g 1 se, ii N' e s t,/ \\ 7 v i If 9 [Y l f 1) a a 16 270'-- 5 -E jl - 908 .15 _ / ff m 54_ . g\\ M // L g\\ // g / 6 \\ \\ / 6 7 f/ N innousm-Wau. 240,/ 1% / i Camer. \\ N N ' % g.. N s ~ 3 i 1 I j g-210' / m\\ 6 M ' 13 le 3 / 180' s p. j 1 12 e' STA.20 E FancroseAPMit $PictMu$ M IGRFACE DEP08tt Setcsmans 9, Figure E.7c. Safe End N2A e: 4: __.- _,-.- - -. ~ - - -.... - - _... - _ _ _ _ - _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

7. i.,

,,! l' 3

797 i 9 DUANE ARNOLD ENERGY CENTER - NO BENDING 6 ....g....l .g....g.._

O -

.~ CWICAL CRAM SE = 19.W IN 5 CASE 2: PPISSURE + mDE 4 lO'- r a. o. . w 2 3 e 4 CASE 1: P .y lO PRESSURE ONLY 2 a i t-10 I'''' f-O O 2 4 6 8 10 i CRACK LENGTH, INCHES O-l l 0 -O Figure E.7d. l

O k

3 E K _ EY S ~ ~S i i i i r-ir1-i r-i i i iiij 10.0 x g-M( )x 5.0 x 1 x -r 9( Q-U cr) p [h m O/ O g ~ H O X O g X s y O 1.0 g X ~ x x g [_d 0.5 X M Slit Width =0.381 m H ~ ~ O Slit Width =0.254 m O X Slit Width =0.127 m ~ g m CL I O.1 O.1 0.5 1 5 10 MEASURED LEAK RATE (GPM) 1 Figure E.8. UC Berkeley Slit Data vs PICEP Prediction 9 O e e e e e e. ^ e e y e

C/\\ A1 A \\ - I .. G L' .C 8C<S. ^' 1.00 i ,ii. i i G---X 0.50 x n^xinun tExx etow m 0 AVERAGE LEAK FLOW r { Q_ t_o 4 g, c x F-< c x E M O.10 W ] J o j W O.05 F-o. o D x W E Q-1 I 1 i O.01 O.01 0.05 0.1 0.5 1 i MEASUREO LEAK RATE (GPM) i Figure E.9. Canadian Fatigue Crack Data vx. PICEP Prediction

O - n. m. ,.,U+- ,[e *: # .a^ { JD' Q ' I' 'QL*M&, -%s k'[ 4

' Q

$$, M'j% , jo ((,[ sjJ 'Nl=A: ig.Q' '~ 'L - .~ ,;;hRQW ~ f? :: ... ?a n., . - o. en ~, a.s g?, ' ^ ,ye ~-? i!? ' ' '. '$ < y.'-h;,fk m -+ m.w: WO ' ' ;L nffj$Na .p ,;f. + %.a o m.. O x;a., d %.s%..4 -{ y g:p m e, "up - t. .,R, f, y. j P= i __._._. ~ h m. M w f A, f/ --. $$ bo .r.y ,k, ..,. y q' m.4 m,i.x. q# w.. . ~... gs.w J~4 a ~.4.+h. " f. > some., e a A #,.. J.l5 : ~?.*y yy av. ~ , N, p. .,L a, p g__ f ' ;an> g. c- .r - -)... . 4 r t,yj -lM l s n:,,&, b, S;;gg, ~L L* 't

• I nh; >:t W

_,c ^ s... C < wnnm, l' - Q;m,~. Q :- e 'K,v-T m.r.gdq ::n.; . Ruptuu 7: e. a, ~ y s _. k nm - ,,,,,_, g pg 4 .( _.. ~a t 9... .- m w. ;, s.c- ~. .r g e s 't,f. $'y t ~ s / s-I, . ~ ,' 1,4..., iM",, C ] s '% w as '/ S.' S y ~fi ~..'-. ,-%s. O 8% l ] e.: .x. p:r r.. p.

