NG-17-0111, Duane Arnold Energy Center, Revision 24 to Updated Final Safety Analysis Report, Chapter 3, Design of Structures, Components, Equipment, and Systems
Text
{{#Wiki_filter:°
°
°
°
°
UFSAR/DAEC - 1 T3.1-1 Revision 14 - 11/98 Table 3.1-1 Sheet 1 of 3 REACTOR COOLANT PRESSURE BOUNDARY Influent Lines Inside Drywell Outside Drywell
Feedwater a CV MOCV a. HPCI return b -- CV-MOV
- b. RCIC return b -- CV-MOV
- c. Cleanup return b -- CV-MOV RHR return to recirc.
c CV MOV Core spray c CV MOV CRD return d CV CV Standby liquid control e CV CV Other f -- -- a. MO1908 bypass CV -- Key: CV = check valve MOCV = motor-operated check valve MOV = motor-operated valve
aThat portion of the feedwater line that forms part of the reactor coolant pressure boundary and penetrates the primary containment has two isolation valves. The isolation valve inside the containment is a simple check valve. The isolation valve outside the containment is a stopcheck globe valve. Should a break occur in the feedwater line, the check valves prevent a significant loss of inventory and offer immediate isolation. During the postulated LOCA, it is desirable to maintain reactor coolant makeup from all sources of supply. For this reason, the isolation valve outside of containment does not automatically isolate on signal from the protection system. However, the valve is capable of being remotely closed from the control room to provide long-term leakage protection upon operator judgment that continued makeup from the feedwater source is
unnecessary.
bInfluent lines that form a portion of the reactor coolant pressure boundary but do not penetrate the primary containment must adequately reflect the importance to safety of isolating these piping systems. Pipes of this type include those portions of the RCIC, the reactor water cleanup (RWCU), and the HPCI lines that tie into the feedwater lines.
Each of these lines has two isolation valves in series. The first of these is a check valve, and the other is a motor-operated, automatic and remote manually actuated valve. The RCIC and HPCI lines are closed during normal operation, whereas, the RWCU line is open during operation and isolates from the reactor coolant pressure boundary upon receipt of signals from the protection system. UFSAR/DAEC - 1 T3.1-2 Revision 13 - 5/97 Notes - Table 3.1-1 Sheet 2 of 3
cThe RHR return lines to recirculation system and the core spray lines have check valves inside the containment that provide for immediate isolation in the event of a break upstream of these valves. In addition, the isolation valves outside the containment are normally closed, automatic and remote manually actuated valves designed to provide long-term leakage control in the event of a break in these lines. For the postulated LOCA, the protection system will initiate automatic opening of the injection valves at the appropriate time to ensure that acceptable fuel design limits are not exceeded.
dThat portion of the CRD return line that forms part of the reactor coolant pressure boundary and penetrates the primary containment has two isolation valves. Both valves are simple check valves and are located inside as well as outside the primary containment.
Criterion 55 states that a simple check valve may not be used as the automatic isolation valve outside the containment. During the postulated LOCA, it is desirable to maintain reactor coolant makeup from all sources of supply. For this reason, valves that automatically isolate upon signal from the protection system are not included in the design of this system. Should a break occur in the CRD return line, the check valves
would prevent significant loss of inventory and offer immediate isolation.
eThe standby liquid control line uses a simple check valve as the isolation valve inside as well as outside the primary containment. Criterion 55 states that a simple check valve may not be used as the automatic isolation valve outside the containment; however, should insertion of the liquid poison become necessary, it is imperative that the injection line be open. In the design of this system, it has been the accepted practice to omit an automatic valve that opens on signal as this introduces a possible failure mechanism. As a means of providing assurance for reliable timely actuation, an explosive valve is used. In this manner, the availability of the line is ensured. Because the standby liquid control line is a normally closed, nonflowing line, rupture of this line is very remote; however, should a break occur, the check valves provide positive actuation for immediate isolation.
f Other a. The MO1908 Bypass Lines is a 1/2 line bypassing MO1908 and completely contained within Primary Containment. This line has a check valve which allows flow only in the influent direction and is intended to relieve pressure between MO1908 and MO1909 during a LOCA. The pressure is relieved back to the Reactor Vessel. MO1908 is located on the RHR Shutdown Cooling Suction Line which is an effluent line.
UFSAR/DAEC - 1 T3.1-3 Revision 13 - 5/97 Notes - Table 3.1-1 Sheet 3 of 3
CRD Insert and Withdraw Lines
Criterion 55 concerns the reactor coolant pressure boundary penetrating the primary reactor containment. As shown in Table 3.2-4, the CRD insert and withdraw lines are not part of the reactor coolant pressure boundary.
The basis to which the CRD lines are designed is commensurate with the safety importance of isolating these lines. Since these lines are vital to the scram function, their operability is of utmost concern.
In the design of this system, it has been accepted practice to omit automatic valves for isolation purposes as this introduces a possible failure mechanism. As a means of providing positive actuation, manual shutoff valves are used. In the event of a break on these lines, the manual valves may be closed to ensure isolation. In addition, a ball valve located in the insert line is designed to automatically seal this line in the event of a break.
Finally, several breaks and combinations of breaks in the CRD lines have been postulated
and analyzed (see Section 4.6.2). The results of these analyses indicate that the worst situation causes a leak rate that is negligible compared to the makeup capability.
TIP System
Since the TIP system lines do not comprise a portion of the reactor coolant pressure
boundary, GDC 55 is not directly applicable to this specific class of lines. The basis to which these lines are designed is more closely described by GDC 54, which states in effect, that the isolation capability of a system be commensurate with the safety importance of that isolation. However, since the TIP lines communicate directly with the primary containment via the relief valves on the TIP indexer, conformance with GDC 56 must be addressed. GDC 56 can be satisfied by classifying these lines as nonessential instrument lines, subject to the acceptance criteria of Regulatory Guide 1.11. Such conformance has been demonstrated in NEDC-22253, BWR Owners Group Evaluation of Containment Is olation Concerns, October 1982. These and other safety features are described in the following paragraphs.
When the TIP system cable is inserted, the ball valve of the selected tube opens automatically so that the probe and cable may advance. A maximum of four valves may be opened at any one time to conduct the calibration, and any one guide tube is used, at most, a few
hours per year.
If closure of the line is required during calibration, a signal causes the cable to be retracted and the ball valve to close automatically after completion of cable withdrawal. To ensure isolation capability if a TIP cable fails to withdraw or a ball valve fails to close, an
explosive, shear valve is installed in each line. Upon receipt of a signal, this explosive valve will shear the TIP cable and seal the guide tube. UFSAR/DAEC - 1 T3.1-4 Revision 13 - 5/97 Table 3.1-2 REACTOR COOLANT PRESSURE BOUNDARY Effluent Lines Inside Drywell Outside Drywell Main steam NOV NOV Reactor water cleanup MOV MOV RHR shutdown cooling MOV MOV Main steam drain MOV MOV RCIC turbine steam MOV MOV HPCI turbine steam MOV MOV Key: NOV = nitrogen-operated valve MOV = motor-operated valve
- UFSAR/DAEC-1 TABLE 3.2-1 INDEX T3.2-1 Revision 22 - 5/13 Sys # SysName SysCode I Reactor System B11 II Nuclear Boiler System B21 III Recirculation System B31 IV CRD Hydraulic System C11 V Standby Liquid Control System C41 VI Neutron Monitoring System C51 VII Reactor Protection System C71 VIII Process Radiation Monitors D11 IX RHR System E11 X Low Pressure Core Spray E21 XI HPCI System E41 XII RCIC System E51 XIII Fuel Service Equipment F11 XIV Reactor Vessel Service Equipment F13 XV In-Vessel Service Equipment F14 XVI Refueling Equipment F15 XVII Storage Equipment F16 XVIII Radwaste System G11 XIX Reactor Water Cleanup System G31 XX Fuel Pool Cooling and Cleanup System G41 XXI Control Room and Remote Shutdown Panels H11 C61 XXII Local Panels and Racks H21 XXIII Offgas System N62 XXIV Emergency Service Water E13 XXV RHR Service Water System E12 XXVI RBCCW P42 XXVII Well Water System P46 XXVIII Pneumatic Systems T48 P50 XXIX Diesel Generator Systems R43 XXX Containment Atmosphere Control System T48 XXXI Standby Gas Treatment System T46 XXXII ECCS Equipment Area Cooling System T41 XXXIII Power Conversion System N11 N21 XXXIV Condensate Storage and Transfer System P11 XXXV Auxiliary a-c Power System R20, R22-24 XXXVI 125/250 Volt d-c Power System R42 XXXVII River Water Supply W10 XXXVIII Not Used XXXIX HVAC XXXX Miscellaneous Components UFSAR/DAEC-1 TABLE 3.2-1 INDEX (by SysName)
T3.2-2 Revision 22 - 5/13 Sys # SysName SysCode XXXVI 125/250 Volt d-c Power System R42 XXXV Auxiliary a-c Power System R20, R22-24 XXXIV Condensate Storage and Transfer System P11 XXX Containment Atmosphere Control System T48 XXI Control Room and Remote Shutdown Panels H11 C61 IV CRD Hydraulic System C11 XXIX Diesel Generator Systems R43 XXXII ECCS Equipment Area Cooling System T41 XXIV Emergency Service Water E13 XX Fuel Pool Cooling and Cleanup System G41 XIII Fuel Service Equipment F11 XI HPCI System E41 XXXIX HVAC XV In-Vessel Service Equipment F14 XXII Local Panels and Racks H21 X Low Pressure Core Spray E21 XXXX Miscellaneous Components VI Neutron Monitoring System C51 II Nuclear Boiler System B21 XXIII Offgas System N62 XXVIII Pneumatic Systems T48 P50 XXXIII Power Conversion System N11 N21 VIII Process Radiation Monitors D11 XVIII Radwaste System G11 XXVI RBCCW P42 XII RCIC System E51 VII Reactor Protection System C71 I Reactor System B11 XIV Reactor Vessel Service Equipment F13 XIX Reactor Water Cleanup G31 III Recirculation System B31 XVI Refueling Equipment F15 XXV RHR Service Water System E12 IX RHR System E11 XXXVII River Water Supply W10 XXXI Standby Gas Treatment System T46 V Standby Liquid Control System C41 XVII Storage Equipment F16 XXVII Well Water System P46 UFSAR/DAEC-1 TABLE 3.2-1 DAEC Classification of Components in System T3.2-3 Revision 22 - 5/13 Scope Quality of Safety Construction Quality Assurance Seismic Supply Class Code Code Group Req. Category PO Date Comments Principle Component (a) (b) Class (c) Class (d) (e)(f) (g) Footnotes (h) I Reactor System B11 1 Reactor vessel GE/C 1 A ASME Section III, 1965 Edition, Summer A B I - 1d No 1967 Addenda 2 Reactor vessel support skirt GE 1 A ASME Section III, 1965 Edition, Summer A B I - 1d No 1967 Addenda 3 Reactor vessel appurtenances, GE 1 A ASME Section III, 1965 Edition, Summer A B I - 1d No pressure retaining portions 1967 Addenda 4 CRD Housing Supports (Shoot-out GE 2 - - - B I - - No Steel) 5 Reactor internal structures, GE 2 A ASME Section III, 1965 Edition, Summer NA B I - 1d,1u No engineered safety features 1967 Addenda 6 Reactor internal structures, other GE Other - - NA B I - 1u No 7 Control rods GE 2 - - NA B I - - No 8 Control rod drives GE 1 1 Section III, Class 1 appurtenances A B I - - No 9 Core support structure GE 2 A ASME Section III, 1965 Edition, Summer NA B I - 1d,1u No 1967 Addenda 10 Power range detector hardware GE 2 - - B B I - 1a,1b No 11 Fuel assemblies GE 2 - - NA B I - - No 12 RX Vessel Stabilizer GE 2 A ASME Section III, 1965 Edition, Summer NA B I - - No 1967 Addenda 13 Refueling bellows GE Other - - - - NA - - No II Nuclear Boiler System B21 1 Vessels, level instrumentation GE 1 - - A B I - - No condensing chambers 2 Vessels, N2 accumulators B 2 2 ASME Section III
- B I 02/26/73 - No 3 Piping, relief valve discharge B 3 3 USAS B31.7-1969 C B I 07/30/70 - No 4 Piping, main steam within outermost GE 1 - ANSI B31.1.0 + Code Cases N2, N7, N9, A B I 12/05/69 1a,1b No isolation valve N10 5 Pipe supports, main steam GE 1 1 Requirements for Class 1 piping supports in A B I 02/26/71 - No ANSI B31.7.
6 Pipe restraints, main steam B 2 - - NA B I - - No 7 Piping, other within outermost B 1 1 USAS B31.7-1969 A B I 07/30/70 1a,1b No isolation valves 8 Piping, instrumentation beyond B Other - USAS B31.1.0-1967
- B - - 1a,1b No outermost isolation valves
UFSAR/DAEC-1 TABLE 3.2-1 DAEC Classification of Components in System T3.2-4 Revision 22 - 5/13 Scope Quality of Safety Construction Quality Assurance Seismic Supply Class Code Code Group Req. Category PO Date Comments Principle Component (a) (b) Class (c) Class (d) (e)(f) (g) Footnotes (h) 9 Safety valves GE 1 1 ANSI B31.1.0, Addenda and applicable A B I 12/30/69 1f No code cases or NDE standards of B31.1. Code Cases N2, N7, N9, N10 except that the acceptance standards for Class 1 valves in the Draft ASME Code for Pumps and Valves for Nuclear Power may be applied. Use ANSI 16.5 or MSS-SP-66 for design. 10 Relief valves GE 1 1 ASME Code, Section III, 1968 Edition, A B I 12/30/69 1f No Article 9, with 1968 Winter Addenda. 11 Valves, main steam isolation valves GE 1 1 ANSI B31.1.0, Addenda and applicable A B I 10/15/69 1f No code cases or NDE standards of B31.1. Code Cases N2, N7, N9, N10 except that the acceptance standards for Class 1 valves in the Draft ASME Code for Pumps and Valves for Nuclear Power may be applied. Use ANSI 16.5 or MSS-SP-66 for design. 12 Valves, other, isolation valves and B 1 1 ASME Code for Pumps and Valves for A B I 10/16/70 1f No within Nuclear Power 13 Valves, instrumentation beyond B Other - USAS B31.1.0-1967
- - - - 1a,1b,1f No outermost isolation valves 14 Mechanical modules, instrumentation, GE 2 - - B B I - 1c No with safety function 15 Electrical modules with safety function GE 2 - - NA B I - 1c No 16 Cable, with safety function B 2 - - - B I - - No III Recirculation System B31 1 Piping GE 1 - ANSI B31.1.0 + Code Cases N2, N7, N9, A B I 12/05/69 1a,1b No N10 2 Pipe suspension, recirculation line GE 1 1 Requirements for Class 1 piping supports in A B I 02/26/71 - No ANSI B31.7.
3 Pipe restraints recirculation line GE 2 - - NA B I - - No 4 Pumps GE 1 1 Draft ASME Code for Pumps and Valves A B I 11/22/68 1e No for Nuclear Power or NDE and acceptance requirements of ANSI B31.1 Code Cases N7,N9,N10 + Design Guide for sizing pressure parts in ASME Boiler and Pressure Vessel Code [1968] Section III, Class C
UFSAR/DAEC-1 TABLE 3.2-1 DAEC Classification of Components in System T3.2-5 Revision 22 - 5/13 Scope Quality of Safety Construction Quality Assurance Seismic Supply Class Code Code Group Req. Category PO Date Comments Principle Component (a) (b) Class (c) Class (d) (e)(f) (g) Footnotes (h) 5 Valves GE 1 1 Requirements for Class 1 valves in Draft A B I 03/20/70 1f No ASME Code for Pumps and Valves for Nuclear Power and requirements applicable to valves in ASME Section III
[1968 Edition plus addenda], articles 1 and
- 8. 6 Motor, pump GE Other - NEMA Standards NA D I - 1x No 7 Electrical modules, with safety GE 2 - - NA B I - 1c No function 8 Cable with safety function B 2 - - - B I - - No IV CRD Hydraulic System C11 1 Valves, isolation, water return line B 1 1 ANSI B31.1.0 or ASME Code for Pumps A B I 12/19/72 1f No and Valves for Nuclear Power, Class 1, or ASME Section III, 1971 Edition, Class 1 2 Valves, scram discharge volume lines GE/B 2/2 2 ASME Code for Pumps and Valves for B/B B/B I/I 12/19/72 1f No Nuclear Power 3 Valves insert and withdraw lines GE/B 2/2 2 ANSI B31.1.0 or ASME Code for Pumps B/B B/B I/I 12/19/72 1f No and Valves for Nuclear Power or ASME Section III, 1971 Edition, Class 1 4 Valves, other B Other 2 ASME Section III, 1971 Edition D D NA 12/19/72 1a,1f,1l Yes 5 Piping, water return line within B 1 1 ASME Section III, 1971 Edition A B I 12/19/72 - No isolation valves 6 Piping, scram discharge volume lines B 2 2 ASME Section III, 1971 Edition B B I 12/19/72 - No 7 Piping, insert and withdraw lines B 2 2 ASME Section III, 1971 Edition B B I 12/19/72 - No 8 Piping, other B Other 2 ASME Section III, 1971 Edition D D NA 12/19/72 1a,1b,1l Yes 9 Hydraulic control unit GE 2 - - Special B I - - Yes 10 Electrical modules, with safety GE 2 - - NA B I - 1c No function 11 Cable, with safety function B 2 - - - B I - - No V Standby Liquid Control System C41 1 Standby liquid control tank GE 2 - API-650 and ASME Section VIII, Div. 1 B B I - - Yes 2 Pump GE 2 2 ASME Code for Pumps and Valves for B B I - 1e No Nuclear Power 3 Pump motor GE 2 - NA B I - - No 4 Valves, explosive GE 2 2 ASME Code for Pumps and Valves for B B I - 1f No Nuclear Power 5 Valves, isolation and within B 1 1 ASME Code for Pumps and Valves for A B I 10/16/70 1f No Nuclear Power 6 Valves, beyond isolation valves B 2 2 ASME Code for Pumps and Valves for B B I 10/16/70 1f No Nuclear Power 7 Piping, within isolation valves B 1 1 USAS B31.7-1969 A B I 07/30/70 1a,1b No UFSAR/DAEC-1 TABLE 3.2-1 DAEC Classification of Components in System T3.2-6 Revision 22 - 5/13 Scope Quality of Safety Construction Quality Assurance Seismic Supply Class Code Code Group Req. Category PO Date Comments Principle Component (a) (b) Class (c) Class (d) (e)(f) (g) Footnotes (h) 8 Piping, beyond isolation valves B 2 2 USAS B31.7-1969 B B I 07/30/70 1a,1b,1m No 9 Electrical modules, with safety GE 2 - - NA B I - 1c No function 10 Cable, with safety function B 2 - - - B I - - No 11 Test Tank GE Other - - D D NA - - No 12 Piping, between test tank and its B Other - USAS B31.1.0 D D NA - - No isolation valves VI Neutron Monitoring System C51 1 Piping, TIP GE 2 - - B B I - 1a,1b No 2 Valves, isolation, TIP subsystem GE 2 - - B B I - 1f No 3 Electrical modules, IRM and APRM GE 2 - - NA B I - 1c,1w Yes 4 Cable, IRM and APRM B 2 - - - B I - - No VII Reactor Protection System C71 1 Electrical modules GE 2 - - NA B I - 1c No 2 Cable B 2 - - - B I - - No VIII Process Radiation Monitors D11 1 Electrical modules for main steam GE 2 - - NA B I - 1c No line and reactor building ventilation monitors 2 Cable for main steam line and reactor B 2 - - - B I - - No building ventilation monitors IX RHR System E11 1 Heat exchangers, primary side GE 2 B ASME Section III, Class B and TEMA-C B B I 08/15/69 - No 2 Heat exchangers, secondary side GE 3 - ASME Section VIII, Div. 1, and TEMA-C C B I 08/15/69 - No 3 Piping, within outermost LPCI &
B 1 1 USAS B31.7-1969 A B I 07/30/70 1a,1b No shutdown cooling isolation valves 4 Piping, other B 2 2 USAS B31.7-1969 B B I 07/30/70 1a,1b No 5 Pumps GE 2 2 Draft ASME Code for Pumps and Valves B B I 09/17/69 1e No for Nuclear Power or NDE and acceptance requirements of ANSI B31.1 Code Cases N7,N9,N10 + Design Guide for sizing pressure parts in ASME Boiler and Pressure Vessel Code [1968] Section III, Class C 6 Pump motors GE 2 - - NA B I - - No 7 Valves, isolation, LPCI & shutdown B 1 1 ASME Code for Pumps and Valves for A B I 10/16/70 1f No cooling lines Nuclear Power 8 Valves, isolation, other B 2 2 ASME Code for Pumps and Valves for B B I 10/16/70 1f No Nuclear Power 9 Valves, beyond isolation valves B 2 2 ASME Code for Pumps and Valves for B B I 10/16/70 1f No Nuclear Power 10 Mechanical modules GE 2 - - B B I - 1c No 2011-008 UFSAR/DAEC-1 TABLE 3.2-1 DAEC Classification of Components in System T3.2-7 Revision 22 - 5/13 Scope Quality of Safety Construction Quality Assurance Seismic Supply Class Code Code Group Req. Category PO Date Comments Principle Component (a) (b) Class (c) Class (d) (e)(f) (g) Footnotes (h) 11 Electrical modules, with safety GE 2 - - NA B I - 1c No function 12 Cable, with safety function B 2 - - - B I - - No X Low Pressure Core Spray E21 1 Piping, within outermost isolation B 1 1 USAS B31.7-1969 A B I 07/30/70 1a,1b No valves 2 Piping, beyond outermost isolation B 2 2 USAS B31.7-1969 B B I 07/30/70 1a,1b No valves 3 Piping, floodup line to condensate B Other - USAS B31.1.0 C B I - 1a,1b No storage tank 4 Pumps GE 2 2 Draft ASME Code for Pumps and Valves B B I 09/17/69 1e No for Nuclear Power or NDE and acceptance requirements of ANSI B31.1 Code Cases N7,N9,N10 + Design Guide for sizing pressure parts in ASME Boiler and Pressure Vessel Code [1968] Section III, Class C 5 Pump motors GE 2 - - NA B I - - No 6 Valves, isolation and within B 1 1 ASME Code for Pumps and Valves for A B I 10/16/70 1f No Nuclear Power 7 Valves, beyond outermost isolation B 2 2 ASME Code for Pumps and Valves for B B I 10/16/70 1f No valves Nuclear Power 8 Valves, floodup line to condensate B Other - USAS B31.1.0 C B I - 1f No storage tank 9 Electrical modules with safety function GE 2 - - NA B I - 1c No 10 Cable, with safety function B 2 - - - B I - - No XI HPCI System E41 1 Piping, within outermost isolation B 1 1 USAS B31.7-1969 A B I 07/30/70 1a,1b No valves 2 Piping, beyond outermost isolation B 2 2 USAS B31.7-1969 B B I 07/30/70 1a,1b No valves 3 Piping, return test line to condensate B Other - USAS B31.1.0 D D NA 07/30/70 1a,1b,1r No storage tank beyond second isolation valve 4 Pumps GE 2 2 Draft ASME Code for Pumps and Valves B B I 07/31/69 1e No for Nuclear Power or NDE and acceptance requirements of ANSI B31.1 Code Cases N7,N9,N10 + Design Guide for sizing pressure parts in ASME Boiler and Pressure Vessel Code [1968] Section III, Class C
UFSAR/DAEC-1 TABLE 3.2-1 DAEC Classification of Components in System T3.2-8 Revision 22 - 5/13 Scope Quality of Safety Construction Quality Assurance Seismic Supply Class Code Code Group Req. Category PO Date Comments Principle Component (a) (b) Class (c) Class (d) (e)(f) (g) Footnotes (h) 5 Valves, isolation and within B 1 1 ASME Code for Pumps and Valves for A B I 10/16/70 1f No Nuclear Power 6 Valves, return test line to condensate B Other - USAS B31.1.0 D D NA 10/16/70 1f No storage beyond second isolation valve 7 Vacuum pump discharge line B Other - USAS B31.1.0 D D NA - 1a,1b No 8 Valves, other B 2 2 ASME Code for Pumps and Valves for B B I 10/16/70 1f No Nuclear Power 9 Turbine GE 2 - NEMA Standards for Mechanical Drive NA B I 07/29/69 1g Yes Steam Turbine 10 Electrical modules, with safety GE 2 - - NA B I - 1c No function 11 Cable, with safety function B 2 - - - B I - - No XII RCIC System E51 1 Piping, within outermost isolation B 1 1 USAS B31.7-1969 A B I 07/30/70 1a,1b No valves 2 Piping, beyond outermost isolation B 2 2 USAS B31.7-1969 B B I 07/30/70 1a,1b No valves 3 Piping, return test line to condensate B Other - USAS B31.1.0 D D NA 07/30/70 1a,1b,1r No storage tank beyond second isolation valve 4 Pumps GE 2 2 Draft ASME Code for Pumps and Valves B B I - 1e No for Nuclear Power or NDE and acceptance requirements of ANSI B31.1 Code Cases N7,N9,N10 + Design Guide for sizing pressure parts in ASME Boiler and Pressure Vessel Code [1968] Section III, Class C 5 Valves, isolation and within B 1 1 ASME Code for Pumps and Valves for A B I 10/16/70 1f No Nuclear Power 6 Valves, return test line to condensate B Other - USAS B31.1.0 D D NA 10/16/70 1f No storage beyond second isolation valve 7 Vacuum pump discharge line B Other - USAS B31.1.0 D D NA - 1a,1b No 8 Valves, other B 2 2 ASME Code for Pumps and Valves for B B I 10/16/70 1f No Nuclear Power 9 Turbine GE 2 - NEMA Standards for Mechanical Drive NA B I 07/29/69 1g Yes Steam Turbine 10 Electrical modules, with safety GE 2 - - NA B I - 1c No function 11 Cable, with safety function B 2 - - - B I - - No XIII Fuel Service Equipment F11 1 Fuel preparation machine GE 3 - - NA B I - - No UFSAR/DAEC-1 TABLE 3.2-1 DAEC Classification of Components in System T3.2-9 Revision 22 - 5/13 Scope Quality of Safety Construction Quality Assurance Seismic Supply Class Code Code Group Req. Category PO Date Comments Principle Component (a) (b) Class (c) Class (d) (e)(f) (g) Footnotes (h) 2 General purpose grapple GE 2 - - NA B I - - No XIV Reactor Vessel Service Equipment F13 1 Steam line plugs GE 3 - - NA B I - - No 2 Dryer and separator sling and head GE 2 - - NA B I - - No strongback XV In-Vessel Service Equipment F14 1 Control rod grapple GE 2 - - NA B I - - No 2 Lightweight Work Platform GE Other - - NA D I - - No 3 360 Degree Work Platform GE Other - - NA B I 12/31/07 - No XVI Refueling Equipment F15 1 Refueling equipment platform GE 2 - - NA B I - - No assembly XVII Storage Equipment F16 1 Fuel storage racks GE 2 - - NA B I - - No 2 Defective fuel storage container GE 3 - - - B I - - No XVIII Radwaste System G11 1 Tanks, Atmospheric GE/B Oth/Oth API-650 or AWWA-D100 or ANSI B96.1 C&D D/D NA/NA 07/16/70 1h Yes or equivalent plus NDE per ASME Section VIII Div. 1. 2 Heat exchangers GE/B Oth/Oth - - D/D D/D NA/NA - - No 3 Piping and valves, containment B 2 2 USAS B31.7-1969 B B I - 1a,1b No isolation 4 Piping, other B Other 3 USAS B31.7-1969 C&D D NA - 1a,1b,1j Yes 5 Pumps GE Other 3 ANSI B31.1.0 or ASME Code for Pumps C&D D NA 07/03/72 1e,1j No and Valves for Nuclear Power, Class 3, or ASME Section III, 1971 Edition, Class 3 6 Valves, flow control and filter system GE Other 3 ASME Code for Pumps and Valves for C&D D NA - 1f,1j No Nuclear Power 7 Valves, other B Other 3 ANSI B31.1.0 or ASME Code for Pumps C&D D NA - 1f,1j No and Valves for Nuclear Power, Class 3, or ASME Section III, 1971 Edition, Class 3 8 Mechanical modules GE Other - - C&D D NA - 1c,1j No XIX Reactor Water Cleanup System G31 1 Vessels: filter/demineralizer GE Other C ASME Section III C B I 12/17/70 1t No 2 Heat exchangers GE Other - ASME Section VIII and TEMA-C C B I 09/25/67 1t No 3 Piping, within outermost isolation B 1 1 USAS B31.7-1969 A B I 07/30/70 1a,1b No valves 4 Piping, beyond outermost isolation B Other 3 USAS B31.7-1969 C B I 07/30/70 1a,1b,1o, Yes valves 1t 5 Pumps GE Other 3 ANSI B31.1.0 or ASME Code for Pumps C B I 12/23/69 1e,1t No and Valves for Nuclear Power, Class 3, or ASME Section III, 1971 Edition, Class 3
UFSAR/DAEC-1 TABLE 3.2-1 DAEC Classification of Components in System T3.2-10 Revision 22 - 5/13 Scope Quality of Safety Construction Quality Assurance Seismic Supply Class Code Code Group Req. Category PO Date Comments Principle Component (a) (b) Class (c) Class (d) (e)(f) (g) Footnotes (h) 6 Valves, isolation valves and within B 1 1 ANSI B31.1.0 or ASME Code for Pumps A B I - 1f No and Valves for Nuclear Power, Class 1, or ASME Section III, 1971 Edition, Class 1 7 Valves, beyond outermost isolation GE/B Oth/Oth 3 ANSI B31.1.0 or ASME Code for Pumps C/C B I - 1f,1o,1t No valves and Valves for Nuclear Power, Class 3, or ASME Section III, 1971 Edition, Class 3 8 Mechanical modules GE Other - - - B I - 1c,1t No XX Fuel Pool Cooling and Cleanup System G41 1 Vessels, filter/demineralizers GE Other - ASME Section VIII, Div. 1. C D NA - - No 2 Vessels, other B Other - ASME Section VIII, Div. 1. C D NA - - No 3 Heat exchangers GE Other - ASME Section VIII, Div. 1, and TEMA-C C D NA 12/23/69 - No 4 Pumps GE Other 3 ANSI B31.1.0 or ASME Code for Pumps C D NA - - No and Valves for Nuclear Power, Class 3, or ASME Section III, 1971 Edition, Class 3 5 Piping B Other 3 USAS B31.7-1969 C D NA - 1a No 6 Valves B Other - ANSI B31.1.0 or ASME Code for Pumps C&D D NA - 1f No and Valves for Nuclear Power, Class 3, or ASME Section III, 1971 Edition, Class 3 XXI Control Room & Remote Shutdown Panels H11, C61 1 Electrical modules, with safety GE 2 - - NA B I - 1c No function 2 Cable, with safety function B 2 - - - B I - - No XXII Local Panels and Racks H21 1 Electrical modules, with safety GE/B 2/2 - - B/- B/B I/I - 1c No function 2 Cable, with safety function B 2 - - - B I - - No XXIII Offgas System N62 1 Tanks GE Other - AWWA D100 or API-650
- D NA 10/21/71 1h No 2 Heat exchangers GE Other - Section III & TEMA-C - D NA 11/20/72 - No 3 Piping B Other 3 ASME Section III-1971 - D NA 07/27/72 1k Yes 4 Pumps GE Other - - D D NA - 1e,1k Yes 5 Valves, flow control B Other - - D D NA - 1f,1k Yes 6 Valves, other B Other - - D D NA 11/29/71 1f,1k Yes 7 Mechanical modules GE Other - - D D NA - 1c No 8 Pressure vessels GE Other - - D D NA 10/21/71 - No XXIV Emergency Service Water E13 1 Piping B 3 - ANSI B31.1.0-1967 D+QA B I 07/30/70 1y Yes 2 Pumps B 3 3 ASME Code for Pumps and Valves for C B I - 1e No Nuclear Power
UFSAR/DAEC-1 TABLE 3.2-1 DAEC Classification of Components in System T3.2-11 Revision 22 - 5/13 Scope Quality of Safety Construction Quality Assurance Seismic Supply Class Code Code Group Req. Category PO Date Comments Principle Component (a) (b) Class (c) Class (d) (e)(f) (g) Footnotes (h) 3 Pump motors B 3 - - B I - - No 4 Valves B 3 3 ASME Code for Pumps and Valves for C B I 10/16/70 1f No Nuclear Power 5 Electrical modules, with safety B 3 - - - B I - 1c No function 6 Cable, with safety function B 3 - - - B I - - No XXV RHR Service Water System E12 1 Piping B 3 3 USAS B31.7-1969 C B I 07/30/70 1p, 1y No 2 Pumps B 3 3 ANSI B31.1.0 or ASME Code for Pumps C B I - 1e No and Valves for Nuclear Power, Class 3, or ASME Section III, 1971 Edition, Class 3 3 Pump motors B 3 - - - B I - - No 4 Valves B 3 3 ASME Code for Pumps and Valves for C B I 10/16/70 1f,1p No Nuclear Power 5 Electrical modules, with safety B 3 - - - B I - 1c No function 6 Cable, with safety function B 3 - - - B I - - No XXVI RBCCW P42 1 Piping, and valves forming part of B 2 B ASME Nuclear Vessels Code Section III, B B I 07/30/70 - No primary containment boundary Extension of Containment Code Cases 1425, 1426 & 1427 2 Piping and valves inside drywell B Other - USAS B31.1.0 D B I - 1s No 3 Piping and valves, other B Other - USAS B31.1.0 D - NA - - No XXVII Well Water System P46 1 Piping and valves forming part of B 2 B ASME Nuclear Vessels Code, Section III, B B I - - No primary containment boundary Extension of Containment Code Cases 1425, 1426 & 1427 2 Piping and valves inside drywell B Other - USAS B31.1.0 D B I - 1s No 3 Piping and valves, other B Other - USAS B31.1.0 D - NA - - No XXVIII Pneumatic Systems T48, P50 1 Nitrogen vessels, accumulators, B 2 2 ASME Section III
- B I 02/26/73 - No supporting safety-related systems 2 Nitrogen piping and valves in lines B 2 2 ASME Section III - B I - - No between above accumulators and safety-related systems 3 Nitrogen piping and valves forming B 2 2 ASME Section III - B I - 1b No part of containment boundary 4 Instrument air vessels, accumulators, B 3 - - - B I 02/26/73 - No supporting safety-related systems
UFSAR/DAEC-1 TABLE 3.2-1 DAEC Classification of Components in System T3.2-12 Revision 22 - 5/13 Scope Quality of Safety Construction Quality Assurance Seismic Supply Class Code Code Group Req. Category PO Date Comments Principle Component (a) (b) Class (c) Class (d) (e)(f) (g) Footnotes (h) 5 Instrument air piping and valves in B 3 - - - B I - - No lines between above accumulators and safety-related systems XXIX Diesel Generator Systems R43 1 Day tanks B 3 - API-650 or AWWA-D100 or ANSI B96.1 C B I - - No or equivalent plus NDE per ASME Section VIII, Div. 1. 2 Piping and valves, fuel oil system and B 3 3 USAS B31.7 for pipe and ASME Code for C B I - - No diesel service water system Pumps and Valves for Nuclear Power 3 Pumps, fuel oil system and diesel B 3 3 ASME Code for Pumps and Valves for C B I - - No service water system Nuclear Power 4 Pump motors, fuel oil system and B 3 - - - B I - - No diesel service water system 5 Diesel-generators B 2 - - - B I - - No 6 Electrical modules with safety function B 3 - - - B I - 1c No 7 Cable, with safety function B 3 - - - B I - - No 8 Diesel fuel storage tanks B Other - API-650 or AWWA-D100 or ANSI B96.1 C B I - - No or equivalent plus NDE per ASME Section VIII, Div. 1. 9 Diesel Air Start System B 2 - - - B I - - No XXX Containment Atmosphere Control System T48 1 Piping and valves from primary B 2 B ASME Nuclear Vessels Code Section III, B B I 07/30/70 - No containment through outer isolation Extension of Containment Code Cases valve 1425, 1426 and 1427 XXXI Standby Gas Treatment System T46 1 All components with safety function, B 3 - - - B I - - No including offgas stack dilution fans XXXII ECCS Equipment Area Cooling System T41 1 All components with safety functions B 3 - - - B I - - No XXXIII Power Conversion System N11, N21 1 Main steam piping from outboard B Other - USAS B31.1.0 D+QA B I 07/30/70 1a Yes MSIV to turbine stop valves and branch line piping up to and including first valve 2 Steam piping and valves, other B Other - USAS B31.1.0 D+QA D NA 07/30/70 1a No 3 Reactor feedwater piping and valves, B 1 1 USAS B31.7-1969 A B I 07/30/70 1a,1b,1f No RPV to outermost isolation valve 4 Reactor feedwater piping and valves, B Other - USAS B31.1.0 D+QA D NA 07/30/70 1a,1f Yes other 5 MSIV-LTS piping and valves IE Other - USAS B31.1.0 D+QA D A - - Yes XXXIV Condensate Storage and Transfer System P11 UFSAR/DAEC-1 TABLE 3.2-1 DAEC Classification of Components in System T3.2-13 Revision 22 - 5/13 Scope Quality of Safety Construction Quality Assurance Seismic Supply Class Code Code Group Req. Category PO Date Comments Principle Component (a) (b) Class (c) Class (d) (e)(f) (g) Footnotes (h) 1 Condensate storage tank B Other - API-650 plus augmented NDE of welds D+QA - NA 07/30/70 1i Yes 2 Piping and valves B Other - USAS B31.1.0 D D NA - 1f,1q No 3 Other components B Other - - D D NA - 1q No XXXV Auxiliary a-c Power System R20, R22-24 1 All components with safety function B 2 - - - B I - - Yes XXXVI 125/250 Volt d-c Power System R42 1 All components with safety function B 2 - - - B I - - No XXXVII River Water Supply W10 1 Piping, pumps and valves B 3 - ANSI B31.1.0 D+QA B I - 1f, 1y Yes 2 Intake traveling screen, trash rakes B 3 - - - B I - - No 3 Pump motors B 3 - - - B I - - No XXXVIII Not Used - - No XXXIX HVAC 1 Control room
- - B I - - No 2 Pump house - - B I - - No 3 Emergency diesel generator room - - B I - - No 4 Reactor building secondary - - B I - - No containment isolation dampers 5 Battery rooms - - B I - - No 6 Intake structure - - B I - - No 7 Essential switchgear rooms - - B I - - No XXXX Miscellaneous Components 1 Reactor Building Crane B 3 - - - B I - - No 2 Containment Penetrations for Process B 2 - - - B I - - No Piping and Electrical General General n/a Yes UFSAR/DAEC-1 TABLE 3.2-1 DAEC Classification of Components in System Footnotes T3.2-14 Revision 22 - 5/13 Footnote # Footnote Systems a GE = General Electric; B = Bechtel; C = CB&I; IE = Iowa Electric(=DAEC); DAEC = Duane Arnold Energy Center b 1, 2, 3, "other" = safety classes defined in Section 3.2.4; "unc" = unclassified as defined in Section 3.2.4. c The equipment shall be constructed in accordance with the codes listed in Table 3.2-2, if no Code of Construction is provided in this table. The term "construction," as used in this UFSAR, includes provisions for design, materials, fabrication, erection, testing and inspection.