/

M O -s. d

\\

l j ~ s. . J g .T -. g 4 ~ + e /, 9: .y 'f..

• 34=.

j, 'a ,.{.- -ue.,~, g,.; l S. Figure E.10 Photograph of Pipe with a Near-Critical Circumferential Crack 4! a..

Ol 1 i s >4 ygA* w). ..y'YO:r:

}: f',. '. +. v ~y y-y.?m

- ;,9.,j.g; g%2% :g'RQ.hl;Q.;R,. , }l ~';t. ^ ' ~ ~ ~ ~ s

w. o; ^ -

• g;. 'vf x. ;> > c

. ;a; -,,. .,....w g; 1 ..m w +- r. .4.,, 3 qe. \\iU;3 j g;afN* g?{ ' W-p -% +. - 6 ay;h,. z,.Q; 7htm.,), %.q e m.y% e

.c.

,nr' 't, ,. 's t f 'f;W,e W: \\ p,? +?

• w.cg A

. ;y - V __._..;,, s +, ? rp-. W ; Y.,g :, W, v n

a - -,, -

, r.. k,c,._ pac.m m um, - - %..,w .u-g% s f 1,.4, s s.., o A.. %. ' t J 4 4 h f.(. : 3 av . x a.+ m..,, 9 6 4 4,. y... -.. s s l g.vr+t +mm m )w. -a.. le.. ??v r y ,v, %, ; t.. --y s v,i- : w 1,, r... r caua.w, ' i ' [J: 4 3 ' l [.f i n: l ( & fi f.'

• l' i, I? s =0,W. ( $/jt/;f-j

-'N c,.- e t ~.. ' ' g L % n m +n i:'..; . As.+ar u 2 x. X 6,%: - i '-

h'-#. e 4-m t,3, ~w g.-

,.

• w..-.. u. n -

.-44 .mn e. .. -', *m., : p p. .*'%.W g 1 ' +d- ~ 9-t i ..g~' l,c... e,.; ,. m-A % r. -,c t; e ) 9 t'. r J O~ .r. .N i '\\ f 8 6 g S ."* \\ } , - l4 41 s., 9 1, . _1 '.1-s = w-i 9; l Figure E.10 Photograph of Pipe with a Near-Critical Circumferential Crack 9 1

T' ~ A7l;',[]Nji' N A" ()xj[l[ A ff (E.ECI CIRCIF i. i i y i i i 11.0125 / ~ s / 0.0100 / y / s PICEP CALCULATION WITH / 05 0.0075 INITIAL OPENING ~ h-X / l.' ' ~ / A ~ /y'/ cc y' ~~ 0.0050 ,/ y f, / iii \\PICEPCALCULATIONWITHOUT INITIAL PENING 0.0025 s / / I I ~ O.0000 ' ' O b 10 15 20 APPLll~ l:1 STRESS (ksi) Figure P. 11 ANL IGSCC Data vs PICEP Prediction

lilj 'i'.'Q/\\f- /\\ ('lji[][\\ \\l N/\\ 1;) ' /\\ /\\ 3 /\\ i 11.1.16 ,iii e i ii iiii i8 8 I l.0b / ~ / / / (.l.0 4 n j (9 PICEP CALCULATION / / O.03 EiE,' x, ? ANL DATA y~ Ir yk v ' ICEP CALCULATION WITHOUT - l}.02 .. -c' P gii , \\ ITIAL OPENING IN , /,,- / / .. --,-/ - f----. U.01 }, s ~ ~ ~ ' ' , i. 0~ 0 U .?. F. b 7.5 10 1 2. 5-Al >l'L ll l'i 'J. I kl SS Ikn i) Figure E.12. ANL Fatigue Crack Data vs PICEP Prediction O e e e e e. e e e e e

.:.0 O O O O .O ~ O-O. U~ 'O-4 cut'0 et a ROUN J 5mm ~~UBE L/J=3001 ~ 5.0 ~ ~ X SUBC00 LED WATER 2 E SATURATED WATER MI SATURATED STEAN E L13 F M <o' 1.0 y LLI _J 0.5 o LiJ H + O ~ O W + 0' Q_ O.1, f O.1 0.5 1 5 ' MEASURED LEAK RATE (GPM) Figure E.13. CREC Data vs. PICEP Prediction

o [ W w o 0 7 l F 1 0 0 e 1 t a ~ R k a e L A d F e t ~ 5 c i 3 0 d = e T A 8 r / P N /Y E e S h t O C l n l o g N n e 0 i I. l 0l o \\ o I c T bu G S N f l l E o e i t I c C e i K f f 4 C E 1: A c i U R rt C e m e S ara P T 0 2 14 E e ~ er J u g C i F ? T e O r - ~. E 0 8 6 4 2 O 1 e E a e i. _ < "~ x < w 1_ ui O l lll ,ll1ll 'l!lllll l

0 ~ 1 3 3 3 v 3 4 2 5 wo l l F 5 A e s 8 t j /'/ / a 's. / R / k 5 a /[/ Le l 3 5 1 d A e 9 c = t Ff g, l 4 L i 4 l de l r l 6C P N no \\ I I D l 'l. T / i L [ I F O f 'j f N o ' 7 I

c. '

E tc s e I f I 4 f K E I C c i I A r t R em C a [] ra P ~ ~ 4 51 l, -2 ~ E e l r u g i F i 4 O 7 8 6 4 2 0 I 70 tg M<LL 5l _ i.

l. O S ~ - ~ ~ w ~ o l l k 0 F O e / R 1 ta e k ae L de t A / / P c i 8 de e r n S o N / il e l r / N e e O / l u C s / s / r 7 1, P 6l / f i T o l. / R D t c l. N e / f /,'"~~ E f e E a i I a c s p i K r t 0 0 /~~ 4 C e 0 0 0 m 8 2 0 A a i' 1 1 6 r = = = P P P R a /# C S P 6 ~ 1 ~ 0 e E r 2 u G g i ~ F C E r 2-E_ ..~~ O 0 8 6 4 p-0 1 8 r{h f_ <* y<oa t f 8 l\\lll11I l 1I

3 D D l l APPENDIX E BVPS-2 LEAK RATE CURVES 3 D D D e O

O APPE3fDIX F. BVPS-2 LEAK RATE CURVES O The following leak rate curves give the leak rate (in gallons per cinute) from a crack in the BOP piping as a function of crack length (in inches). The leak rate curves corraspond to the lines cnd breaks tabulated in previous sections of this report. 9 O S: O O' 0; l e e F-1 4' l l

O O O O O O O O O O O. BYPS-2 SIS 6-INCHLINE,NORMALLOADLEAXAGE,CIRCUN ......................................................................................................,....................j 10, i L 1 E A .....................i....................].....................j.....................i..................j.....................' 8. U n R 4, P 28 .....................e .....................e.................................................. f 4 / e. t i I t 1 O, I-- ,00 2,00 4,00 6,00 CRACXLENGTH(INCHES)

m L./ a.........................................,.......................................... i. i. O i i. = = = =. e =* E. ap

=2 :

.as g a. ............y............. g:hC:. e m O !. i Mi 6 O: i i i a; 24: 3------------**-------------3-------------:------ i. a: i. M:

3 :

V e sC:3n : cso a: M sC>. z G3 O . ::3

• i.............*.............

.............:..............i.....,8..........GB Z O A :: e: 'u:t* v i

El::

=C:> : m-e 2::: O Z M

3 M*

g ............................A............ Z.: DC .3: 3 i O a l =

E:: :

5:C O: O Z: M: i 53 .as g 7........... 9............ cu Cc: t O: 'l f.aC : CNE: .= g e-c.o.: i g:, .............................:........................ 7..............

l::> ;

EX:t : O g O 8 8 8 8 g O E cs3 e cs3 C!D CD Ck:t ch3 CE3 CD CE3 C"3 CE3 CD l --e G~

3Mc::nc sz::a n-.M o::3 r 9

au esa ,.........................................,.........................................m ,o i e w Ei i i i. i ~ ' O vi !..............i.......... ..!............ i..............i..............3 eo O b ~ ~ d i i i m :. c,o 3i i i a r.-a X" a *:

.............i...............*..............:............................ m "U

!gs O

a :

i i

=

g::: o x ~: Mi. i i d O "hs". m g; -O i.