d B = The equipment shall meet the quality assurance requirements of 10CFR50, Appendix B, in accordance with the quality assurance program described in Chapter 17. D = The equipment shall be constructed in accordance with the quality assurance requirements consistent with the good practices for steam power plants. e I = The equipment shall be constructed in accordance with the seismic requirements for the safe shutdown earthquake, as described in Section 3.7, Seismic Design. Seismic adequacy of certain equipment was verified by Seismic Qualification Utility Group (SQUG) methodology. The NRC issued a Safety Evaluation on the resolution of USI A-46 at the DAEC on July 29, 1998. NA = The seismic requirements for the safe shutdown earthquake are not applicable to the equipment. A = Seismic Adequate for MSIV-LTS. f Portions of 'non-seismic category I' piping (seismic category NA) passing through rooms containing safeguard equipment are seismically supported as seismic category I. g Date on the purchase order for the component. Where provided, this can be used to establish the code edition and addenda in effect for the component. h A "yes" in this column signifies there is a comment regarding the item at the end of Table 3.2-1. 1a The following items are applicable to instrument, sampling or small bore (3/4" NPS and smaller), as noted: **(1)Lines 3/4" and smaller which are part of the reactor coolant boundary shall be Safety Class 2. **(2)All instrument lines which are connected to the reactor coolant pressure boundary and are utilized to actuate safety systems shall be Safety Class 2 from the outer isolation valve or the process shutoff valve (root valve) to the sensing instrument. **(3)All instrument lines which are connected to the reactor coolant pressure boundary and are not utilized to actuate safety systems shall be Quality Group D from the outer isolation valve or the process shutoff valve (root valve) to the sensing instrumentation. **(4) All other instrument lines through the root valve shall be of the same classification as the system to which they are attached, except those lines that contain an excess flow check valve (EFCV) are classified as Quality Group D beyond the EFCV. See Figure 3.2-2. **(5) All other instrument lines beyond the root valve, if used to actuate a safety system, shall be the same classification as the system to which they are attached. **(6) All other instrument lines beyond the root valve, if not used to actuate a safety system, shall be quality Group D. **(7) All sample lines from the outer isolation valve or the process root valve through the remainder of the sampling system shall be Quality Group D. I-10, II-7, II-8, II-13, II-4, III-1, IV-4, IV-8, V-7, V-8, VI-1, IX-3, IX-4, X-1, X-3, X-2, XI-1, XI-3. XI-2, XI-7, XII-2, XII-7, XII-1, XII-3, XVIII-4, XVIII-3, XIX-4, XIX-3, XX-5, XXXIII-4, XXXIII-1, XXXIII-3, XXXIII-2, 1b ANSI B31, Code Case 78 applies for B31.7 Class 1 and Class 2 pipe fittings 3/4" nominal pipe size (NPS) and smaller. I-10, II-7, II-8, II-13, II-4, III-1, IV-8, V-8, V-7, VI-1, IX-4, IX-3, X-2, X-3, X-1, XI-2, XI-3, XI-7, XI-1, XII-3, XII-1, XII-7, XII-2, XVIII-3, XVIII-4, XIX-3, XIX-4, XXVIII-3, XXXIII-3, 1c A module is an assembly of interconnected components which constitute an identifiable device or piece of equipment. For example, electrical modules include sensors, power supplies, and signal processors. Mechanical modules include turbines, strainers, and orifices. II-14, II-15, III-7, IV-10, V-9, VI-3, VII-1, VIII-1, IX-10, IX-11, X-9, XI-10, XII-10, XVIII-8, XIX-8, XXI-1, XXII-1, XXIII-7, XXIV-5, XXV-5, XXIX-6, 1d GE Specification 21A1100AS (Ref. 243) adds the following code requirements to the Reactor Vessel: The Winter 1967 Addenda to the ASME Code Section III is not to be included as a basis for purchase of this vessel, except as follows: 1) Charpy impact teats per N-331.2 of the Winter 1967 Addenda will be furnished; 2) Welds are to be ultrasonically examined using the angle beam method described by N-625 of Winter 1967 Addenda; 3) The changes to Article 4-Design by the Winter 1967 Addenda are included; 4) The addition of Appendix IX - Quality Control and Nondestructive Examination Methods in included. I-1, I-2, I-3, I-5, I-9, UFSAR/DAEC-1 TABLE 3.2-1 DAEC Classification of Components in System Footnotes T3.2-15 Revision 22 - 5/13 Footnote # Footnote Systems 1e For pump designs, the applicable class, section, or subsection of the referenced ASME B&PV Code is used as a guide in calculating the thickness of pressure-retaining portions of the pump and in sizing cover bolting. For example, use ASME Section III, Class C, 1968 Editions, for a design guide for Quality Group A & B pumps. For Quality Group D below 150 psig and/or 212 deg. F, manufacturer's standard pump service intended may be used. III-4, V-2, IX-5, X-4, XI-4, XII-4, XVIII-5, XIX-5, XXIII-4, XXIV-2, XXV-2 1f ANSI B16.5 or MSS-SP-66 apply for valves (Note MSS-SP-66-1964 was withdrawn from publication in favor of ANSI B16.34-1973) II-9, II-10, II-11, II-12, II-13, III-5, IV-1, IV-2, IV-3, IV-4, V-4, V-5, V-6, VI-2, IX-8, IX-9, IX-7, X-7, X-6, X-8, XI-8, XI-6, XI-5, XII-8, XII-6, XII-5, XVIII-6, XVIII-7, XIX-6, XIX-7, XX-6, XXIII-6, XXIII-5, 1g The RCIC and HPCI turbines do not fall within the applicable design codes. To assure that the turbines are fabricated to the standards commensurate with their safety and performance requirements, General Electr ic has established specific design requirements for these components. XI-9, XII-9, 1h Existing API/AWWA standards and supplementary requirements apply. Tanks are to be constructed to meet the intent of API Standards 620 or 650 or AWWA Standard D100 for those fuel, oil, or water storage tanks. XVIII-1, XXIII-1, 1i The condensate storage tank will be designed, fabricated and tested to meet the intent of API Standard 650. In addition, the specifications for this tank will require 100% surface examination of the side wall to bottom joint and 100% volumetric examination of the side wall weld joints. XXXIV-1, 1j ASME Section VIII, Division I, and USAS B31.1.0 apply downstream of the outermost isolation valves. XVIII-4, XVIII-5, XVIII-6, XVIII-7, XVIII-8, 1k The gaseous radwaste system piping, pumps and valves containing gaseous radwaste shall be constructed in accordance with the applicable codes of Quality Group D XXIII-3, XXIII-4, XXIII-5, XXIII-6, 1l Some of this piping was also constructed to B31.1.0 IV-4, IV-8, 1m Some lines, such as ECB-9 (drain to filter/demineralizer), are class 3, non-seismic. V-8, 1n DELETED 1o Lines DCB-1 and DCB-2 are nuclear class 3, according to Bechtel Specification M-190 XIX-4, XIX-7, 1p The RHRSW backwash line (GBD-62 and GBD-63) is non-seismic, according to Bechtel Specification M-190, Sheet 23A. XXV-1, XXV-4, 1q Portions of this system which supply suction for HPCI, RCIC, and Core Spray from the condensate storage tank are seismic category I. XXXIV-2, XXXIV-3, 1r The return line to the condensate storage tank classified as "Q" by Bechtel in the Q-list (Ref. 225) and was built that way by Bechtel. However, these lines are actually Quality Group D, with no QA requirement. That is the way these lines are classified in this table. XI-3, XII-3, 1s The Bechtel Q-list (Ref. 225, item 2.4365) notes this item as Q. However, the entry refers to Bechtel Specification M-119 (Reference 252). This document addresses Seismic Category I supports only. Therefore, only the supports in this item have a requirement for quality assurance and are Seismic Category I. XXVI-2, XXVII-2, 1t Bechtel Q-list (Ref. 225, item 2.1510) notes this item as Q. However, only pipe hangers and supports provide a specification (M-119, Ref. 252) as a reference. Therefore, only pipe hangers and supports for this item are seismic category I and have special Quality Assurance requirements. XIX-1, XIX-2, XIX-4, XIX-5, XIX-7, XIX-8, 1u See GE document NEDC-31853 (Ref. 2) "Duane Arnold Design Safety Standards", Appendix A, for stress and deformation limits for this item. I-5, I-6, I-9, 1v DELETED 1w This item includes the 24 volt D.C. Power System. VI-3, UFSAR/DAEC-1 TABLE 3.2-1 DAEC Classification of Components in System Footnotes T3.2-16 Revision 22 - 5/13 Footnote # Footnote Systems 1x The recirc pump motors are not seismic category I. However, per 21A9213 sections 4.1.8.3.5 & 4.1.6, they must be seismically supported so as not to breach the integrity of the reactor coolant pressure boundary. III-6, 1y De-icing lines located in the Intake Structure are non-seismic Category I. XXIV-1, XXV-1, XXXVII-1, UFSAR/DAEC-1 TABLE 3.2-1 DAEC Classification of Components in System Comments T3.2-17 Revision 22 - 5/13 Sys# / Comp# Comments IV 4 1 The design and construction specifications for the hydraulic control unit (HCU) do invoke such codes and standards as can be reasonably applied to individual parts in developing required quality levels, but these codes and standards are supplemented with additional requirements for these parts and for the remaining parts and details. For example: (1) all welds are liquid-penetrant inspected; (2) all socket welds are checked for minimum engagement and end gap between pipe and socket bottom by a marking technique; (3) all welding was performed by qualified welders; (4)all work was done per written procedures. The following examples are typical of the problems associated with codes designed to control field-assembled components when applied to the design and production of factory fabricated specialty components: **1. The HCU nitrogen gas bottle is a spun forging that is mechanically joined to the accumulator. It stores the energy required to scram a drive at low vessel pressure. It has been code stamped since its introduction in 1966, although its size exempts it from mandatory stamping. It is constructed of a material listed by the ASME B&PV Code, Section VIII, that was selected for its strength and formability. **2. The scram accumulator is joined to the HCU by a split flange joint chosen for its compact design to facilitate both assembly and maintenance. Both the design and construction conform to the B31.1.0 piping code. This joint, which requires a design pressure of 1750 psig, has been proof tested to 10,000 psi. **3. The accumulator nitrogen shutoff valve is a 6,000 psi cartridge valve whose copper alloy material is listed in the ASME B&PV Code, Section VIII. The valve was chosen for this service partly because it is qualified by the U.S. Navy for submarine service. **4. The directional control valves are solenoid pilot-operated valves that are subplate mounted on the HCU. The valve has a body specially designed for the HCU, but the operating parts are identical to a commercial valve with a proven history of satisfactory service. The pressure retaining parts are stainless steel alloys chosen for service, fabrication and magnetic properties. The manufacturer cannot substitute a code material for that used for the solenoid core tube. The foregoing examples are not meant to justify one pressure integrity quality level or another, but to demonstrate that the codes and standards invoked by those quality levels are not strictly applicable to special equipment and part designs. Group D classification is generally applicable because the codes and standards invoked by that cla ssification contain clauses that permit the use of manufacturer's standards and proven design techniques that are not explicitly defined within those codes. This was supplemented by the quality control techniques described above. IV 4 2 Bechtel built these items to ASME Class 2 standards, based on the piping classes given on the P&ID (see paragraph 3.2.7). However, based on the QA classification of these items, they should have been Class 3. The higher class is shown on this table, although Class 3 is justifiable and would make more sense with the QA Group D designation. IV 8 1 The design and construction specifications for the hydraulic control unit (HCU) do invoke such codes and standards as can be reasonably applied to individual parts in developing required quality levels, but these codes and standards are supplemented with additional requirements for these parts and for the remaining parts and details. For example: (1) all welds are liquid-penetrant inspected; (2) all socket welds are checked for minimum engagement and end gap between pipe and socket bottom by a marking technique; (3) all welding was performed by qualified welders; (4)all work was done per written procedures. The following examples are typical of the problems associated with codes designed to control field-assembled components when applied to the design and production of factory fabricated specialty components: **1. The HCU nitrogen gas bottle is a spun forging that is mechanically joined to the accumulator. It stores the energy required to scram a drive at low vessel pressure. It has been code stamped since its introduction in 1966, although its size exempts it from mandatory stamping. It is constructed of a material listed by the ASME B&PV Code, Section VIII, that was selected for its strength and formability. **2. The scram accumulator is joined to the HCU by a split flange joint chosen for its compact design to facilitate both assembly and maintenance. Both the design and construction conform to the B31.1.0 piping code. This joint, which requires a design pressure of 1750 psig, has been proof tested to 10,000 psi. **3. The accumulator nitrogen shutoff valve is a 6,000 psi cartridge valve whose copper alloy material is listed in the ASME B&PV Code, Section VIII. The valve was chosen for this service partly because it is qualified by the U.S. Navy for submarine service. **4. The directional control valves are solenoid pilot-operated valves that are subplate mounted on the HCU. The valve has a body specially designed for the HCU, but the operating parts are identical to a commercial valve with a proven history of satisfactory service. The pressure retaining parts are stainless steel alloys chosen for service, fabrication and magnetic properties. The manufacturer cannot substitute a code material for that used for the solenoid core tube. The foregoing examples are not meant to justify one pressure integrity quality level or another, but to demonstrate that the codes and standards invoked by those quality levels are not strictly applicable to special equipment and part designs. Group D classification is generally applicable because the codes and standards invoked by that cla ssification contain clauses that permit the use of manufacturer's standards and proven design techniques that are not explicitly defined within those codes. This was supplemented by the quality control techniques described above. IV 8 2 Bechtel built these items to ASME Class 2 standards, based on the piping classes given on the P&ID (see paragraph 3.2.7). However, based on the QA classification of these items, they should have been Class 3. The higher class is shown on this table, although Class 3 is justifiable and would make more sense with the QA Group D designation.
UFSAR/DAEC-1 TABLE 3.2-1 DAEC Classification of Components in System Comments T3.2-18 Revision 22 - 5/13 Sys# / Comp# Comments IV 9 1 The design and construction specifications for the hydraulic control unit (HCU) do invoke such codes and standards as can be reasonably applied to individual parts in developing required quality levels, but these codes and standards are supplemented with additional requirements for these parts and for the remaining parts and details. For example: (1) all welds are liquid-penetrant inspected; (2) all socket welds are checked for minimum engagement and end gap between pipe and socket bottom by a marking technique; (3) all welding was performed by qualified welders; (4)all work was done per written procedures. The following examples are typical of the problems associated with codes designed to control field-assembled components when applied to the design and production of factory fabricated specialty components: **1. The HCU nitrogen gas bottle is a spun forging that is mechanically joined to the accumulator. It stores the energy required to scram a drive at low vessel pressure. It has been code stamped since its introduction in 1966, although its size exempts it from mandatory stamping. It is constructed of a material listed by the ASME B&PV Code, Section VIII, that was selected for its strength and formability. **2. The scram accumulator is joined to the HCU by a split flange joint chosen for its compact design to facilitate both assembly and maintenance. Both the design and construction conform to the B31.1.0 piping code. This joint, which requires a design pressure of 1750 psig, has been proof tested to 10,000 psi. **3. The accumulator nitrogen shutoff valve is a 6,000 psi cartridge valve whose copper alloy material is listed in the ASME B&PV Code, Section VIII. The valve was chosen for this service partly because it is qualified by the U.S. Navy for submarine service. **4. The directional control valves are solenoid pilot-operated valves that are subplate mounted on the HCU. The valve has a body specially designed for the HCU, but the operating parts are identical to a commercial valve with a proven history of satisfactory service. The pressure retaining parts are stainless steel alloys chosen for service, fabrication and magnetic properties. The manufacturer cannot substitute a code material for that used for the solenoid core tube. The foregoing examples are not meant to justify one pressure integrity quality level or another, but to demonstrate that the codes and standards invoked by those quality levels are not strictly applicable to special equipment and part designs. Group D classification is generally applicable because the codes and standards invoked by that cla ssification contain clauses that permit the use of manufacturer's standards and proven design techniques that are not explicitly defined within those codes. This was supplemented by the quality control techniques described above. V 1 1 The standby liquid control storage tank is designed, fabricated, inspected, and tested to meet the intent of API Standard 650 and the ASME B&PV Code, Section VIII, Division 1. All butt welds are given spot radiographic examination. Liquid-penetrant inspection is conducted per the ASME Code, Section VIII, Division 1, on the following welds: **(1)All tank nozzle welds below and including the overflow nozzle are examined internally and externally to the tank. **(2)A ll fillet and socket welds receive a random examination. V 1 2 The construction of the accumulator is in accordance with the requirements of the ASME B&PV Code, Section VIII, Division 1. An ASME stamp is required. Other codes applied to the accumulator are as follows: **(1)ANSI B16.11 "Forged Steel Fittings, Socket Welded and Threaded". **(2)AND 10050 "Bosses, Standard Dimensions for Gasket Seal Straight Thread". VI 3 1 See DAEC letter to the USNRC, NG-91-2652, dated 8/27/91, for inclusion of the 24 V D. C. Power Supply with this item. XI 9 1 The HPCI turbine is categorized as machinery and thus does not fall within the classification groups as earlier identified. To ensure that the turbine was fabricated to the standards commensurate with its performance requirements, General Electric has established specific design requirements for this component, as follows: **(1)All welding was qualified in accordance with Section IX of the ASME B&PV Code. **(2)All pressure retaining castings and fabrications were hydrotested to 1.5 x design pressure. **(3)All high pressure castings were radiographed according to ASTM E-94 (20% coverage, minimum), ASTM E-142 (severity level 3), ASTM-71, ASTM 186, or ASTM-280. **(4)As-cast surfaces were magnetic particle or liquid-penetrant tested according to the ASME B&PV Code, Section III, 1968 Edition, paragraph N 323.3 or N323.4. **(5)Wheel and shaft forgings were ultrasonically tested according to ASTM A388. **(6)Butt welds were radiographed according to the ASME B&PV Code, Section III, 1968 Edition, paragraph N624, and magnetic particle or liquid penetrant tested according to ASME Section III, paragraph N626 or N627. **(7)Notification made on any major repairs and records maintained. **(8)Record system and traceability according to the ASME B&PV Code, Section III, 1968 Edition, IX-225. **(9)Control and identification according to the ASME B&PV Code, Sec tion III, 1968 Edition, IX-226. **(10)Procedures conform to the ASME B&PV Code, Secti on III, 1968 Edition, IX-300. **(11)Inspection personnel are qualified according to the AS ME B&PV Code, Section III, 1968 Edition, IX-400.
(APED-A61-052, Comment 10)
XII 9 1 The RCIC turbine is categorized as machinery and thus does not fall within the classification groups as earlier identified. To ensure that the turbine was fabricated to the standards commensurate with its performance requirements, General Electric has established specific design requirements for this component, as follows: **(1)All welding was qualified in accordance with Section IX of the ASME B&PV Code. **(2)All pressure retaining castings and fabrications were hydrotested to 1.5 x design pressure. **(3)All high pressure castings were radiographed according to ASTM E-94 (20% coverage, minimum), ASTM E-142 (severity level 3), ASTM-71, ASTM-186, or ASTM-280. **(4)As-cast surfaces were magnetic particle or liquid-penetrant tested according to the ASME B&PV Code, Section III, 1968 Edition, paragraph N 323.3 or N323.4. **(5)Wheel and shaft forgings were ultrasonically tested according to ASTM A388. **(6)Butt welds were radiographed according to the ASME B&PV Code, Section III, 1968 Edition, paragraph N624, and magnetic particle or liquid penetrant tested according to ASME Section III, paragraph N626 or N627. **(7)Notification made on any major repairs and records maintained. **(8)Record system and traceability according to the ASME B&PV Code, Section III, 1968 Edition, IX-225. **(9)Control and identification according to the ASME B&PV Code, Sec tion III, 1968 Edition, IX-226. **(10)Procedures conform to the ASME B&PV Code, Secti on III, 1968 Edition, IX-300. **(11)Inspection personnel are qualified according to the AS ME B&PV Code, Section III, 1968 Edition, IX-400.
(APED-A61-052, Comment 10)
UFSAR/DAEC-1 TABLE 3.2-1 DAEC Classification of Components in System Comments T3.2-19 Revision 22 - 5/13 Sys# / Comp# Comments XVIII 1 1 Unprocessed liquid radioactive waste piping and equipment pressure parts installed prior to January 1, 1983, were classified as Quality Group C. Unprocessed liquid radioactive waste piping and equipment pressure parts installed subsequent to January 1, 1983, may be included in Quality Group D with added quality control (D+QA) in accordance with the design guidance contained in Regulatory Guide 1.143, Revision 1, modified as follows: **(1)Paragraphs C.1.1.3, C.2.1.3, and C.3.1.3 - The commitment is limited to the seismic design methods used in the original construction of the DAEC and is not upgraded to Regulatory Guide 1.143, Revision 1, requirements. **(2)Paragraph C.4.3 - Systems will be fabricated in accordance with good operability, maintenance, and repairability practices. **(3)Paragraph C.6 - All of pa ragraph C.6 is replaced in its entirety by the following sentence. "All safety related systems or portions of systems shall be designed, fabricated and installed in accordance with Quality Level II requirements." XVIII 4 1 Unprocessed liquid radioactive waste piping and equipment pressure parts installed prior to January 1, 1983, were classified as Quality Group C. Unproce ssed liquid radioactive waste piping and equipment pressure parts installed subsequent to January 1, 1983, may be included in Quality Group D with added quality control (D+QA) in accordance with the design guidance contained in Regulatory Guide 1.143, Revision 1, modified as follows: **(1)Paragraphs C.1.1.3, C.2.1.3, and C.3.1.3 - The commitment is limited to the seismic design methods used in the original construction of the DAEC and is not upgraded to Regulatory Guide 1.143, Revision 1, requirements. **(2)Paragraph C.4.3 - Systems will be fabricated in accordance with good operability, maintenance, and repairability practices. **(3)Paragraph C.6 - All of pa ragraph C.6 is replaced in its entirety by the following sentence. "All safety related systems or portions of systems shall be designed, fabricated and installed in accordance with Quality Level II requirements." XIX 4 1 Segments of the RWCU piping have been replaced with IGSCC resistant materials. XXIII 3 1 The construction codes used for Offgas pipe, pumps, and valves were USAS B31.7; ASME Sections III and VIII; and the Draft Pump and Valve Code. To identify the correct code for a component it is necessary to research the receiving inspection files (File Q2.321). There are two reasons that several codes and dates were applied. The first reason is that the Offgas design was changed during Procurement/Construction. The second reason is that the codes were changing rapidly during the period of the design. The project correspondence which records when the Offgas System construction code was changed is given in APED-N62-076. XXIII 4 1 The construction codes used for Offgas pipe, pumps, and valves were USAS B31.7; ASME Sections III and VIII; and the Draft Pump and Valve Code. To identify the correct code for a component it is necessary to research the receiving inspection files (File Q2.321). There are two reasons that several codes and dates were applied. The first reason is that the Offgas design was changed during Procurement/Construction. The second reason is that the codes were changing rapidly during the period of the design. The project correspondence which records when the Offgas System construction code was changed is given in APED-N62-076. XXIII 5 1 The construction codes used for Offgas pipe, pumps, and valves were USAS B31.7; ASME Sections III and VIII; and the Draft Pump and Valve Code. To identify the correct code for a component it is necessary to research the receiving inspection files (File Q2.321). There are two reasons that several codes and dates were applied. The first reason is that the Offgas design was changed during Procurement/Construction. The second reason is that the codes were changing rapidly during the period of the design. The project correspondence which records when the Offgas System construction code was changed is given in APED-N62-076. XXIII 6 1 The construction codes used for Offgas pipe, pumps, and valves were USAS B31.7; ASME Sections III and VIII; and the Draft Pump and Valve Code. To identify the correct code for a component it is necessary to research the receiving inspection files (File Q2.321). There are two reasons that several codes and dates were applied. The first reason is that the Offgas design was changed during Procurement/Construction. The second reason is that the codes were changing rapidly during the period of the design. The project correspondence which records when the Offgas System construction code was changed is given in APED-N62-076. XXIV 1 1 Emergency service water system meets the pressure integrity requirements of Quality Group D, including the additional quality assurance requirements for seismic category I piping. All inspection records will be retained according to the Quality Assurance Program of Chapter 17. These records include data pertaining to the qualification procedures and examination results. XXXIII 1 1 For Main Steam and Turbine Bypass piping and valves, all inspection records were retained according to the Quality Assurance Program of Chapter 17. These records include data pertaining to the qualification of inspection personnel, examination procedures, and examination results. XXXIII 1 2 Turbine Stop, Control, and Bypass Valves: A certification was obtained from the vendors of these valves indicating that all cast pressure-retaining parts of a size and configuration for which volumetric examination methods are effective have been examined by radiographic methods by qualified personnel. Ultrasonic examination to equivalent standards are used as an alternative to radiographic methods. XXXIII 1 3 The main steam piping between the outermost containment isolation valves up to the turbine stop valves, the main turbine bypass piping up to the turbine bypass valves and all branch line connected to these portions of the main steam and turbine bypass piping up to the first valve capable of timely actuation are classified as Quality Group D and meet the additional quality assurance requirements for "critical" piping, as stated in Section 17.1.8.1, Schedule IV. XXXIII 1 4 The first valve capable of timely actuation in branch lines connected to the main steam lines between the outermost containment isolation valves and turbine stop valves and connected to the turbine bypass valves meets all of the pressure integrity requirements of Quality Group D, including the additional quality assurance requirements for "critical" piping, as stated in Section 17.1.8.1, Schedule IV.
UFSAR/DAEC-1 TABLE 3.2-1 DAEC Classification of Components in System Comments T3.2-20 Revision 22 - 5/13 Sys# / Comp# Comments XXXIII 1 5 All inspection records for the main steam and turbine bypass piping and the first valve in the branch lines c onnected to this piping were retained according to the Quality Assurance Program of Chapter 17. These records include data pertaining to the qualification of inspection personnel, examination procedures, and examination results. XXXIII 4 1 Materials used in feedwater control valves are as follows: **(1)Valve body is ASTM A105 Gr. II. **(2)Valve bonnet is ASTM A105 Gr. II and A234 Gr. WPB. XXXIII 4 2 Examination and testing requirements for the feedwater control valves are as follows: **(1)All pressure retaining castings are radiographed, after final heat treatment, in accordance with the ASME B&PV Code, Section III, Appendix IX, paragraph 330 and ASTM E142. Discontinuities are judged by ASTM E71, E186, and E280. **(2)All accessible surfaces of all pressure retaining castings are examined in finished condition, after final heat treatment, by either liquid penetrant methods per paragraph N323.4 or magnetic particle methods per paragraph N323.3, with acceptance criteria per paragraph N323.4 of the Summer 1969 addenda to ASME Section III. **(3)All pressure retaining forgings are examined in the as-furnished condition by the ultrasonic method per paragraph N322 of the Summer 1969 Addenda to ASME, Section III. XXXIII 5 1 Seismic Adequate Category (A) as added for MSIV-LTS was based on original construction of pipe as "critical" using the same records as seismic category I pipe. A seismic evaluation based on walkdowns and comparative analysis was performed in lieu of a formal seismic analysis. XXXIII 5 2 The QA Requirements for MSIV-LTS are driven by the SER for Amendment 207 to License No DPR 49 (Docket No. 50-331). XXXIV 1 1 The condensate storage tank was designed, fabricated, and tested to meet the intent of API Standard 650. In addition, the specifications for this tank require (1) 100% surface examination of the side wall to bottom joint and (2) 100% volumetric examination of the side wall weld joints. XXXIV 1 2 Page T3.2-5 of the UFSAR (Ref. 233) says that the CST is non-seismic. XXXV 1 1 The Auxiliary a-c Power System is composed of the 4160 VAC Switchgear, the 480 VAC Load Centers, the 480 VAC Motor Control Centers, and the Instrument AC Control Power System. XXXVII 1 1 The River Water Supply System meets the pressure integrity requirements of Quality Group D, including quality assurance requirements for seismic category I. Inspection records will be retained according to the Quality Assurance Program of UFSAR Chapter 17. These records include data pertaining to the qualification procedures and examination results. General 1 B31.1.0 and B31.7 were originally published as USA Standards (USAS), but are now designated as ANSI Standards. General 2 See additional material examination requirements of Section 17.1.8.1 for piping and valves. General 3 Code effective date is obtained by the Purchase Order date for the particular component (see Table 3.2-1) or by referring to Table 3.2-2.
UFSAR/DAEC-1 T3.2-21 Revision 22 - 5/13 This page intentionally left blank.
UFSAR/DAEC-1 TABLE 3.2-2 Sheet 1 of 2 CLASSIFICATION AND CODE COMPLIANCE REQUIREMENTS T3.2-22 Revision 22 - 5/13 {{{ For items which do not have a specific construction code listed In Table 3.2-1, the following codes, including their addenda and applicable code cases, in effect at the time of component purchase order date, have been applied. In case of a conflict between this table and Table 3.2-1, Table 3.2-1 shall govern. Safety Class (SC) or Quality Group Classification (c) Components (d) Components Ordered Before Jan. 1, 1970 Components Ordered on or after Jan. 1, 1970 and before July 1, 1971 Components Ordered on or After July 1, 1971 SC-1 Group A Vessels Piping Pumps (e) & valves (g) Heat Exchangers ASME Section III, '68 Ed., Classes A, C; USAS B31.1.0;(a)(l) USAS B31.1.0;(a)(l) TEMA Code (f) ASME Section III, '68 Ed., Class A; USAS B31.7, Class I;(b)(k)(l)(h) ASME NP&VC Class I;(m) TEMA Code (f) ASME Section III, '71 Ed., Class 1; NA&NB subsections;(h) TEMA Code (f) SC-2 Group B Vessels Piping Pumps (e) & valves (g) Heat Exchangers
Tanks ASME Section III, '68 Ed., Classes B, C;(i) USAS B31.1.0;(a)(l) USAS B31.1.0;(a)(l) TEMA Code;(f) (j) ASME Section III, '68 Ed., Classes B, C;(i) USAS B31.7, Class II;(b)(k)(l) NP&VC Class III;(m) TEMA Code (f) (j) ASME Section III, '71 Ed., Class MC (i) or 2; NA&NC subsections; TEMA Code (f) (j) SC-3 Group C Vessels Piping Pumps (e) & valves (g) Heat Exchangers
Tanks ASME Section VIII, '68 Ed., Div. 1; USAS B31.1.0;(a)(l) USAS B31.1.0;(a)(l) TEMA Code (f) (j) ASME Section VIII, '68 Ed., Div. 1; USAS B31.7, Class III;(b)(k)(l) NP&VC Class III;(m) TEMA Code (f) (j) ASME Section III, '71 Ed., Class 3; NA&ND subsections; TEMA Code (f) (j) Power Plant (piping systems) Group D Vessels Piping Pumps (e) & valves (g) Heat Exchangers
Tanks ASME Section VIII, '68 Ed., Div. 1; USAS B31.1.0;(a)(l) USAS B31.1.0;(a)(l) TEMA Code (f) (j) ASME Section VIII, '68 Ed., Div 1; USAS B31.1.0;(a)(l) USAS B31.1.0;(a)(l) TEMA Code (f) (j) ASME Section VIII, '71 Ed., Div 1; USAS B31.1.0;(a)(l) USAS B31.1.0;(a)(l) TEMA Code (f) (j) UFSAR/DAEC-1 TABLE 3.2-2 Sheet 2 of 2 CLASSIFICATION AND CODE COMPLIANCE REQUIREMENTS T3.2-23 Revision 22 - 5/13 a USAS B31.1.0-1967 plus applicable code cases. Requirements of ANSI Nuclear Code Cases N-2, N-7, N-9, and N-10 are applicable for Group A (RCPB) compone nts ordered before January 1, 1970. b USAS B31.7-1969 plus a pplicable code cases c For detailed piping/equipment cla ssification, refer to Table 3.2-1. d Components required to be stamped to Sec tion III of the ASME B&PV Code are stamped with the applicable ASME Code symbol, and the required third-party inspection was performed by an authorized Inspector. e For pump designs, the applicable class, s ection, or subsection of the referenced AS ME B&PV Code is used as a guide in calculating the thickness of pressure-retaining portions of the pump and in sizing cover bolting. For example, use ASME Section III, Class C for Group A&B pump design guide. For Group D below 150 psig and/or 212°F, Manufacturer's Standard pump for service intended may be used. f Tubular Exchanger Manufacturer's Association (TEMA) Code Requirements were applied using classes appropriate to each heat exchanger's duty cycle. g ANSI B16.5 or MSS-SP-66 apply for valv es (note MSS-SP-66 1964 was withdrawn fr om publication in favor of ANSI B16.34-1973). h Class I nuclear piping, pumps, and valves purchased after January 1, 1970, will meet the provision of ASME B&PV Code Section III, paragraph N-153 for stamping and third-party inspection. i Metal containment vessel and penetrations (extensions of containment) are ASME, Section III, stamped Class B or MC (subsection NA&NE), and the required third-party inspection shall be performed by an authorized inspector. j Existing API/AWWA standards and supplementary requirements ap ply. Tanks are to be designed, fabricated, constructed, and tested to meet the intent of API Standa rds 620 (Recommended Roles for Design and Cons truction of Large Welded Low Pressure Storage Tanks) or 650 (Welded Steel Tanks for Oil Storage) or AWWA Standard D100 (standard for steel tanks, stand pipes, reservoirs, and elevated tanks for water storage) for these fuel, oil, or water storage tanks. k ANSI B31, Code Case 78 applies for B31.7 Class I and Class II pipe and fittings 3/4 inch nominal pipe size and smaller. l These codes were originally published as USA Standard s (USAS), but are now designated as ANSI Standards. m ASME Code for Pumps and Valves for Nuclear Power (or Nuclear Pump and Valve Code). This was incorporated into ASME Section III after the construction of DAEC. UFSAR/DAEC-1 TABLE 3.2-3 SEISMIC CATEGORY I STRUCTURES T3.2-24 Revision 22 - 5/13 Reactor Building
Drywell (including reactor vessel pedestal)
Wetwell (torus)
Control Building
Intake Structure
Turbine Building (portion containing emergency diesel generators)
Pump House (portion containing residual heat removal and emergency service water systems)
Offgas Stack
Notes: (a) Structures, systems, and components not listed as Seismic Category I in Table 3.2-1 or above are nonseismic. (b) 10 CFR 50 Appendix B QA program is applied to Seismic Category I structures.
UFSAR/DAEC-1 TABLE 3.2-4 DESIGN REQUIREMENTS FOR SAFETY CLASSES 2 AND 3 ELECTRIC SYSTEMS AND COMPONENTS T3.2-25 Revision 22 - 5/13 PROTECTION CLASS 1E Components Modules Sensors Systems (a)CableConnectors Switch Gear Transformers DieselSystems Motors Valve ActuatorsPenetrationsIEEE-323 IEEE-323 IEEE-344 IEEE-323 IEEE-344 IEEE-279 (b) (b) IEEE-344IEEE-344 (b) IEEE-308IEEE-323 IEEE-334(c) IEEE-344 IEEE-323IEEE-344 Notes: (a) IEEE-279 shall apply only to those Safety Class 2 or 3 systems and components which actuate reactor trip or, in the event of a serious reactor accident, actuate engineered safeguards. (b) Design requirements had not been developed for this Condition of Design by the applicable code at the time DAEC was designed. Design requirements are to be developed for the specific component. (c) GE Scope of Supply
UFSAR/DAEC-1 TABLE 3.2-5
SUMMARY
OF SAFETY CLASS DESIGN REQUIREMENTS T3.2-26 Revision 22 - 5/13 Safety Class Design Requirements 1 2 3 Other Quality Group Classification (a) A B C or D + QA D + QA C or D Quality Assurance Requirement(b) B B B B B or D Seismic Category (c) I I I NA I or NA
Notes:
(a) The equipment shall be constructed in accordance with the in dicated code listed in Table 3.2-1 or 3.2-2. (b) B - The equipment shall be constructed in accordance with the quality assurance requirements of 10CFR50 Appendix B and the Quality Assurance Program described in Chapter 17. D - The equipment shall be constructed in accordance with the Quality Assurance requirements consistent with good practice for steam power plants. (c) I - The equipment is this seismic category shall be constructed in accordance with the seismic requirements of the safe shutdown earthquake as described in Subsection 3.2.1 and Section 3.7. NA - The seismic requirements for the safe shutdown earthquake are not applicable to the equipment of this classification.
.IIN,Tr2UMENTLINE'IFIC,ATION......MAINLlNEI@MAINLINe--iUBHU,r&.FLAQE*/1'UeeIflnlI0e.F.c.v.re?TCONN.rU!INt,'!,.,OGk:frWELDORFLARELE55TUBEFITTING.SDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTProcessLineCodeFigure3.2-2Revision9-6/91 UFSAR/DAEC - 1 3.3-1 Revision 13 - 5/97 3.3 WIND AND TORNADO LOADINGS
3.3.1 WIND LOADINGS
All structures are designed for wind loads in accordance with ASCE Paper 3269 "Wind Forces on Structures." l Wind pressure as it applies to the structures is as follows (See also Section 2.3):
Wall Load Height (ft) Basic Velocity (mph) Dynamic Pressure (Including 1.1 Gust Factor) q (psf) Pressure 0.9 q (psf) Suction 0.4 q (psf) Roof Load Suction 0.6 q (psf) 0-50 105 34 31 14 20 50-150 125 48 43 19 29 150-400 145 65 59 26 39
Whenever wind loads are combined with other loads, a 33% increase in allowable stresses is permitted.
3.3.2 TORNADO LOADINGS
3.3.2.1 Applicable Design Parameters
The design-basis tornado consists of a tornado with a maximum tangential velocity of 300 mph as shown in Figure 3.3-1, traveling with a maximum transverse velocity of 60 mph. The loadings created by the design-basis tornado are reflected in the
following two tornado design criteria used in the design of tornado-resistant structures:
- 1. The velocity components are applied as a uniform 300 mph wind on the structure.
- 2. The pressure differential is applied as a 3 psi positive (bursting) pressure occurring in 3 sec.
Although the design-basis tornado consists of a tornado with a maximum transverse velocity of 60 mph, the tornado phenomenon is so complex that considerable engineering judgment is necessary in the determination of a wind-loading criterion. Therefore, the design-basis tornado velocity components are conservatively applied as a 300 mph wind on the structure using the applicable portions of the wind design methods
described in ASCE 3269 1 particularly for shape factors. Variation of wind velocity with height is not used. The average wind-loading approach is chosen since it is almost impossible for a structural designer to apply a vortex-shaped loading on a building. UFSAR/DAEC - 1 3.3-2 Revision 13 - 5/97 Shape factors from ASCE Paper 3269 are used since they are the best currently available information, even though velocities as high as 300 mph are not considered in the paper. The radial and vertical components are neglected since they act parallel to the walls of the structure. In order to determine the conservatism of this approach, the tornado criteria
have been checked for various critical structures against the design-basis tornado (an extrapolation to the design velocities of Hoecker's 2 velocity profiles derived from movies of the Dallas tornadoes of April 2, 1957). The average wind loading on the DAEC reactor building due to the design-basis tornado with a transverse velocity of 60 mph is approximately that due to a 215-mph wind, which is 85 mph below the design wind loading. Figure 3.3-2 shows a comparison of the tornado wind loading criteria to the design-basis tornado superimposed on a reactor building wall.
The tornado model used on past jobs specifies a maximum tangential velocity of 300 mph and a maximum translational velocity of 60 mph as shown in Figure 3.3-1. The tangential velocities vary with height and distance from the tornado center. When this tangential velocity distribution is superimposed on the reactor building, and the wind
loads integrated over the surface of the structur e, the resultant average wind load is about 155 mph. The 215-mph wind loading mentioned accounts for the 60-mph additional
force due to the translational velocity. The above forces represent the actual loads on the structure if the tornado model actually struck the building. The actual loading condition is
shown in Figure 3.3-2 by the solid lines.
The above discussion is provided only as a justification of the 300-mph wind load criterion. The tornado model with a 300-mph maximum velocity component results in an average wind load of 215 mph. This 215 mph is considered to be a realistic value of the effective wind force on the structure. The increase in the criterion to 300 mph accounts for the possible excess load effects due to model assumptions, assumptions in the structural analysis, simplifications in calculations, and effects of construction sequence and methods. Since wind pressure is a function of velocity squared (V 2), the increase from 215 to 300 mph results in a pressure increase from 120 to 230 psf. Thus, the margin of safety implied by the increase in loading criterion is about 2.0. In addition, an increase in design loads is provided by simultaneously applying a differential (bursting) pressure due to the 3-psi pressure drop. This pressure adds to the suction force
of the wind to provide one of the critical loading conditions. The wind and depressurization forces are directly combined even though they do not occur simultaneously. This simplification adds to the margin of safety of the structure. UFSAR/DAEC - 1 3.3-3 Revision 13 - 5/97 Where failure could affect the operation and functions of the primary containment and reactor primary system, and for structures housing equipment necessary for the safe
shutdown of the reactor, the following tornado effects are considered in the design of
these structures (see also Sections 2.3 and 3.5.1):
- 1. External wind forces resulting from a tornado funnel having a horizontal peripheral tangential velocity of 300 mph and a transverse velocity of 60 mph.
The wind force on the structure was considered as a static load and calculated on the basis of a 300-mph horizontal wind applied over the full height of the
projected area.
- 2. Differential pressure between inside and outside of fully enclosed areas of 3 psi (bursting). Means are included in the actual design of the structures to limit
excessive pressure differentials.
- 3. Missile equivalent to a 4 in. x 12 i
- n. x 12 ft long wood plank (108 lb) traveling end-on at 300 mph, or a passenger auto (4000 lb) flying through the air at 50 mph and at not more than 25 ft aboveground with a contact area of 20 ft
- 2.
- 4. A torsional moment resulting from applying the wind specified in item 1 above on one-half of the structure and a wind velocity equal to one-half that specified in item 1 above applied to the other half of the building in the opposite direction.
Note: The effects of items 1, 2, and 3 were considered as acting simultaneously.
All structures housing equipment necessary for a safe shutdown are designed to withstand a tornado-induced depressurization rate of 1 psi/sec for 3 sec. To accomplish this design objective, all nonvented compartments are checked to verify that they are
capable of withstanding an internal (bursting) pressure of 3 psi.
The margin of safety in the tornado wind criteria is in the assignment of the large wind force and the simultaneous loading application. A concrete building is designed using the normal ACI code provisions and methods for ultimate strength design. This specifically includes the appropriate capacity reduction factor (0). The only modification to the ACI provisions is in the assignment of load factors. The load factors for the design equation are always assigned as 1.0.
The implied margin of safety is verified by the lack of structural damage to
buildings that have withstood tornadoes. Steel structures are designed using traditional elastic methods of analyses and allowable stresses of 1.5 f s with 0.9 F y as the upper limit. This is consistent with the design philosophy of structures under the DBE. UFSAR/DAEC - 1 3.3-4 Revision 13 - 5/97 3.3.2.2 Determination of Forces on Structures
Once the loads have been established, the design requires a determination of the
pressure coefficients. The forces on a square-sided building are 185 psf pressure and 115 psf suction. The outward pressure due to depressurization ranges from 50 psf to over 400 psf depending on compartment geometry. Thus, the total suction pressure on a wall ranges from 165 psf to over 600 psf. Normal wind loads specify a suction pressure of about 20 psf. Using the load factors of the ACI for normal wind loads and those
proposed for the tornado wind loads, the resultant loads are 25 psf (1.25 x 20 psf) and 165 to 600 psf, respectively. The damage of concrete structures by tornadoes is almost nonexistent. Buildings and structures designed for normal wind loads have been in the
direct path of tornadoes and sustained no structural damage. Thus by increasing design loads by at least 6.6 times (165/25 = 6.6), the structural integrity of the building is
ensured.
For those compartments that are vented, a flow analysis of all air volumes and interconnecting vent areas was performed, and the maximum transient pressure
differential across every wall, floor, and roof was calculated using the principles of fluid mechanics to determine its maximum transient pressure differential. Finally, each structural component was checked to ensure that it could withstand the maximum calculated transient pressure differential that it would experience.
3.3.2.3 Effect of Failure of Structures or Components Not Designed for Tornado Loads
The structural steel frames of the reactor building upper superstructure are designed to withstand the pressure corresponding to a 300-mph wind.
The reactor building siding and roof decking, however, have been designed for the normal wind loading. When this design velocity is appreciably exceeded, the siding and decking may blow off and expose the refue ling floor and parts of the reactor building not required for safe shutdown.
If tornado winds traverse the site, the reactor is capable of being shut down and secured in a safe shutdown mode. Superstructure damage could be incurred to the reactor building, turbine building, and incoming power lines without affecting the ability to shut down the reactor, to maintain the integrity of the primary containment, and to provide adequate core cooling. Simultaneous damage to all of these structures is not expected. Components that directly affect the ultimate safe shutdown of the plant are located either
under the protection of reinforced concrete or underground.
UFSAR/DAEC - 1 3.3-5 Revision 13 - 5/97 REFERENCES FOR SECTION 3.3
- 1. American Society of Civil Engineers, "Wind Forces on Structures Paper No. 3269, Final Report of Task Committee on Loads and Stresses, Structural
Division," Transactions ASCE , 1961.
- 2. W. H. Hoecker, Jr., "Wind Speed and Air Flow Patterns in the Dallas Tornado of April 2, 1957, "Monthly Weather Review , 1960.
500400300200100o100:r::r:I:r:a..a..a..a..00000101010r0I500I(\Ja..Ia..::r:a..Ia..a..1000r---010(\J00(\Jr---(\J0(\J400UJUJu...300t?:r:gCDUJ:r:200DISTANCEFROMTORNADOCENTER(FEET)DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTTangentialVelocityDistributionFigure3.3-1 A-0/'DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTComparison-Design-BasisTornadoandTornadoWindLoadingonReactorBuildingFigure3.3-2
I!')C.TERIORT"'(I=',DOORSTOPL.OGSill-,-----r-l.---/f./illE)'TI!RIOREL.e.VAT/Ot-J..OOORDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTTypicalStoplogArrangementsFigure3.4-1Revision1-6/83 UFSAR/DAEC-1 3.5-1 Revision 23 - 5/15 3.5 MISSILE PROTECTION
3.5.1 MISSILE SELECTION AND DESCRIPTION
3.5.1.1 Internally Generated Missiles (Outside Containment)
Wheel average tangential stresses in the HPCI and RCIC turbines are sufficiently
low that wheel failure is not predicted even at the theoretical runaway condition of 200% rated speed. Therefore, failure of these turbine wheels is considered so improbable as to be of no consequence with respect to becoming a potential missile or affecting safe
shutdown of the plant. In addition, the HPCI and RCIC are located in separate concrete rooms within the reactor building.
3.5.1.2 Internally Generated Missiles (Inside Containment)
3.5.1.2.1 Recirculation Pump Overspeed Missile
3.5.1.2.1.1 Description
The subject of the effects of destructive overspeed of the recirculation pump
following a design-basis LOCA, which is an ACRS concern, l is discussed below.
The DAEC drywell piping layout has been examined to determine the effect of missiles formed in the recirculation system following a postulated pipe rupture. The source of these postulated missiles is the breakup of either the recirculation pump motor or the recirculation pump impeller. The effects of the overspeed of the recirculation pump motor could be avoided by means of a decoupling device between the recirculation pump and the pump motor. From the results of an extensive analysis, the probability that any serious damage would result from the pump motor or impeller missiles is considered to
be negligibly low.
3.5.1.2.1.2 Effects of Recirculation Pump Overspeed
General Electric has submitted a licensing topical report 2 to the NRC, which contains a discussion of the theoretical potential for damage resulting from missiles originating in a recirculation system pump due to destructive overspeed caused by the pump acting as a turbine following a postulated pipe rupture in the recirculation system. Although the report indicates that destructive overspeed of the pump and motor could occur and missiles resulting from the overspeed could be generated, the probability of these missiles causing-damage is extremely low. The report also indicates that the effects of overspeed of the motor could be avoided by incorporating a decoupling device between the pump and motor, but this has not been done. In order to evaluate the effects of missiles produced by pump overspeed, it is necessary to determine if such missiles could escape the piping system and, if so, to determine whether the missiles would strike safety-related components with enough energy to penetrate or otherwise cause enough damage UFSAR/DAEC-1 3.5-2 Revision 23 - 5/15 to impair operation. Such an evaluation was made including a study of the probability of serious damage from the pump missiles. The DAEC plant layout has been compared with for which the probability of such an event was calculated.
and therefore, the results of the study are considered applicable to the DAEC. In addition, it must be noted that the results given in this evaluation are actually conservative if applied to the DAEC for the following reasons:
- 1. The DAEC has an improved pipe restraint system as compared to the evaluation model
- 2.
- 2. The DAEC recirculation pipe diameter of 22 in. is smaller than the evaluation model of 28 in., resulting in smaller missiles and less energy.
Consequently, the results given in Reference 2 represent the bounded values for
the DAEC.
3.5.1.2.1.3 Analysis
The probabilistic analysis of the release of radioactive material to the environment due to the formation of recirculation pump missiles involves a hypothetical
series of events that include the following:
- 1. The instantaneous double-ended break of the recirculation line at a
specific location followed by,
- 2. The formation and escape of a missile of very high energy which is,
- 3. Directed at the containment or a pipe at a very specific angle and,
- 4. Then impacts at its sharpest edge orientation, thus possibly endangering containment integrity.
This highly unlikely series of events has such a low probability of occurrence that it is comparable to other low- probability events that are considered negligible.
Given the design-basis LOCA, consider the random nature of the subsequent events of a missile-induced pipe break. First, the recirculation-line break must occur at
one of several locations out of over 40 locations at which it is equally probable that a break could occur. Next, the pump impeller must break up in a certain way, and a large missile must enter the piping. The missile must then leave the pipe in a certain direction out of a virtually infinite number of possible directions within a cone of dispersion. The UFSAR/DAEC-1 3.5-3 Revision 23 - 5/15 missile must not lose energy due to impact with nonvital piping or other objects, or the containment, and must finally strike a pipe at a certain angle and specific orientation. The total probability of damage to piping due to recirculation pump missiles is presented in Table 3.5-1. The unlikely and random occurrence of the missile-induced pipe break is of the same probability as the random failure of a second active component.
The probability of this series of events occurring in conjunction with a LOCA and releasing radioactive material to the environment is summarized in Table 3.5-2. For comparison, the probability of radioactive material release from all other causes is also
included.
It can be seen from Table 3.5-2 that the probability of radioactive material release to the atmosphere is only slightly increased when recirculation pump missiles are considered and that the overall probability of release is still very low.
The probability of the dual-break event (recirculation pipe break and containment break or other pipe break) is comparable to the probability of other events normally considered negligible. For example, the probability of the LOCA plus a single failure of the LPCI injection valve plus the break of a core spray line by a randomly directed recirculation pump missile is calculated to the 1.1 x 10-10 per year. The corresponding probability of a LOCA plus a single failure of the LPCI injection valve plus the random failure of another active component in the core spray system is 2.3 x 10-10 per year. Thus, the dual-break probability is smaller than the probability of other random double active component failures.
In summary, the breaking of a vital pipe by a hypothetical recirculation pump missile randomly directed is of the same nature as the random failure of two active components. The low probability of this failure in conjunction with the "double-failure" nature of the occurrence makes it outside the scope of the design basis of the DAEC.
3.5.1.3 Turbine Missiles
are designed to withstand the loading due to missiles generated by the failure of the turbine-generator. Further details on missile generation probability, missile characteristics, penetration ability of missiles, and protection against missiles are provided below.
The GE turbine for the DAEC is a three-casing, tandem-compound, four flow exhaust, 1800-rpm unit with 38-in. last-stage blades. Connected to the turbine shaft is the
ac generator, and coupled to the generator is the exciter.
The original low-pressure rotors have been replaced by the "monoblock" design. This
design has wheels which are integral to the shaft, while the original design had separate wheels which were shrunk onto the shaft. The missile analysis below, which considers a last
stage wheel failure, was applicable to the original design. The probability of 3.5-4 Revision 23 - 5/15 wheel missile generation is significantly lower with the monoblock design, and any postulated wheel missile would have less potential for causing damage since the monoblock design precludes failures which would result in large wheel fragments.
Therefore the analysis below represents a highly conservative upper bound for the monoblock rotor. The bucket vane analysis below is applicable to the monoblock rotor, since it uses the same bucket design as the original rotor. The bucket vane analysis is also applicable to the long shank bucket design installed on the 5 th stage turbine wheels of both low pressure rotors. The long shank design has the same type of connection as the original bucket design.
The probability of the occurrence of turbine missiles has been evaluated by GE to be very low. Proven control system reliability and equipment redundancy make a turbine runaway highly improbable. Even assuming a turbine runaway, modern test procedures of disk forgings virtually eliminate the possibility of wheel failure, except at above 169%
of rated speed.
Based on the GE report 4, the following 40-year interval probabilities of a wheel failure have been calculated:
Low speed failure (below 127% speed), 3.5 x 10 -7 Runaway failure (at higher speed), 2.1 x 10 -7 Total lifetime failure probability, 5.6 x 10 -7 The corresponding average annual probability of a wheel failure is found to be 1.4
x 10-8. However, recognizing the potentially serious consequences of turbine missiles, it was postulated in this study that turbine missiles are generated and their effect on the
plant safety was evaluated.
In the case of the GE low-pressure turbine, two missiles were postulated. The first, and potentially the most dangerous because of its size, was the last-stage wheel
- 5. Typically, it is also the most highly stressed and hence the most probable candidate for failure. The second missile considered was the bucket vane. There is a good possibility that it would lose most of its energy and would be deformed after penetrating the inner
casing and the hood.
General Electric studies indicate that in the case of a last-stage wheel, 120-degree fragments are potentially more damaging than either 90- or 180-degree fragments
- 5. The properties of a 120-degree wheel fragment are shown in Table 3.5-3 and the properties of a bucket vane are shown in Table 3.5-4.
Local penetration effects were determined by the modified Petry Formula
- 6. According to the formula, a missile will penetrate a reinforced-concrete slab of infinite thickness the following amount:
2013-019 3.5-5 Revision 23 - 5/15 D = K A P V' (3.5-1)
where
K = penetration coefficient of concrete experimentally determined (see Table 3.5-5)
A P = sectional pressure of the missile obtained by dividing the missile weight by the frontal area (lb/ft)
V = striking velocity of the missile (ft/sec)
V 1 = Log 10 1 + V 2 215,000
Amirikian reports Navy experiments that resulted in a formula to calculate
penetration into concrete slabs of finite thickness:
D' = D [1 + e-4(T/D - 2)] (3.5-2)
where
D = penetration depth in an infinite slab, ft D' = penetration depth in a finite thickness slab, ft
T = slab thickness, ft
The rearrangement of this equation shows that D = T/2 gives complete penetration. Therefore, from Equation 3.5-1, the thickest slab that will be perforated by a missile is:
T = 2 K A P V' (3.5-3)
Both low trajectory missiles (LTM), where the direction of impact is predominately horizontal, and high trajectory missiles (HTM), where the direction of impact is vertical, were considered. Air dr ag, which has the effect of significant energy loss for the missile, was considered for the HTM.
The results of the penetration ability of the postulated turbine missiles are summarized in Table 3.5-6.
The penetration depths shown in Table 3.5-6 represent the minimum concrete thickness required to stop the respective missiles.
All equipment required for a safe shutdown of the plant was reviewed regarding its vulnerability to the postulated turbine missiles. 3.5-6 Revision 23 - 5/15 Table 3.5-6 lists the minimum concrete wall thickness to prevent penetration by the missiles under consideration. All LTM generated will be contained by the heavy beams of the turbine foundation. A summary of protection provided for safety-related equipment located in the various buildings is given in the
following sections.
3.5.1.3.1 Reactor Building
The exterior wall of the reactor building directly in the path of a turbine missile varies in thickness which will provide adequate protection. See also Section 10.2.
3.5.1.3.2 Spent-Fuel Storage Pool
Damage of the pool by low trajectory turbine-generated missiles is not possible due to the structural and shielding requirements that result in
As noted in Section 3.5 1.3, the total lifetim e probability of a wheel failure is 5.6 x 10 -7. In view of the low probability, no further analysis is required.
3.5.1.3.3 Control Building
The control building is located outside the path of a turbine-generated missile. provide adequate protection.
Based on the conservative assumption that a secondary missile, spalling of concrete, will be generated from a postulated turbine missile intersecting the reactor
building east-wall, the 40-year interval probability of occurrence is 5.6 x 10 -7 as evaluated above.
Recognizing the potential serious consequence of secondary missiles, all systems and equipment required for safe shutdown of the plant were reviewed regarding their vulnerability to a postulated secondary missile.
An inventory of all systems required for safe shutdown of the plant shows that there are none that are vulnerable to a secondary missile. 3.5-7 Revision 23 - 5/15 3.5.1.3.4 Diesel-Generator Rooms
Accordingly, no turbine missile will disable the diesel-generator.
3.5.1.3.5 Area of the Drywell Shield Plugs
The only missiles that can strike in the area of the shield plugs are the 4 in. x 12 in. x 12 ft plank at 300 mph and the missile generated by a turbine-generator failure (see Section 3.5.1.2). The plank missile is postulated to travel in a horizontal direction and
thus cannot strike the horizontal exposed surface of the plug. An analysis for the unlikely event of a turbine-generator failure has been made. Using the modified Petry Formula, it
was found that thus penetration is not possible.
The lower surface of the bottom plug is sheated with 0.25-in. stainless steel plate so that any localized concrete spalling will not result in concrete falling into the drywell.
3.5.1.4 Missiles Generated by Natural Phenomena
A missile equivalent to a 4 in. x 12 i
- n. x 12 ft long wood plank (108 lb) traveling end-on at 300 mph and a passenger auto (4000 lb) flying through the air at 50 mph and at not more than 25 ft aboveground with a contact area of 20 ft 2 generated by the design basis tornado were considered in the design of the DAEC (see Section 3.3.2.1).
3.5.1.5 Missiles Generated by Events Near the Site
The closest commercial rail line to the DAEC which could carry hazardous material on a routine basis is approximately 3.5 miles west of the site. No hazardous material is manufactured in the vicinity of the site.
There are no known mineral mines or petroleum wells located within 5 miles of the plant site. There is only one operational quarry within 5 miles of the site. It is located approximately 3 miles southwest of the DAEC site. The potential hazard from the quarry
is judged to be insignificant (see Section 2.2.3).
A survey of offsite facilities at the time of initial FSAR submission (1972) concluded that there were no offsite facilities that might have a significant effect on the
safe operation of the plant.
3.5-8 Revision 23 - 5/15 The hazard analysis contained in Reference 2 determined the gas pipelines around the DAEC site present negligible risk to the safe operation of DAEC. 3.5.1.6 Aircraft Hazards Aircraft hazards from airports, military aviation and federal airways in the
vicinity of the DAEC are discussed in Section 2.2.2.5, 2.2.2.5.1 and 2.2.2.5.2. At the time of the initial FSAR, the private airport for light planes were approximately 2 miles west in Shellsburg, Iowa. Such a facility was designated on the most recent edition of the Dubuque Sectional Aeronautical Chart and also on the 1968 USGS topographical map, Shellsburg, Quadrangl
- e. Investigation and discussion with Cedar Rapids area aviation organizations indicat ed that a private restricted airstrip was operated at one time at the above mentioned location by an individual who owned a single light airplane, but the owner had dismantled the hangar and later died. The air strip was no
longer in use, but was still designated on the aeronautical charts because apparently no cancellation had been initiated. During the course of the investigation, it was determined that another small landing strip not shown on the most recent aeronautical charts exists at a location approximately 4 miles southeast of the plant. Discussion with the airfield manager indicated that approximately 10 single engine light airplanes were stationed at the strip. The maximum gross weight of these aircraft was estimated to be 3000 lb. Runway orientation is north-south (300U ft turf) and east-west (2600 ft turf) according to
the "Iowa Airport Directory 1972-73," published by the Iowa Aeronautical Commission. The facility was an uncontrolled field, and accordingly there was no record kept of the number of takeoffs and landings. An estimate of the number of takeoffs and landings was made based upon the assumption that between May 1 and November 1 each of the 10 planes averaged four movements per weekend. During the winter months it was estimated
that four of the planes are active on six weekends, again at an average of four movements per weekend. These assumptions were felt to be conservative and were based upon discussions held with the owner. Accordingly, it was conservatively estimated that there were less than 1500 movements per year from this airfield. The airfield owner indicated that at the time of contact there were no plans for significantly expanding the scope of operations at this facility. No other airfields were known at that time to exist within 5 miles of the DAEC. At the time of the initial FSAR the potential for aircraft accidents at the facility was considered extremely improbable due to the relatively small number of airplane movements at the small landing strip described in Section 2.2.2.5.
3.5-9 Revision 23 - 5/15 However, under no conditions would the ability to safely shut down the reactor be compromised. Thus, the effects of an aircraft accident would be possible loss of generating capacity and However, the capability to safely shutdown the reactor would be maintained.
Aircraft hazards from airport , military aviation and federal airways in the vicinity
of the DAEC are discussed in Section 2.2.2.5, 2.2.2.5.1 and 2.2.2.5.2. The conditional probability of core damage from an aircraft crash into the site is evaluated and documented in Appendix B of Reference 7. This analysis was prepared in response to Generic Letter 88-20 and submitted to the NRC as part of DAEC's Individual Plant Examination External Events. This analysis found the core damage frequency to be
below the NUREG-1407 screening criterion of 1E-6/yr. Therefore, the analysis concludes that the design of the DAEC plant is appropriate of its sitting (i.e., with respect to aviation hazards) such that the contribution to overall plant risks from aircraft crashes
in and around the plant is not significant.
3.5.1.7 Propane Storage Tank
An auxiliary boiler propone gas pilot supply tank
. Evaluation of the tank shows that the tank does not present an immediate risk to plant safety, but an accident sequence (vehicle impact) could conservatively be postulated with a frequency
greater than the NUREG-1407 screening criterion of 1E-6/yr. thereby reducing the overall risk contribution to less than the NUREG-1407 screening criterion of 1E-6/yr.
3.5.2 STRUCTURES, SYSTEMS, AND COMPONENTS TO BE PROTECTED FROM
EXTERNALLY GENERATED MISSILES
Reactor building walls and slabs and control room walls and slabs are designed to withstand the loading due to missiles generated by the failure of the turbine-generator (see Section 3.5.1.3). The thickness of the structural components is adequate to prevent complete penetration by a missile.
Structures housing equipment necessary for the safe shutdown of the reactor are designed to withstand the design-basis tornado-generated missiles (see Section 3.3.2.1).
3.5.3 BARRIER DESIGN PROCEDURES
Within containment the following measures have been taken to ensure that the damage caused by any single missile that might be generated from a component failure will not remove more than one redundant subsection of a vital safety system from
service. 3.5-10 Revision 23 - 5/15 The separation of redundant safety-related components has been used as a basic criterion throughout the design of this plant to optimize the independence of these components. The design intent was to separate redundant counterparts by 180º in azimuth and, where this was not possible, to maintain the maximum possible separation. An elaborate truss system has been used inside the drywell to restrain large pipes; this truss system has also furnished a considerable amount of structural material which acts as a protective barrier and prevents a missile from traveling a very great distance. Section 6.2.1 discusses in more detail the means by which primary containment integrity is
protected.
Missile protection for the EDGs is accomplished by separation, as described in Section 8.3.1.3. 2012-012 3.5-11 Revision 23 - 5/15 REFERENCES FOR SECTION 3.5
- 1. ACRS letter on Duane Arnold Energy Center Operating License from H. G.
Mangelsdorf to Dixy Lee Ray, dated March 13, 1973.
- 2. General Electric Company, Analysis of Reciruclation Pump Overspeed n a Typical General Electric Boiling Water Reactor, NEDO-10677, October 1972
- 3. "Probabilistic Analysis of the Effects of Missiles Formed in the Recirculation System Following Postulated Pipe Failure," filed with James A. Fitzpatrick Nuclear Power Plant, Docket No. 50-333, from Asa George, of PASNY, to J. F.
Stolz, AEC, dated April 10, 1973.
- 4. General Electric Company, Probability of Turbine-Generator Rotor Failure Leading to Ejection of External Missiles, memo report, February 1971.
- 5. E. E. Swicky, An Analyses of Turbine Missiles Resulting from Last-Stage Wheel Failure, TR675L211, General Electric, Schenectady, New York, 1967.
- 6. A. Amirikian, Design of Protection Structures, Nav. Docks P-51, Bureau of Yards and Docks, Department of the Navy, Washington, DC, 1959.
- 7. IES Utilities Inc., Individual Plant Examination of External Events, Transportation and Nearby Facility Hazards Evaluation of the DAEC Site, December 1995.
UFSAR/DAEC-1 T3.5-1 Revision 13 - 5/97 Table 3.5-1 TOTAL PROBABILITY OF DAMAGE TO PIPING DUE TO RECIRCULATION PUMP MISSILES (with LOCA probability of 1 x 10 -5 per year)
Occurrence Total Probability (per year) Penetration of main steam line 0.63 x 10 -8 Disablement of main steam isolation valve 1.0 x 10 -8 Penetration of feedwater line 2.0 x 10 -7 Penetration of core spray line 6.5 x 10 -8 Disablement of high-pressure coolant injection 1.5 x 10 -8 Disablement of reactor core isolation cooling 0.5 x 10 -8 UFSAR/DAEC-1 T3.5-2 Revision 13 - 5/97 Table 3.5-2 PROBABILITY OF RELEASE OF RADIOACTIVE MATERIAL TO ENVIRONMENT DUE TO RECIRCULATION PUMP MISSILES
Events Probability per Year of Recirculation Pump Missiles (with LOCA a) All Other Causes Except Pump Missiles
Loss of containment
integrity 30 x 10-9 1 x 10-9 Fuel damage b and loss of containment integrity 63 x 10-13 2.1 x 10-13 aGiven a design-basis LOCA probability of 1 x 10 -5 per year. bFuel damage probability based on "realistic" core heatup assumptions is 2.1 x 10-4 per year.
UFSAR/DAEC-1 T3.5-3 Revision 13 - 5/97 Table 3.5-3 PROPERTIES OF A 120-DEGREE SEGMENT FROM A 38-IN. WHEEL a Parameter Value Fragment angle 120 degrees Fragment weight 5944 lb Minimum proj. area 3.657 ft 2 Maximum proj. area 8.368 ft 2 Failure speed, % of 1800 rpm 169 Initial velocity 666.8 ft/sec Energies Initial, transition 41.0 x 10 6 ft-lb Initial, rotation 23.5 x 10 6 ft-lb Outside turbine casing, translation 20.5 x 10 6 ft-lb After air drag (vertical trajectory) 16.3 x 10 6 ft-lb aSee Reference 5. UFSAR/DAEC-1 T3.5-4 Revision 13 - 5/97 Table 3.5-4 PROPERTIES OF A BUCKET VANE
Parameter Value Vane weight 27.25 lb Vane tip area 1.41 in. 2 Vane root area 4.36 in. 2 Vane maximum thickness 0.920 in. Failure speed, % of 1800 rpm 169 Failure velocity 1600 ft/sec Failure energy 1 x 10 6 ft-lb Estimated velocity after penetrating casing 1100 ft/sec Estimated energy after penetrating casing 0.5 x 10 6 ft-lb UFSAR/DAEC-1 T3.5-5 Revision 13 - 5/97 Table 3.5-5 PENETRATION COEFFICIENTS Material K 3000-psi reinforced concrete 0.00476 6000-psi reinforced concrete 0.00282
Table 3.5-6 PENETRATION ABILITY OF POSTULATED TURBINE MISSILES
Concrete Compressive Strength (psi) 38-in. Wheel LTM 38-in. Wheel HTM Bucket Vane LTM Bucket Vane HTM 6000 1 ft 9 in. 1 ft 6 in. 0 ft 6 in. 0 ft 6 in. 3000 3 ft 0 in. 2 ft 6 in. 1 ft 0 in. 1 ft 0 in.
UFSAR/DAEC - 1 3.6-1 Revision 23 - 5/15 3.6 PROTECTION AGAINST DYNAMIC EFFECTS ASSOCIATED WITH THE POSTULATED RUPTURE OF PIPING 3.6.1 POSTULATED PIPING FAILURES IN FLUID SYSTEMS OUTSIDE CONTAINMENT
3.6.1.1 Design Bases
Not required for FSAR.
3.6.1.2 Description
A detailed analysis has been made of all high-energy lines located outside containment at the DAEC to determine if the rupture of any such line would impair the ability of the plant to be shut down and maintained in a safe shutdown condition.
This analysis was conducted in accordance with the criteria set forth in the attachment to the AEC letter of December 15, 1972, "General Information Required for
Consideration of the Effects of a Piping System Break Outside Containment," and supplemented by analysis utilizing the revised guidance contained in NRC Generic Letter 87-11 as well as the loss of offsite power criterion in subsection 3.b.(1) from the Branch Technical Position SPLB 3-1 attachment to Standard Review Plan No. 3.6.1, Revision 2.
These figures are not revised to show subsequent changes.
3.6.1.2.1 Assumptions
- 1. The assumed modes of pipe failure are as follows:
- a. Circumferential breaks, Figure 3.6-42, are those breaks that are perpendicular to the pipe axis. The break area is taken to be the same
as the internal cross-sectional area of the pipe unless the pipe is adequately restrained to prevent relative motion of the two sides of the break. Dynamic forces resulting from such breaks are assumed to separate the piping axially and cause whipping in any direction normal to the pipe axis, unless the pipe is restrained to prevent such motion.
- b. Longitudinal breaks, Figure 3.6-43, are those breaks that are parallel to the pipe axis. The break area is equal to the effective cross-sectional flow area upstream of the break location. Dynamic forces resulting from such breaks are assumed to cause lateral pipe movements in the direction normal to the pipe axis, unless the pipe is restrained to prevent such motion.
UFSAR/DAEC - 1 3.6-2 Revision 23 - 5/15 c. Critical size cracks are those breaks that are taken to be one-half the pipe diameter in length and one-half the wall thickness in width.
- 2. Circumferential and longitudinal breaks have been assumed to occur at the following locations in each piping run or branch run of Seismic Category I piping. Only a single break has been assumed to occur.
- a. Terminal ends.
- b. Any intermediate locations between terminal ends where either the circumferential or longitudinal stresses derived on an elastically calculated basis under the loadings associated with seismic events and
operational plant conditions exceed 0.8 (S h + S A), where S h is the stress calculated by the rules of NC-3600 and ND-3600 for Class 2 and 3 components, respectively, of the ASME Code, Section III, Winter 1972 Addenda, and S A is the allowable stress range for expansion stress calculated by the rules of NC-3600 of the ASME
Code, Section III, or the Code for Pressure Piping, ANSI B31.1.0, 1967, or the expansion stresses exceed 0.8 S A.
- c. Two intermediate locations in addition to those determined by the above, selected on the basis of highest stress determined by taking the sum of normal operation stresses and seismic stresses.
Note: The above criteria for intermediate breaks has been relaxed by use of the criteria in NRC GL 87-11. This applies to the RCIC and HPCI steam supply piping, RWCU
return piping, and the Feedwater lines between the drywell penetrations and the
outboard whip restraints.
- 3. For Nonseismic high-energy piping, a single circumferential and longitudinal break has been postulated at any location.
- 4. A critical size crack has been postulated to occur at any location along the length and at any point around the circumference of a pipe carrying high-energy fluid. This criterion is not applied to the HPCI and RCIC steam supply lines or the RWCU return piping from the regenerative heat exchanger to the feedwater system, or the Feedwater lines between the drywell penetrations
and the outboard whip restraints, which have been analyzed pursuant to the
revised criteria in NRC GL 87-11.
- 5. The postulated break has been conservatively assumed to occur during normal steady-state operating conditions at rated power.
- 6. Loss of offsite ac power has been assumed to occur concurrently with the postulated failure of the high-energy pipe except where the revised criteria of
NRC Generic Letter 87-11 have been applied using the criterion in the
Standard Review Plan (SRP) as indicated in Section 3.6.1.2. This is, in UFSAR/DAEC - 1 3.6-3 Revision 23 - 5/15 general, a conservative assumption. However, in some cases, a loss of normal ac power actually mitigates the effects of the postulated pipe break (e.g., failure in the feedwater system). In those cases, it has been conservatively assumed that offsite power is not lost.
Use of the loss of offsite power guidance from the SRP with the implementation of Generic Letter 87-11 is a departure from the conventional
practice used in analyses at DAEC; however, it is consistent with the guidance in the Standard Review Plan, Revision 2. This methodology was followed
since Generic Letter 87-11 was initiated to provide a revision to the guidance
given in the SRP and by following these guidelines, greater consistency resulted between the DAEC practice and the methodology provided in the SRP. This practice may be followed if other opportunities arise for performing revisions to earlier analyses.
- 7. No other accident has been assumed to occur concurrently with the pipe failure outside the containment.
- 8. A single failure of an active component has been assumed to occur in analyzing the accident and the ability to safely shut down the plant.
3.6.1.2.2 General Approach
The analysis has been conducted using the following general procedure:
- 1. The high-energy piping systems were identified using the criteria that the service temperature is greater than 200ºF and the design pressure is greater than 275 psig. The systems meeting those criteria are the following:
- a. Main steam.
- b. Feedwater.
- c. HPCI steam.
- d. RCIC steam.
- e. Reactor water cleanup (RWCU).
- f. High-energy sampling and instrument sensing lines.
- g. HPCI discharge piping in the steam tunnel upstream of the normally closed inboard isolation valve (MO2312).
- 2. The systems required for safe shutdown of the reactor for the postulated pipe failure of each of the high-energy systems were identified. By verifying that these systems are maintained operational in the event of the postulated UFSAR/DAEC - 1 3.6-4 Revision 23 - 5/15 failure of the piping systems, it is ensured that the plant can be safely shut down and maintained in a safe shutdown condition.
- 3. physical arrangement of each high-energy piping system was investigated to determine the potential effects of pipe whip or jet impingement on structures, systems, and components required for safe shutdown of the plant.
- 4. The electric cables that could be broken by either pipe whip or jet impingement were identified. Each cable that could be broken was tabulated and the effects of its loss analyzed with respect to the ability to safely shut
down. The loss of required redundancy was considered. The results of that
study were incorporated in the analyses of shutdown capability that were included in the study of each high-energy system.
- 5. The effects of steam pressurization of the compartments that could be pressurized by the failure of any of the identified high-energy lines were
investigated.
- 6. The environmental effects of postulated ruptures of any of the high-energy lines were evaluated.
- 7. A site study was made of each area of concern to verify that no other potential problems exist. Vent areas between compartments were quantitatively
checked. 8. For those problem areas identified by the above process, alternative corrective measures were considered and a method of correction decided upon.
- 9. The modification to install a HPCI High Pressure Keep Fill system introduced the possibility for the HPCI discharge piping, in the steam tunnel, upsteam of the normally closed inboard isolation valve (MO2312), to at times, meet the criteria of a high energy line. The maximum possible energy release as a
result of a HPCI discharge line break is bounded by that of a feedwater line break in the steam tunnel, and therefore, does not require analysis.
3.6.1.2.3 Inherent Safety Features of the DAEC
The seismic classification of the various structures and systems is given in Section 3.2.1.
Safeguard equipment is located in Seismic Category I structures with redundant equipment being physically separated by distance as well as Seismic Category I walls.
UFSAR/DAEC - 1 3.6-5 Revision 23 - 5/15 By the nature of this type of reactor plant, The shielding walls also mitigate the effects of a high-energy pipe failure by providing physical separation between the high-energy lines and other equipment.
Using the criteria previously outlined, it has been determined that no postulated high-energy pipe failure can cause damage from pipe whip, jet impingement, external overpressurization, or environmental conditions to the control room complex.
There are no high-energy lines in the vicinity of the diesel-generator rooms. No postulated high-energy pipe failure can cause damage from the pipe whip, jet impingement, or environmental conditions to the onsite emergency ac power supply.
Information on pressure integrity of piping, applicable codes, specifications, and piping classification is presented elsewhere in this chapter. The quality control and inspection programs that have been used for piping systems outside the containment are
specified in this chapter and in Chapter 17.
The following general comments apply to the design of the DAEC relating to the environmental effects of a postulated high-energy pipe failure in addition to those in Section
3.11.
- 1. All Class 1E cabling used throughout the DAEC has been selected to comply with the environmental specifications as outlined in Section 7.3.5 (safety-
related) based on testing and/or analysis.
- 2.
- 3. The valve motor operators used outside the primary containment are similar to those used inside the primary containment.
The used within the primary containment under high-temperature saturated steam conditions is described in Section
3.11.
- 4. Safeguard equipment is protected from the direct effects of postulated pipe failures which could cause the environment in which it is located to become environmentally harsh per Reference 9 of Section 3.11. Plant specific high energy-line break evaluations have demonstrated that the possibility exists for This scenario however, does not prevent the mitigation of the event since the redundant train of RHR is UFSAR/DAEC - 1 3.6-6 Revision 23 - 5/15 available. No high-energy line unrelated to the equipment is routed through a safeguard equipment room.
- 5. The criteria for routing of electric cabling are presented in Chapter 8. Factors such as physical separation and encasement of cabling in conduit further mitigate the effects of the postulated pipe failure on the ability of the
plant to be safely shut down.
- 6. There are no safeguard instrument panels located in compartments through which high-energy piping is routed, nor ar e there direct line-of-sight paths of communication (doors, wall penetrations, etc.) between high-energy pipes
and safeguard panels. Therefore, the direct effects of pipe whip and jet impingement on such panels can be neglected.
- 7. As such, it is highly unlikely that steam from a high-energy pipe failure would be drawn into the control room.
3.6.1.2.4 Analysis
3.6.1.2.4.1 Structural Loading. The methods used to evaluate the adequacy of the structures that could be affected by the postulated failure of a high-energy line outside the primary containment are the same as those presented in Section 3.8.
3.6.1.2.4.2 Jet Impingement Loading. The following formulas were used to evaluate the various categories of jet impingement loading:
- 1. Jet impingement from a circumferential high-energy pipe failure as a function of distance x from the break:
P x= K j {P o - P a} [1 + 2x tan/D]2 where
P x = pressure of jet as a function of distance Kj = thrust coefficient
= 1.26 for steam or two-phase flow = 2.0 for subcooled water such as feedwater or cleanup flow
P o = pressure in pipe just upstream of break (psi) x = distance from break (in.) D = pipe inside diameter (in.)
P a = ambient pressure outside of pipe before break (psi)
= half angle of jet dispersion = 10 degrees (tan = 0.17633)
UFSAR/DAEC - 1 3.6-7 Revision 23 - 5/15 2. Jet impingement from a longitudinal high-energy pipe failure as a function of distance x from the break:
P x = K j {P o - P a} (1 + 2x tan /) (1 + 2xtan /w) where the terms are as defined above and
= crack length (in.) = 2D w = crack width (in.) = A/
- 3. Jet impingement from a critical crack fa ilure in a high-energy pipe as a function of distance x from the break is assumed to be the same as for a longitudinal failure.
It has been assumed that jet impingement forces are developed in zero time and that any changes in the forces as a function of time are negligible. Jet impingement is assumed to cease when the high-energy line is isolated.
3.6.1.2.4.3 Factors to Account for Target Shape.
- 1. For flat or concave surfaces, the effective area of jet interaction was taken to be equal to the projected area of the target normal to the jet or the expanded area of the jet normal to the target, whichever was smaller.
- 2. For circular targets, such as pipes, the effective area of interaction was taken to be 0.6 of the cross-sectional area of the target normal to the jet or the expanded area of the jet normal to the target, whichever was smaller.
Therefore, the force on a circular target is given by:
F = 0.6 P X A T where
P x = jet impingement pressure as calculated by the appropriate formula discussed above A T = cross-sectional area of the target normal to the jet
3.6.1.2.4.4 Jet Thrust Forces. The reaction forces acting on the pipe caused by the momentum change of fluid flowing through th e break were calculated using the following relationship:
F j = K j A b (P o - P a) where K j , P o , and P a are as defined above and
F j = jet reaction force (lb) A b = area of break (in.
- 2)
UFSAR/DAEC - 1 3.6-8 Revision 23 - 5/15 3.6.1.2.4.5 Compartment Pressure Analysis Model. A postulated high-energy pipe rupture is the expulsion of high-energy steam or a steam-water mixture out of the ruptured pipe into the surrounding compartment. As the pressure builds up within the compartment, the steam-air-water mixture flows through openings to relieve the resultant pressure. The equilibrium pressure achieved depends on the number and shape of the openings between the compartments, the volume of each compartment, and the blowdown rate from the broken pipe. Differential-pressure analyses were made to calculate the pressure responses of two compartments during the postulated event. The calculations include mass and energy balances of the two-phase, two-compartment, steam-air-water mixture as high-energy fluid enters the compartments during the event and passes through the various compartment vent openings. There are no provisions in the calculations for heat transfer since heat transfer would have a negligible effect on compartment pressures for the short time following the rupture within which peak differential pressure occurs.
The analyses conducted yielded pressure versus time relationships that were used to determine if the structures surrounding the affected compartments or vital equipment within the compartments were capable of withstanding the pressurization. In all the cases the
analyses indicated that sufficient vent area was available to prevent overpressurization.
3.6.1.3 Safety Evaluation
In each of the system evaluations described below, the following systems were
analyzed with regard to their required functi ons for the safe shutdown of the plant. The applicability of each of these systems was determ ined using the criteria set forth in Sections 6.3 and 7.3 and Chapter 15.
- 1. Pressure relief system.
- 2. Reactor scram protection.
- 3. Reactor vessel isolation.
- 4. Required core cooling (RHR, HPCI, RCIC, ADS, and emergency service water).
In all cases, the system evaluation included the electric power supplies and the instrumentation associated with a particular system.
UFSAR/DAEC - 1 3.6-9 Revision 23 - 5/15 Pipe failure in the main steam system outside the primary containment is discussed in Chapter 15. The design of the main steam lines, flow restrictors, and isolation valves is presented in Chapter 5. As described in Section 7.3, the main steam lines will automatically isolate in the event of a postulated failure.
With respect to pipe whip, the postulated main steam line break locations are the terminal ends and the two highest stressed intermediate points. Specifically, these locations
are the following:
- 1. The connections at the downstream side of the outboard main steam isolation valves. It is assumed that at these locations either a longitudinal or a circumferential break could occur at the junction of the steam pipe to the
isolation valve.
- 2. 3. The connections at the turbine stop valves.
As part of the design of the DAEC to ensure the protection of the outboard main steam isolation valve in the unlikely event of a pipe rupture in the steam tunnel or the turbine building, pipe whip restraints have been placed in the steam tunnel around the main steam
and feedwater piping as shown in Figure 3.6-44. These restraints preclude damage from pipe whip due to a pipe rupture in this area from having any unacceptable effect on the ability of
the plant to be safely shut down.
To mitigate the effects of a pipe break down stream of the outboard isolation valve, restraining structures have been placed in the area immediately external to the containment.
The major component in this system of restraints is the it is capable of restraining a full spectrum of loading conditions, including the rupture of the main steam/feedwater line.
to control the movement of the piping in the unlikely event of a pipe rupture in this area. These
structures are capable of restraining the full thrust force of the ruptured pipe.
As discussed in Section 7.3, the HPCI, RCIC, and main steam lines are equipped with
differential-pressure switches that isolate the respective lines on high flow. In the event of a feedwater line break, a low suction pressure alarm and feedwater pump trip is initiated and a reactor scram will occur on low water level.
With regard to the possible effect of jet impingement, The possible effects of jet impingements on this conduit have been analyzed; no protection is
required.
2013-012 UFSAR/DAEC - 1 3.6-10 Revision 23 - 5/15 It is concluded that whipping of the main steam lines from the postulated failure will not result in the failure of any other high-energy lines or equipment required for safe
shutdown of the plant.
It has been determined by analysis that the steam tunnel structure can withstand the combined effects of dead loads and steam tunnel peak pressurization plus the jet impingement loads from the steam line break.
A blowout panel in the entrance of the steam tunnel to the turbine building prevents
any excessive pressure buildup in the tunnel.
Jet impingement and environmental effects from a critical crack in the main steam lines have been investigated. The critical cracks have been assumed to occur at any point along the main steam pipes and at any point on the circumference of the pipe. It has been assumed that such breaks can occur at the most adverse location with respect to structures or components required for safe shutdown. It has been determined that no such failures in the main steam lines will adversely affect the ability to safely shut down the plant. The analysis leading to this conclusion included a study of the effects of jet impingement on the adjacent pipes; on the floor, walls, and ceiling of the tunnel; and on the
It is concluded that the plant can be safely shut down following a postulated main steam line failure outside the primary containment.