• @ ' *:E:

i. i. M i O = m. I O e :: i i w: ce: i cs3 g:j .j.............j..............:..............i.............*............ .m N: i i cu si l i imi. i. i. i i. g i. m O 7 W W W W m W G G3 cE3 GR G3 -4 O a r.wa.cz:::c m.c x: s r.w a oo x 0 1

O E ..............g........................... i cza g .-e i i N.. i g.............g...... ............-............a............. E.: s. m. ...............;..............!............!............^! O! Em1

• 3 3

IOD i i. i. C i. .= g.3 3.............;..............j..............g...........y4............. g N a: .i n glll3 . g, a:,r) h.! @M .g3 :l:g: a. ..............:.............y.. g v ai i i i eZ O g v lllEl: i ,...........................+............y............ g O

25:

gm3 f g.g

.............:..............:..............:..............:............. @ -lBC w.

-O i i i 'atf* <:x: AC 3 3 O e. I S os; W :: i .i sc,r3 ig.: j j.............:..........................j..............i.............. 12 9 o3 e3 g g o 0 a r g3 e ~ C53 .-4 9 O

'O .s..................................................................._............... .= O = i 3 i csp .csg z. ............y............. s. Ci i N i i co i. O 8: 8 : i. i U: g.4............ 5:as3 = 0 :: i i. i. 6: AC: .O i i i Gb A :. .CER .............................y............. A.- e e c,o 6: I;mr.3 c:>. ,, ;3 :

E::

4> i i ..............:..............i..............:.......................... 0 a: v g:g : ,c:p : 's =I: 's's M z :: 8: "J a:33 Z .gga g;,,,3 g;; 3 - ............................e............. -d O E i. i i i. '42* hc = a :. 8> .CE:

E::.

v: :n : N Z! .j.............j.............3..............i.............3.............. = g .,O c.: i i i i -s: Ga':8 ! M :. g.............j.............. .............:..............:..............G 4s0 : i i 8NI l os: i <O j l tV3 : = l C3.* : i i

n..

..............................e............. l A. ) i g g g g ,o cs: cs3 C53 153 t$3 c53 C53 G G G G3 G -4 0

s sw.cs::::c::

s:=: <:c m.-. s:ar.: Os2z O

O i 5 j i O ~ E: j j i i i CEB . CD = 52::: i g e e.--e. o: 5:ma " i i i. CE: i. O

ng;

O ............y Mi I ~

a!

i i i i N 8C:4 :. m e .cg: Saa o. M! CEB a.5 .:..............:..............i.............i.......eg..:............_. Cum z 9 i. i. i. j i '42* w o: i

E::

z: IP-* O z g;ma g;ma :. .g..............~..............:.............j.............[.....;....._ M g

z..

M:*

g i. i O .ct: hi N ~ v zi i. i cp i.............:..............:.............i..............i............ 4 S 9 w i.

w. :

M a:.c. O* I s2:: i i .i N i 3 O .............i.............*.......................... g* iEi. i. i. i i i. e q g i. CD e csz CEB G3 CD CD CER