3.6.1.3.2 Turbine Building
As the main steam lines enter the turbine building, they are routed as shown in The reactor protection cabling for the low condenser vacuum and turbine stop valve fast-closure signals for main steam isolation valves This situation presents no problem, however, as backup isolation signals exist for the main steam isolation valves as discussed in Section 7.3.
As the main steam lines approach the turbine stop valves, their configuration is such
that a rupture could cause a line to whip toward the control building. This situation has been
analyzed, and the presence of two reinforced-concrete block walls, a corridor, the reinforced-concrete wall of the control building, and the floor slab at ensures that no loss of structural integrity of the control building will occur.
The only other potential problem area in the turbine building was found to be
UFSAR/DAEC - 1 3.6-11 Revision 23 - 5/15 For areas potentially subject to adverse effects of a postulated feedwater pipe break, this cabling is routed in rigid steel conduit outside zones of potential pipe impact and with sufficient separation/protection to preclude detrimental damage from jet impingement. It is concluded that the plant can be safely shut down following a postulated feedwater pipe break outside the containment.
3.6.1.3.3 Reactor Building
In the reactor building, the HPCI and RCIC steam lines were analyzed and break locations determined as shown in Based upon the revised criteria for determining break locations, which is contained in NRC GL 87-11, the HPCI and RCIC steam lines are postulated to break only at the terminal ends. Thus, only the steam turnnel and HPCI and RCIC equipment rooms are affected. It was found that at various points ruptures could cause damage to RHR service water and/or emergency service water systems. In all cases, the rupture would result in having to shut down the equipment located in one corner room (one-half of the RHR system) as a result of a loss of cooling water or service water to the equipment.
The loss of the redundant service water loop and thus the ability to shut down and cool down the plant is dependent on the ability to manually operate the RHR heat exchanger discharge valve in the redundant loop. This valve is the only single active component in each of the service water trains whose failure would result in the loss of that train. These
valves are located so that they would be accessible in the event of a rupture that would disable the other train. The location of these valves and the cross-connection capability of the RHR and emergency service water systems as described in Section 5.4.7 ensure that sufficient time would be available to restore the ability to shut the plant down and maintain it
in a safe shutdown condition.
A rupture of the RCIC steam line in the RCIC room would result in damage only to the RCIC system itself.
A rupture of the HPCI steam line inside the HPCI room could result in damage to the emergency service water system or one of the RHR service water valves described
previously. In this case, the above discussion relative to the RHR heat exchanger valve
would also apply.
2013-016 2013-016 UFSAR/DAEC - 1 3.6-12 Revision 23 - 5/15 The RWCU system is described in Section 5.4.8. The system contains high-energy fluid in only a portion of its piping.
A postulated failure in the cleanup system will result in a single-ended failure. A check valve in the return line immediately upstream of the connection into the feedwater piping would prevent backflow from the return side of the break. Automatic isolation of the cleanup system is initiated by various line break sensing instruments. These are identified in
Section 5.4.8.
It has been determined that pipe whip or jet impingement from pipe failure at the postulated break points will not adversely affect the ability to safely shut down the plant. The substantial shielding walls surrounding each of the equipment cells prevent damage to any equipment outside the cell containing the postulated break. The cleanup system equipment cells do not contain any components not related to the cleanup system that are
required for safe plant shutdown.
Critical cracks in the RWCU line have been evaluated and will not adversely affect the ability to safely shut down the plant.
It is concluded that the plant can be safely shut down and maintained shut down following a postulated RWCU system line failure outside the primary containment.
All instrument and sample lines in the DAEC are routed such that their failure would not result in any damage to equipment necessary for safe shutdown.
The postulated rupture of an instrument or sample line would be a single-ended
failure that would not result in any direct effects on structures, systems, or components required for shutdown. Pipe whip and jet impingement effects would be negligible due to the small size of these lines and the fact that they are completely enclosed in protective channels. Therefore, the only potential effects of a postulated line failure are environmental in nature and would not have any adverse effect on the ability of the plant to be safely shut down.
Instrument line isolation considerations and the effect of an instrument line rupture
are discussed in Section 6.2.4.
From the discussions presented above, it is concluded that the design of the high-energy piping systems outside containment on the DAEC is such that their failure would not result in the inability of the plant to be shut down and maintained in a safe shutdown
condition.
UFSAR/DAEC - 1 3.6-13 Revision 23 - 5/15 3.6.2 POSTULATED PIPING FAILURES IN FLUID SYSTEMS INSIDE CONTAINMENT
3.6.2.1 Design Bases (PSAR)
Not required for FSAR.
3.6.2.2 Description
Pipe failures inside containment can give rise to pipe movement and jet impingement effects. Protecting against pipe movement and jet impingement involves the consideration of the "source" and the "target." The source investigation must include the determination of where a pipe may fail and the means to compute the resultant forces. The failed pipe can result in a pipe movement in reaction to the expulsion of a high-pressure fluid or jet-like impingement of the fluid itself. The target investigation must include identification of components or systems that are considered essential, and the means to determine how much force or energy a target can receive without the impairment of function.
Table 3.6-1 is a matrix that was developed to display the interrelationships that exist between the identified potential sources of pipe movement and the systems, components, and structures important to safety. These interrelationships were investigated using design drawings and a model of the DAEC drywell.
as an aid in picturing the restraint system, pipe routing, and physical arrangement within the drywell area.
3.6.2.2.1 Assumptions
- 1. General Assumptions
- a. The pipe break was assumed to occur in combination with the DBE and loss of offsite power.
(1) Direct current and standby ac power were available.
(2) No credit was taken for the integrity of any piping or the functioning of any system not classified as Seismic Category I.
- b. The single failure of an active component (i.e., the single-failure criterion) was applied after the given rupture and its immediate effects
had occurred.
- c. The plant was in normal operation before the initiating event.
UFSAR/DAEC - 1 3.6-14 Revision 23 - 5/15 d. In this analysis, the loop selection logic is assumed to be inoperable. Since it is required for the recirculation-line break only, protecting the core spray loops from a recirculation-line break is
sufficient.
- 2. Source Piping Assumptions
- a. All pipes were assumed to be susceptible to either a circumferential or slot failure throughout that portion of their length that is normally exposed to reactor pressure. Failures were assumed to be nonmechanistic; no attempt was made to specify how they occurred.
- b. In those portions of systems where the disposition of valves and/or check valves is such that a sustained jet will not result subsequent to a pipe rupture, pipe movement or jet spray was not considered.
- c. Pipe trajectories were based on the reaction of the source pipe to a circumferential rupture. Various pivotal points for the pipe were chosen and the resultant movement evaluated.
- 3. Target Impingement Assumptions
- a. A line of a given size and schedule will not cause a line of equal or larger size and schedule to fail to perform its function if the smaller
should strike the larger.
- b. Any component (except the containment vessel wall) within the jet spray of the postulated rupture was not considered functional following the event. Ten pipe diameters radially from the rupture and
a 10-degree half-angle of dispersion were used as guideline values for determining the extent of the jet spray.
- c. Valves that are closed during normal operation were not required to have protection to ensure closure during the isolation process.
- d. The local temperature of the primary steel containment under jet impingement was taken as 300F. In areas where local deformation is not permitted, the allowable stresses corresponding to this temperature
were used.
3.6.2.2.2 General Approach
The following systems were identified as essential to plant safety:
- 1. Safety-related systems required to shut down the reactor.
UFSAR/DAEC - 1 3.6-15 Revision 23 - 5/15 2. Safety-related systems required to isolate the reactor vessel and primary containments.
- 3. Safety-related systems required for cooling the core.
- 4. Safety-related systems required for containment integrity.
The study was conducted by analyzing, for each potential source, the consequences of
a rupture as it affected the essential targets. In general, the following four solutions satisfied
the objective:
- 1. The source pipe is restrained.
- 2. The source and target are separated by a great enough distance such that the source cannot affect the target.
- 3. The source cannot affect the target due to the interference provided by intervening structures.
- 4. If contact can be made by a jet or the pipe traveling on a hypothesized trajectory, the consequences will not prevent the safety-related functions from being accomplished, that is, the containment vessel wall will not be penetrated or adequate system/component redundancy exists to ensure the completion of
a required function.
3.6.2.2.3 Inherent Safety Features of the DAEC
There are five inherent design features that further minimize the effects of pipe failures:
- 1. Based on conservative piping design using proven engineering practices, the proper choice of piping materials and loadings, and conservative quality control standards and procedures for piping, materials, fabrication, and installation, it is most unlikely that pipes will catastrophically fail.
- 2. The structures, including the primary containment, are conservatively designed. The primary containment vessel is completely enclosed in a
reinforced-concrete structure having a thickness of 4 to 7 ft.
- 3. A dimensionally controlled gap is provided to permit the growth of containment due to temperature and pressure. A nominal gap of 2 in. is provided up to the drywell head so that the containment may expand or deform the distance without suffering a "tear." This feature has also been evaluated in areas subject to jet force and has been determined to be adequate.
Since concrete is not available at the vent openings, deflection plates have
been placed across these openings for jet protection. UFSAR/DAEC - 1 3.6-16 Revision 23 - 5/15 4. If a pipe leak should occur, means for detecting even small leaks are available so that proper action can be taken before they develop into
appreciable breaks.
- 5. The design provides for the consideration of system redundancy to optimize the reliability of accomplishing essential functions. The drywell design has used this concept and the preliminary study to ensure that the independence of redundant systems is maintained.
3.6.2.2.4 Analysis
3.6.2.2.4.1 Jet Force. The jet force is that force resulting from the sudden expansion of steam or water through a given rupture as measured at the jet exit plane.
F J = KP V A B where
F J = jet force
K = thrust multiplication factor to account for the effect of the change in momentum of the escaping medium. K = 1.2 for saturated steam or water
P v = reactor vessel normal operating pressure (1050 psia)
A B = area of break, assumed equal to free flow area of the pipe
The jet is assumed to expand in the shape of a truncated cone with a half-angle of dispersion equal to 10 degrees. This dispersion spreads the total force over a greater area as radial distance from the rupture increases. No quantitative use was made for the reduction in jet forces due to flow restrictions, and, since a loss of function was assumed, no use was made of the reduction in jet forces due to target shape.
3.6.2.2.4.2 Restraint Spacing. Pipe movement is restricted by limiting the distance between pipe restraints, where required, to some dimension less than the critical plastic hinge
length of the pipe.
Restraints have been added on any pipe which, by analysis, could circumferentially fail and either hit the drywell liner with an unacceptable amount of energy or hit some component whose loss is unacceptable. The restraints added to the RHR suctions and
discharge on the recirculation loop fall into this category.
UFSAR/DAEC - 1 3.6-17 Revision 23 - 5/15 1. Rupture at an elbow Where the assumed rupture is at a pipe elbow or fitting for a circumferential
pipe break, the critical plastic hinge length (L l) is determined by the moment-resisting capabilities of the pipe (M p) and the magnitude of the blowdown jet thrust (F j), assuming the pipe acts as a simple cantilevered member. The critical plastic hinge length of this condition is determined as
L 1 = M p F J (DLF) The dynamic load factor (DLF) used in the loading design of these systems is dependent on the ratio of maximum allowable deflection to yield
deflection (u), and the ratio of load duration to the period of the structure. Based on the structures and loads expected, a value of 1.25 was determined to
be conservative and was used in the design.
M p is equal to the plastic moment of the pipe based on the following formulas:* M P = Sy I 1.27 (recirculation line) R
M P = Sy 4/3 (R 3 - r 3) (other restrained piping) where
I = Moment of inertia of the pipe = (R 4 - r 4) 4 Sy = yield strength of the pipe material at operating conditions.
R = pipe outside radius
r = pipe inside radius
- 2. Longitudinal Rupture
The second case considered is a longitudinal pipe rupture located between the restraints. The most severe loading condition for this analysis is when the rupture occurs midway between the restraint brackets. The critical plastic
hinge length (L
- 2) is analyzed if the pipe is considered a simply supported member, determined by L 2 = M p F J (DLF) The pipe restraints are designed to allow for normal pipe movements due to temperature and OBE requirements.
- The two formulas produce results that are virtually identical.
UFSAR/DAEC - 1 3.6-18 Revision 23 - 5/15 3.6.2.2.4.3 Restraint Loading. The magnitude of the pipe restraint loads for the bracket design is determined by the following formula:
F 2 = KPA x C
where
k = thrust multiplication factor for the primary two-phase steam-water mixture of 1.20
P = operating pressure of the reactor vessel of 1050 psia
A = free flow area of the pipe
C = load factor
3.6.2.2.4.4 Recirculation Line Restraint Criteri
- a. The recirculation piping loops are restrained against pipe movement, in the event that the recirculation-line ruptures, by a system of pipe restraint brackets. Both longitudinal (axial) and circumferential (guillotine) type pipe ruptures were considered in the design of the pipe restraint system. The pipe ruptures were assumed to occur anywhere in the system, and the consequential damage of the whipping pipe was evaluated. The restraints are located and spaced in such an arrangement on the recirculation piping system as to protect the primary containment pressure boundary, to ensure that the DBA break area is not exceeded, and to ensure essential component protection with sufficient emergency core cooling capability.
The spacing of the pipe restraints is established at less than the distance between that restraint and the primary containment drywell shell plate. The ruptured recirculation loop pipe is restrained from touching the containment shell plate everywhere within the drywell.
The pipe restraint system is also designed to limit excessive pipe movements in areas
where the ruptured recirculation loop pipe could strike an adjacent pipe which, if ruptured, would exceed the DBA break area. In this analysis, the recirculation pipe is not allowed to
touch an adjacent pipe of sufficient size so that the total combination of break areas of each pipe (including any other adjacent pipes which the second or third piping systems might rupture due to cascading effects) does not exceed the DBA break area. This excessive pipe movement is restricted in these areas by limiting the distance between restraints to some dimension less than the critical plastic hinge length of the pipe.
UFSAR/DAEC - 1 3.6-19 Revision 23 - 5/15 For the recirculation line, the dynamic load factor was found to vary for each pipe size, pipe system arrangement, pipe restraint location, and various other conditions such as locations of hangers, pumps, valves, etc. This value varied from 1.1 to 1.7 for various configurations modeled, and an average value of 1.50 was established to standardize restraint
hardware design and construction.
The pipe restraint hardware design was established to limit the maximum allowable
stresses in the bracket to 150% of AISC C ode allowable and 90% of the breaking strength of the cables for this infrequent loading condition. The restraint brackets are fabricated of A-36 material, and the cables are Type 6 x 26 G IWRC extra-improved plow steel
construction.
3.6.2.2.4.5 Main Steam, Feedwater, and HPCI Steam Line Restraint System. The main steam lines, feedwater lines, and HPCI steam line are restrained to prevent excessive pipe movement, that is, restraint spacing is such as to prevent the formation of a plastic
hinge.
Piping restraints and supports installed for the express purpose of preventing pipe motion are designed so that the stresses in the restraint and its supporting structure remain below the material minimum specified yield stress for the combinations of live and dead
loads discussed in Section 3.8.
3.6.2.2.4.6 Containment Impingement. Since the physical integrity of the containment is paramount, no pipe is permitted to strike the containment with enough energy to penetrate it. Restrained lines cannot move, and thus meet this criterion; all other lines were analyzed to ensure that a moving pipe c ould not travel a sufficient distance to acquire the energy required to penetrate the containment.
- 1. Energy imparted to a moving pipe
- a. The energy of the moving pipe at the point of contact is E T = F j S where
F j = the jet force = KP V A B S = distance traveled by the pipe, i.e., arc of motion of the pipe = L L = unattached length of pipe
= angle through which the pipe can move
No credit was taken for the energy dissipated due to strain hardening.
UFSAR/DAEC - 1 3.6-20 Revision 23 - 5/15 b. The energy dissipated after the formation of the plastic hinge is E p = M p
- c. Impact energy (E
- 1)
The impact energy is the energy available at the point of impact
E l = E T - E p
- 2. Energy required to penetrate containment
The Stanford Research Institute (SRI) formula1, 2 for the penetration of steel plates was used.
U = D m T u (O.344 t 2 + 0.00806 Wt) where
U = critical penetration energy, ft-lb f D m missile diameter; when the outer circumference of the pipe is tangent to the outer surface of the plate, the missile diameter is assumed to be the chord formed by the outer surface of the plate and the outside diameter. = 2 (2RT - T 2 )1/2 for R<T = 2R for R T T u = ultimate strength (psi) = 70,000 psi
t = plate thickness (in.)
W = window width = 8D m R = outside radius of pipe
- 3. Penetration of the Containment
The potential for penetration of the containment was determined by equating the energy required to penetrate the containment and the impact energy. For each given pipe size, containment thickness, and unattached length of pipe, there is a minimum distance the pipe must move to satisfy this condition. This value provided an estimate of the piping drywell separation that is critical and indicated where more detailed evaluation was required.
UFSAR/DAEC - 1 3.6-21 Revision 23 - 5/15 3.6.2.2.4.7 Separation. Maintaining the independence of redundant safety systems and components is greatly enhanced by separating the redundant components so that no single postulated event can prevent the safety-related functions from occurring.
The concept of the separation was used throughout the design of the plant. As a specific example, the emergency core cooling systems were analyzed to identify precisely
wherein the redundancy exists. The core cooling function requires both depressurization of
the reactor and low-pressure injection. For large breaks, the reactor will depressurize itself; for small breaks the HPCI-automatic depressurization system (ADS) combination is redundant. Low-pressure coolant injection is by the core spray-LPCI combination; for
recirculation line breaks, core spray loop A, core spray loop B, and LPCI functions are redundant; for all other breaks, core spray A, core spray B, LPCI-A, and LPCI-B are redundant. The protection requirements are summarized below:
- 1. HPCI-ADS
- a. No small steam or water line break can be allowed to cause the loss of two ADS valves or cause the loss of high-pressure coolant injection
and one ADS valve.
- b. No HPCI break can be allowed to cause the loss of any ADS valve.
- c. No intermediate size water break can be allowed to cause the loss of any ADS valve.
- 2. Core Spray-LPCI Systems
- a. No recirculation line or LPCI rupture can be allowed to cause the loss of either core spray loop.
- b. No core spray rupture can be allowed to cause the loss of the LPCI function.
- c. No rupture can be allowed to cause total failure of the core spray system or the loss of one core spray loop and the LPCI function simultaneously.
For any pipe break, at least one RHR loop, including its RHR heat exchanger and service water pump, must remain operable for suppression pool cooling.
UFSAR/DAEC - 1 3.6-22 Revision 23 - 5/15 For the DAEC, the following apply:
- 1. The small steam break is less than 0.09 ft 2 (4.1 in. I.D.).
- 2. The small water break is less than 0.05 ft 2 (3 in. I.D.).
- 3. The intermediate water break is greater than 0.05 ft 2 and less than 0.17 ft 2 (5.6 in. I.D.).
3.6.2.2.4.8 Pipe Ruptures Within the Reactor Shield. Incorporated into the design of the reactor shield is the capability to withstand, without failure, the internal pressure and coincident jet impingement loads resulting from failures of high pressure lines in the shield space region (from the outside diameter of the r eactor vessel). A failure of the reactor vessel (including nozzles) is not considered credible; however, the consequences of safe-end
failures are given full consideration. Safe-ends, even though attached by the reactor vessel manufacturer, are not considered to be an integral part of the reactor vessel but are regarded as a transition piece between the reactor vessel and the primary piping. Although steps have
been taken to effectively preclude safe-end failures, the design criteria developed for the reactor shield consider a full spectrum of breaks up to a double-ended recirculation-line
break at the nozzle to safe-end weld.
Maximum Internal Pressure Buildup
Maximum internal reactor shield static pressure, for any credible event, occurs following a circumferential rupture of the recirculation outlet line at the nozzle to safe-end
weld (shown in Figure 3.6-47). This break area is 2.18 ft 2 (22-in. pipe I.D.).
The penetrations are designed to prevent a full break from pressurizing the shield
space. Since this break occurs at about 12 in. from the outside diameter of the shield, only a fraction of the break flow from the cross-sectional area finds its way into the annular space between the shield and the reactor vessel. Based on the maximum permitted movement of the pipe due to limits set by the shield penetration, as shown in Figure 3.6-48, the effective area
of the cross section blowing down into the annulus is 1.61 ft 2 (74% of the cross-sectional area). A conservative break size of 2.18 ft 2 was assumed.
The parameters used in the pressure response analysis of the reactor shield included a reactor cavity volume of 3136 ft
- 3. The vent area to the drywell was conservatively taken as 133 ft 2. This area is based on the assumption that all openings in the shield have been closed with penetration plugs similar in configuration to those represented in Figure 3.6-46. Only a portion of the openings have plugs, however, so that the actual computed vent area is approximately 269 ft
- 2. As a further conservatism, all vent openings were treated as orifices.
UFSAR/DAEC - 1 3.6-23 Revision 23 - 5/15 Moody's slip-flow model (Reference 3) of two-phase maximum flow with subcooled liquid was used to determine a maximum blowdown rate of 9316 lb/sec ft
- 2. This equates to a constant blowdown rate of 14,999 lb/sec for the 1.61 ft 2 flow area considered. The enthalpy of the flow was 529 Btu/lb.
A conservative calculation of the vent area from the shield space combined with these
criteria results in a peak pressure of 18.98 psi.
Combination Jet Loads and Internal Pressure Buildup
As indicated by Figure 3.6-47, a mechanism whereby significant direct impingement loads can arise on the structural components of the recirculation line outlet shield penetration cannot be identified. However, for smaller reactor coolant lines, it is conceivable that a
nozzle to safe-end weld will be located far enough inside the shield penetration plugs to allow a significant impingement load. It also should be noted that only the recirculation nozzles and the two reactor water level instrumentation nozzles at the top of the core will have shield penetration plugs. The shield openings at other nozzles away from the core region can be made large enough for inservice inspection without removable plugs.
All lines penetrating the reactor shield are listed in Table 3.6-2. For the recirculation inlet line configuration, shown in Figure 3.6-49, a jet is postulated that emanates from a double-ended rupture at the safe-end weld and impinges on the penetration shield plug in an outward manner with an average angle of approximately 31 degrees from the horizontal. As an additional margin of conservatism, it was assumed that the jet force impinging on the shield plugs results from the jet pressure times the projected cross-sectional area of the line.
The impingement force on the blockout is given by the following:
F I = K d K t K a P v A b where
F I = impingement force K d = dynamic load factor K t = thrust multiplication factor K a = correction for target orientation (cos ) P v = pressure 1065 psia (dome pressure of 1055 psia + 10 psi hydrostatic head) A b = effective area of break = angle of impingement
UFSAR/DAEC - 1 3.6-24 Revision 23 - 5/15 The structural rigidity of the shield plug results in a uniformly distributed jet load over the surface area of the inside face of the pl ug. Thus, for the recirculation inlet line break described above, a jet load of 112,590 lb is distributed over a surface area of
2000 in.2 . The peak stagnation pressure in the annulus resulting from 100% of this break pressurizing the shield annulus is 3.8 psi. The coincident jet and pressure loading is 60.2 psi. For smaller lines, the coincident pressure and jet force loadings are less severe. Thus, it may be concluded that an equivalent design pressure of 70 psi ensures an adequate
capability to withstand credible jet force and pressure loadings without intolerable
consequences, as shown on Table 3.6-3.
This configuration was conservatively analyzed to determine the capability of the
shield wall to withstand pressures generated in the annulus between the reactor pressure vessel and the shield. The following criteria are used to estimate shield wall capability:
- 1. That only the two steel plates, acting as a thin cylindrical shell, resist the pressure forces with no credit for wide flange beam or concrete strength.
- 2. That the shear stress in the welds is taken as 1.5 times the normal code allowable.
- 3. That the pressure differential across the shield wall is a constant load although the differential pressure would continually decrease as the drywell is
pressurized.
For these assumptions, the shield wall is capable of withstanding a differential
pressure of 96 psig.
Based on the above, the static pressures and appropriate jet impingement pressures
have been evaluated on all shield plugs. These values are shown in Table 3.6-3. The calculated values were increased by 20% to give design values with an additional margin of
safety. All plugs are restrained for the design values.
3.6.2.2.5 Conclusions
Table 3.6-1 summarizes how the essential targets are protected from potential sources of pipe movement.
- 1. Restraint Design
UFSAR/DAEC - 1 3.6-25 Revision 23 - 5/15 The central feature of the drywell restraint system is a Vierendeel truss network located around the pressure vessel, cantilevered from the biological shield. The main steam and feedwater pipes are run between the vertical members of the truss. The HPCI steam line has ring-type restraints attached to the end vertical truss member. All restraints are anchored either to the biological shield or the truss system.
The largest loading on the truss network results from a rupture of a main steam line. For this situation, the calculated loading at the limiting point is
within the allowable stress (90% of yield strength).
Figure 3.6-46, Sheets 1 through 4, provides photographs of the restraint system within the DAEC drywell.
- 2. Pipe Penetrations
The containment pipe penetrations are designed to withstand the normal containment environmental conditions that may prevail during plant operation
and to retain their integrity during all postulated accidents.
- 3. Jet Force on Containment Wall
The drywell for the DAEC meets the following requirements for jet forces:
Location Jet Force Maximum (lb) Interior Area Subjected to Jet Force, (ft
- 2) Force per Unit Area (psi)
Spherical
portion of
drywell 393,000 2.19 (recirc. line) 1245 Cylinder
and sphere
to cylinder
transition 325,000 1.80 (main steam) 1245
Assuming:
- a. 1050 psia reactor pressure.
- b. 1.2 thrust multiplication factor.
- c. Interior area = inside area of pipe considered.
UFSAR/DAEC - 1 3.6-26 Revision 23 - 5/15 4. Pipe Penetration of Containment Wall
To penetrate the containment, a pipe must acquire a kinetic energy greater than that which can be absorbed by the containment itself. The energy consideration concept is used because of the impulse-like characteristic of the impact. Using this approach, and assuming a drywell thickness of 0.75 in. (the thinnest plate any potential source passes through), the following table has
been derived: Pipe Size (in.) Sch. 80 Distance Required To Penetrate Containment (ft) 1 26.4 2 12.1 3 8.0 4 5.7 6 3.6 8 2.6 10 2.0 12 1.6 16 1.3 20 1.0 22 0.9
For pipes with an unsupported span less than 6 ft or for thicker containment plates, the allowed travel distances are greater. These situations
were evaluated on a case basis.
The above table is based on the following assumptions:
- a. The length of a pipe is the straight-line distance between restraints or between penetrations about which a pipe can hinge. If a pipe ruptures circumferentially, it will travel in a path similar to an arc of a
circle with the unattached length acting as the radius. The actual
unattached length depends on the hinge point and the particular pipe
configuration.
- b. The distance required to penetrate the containment represents the distance the end of a pipe must travel to pose a hazard to containment
integrity. If the end of the pipe can travel a greater distance, the pipe must be restrained.
UFSAR/DAEC - 1 3.6-27 Revision 23 - 5/15 c. A single value of the minimum distance to travel can be used for longer pipe lengths because of the asymptotic nature of the distance
required versus the length of pipe curves of the various sizes of pipe.
- 5. Containment Overpressure
The containment will not be overpressurized as long as the combined pipe break areas are less than the design-basis accident of 2.523 ft
- 2. On the DAEC, only the recirculation line, main steam line, and RHR suction and discharge to the recirculation line have the minimum inside diameter that can create an overpressure condition in combination with another smaller pipe break.
Line Nominal Diameter (in.) Inside Diameter (Sch. 80) (in.) Inside Diameter of Maximum Permitted Secondary Break (in.)
Recirculation
line 22 19.75 8.5 Main steam line 20 18.0 11.8 RHR discharge 20 18.0 11.8 RHR suction 18 16.1 14.2
Since these lines will be restrained, they will not be able to sever another line and possibly overpressurize the containment.
- 6. Small Lines
Lines 1 in. or less in diameter are not considered as potential sources of pipe movement or jet spray. Based on the previously discussed criteria, a 1-in. line would have to travel 26.4 ft unimpeded to penetrate a 0.75-in. containment plate. A circumferential rupture of 1-in. diameter would cause a jet to form, which, according to the model on jet forces, will decrease to 45 psi at a distance of 1 ft from the break.
Based on the inherent design features, the conservatism used throughout the analyses as confirmed by the independent evaluation, and the restraint system used, it was found that the effects of a pipe rupture inside the drywell can be accommodated so that containment integrity and adequate emergency core cooling are ensured.
UFSAR/DAEC - 1 3.6-28 Revision 23 - 5/15 3.6.2.2.6 Reanalysis of Recirculation Piping Restraints
General Electric conducted a series of tests 5 on restraint cables to compare actual performance with predicted characteristics. The force-deflection curves for 1-in. cables were plotted and scaling relationships determined such that this data could be extrapolated to other cable sizes. This extrapolated data was then compared to the characteristics used in the previous FSAR analysis. This comparison found a significant discrepancy between assumed
and actual characteristics, with actual characteristics having considerably less strain energy capability to arrest the motion of a whipping pipe.
Portions of the recirculation loop pipe-whip analysis were then reperformed to determine if the cables would still be adequate to perform their design function. Fluid blowdown forces were recalculated using current technology, as a refinement to the approximate methods used in the earlier analysis. For circumferential breaks at terminal points, simplified, nonlinear, time-history response of the pipe and cables to the revised blowdown forces was determined, using actual cable property data. This dynamic analysis
found that the cables could fail under the applied loads.
Upgrade Program
Since the previous evaluation was performed, pipe rupture criteria have undergone considerable evolution. Currently, for nuclear grade piping systems, pipe ruptures are postulated on a mechanistic basis. Terminal ends, tees, and intermediate high stress points are considered the most likely locations where the pipe would break if an accident were to
occur. It was thus considered prudent to upgrade those restraints, which on the basis of break probability, would be the most likely restraints to be loaded. On this basis the following
restraints have been upgraded:
Nine cable restraints on the loops A & B suction, header, and riser lines were
replaced by a new design which uses A304 stainless steel rods as the energy absorbing element.
Two cable restraints on the discharge risers were removed and the existing frame supports were shimmed. (These frames also function as pipe-whip restraints for the main steam pipe.) The final gap between the pipe and shimmed frame is consistent with other pipe-whip frames on the system which are structurally similar and designed for the
appropriate loads.
The new pipe-whip restraints have been designed to appropriate regulatory and industry standards. All work (including material procurement, fabrication, and installation) was performed to quality standards commensurate with those for Seismic Category I
structures.
Figures 3.6-50 through 3.6-54 show the new modifications for the restraints;
UFSAR/DAEC - 1 3.6-29 Revision 23 - 5/15 3.6.3 INDEPENDENT EVALUATION OF THE MAIN STEAM AND RECIRCULATION LINE RESTRAINTS INSIDE CONTAINMENT
This section presents the results of the analytical studies performed by Nuclear Services Corporation to evaluate the adequacy of existing pipe rupture restraint systems in
1972 that were designed by Bechtel and GE for the DAEC. Restraint designs were evaluated by considering circumferential and longitudinal breaks of the main steam line for Bechtel
restraints and of the recirculation lines for GE restraints.
Nonlinear dynamic analyses of the main steam and recirculation lines were performed
to evaluate the adequacy of existing pipe restraints. The restraints are required to protect the primary containment from the results of postulated circumferential and longitudinal breaks in the piping systems. The time histories of maximum strain in the pipe, maximum reaction force in the restraints, and the envelope of maximum deflections of the pipe were determined.
The coupled nonlinear dynamic analysis of the pipe-restraint system takes into account the dynamic nature of the rupture force, impact effects due to the clearance between the restraint and pipe, and elastic-plastic deformation of the pipe-restraint system.
3.6.3.1 Break Locations
Break locations on the piping systems were postulated based on the following guidelines:
- a. Locations that yield potential damage to the containment integrity.
- b. Locations that maximize the loading on the restraint.
- c. Locations that maximize the unsupported span for the pipe.
- d. Locations that help to evaluate the adequacy of different designs of the restraint system, and for each design evaluate different load-resisting paths.
- e. Locations that contain representative points of maximum stresses in the piping system.
The combined primary plus secondary stresses at all locations in the main steam lines due to these same load combinations are less than 1.8 S
- h.
Larger numbers of break locations were postulated for the recirculation lines than for the main steam lines because of the higher stresses and greater number of different configurations of the pipe-restraint system in the recirculation lines.
UFSAR/DAEC - 1 3.6-30 Revision 23 - 5/15 Ten break locations were considered in the analyses and are described as follows:
- 1. Bechtel Restraint Design
Main steam line
- a. Circumferential break at the reactor vessel nozzle (Ml).
- b. Circumferential break at the horizontal header (M2).
- c. Longitudinal break in the radial direction adjacent to restraint (M3).
- d. Longitudinal break in the tangential direction adjacent to restraint (M4).
Refer to Figure 3.6-57 for break location.
- 2. General Electric Restraint Design
Recirculation Suction Line
- a. Circumferential break at the reactor vessel nozzle (Rl).
- b. Longitudinal break in the radial direction adjacent to restraint (R2).
- c. Longitudinal break in the tangential direction adjacent to restraint (R3).
Recirculation Discharge Line
- a. Circumferential break at the discharge elbow (R4).
Recirculation Riser
- a. Circumferential break at the reactor vessel nozzle (R5).
Recirculation Header
- a. Circumferential break at the riser/header tee (R6).
Refer to Figure 3.6-58 for break locations.
UFSAR/DAEC - 1 3.6-31 Revision 23 - 5/15 3.6.3.2 Restraint System Description 3.6.3.2.1 Bechtel Restraint Design
The typical Bechtel restraint design evaluated in the pipe rupture analyses is shown in Figure 3.6-59. The restraint is a closed horseshoe-shaped structural bracket around the pipe.
The bracket cross section consists of inner and outer flanges and a web, all fabricated from
A-441 structural steel plates. These restraints are anchored either to the shield wall or to structural steel framing.
3.6.3.2.2 General Electric Restraint Design
The typical GE restraint design, shown in Figure 3.6-60, consists of two 6 x 26 G IWRC cables, around the pipe and anchored to a structural steel bracket at the cable anchor
plate. The structural steel bracket is fabricated from A-36 steel plates. The movement of the pipe in the outward (radial) direction is resisted by the cable, whereas the structural steel bracket provides resistance to pipe movement in the lateral direction.
3.6.3.3 Design Criteria
3.6.3.3.1 Systems Criteria
The routing of the piping and the location of the pipe rupture restraints, and the structural steel floor framing providing the resistance to pipe movement, were taken from
Bechtel and GE DAEC design drawings. Pipe properties, system temperature, pressure, and material specifications are summarized in Table 3.6-4 through 3.6-20. The stress-strain curves used in the analyses for the different materials are based on information from standard
references. The actual and the idealized stress-strain curves for the slow rate of the loading
are shown in Figures 3.6- 61 through 3.6-63. For ra pid strain rate effects, the ordinates of the stress-strain curves for the ductile materials have been increased by 15% per References
6 and 7. The clearances between the piping and the restraints that are used in the analyses are summarized below.
Pipe Rupture Restraint Clearances
Location Clearance (in.)
Main steam line and restraint 0.5
Main steam line and structural steel floor framing at 13.0
Recirculation line and cable restraint 0.9
UFSAR/DAEC - 1 3.6-32 Revision 23 - 5/15 Recirculation line and structural steel bracket restraint 1.2 Recirculation discharge line and structural steel floor framing at 18.0
A damping value of 2% of critical damping is used in the analyses. 8 3.6.3.3.2 Loading Criteria
The rupture force time curves for the circumferential and longitudinal breaks in the piping systems were generated using information furnished in Reference 9, and are shown in Figures 3.6-64 through 3.6-71. The time-dependent nature of the rupture force is due to the effects of wave propagation phenomena on the piping. In general, there are three principal
periods of interest as described below:
- 1. Initial period, from the instantaneous occurrence of the break until the flow front reaches the first change in direction in the piping system. During
this period the rupture force is given by
F 1 = PA e (3.6-1) where
F 1 = initial rupture force
P = system normal operating pressure
A e = effective flow area of the piping system = KCA
K = break area factor relating effective area A e to the pipe flow area A
= 1.0 for circumferential break = 2.0 for longitudinal break
C = discharge coefficient
= 1.0 for circumferential break = 0.6 for longitudinal break
The time span for the initial period is given by
t 1 = L (3.6-2) V
UFSAR/DAEC - 1 3.6-33 Revision 23 - 5/15 where t 1 = duration of the initial period
L = distance from break point to first elbow for a circumferential break
= twice the break length for a longitudinal break
V = sonic velocity
= 1600 fps for steam = 4000 fps for water
- 2. Intermediate period, during which time there is an unbalanced force in the piping system from the break point to the first elbow. This force is due to the momentum stored in the pipe and is given by L F 2 = A e (v) dz (3.6-3) t where
F 2 = rupture force during intermediate period
L A e (v) dz = 0.7 F 1 for saturated steam t
= P sat F 1, for saturated water P The intermediate period exists until a reflected wave returns from the reservoir. Therefore the duration for the intermediate period is given by
t 2 = L 2 (3.6-4) V
where
L 2 = twice the distance from the first change of direction to the reservoir
For the longitudinal breaks, there are two intermediate periods corresponding to the force time curve for each direction. The force time curve for each direction has been superimposed thus giving the overall force time
history for this type of break.
UFSAR/DAEC - 1 3.6-34 Revision 23 - 5/15 3. Steady-state period takes place at the end of the intermediate period. For frictionless steady-state flow the rupture force is given by
F 3 = 1.26 F l for saturated steam (3.6-5) = 1.26 F 1 for saturated water
To account for the frictional losses and the presence of flow restrictions in the pipe, the coefficients in the above equations were reduced using the
procedure described in Reference 10.
The detailed calculation of the rupture force time history for various break locations is summarized in Tables 3.6-13 through 3.6-20.
3.6.3.4 Analytical Procedure
3.6.3.4.1 Description of the Pipe Rupture Phenomena
The piping system responds to the rupture by moving in the direction of the rupture force. The pipe and restraint system will undergo elastic and/or plastic deformation until either the input energy due to the time-dependent force equals the energy absorbed by the restraint system or the system fails. The elastic and/or plastic deformation of the pipe and restraint system varies in degree, depending on the geometry (spacing of the restraints, clearances, etc.) and the load- or energy-carrying capacity of the system.
3.6.3.4.2 Mathematical Model
- 1. Lumped Parameter Idealization
The continuous piping system was idealized as an assembly of arbitrary segments that are assumed to be rigid. The mass and mass moment of inertia of each segment were lumped at the center of gravity of each segment while the elastic properties of segments were concentrated at points between segments. Load deformation characteristics of the piping system were
represented by the shear and bending springs. The stiffness properties of these springs are dependent on the effective shear area and moment of inertia of the
pipe.
- 2. Boundary Conditions and External Restraints
The stiffness characteristics of the springs at the ends of the model represent the flexibility of the remainder of the piping system. The stiffness of piping elbows and branch connections was modified to account for local deformation effects by the flexibility factors suggested in Reference 11. The
pipe rupture restraints were represented by their load deflection characteristics. The mathematical models of the pipe-restraint system for UFSAR/DAEC - 1 3.6-35 Revision 23 - 5/15 the various types of breaks analyzed are shown on Figures 3.6-72 through 3.6-79.
3.6.3.4.3 Mathematical Formulation of the Analysis
Considering the translation and rotational degrees of freedom of the nonlinear spring-mass system and assuming a viscous form of damping (velocity proportional), the equation of equilibrium is expressed in matrix form as follows:
MX + C T DCX + C T[f(CX)] = F (t) (3.6-6) where M = diagonal mass matrix
X = absolute coordinate of mass m i C = rectangular matrix relating relative (spring) coordinate to absolute coordinate
D = diagonal matrix of damping coefficients
f(CX) = force-deflection relationships for springs
F (t) = forcing function
The equations of motion were integrated using Newmark's "BETA" parameter method 12. In this step-by-step integration method, the value of
= 0.25 was selected. This corresponds to a uniform value of acceleration during the time interval to the mean of the initial and final values of acceleration. Time intervals were determined using a test on the number of significant figures desired for accuracy in the iterative procedure that converges on acceleration. The above-mentioned calculations were carried out using the NONLIN computer program
- 13. The output of the analysis consists of displacements, velocities, accelerations, and spring forces as a function of time. The envelopes of these quantities were summarized at the end of the output. From the displacement time histories, the time histories of the maximum strains in the pipe were calculated.
3.6.3.4.4 Load Deflection Characteristics for Nonlinear Springs
3.6.3.4.4.1 Piping. The load deflection characteristics, for the lumped bending and shear springs representing the pipe, were calculated using the idealized stress-strain curves for the pipe material and assumed stress distribution across the pipe cross section as shown in
Figures 3.6-80 and 3.6-81. In developing these load deflection characteristics, the pipe cross section was assumed to remain circular.