==.4 O a s:=a.cz:::m::: sm.: ::c:E-- ::=a O s2 E G. 1

h D D 1 b APPENDIX Q J FLET VERIFICATION CALCULATIONS D D D D e l \\

O APPENDIX 9. FLET VERIFICATION CALCULATIONS O G.1 INTRODUCTION FLET (Ekaw Evaluation by I. earing Instability) [Q d] is software. that computes the load for growth of an existing crack or the load for crack-growth instability. The program considers g plasticity in the material near the flaw, a feature that makes it cpplicable to the ductile materials used in the pressure boundary of nuclear power plants. G.2 THE FRACTURE MECHANICS OF FLET g Prior to EPRI-supported research in elastic-plastic fracture starting in 1979 no simple general estimation f ormulas existed for flawed ductile materials. Each crack geometry and material was considered on a separate basis and generally required cxpensive elastic-plastic finite element analysis to determine a g flawed component's fitness-for-service. The methodology contained in FLET is the simplified elastic-plastic fracture mechanics of Kumar, German and Shih (Qt2] developed with EPRI support. The accuracy of the estimation formulas and procedures has been extensively checked by g comparison with specimen tests and elastic-plastic finite element calculations. In later work, directed to cylinders, German et al. [Eml] compared estimation scheme results with finite element calculations and pipe experiments and concluded that pipes can be modeled with sufficient accuracy by (these) 2-dimensional Gngineering approach solutions." Additional piping solutions g were added by Kumar et al. (G.4, G.5) that further confirmed the cccuracy of the technology by comparison with finite element calculations. More recently, Zahoor and Gamble [G.6] have extended the verification of the methodology for piping by comparing crack G initiation and instability data from tests of 20 pipes containing both circumferential through-the-wall and surface cracks. In both cases they conclude that the estimation procedures provide sufficient accuracy for engineering applications. O G.3 DESCRIPTION OF THE FLET SOFTHARE EPRI developed the program in early 1984 for use with round robin piping calculations held in San Antonio, Texas on June 22-23 [G. 7 ]. Development has continued and the EPRI results reported here use the current version. The code contains the estimation fracture mechanics methods described above. O G-1

O The program includes all of the piping solutions given in References G.2 through G.5. 9 For the leak-before-break piping application considered here, the colutions for the circumferential through-the-wall cylinder crack in combined tension and bending are the cost important. The software contains the tables of h functions with appropriate interpolation algorithms and performs computer simulation of the graphic procedures described in detail in Reference G.2. O Three solution options are available to the user. These are the J-integral tearing instability analysis (JT), the ductile fracture failure analysis diagram analysis (DPFAD), and the J-integral load analysis (JL). Except for small differences, usually each of these methods leads to the same crack initiation O cnd instability loads (G.71. For the leak-before-break analyses in this report, the DPFAD option is used to account for the crack growth between initiation and the instability load. G.4 FLET VERIFICATION CALCULATIONS O The following discussion concerns the accuracy of the fracture mechanics methodology discussed above and implemented in the EPRI FLET computer program. Three different problems are compared for load at initiation and at instability. O G.4.1 J at Initiation - NUREG 1061 Volume 3, Figure A-7 This figure compared experimental initiation J values for 8-inch diameter SA 106 Grade B steel pipes containing various length through-the-Wall circumferential cracks with estimation solutions by the NRC Pipe Crack Study Group (G. 8 ]. The load is pure O bending. A recalculation of the data using the FLET code with identical input and the corresponding plot of EPRI results is shown in Figure G.1. The plot also compares calculations by Zahoor and Gamble using a modified estimation formula (G.6] appropriate for O this problem with a similar FLET calculation. This modification corrects the plastic part of the J-integral formulas used for the circumferential through-the-wall pipe flaw. The effect is to reduce the conservatism in calculation of crack-driving force as shown in Figure G.1. The initiation J values are also given in Table G.1. The FLET results agree well with Reference G.6 and reasonably model the experimental results. O O G-2

4 O G.4.2 Crack-Growth Instability Loads for Flawed Neldments EPRI compares FLET results for the load-carrying ability of flawed austenitic stainless steel flux weldments with calculation O cade in support of the evaluation procedures found in the ASME Boiler and Pressure Vessel Code Section XI, INB-3640 (G. 9 ]. The EPRI calculations use identical material properties and the same through-the-wall crack for comparison with the ASME calculation. The analyses here are for a 28-inch diameter pipe in pure bending. Results are shown in Figure G.2 for a submerged arc weldment and in Figure G.3 for a shielded metal arc weldment. FLET agrees well with the ASME results for the SAN weld and is somewhat more conservative than the ASME SMAN calculation. Both analyses use the unmodified estimation formulas. g G.4.3 Initiation and Instability Loads in Combined Tension and Bending 4 These calculations compare loads computed with the three-dimensional finite element program ABAQUS [G.10), the NO-BREAK g. program (G.11 ] (which uses an estimation procedure similar to FLET), and the FLET program. The comparison calculations are i given in Reference G.12. EPRI presents the results in terms of multiples of faulted load that are added to the Service level A loads. EPRI assumed the g level A load is 564,670 lbs tension, and the faulted load is 36,000 lbs tension and 3,800,000 in-lb bending moment. These service level A and faulted loads correspond to the normal loads and the SSE load respectively of this report for the i cpecific line considered. g Table G.2 compares calculations for J/Jic, tearing modulus, and I load at instability, Table G.3 compares calculations for J/Jic in greater detail, and Table G.4 compares load (J) and the tearing modulus (T) at instability. 9, i The results for this set of calculations show fair agreement with l the ABAQUS and NO-BREAK calculations. l \\ G.5 CONCLUSIONS O These comparisons show reasonable agreement between the FLET j program, other calculational procedures, and experimental data. l EPRI concludes that the estimation fracture mechanics as i implemented in FLET is acceptable for leak-before-break cngineering calculations when used with the recommended factors G of safety. i S: G-3

3 REFERENCES G.1

Okamoto, A.,

and Norris, D., FLET: PIPE CRACK INSTABILITY PROGRAM. EPRI Draft Report, March 1986. G.2 Kumar. V.,

German, M.D.,

Shih, C.F.,

An Engineering Approach of Elastic-Plastic Fracture Analysis, EPRI Report NP1931, July 1981. ~ G.3 German, M.D. et al., Elastic-Plastic Fracture Analysis of Flawed Stainless Steel Pipes, EPRI Report 2608-LD, September 1982. G.4

Kumar, V.,

German, M.D.,

Hilkening, H.H.,

Andrews,

) H.R., de

Lorenzi, H.G.,

and Mowbray, D.F., Advances in Elastic-Plastic Fracture Analysis, EPRI Report NP3607, August 1984. G.5 "New Estimation Scheme Solutions," personal communication from V. Kumar to D.M. Norris, EPRI. J February 1986. G.6 Zahoor. A., and Gamble, R.M., Evaluation of Flawed Pipe Experiments, Final Report. EPRI Project RP 2457-8, July 1986 (in press). C) G.7

Kanninen, M.F.,

editor, Proc. CSNI/NRC Horkshop on Ductile Piping Fracture Mechanics, San Antonio, Texas, '21-22 June 1984 (draft report to be published as USNRC NUREG report). G.8 Report of the U.S. Nuclear Regulatory Commission Piping C) Review Committee. Evaluation of Potential for Pipe

Breaks, NUREG-1061 Volume 3 USNRC, Hashington D.C.,

November 1984. G.9 Evaluation of Flaws in Austenitic Steel Piping, prepared by Section XI Task Group for Piping Flaw 3 Evaluation. ASME boiler and Pressure Vessel Code Committee EPRI Report NP-4690-SR, Electric Power Research Institute, Palo Alto, California, July 1986. G.10 ABAQUS-ND: A Finite Element Code for Nonlinear Dynamic Analysis, EPRI NP-1552 CCM, Vols. 1-4, August 1980. J G.11 NO-BREAK, Fracture Mechanics Software for_ Piping Flaw Evaluation and Leak-Before-Break Analysis. User's Manual. Revision I, Novetech Corporation. Rockville, MD, January 1986. O G.12 Personal communication from B. Kee of Ontario Hydro to D.M. Norris of EPRI June 1986. 3 G-4

O TABLE G.1 O COMPARISON OF INITIATION J FOR EIGHT-INCH DIAMETER FERRITIC STEEL PIPES 9 J-T METHOD DPFAD METHOD K G.6

FLEI, Experiment Experiment Original Improved Original Improved Number EPRI EPRI EPRI EPRI 3