- 1. Bending Spring
The moment carrying capacity of the pipe, at the stress level corresponding to the yield and ultimate strain, is given by UFSAR/DAEC - 1 3.6-36 Revision 23 - 5/15 M p = y z p (3.6-7)
M1.Ou = S[(Z p - 1) y + 1.0u] (3.6-8) S
where
M p = plastic moment carrying capacity the pipe
M1.0u = moment carrying capacity of the pipe at ultimate strain
y = stress at yield
1.0u = stress corresponding to ultimate strain
Z p = plastic section modulus of the pipe = 4r 2 t S = elastic section modulus of the pipe = r 2 t The rotation of the bending spring for the moments calculated above is given by p = M p (3.6-9) K 1.0u = (M1.0u - M p) L + p (3.6-10) E st I where p = rotation of the bending spring at yield 1.0u = rotation of bending spring at ultimate strain K = elastic stiffness of bending spring
EI L
E = elastic modulus
L = distance between two mass points
I = moment of inertia of the pipe UFSAR/DAEC - 1 3.6-37 Revision 23 - 5/15 E st = strain hardening modulus
- 2. Shear Spring
The load carrying capacity of the shear spring for the shear stress distribution as shown in Figure 3.6-81 is given by
F y = yA/2 (3.6-11)
F1.0u = 1.0uA/2 (3.6-12)
where
F y = shear force carrying capacity of the pipe at yield
F1.0 u = shear force carrying capacity of the pipe at ultimate strain A = effective shear area of the pipe cross section
The deflection of the shear spring, corresponding to its load carrying capacity as computed by equations 3.6-11 and 3.6-12, is given by y = F y (3.6-13) Ks 1.0u = 1.0u y (3.6-14) y where
K s = elastic stiffness of shear spring = AG L
G = shear modulus
y = deflection of the shear spring at yield 1.0u = deflection of the shear spring at ultimate strain u = ultimate strain value
UFSAR/DAEC - 1 3.6-38 Revision 23 - 5/15 3.6.3.4.4.2 Restraint System
- 1. Bechtel Restraint Design
The load deflection characteristics for the typical restraint design were calculated by first idealizing the restraint system as an equivalent frame structure and then performing elastic and plastic analyses of this equivalent frame. The elastic stiffness values, used in calculating the load deflection
characteristics for the longitudinal and lateral direction, were furnished by
Bechtel.
- 2. General Electric Restraint Design
The movement of pipe in a radial direction away from the reactor vessel is resisted by the cable. The load deflection characteristics of the restraint for this direction of pipe movement were calculated by idealizing the restraint system as a series of springs, representing the cable, cable anchor plate, restraint structure (GE), and the support structure (Bechtel), through which the
restraint is anchored to the shield wall. As the load is increased, each spring will undergo elastic-plastic deformation to a different degree depending on its
own load deflection property. This interaction effect between various springs was considered in calculating the total load deflection characteristics for the
restraint.
The movement of pipe in the tangential direction with respect to the reactor vessel is resisted by the structural steel bracket. The effects of combined shear
and bending were included in calculating the load deflection characteristics of the restraint for this direction of pipe movement.
The typical load deflection curves for the pipe and restraints are shown in
Figures 3.6-82 through 3.6-91. The clearance between pipe and restraint is included in the load deflection curves for the restraints.
3.6.3.5 Discussion of Results
The maximum response results of the pipe and the restraint system are summarized
on Tables 3.6-21 through 3.6-30. The load carrying capacity of the pipe and restraint at different strain levels is noted for comparison purposes.
3.6.3.5.1 Bechtel Restraints
The time histories of maximum strain in the pipe, maximum reaction force in the restraint, and the envelope of maximum deflections for the circumferential and longitudinal breaks of the main steam line are shown in Figures 3.6-92 through 3.6-103. The cyclic high-
frequency nature of the elastic response should be noted in Figures 3.6-98, 3.6-99, 3.6-101, and 3.6-102. In addition, in comparing Figures 3.6-66 and 3.6-102, it should be UFSAR/DAEC - 1 3.6-39 Revision 23 - 5/15 noted that the maximum reaction force on the restraint is 2.7 times the rupture force. This magnification factor accounts for the impact effect on the elastic system due to the clearance between the pipe and restraint system. The summary of the response results as shown on Tables 3.6-21 through 3.6-24 indicates that the maximum strain in the pipe and restraint is within 50% of their ultimate values.
3.6.3.5.2 General Electric Restraints
- 1. Recirculation Suction Line
The time histories of the pipe and restraint system response for the circumferential and longitudinal breaks of the recirculation suction line are
shown in Figures 3.6-104 through 3.6-112. The response results are summarized on Tables 3.6-25 through 3.6-27. It can be seen that their maximum values are within the load carrying capacity of the pipe and the restraint system.
- 2. Recirculation Discharge Line
The response results for the recirculation discharge line are shown in graphic form in Figures 3.6-113 and 3.6-114 and are summarized in tabular form on Table 3.6-28. Although the maximum strain in the pipe approaches the ultimate strain for the pipe material, the movement of the pipe will not degrade the containment integrity or the adequacy of the emergency core cooling system due to the inherent distance of the pipe's travel path from these systems.
- 3. Recirculation Riser
The time histories of the response results for the recirculation riser are shown in Figures 3.6-115 and 3.6-116. The summary of the results is shown on Table 3.6-29. Although the maximum strain in the pipe approaches the ultimate strain value for the pipe material, the movement of the pipe will not degrade the containment integrity.
- 4. Recirculation Header
The response results for the recirculation header are shown in graphic form in Figures 3.6-117 through 3.6-119 and are summarized in tabular form on Table 3.6-30. It can be seen that the maximum strain in the pipe and the restraint system is within 50% of their ultimate values.
UFSAR/DAEC - 1 3.6-40 Revision 23 - 5/15 REFERENCES FOR SECTION 3.6
- 1. C.V. Moore, "The Design of Barricades for Hazardous Pressure Systems,"
Nuclear Engineering and Design, Vol. 5, pp. 81-97, 1967.
- 2. A.L. Gluckman, "Some Notes on Dynamic Structural Problems in the Design of Nuclear BWR Stations," Nucl ear Structural Engineering, Vol. 2, 1965.
- 3. F. J. Moody, "Maximum Flow Rate of a Single Component, Two-Phase Mixture," ASME Paper No. 64-HT-35.
- 4. Not Used.
- 5. General Electric Design Report 22A4046, Recirculation System Pipe Whip Restraint for BWR 4, 218 and 215 Mark I and Mark II Product Line Plant, General Electric Document No. 22A4046.
- 6. N. R. Rao, M. Lohrmann, and L. Tall "Effects of Strain Rate on Yield Stress of Structural Steels," Journal of Materials, 1966.
- 7. N. M. Newmark, Design of Structures for Dynamic Loads Including the Effects of Vibration and Ground Shock, Bechtel Corporation Lectures, 1969.
- 8. N. M. Newmark, Seismic Design Criteria for Nuclear Reactor Facilities, Fourth World Conference on Earthquake Engineering, Chile, 1969.
- 9. General Electric Company, Systems Criteria and Application for Protection Against the Dynamic Effects of Pipe Break, Document No.
22A2625, 1972.
- 10. F. J. Moody, Prediction of Blowdown Thrust and Jet Forces, ASME Paper No. 69HT-31, 1969.
- 11. American Society of Mechanical Engineers, "Nuclear Power Piping Code," ANSI B31.7, 1969.
- 12. N. M. Newmark, "A Method of Computation for Structural Dynamics,"
Journal of Engineering Mechanics Division, ASCE, 1959.
- 13. Nuclear Services Corporation, "NONLIN" Computer Code for Nonlinear Dynamic Analysis, 1972.
- 14. Generic Letter 87-11, "Relaxation in Arbitrary Intermediate Pipe Rupture Requirements," June 19,1987.
UFSAR/DAEC - 1 T3.6-1 Revision 12 - 10/95 Table 3.6-1
UFSAR/DAEC - 1 T3.6-2 Revision 12 - 10/95 Table 3.6-1 A A
UFSAR/DAEC - 1 T3.6-3 Revision 12 - 10/95 Key to Table 3.6-1
UFSAR/DAEC - 1 T3.6-4 Revision 12 - 10/95 Notes to Table 3.6-1
UFSAR/DAEC - 1 T3.6-5 Revision 12 - 10/95 Table 3.6-2 PRINCIPAL LINES PENETRATING THE REACTOR SHIELD Size (in.) Number System Reactor Flow Direction 22 2 Recirculation Out 10 8 Recirculation In 10 4 Feedwater In 8 2 Core spray None a 2-1/2 1 CRD hydraulic None a 2 5 Instrumentation None a Note: The jet pump instrument penetration was omitted from this listing since it does not fit the nozzle line definitions considered here. In any case, it represents no significant source for jet force. a Normally no flow.
UFSAR/DAEC - 1 T3.6-6 Revision 17 - 10/03 Table 3.6-3
SUMMARY
OF PRESSURES ON SACRIFICIAL SHIELD PLUGS Penetration Static Pressure Jet Pressure Total Pressure Design Pressure Recirc. outlet N1 18.98 0.0 18.98 20 psi Recirc. inlet N2 3.8 56.4 60.2 70 psi Inst. N16 0.0 4.6 4.5 20 psi a a N1 failure governs.
UFSAR/DAEC - 1 T3.6-7 Revision 12 - 10/95 Table 3.6-4 PIPE DATA: SECTION PROPERTIES
Description
Schedule Size O. D. (in.) Wall Thickness (in.) Weight of Pipe (lb/ft) Weight of Water (lb/ft) Weight of Insulation (lb/ft) Total Weight (lb/ft) Main steam
line 80 20 1.031 209 -- 24 233 Recirculation
suction line -- 22 0.903 204 138.5 26 368.5 Recirculation
discharge line -- 22 1.038 232 135.7 26 393.7 Recirculation
riser -- 10.75 0.542 59.1 31.8 15.3 106.2 Recirculation
header -- 16 0.771 125.4 71.1 20.2 216.7 PIPE DATA: MECHANICAL AND PHYSICAL PROPERTIES
Description Normal Operating Pressure (psig) Temp. (F) Matl. ASTM Spec. MOD. OF ELAST. E. COLD E. HOT (10 6 psi) (10 6 psi) Yield Stress at Oper. Temp. u (ksi) Ultimate Stress 1.0u (ksi) Main steam line 1020 562 A155 KCF 70 27.9 26 29.45 47.0 Recirculation
suction line 1032 532 A358 TP 304 28.3 25.8 18.8 63.0 Recirculation
discharge line 1200 532 A358 TP 304 28.3 25.8 18.8 63.0 Recirculation
riser 1200 535 A358 TP 304 28.3 25.8 18.8 63.0 Recirculation
header 1200 535 A358 TP 304 28.3 25.8 18.8 63.0 UFSAR/DAEC - 1 T3.6-8 Revision 12 - 10/95 Table 3.6-5 Sheet 1 of 2 MAIN STEAM LINE MATHEMATICAL MODELS (M1), (M3), AND (M4) MASS AND STIFFNESS PROPERTIES Spring Constants Mass No. Spring No. Elevation (ft) Mass (lb. sec.2) in. Mass Moment of Inertia (lb sec 2 in.) Lb x 10 6 in. Lb in. x 10 6 rad 1 1.86 2 108.5 1 13.28 2 2252.3 3 3.12 4 659.0 3 8.67 4 1470.9 5 2.71 6 793.4 5 7.46 6 1264.0 7 3.02 8 1056.0 7 7.08 8 1201.0 9 3.02 10 1056.0 9 6.83 10 1158.0 11 3.24 12 1285.0 11 6.59 12 1117.0 13 3.24 14 1285.0 13 7.42 14 1258.9 UFSAR/DAEC - 1 T3.6-9 Revision 12 - 10/95 Table 3.6-5 Sheet 2 of 2 MAIN STEAM LINE MATHEMATICAL MODELS (M1), (M3), AND (M4) MASS AND STIFFNESS PROPERTIES Spring Constants Mass No. Spring No. Elevation (ft) Mass (lb. sec.2) in. Mass Moment of Inertia (lb sec 2 in.) Lb x 10 6 in. Lb in. x 10 6 rad 15 2.51 16 648.0 15 8.95 16 1517.0 17 2.26 18 494.4 17 9.44 18 1601.0 19 2.26 20 494.4 19 8.09 20 1372.8 21 3.02 22 1057.0 21 9.44 22 1601.6 23 2.51 24 189.0 23 0.002 24 0.086 30 33.3 M3 31 231.0 M3 30 1.96 M4 31 1201.2 M4 UFSAR/DAEC - 1 T3.6-10 Revision 12 - 10/95 Table 3.6-6 Sheet 1 of 2 MAIN STEAM LINE MATHEMATICAL MODELS (M2) MASS AND STIFFNESS PROPERTIES Spring Constants Mass No. Spring No. Elevation (ft) Mass (lb. sec.2) in. Mass Moment of Inertia (lb sec 2 in.) Lb x 10 6 in. Lb in. x 10 6 rad 1 2.51 2 189.0 1 9.44 2 1601.6 3 3.02 4 1057.0 3 8.09 4 1372.8 5 2.26 6 4.94 5 9.44 6 1601.6 7 2.26 8 494.4 7 8.95 8 1517.0 9 2.51 10 648.0 9 7.42 10 1258.9 11 3.24 12 1285.0 11 7.59 12 1117.0 13 3.24 14 1285.0 UFSAR/DAEC - 1 T3.6-11 Revision 12 - 10/95 Table 3.6-6 Sheet 2 of 2 MAIN STEAM LINE MATHEMATICAL MODELS (M2) MASS AND STIFFNESS PROPERTIES Spring Constants Mass No. Spring No. Elevation (ft) Mass (lb. sec.2) in. Mass Moment of Inertia (lb sec 2 in.) Lb x 10 6 in. Lb in. x 10 6 rad 13 6.83 14 1158.0 15 3.02
16 1056.0 15 7.08 16 1201.0 17 3.02
18 1056.0 17 7.46 18 1264.0 19 2.71
20 793.4 19 8.67 20 1470.9 21 3.12
22 659.0 21 13.28 22 2252.3 23 1.86
24 108.5 23 1.96 24 1201.2 UFSAR/DAEC - 1 T3.6-12 Revision 12 - 10/95 Table 3.6-7 RECIRCULATION SUCTION LINE MATHEMATICAL MODEL (R1) MASS AND STIFFNESS PROPERTIES Spring Constants Mass No. Spring No. Elevation (ft) Mass (lb. sec.2) in. Mass Moment of Inertia (lb sec 2 in.) Lb x 10 6 in. Lb in. x 10 6 rad 1 6.32 2 825.40 1 14.9 2 3121.9 3 3.93 4 513.30 3 13.7 4 2861.7 5 2.387 6 311.70 5 7.68 6 1604.7 7 5.728 8 2792.80 7 6.23 8 1300.87 9 4.773 10 1696.90 9 6.85 10 1460.87 11 4.773 12 1696.90 11 6.85 12 1430.87 13 4.773 14 1696.90 13 6.85 14 1430.87 15 4.773 16 1696.90 15 6.85 16 1430.87 17 4.773 18 1696.90 17 8.74 18 1826.6 19 2.386 20 311.60 19 0.067 20 295.0 UFSAR/DAEC - 1 T3.6-13 Revision 12 - 10/95 Table 3.6-8 RECIRCULATION SUCTION LINE MATHEMATICAL MODEL (R2) MASS AND STIFFNESS PROPERTIES Spring Constants Mass No. Spring No. Elevation (ft) Mass (lb. sec.2) in. Mass Moment of Inertia (lb sec 2 in.) Lb x 10 6 in. Lb in. x 10 6 rad 1 6.32 2 825.40 1 14.9 2 3121.9 3 3.93 4 513.30 3 13.7 4 2861.7 5 2.387 6 311.70 5 7.68 6 1604.7 7 5.728 8 2792.80 7 6.23 8 1300.87 9 4.773 10 1696.90 9 6.85 10 1430.87 11 4.773 12 1696.90 11 6.85 12 1430.87 13 4.773 14 1696.90 13 6.85 14 1430.87 15 4.773 16 1696.90 15 6.85 16 1430.87 17 4.773 18 1696.90 17 8.74 18 1826.6 19 2.386 20 311.30 19 0.067 20 295.0 25 3.5 26 508.0 UFSAR/DAEC - 1 T3.6-14 Revision 12 - 10/95 Table 3.6-9 RECIRCULATION SUCTION LINE MATHEMATICAL MODEL (R3) MASS AND STIFFNESS PROPERTIES Spring Constants Mass No. Spring No. Elevation (ft) Mass (lb. sec.2) in. Mass Moment of Inertia (lb sec 2 in.) Lb x 10 6 in. Lb in. x 10 6 rad 1 6.32 2 825.40 1 14.9 2 3121.9 3 3.93 4 513.30 3 13.7 4 2861.7 5 2.387 6 311.70 5 7.68 6 1604.7 7 5.728 8 2792.80 7 6.23 8 1300.87 9 4.773 10 1696.90 9 6.85 10 1430.87 11 4.773 12 1696.90 11 6.85 12 1430.87 13 4.773 14 1696.90 13 6.85 14 1430.87 15 4.773 16 1696.90 15 6.85 16 1430.87 17 4.773 18 1696.90 17 8.74 18 1826.6 19 2.386 20 311.60 19 0.067 20 295.0 25 0.59 26 508.0 UFSAR/DAEC - 1 T3.6-15 Revision 12 - 10/95 Table 3.6-10 RECIRCULATION SUCTION LINE MATHEMATICAL MODEL (R4) MASS AND STIFFNESS PROPERTIES Spring Constants Mass No. Spring No. Elevation (ft) Mass (lb. sec.2) in. Mass Moment of Inertia (lb sec 2 in.) Lb x 10 6 in. Lb in. x 10 6 rad 1 4.96 2 647.80 1 15.65 2 3227.2 3 2.55 4 333.03 3 15.65 4 3227.2 5 2.55 6 333.03 5 15.65 6 3227.2 7 2.55 8 333.03 7 10.47 8 2151.5 9 5.10 10 1813.56 9 7.82 10 1613.6 11 5.10 12 1813.56 11 8.53 12 1760.0 13 4.25 14 1122.0 13 0.013 14 2120.0 UFSAR/DAEC - 1 T3.6-16 Revision 12 - 10/95 Table 3.6-11 RECIRCULATION SUCTION LINE MATHEMATICAL MODEL (R5) MASS AND STIFFNESS PROPERTIES Spring Constants Mass No. Spring No. Elevation (ft) Mass (lb. sec.2) in. Mass Moment of Inertia (lb sec 2 in.) Lb x 10 6 in. Lb in. x 10 6 rad 1 .645 2 67.2 1 3.51 2 230.77 3 .549 4 165.2 3 3.73 4 243.3 5 .549 6 165.2 5 3.73 6 243.3 7 .549 8 165.2 7 3.73 8 243.3 9 .549 10 165.2 9 3.73
10 243.3 11 .549 12 165.2 11 3.73
12 243.3 13 .549 14 165.2 13 3.73
14 243.3 15 .549 16 165.2 15 0.004 16 21.7 UFSAR/DAEC - 1 T3.6-17 Revision 12 - 10/95 Table 3.6-12 RECIRCULATION SUCTION LINE MATHEMATICAL MODEL (R6) MASS AND STIFFNESS PROPERTIES Spring Constants Mass No. Spring No. Elevation (ft) Mass (lb. sec.2) in. Mass Moment of Inertia (lb sec 2 in.) Lb x 10 6 in. Lb in. x 10 6 rad 1 .888 2 52.461 1 10.01 2 1452.20 3 .5614 4 23.015 3 15.85 4 2299.32 5 .5614 6 23.015 5 15.85 6 2299.32 7 .5614 8 23.015 7 15.85 8 2299.32 9 .5614 10 23.015 9 15.85 10 2299.32 11 .4679 12 17,465 11 19.02 12 2759.18 13 .5614 14 23.015 13 15.85 14 2299.32 15 .5614 16 23.015 15 15.85 16 2299.32 17 .5614 18 23.015 17 15.85 18 2299.32 19 .795 20 42.197 19 0.72 20 479.0 UFSAR/DAEC - 1 T3.6-18 Revision 14 - 11/98 Table 3.6-13 MAIN STEAM LINE CIRCUMFERENTIAL BREAK M1 AT REACTOR VESSEL NOZZLE RUPTURE FORCE TIME HISTORY Data: P o = 1020 psi A = 252.7 in. 2 , K = 1.0; C = 1.0; A e = KCA = 252.7 in. 2 A R = Restriction flow area (velocity limiter) = 0.25A L 1 = Distance from break to first elbow = 4.25 ft L 2 = Twice the distance from break to first change of direction (at velocity limiter) = 174 ft
Assumptions: 1. Neglect frictional losses
- 2. Choking occurs at the velocity limiter
- a. Initial Thrust
F 1 = 1.0 P o A e = 1020 x 252.7
= 258 kips 1000
(Table 11-1, Reference 9)
t 1 = L 1 = 4.25 = 0.0027 sec V 1600
- b. Intermediate Thrust
F 2 = 0.7 P o A e = 0.7 x 2.58 = 180 kips (Figure 11-6B, Reference 9)
t 2 = L 2 + t 1 = 174 + 0.0027 = 0.11 sec V 1600
- c. Steady-state Thrust
F 3 = 0.4 P o A e = 0.4 x 258 = 103 kips (Figure 11-6B, Reference 9) UFSAR/DAEC - 1 T3.6-19 Revision 12 - 10/95 Table 3.6-14 MAIN STEAM LINE CIRCUMFERENTIAL BREAK M2 AT HEADER RUPTURE FORCE TIME HISTORY Data: P o = 1020 psi A = 252.7 in. 2; K = 1.0; A e = KCA = 252.7 in. 2 L 1 = Distance from break to first elbow = 3.0 ft L 2 = Twice the distance form break to first change of direction (at reactor vessel) = 100 ft
- a. Initial Thrust
F 1 = 1.0 P o A e = 1.0 x 1020 x 252.7
= 258 kips 1000
(Table 11-1, Reference 9)
t 1 = 3.0 = 0.00188 sec 1600
- b. Intermediate Thrust
F 2 = 0.7P o A e = 0.7 x 258 = 180 kips (Figure 11-6B, Reference 9)
t 2 = L 2 + t 1 = 100 + 0.0018 = 0.0625 sec V 1600
- c. Steady-State Thrust
Friction losses:
Pipe (50 ft length) L
= 33 D
Three elbows (3 x 20) = 60
L = 99 D
FL = 0.012 x 99 = 1.2 D
F 3 = 160,000 x 252.7 = 281 kips (Figure 11-4, Reference 9) 144 UFSAR/DAEC - 1 T3.6-20 Revision 14 - 11/98 Table 3.6-15 Sheet 1 of 2 MAIN STEAM LINE LONGITUDINAL BREAK M3 & M4 RUPTURE FORCE TIME HISTORY Data: P o = 1020 psi
A = 252.7 in. 2; K = 2.0 ; C = 0.6; A e = KCA = 303 in. 2 A R = Restriction flow area (velocity limiter) = 0.25 A
L 1 = Twice the break length = 2 x 5 = 10 ft
L 21 = Twice the distance form break to reactor vessel = 2 x 20 = 40 ft
L 22 = Twice the distance form break to velocity limiter = 134 ft
Assumptions: 1. Neglect frictional losses
- 2. Choking occurs at the velocity limiter
- a. Initial Thrust
F 1 = P o A e = 1020 x 303
= 309 kips 1000
t 1 = 10 = 0.0063 sec 1600
- b. Intermediate Thrust
First intermediate thrust from reactor vessel side
F 21 = 0.7 P o A e = 0.7 x 309 = 216 kips (Figure 11-6B, Reference 9)
t 21 = L 21 + t 1 = 40 + 0.0063 = 0.031 sec V 1600
Second intermediate thrust from reactor vessel side and velocity limiter side
F 22 = 1/2(1.26 P o A e + 0.7 P o A e) (Table 11-2 and Figure 11-6B, Reference 9)
F 22 = 1/2 1.96 x 1020 x 303
= 303 kips 1000
t 22 = L 22 + t 1 = 134 + 0.0063 = 0.09 sec V 1600 UFSAR/DAEC - 1 T3.6-21 Revision 12 - 10/95 Table 3.6-15 Sheet 2 of 2 MAIN STEAM LINE LONGITUDINAL BREAK M3 & M4 RUPTURE FORCE TIME HISTORY
- c. Steady-State Thrust
Combination of steady-state thrust from reactor vessel side and velocity limiter side
F 3 = 1/2(1.26 P o A e + 0.4 P o A e) (Table 11-2 and Figure 11-6B, Reference 9)
= 1/2 1.66 x 1020 x 303 = 257 kips 1000 UFSAR/DAEC - 1 T3.6-22 Revision 12 - 10/95 Table 3.6-16 RECIRCULATION SUCTION LINE CIRCUMFERENTAIL BREAK R1 AT REACTOR VESSEL NOZZLE RUPTURE FORCE TIME HISTORY Data: Po = 1032 psi; T = 532F P sat = 900 psi; h = 526; h sat = 547; h sub = 21 A = 320 in.
2; K = 1.0; C= 1.0; A e = KCA = 320 in. 2 A R = Restriction flow area (jet pumps) = 55.4 in. 2 L 1 = Distance from break to first elbow = 6 ft
L 2 = Twice the distance from break to first change of direction (at jet pumps) = 212 ft
Assumptions: 1. Saturated water blowdown since h sub < 22 (page 11-7, Reference 9) 2. Choking occurs at jet pump
- 3. Neglect frictional losses in the line
- 4. Neglect losses in recirculation pump
- a. Initial Thrust
F 1 = 1.0 P o A e = 1032 x 320
= 330 kips (Table 11-1, Reference 9) 1000
t 1 = L 1 = 6 = 0.0015 sec V 4000
- b. Intermediate Thrust
F 2 = P sat F 1 = 900 x 330 = 288 kips (Table 11-1, Reference 9) P o 1032
t 2 = L 2 + t 1 = 212 + 0.0015 = 0.0545 sec V 4000
- c. Steady-State Thrust
F 3 = 0.72 P o A e = 0.72 x 1032 x 320
= 238 kips (Figure 11-5, Reference 9) 1000 UFSAR/DAEC - 1 T3.6-23 Revision 12 - 10/95 Table 3.6-17 Sheet 1 of 2 RECIRCULATION SUCTION LINE LONGITUDINAL BREAK R2 AND R3 RUPTURE FORCE TIME HISTORY Data: Po = 1032 psi; T = 532F P sat = 900 psi; h = 526 Btu/lb; h sat = 547; h sub = 21 A = 320 in.
2 , K = 2.0; C = 0.6; A e = KCA = 384 in. 2 A R = Restriction flow area (jet pumps) = 55.4 in. 2 L 1 = Twice the break length = 10 ft
L 21 = Twice the distance from break to reactor vessel = 40 ft
L 22 = Twice the distance from brake to jet pump = 180 ft
Assumptions: 1. Saturated water blowdown since h sub < 22 2. Choking occurs at jet pump
- 3. Neglect frictional losses in the line
- 4. Neglect losses in recirculation pump
- a. Initial Thrust
F 1 = P o A e 1032 x 384
= 396 kips (Table 11-1, Reference 9) 1000
t 1 = L 1 = 10 = 0.0025 sec V 4000
- b. Intermediate Thrust
First intermediate thrust
F 21 = P sa F 1 (Table 11-1, Reference 9) P o
= 900 x 396 = 346 kips 1032
t 21 = L 21 + t 1 = 40 + 0.0025 = 0.0125 sec V 4000
UFSAR/DAEC - 1 T3.6-24 Revision 12 - 10/95 Table 3.6-17 Sheet 2 of 2 RECIRCULATION SUCTION LINE LONGITUDINAL BREAK R2 AND R3 RUPTURE FORCE TIME HISTORY Second intermediate thrust from reactor vessel nozzle side and jet pump side
F 22 = 1.26 P o A e + P sat F 1 (Table 11-2 and Table 11-1, Reference 9) 2 P o 2
= 1.26 x 1032 x 384 + 900 x 396 2 x 1000 1032 2
= 250 + 173 = 423 kips
t 22 = 180+ 0.0025 = 0.0475 sec 4000
- c. Steady-State Thrust
Combination of thrust from reactor vessel side and jet pump side
F 3 = 1.26 P o A e + 0.72 P o A e (Table 11-2 and Figure 11-5, Reference 9) 2 2
= 1.26 x 1032 x 384 + 0.72 x 1032 x 384 2 x 1000 2 x 1000
= 250 + 142 = 392 kips UFSAR/DAEC - 1 T3.6-25 Revision 14 - 11/98 Table 3.6-18 RECIRCULATION DISCHARGE LINE CIRCUMFERENTIAL BREAK R4 AT DISCHARGE ELBOW RUPTURE FORCE TIME HISTORY Data: P o = 1032 + pressure rises in the pump = 1200 psi
T = 535F P sat = 923 psi, h = 530; h sat = 528; h sub = 2 A = 311.6 in. 2; K = 1.0; C = 1.0; A e = KCA = 311.6 in. 2 A R = Restriction flow area (jet pump) = 55.4 in. 2 L 1 = Distance from break to first elbow = 3 ft
L 2 = Twice the distance form break to first change of direction (at jet pump) = 88 ft
Assumptions: 1. Saturated water blowdown since h sub < 22 (page 11-7, Reference 9)
- 2. Neglect frictional losses in the line
- 3. Choking occurs at jet pumps
- a. Initial Thrust
F 1 = P o A e = 1200 x 311.6
= 374 kips (Table 11-1, Reference 9) 1000
t 1 = 3 = 0.0008 sec 4000
- b. Intermediate Thrust
F 2 = P sat F 1 = 923 x 374 = 287 kips (Table 11-1, Reference 9) P o 1200
t 2 = 88 + 0.0008 sec = 0.0228 seconds 4000
- c. Steady-State Thrust
F 3 = 0.72 F 1 (Figure 11-5, Reference 9)
F 3 = 0.72 x 374 = 269 kips UFSAR/DAEC - 1 T3.6-26 Revision 14 - 11/98 Table 3.6-19 RECIRCULATION RISER CIRCUMFERENTAIL BREAK R5 AT REACTOR VESSEL NOZZLE RUPTURE FORCE TIME HISTORY Data: P o = 1200 psi; T = 532F h = 530; P sat = 923 psi; h sat = 528; h sub = 2 A = 73.34 in. 2; K = 1.0; C = 1.0
A e = KCA = 73.34 in. 2 L 1 = Distance from break to first elbow = 1.5 ft
L 2 = Twice the distance from break to first change of direction (at recirculation header) = 31 ft
Assumptions: 1. Saturated water blowdown
- 2. Neglect frictional losses in the line
- 3. Recirculation header as the reservoir
- a. Initial Thrust
F 1 = P o A e (Table 11-1, Reference 9)
F 1 = 1200 x 73.34
= 88 kips 1000
t 1 = 1.5 = 0.0004 sec 4000
- b. Intermediate Thrust
F 2 = P sat F 1 (Table 11-1, Reference 9) P o
= 923 x 88 = 68 kips 1200
t 2 = 31 + 0.0004 = 0.00815 sec 4000
- c. Steady-State Thrust
F 3 = 1.26 P o A e (Table 11-2, Reference 9) = 1.26 x 1200 x 73.34
= 111 kip 1000 UFSAR/DAEC - 1 T3.6-27 Revision 12 - 10/95 Table 3.6-20 RECIRCULATION HEADER CIRCUMFERENTIAL BREAK R6 AT RISER/HEADER TEE RUPTURE FORCE TIME HISTORY Data: P o = 1200 psi; T = 532F P sat = 923 psi; j = 530; h sat = 528; h sub = 2 A = 73.34 in.