3,680. 13.261. 6,100. 13,230. 6,058. 7 5,400. 12,474. 5,800. 12,420. 5,802. 8 4,420. 7,579. 3,750. 7,512. 3,727. 11 2,340. 7,565. 3,650. 7,518. 3,571. O 12 3.110. 11,802. 5,300. 10,700, 4,826. 14 4,300. 8,013. 3,650. 7,960. 3,589, 15 2,850. 10.510, 4,720. 10,400, 4,621. O l e O O O ,,.--.--,.----.,-.-__.--.,.-..,._.-...-,.,__.-.c

b TABLE G.2 ) COMPARISON OF J/JIc, TEARING MODULUS, AND LOAD AT INSTABILITY J INSTABILITY ABAQUS NO-BREAK DPFAD METHOD (J-T) (FLET) IMPROVED ORIGINAL IMPROVED EPRI EPRI EPRI J/JIo 4.857 4.667 4.005 3.992 D T 14.750 15.950 na na D Faulted Load 5.194 5.374 4.734 5.220 Multiple J m J e 3

O' TABLE G.3 8 COMPARISON OF J/JIo 9 Faulted Load Multiple ABAQUS NO-BREAK DPFAD METHOD g (J-T) (FLET) IMPROVED ORIGINAL IMPROVED EPRI EPRI EPRI 1 0.141 0.130 0.151 0.149 g 2 0.415 0.358 0.436 0.414 3 0.967 0.814 0.989 0.872 4 2.033 1.712 2.192 1.746 g 5 4.092 3.450 4.676 3.349 6 8.041 6.704 9.633 6.272 e O O 1 O O

a O TABLE G.4 O COMPARISON OF TEARING MODULUS O i Faulted Load Multiple ABAQUS NO-BREAK FLET

• O (J-T)

(J-T) IMPROVED ORIGINAL IMPROVED EPRI EPRI EPRI O i 0.37 0.30 0.33 0.32 2 1.10 1.00 0.95 0.89 3 2.62 2.44 2.40 2.05

O 4

5.75 5.44 5.61 4.27 5 12.16 11.58 12.61 16.69 6 25.07 23.35 26.74 28.62 O o not applicable to the DPFAD method 'O 4 i I .O i .O

1 O e 6.0 i I . 4 Experimental Data I e EPRI Estimation Scheme (Ref.. G. 8) e 4.0 GImproved EPRI Estimation Scheme " A .o AFLET EPRI Estimation Sche e g5 3 AFLET Improved Scheme i m 1 l 5 8 l I t g I l h 2.0 3 i ee I 1.0 1 g o8 /, 2 0.6 v i g f Estimated Scatter Band j 0.4 I g j i l l l o% l l 0 l i I i l 0.2 g 0.7 0.8 0.9 1.0 Experimental Moment at Crack initiation / Limit Load 9 Figure G.1 Comparison of experimental data from pipe tests with calculations O O

D-D 3 SAU % 0 teroncen 4 1.o - 16 28 = 34 f

a. -

/// ll / 1 . o.s m l 0 g l o.4 0FLET calculations g o.2 o l I I l o at a2 0.3 0.4 0.5 0 rr.etion. circume.,.ne. < orir i Figure G.2 Ratio of Instability Moment to Limit Moment O as a Function ofCircumferential Throughwall Flaw Length for a 28-Inch Diameter Pipe, SAW at 550*F -O O

C, O SMAW 9 08 Asial Loed ] i d s 30. j m LD I 0.4 = a FLET calculations 0.2 = 0 l l I l 0 0.1 0.2 0.3 0.4 0.5 Fraction of Circumference (GlW) i Figure G.3 Ratio of Instability Load to Limit Load as a Function of G Circumferential Flaw Length for Bending and Axial Loadings, 28-Inch Diameter Pipe, SMAW at 550*F O 9: i O I .}}