2; K = 1.0;C = 1.0 A e = KCA = 73.34 in. 2 L1 = Distance from break to header
Assumptions: 1. Saturated water blowdown
- 2. Recirculation header as the reservoir
- 3. Neglect intermediate thrust
- a. Initial Thrust
F 1 = P o A e (Table 11-1, Reference 9)
F 1 = 1200 x 73.34
= 88 kips 1000
t 1 = 1.5 = 0.0004 sec 4000
- b. Steady-State Thrust
F 3 = 1.26 P o A e (Table 11-2, Reference 9)
= 1.26 x 1200 x 73.34 = 111 kips 1000 UFSAR/DAEC - 1 T3.6-28 Revision 12 - 10/95 Table 3.6-21 RESTRAINT DESIGN - BECHTEL MAIN STEAM LINE CIRCUMFERENTIAL BREAK (M1) AT REACTOR VESSEL NOZZLE Parameter 0.5-in. Clearance 2% Damping
Pipe Maximum moment (kips in.) 13.5 x 10 3 Percent of moment carrying capacity at yield (M p = 12.6 x 10 3 kips in.) 107% Percent of moment carrying capacity at 1.0u (M1.0u = 17.5 x 10 3 kips in.) 77% Maximum strain level (in./in.) 0.0343
Percent of ultimate strain
(u = 0.18 in./in.) 20% Maximum deflection (in.) 18
Restraint Maximum restraint load (kips) 615
Percent of load carrying capacity of restraint at yield (930 kips) 66% Percent of load carrying capacity of restraint at
1.0u (1190 kips) 52% Maximum strain level (in./in.) 0.0013
Percent of ultimate strain
(u = 0.18 in./in.) 0.72% UFSAR/DAEC - 1 T3.6-29 Revision 12 - 10/95 Table 3.6-22 RESTRAINT DESIGN - BECHTEL MAIN STEAM LINE CIRCUMFERENTIAL BREAK (M2) AT HORIZONTAL HEADER Parameter 0.5-in. Clearance 2% Damping
Pipe Maximum moment (kips in.) 13.3 x 10 3 Percent of moment carrying capacity at yield (M p = 12.6 x 10 3 kips in.) 106% Percent of moment carrying capacity at
1.0u (M 1.0u = 17.5 x 10 3 kips in.) 76% Maximum strain level (in./in.) 0.0213
Percent of ultimate strain
(u = 0.18 in./in.) 12% Maximum deflection (in.) 14
Restrainment
Maximum restraint load (kips) 646
Percent of load carrying capacity of restraint at yield (1150 kips) 56% Percent of load carrying capacity of restraint at
1.0u (1344 kips) 48% Maximum strain level (in./in.) 0.0012
Percent of ultimate strain
(u = 0.18 in./in.) 0.67% UFSAR/DAEC - 1 T3.6-30 Revision 12 - 10/95 Table 3.6-23 RESTRAIN DESIGN - BECHTEL MAIN STEAM LINE LONGITUDINAL BREAK (M3) IN RADIAL DIRECTION Parameter 0.5-in. Clearance 2% Damping
Pipe Maximum moment (kips in.) 12.0 x 10 3 Percent of moment carrying capacity at yield (M p = 12.6 x 10 3 kips in.) 95% Percent of moment carrying capacity at
1.0u (M1.0u = 17.5 x 10 3 kips in.) 69% Maximum strain level (in./in.) 0.0014
Percent of ultimate strain
(u = 0.18 in./in.) 0.8% Maximum deflection (in.) 0.69
Restraint Maximum restraint load (kips) 793
Percent of load carrying capacity of restraint at yield (930 kips) 85% Percent of load carrying capacity of
restraint at 1.0 u (1190 kips) 67% Maximum strain level (in./in.) 0.0017
Percent of ultimate strain
(u = 0.18 in./in.) 0.94% UFSAR/DAEC - 1 T3.6-31 Revision 12 - 10/95 Table 3.6-24 RESTRAINT DESIGN - BECHTEL MAIN STEAM LINE LONGITUDINAL BREAK (M4) IN TANGENTIAL DIRECTION Parameter 0.5-in. Clearance 2% Damping
Pipe Maximum moment (kips in.) 12.0 x 10 3 Percent of moment carrying capacity at yield (M p = 12.6 x 10 3 kips in.) 95% Percent of moment carrying capacity at
1.0u (M 1.0u = 17.5 x 10 3 kips in.) 69% Maximum strain level (in./in.) 0.0009
Percent of ultimate strain
(u = 0.18 in./in.) 0.6% Maximum deflection (in.) 0.72
Restraint Maximum restraint load (kips) 850
Percent of load carrying capacity of restraint at yield (1050 kips) 81% Percent of load carrying capacity of
restraint at 1.0u (1344 kips) 63% Maximum strain level (in./in.) 0.0016
Percent of ultimate strain
(u = 0.18 in./in.) 0.89% UFSAR/DAEC - 1 T3.6-32 Revision 14 - 11/98 Table 3.6-25 RESTRAINT DESIGN - GENERAL ELECTRIC RECIRCULATION SUCTION LINE CIRCUMFERENTIAL BREAK (R1) AT REACTOR VESSEL NOZZLE Parameter 0.9-in. Clearance 2% Damping
Pipe Maximum moment (kips in.) 15.38 x 10 3 Percent of moment carrying capacity at
1.0u (M 1.0u = 24.7 x 10 3 kips in.) 62% Maximum strain level (in./in.) 0.124
Percent of ultimate strain
(u = 0.39 in./in.) 32% Maximum deflection (in.) 20.3
Restraint Cable Maximum restraint load (kips) 554
Maximum load/cable (kips) 148
Percent of breaking strength (159 kips) 93%
Percent of ultimate strength (173 kips) 86%
Cable anchor plate
Maximum load/plate (kips) 148
Percent of load carrying capacity at
1.0u (196 kips) 76% Maximum strain level (in./in.) 0.075
Percent of ultimate strain
(u = 0.18 in./in.) 42% UFSAR/DAEC - 1 T3.6-33 Revision 12 - 10/95 Table 3.6-26 RESTRAINT DESIGN - GENERAL ELECTRIC RECIRCULATION SUCTION LINE LONGITUDINAL BREAK (R2) IN RADIAL DIRECTION Parameter 0.9-in. Clearance 2% Damping
Pipe Maximum moment (kips in.) 8.8 x 10 3 Percent of moment carrying capacity at
1.0u (M1.0u = 24.7 x 10 3 kips in.) 36% Maximum strain level (in./in.) 0.0019
Percent of ultimate strain
(u = 0.39 in./in.) 0.5% Maximum deflection (in.) 1.6
Restraint Cable Maximum restraint load (kips) 536
Maximum load/cable (kips) 143
Percent of breaking strength (159 kips) 90%
Percent of ultimate strength (173 kips) 83%
Cable anchor plate
Maximum load/plate (kips) 143
Percent of load carrying capacity at
1.0u (196 kips) 73% Maximum strain level (in./in.) 0.056
Percent of ultimate strain
(u = 0.18 in./in.) 31% UFSAR/DAEC - 1 T3.6-34 Revision 12 - 10/95 Table 3.6-27 RESTRAINT DESIGN - GENERAL ELECTRIC RECIRCULATION SUCTION LINE LONGITUDINAL BREAK (R3) IN TANGENTIAL DIRECTION Parameter 0.9-in. Clearance 2% Damping
Pipe Maximum moment (kips in.) 11.9 x 10 3 Percent of moment carrying capacity at yield (M p = 8.7 x 10 3 kips in.) 137% Percent of moment carrying capacity at
1.0u (M1.0u = 24.7 x 10 3 kips in.) 48% Maximum strain level (in./in.) 0.00087
Percent of ultimate strain
(u = 0.39 in./in.) 0.2% Maximum deflection (in.) 2.1
Restraint Maximum restraint load (kips) 518
Percent of load carrying capacity of restraint at yield (492 kips) 105% Percent of load carrying capacity of
restraint at 1.0u (917 kips) 56% Maximum strain level (in./in.) 0.044
Percent of ultimate strain
(u = 0.18 in./in.) 24% UFSAR/DAEC - 1 T3.6-35 Revision 12 - 10/95 Table 3.6-28 RESTRAINT DESIGN - GENERAL ELECTRIC RECIRCULATION DISCHARGE LINE CIRCUMFERENTIAL BREAK (R4) AT DISCHARGE ELBOW Parameter 1.2-in. Clearance 2% Damping
Pipe Maximum moment (kips in.) Load carrying capacity of pipe exceeded.a Percent of moment carrying capacity at
1.0u (M1.0u = 28.0 x 10 3 kips in.) a Maximum strain level (in./in.) a Percent of ultimate strain (u = 0.39 in./in.) a Restraint Maximum restraint load (kips) 500 Percent of load carrying capacity of restraint at yield (492 kips) 102% Percent of load carrying capacity of
restraint at 1.0u (917 kips) 54% Maximum strain level (in/in.) 0.031
Percent of ultimate strain
(u = 0.18 in./in.) 17% a Due to the high local strain and collapse of the pipe cross section, as the load carrying capacity of pipe is exceeded, the pipe will hinge around the restraint traveling in a path similar to an arc of a circle with the unattached length acting as the radius. This travel path of the recirculation discharge line will not degrade the containment integrity or the adequacy of the emergency core cooling system. UFSAR/DAEC - 1 T3.6-36 Revision 12 - 10/95 Table 3.6-29 RESTRAINT DESIGN - GENERAL ELECTRIC RECIRCULATION RISER CIRCUMFERENTIAL BREAK (R5) AT REACTOR VESSEL NOZZLE Parameter 0.9-in. Clearance 2% Damping
Pipe Maximum moment (kips in.) Load carrying capacity of pipe exceeded.a Percent of moment carrying capacity at
1.0u (M 1.0u = 3.4 x 10 3 kips in.) a Maximum strain level (in./in.) a Percent of ultimate strain (u = 0.39 in./in.) a Restraint Cable Maximum restraint load (kips) 220
Maximum load/cable (kips) 64
Percent of breaking strength (79 kips) 81%
Percent of ultimate strength (86 kips) 74%
Cable anchor plate
Maximum load/plate (kips) 64
Percent of load carrying capacity
at 1.0u (94 kips) 68% Maximum strain level (in./in.) 0.038
Percent of ultimate strain
(u = 0.18 in./in.) 21% a Due to the high local strain and collapse of the pipe cross section, as the load carrying capacity of pipe is exceeded, the pipe will hinge around the restraint traveling in a path similar to an arc of a circle with the unattached length acting as the radius. This travel pa th of the recirculation risers will not degrade the containment integrity or the adequacy of the emergency core cooling system. UFSAR/DAEC - 1 T3.6-37 Revision 12 - 10/95 Table 3.6-30 RESTRAINT DESIGN - GENERAL ELECTRIC RECIRCULATION HEADER CIRCUMFERENTIAL BREAK (R6) AT RISER/HEADER TEE Parameter 1.2-in. Clearance 2% Damping Pipe Maximum moment (kips in.) 7.7 x 10 3 Percent of moment carrying capacity at yield (M p = 3.9 x 10 3 kips in.) 197% Percent of moment carrying capacity at
1.0u (M 1.0u = 11.0 x 10 3 kips in.) 70% Maximum strain level (in./in.) 0.145
Percent of ultimate strain
(u = 0.18 in./in.) 37% Maximum deflection (in.) 37.8
Restraint Maximum restraint load (kips) 291
Percent of load carrying capacity of restraint at yield (283 kips) 103% Percent of load carrying capacity of
restraint at 1.0u (440 kips) 66% Maximum strain level (in./in.) 0.038
Percent of ultimate strain
(u = 0.18 in./in.) 21%
RESTRAINTDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTCircumferentialPipeRuptureFigure3.6-42' '*RESTRAINTSIDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTLongitudinalPipeRuptureFigure3.6-43 SHEET1FIGURE3,6-44DUANEARNOLDENERGYCENTERNEXTERAENERGYDUANEARNOLD,LLCUPDATEDFINALSAFETYANALYSISREPORTNOTE:MAINSTEAMANDFEEDWATERPIPERESTRAINTSTEAMTUNNEL:,"roRESTRAINTLOCATIONNQTES,I.AiClD1o$lIlH$;/'l:luIIsr",11$(>(;11:<<'VIVA'e""';e..Ilt.!.I'IJ!I.I).$81IIVI1t.CMelAIXlVNO&/.#.0.S.$HINS6#rA::KWI"C05.RE:F£RE:.NCE:.DW'GS."'1';0/11'4I"rMI$$t:CW4t.'?elf'-lr.Aj-/M8p,rPlNG'M/.()/I/t.t.7l1t:.("'.:.:.Ar-,UtJP/PIN4$l£Q7oN4C,erva8J,t':O'-__.-AI-&,S?NOTE:THISREPRODUCIBLEWASMADEFROMDWGC'556u.':nOU.,,,,,M..DETAIL7w.t.EII-z'M**DETAIL2SCAltFO'GUTINSlIl,AnOIl1QJ;.WR:iV--"FORBOX,SEEDETAIL8HIELDToO£tERIlINl:ToSUIT'ASaUILT"COMlllTlQ1I....*.,.,6seAn(2-1"0'Atr'..Z*x2IZ*I!l"ElOX'"Tt-P4COIlt£IiIl;t.'a:wrI!\+}W""'""<"".JDETAILSECH*Hid9ZlliJbWDETAIL8$CAl[rJ;*1:0'-4PIPERESTI<AIN'r.5I?V/.l:./!J'aIfU\llAU:..E"Iv.l'-2SPrI'OPf!",",.0d!llJ!JP':JIPIl44UJ!UYA7I(JItIJt'l'/.t'-I(seEO&tfe*.,&/.BECH-M366REV.2REVISION2008/09 SHEET2-09/07REVISION19FIGURE3.6-44MAINSTEAMANDFEEDWATERPIPERESTRAINTSTEAMTUNNEL,,III!fi'III'1'!/!i..?ti,.:4'---t---I-+"'-I'i1'..r:'>RESTRAINTLOCATIONDUANEARNOLDENERGYCENTERFPLENERGYDUANEARNOLD,LLCUPDATEDFINALSAFETYANALYSISREPORTNore-sl/.4DI)',SHIMS?JI?'Jf(HJ!ViM'AS"""ASGSreeLOitS<:/VlvJ'lUFNI:S,At'WII.05'UIIhWe'i.lI.",A/lQVAlD,(/,,11.0,s*.:wINS'"IJIPrl1CJ:;W/1t..tU.IN'ft.fRE.'-NceOW"s,feM:1'C/(6l,()(J.-/Nf.i!Uf>l'.1t::.:.:f.>tl'i/,ft;UCf/QI.IIMI.O,.,eL.8\f!o::**.***M-251.<'UlCtfR..""A141ft..1P:t;'fIN,""*.*__**AA/./fI}ItI...$I{."'2__._._.._.,,<;:'-<<>doNOTE'THISREPROOUCIBLEWASMADEFROMDWGC-5594',:'!I.'i!()}/2I1tJm,:O,f/UINl1.AIKI-KJ,(:lfO"lufaIh,,'m011IsC;,llI1n:rt,d5t:-CTIONG.E.$IUCOJ./$J!A/J!/t4U4/lO(1N()I'I.'AMr,rYIf=vOOIJ.II.IS",/'/cHI)/(e,Jrlo.c.r.<'1.-4'1"!.'",,.,/1(/f,-fst;!8t..oWOl/rPAIVc-LSel..EV4ilONUW:1/./6WI!'Sr.ft.'/l(>MLFr,-,p{'1NGetlJo.c:IN/.O.-SECTIONS/i-CTION.ft.tV,4'1'JOAI.TY'-'CI..4MPO,f.l'AILEDJf...L-..'NP*{'.I'-o,*OET"'ILHIN.Pl.JH!wlaoe;/.0o.c.,INOrE-,1111:PleUJmIMr,:,/.,t,<,qIN(I.):'ill"&In.r(;'(W(J/f'I()N.i+!I,...'.I'A*X?8<<.""'.i'!IF"Ct;;:-r;;;c,***;*:'."1*I,1N!.4t:1'l>/If-__.____5:.0'1'j0illil..'REV.2./I/r,<'ct.1\<'II,t-lPIPEHEsrllAINrsW4LLIl',l/J.,WALL..flVl'"IJIIf.qPIfH4ND....,1.0.......BECH-M367Il!ll'lk:1I:.....>-"-+-(,,.4'?8*'I!;MJ.!8CJ.rS-4:!/%1"fI!'II'7.""--',,L
",'"--,.,",'7,',0"---OJ."...."';-',,'\7',----------Io,-05'-a"DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationLinePenetrationofSacrificialShieldFigure3.6-47 OPENINGINSAC.SHIELD9"FLOWDISCHARGEAREA=1.61SOFTMAX.DISPLACEDPOSITIONOFPIPEATFAILURER=20"l.R.=10,1511O.R.=11.00"-----'*DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTPipeBreakinReactorShieldRegionFigure3.6-48 '","o,,\\,PVIJ>"'"'"zoCr"',I)','--J'I-L,),-,,"'o"-----.,\)._J,*I:>'DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationInletPenetrationofSacrificialShieldFigure3.6-49 ..-.........-----1-_.---...-_.------DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTU-BoltTypePipeWhipRestraintSuctionLine-SideElevationFigure3.6-50 --lIIItDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYIUPDATEDFINALSAFETYANALYSISREPORTU-BoltTypePipel.JhipSuctionLine-FrontElevationFigure3.6-51
r------.........I_____J*------FDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTU-BoltTypePipeWhipRiserLine-SideElevationFigure3.6-52
...rrra*(11\.<le+/-lli,,)%(DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTU-BoltTypePipeWhipRestraintRiserLine-SectionFigure3.6-53 ,IIIII!r+---i-----'-!---i-----h+-"---i----#l...!---+------o,';u.gDUANEARNOLDENERGY,CENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTI304StainlessSteelRod/ChannelBeamTypePipeWhipRestraint-SideElevationFigure3.6-54 88-<p-_._--II*I:::r.--iI+IIIII_____lOOPBit,'IfIlOOPA*ExistingrestraintsarefabricatedfrombeamsectiQrsonly(e.g.,nocablesused).Therefore,theserestraintswillnotbemodified.**LoopBsameasLoopAunlessnotedotherwise.DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTLocationforNewRestraintstoReplaceExistingCablesFigure3.6-55 REACiORNOZZLE(TYP.)RECIRCULATIONHEADERO._"_"_"/!IO°-.,../'.........../",./"'-..!/.,\.:...."r====iU*1\*,.'/"I'............"/I27':<'-+-:.....x-..:i((0""-":.,-180°\IIJUJ:0::*tlce*&RISERRECIRCULATIONRECIRCULATIONSUCTIONLINERECRDISCHARGELINE*CombinedPrimary+SecondaryStressl.aShIICombinedPrimary+SecondaryStress2.4ShDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTIsometricSketch,RecirculationSuctionandDischargeLineFigure3.6-56 RESTRAINT#26RESTRAINT#27RESTRAINT#28LOOPALOOPBDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRestraintDesign,BechtelCorporation,MainSteamLineIsometricSketchFigure3.6-57 RESTRAINT#21RECIRCULATIONSUCT!ONLINERESTRAINT22RESTRAINT#23*O...-"-"/90*"'"'.:--X*,/.................../'\I'.I**.,X'j./'...............'1"'./'I.............../:21r/**-.._..+-...-,.\cmcr;@INOZZLE'\....J.(TYP,).V>*Ic::I:!3'I.c::.i*VRecirculationDischargeLineRESTRAINT#24DUANEARNOLDENERGYCENTERIESUTILITIES,INC.UPDATEDFINALSAFETYANALYSISREPORTRestraintDesign!GeneralElectricCompany,RecirculationLineIsometricSketchFigure3.6-58Revlslon14-11/98 WQ......J....J:io....Jw-:cV><<,<<Zo-t;WV>\\\iI-Q.."",uSIV-I:cwQ..I-'">-ZI-<<'"Ell....JI.....:>:III/,1,II"j:,II,'"N\DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTMainSteamLineRestraintDesign,BechtelCorporationFigure3.6'-59 ""I"""Z0,I-UL.UV1---1<!l'"NX'"I-V1L.U...J/Xl;5U""'"zoUL.U""ClDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationLineRestraintDesign,GeneralElectricCompanyFigure3.6-60 ...<.D0...'"...<<:<<:0::E::EI--I--V>V><<:<<:!,I0'"0,,I.z'-,,zIJ)zz,0I<<:NZ...-""......-I-a<<:/V>""LO/,I--<<:/V>V>/V>::E/L.LII--/""V>/l-e(\V>/Cl/\\L.LIIN,\...J/<<:0I'\L.LIV>Clc..I\\0<.D\0\x'\.\\en"-N\\\IIL.LI--------'--00I--00000NOL.LI0CO<.D...N.....V>Nu..*u..00ISd£OlNIDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTStressStrainCurvesforStructuralSteelFigure3.6-61 "'oo0:>ooooN.z\\\\\\\\\\0\\,\\\\\u..08"'0::0:wI-0:u..:0:0ffi0000I-""""'"0'"M0"M""0ww....6-------;t------0:>"'DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTStressStrainCurvefor304StainlessSteelFigure3.6-62 8580757065:r:to60zUJr=V>55"":'S50'"co'":345SI-40zUJuffi35,:330252015105o.2.4.6.81.01.21.41.61.82.02.2PERCENTELONGATION(OFORIGINALLENGTH)DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTLoad-ElongationCurvefor6x26GIWRCCableFigure3.6-63 o'"o'"o,,ocoLnN_*.-Vl0Z0U,.......,NVl0Z00LU""l-N00000NIDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTMainSteamLineCircumferentialBreakMlatReactorVesselNozzle,RuptureForceTimeCurveFigure3.6-64 ocoN-o-';o'"-oooCO'"NcoCO-8oV)ozouUJV)Z-DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTMainSteamLineCircumferentialBreakM2atHorizontalHeader,RuptureForceTimeCurveFigure3.6-65 I--....<0'"-M0M-\0'"-0MI---oMoenooVIClZoUu.JMVI\0oZo_ou.J:<:-I-'"oooooSdDINIDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTMainSteamLineLongitudinalBreakM3andM4,RuptureForceTimeCurveFigure3.6-66 roMoeo<teo00V>ClZ0ULUV>00Z00NLU:;:l-eor000Noooo----._.._----------------==....JODUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationSuctionLineCircumferentialBreakRlatReactorVesselNozzle,RuptureForceTimeCurveFigure3.6-67 o-'"oN'"'"o.nooo.n-;::!oo--I-Vlo.n0-NZ000ULLI0VlZ-LLI::E-f-N000-00NIDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationSuctionLineLongitudinalBreakR2andR3,RuptureForceTimeCurveFigure3.6-68 -en<DNoMoMN00V)ClZ0UU.l"V)00NZU.l'"....000000'----.._..."-"'-"---"-'-.._._'--._.-......-....---------_...----_.NINoooooDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationDischargeLineCircumferentialBreakR4atDischargeElbowRupture,ForceTimeCurveFigure3.6-69 cooooco'".,.oooocoCOooo-_._---".-..-.-_____0_;-------------------===-10Sd!>1NI(/)ClZoUW(/)ZDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationRiserCircumferentialBreakR5atReactorVesselNozzle,RuptureForceTimeCurveFigure3.6-70 aaOJOJ--_.._-if)Clza.".uaw.Jaif)azaw.J::E:I-aaaa0Sci])1NI3:JMO.:lDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationHeaderCircumferentialBreakR6atRiser/HeaderTee,RuptureForceTimeCurveFigure3.6-71. FtBENDINGSPRINGSPRINGSREPRESENTINGTHEBOUNDARYCONDITIONSELEVATION825.0'RESTRAINT#26RESTRAINT#27RESTRAINT#28RESTRAINT#29DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTMainSteamLineCircumferentialBreakatReactorVesselNozzle,MathematicalModelMlFigure3.6-72 SPRINGSREPRESENTINGTHEBOUNDARYCONDITIONSBENDINGSPRINGSHEARSPRINGFt---.......1RESTRAINT#26RESTRAINT#27RESTRAINT#28ELEVATION777.5'DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTMainStreamLineCircumferentialBreakatHorizontalHeader.MathematicalModelM2Figure3.6-73 BOUNDARYCONDITIONS31ELEVATION825.0'RESTRAINT#27RESTRAINT#2617233....-1j---lI""'@SHEARSPRINGFt-------l...nBENDINGSPRINGDRYCONDITIONSSPRINGSREPRESENTINGTHEDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTMainSteamLineLongitudinalBreakatMassPoint9,MathematicalModelM3andM4Figure3.6-74 CG9RESTRAINT#23r]7)RESTRAINT#24Pi/\I\J\4iCDFtELEVATION783,5'2CDRESTRAINT#2143CVBENDINGSPRING7RESTRAINT#229SHEARSPRINGBOUNDARYCONDITIONSSPRINGSREPRESENTINGTHEDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationSuctionLineCircumferentialBreakatReactorVesselNozzle,MathematicalModelRlFigure3.6-75 RESTRAINT#240-+--..RESTRAINT#23..-+--.r=v\JV\IiSPRINGSREPRESENTINGTHECDBOUNDARYCONDITIONSELEVATION783.5'210RESTRAINT1/2143(?)BENDINGSPRING7oRESTRAINT1/22Ft@SHEARSPRINGSPRINGSREPRESENTINGTHEBOUNDARYCONDITIONSDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationSuctionLineLongitudinalBreakatMassPoint9,MathematicalModelR2andR3Figure3.6-76 SPRINGSREPREENTINGTHEOUNDARYCONDITIONS@@1411BENDINGSPRING@129SHEARSPRINGCDRESTRAINT#15\............-7C!'G)86G)643r;;:-.G)'-:..'421G)ELEVATION748.75'Ft('-.2DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationDischargeLineCircumferentialBreakatDischargeElbow,MathematicalModelR4Figure3.6-77 FtBENDINGSPRING1RESTRAINT#17SHEARSPRING879-l--e-I1515DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationRiserCircumferentialBreakatReactorVesselNozzle,MathematicalModelR5Figure3.6-78 "'N.-.-z0-t-o.-"V>,>-w-'wW:I:V>9JN'-'....e>..-G'0QN""t-ZCO-'"8'"t-V>W0'"N..-'0co.-'"20-W'"0:I:0.-Nt-V>'"'"ZV>20_20-t-0'"20_Zwt-wV>_rowo"'200.-0wu'",..V>'""''200_z'"=>0.-0V>'"DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationHeaderCircumferentialBreakatRiser/HeaderTee,MathematicalModelR6Figure3.6-79 aa1.Ocu1--*"1IJIIIJJI:;;:I;;;I/r1.Os1JIdealizedStressStrainCurve.OsuPMoment-RotationCurveAssumedBendingStressDistributionDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTLoadDeflectionCurveforNonlinearSpringBendingSpringFigure3.6-80 rJ1.DEY.1.DEYIdealizedStressStrainCurveF1.DEY0y0l.OEYLoad-DeflectionCurveAssumedShearStressDistributionDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTLoadDeflectionCurveforNonlinearSpringShearSpringFigure3.6-81 SHEARSPRINGFORCE(LBS.)F"0.452x106Y-7.88-0.0587.88"-0.830x106DEFLECTION'<5'(INCHES)BENDINGSPRINGMOMENT(LBS-IN)ROTATION(RADIANS)"-12.56x106_6M1.Oeu--17.45x10-0.9620.962M1,Oeu"17.45x106Mp"12.56x106DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTMainSteamLineTypicalLoadDeflectionCurveforPipeFigure3.6-82 F=-0.930x106YII....F--1.189x106:1.0eu-IDEFLECTION'6'(INCHES)8.40.50.582FORCE(LBS)-8.4-.582-.5RADIALDIRECTIONF1.189x1061.0eu=TANGENTIALDIRECTIONFORCE(LBS)F=-1.051x106y=-1.344x1061.Oeu6F10=1.344x10.euF=1.051x106y-5.22-.55-.5.5.555.22DEFLECTION'6'(INCHES)DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTMainSteamLineTypicalLoadDeflectionCurveforRestraintFigure3.6-83 SHEARSPRINGFORCE(LBS)Fl0=1.08x106.*u-8.42-0.040.048.42DEFLECTION'6'(INCHES)Fy=-0.323x10661.O*u=-1.08x10BENDINGSPRINGMOMENT(LBS-IN)Ml.0*u=e4.70x106M=8.69x106P2.3-0.0050.005-2.3......ROTATION(RADIANS)6Mp=-8.69x10M10=-24.70x106.*uDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationSuctionLineTypicalLoadDeflectionCurveforPipeFigure3.6-84 RADIALDIRECTIONFORCE(LBS)0899,9-1.634DEFLECTION'6'1.1912,02(INCRES)6F1,Oeu=0.594x10.fy=0.46x106TANGENTIALDIRECTIONFORCE(LBS)_6F1,OEU-0.917x10Fy=0.492x106-13.7-1.291-1.21.21.29113.7Fy=-0.492x106=-0.917x1061,OEUDEFLECTION'6'(INCHES)DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPOATEOFINALSAFETYANALYSISREPORTRecirculationSuctionLineTypicalLoadDeflectionCurveforRestraintFigure3.6-85 SHEARSPRINGFORCE(LBS)4.5_6Fl.O*U--1.237x10-4.95F=1.237x1061.0EuBENDINGSPRINGMOMENT(LB-IN)1.353Mp=-9.86x106M1.0Eu=-28.03x106-1.53ROTATIONII(RADIANS)M=9.86x106PM=28.03x1061.0,uDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationDischargeLineTypicalLoadDeflectionCurveforPipeFigure3.6-86 -13.76F10=0.917x10**u-1.291-1.2FORCE(LBS)1.21.291F=-0.492x106y._6Fl0--0.917x10**u'6'DEFLECTION(INCHES)DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationDischargeLineTypicalLoadDeflectionCurveforRestraintFigure3.6-87 SHEARSPRINGFORCE(LBS)_6Fl.O£u-0.315x10Fy=0.094x106-9.08-0.02520.02529.08--+-----1---;1---+-----1-......DEFLECTION'6'(INCHES)6Fy=-0.094x10_6Fl.OEu--0.315x10BENDINGSPRINGMOMENT(LBS-IN)-22323--t------ir--f--+-----"-'-.;p:--.-ROTATION(RADIANS)Mp=-1.222x1066M10=-3.469x10.WDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationRiserTypicalLoadDeflectionCurveforPipeFigure3.6-88 FORCE(LBS)'0IF=-0612x106y*-+---+-f----.----*!----.---.-.L---\!--illio....DEFLEcnON0.91.0041.071.5B4(INCHES)-3.504IF=0.22x106y6Fl0=0.274x10**u-4.0IDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationRiserTypicalLoadDeflectionCurveforRestraintFigure3.6-89 SHEARSPRINGFORCE(LBS)6F10=0.668x10*EU-4.5DEFLECTION'6'-f1--o---t-f::-+/-:-;:------:4;-"".r.:-5-+-(INCHES)Fy=-0.199x1066F10=-0.668x10*EUBENDINGSPRINGMOMENT(LBS-IN)0.75M=-3.867x106PM=-10.99x1061.0EU-0.756M10=10.99x10*EUMp=3.867x106-----1----+-1--+---+---111<-ROTATION(RADIANS)DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationHeaderTypicalLoadDeflectionCurveforPipeFigure3.6-90 FORCE(LBS)12.81.273F=-0.283x106Y1.2Fl.OEU=-0.44x106-1.273-1.2-12.8'6'DEFLECTION(INCHES)DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationHeaderTypicalLoadDeflectionCurveforRestraintFigure3.6-91 o'"MogMoo'"NoooNoo'"ogoo'"oo'NI/'NINIWnWIXVW'r=-=="-----------------,DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTMainSteamLineCircumferentialBreakMlatReactorVesselNozzle,MaximumStrainTimeHistoryinPipeFigure3,6-92 V)'"....'"'"->('-'"en-...,...":::>"wi-'"""'"!;;:....zc<""....V)<oJ""u..'">-....ui-c<0-c<U'"Z>-""""c<u'"c<'"....I-i---,IIIII'"'"'"'"'"'"'"'"'"'"'"'">(>(>(>(>(>(""'"....0')N'"0')ex>N""N....NNN'"Nex>.""....N'"'":ex>!i'"U<oJVlZ!;l!.......'"N'"'"'591NI,DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTMainSteamLineCircumferentialBreakf'UatReactorVesselNozzle,MaximumReactionForceTimeHistoryinRestraintNo.26Figure3.6-93 MASSPOINTS823'-10"820'-0"810'-0"9)-----+f800'-0"111---.-.1H13l---f-i790'-0"151---H@---tt19)--tl780'-0"777'-5"EL.770'-0"-2024681012141618DEFLECTIONININCHESDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTMainSteamLineCircumferentialBreakMlatReactorVesselNozzle,EnvelopeofMaximumDeflection"Figure3.6-94 00-oIIS'wIIIIII'"0'"0'"0N0.....'"N0NN000000'NI/'NININIVM1S'XVWJI-0MI-ooNI-<0NI-.".NNN-0N.J-:::----;'!N.]VIClZ0UUJooVI0Z--I-----I-gI---...,III0'"0'".....'"N000Cl00DUANEARNOLDENERGYCENTERIESUTILITIES,INC.UPDATEDFINALSAFETYANALYSISREPORTMainSteamLineCircumferentialBreakM2atHorizontalHeader,MaximumStrainTimeHistoryinPipeFigure3.6-95Revision1411/98 V)""...J'"'"'"'"x'"'"et)N'"H'":>NW'"'"l-N<<I-Z-NNl-V)UJ'"'"N.....'">-I-et)---;u<<"-<<u'"<.0Z-'"'"<<u'"<<N'"...J'"V)'"Z'"""U'"UJV)Z-'"0!;l!-I-'"0N00'"'"-00XX"-'"'"ox'"x'"'"'"'"000xxXMN'581NIDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTMainSteamLineCircumferentialBreakM2atHorizontalHeader,MaximumReactionForceTimeHistoryinRestraintNo.29Figure3.6-96 820.0'810.0'800.0'790.0'780.0'EL.770.0'-2o246810121416DEFLECTIONININCHESDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTMainSteamLineCircumferentialBreakM2atHorizontalHeader,EnvelopeofMaximumDeflectionFigure3.6-97 Vlco-'0Mon0XCON0'"'"N"i;l0...N!;;;N!z:N-I-0VlNUJ'".....0CO>-I--U<l;'"D.<l;U'"z:...-'"<l;UN0-'0.CO0Vl<::>'"z:<::>0.UUJVl...0z:-UJ:E-NI-00onononononononon00000000><><><><X><><><CO.....'"on...MN'S81NIDMO;jNOllJ\i3MDUANEARNOLDENERGYCENTERIQWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTMainSteamLineLongitudinalBreakM3inRadialDirection,MaximumReactionForceTimeHistoryinRestraintNo.27Figure3.6-98 C)00--roNCO...U)NC)*...."NC)...N....N...<c<C)V>NLUCO-J=>U)..."...NC)...COC)V>0U)C)C)OJLUV>"C)...NC)C)*NI/*NINIDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTMainSteamLineLongitudinalBreakM3inRadialDirection,MaximumStrainTimeHistoryinPipeFigure3.6-99 MASSPOINTS823.75'820.0'810.0'800.0'790.0'780.0'EL.&&-.4-.20.2.4.6.81.01.2DEFLECTIONININCHESDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTMainSteamLineLongitudinalBreakM3inRadialDirection,EnvelopeofMaximumDeflectionFigure3.6-100 0M.Z-"-00'"Z-00<D'"0"v'"'"W0'"'"Z-I-0Vl'"UJ00-I-...J::><DV'"0Z000U0UJVlZ-<D0UJ:;;;-l-V0'"00'"00......<D'".,.M'"000000000000000000000000000'NU'NININIlIlllS'X\IWDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTMainSteamLineLongitudinalBreakM4inTangentialDirection,MaximumStrainTimeHistoryinPipeFigure3.6-101 -JU10...0XM<t<tM00"W"'0....<t<<""<<WU-00.>-""00<><<...<<<><0"'2:..."">-c<c<<<<t<>...<><<0-J....0000<>2:0<>W"'02:""W<t""00<><>U1U1U1U1U1U1U1U10CO<><>CO<>CO0...............XXXXXXXXXCOCO"'"'U1..M'salNIDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTMainSteamLineLongitudinalBreakM4inTangentialDirection,MaximumStrainTimeHistoryinRestraintNo.27Figure3.6-102 MASSPOINTS.8.6.4.2o-.4-.2790.0'800.0'810.0'825.0'820.0'780.0'EL.777.5'DEFLECTIONININCHESDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTMainSteamLineLongitudinalBreakM4inTangentialDirection,EnvelopeofMaximumDeflectionFigure3.6-103 C>00"""CO""0-00"',C>*.--.<Tw*C>..-....,""<<C>onLW<<:ECO""..-*oJ=>"'..-*<T..-..-*C>..-*COC>onC>"'C)C>L)LWon<T""C>LW:E""C)*C>'N!/'N!N!'XVWDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationSuctionLineCircumferentialBreakRlatReactorVesselNozzle,MaximumStrainTimeHistoryinPipeFigure3.6-104 Vl'"...JVlo":>wooN<co<toNoo1.0LOInLOLOo0000xxxxX1.0<o::tMN'salNIDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationSuctionLineCircumferentialBreDkR1atReactorVesselNozzle,MaximumReactionForceTimeHistoryinRestraintNo.21Figure3.6-105 780.0'770.0'9@760.0'1315EL.750.0'---,-MASSPOINTS-10-5o5101520DEFLECTIONININCHESDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationSuctionLineCircumferentialBreakRlatReactorVesselNozzle,EnvelopeofMaximumDeflectionFigure3.6-106 0McoN<0N...NNN0Nco<0....N,0co0V")0z0U<0'"0V")z-...'"0::E-....N0,0'"MoIIII0CO<0...N0CO<0...NN555000000000000000000000'NI/'NININllfM1S'XIfWDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationSuctionLineLongitudinalBreakR2inRadialDirection,MaximumStrainTimeHistoryinPipeFigure3,6-107 Vlco...JU">0X<T'"0U">M"::lWCO0N'"Nl-.Z""<T'"l-NVlUJ'"U-N0N>-I-0U""N"-""U'-"COVlZ0Z>-0'"U'"UJ""'"VlUZ0""UJ0...J_<T""l-N0.CO0'"0<T0N00U">U">U">U">U">U">000000XXXXXX'"U"><TMN'S81NIDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationSuctionLineLongitudinalBreakR2inRadialDirection,MaximumReactionForceTimeHistoryinRestraintNo.22Figure3.6-108 780.0'770.0'MASSPOINTS760.0'11EL.750.0'.j-7-'.._.19-0.50.51.52.5-1.001.02.03.0DEFLECTIONININCHESDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationSuctionLineLongitudinalBreakR2inRadialDirection,EnvelopeofMaximumDeflectionFigure3.6-109 --...,.M.-OJZN---Z1-<0-N0>M0<?I-q-"N.::lWI-N0NZ-1-0iiiNl-V>.....I-OJ"-->-I-1-<0....J::::>--::::,..1-;::I-N./1-0-V>I-OJ'"0Z'"U.....<::V>1-<0Z0-I-q--0I-r-N00IIIIIJIJI0>OJ....<0'"q-MN-00'"00000000'"00'"000000'"00000*NI/*NININI\IlJ1S*X\IWDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationSuctionLineLongitudinalBreakR3inTangentialDirection,MaximumStrainTimeHistoryinPipeFigure3.6-110
- v>5InCl'""(;lCl[\IIIIIILOLOLOLOLOLOClClClClClCl><><><><><><'"In....""N'sa1NIDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationSuctionLineLongitudinalBreakR3inTangentialDirection,MaximumReactionForceTimeHistoryinRestraintNo.22Figure3.6-111 780.0'770.0'760.0'EL.750.0'MASSPOINTS-2-1o123DEFLECTIONININCHESDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDSAFETYANALYSISREPORTRecirculationSuctionLineLongitudinalBreakR3inTangentialDirection,EnvelopeofMaximumDeflectionFigure3.6-112 z......z""'"0II"W0z<0::l-V>""'\\\l-I...JI=>u..oI-c..<:3'-"0z""'0>-""'0::""'0::U<Xu""'0,,",<c..o...Jc..o....o'"oN0'"'"N'"N,....N,NN0N,'"'"....,N0,'"0,V>0Z0'"U0""'V>Z....0""':>:l-N00'NI/'NINIXVWDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationDischargeLineCircumferentialBreakR2atDischargeElbow,MaximumStrainTimeHistoryinPipeFigure3,6-113
'"w'"wwuxwwa-a-u..'">-....-gu""a-""u-'"z>-'"-'"""u'"""-'"-'-J>i-2i-'"'"-"'""'-:nTto'InlLnlLn'lL')'o00000xxxxxX\.01O'<:T('t")N*S81NI<0'":...--;:::S'"'"'"'"Z<0'"'"uw'"z...'"w:0;:....N'"'"DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationDischargeLineCircumferentialBreakR4atDischargeElbow,MaximumReactionForceTimeHistoryinRestraintNo.15Figure3.6-114 0'":z.....<X)N:z'"'"'"N011<T::INW*0,NNz:*....0V'lNu.J<X)......J::::l'"<T*N*0*V'l<X)00z:0,Uu.J....V'l....'"....0:zu.J""<T-0....N00....'"Ln<T'"N0000000'NII'NININI\flJ1S'X\fWDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationRiserCircumferentialBreakR5atReactorVesselNozzle,MaximumStrainTimeHistoryinPipeFigure3.6-115 V1co-J'"00'"".,......CONN""'"WN:3t;:.,.N....Z-;:2N....NV1UJ'"w..00N>-....-UCO<<"-<<u'"'"Z->-'"'"<<.,.u0UJ"--JN-"-w..00>-.....-u00:"-ro<50,'"ZZ'0'"u>-*0UJOXV1"'0<5i:!iz-UJ.,.OUJ0LU<<U::E:0><--JUJ....N00'"'"'"'"'"'"'"'"'"'"0000000000""""""""""0ro'".,.N0<Xl'".,.NN'S81NIDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationRiserCircumferentialBreakR5atReactorVesselNozzle,MaximumReactionForceTimeHistoryinRestraintNo.17Figure3.6-116
- z:..........*z:....'"M*o*0MOJ'"'"'"....'"'"'"0'"OJ(I")0Z0UW'"(I")*Z....W....:E....*I-'"0*OJ0<00*....0'"00o'".o'"oo'"ooMoo'NI/'NINIwnwlxvwDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationHeaderCircumferentialBreakR6atRiser/HeaderTee,MaximumStrainTimeHistoryinPipeFigure3.6-117 Vlco....Ja'"MaxcoaN......II<0N"Wq...N!z-N;;'iN....VlUJa'"Nu..a>-....co-u<<<>.<<<0u'"z-i;;;...'"<<uQ<<N0....JaVlQ'"Z0auUJVl<0Za-UJ:E-.......aNaa'"a'"ax...XMXNx'salNI,-------------------,DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationHeaderCircumferentialBreakR6atRiser/HeaderTee,MaximumReactionForceTimeHistoryinRestraintNo.21Figure3.6-118 S3H3NININOI1331d30DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTRecirculationHeaderCircumferentialBreakR6atRiser/HeaderTee,EnvelopeofMaximumDeflectionFigure3.6-119 UFSAR/DAEC-1 3.7-1 Revision 14 - 11/98 3.7 SEISMIC DESIGN
General Electric (GE) and Bechtel Corporation were responsible for the seismic design of all structures, systems and components of the plant that were related to safety.
The GE Atomic Power Equipment Department (APED) organizations who had responsibility for the seismic design of safety-related systems and structures in the NSSS
were Requisition Engineering and Design Engineering. The seismic design responsibility was assigned to the functional groups for component designs or systems designs who are responsible for the equipment and/or structure design. Those functional
groups were responsible to the Manager, Design Engineering.
The Bechtel DAEC project design organization consisting of the Mechanical
Group, Layout Group, Civil Group, and the Electrical Group, in parallel with the Bechtel Power and Industrial Piping Stress Group, had the responsibility for the seismic design of all balance-of-plant (BOP) structures, systems, and components related to safety. Dames and Moore performed the site seismology studies. These studies were reviewed and checked by Bechtel's Soils and Geology Department. Chicago Bridge and Iron performed the primary containment stress analyses. John A. Blume and Associates, Engineers, performed the dynamic analysis of all Seismic Category I structures and other structures housing Seismic Category I equipment.
Design organizations of GE APED were responsible for proper application of seismic design loads and conditions to the design of equipment components and piping in
the NSSS scope. Analytical assistance was available within GE Design Engineering from analytical components specialized in seismic design. An Engineering Practices and Procedures Manual defined explicitly in writing all Design Engineering and Application Engineering responsibilities, including seismic. The Manager, Design Engineering, had overall responsibility for the adequacy of the seismic design of the GE product.
The dynamic analysis of plant structures was performed by John A. Blume Engineering after the location of major component masses and structural concepts was determined by the involved Bechtel project groups. The resulting floor response spectrum curves were submitted to Bechtel for review and approval. On approval the seismic report was transmitted to the GE APED Project Organization and other project
groups within Bechtel through the Civil Group, which coordinated and had overall responsibility for the BOP seismic design. The Civil group also had responsibility for
plant structural design. The Mechanical and Electrical Groups had responsibility for obtaining vendor seismic design analyses or test results of safety-related equipment and instrumentation; the Layout Group provided input on plant piping layout to the Piping Stress Group, which performed piping stress calculations for Seismic Category I piping systems.
The mechanism for the interchange of needed design information and changes thereto and the coordination of the various facets of the seismic design among the involved design organization components a nd/or groups is shown in Figure 3.7-1. UFSAR/DAEC-1 3.7-2 Revision 14 - 11/98 The system shown in Figure 3.7-1 is a pattern of interrelationships and checks from which an interactive process evolved that ensured proper plant seismic design for structures, systems, and components related to safety.
Within GE APED Design Engineering, the design engineer was ultimately responsible for the implementation of the seismic design requirements. Within the
Bechtel DAEC project organization, the engineer responsible for the safety-related equipment, supported by engineers qualified in seismic analysis within the Civil Group and occasionally by qualified engineers of John A. Blume Engineering, was responsible for the implementation of the seismic design requirements.
For GE APED components, the adequacy of seismic design was the responsibility of the individual design engineer. Within Bechtel, the seismic certification of safety-related equipment was the responsibility of the design group procuring the equipment. Within each group, one or more engineers coordinated the transfer of vendor seismic certification (analyses, tests, or documentation of suitable performance in comparable vibrational environments) to the Civil Group for engineering review and approval.
Seismic design requirements were incorporated in the purchase specifications of all safety-related equipment.
For Bechtel-supplied Seismic Category I equipment, the supplier was given three options. First, the supplier determined the fundamental natural frequency: if it was
found to be greater than 33 Hz, he could then use the floor horizontal acceleration values; if found to be less than 33 Hz, he was provided with the appropriate acceleration value from the response curve for 1% critical damping. Second, if the supplier chose not to determine the natural frequency, he was provided and instructed to use the peak value of the 1% critical damping response curve for molded supports, and 2% critical damping
response curves for bolted supports, for the appropriate floor level. In all cases, the vertical seismic component was two-thirds of the foundation level horizontal acceleration for the operating-basis earthquake (OBE) or design-basis earthquake (DBE), applied simultaneously with the horizontal component. Third, the supplier was allowed to carry out a dynamic test.
The vendor was required by the purchase specification to submit the test data and/or the seismic analysis for the buyer's approval as a condition of acceptance of the equipment for the intended function. The vendor could use test data on the particular components or equipment, applicable data from previously tested comparable equipment, performance data from comparable equipment which, during normal operating
conditions, had been subjected to equal or greater shock loadings, and/or suitable
analytical results.
This data was then referred, for NSSS equipment, to the responsible design
engineer within GE APED or, for BOP safety-related equipment, to the Bechtel Civil Group, which reviewed the methods, procedures, and results for compliance with the UFSAR/DAEC-1 3.7-3 Revision 14 - 11/98 criteria. If rejected, the vendor was requested to resubmit an amended report and/or perform structural modifications to the equipment. The whole sequence of submittal and review was repeated, if necessary, to achieve conformance with the design criteria.
3.7.1 SEISMIC INPUT
The foundation level accelerations for the OBE and the DBE are described in Section 2.5.2.
3.7.1.1 Design Response Spectra
In addition to the identification of the maximum foundation level acceleration that might occur at the site, the distribution of frequencies of the foundation level motion was considered. To represent the frequency distribution of foundation level motion resulting from postulated earthquakes, the concept of the response spectrum was used.
A response spectrum is a plot of the maximum value of some parameter, such as acceleration, experienced by a single-degree-of-freedom damped spring mass system subjected to an input motion. The maximum value is expressed as a function of the natural period and damping of the system. The foundation level response spectra for the OBE and DBE are shown in Figure 2.5-8.
3.7.1.2 Design Time History
One earthquake time history was developed for use as the input motion in performing the seismic analysis of structures. The time history was developed so that response spectra generated from it would yi eld no significant deviations below the OBE response spectra for the range of damping considered in the analyses. Comparative plots of the acceleration response spectra generated from the modified earthquake time history and the OBE response spectra are shown for damping ratios of 0.005, 0.010, 0.020, and 0.050 in Figures 3.7-2 through 3.7-5. These comparative plots were presented in the PSAR and accepted by the NRC. This time history represents the site specific (bedrock) earthquake motion and was used as the input for the seismic analysis of structures. The mathematical model for each seismically analyzed structure included the soil-structure interaction effects as described in Section 3.7.2.4. Therefore, the use of one input motion for the seismic analysis of structures is acceptable.
The DBE spectrum is a multiple of the OBE spectrum. Therefore, the DBE values are obtained by multiplying the OBE values by the same multiple, thus ensuring the same match with the site spectrum for the DBE as was obtained for the OBE.
3.7.1.3 Damping Values
UFSAR/DAEC-1 3.7-4 Revision 14 - 11/98 For structures applying to BOP systems comprised of elements possessing different damping characteristics, a composite damping value was selected for each natural mode of vibration. The composite damping values were determined by first plotting and inspecting the mode shapes. Then weighted damping values were determined based on the estimated degree of participation of each element in each mode. These weighted, or composite, damping values were used in the dynamic analysis of the system by the modal superposition method.
For NSSS systems, to account for damping in different elements of a coupled system, a modal damping value for each mode of the coupled system was obtained by using the kinetic energy of the different elements as a weighting function.
Structures were analyzed using modal superposition techniques. Element or material-associated damping values are given in Table 3.7-1. "Composite" or modal damping values in structural systems comprised of different element material types were selected on the basis of an inspection of the significant mode shapes. If a particular mode indicated the response to be of a single-element type, the damping value corresponding to that element type was assigned to that mode. If a particular mode indicated the response to be of several element types, the damping value for that mode was estimated on the basis of the degree of participation of the different elements. In all cases, the damping
values were selected conservatively.
3.7.1.4 Supporting Media for Seismic Category I Structures
The reactor building mat was designed to span a 16-ft hypothetical cavity at a
depth of 2 ft below the rock surface in acco rdance with Section 1.8 of the PSAR, which was submitted with PSAR Amendment 10 on November 26, 1969.
3.7.2 SEISMIC SYSTEM ANALYSIS
3.7.2.1 Seismic Analysis Methods
The design of the Seismic Category I structures, the reactor pressure vessel, and the drywell is based on a dynamic analysis using the time-history method. The design of other Seismic Category I equipment supported within the structures is typically based on a dynamic analysis using the response spectrum method.
The structures were analyzed for the foundation level accelerations described in Section 2.5.2. In the dynamic analyses of Seismic Category I structures and components, equivalent discrete-mass mathematical models were developed to represent the physical characteristics of the structural systems. The models included rocking and translation of the structures on their foundation materials. In the analyses, all modes that contributed
significantly to the total response were used. The earthquake conditions were applied to the structures in the direction of each of their principal axes. Stresses resulting from UFSAR/DAEC-1 3.7-5 Revision 14 - 11/98 horizontal and vertical accelerations were considered to be acting simultaneously and added algebraically. Structures, including the reactor pressure vessel and the drywell, were analyzed by the time-history method. These analyses considered the effects of the dynamic coupling between buildings and equipment. In the time-history method of dynamic analysis, the response of each mass for each individual mode at each increment of time is computed, and the total response for each increment of time is obtained by adding together the response of each mode at a particular instant of time. The maximum response is then the maximum absolute value of the time history of the total response.
Other Seismic Category I mechanical and electrical equipment, and piping supported in or on major structures, were generally analyzed by the response spectrum method of dynamic analysis. The response spectrum method uses the single-mass response spectra. In this method, the maximum response for each quantity for each mode is computed and the modal responses are combined to determine the total response.
3.7.2.1.1 Mathematical Procedures
The equation of motion of a multidegree-of-freedom discrete-mass damped system subjected to an arbitrary ground motion ()vgt , can be written as
()()()()mvtcvtkvtm ovgt++= (3.7-1)
where m = mass matrix c = damping matrix k = stiffness matrix o = unit vector ()vgt = ground acceleration ()()()vtvtvt ,, = matrices of displacement, velocity and acceleration, respectively Using the orthogonality relations and expressing the displacement, velocities and accelerations in normal coordinates (i.e., ()()vtYt, ()()vtYt, and ()()vtYt,), the above couple equations of motion (Equation 3.7-1) may then be rewritten as the following uncoupled, normal equation of motion:
()()()()M r Y r t r M r r Y r tK r Y r t r M rvgt++=2 (3.7-2) where M rr t m = generalized mass for the r th mode UFSAR/DAEC-1 3.7-6 Revision 14 - 11/98 r r t cr r r t m r=2 = damping ratio for the r th mode K r T r K r= = generalized stiffness for the r th mode r r t m o r t m r= = participation factor for the r th mode r = undamped circular frequency of the r th mode r = mode shape matrix for the r th mode r t = transposition of r The undamped circular frequencies, , may be calculated from K r m=20 (3.7-3) and the mode shape matrix for the r th mode can be obtained from []K r m r=20 (3.7-4)
The solution of the differential Equation 3.7-2, for the case of at-rest initial
conditions is
()()()()Y r t r r t v g e rr t r td=1 2 0 1 2sin (3.7-5)
For small damping ratios, r, the above solution may be approximated by
()()()()[]Y r t r r o t v g e rr t r td=sin (3.7-6)
There are basically two methods of dynamic analysis that can be used to solve multidegree-of-freedom problems: the time-history method and the response spectrum method.
3.7.2.1.1.1 Time-History Method of Analysis
UFSAR/DAEC-1 3.7-7 Revision 14 - 11/98 If the ground motion acceleration time history, ()vgt , is known, Equation 3.7-6 can be solved by a numerical step-by-step integration procedure. Y r(t) is computed as a function of time for r=1,2,3 . . . n, where n is the total number of degrees of freedom of the system. The modal displacements, v r(t), at time t for the r th mode, can then be calculated from
()()V r t r Y r t (3.7-7)
The total displacement, v(t), of the structure at any time t may be obtained by adding the individual modal displacements at time t
()()()()VtVtVtV n t=++12...... (3.7-8)
Once the time histories of the displacements have been determined, the time histories of shears and moments can be determined by conventional structural analysis procedures. The maximum values of the time histories of the shears and moments are determined and then used for design.
3.7.2.1.1.2 Response Spectrum Method of Analysis
If the ground motion time history is not available and the design earthquake is specified in terms of a response velocity spectrum, Equation 3.7-6 can be written
()Y r t r S vr r max= (3.7-9) where S v r = spectral velocity for the r th mode. ()()()[]S vr t v g e rr t r td=0sin max (3.7-10)
The maximum modal displacements, V rmax , for the r th mode are v rr r S vr rmax (3.7-11) If the design earthquake is specified in terms of a response acceleration spectrum instead of a velocity spectrum, the maximum modal displacements, V rmax , of the structure for the r th mode are
UFSAR/DAEC-1 3.7-8 Revision 14 - 11/98 v rr r S ar rmax2 (3.7-12) where S a r = spectral acceleration for the r th mode With maximum modal displacements known, the other modal quantities such as shears and moments can be computed for each mode by conventional structural analysis
procedures.
3.7.2.1.2 Example of Building Analysis
Figure 3.7-6 shows the model that was used in the analysis of the reactor building and the major equipment housed within it. The building was represented as a series of masses lumped at the floor levels and connected by weightless elastic elements representing the flexural characteristics of the building. The deformation of the building foundation material was represented by translational and rocking springs. The spring constants were calculated based on formulas published by Whitman and Richart. 1 The drywell, reactor pressure vessel, shield wall, and pedestal were represented by lumped-mass models, and elastic and rigid elements were used to represent the interconnections among the components of the resulting composite structure.
The natural mode shapes and periods of vibration were determined as described in Section 3.7.2.1.1. Damping values were assigned to each mode based on the relative deformations of the different materials within the structure as described in Section 3.7.1.3. For this example, the natural periods of vibration and assigned damping values for the significant modes of the reactor building model are shown in Table 3.7-2..
By applying the appropriate earthquake time history at the base of the model, time histories of the individual modal responses were determined for the structure. From these results, time histories of the structural accelerations, deflections, shears, and moments were determined. The maxima of these values were used in the design of the structure. For this example, the maximum accelerations, displacements, shears, and moments resulting from applying the OBE time history to the reactor building model are shown in
Tables 3.7-3 through 3.7-7.
Floor response spectra for building floors and other support points within the structure were generated as described in Section 3.7.2.5. For this example, the response spectra for damping ratios of 0.005, 0.010, 0.020, and 0.050 derived from the motions of
the reactor building floor at elevation 855 ft 0 in. (mass point 2) caused by the OBE are shown in Figure 3.7-7.
3.7.2.2 Natural Frequencies and Response Loads
UFSAR/DAEC-1 3.7-9 Revision 14 - 11/98 Typical determination of the natural periods of vibration at various locations in the reactor building and the associated movements, forces, and moments is described in Section 3.7.2.1.2. 3.7.2.3 Procedure Used for Modeling
A typical model is described in Section 3.7.2.1.2.
3.7.2.4 Soil-Structure Interaction
The seismic analysis of structures must account for the site specific (bedrock) earthquake motion and the effects of the load transfer system. This can be accomplished with a model that is anchored at bedrock with the soil, structure and other mass/stiffnesses explicitly included. Soil-structure interaction effects are represented by the equivalent springs between the base of the model and the building foundation (see Figures 3.7-6, 3.7-8 and 3.7-9 for examples). This allows the bedrock motion to be the input for the analysis, with the amplification effects being directly calculated. The bedrock response spectra are assumed to be the same as the foundation level response spectra for structures supported on bedrock (i.e., no amplification). As with any analysis, simplification of the model is allowed if appropriate conservatism is applied. For the seismic analysis of a building, a simple model is one that is anchored at the foundation level with the input motion defined for the foundation level to account for the effects of soil-structure interaction. This simple model eliminates the need to model the soil properties. Appropriate conservatism is included in the foundation level response spectra for structures supported on soil. The foundation level response spectra are described in Section 2.5.2. 3.7.2.5 Development of Floor Response Spectra
Floor response spectra were derived for use in the analysis and design of Seismic Category I mechanical and electrical equipment and piping supported within the
structures. The floor response spectra were obtained as follows: the building was subjected to the developed earthquake time history, and the corresponding output acceleration time histories at the floors or points of interest were determined. These acceleration time histories were then used to derive single-degree-of-freedom system
response spectra, which are the floor response spectra, for each floor or point of interest.
The expected variations in the parameters used in the analyses were considered for both the design of the structures and the determination of floor response spectra. A minimum shift in building frequency of +/-10% from the variation in structural properties was included. A possible variation in the soils moduli (for till and backfill) of +/-50% was
used. The use of these values resulted in a conservative design.
UFSAR/DAEC-1 3.7-10 Revision 14 - 11/98 3.7.2.6 Three Components of Earthquake Motion
The methods used to combine the components of earthquake motion are described
in Section 3.7.2.1.
3.7.2.7 Combination of Modal Responses
Since it is not known at which time the maximum modal responses computed in Section 3.7.2.1.1.2 occur, an approximate method had to be used to combine the modal responses to obtain the total response. One method of combination is to compute the sum of the absolute values of the response of each mode. However, the most commonly used method of combination is the root-mean-square method. In this approximation, the total response is obtained by computing the square root of the sum of the squares of the modal maxima. For example, the total displacements of the structure, v tot, are computed from v tot v rvvv n==++2 1 2 2 22 12 12maxmaxmax.....max// (3.7-13) The approximate values of the total shear and moment can be computed in a similar manner.
If several controlling frequencies in an eigenvalue solution are found to lie close together, the root-mean-square method of modal combination was used. In these cases, the modal maxima were combined by direct summation, or the system was analyzed by the time-history method. A technical justification for the use of the root-mean-square method is presented in Section 3.9.3.1.5.
3.7.2.8 Interaction of Nonseismic Structures with Seismic Category I Structures
Seismic design of Nonseismic structures follows prudent engineering practice, in accordance with the Uniform Building Code (UBC), 1970. Although the site is in Seismic Zone 0, criteria for Zone 1 were applied.
Allowable stresses for Nonseismic structures are as specified in Section 3.9.4.
Seismic Category I to Nonseismic structure interfaces are designed so that there will be no functional failure of Seismic Category I structures because of the possible failure of Nonseismic structures. This is accomplished by physically separating Nonseismic structures from Seismic Category I structures.
Those portions of Nonseismic structures housing Seismic Category I equipment were designed in accordance with Seismic Category I design criteria. In addition, the
structures are investigated as a whole to ensure that failure in other areas would not endanger the areas housing Seismic Category I equipment.
3.7.2.9 Effects of Parameter Variations on Floor Response Spectra
UFSAR/DAEC-1 3.7-11 Revision 14 - 11/98 The effect of uncertainties in building frequencies and in soil properties was
considered as described in Section 3.7.2.5.
3.7.2.10 Use of Constant Vertical Static Factors
For those structures that were determined to be rigid in the vertical direction, vertical accelerations were applied as static coefficients. For those items that are not rigid in the vertical direction, dynamic analyses were performed using the time-history method or response spectrum method, as appropriate.
Structures and components with a fundamental natural frequency greater than 30 Hz were considered to be rigid and were designed to resist the maximum acceleration of
their support applied as a static coefficient. A structure's vertical response was determined by analyzing a mathematical model developed to represent the vertical physical characteristics of the soil-structure systems. Soil-structure interaction effects are represented by the equivalent springs shown at the base of each model (see Figures 3.7-6, 3.7-8 and 3.7-9 for examples). The masses in the model include the weights of the building structure and appropriate equipment loads. The stiffness of the springs between masses was based on the stiffness of the walls and columns between floors. The structures were determined to be rigid in the vertical direction, which resulted in no amplification of vertical accelerations.
UFSAR/DAEC-1 3.7-12 Revision 14 - 11/98 3.7.2.11 Method Used to Account for Torsional Effects
For asymmetrical buildings, the translational and torsional seismic responses were determined by performing a dynamic analysis where the building was represented by a three-dimensional lumped-mass model. As an example, Figure 3.7-9 shows the model that was used in the analysis of the control building. This model consists of the building masses lumped at the floor levels and interconnected by flexible shear elements representing the walls and by rigid diaphragms representing the floors and roof. The structure is nearly symmetrical about the north-south axis, but is asymmetrical about the east-west axis because of the braced frame on the south wall. The properties of the sand fill on which the building is founded were modeled by a series of springs that represent the resistance of the sand fill to forces imposed on it by the motion of the building. Horizontal, vertical, and rotational motions of the foundation were considered in determining the foundation stiffness. The spring constants used were computed from formulas derived from elastic half-space theory and published by Whitman and Richart. 1 The range of spring values used was based on a possible variation of soils moduli of +/-50% as described in Section 3.7.2.5. The earthquake motions postulated for the DAEC site, as presented in the PSAR and approved by the NRC, did not include spectra for torsional motions. Therefore, no torsional dynamic analyses of symmetrical buildings were performed. However, the symmetrical structures were designed for the effects of a torsional moment equal to the
story shear acting with an eccentricity between the centers of mass and rigidity of 5% of the appropriate building dimension. The torsional effects had no significant influence on
the designs.
3.7.2.12 Comparison of Responses
Time-history analyses were performed for all structures. Modal response spectrum analyses were not performed because the time history produced spectra that
were conservative relative to the criteria response spectra.
3.7.2.13 Methods for the Seismic Analysis of Dams
This subject does not apply to the DAEC. See Section 2.4.4.
3.7.2.14 Analysis Procedure for Damping
Structures were analyzed using modal superposition techniques. Element or material-associated damping values are given in Table 3.7-1. "Composite" or modal damping values in structural systems comprised of different element material types were selected on the basis of an inspection of the significant mode shapes and on the assumption that the contribution of each material to the composite effective modal damping is proportional to the elastic energy induced in each material. The following UFSAR/DAEC-1 3.7-13 Revision 14 - 11/98 criteria and procedures were applied on a mode-by-mode basis to evaluate and conservatively determine composite damping values:
- 1. Where a particular mode primarily indicated the response to be of a single element type, the damping value corresponding to that element type was assigned to that mode. Where all but a negligible amount of the elastic energy was induced in, for example, concrete or soil, the damping value appropriate to these materials was applied. Similarly, where a lightly damped material exhibited a major portion of the elastic energy to be of the mode, a conservative choice was made to use the damping value of that material for that mode. In most cases for the DAEC, the modes were well defined according to material types, and composite damping values could be selected by a visual inspection of mode shapes with no numerical computations
required.
- 2. In a few instances, the above criteria could not be applied because a particular mode indicated the response to be of several element types. The damping value for that mode was conservatively estimated on the basis of the degree of participation of the different elements. The elastic energy induced in the different elements was estimated, and composite damping values were
assigned in proportion to the elastic energy. The approach described above is consistent with currently accepted techniques, and in all cases, the damping values were selected conservatively. The use of this approach has resulted in a design that can conservatively resist the seismic motions
postulated.
3.7.3 SEISMIC SUBSYSTEM ANALYSIS
Seismic Category I equipment was examined to ensure its ability to withstand design loading requirements. Experienced designers examined the equipment using appropriate techniques to determine which specific portions of the systems and components required further examination. These techniques fit into two general categories: (1) normal analytical techniques, consisting of empirical design methods as defined by appropriate design codes; and (2) special techniques, used to supplement code
calculations or to cover conditions not considered by existing codes.
- 1. Normal Design Techniques
Seismic Category I equipment (e.g., piping, valve bodies, and pump cases) was designed in accordance with applicable industrial codes. Some codes used empirical design methods for equipment that could not be sized by conventional rational stress analysis methods and did not require a detailed stress analysis for primary design work. This equipment was designed to meet a detailed functional requirement specification. The design is supported
by field and test experience. UFSAR/DAEC-1 3.7-14 Revision 14 - 11/98
- 2. Special Supplemental Methods
Some complex equipment is normally sized by rational stress analysis techniques and requires supplemental criteria in areas where industrial codes
do not apply.
For piping systems designed to the rules of ANSI B31.7, that is, Nuclear Classes II and III and Seismic Category I Piping, the stresses resulting from earthquake loads
have been considered as follows:
- 1. OBE The vectorial combination of all longitudinal primary stresses does not exceed 1.2 times the hot allowable stress (S h). 2. DBE The vectorial combination of all longitudinal primary stresses does not exceed the material yield stress at temperature unless higher allowable limits can be calculated by the methods outlined in Section 3.8.
The seismic evaluation of piping is discussed in detail in Section 3.7.3.8.
3.7.3.1 Seismic Analysis Methods
3.7.3.1.1 Bechtel-Supplied Equipment
All BOP Seismic Category I mechanical equipment was analyzed according to the General Project Seismic Requirements that were developed by Bechtel and the seismic
consultants on the DAEC.
In most cases, the analysis of the equipment was made by the equipment vendor and reviewed by Bechtel; however, a few items were analyzed by Bechtel in the manner
of a typical calculation.
The basis for ensuring proper functioning of the equipment during a seismic event was that the equipment should not fail or be subject to misoperation or malfunction during or after a seismic disturbance resulting from the OBE, nor should the equipment fail or be subject to misoperation or malfunction after a seismic disturbance resulting from the DBE. Operability during an OBE event was ensured by requiring that the maximum stresses from combined seismic and normal loads should not exceed the
applicable code allowable stresses without the usual one-third increase of allowable stresses for short-term loadings. The induced displacements were also required to not
exceed those required for safe operation. Operability after a DBE event was ensured by requiring that the maximum stress from combined seismic and normal loads should not UFSAR/DAEC-1 3.7-15 Revision 14 - 11/98 exceed 90% of the yield stress of the material and that the induced displacements should not cause failure, malfunction, or prevent safe shutdown.
Three methods were permitted to ensure compliance with the project seismic requirements: frequency-not-determined analysis, frequency-determined analysis, or verification by testing. In all cases, the vertical seismic component was two-thirds of the foundation level horizontal acceleration for the OBE or DBE, applied simultaneously with the horizontal component.
For the frequency-not-determined method of analysis, the equipment was analyzed for a static coefficient equal to 1.5 times the peak acceleration of the floor response spectra. An additional discussion of this analysis method is found in Section 3.7.3.10. The stresses due to this static coefficient were combined with the normal
stresses as indicated above.
For the frequency-determined method of analysis, the fundamental natural frequency of the equipment was determined by analysis. If the natural frequency was found to be greater than 33 cycles/sec (Hz), the equipment was categorized as "rigid" and
analyzed for a constant acceleration equal to the peak support acceleration. If the natural frequency was found to be less than 33 cps, the equipment was categorized as "flexible" and analyzed for accelerations determined from a response acceleration spectrum appropriate for the mounting location of the equipment. Whether the equipment was rigid or flexible, the stresses due to the seismic accelerations were combined with the normal stresses as indicated above.
For items that were analyzed by testing, the equipment was subjected to vibration levels equivalent to those represented by the appropriate response acceleration spectrum for the mounting location of the equipment. For the OBE accelerations, the equipment
was required to be operable during and after testing. For the DBE accelerations the equipment was required to be operable after the test.
It was found that certain types of equipment were most readily analyzed using the frequency-not-determined method. The Seismic Category I heating and ventilating duct systems typify this equipment. The complex geometries of a duct system make frequency determination extremely difficult; however, the small component weights make the induced seismic loads low. Thus, the Seismic Category I heating and ventilating ducts could be easily analyzed using the frequency-not-determined method and the resulting seismic and normal stresses kept within permitted values.
Equipment that was best analyzed by the frequency-determined method is typified by the several small Seismic Category I pumps used throughout the plant. The components of these pumps are relatively simple, and fundamental natural frequencies were determined using conventional methods of analysis. Invariably, the natural frequencies of these items fell in the rigid range with the resulting low seismic accelerations. Because of the inherent conservatism in the design of this type of UFSAR/DAEC-1 3.7-16 Revision 14 - 11/98 equipment, and the low seismic coefficients required, the seismic stresses were of little consequence in the design of this equipment.
As of December 1972, no Seismic Category I mechanical equipment had been analyzed by testing; however, should testing be required, the methods used will be as
described above.
3.7.3.1.2 General Electric-Supplied Equipment
The design of the Seismic Category I mechanical equipment must meet the design criteria set forth in this section. Proper functioning of Seismic Category I equipment during a seismic event is ensured by the fact that such equipment is analyzed for combinations of dead loads, live/operating loads, and seismic loads. The stresses resulting from such load combinations are less than allowable design stresses. For the
MSIV, the structural acceptance criteria 3.8.4.5 is used for non-pressure boundary components. The results of the various loads analyses are included in Tables 3.7-8
through 3.7-20.
For dynamic analysis, Seismic Category I equipment is represented by models that consist of discrete masses connected by weightless springs. The criteria used to lump masses are the following:
- 1. Masses are chosen so that all significant modes are included.
- 2. A mass is located at each significant concentrated weight.
- 3. If the equipment has an overhang span whose flexibility is significant, an overhanging mass is used.
- 4. When a mass is located between two supports, it is located near the point of maximum displacement. This tends to conservatively lower the natural frequencies of the equipment. Similarly, in the case of live loads or variable support stiffness, the location of the load and the magnitude of support stiffness are chosen so as to lower the frequency. This ensures conservative dynamic
loads.
3.7.3.2 Determination of Number of Earthquake Cycles
The stress in the piping is assumed to be occurring with a frequency of 20 Hz. This value is considered to be conservative, since it is greater than the significant natural frequencies of both the containment structures and the piping systems. The time duration of both the OBE and the DBE is assumed to be 15 sec. Two occurrences of the OBE and one occurrence of the DBE are assumed.
UFSAR/DAEC-1 3.7-17 Revision 14 - 11/98 These values are also considered to be conservative. The number of earthquake cycles is, therefore, 600 for the OBE and 300 for the DBE. These numbers are specified
in the Bechtel design specification for ASME Code, Section III, Nuclear Class I piping.
The following criteria were used in the design of Class I systems:
- 1. DBE a. Number of assumed DBE in the life of piping system is <
- 1.
- b. Duration of strong motion vibration for each DBE is 15 sec.
- c. Number of cycles of the piping system for each DBE is 300.
- d. Total lifetime number of cycles of the piping system is 300.
- 2. OBE a. Expected number of equivalent OBE in the life of the piping system is 2.
- b. Average duration of strong motion vibration OBE is 15 sec.
- c. Average number of cycles of the piping system for each OBE is 300.
- d. Total lifetime number of cycles of the piping system is 600.
The OBE was considered to act concurrently with upset condition when making the ANSI B31.7 analysis, and the DBE was considered an emergency condition.
3.7.3.3 Procedure Used for Modeling
Analytical models used for structures and components are described in Sections 3.7.2.1 and 3.7.3.1. Models used for piping are described in Sections 3.7.3.8.
3.7.3.4 Basis for Selection of Frequencies
Frequencies for seismic evaluations were selected in accordance with Sections
3.7.2.1.1, 3.7.3.1 and 3.7.3.8.
3.7.3.5 Use of Equivalent Static Load Method of Analysis
The conservative simplification of using equivalent static loads is described in
Sections 3.7.3.1.1 and 3.7.3.8.4.
3.7.3.6 Three Components of Earthquake Motion
UFSAR/DAEC-1 3.7-18 Revision 14 - 11/98 The method used to combine the components of earthquake motion are described in Section 3.7.2.1.
3.7.3.7 Combination of Modal Responses
Responses were combined as described in Section 3.7.2.7.
3.7.3.8 Analytical Procedures for Piping
The piping has been analyzed for the effects of thermal loads combined with
deadweight and external forces. The calculated bending and torsional stresses are combined in accordance with the requirements of ANSI B31.1.0, 1967, Power Piping
Code. Flexibility and stress intensification factors were applied in accordance with ANSI B31.1.0. Several thermal cycles were evaluated, and all critical points were compared to the expansion stress limits of ANSI B31.1.0. In addition, events with very low probability of occurrence were analyzed, and stresses of all critical points were evaluated to the limits defined in this section. The point of highest stress for each load combination is located, and the calculated stress is compared to the allowable stress in Tables 3.7-21
through 3.7-26.
The piping systems were dynamically analyzed using the response spectrum method. Input to the dynamic analysis was damped acceleration response spectra for the applicable floor elevation. The percentage of critical damping for all modes is 0.5% for
the OBE and 1.0% for the DBE as given in Table 3.7-1.
The continuous piping system was mathematically idealized as an assembly of elastic structural members connecting discrete nodal points. Nodal points were placed in such a manner as to isolate particular types of piping elements, such as straight runs of pipe, elbows, and valves, for which force deformation characteristics can be categorized.
Nodal points were also placed at all discontinuities, such as piping supports, concentrated
weights, branch lines, and other critical points where stress calculation was desired.
The mathematical model was a lumped mass, multi-degree of freedom model. The distributed piping mass was "lumped" at the system nodal points. Valves were considered as lumped masses on the pipe, and valve operators are considered as lumped masses acting at the center of gravity of the operator. Inertia forces at the lumped masses resulted from the seismic-induced accelerations of the system supports.
Total system response, in terms of forces, moments, and seismic stresses, was obtained by the root-mean-square combination of the individual modal values for each
direction of earthquake excitation input. The vertical direction earthquake was assumed to act simultaneously with either direction horizontal earthquake. The results of the seismic analyses were combined with other loading conditions, and combined stresses were computed in accordance with ANSI B31.1.0.
UFSAR/DAEC-1 3.7-19 Revision 14 - 11/98 Constant load factors were not used in a multimass dynamic analysis such as piping seismic analysis. Inputs to such dynamic analyses are taken from the reactor
building floor response spectra.
3.7.3.8.1 Balance-of-Plant Piping Analysis
BOP safety-related piping systems used the response spectrum technique to compute shears, moments, stresses, deflections, and/or accelerations for each seismic-excited piping mode. The piping system was idealized as a mathematical model consisting of lumped masses separated by elastic members. The lumped masses were carefully loaded so as to adequately represent the dynamic and elastic properties of the piping system. The three-dimensional stiffness matrix of the mathematical model was determined by the direct stiffness method.
The mass matrix was also calculated. After the stiffness and mass matrices of the mathematical model were calculated, the natural frequencies of vibration and corresponding mode shapes were determined using the following equation:
[]KW N M N=20 (3.7-14) where K = stiffness matrix W N = natural circular frequency for the N th , mode M = mass matrix 0 = zero matrix N = mode shape matrix for the N th mode The mode shapes were normalized according to the following equation:
=N T M N 1 (3.7-15)
UFSAR/DAEC-1 3.7-20 Revision 14 - 11/98 The maximum response of each mode was found through the following equation: Y N t N TMDSa W N M N ()max=2 (3.7-16) where Sa = Spectral acceleration value for the N th mode D = earthquake vector matrix, used to introduce earthquake direction to the response analyses N T = transposition of the N th mode shape M N = generalized mass of the N th mode shape Y N = generalized coordinate for the N th mode Using the maximum generalized coordinates for each mode, the maximum deflections associated with each mode are calculated using
()V N N Y N tmax (3.7-17)
The root-mean-square method is used to combine the total modal responses as
indicated by
[]ViViViVi N=+++1 2 2 22 12......./ (3.7-18) where Vi = deflection at i th point due to the response of N modes Vi N = deflection at i th point due to N th mode Once the appropriate deflections were determined for each mass and each mode, the effective applied forces for each mode were computed using the following equation:
Q N KV N= (3.7-19) where Q N = inertial forces due to mode N UFSAR/DAEC-1 3.7-21 Revision 14 - 11/98 The accelerations for each mode were calculated by the following equation: a N MQ N=1 (3.7-20) where a N = accelerations due to N th mode M1 = the inverse of mass matrix After the effective forces have been determined, the internal forces (thrusts and shears) and moments for each mode were calculated using
S N bQ N= (3.7-21) where S N = internal forces and moments due to the N th mode b = force transformation matrix The internal forces (thrusts and shears) and moments were combined on the same basis as Equation 3.7-18. The stresses were then computed from the internal forces and moments and in accordance with ANSI B31.7.
The percentage of critical damping for all modes is 0.5 for the OBE and 1.0 for
the DBE.
The criteria for the selection and location of snubbers and dampers for Seismic
Category I piping were as follows:
- 1. The use of snubbers or dampers was limited to those locations where unacceptable thermal expansion stresses would result from the use of a rigid
translational restraint.
- 2. The snubbers and dampers have provision for thermal movement and can limit translational movement during the earthquake.
- 3. The snubbers and dampers were selected to sustain the seismic reaction resulting from the DBE at the point of attachment to the piping.
3.7.3.8.2 Computer Codes
The computer codes that were used by the San Francisco Power Division of Bechtel in the seismic stress analysis of safety-related piping are the following:
- 1. ME 632 - "Seismic Analysis of Piping Systems," Bechtel.
UFSAR/DAEC-1 3.7-22 Revision 14 - 11/98 2. ME 101 - "Leap" - "Linear Elastic Analysis of Pipe," Bechtel.
- 3. PISOL - EDS Nuclear, Inc.
- 4. NUPIPE - Nuclear Services Corporation.
- 5. SAPIPE - PMB Systems Engineering, Inc.
- 6. TPIPE - PMB Systems Engineering, Inc.
The computer programs listed above have been verified as follows:
- 1. ME 632 has been verified using PISOL, PIPESD, and TPIPE.
- 2. ME 101 has been verified using ME 632, TPIPE, and SUPERPIPE.
- 3. PISOL has been verified using NUPIPE, PIPESD, ADLPIPE, and ME 101.
- 4. NUPIPE has been verified using ADLPIPE. (In the verification, the algebraic summation option in ADLPIPE was not used.)
- 5. SAPIPE has been verified using PISOL.
- 6. TPIPE has been verified using PISOL and ME 632.
3.7.3.8.3 Recirculation Piping and Nozzle Analysis
3.7.3.8.3.1 Recirculation Piping
The dynamic analysis that was performed on the recirculation piping systems determined the inertia effects of seismic loading. The primary objective of this analysis was to demonstrate that the recirculation piping loop A and loop B and their RHR piping meet the requirements of ANSI B31.1.0, 1967.
Because dynamic effects due to flow-induced vibration are insignificant, an analysis of this type was not performed. Water-hammer-effects analysis in the recirculation piping system was also considered and found to be negligible.
The seismic analyses were performed by using the method of response spectrum superposition. In this method, the maximum acceleration response of each mass of the mathematical model was computed for each significant mode using an appropriate acceleration response spectrum corresponding to the reactor building motion criteria. The computation for the acceleration response spectrum of each mass, which is the function of the maximum acceleration parameter versus the period of vibration, was based on a single-degree-of-freedom system s ubjected to the appropriate force vibration for an assumed critical damping ratio. All the significant mode shapes with frequencies UFSAR/DAEC-1 3.7-23 Revision 14 - 11/98 less than 33 Hz were included in determining the seismic responses. Each seismic response parameter (inertia forces, displacements, member forces, reactions, etc.) was calculated using the sum of squares methods to determine the resultant value of the respective response parameters.
The seismic (inertia loads) cases were as follows:
- 1. NCASE 1 and 2 OBE cases for vertical accelerations (Y direction) acting concurrently with horizontal north-south accelerations (Z direction) and for the same vertical
accelerations acting concurrently with the horizontal east-west accelerations (X direction), respectively. The OBE cases assume an 0.5% damped acceleration response spectrum.
- 2. NCASE 3 and 4 DBE cases for the same combined accelerations as OBE NCASE 1 and 2, respectively, except that the acceleration response spectra are for the DBE which are two times the OBE, using 1.0% of critical damping.
The seismic anchor movement load case was based on the OBE and is considered in the analyses to determine the effects of relative earthquake displacements on the piping and supports.
The horizontal displacements for the reactor pressure vessel nozzles and the
anchors located just outside the drywell penetrations were based on the maximum value at the corresponding elevation. There were no relative vertical displacements.
The loading conditions for thermal, weight/pressure, external forces, and seismic loadings for various normal and upset loadings, and loadings having a very low probability of occurrence, are individually considered and subsequently combined as to satisfy the rules of ANSI B31.1.0, 1967, and the following stress criteria:
3.7.3.8.3.2 Stress Criteria See Tables 3.7-21 through 3.7-26.
UFSAR/DAEC-1 3.7-24 Revision 20 - 8/09 3.7.3.8.3.3 Drawings
The following drawings provide piping isom etric views of the recirculation line. Drawing No. C-518 M-116 M-332 M-333 M-338 M-339 M-340 M-341 M-352 M-353 M-357 APED-B11-2655-97 APED-B31-9(l)
APED-B31-9(2)
APED-B31-15(l)
APED-B31-15(2)
APED-B31-23(l)
3.7.3.8.3.4 Nozzle Analysis
Vessel Design Conditions
- 1. Design pressure, 1250 psig at vessel elevation 0.0.
- 2. Design temperature, 575
°F.
- 3. Normal operating pressure - 1005 psig at top of vessel.
- 4. Normal operating temperature - Saturation temp at 1005 psig.
The calculated reactions are shown in Table 3.7-27.
The recirculation inlet nozzle design was found adequate in accordance with the requirements of ASME Code, Section III, Nuclear Vessels, and of General Electric
Specification 2lA1100AS.
UFSAR/DAEC-1 3.7-25 Revision 19 - 9/07 3.7.3.8.4 Equivalent Dynamic Analysis
There are two types of analysis in this category, as follows:
- 1. Analysis using first mode greater than the peak value.
- 2. Analysis using a modified spectrum curve.
Both of these approaches result in charts and tables showing span lengths and
restraint forces for various building elevations.
3.7.3.8.4.1 First Mode Greater than Spectrum Peak
A piping system may be considered seismically acceptable if it can be divided into a series of simple spans. These spans are limited by guides that are specified in the form of vertical and lateral restraints at each change of direction, at all concentrated masses (e.g., valves), at all extended masses, at each tee, and at a maximum spacing on
straight runs of piping defined by the following criteria. The fundamental frequency of multispan piping systems supported as stated above is greater than or equal to the fundamental frequency of a simple beam of maximum seismic span that is calculated
using UFSAR/DAEC-1 3.7-26 Revision 14 - 11/98 f EI mL=2 4 (3.7-28)
where f = fundamental frequency E = modulus of elasticity I = moment of inertia m = mass per unit length L = maximum seismic span (maximum distance between two seismic guides)
The frequency is chosen so that it is 20% larger than the frequency that defines the rigid side of the spectrum curve as shown in Figure 3.7-10. This is done on a case basis for each elevation. The simple beam formula can be used for the static equivalent load analysis with the peak value and at the spectrum curve and yields conservative results. A static load is then applied to the span to determine the maximum displacement, moment, and restraint force, which is calculated by
VmLSa EI=5 384 4 (3.7-29) MmLSa=0125 2. (3.7-30) RmLSa=2 (3.7-31) where S a = the peak value of spectrum curve V = maximum displacement M = maximum moment R = maximum restraint force
Even though restraints are specified to ensure that the system will be on the rigid
side of the curve, the spectral acceleration associ ated with the peak of the curve is used to obtain restraint loading and piping stresses.
A dynamic analysis was performed to verify the conservatism of the approach.
The piping system chosen for analysis was a general model that included various piping configurations. The model and results are given in Section 3.7.3.8.4.5. A sample chart of limiting span is shown in Table 3.7-28.
3.7.3.8.4.2 Modified Spectrum Method
The piping system is restrained as described above, except that the maximum spacing between two seismic guides on a straight run was determined by dynamic UFSAR/DAEC-1 3.7-27 Revision 14 - 11/98 calculations using a modified spectrum curve. The spectrum curve for a particular building elevation was modified such that the flexible side of the peak of the curve remained constant at the peak spectral acceleration for decreasing frequencies (see Figure 3.7-11). A sample chart of limiting span is shown in Table 3.7-29.
If a dynamic analysis were performed using the above spectrum, the results
would, by inspection, be conservative.
The fundamental frequency of the piping system supported as stated above is greater than the fundamental frequency of a simple beam of maximum seismic span, which was calculated using equation 3.7-28.
A dynamic analysis was then performed on a simply supported beam. The justification of this approach as well as a study demonstrating conservatism by comparing results of this approach with a dynamic analysis of a random piping system
are presented later.
The following is a description of the development of the "modified spectrum method."
The circular frequency of a simple beam is calculated using
()W N N E I mL=2 4 (3.7-32) where W N = natural circular frequency for the N th mode The response spectrum method is then used to find the maximum response of each mode
()()V N S a N NW N V N 2 1 4 21 21 2 21 2 0==, (3.7-33) where S aN = spectral acceleration value of the modified N spectrum curve for the N th mode V N = maximum displacement (at mid-span) due to N th mode The SRSS method is used to combine the modal responses as described in Section 3.7.2.7. VVVV N=++1 2 2 22... (3.7-34)
UFSAR/DAEC-1 3.7-28 Revision 14 - 11/98 where V = maximum displacement due to the response of N modes
After the maximum displacement for each mode is determined, the maximum moment (at mid-span) and the maximum restraint force (at the support) are determined
by
()M NmLSa N N M N 21 4 2 21 3 21 3 2 0==, (3.7-35)
()R N mLSa N N R N 21 4 21 2 21 2 2 0==, (3.7-36) where M N = maximum moment (at mid-span) due to the N th mode R N = maximum restraint force (at the support) due to the N th mode The moments and restraint forces are also combined on the same basis as
Equation 3.7-34.
From the nature of the modified spectrum curve, the spectral acceleration for the first mode is always the largest value of the spectral accelerations of any mode. The first mode frequency for a given span is calculated and the resultant spectral acceleration is obtained. This maximum acceleration is then applied to all higher modes giving
conservative results. The variables in Equations 3.7-34 through 3.7-36 can then be eliminated and the equations reduced to
VmLSa EI=00131 4. (3.7-37) MmLSa=01291 2. (3.7-38) RmLSa=08164. (3.7-39) where Sa = spectral acceleration of modified spectrum curve corresponds to the fundamental frequency of the maximum seismic span V = maximum displacement M = maximum moment R = maximum restraint force
The analysis is made for both horizontal and vertical excitation. The horizontal and vertical responses are then combined by the square root of the sum of the squares.
UFSAR/DAEC-1 3.7-29 Revision 14 - 11/98 The modified spectrum method as described above uses dynamic techniques that give results equal to a full dynamic analysis of a simply supported beam. It has been demonstrated that continuous beams restrained as described above will result in a response lower than that of the simply supported beam.
The modified spectrum method is therefore a dynamic analysis and the SRSS method is applied to the combination of respons es. The larger of the square root of the sum of the squares of one horizontal and the vertical is used in the computations.
The static approach as described in Section 3.7.3.8.4.1 does, however, use the absolute sum method for the combination of responses.
Highest stressed regions are checked and verified for compliance with applicable
codes.
3.7.3.8.4.3 General Guidelines
Although it is difficult to categorize as to which analytical classification a specific piping system will fit, certain generalizations can be made.
The major parts of the larger-diameter piping systems were analyzed using a full dynamic analysis. This was especially true where process fluid temperatures were high. For these systems, the large number of restraints required with the other techniques described would lead to thermal expansion difficulties. This also applies to small-diameter high-temperature systems. For other piping, namely large diameter-low temperature and small diameter-low temperature, one of the equivalent dynamic
approaches was used.
Rigid-range piping techniques were typically reserved for instrumentation and some small-diameter piping. As previously stated, many conditions affect the selection of the appropriate technique. For example, a large-diameter system-low temperature system may be given a rigorous dynamic analysis to reduce the number of restraints required if the system is located such that installation of the restraints would be difficult.
3.7.3.8.4.4 Verification of Simplified Approach
The simplified approach is verified if it can be shown that the fundamental frequency of a piping system restrained as st ated above is greater than or equal to the fundamental frequency of a simply supported beam (pin connected ends called SSB) of maximum seismic span (L).
The fundamental frequency of a multi-equal-span continuous beam is equal to the fundamental frequency of a simply supported beam of the single span length of the continuous beam. 2 For a multi-unequal-span continuous beam, the fundamental frequency of the maximum span using the SSB formula is less than the fundamental frequency of the multi-unequal-span continuous beam. This can be easily proved by UFSAR/DAEC-1 3.7-30 Revision 14 - 11/98 considering a three-span continuous beam. Suppose that the middle span is longer than the two side spans: When the side spans are made smaller, the system approaches the fixed-fixed end case. Suppose that one of the side spans is the longest span: When the middle span is made smaller, the system is approaching the hinged-fixed end case. From the analytical results of a single simply supported beam with various end conditions, it can be concluded that the SSB formula gives the smallest frequency value of the three
cases.2,3 The same argument can be applied to multispan continuous beam. Therefore, the fundamental frequency of a piping system restrained as described above is greater than or equal to the fundamental frequency of a simply supported beam of the maximum seismic span (L).
Peak of response curve method is conservative for piping supported in accordance
with Section 3.7.3.8.4.1.
When seismic spans are limited in length by methods described in Section 3.7.3.8.4.1, the first mode (fundamental) freque ncy falls on the rigid side of the peak of the spectrum curve (Figure 3.7-10). A safety fact or of 1.2 is used to ensure that the first mode is on the rigid side of the peak.
Referring to the spectrum response curve (Figure 3.7-12), it can be seen that the nature of the curve is such that the lst mode will experience a higher spectral acceleration than the higher modes.
It can be shown that the first mode dominates the response for the simply supported beam.
The higher modes contribute less than 4% to the total system response. The coefficient of response is 0.129 for the combination of all modes for bending moments. A static case (consideration of only the lst mode) shows a coefficient of 0.125. The
percentage difference is
01290125012532%....= (3.7-40) Therefore, the use of the l st mode response differs from the total dynamic response by less than 4%.
In the interest of conservatism, however, the spectral acceleration corresponding to the peak of the spectrum response curve is used to determine piping stresses. This practice results in factors of conservatism of 3 to 8, that is,
S ap S a r=3 to 8 (3.7-41) where UFSAR/DAEC-1 3.7-31 Revision 14 - 11/98 S a p = spectral acceleration corresponding to the peak S a r = spectral acceleration corresponding to rigid range
In actuality, the dynamic analysis using the response spectrum method results in a
safety factor of 10.
The modified spectrum method is conservative for piping supported in
accordance with Section 3.7.3.8.4.1.
Owing to the characteristics of the modified spectrum curve (see Figure 3.7-11), the span with the lowest fundamental fre quency will always have a spectral acceleration equal to or greater than spans with a higher frequency.
It can be demonstrated by the same technique as used above that both the fixed-fixed end and fixed-hinged end case result in a lower dynamic response than for the simply supported beam.
Dynamic analyses were done for beams having five, six, and seven equal spans, and in all cases the stresses resulting were smaller than those obtained when the simple beam formula was used. The following is a comparison of the results of the dynamic analysis and the static equivalent load method of a typical problem.
3.7.3.8.4.5 Typical Problem (Example)
As a typical example, a piping system (Figure 3.7-13) has been analyzed by the dynamic analysis using the response spectrum method. The following results show that the static equivalent load method described in Section 3.7.3.8.4.1 yields a very
conservative result:
Dynamic Analysis Static Equivalent Load Method Fundamental frequency (cps) 12.10 11.10 Maximum stress (psi) 200.00 2,031.00 Maximum displacement (in.) 0.01 0.111 Maximum reaction (lb) 10.00 55.80
This typical example analyzed through both the dynamic analysis and the static equivalent load method uses the following criteria:
Number of degrees of freedom = 130 Number of modes considered in dynamic analysis = 30 Least significant period = 0.03 sec
The following is a comparison of the results of the dynamic analysis and modified spectrum method of a typical problem.
UFSAR/DAEC-1 3.7-32 Revision 14 - 11/98 As a typical problem, a piping system (F igure 3.7-14) has been analyzed through the dynamic analysis using the response spectrum method. The following results have shown that the modified spectrum method yields a very conservative result:
Dynamic Analysis Modified Spectrum Method Fundamental frequency (cps) 11.05 6.47 Maximum stress (psi) 1300.00 5703.00 Maximum displacement (in.) 0.05 0.438 Maximum reaction (lb) 16.00 22.00
This typical example analyzed by both the dynamic analysis and the modified spectrum methods (Figure 3.7-12) uses the following criteria:
Number of degrees of freedom = 112 Number of modes considered in dynamic analysis = 20 Least significant period = 0.03 sec
The information presented in this section is contained in Bechtel Corporation
Topical Report BP-TOP-1.
3.7.3.8.5 NSSS Piping
NSSS Class I piping and equipment supplied by General Electric, mainly fall
under four categories as follows:
- 1. Piping (recirculation loop and primary steam piping).
- 2. Equipment such as pumps, heat exchangers, and tanks.
- 3. Instrumentation.
- 4. Reactor pressure vessel internals.
The methods of seismic analysis for the above categories are discussed below.
Piping The piping systems are dynamically analyzed using the response spectrum method. For each of the piping systems, a mathematical model consisting of lumped masses at discrete joints connected together by weightless elastic elements is constructed. For the piping runs, the number of lumped masses is normally adequate to include all the vibration modes with frequencies less than 33 Hz. In addition, masses are lumped at the
points where concentrated weights, like valves, etc. are located. Stiffness matrix and mass matrix are then generated, and natural periods of vibration and corresponding mode shapes are determined. Input to the dynamic analyses are the acceleration response UFSAR/DAEC-1 3.7-33 Revision 14 - 11/98 spectra for the support locations. The increased flexibility of the curved segments of the piping systems is considered. The torsional effects of valves and other eccentric masses are normally considered in the development of a dynamic model.
If the stresses due to these effects can r eadily be shown to be less than 500 psi, the torsional effects of these components are often neglected in the dynamic model. The results for earthquakes acting in the X and Y (vertical) directions simultaneously and the Z and Y directions simultaneously are computed separately. Maximum joint displacements, member forces, and support reactions are determined by a square root of the sum of the square combination of each of these parameters for each mode and for each set of earthquake directions. The member forces thus obtained are combined with the member forces produced by other loading conditions to compute the stresses. The applicable stress and deformation criteria for all the loads including seismic loads are
defined in Section 3.9.
3.7.3.8.6 Piping with Multiple Input Attachments
The relative displacement between piping and equipment under seismic excitation is interpreted as the relative displacement between the supports of piping systems. The following method was used to determine the effect of differential end and support motion on piping systems. The seismic displacements at the ends and at restraints are known from the building analyses. These displacements were applied to the piping restraints and anchors corresponding to the maximum differential displacements that could occur. The analysis was made twice: once for north-south differential displacements and once for east-west differential displacements. For each response quantity considered (i.e., moments or displacements at a point, and restraint force or moment), the largest value of the two analyses was chosen. The displacements and restraint forces and moments were combined with the corresponding quantity from the inertia load analysis. The basis of combination was root mean square, since the maximums of the two quantities would not occur at the same time. The internal moments were used to calculate the stresses in
accordance with ANSI B31.7.
The results of the differential displacement analysis were usually insignificant compared to the results of the inertia force analysis. The differential displacements were usually very small, and most piping systems (especially hot ones) had enough flexibility that these small displacements caused very little distress.
The above discussion applies to BOP equipment and piping; the paragraph below applies to the NSSS equipment and piping.
Criteria and methods of analysis are in cluded in Tables 3.7-21 through 3.7-26. The results obtained from the maximum seismically-induced relative anchor displacements are accounted for by adding to the results from normal seismic analyses.
For piping systems that go between different structures, such as between the reactor building and auxiliary buildings, the spectrum curves at the end points may be UFSAR/DAEC-1 3.7-34 Revision 14 - 11/98 radically different. For these pipes, if intermediate restraints exist between the pipe and one of the two structures, a spectrum curve a ssociated with the restraining structure was used. If no intermediate restraints existed, the average of the spectrum curves associated with the end points was used.
In 1985, the loads at seismic supports were reanalyzed as a part of IE Bulletin 79-
14 activities. The support calculations showed that all supports were operable, but at some supports the component reactions exceeded the design rating. Those supports have been modified to restore the original design factor of safety.
3.7.3.8.7 Seismic Analyses of As-Built Safety-Related Piping Systems
In response to IE Bulletin 79-14, the DAEC inspected as-built safety-related piping systems and found no nonconformances. The inspections described in IE Bulletin 79-14, Items 2 and 3, were conducted for the DAEC by Bechtel using its Procedure for Verifying Conformance of Seismic Analysis to Actual Configuration of Safety-Related Piping Systems. The procedure and the results of the inspections were included in
Reference 4.
3.7.3.9 Multiple-Supported Equipment Components With Distinct Inputs
The only significant components subject to multiple simultaneous seismic input are certain piping subsystems. The evaluation of the effects of such configurations and inputs is presented in Section 3.7.3.8.6.
3.7.3.10 Use of Static Coefficients
Equipment specifications for certain mechanical and electrical equipment allow a static analysis using a coefficient of 1.5 times the peak of the applicable floor response spectrum. This coefficient was derived from a study of different structure configurations and was shown to be conservative.
The study included the following representative simple structural configurations:
- 1. Uniform cantilever comprised of shear elements.
- 2. Uniform cantilever comprised of bending elements.
- 3. Uniform beam with fixed ends.
- 4. Uniform beam with pinned ends.
- 5. Rigid cantilever with a flexible rocking base.
For all cases, it was assumed that the spectral accelerations of the first mode were equal to the peak of the spectrum and that the spectral accelerations of the higher modes were equal to one-half that value. These assumptions are conservative relative to the floor response spectra for the plant. The resultant mass accelerations for each mode were computed using participation factors and then combined by calculating the square root of the sum of the squares of the modal values. The maximum resulting acceleration in the UFSAR/DAEC-1 3.7-35 Revision 14 - 11/98 structure represents the acceleration that is applied to the item of piping or equipment for design purposes. This acceleration is expressed as a coefficient times the peak spectral
accelerations in the following tabulation. (Als o included are the coefficients that would be obtained if only the first mode response is considered.)
Coefficient Case First Mode Only All Significant Modes 1 1.27 1.33 2 1.46 1.50 3 1.29 1.31 4 1.25 1.26 5 1.43 1.46
The maximum coefficient of 1.50 for a slender cantilever represents a reasonable
upper bound.
For the Seismic Category I piping that was analyzed using a static load equivalent, the analysis was based on the simple beam analysis of the maximum seismic span of piping system providing a seismic gui de (restraining the pipe laterally) at each change of direction, at all concentrated masses (e.g., valves), at each tee, and at a maximum spacing (which was determined by the following method) on a straight run of piping.
The fundamental frequency of the small-diameter piping system supported as stated previously is greater than the fundamental frequency of a simple beam of maximum seismic span that is calculated using
f EI mL=2 4 (3.7-42) where f = fundamental frequency E = modulus of elasticity I = moment of inertia m = mass per unit length L = maximum seismic span (maximum distance between two seismic guides)
The static load is equivalent to the peak value of the appropriate spectrum curve. The seismic span is limited by the fundamental frequency (Equation 3.7-42) to be greater
than the frequency of the peak area of the spectrum curve. The piping system is then in the rigid side of the peak area of the spectrum curve. The simple beam formula can be conservatively used for the static equivalent load analysis with the peak value. The UFSAR/DAEC-1 3.7-36 Revision 14 - 11/98 maximum displacement, moment, and restraint force are calculated by equations 3.7-29 through 3.7-31.
The analysis is made for both horizontal and vertical excitation. The horizontal and vertical responses are then combined by the sum of the absolute values method. The stresses are computed in accordance with ASME Code, Section III.
A typical case has been dynamically analyzed using the response spectrum method, with the following results:
Maximum stress = 250 psi Maximum displacement = 0.01 in. Maximum reaction = 10 lb
The simplified analysis as described above shows the following:
Maximum stress = 2031 psi Maximum displacement = 0.111 in. Maximum reaction = 55.8 lb
It is evident from the above discussion that the simplified analysis is conservative.
3.7.3.11 Torsional Effects of Eccentric Masses
The effect of eccentric masses such as valve operators is discussed in Section
3.7.3.8.
3.7.3.12 Buried Seismic Category I Piping Systems and Tunnels
Buried Seismic Category I pipes are laid in a prepared trench and backfilled with select material. The backfill material is compacted to 95% of maximum density as determined by the AASHO T180 Method "D". Field quality control is performed in
accordance with AASHO T147.
Where Seismic Category I piping enters the building near the base, differential movement between the building and soil at the location of pipe penetrations may be considered to be zero. Where Seismic Category I piping enters the secondary containment near the ground surface, flexible or rigid seals are provided as required.
Seismic Category I piping is designed for the maximum relative differential movement that could occur at the support points. Because the stresses resulting from this maximum relative differential movement are not likely to occur in phase with the maximum stresses due to dynamic response of the pipe, if any, the two were combined on a root-mean-square basis.
UFSAR/DAEC-1 3.7-37 Revision 14 - 11/98 The seismic stresses are combined in acco rdance with the design rules of Section 3.8.
3.7.3.13 Seismic Analysis for Reactor Internals
General Electric used the response spectrum method for seismic analyses of piping systems and reactor vessel and internals. The resultant displacements, loads (namely shears and moments), and stresses were computed by the square root of the sum of the square of the corresponding individual model responses (displacement, loads, or
stresses).
A complete discussion of the seismic analysis of the reactor vessel and internals is
presented in Section 3.9.5.2.3.
3.7.3.14 Analysis Procedure for Damping
Damping analysis is discussed in Section 3.7.2.14. UFSAR/DAEC-1 3.7-38 Revision 14 - 11/98 REFERENCES FOR SECTION 3.7
- 1. Whitman, R. V., and Richart, R. E., Jr., "Design Procedures for Dynamically Loaded Foundations," J. Soil Mech. and Fnd. Div., ASCE, SM6, November, 1967.
- 2. Biggs, J. M., "Introduction to Structural Dynamics," McGraw-Hill, 1964.
- 3. Hurty, W. C. and Rubinstein, M. F., "Dynamics of Structures," Prentice-Hall, 1964.
- 4. Letter from L. D. Root, Iowa Electric, to J. G. Keppler, NRC, Region III,
Subject:
Final Response to IE Bulletin No. 79-14 Concerning Seismic Analyses for As-built Safety-related Piping Systems, da ted October 17, 1980 (LDR-80-284).
- 5. Roark, R. J., "Formulas for Stress and Strain," McGraw-Hill, 1954.
- 6. Holtec Report, HI-92889, Licensing Report for Spent Fuel Pool Storage Capacity Expansion DAEC, Transmitted to NRC along with RTS-252, NG-93-0566, dated March 26, 1993.
UFSAR/DAEC-1 T3.7-1 Revision 14 - 11/98 Table 3.7-1 DAMPING VALUES a Percent of Critical Damping Operating-Basis Earthquake Design-Basis Earthquake Containment structure and all internal
concrete structures 2.0 5.0 Other conventionally reinforced
concrete structures; such as shear walls or rigid frames 5.0 5.0 Welded structural steel assemblies 1.0 1.0
Bolted or riveted steel assemblies 2.0 2.0 b Piping systems 0.5 1.0
Foundations, rock or lean concrete backfill, soil 5.0 5.0 a The damping values listed are the damping values originally used. The damping values found in NRC Regulatory Guide 1.61 are also acceptable for use in seismic analyses. b Seismic analysis was performed on certain cab le trays (1L3A, 1M3A, 1M5A and 1N5A) after fireproofing was added in response to 10 CFR 50, Appendix R. A damping value of 5% for a DBE was used for the analysis. UFSAR/DAEC-1 T3.7-2 Revision 14 - 11/98 Table 3.7-2 EXAMPLE OF BUILDING ANALYSIS NATURAL PERIODS OF VI BRATION AND ASSIGNED DAMPING VALUES FOR SIGNIFICANT REACTOR BUILDING MODES Mode Number Period of Vibration (sec) Damping Ratio (%) 1 0.288 2
2 0.228 5
3 0.120 1
4 0.072 5
5 0.059 2
6 0.047 2
7 0.043 2
8 0.036 2
9 0.032 2
UFSAR/DAEC-1 T3.7-3 Revision 14 - 11/98 Table 3.7-3 EXAMPLE OF BUILDING ANALYSIS MODAL RESPONSE, REACTOR BUILDING Mass Point Elevation Acceleration (g units) Displacement (in.) Shear (kips) Moment (ft-kips) 1 0.500 0.581 - - 0 2 0.165 0.122 561 23600 3 0.141 0.099 2580 81000 4 0.117 0.078 4920 176000 5 0.090 0.050 6960 362000 6 0.080 0.017 8250 598000 7 0.070 0.002 8780 955000 Base 0.070 0.002 11240 1070000
UFSAR/DAEC-1 T3.7-4 Revision 14 - 11/98 Table 3.7-4 EXAMPLE OF BUILDING ANALYSIS MODAL RESPONSE, DRYWELL Mass Point Elevation Acceleration (g units) Displacement (in.) Shear (kips) Moment (ft-kips) 8 0.292 0.117 - - 0 9 0.179 0.102 17 249 10 0.148 0.098 26 354 11 0.133 0.089 30 586 12 0.110 0.070 450 4820 13 0.103 0.049 454 9360 14 0.102 0.045 457 11940 15 0.094 0.034 461 16460 16 0.092 0.023 465 21550 17 0.081 0.011 466 26760 18 0.077 0.006 468 31400 7 0.070 0.002 468 35600
UFSAR/DAEC-1 T3.7-5 Revision 14 - 11/98 Table 3.7-5 EXAMPLE OF BUILDING ANALYSIS MODAL RESPONSE, REACTOR PRESSURE VESSEL Mass Point Elevation Acceleration (g units) Displacement (in.) Shear (kips) Moment (ft-kips) 19 0.212 0.115 - - 0 20 0.181 0.107 15 115 - - 0.157 0.091 115 1870 21 0.137 0.083 209 1550 22 0.138 0.077 125 1980 23 0.135 0.069 75 2520 24 0.128 0.060 36 2690 25 0.132 0.050 47 2520 26 0.112 0.038 70 1980 UFSAR/DAEC-1 T3.7-6 Revision 14 - 11/98 Table 3.7-6 EXAMPLE OF BUILDING ANALYSIS MODAL RESPONSE, SACRIFICIAL SHIELD Mass Point Elevation Acceleration (g units) Displaceme nt (in.) Shear (kips) Moment (ft-kips) - - 0.149 0.090 - - 0 27 0.148 0.082 63 687 28 0.144 0.072 20 820 29 0.133 0.057 30 668 26 0.112 0.038 63 444
Table 3.7-7 EXAMPLE OF BUILDING ANALYSIS MODAL RESPONSE, PEDESTAL Mass Point Elevation Acceleration (g units) Displaceme nt (in.) Shear (kips) Moment (ft-kips) 26 0.112 0.038 - - 2360 30 0.098 0.027 154 1210 31 0.082 0.017 179 1460 32 0.077 0.007 197 2850 7 0.070 0.002 208 4300
UFSAR/DAEC-1 T3.7-7 Revision 13 - 5/97 Table 3.7-8 (a) Sheet 1 of 2 STRESS
SUMMARY
FOR NEW FUEL AND PaR SPENT FUEL RACKS Stress (psi) Criteria Loading Condition Location Allowable Calculated a New fuel storage racks Emergency: Column 16,000 2,950 Stresses due to normal, upset, or emergency
loading shall not cause
the racks to fail so as to
result in a critical fuel array Dead loads
Full fuel load in rack
Design-basis
earthquake Base to column welds 11,000 100 Primary stress limit Paper number 3341 and
3342, Proceedings of the
ASCE, Journal of the Structural Division , December 1962 (task committee on light-weight alloys) Channel Support channel to column weld 20,000 6,000 3,150 2,650 PaR Spent fuel storage racks b Emergency "A" At column to base welds 11,000 760 Stresses due to normal, upset, or emergency
loading shall not cause
the racks to fail so as to
result in a critical fuel array Dead loads
Full fuel load in rack
Design-basis
earthquake
Support beam
35,000
33,500 Primary stress limits Paper numbers 3341 and
3342, Proceedings of the
ASCE, Journal of the Structural Division , December 1962 (task committee on lightweight alloys) a These values are calculated for 1.5g static seismic coefficient applied horizontally to the rack. Actual earthquake loads give much lower than these due to the structural stiffness. b Applies to original spent fuel storage racks that were replaced with high-density, PaR racks. UFSAR/DAEC-1 T3.7-8 Revision 13 - 5/97 Table 3.7-8 (a) Sheet 2 of 2 STRESS
SUMMARY
FOR NEW FUEL AND PaR SPENT FUEL RACKS Stress (psi) Criteria Loading Condition Location Allowable Calculated a Emergency conditions Emergency "B" Stress limit = yield
Strength at 0.2% offset Note: Emergency condition "B":
- 1. Loading. In addition to testing the capability of the racks to withstand the loading conditions given in this table, the racks were tested and analyzed to determine their capability to safely withstand the accidental, uncontrolled drop of the fuel grapple from its full retracted position
into the weakest portion of the rack.
- 2. Method of Analysis. The displacement of the vertical column at the ends of the racks was determined by considering the effect of the grapple kinetic energy on the upper structure.
The energy absorbed in the shearing of the rack longitudinal structural member welds was determined. The effect of the remaining energy on the vertical columns was analyzed. Equivalent static load tests were made on the structure to ensure that the criteria were met.
- 3. Results of Analysis. All criteria were met. Analysis showed that the grapple would shear the welds in the area where the impact occurred. The longitudinal structural member bends, but
it does not fail in shear. Grapple penetration into the rack is not sufficient to cause the vertical columns to deflect the fuel into a critical array. Static load testing showed that forces in excess of those resulting from a grapple drop are required to cause the columns to deflect
to the extent that the criteria are violated.
a These values are calculated for 1.5g static seismic coefficient applied horizontally to the rack. Actual earthquake loads give much lower than these due to the structural stiffness. UFSAR/DAEC-1 T3.7-9 Revision 13 - 5/97 Table 3.7-8 (b) Sheet 1 of 2 STRESS
SUMMARY
FOR HOLTEC SPENT FUEL STORAGE RACKS (REF. 6) Stress (psi) Criteria Loading Condition Location Allowable Calculated a Holtec Spent Fuel Storage Racks Emergency "C" Rack to Baseplate Weld 29820 9277 Dead Loads Baseplate to Pedestal Weld (Dimensionless Limit Load Ratio) 1.0 .283 Full Fuel Load in rack Cell to Cell Weld 5271 1222 Design Basis Earthquake Bearing Pad 2380 719 Primary Stress Limit ASME SA-240-304 (Rack Mat'l), ASTM-240, Type 304 (upper part of support feet), ASTM 564-630 (lower part of support feet, age hardened at 1100 °F) Emergency Conditions Emergency "D" Stress Limit = See Note ASME Code, Section III Subsection NF, Yield Strength at 200 °F (Max. Pool Temp.)
a These values are calculated for 1.5g static seismic coefficient applied horizontally to the rack. Actual earthquake loads give much lower than these due to the structural stiffness. UFSAR/DAEC-1 T3.7-10 Revision 13 - 5/97 Table 3.7-8 (b) Sheet 2 of 2 STRESS
SUMMARY
FOR HOLTEC SPENT FUEL STORAGE RACKS (REF. 6) ____________________ These values are based on dynamic analysis using design base earthquake (DBE) response spectrums (horizontal and Vertical). DBE response spectrums were generated by multiplying by two the operating base earthquake (OBE) response spectrums of the Reactor Building floor elevation 812'. The values shown are extracted form the largest rack in the spent fuel pool. Note: Emergency Condition "D":
- 1. Loading. In addition to testing the capability of the racks to withstand the loading conditions given in this table, the free standing Holtec spent fuel storage racks are also analyzed to determine their capability to safely withstand the drop of fuel assembly (680 lbs.) with associated handling equipment (120 lbs.), being carried 18" above the spent fuel rack.
- 2. Method of Analysis. An 800 lbs. fuel assembly plus handling equipment is dropped from 18 " above the top of a storage location and impacts the base of the module. The rack design should ensure that gross structural failure does not occur and the subcriticality of the adjacent fuel assemblies is not violated.
- 3. Results of Analysis. Calculated results show that there will be no change in the spacing between cells. Local deformation of the baseplate will in the neighborhood of the impact will occur, but the dropped assembly will be contained and not impact the liner. In the case when the fuel assembly with the channel is dropped from 18" above the top of the rack, and impacts the top of the rack, it is shown that the damage, if it occur, will be restricted to a depth of less than or equal to 1.09" below the top of the rack. This is above the active fuel region. Analysis of the Local buckling of the fuel cell walls, and in-rack welded joints shows that the maximum stresses are within allowable limits.
UFSAR/DAEC-1 T3.7-11 Revision 14 - 11/98 Table 3.7-9 Sheet 1 of 2 STRESS
SUMMARY
FOR RHR PUMP Criteria Method of Analysis Allowable Stress or Minimum Thickness Required Calculation Closure bolting Bolting loads and stresses calculated per "Rules for Bolting
Flange Connections," ASME
Code, Section VIII, Appendix II Maximum allowable stress = 20,000 psi Maximum calculated stress = 9990 psi Loads: Normal and upset
- Design pressure and temperature, design gasket
load Bolting stress limit
- Allowable working stress per
ASME Code, Section VIII Wall thickness Per rules of ASME Code, part UG, Section VIII Maximum allowable stress - main pump =
14,000 psi Maximum calculated stress = 12,154 psi Loads: Normal and upset
- Design pressure and temperature Stress limit
- ASME Code, Section III Nozzle For the maximum moment due to pipe reaction, the maximum
force shall not exceed the
allowable Force in lb
Moment in ft-lb Loads: Normal plus upset
- Design pressure and temperature, deadweight, thermal expansion, and
operating-basis earthquake Total nozzle stress with these
criteria does not exceed stress limits Normal plus upset
- Suction F= 83,000-1.7M Normal plus upset
- Suction Force = 7,376 lb Moment = 13,894 ft-lb UFSAR/DAEC-1 T3.7-12 Revision 14 - 11/98 Table 3.7-9 Sheet 2 of 2 STRESS
SUMMARY
FOR RHR PUMP Criteria Method of Analysis Allowable Stress or Minimum Thickness Required Calculation Normal and upset (Continued): Discharge
F = 62,000-2.23M Discharge Force = 4985 lb Moment = 14,329 ft-lb Loads: Emergency: Design pressure and temperature, deadweight, thermal expansion, and design-
basis earthquake Emergency
Suction F = 125,000- 1.7M Emergency Suction Force = 9,900 lb Moment = 19,563 ft-lb Stress limit
- ASME Code, Section VIII, for normal and upset, 1.5 of allowable stress for emergency Discharge
F = 92,000 - 2.23M Discharge Force = 7116 lb Moment = 18,609 ft-lb
UFSAR/DAEC-1 T3.7-13 Revision 13 - 5/97 Table 3.7-10 Sheet 1 of 2 STRESS
SUMMARY
FOR RHR HEAT EXCHANGER Criteria Method of Analysis Allowable Stress or Minimum Thickness Required Calculation Closure bolting Bolting loads and stresses calculated per "Rules for Bolted
Flange Connections," ASME
Code, Section VIII, Appendix II
- a. Shell cover bolts
- b. Channel cover bolts
25,000 psi
25,000 psi
23,700 psi
23,900 psi Loads: Normal and upset
- Design pressure and temperature, design gasket
load Bolting stress limit
- Allowable working stress per
ASME Code, Section VIII Wall thickness Shell side, ASME Code, Section III, and TEMA Class A Loads: Normal and upset
- Design pressure and temperature Tube side, ASME Code, Section VIII, and TEMA Class C Stress limit
- ASME Code, Section VIII
- a. Shell
- b. Shell cover
- c. Channel ring
- d. Tubes
- e. Channel cover
- f. Tube sheet 0.6478 in.
0.6407 in
0.6478 in.
0.0548 in.
5.1857 in
5.2815 in. 0.875 in. 0.875 in.
1.000 in.
18 BWG 5.500 in.
5.500 in. UFSAR/DAEC-1 T3.7-14 Revision 13 - 5/97 Table 3.7-10 Sheet 2 of 2 STRESS
SUMMARY
FOR RHR HEAT EXCHANGER Criteria Method of Analysis Allowable Stress or Minimum Thickness Required Calculation Nozzle For the maximum moment due to pipe reaction, the maximum
force shall not exceed the
allowable Force in lb
Moment in ft-lb (maximum in any
direction) Loads: Normal plus upset
- Design pressure and temperature, deadweight, thermal expansion and
operating-basis earthquake Primary stress less than 1.5S m and secondary stress less than
3.0S m Normal plus upset
N1 F = 48,000-1.1M
N2 F = 48,000 - 1.1M
N3 F = 45,000-0.94M
N4 F = 62,000 - 1.0M
Force = later Moment = later
Force = later Moment = later
Force = 6430 lb Moment = 15,257 ft-lb
Force = 3313 lb Moment = 14,043 ft-lb Loads: Emergency: Design pressure and temperature, deadweight, thermal expansion, and design-
basis earthquake Primary stress less than 1.8S m Emergency
N1 F = 74,000 -0.74M
N2 F = 74,00 - 0.74M
Emergency:
Force = later Moment = later
Force = later Moment = later Stress limit
- ASME Code, Section VIII N3 F = 72,000 - 0.6M
N4 F = 108,000 - 0.68M Force = 7378 lb Moment = 20,390 ft-lb
Force = 4959 lb Moment = 19,339 ft-lb UFSAR/DAEC-1 T3.7-15 Revision 14 - 11/98 Table 3.7-11 Sheet 1 of 2 STRESS
SUMMARY
FOR CORE SPRAY PUMP Criteria Method of Analysis Allowable Stress or Minimum Thickness Required Calculation Closure bolting Bolting loads and stresses calculated per "Rules for Bolted
Flange Connections," ASME
Code, Section VIII, Appendix II Maximum allowable stress = 20,000 psi Maximum calculated stress = 9030 psi Loads: Normal and upset
- Design pressure and temperature, design gasket
load Bolting stress limit
- Allowable working stress per
ASME Code, Section VIII Wall thickness Per rules of ASME Code, Part UG, Section VIII Maximum allowable stress - main pump =
14,000 psi Maximum calculated stress = 11,431 psi Loads: Normal and upset
- Design pressure and temperature Stress limit
- ASME Code, Section III Nozzle For the maximum moment due to pipe reaction, the maximum
force shall not exceed the
allowable Force in lb
Moment in ft-lb Loads: Normal plus upset
- Design pressure and temperature, deadweight, thermal expansion, and operating-basis earthquake Total nozzle stress with these criteria does not exceed stress limits Normal plus upset
Suction F = 81,000 - 1.56M Normal plus upset
Suction Force = 1437 lb Moment = 6545 ft-lb UFSAR/DAEC-1 T3.7-16 Revision 14 - 11/98 Table 3.7-11 Sheet 2 of 2 STRESS
SUMMARY
FOR CORE SPRAY PUMP Criteria Method of Analysis Allowable Stress or Minimum Thickness Required Calculation Discharge
F = 61,000 - 2.23M Discharge Force = 1497 lb Moment = 2789 ft-lb Loads: Emergency: Design pressure and temperature, deadweight, thermal expansion, and design-
basis earthquake Emergency
Suction F = 122000 - 1.56M
Discharge F = 92,000 - 2.23M Emergency: Suction Force = 1716 lb Moment = 8207 ft-lb
Discharge Force = 2879 lb Moment = 5640 ft-lb Stress limit
- ASME Code, Section VIII, for normal and upset; 1.5 of allowable stress for emergency UFSAR/DAEC-1 T3.7-17 Revision 13 - 5/97 Table 3.7-12 Sheet 1 of 3 STRESS
SUMMARY
FOR RCIC TURBINE Criteria Method of Analysis Allowable Stress or Minimum Thickness Required Calculation Closure bolting Bolting loads and stresses calculated per "Rules for Bolted
Flange Connections," ASME
Code, Section VIII, Appendix II Maximum allowable stress = 20,000 psi Maximum calculated stress = 6400 psi Loads: Normal and upset: Design pressure and temperature, design gasket
load Bolting stress limit: Allowable working stress per
ASME Code, Section VIII Casing wall thickness Per rules of ASME Code, Part UG, Section VIII Maximum allowable stress = 17,500 psi Maximum calculated stress = 12,700 psi Loads: Normal and upset: Design pressure and temperature Stress limit: ASME Code, Section III Nozzle For the resultant moment due to pipe reaction, the resultant force
shall not exceed the allowable Force in lb Moment in ft-lb Loads: Normal: Design pressure and temperature, deadweight, and thermal expansion Detailed design analysis has demonstrated the limits Inlet F = (2620 - M)/3
Exhaust F = (6000 - M)/3 Inlet Force = 470 lb Moment = 1797 ft-lb
Exhaust Force = 356 lb Moment = 2527 ft-lb UFSAR/DAEC-1 T3.7-18 Revision 13 - 5/97 Table 3.7-12 Sheet 2 of 3 STRESS
SUMMARY
FOR RCIC TURBINE Criteria Method of Analysis Allowable Stress or Minimum Thickness Required Calculation Loads: Normal plus upset
- Design pressure and temperature, deadweight, thermal expansion, and
operating-basis earthquake Normal plus upset
Inlet F = (3000 - M)/2.5
Exhaust F = 3(6000 - M), but
not >8370 lb Normal plus upset
Inlet Force = 633 lb Moment = 2679 ft-lb
Exhaust Force = 849 lb Moment = 3357 ft-lb Loads: Emergency: Design pressure and temperature, deadweight, thermal expansion, and design-
basis earthquake Emergency
Inlet F = (4500 - M)/2.5
Exhaust F = 3(9000 - M), but
not >12,555 lb Emergency: Inlet Force = 803 lb Moment = 3586 ft-lb
Exhaust Force = 1355 lb Moment = 8376 ft-lb Stress limit
- Specified by vendor for normal
loads; ASME Code, Section
VIII, for upset loads; increased 20% for emergency loads Turbine mounting bolt (turbine to baseplate)
Vertical and horizontal forces on mounting bolts calculated as the sum of seismic accelerations on
the turbine and the pipe reaction forces and moments on the
nozzles UFSAR/DAEC-1 T3.7-19 Revision 13 - 5/97 Table 3.7-12 Sheet 3 of 3 STRESS
SUMMARY
FOR RCIC TURBINE Criteria Method of Analysis Allowable Stress or Minimum Thickness Required Calculation Loads: Normal and upset
- Operating-basis earthquake, nozzle loads for operating-
basis earthquake, deadweight, and thermal expansion Tensile and shear stress for bolting materials
are specified in ASME
Code, Section VIII By meeting the nozzle load criteria above, the detailed seismic analyses indicate the mounting bolts satisfy
the allowable stress requirements Loads: Emergency: Design-basis earthquake, nozzle loads for design-basis
earthquake, deadweight, and thermal expansion Tensile stress less than 0.9 yield and shear
stress less than twice allowable by ASME
Code, Section VIII By meeting the nozzle load criteria above, the detailed seismic analyses indicate the mounting bolts satisfy
the allowable stress requirements Stress limits
- ASME Code, Section VIII, allowables for normal and upset loads; for emergency loads 0.9 yield and twice
allowable shear UFSAR/DAEC-1 T3.7-20 Revision 13 - 5/97 Table 3.7-13 Sheet 1 of 2 STRESS
SUMMARY
FOR RCIC PUMP Criteria Method of Analysis Allowable Stress or Minimum Thickness Required Calculation Closure bolting: Bolting loads and stresses calculated per "Rules for Bolted
Flange Connections," ASME
Code, Section VIII, Appendix II Maximum allowable stress = 25,000 psi Maximum calculated stress = 19,288 psi Loads: Normal and upset
- Design pressure and temperature, design gasket
load Bolting stress limit
- Allowable working stress per
ASME Code, Section VIII Wall thickness Per rules of ASME Code, Part UG, Section VIII Maximum allowable stress - main pump =
14,000 psi Maximum calculated stress = 11,960 psi Loads: Normal and upset
- Design pressure and temperature Stress limit
- ASME Code, Section III Volute stress is calculated per Roarck 5 Maximum allowable stress - main pump =
14,000psi Maximum calculated stress = 11,913 psi Nozzle For the maximum moment due to pipe reaction, the maximum
force shall not exceed the
allowable Force in lb Moment in ft-lb Loads: Normal plus upset
- Design pressure and temperature, deadweight, thermal expansion, and
operating-basis earthquake Total nozzle stress with these criteria does not exceed limits Normal plus upset
Suction F = 9400 - 2.50M Normal plus upset
Suction Force = 1029 lb Moment = 2916 ft-lb UFSAR/DAEC-1 T3.7-21 Revision 19 - 9/07 Table 3.7-13 Sheet 2 of 2 STRESS
SUMMARY
FOR RCIC PUMP Criteria Method of Analysis Allowable Stress or Minimum Thickness Required Calculation Discharge F = 9400 - 4.33M Discharge Force = 756 lb Moment = 1231 ft-lb Loads: Emergency: Design pressure and temperature, deadweight, thermal expansion, and design-
basis earthquake Emergency
Suction F = 19,000 - 2.42M
Discharge F = 19,000 - 5.05M Emergency: Suction Force = 1949 ft Moment = 5594 ft-lb
Discharge Force = 1211 lb Moment = 1358 ft-lb Stress limits
- ASME Code, Section VIII, for normal and upset loads; 1.5 allowable stress for emergency
loads UFSAR/DAEC-1 T3.7-22 Revision 13 - 5/97 Table 3.7-14 Sheet 1 of 3 STRESS
SUMMARY
FOR HPCI TURBINE Criteria Method of Analysis Allowable Stress or Minimum Thickness Required Calculation Closure bolting Bolting loads and stresses calculated per "Rules for Bolted
Flange Connections," ASME
Code, Section VIII, Appendix II Maximum allowable stress = 20,000 psi Maximum calculated stress = 18,290 psi Loads: Normal and upset
- Design pressure and temperature, design gasket
load Bolting stress limit
- Allowable working stress per
ASME Code, Section VIII Casing wall thickness Per rules of ASME Code, Part UG, Section VIII Maximum allowable stress = 17,500 psi Maximum calculated stress = 7200 psi Loads: Normal and upset
- Design pressure and temperature Stress limit
- ASME Code, Section III Nozzle For the resultant moment due to pipe reaction , the resultant force
shall not exceed the allowable Force in lb Moment in ft-lb Loads: Normal: Design pressure and temperature, deadweight, and thermal expansion Detailed design analysis had demonstrated the acceptability of
these values Normal: Inlet F = (7570 - M)/3 Normal: Inlet Force = 1760 lb Moment = 6219 ft-lb UFSAR/DAEC-1 T3.7-23 Revision 19 - 9/07 Table 3.7-14 Sheet 2 of 3 STRESS
SUMMARY
FOR HPCI TURBINE Criteria Method of Analysis Allowable Stress or Minimum Thickness Required Calculation Exhaust F = (9930 - M)/3 Exhaust Force = 1,476 lb Moment = 4,684 ft-lb Loads: Normal plus upset
- Design pressure and temperature, and operating-
basis earthquake Normal plus upset
Inlet F = (20,000 - M)/2.5, but not > 5000 lb
Exhaust F = (20,000 - M)/0.8, but not > 11,500 lb Normal plus upset
Inlet Force = 2872 lb Moment = 12,496 ft-lb
Exhaust Force = 2,611 lb Moment = 11,763 ft-lb Loads: Emergency: Design pressure and temperature, deadweight, thermal expansion, and design-
basis earthquake Emergency
Inlet F = (30,000- M )/2.5, but not > 7500 lb
Exhaust F = (30,000 - M)/0.8, but not > 17,250 lb Emergency Inlet Force = 4016 lb Moment = 20,153 ft-lb
Exhaust Force =4,586 lb Moment = 17,150 ft-lb Stress limit
- Specified by vendor for normal
loads; ASME Code, Section
VIII, for upset loads; increased 20% for emergency loads Turbine mounting bolt (turbine to baseplate) Vertical and horizontal forces on mounting bolts calculated as the sum of seismic acceleration on
the turbine and the pipe reaction forces and moments on the
nozzle UFSAR/DAEC-1 T3.7-24 Revision 13 - 5/97 Table 3.7-14 Sheet 3 of 3 STRESS
SUMMARY
FOR HPCI TURBINE Criteria Method of Analysis Allowable Stress or Minimum Thickness Required Calculation Loads: Normal and upset
- Operating-basis earthquake, nozzle loads for operating-
basis earthquake, deadweight, and thermal expansion Tensile and shear stress for bolting materials
are specified in ASME
Code, Section VIII By meeting the nozzle load criteria above, the detailed seismic analysis indicate the mounting bolts satisfy
the allowable stress requirements Loads: Emergency: Design-basis earthquake, nozzle loads for design-basis
earthquake, deadweight, and thermal expansion (Same as for normal and upset loads above.) Stress limit
- ASME Code, Section VIII, allowables for normal and upset loads; for emergency loads, 0.9 yield and twice
allowable shear UFSAR/DAEC-1 T3.7-25 Revision 19 - 9/07 Table 3.7-15 Sheet 1 of 2 STRESS
SUMMARY
FOR HPCI PUMP Criteria Method of Analysis Allowable Stress or Minimum Thickness Required Calculation Closure bolting Bolting loads and stresses calculated per "Rules for Bolted
Flange Connections," ASME
Code, Section VIII, Appendix II Maximum allowable stress - main pump =
20,000 psi; booster pump = 20,000 psi Maximum calculated stress - main pump =
16,960 psi; booster pump = 7180 psi Normal and upset
- Design pressure and temperature, design gasket
load Bolting stress limit
- Allowable working stress per
ASME Code, Section VIII Wall thickness Per rules of ASME Code, Part UG, Section VIII Maximum allowable stress - main pump =
14,000 psi; booster pump = 14,000 psi Maximum calculated stress - main pump = 8700 psi; booster pump
= 3360 psi Loads: Normal and upset: Design pressure and temperature Nozzle stress Volute stress is calculated per
Roarke 5 Main pump = 14,000 psi; booster pump =
14,000 psi Main pump = 7840 psi; booster pump = 2610
psi Stress limit
- ASME Code, Section III Nozzle: For the moment due to pipe reaction, the maximum force
shall not exceed the allowable Force in lb Moment in ft-lb
UFSAR/DAEC-1 T3.7-26 Revision 21 - 5/11 Table 3.7-15 Sheet 2 of 2 STRESS
SUMMARY
FOR HPCI PUMP Criteria Method of Analysis Allowable Stress or Minimum Thickness Required Calculation Loads: Normal and upset
- Normal and upset
- Normal and upset
- Design pressure and temperature, deadweight, thermal expansion, and
operating-basis earthquake Total nozzle stress with these criteria does not exceed stress limits Suction F = 21,000 - 1.83M
Discharge F = 23,000 - 3.17M, but not > 11,500 lb Suction Force = 1738 lb Moment = 3,284 ft-lb
Discharge Force = 947 lb Moment = 4,526 ft-lb Loads: Emergency: Design pressure and temperature, deadweight, thermal expansion, and design-
basis earthquake
Stress limit
- ASME Code, Section VIII, for normal and upset loads; 1.5 allowable stress for emergency
loads Emergency
Suction F = 28,750 - 1.83M
Discharge F = 34,00 - 3.21M Emergency: Suction Force = 2775 lb Moment = 4,813 ft-lb
Discharge Force = 1,077 lb Moment = 4,878 ft-lb
UFSAR/DAEC-1 T3.7-27 Revision 13 - 5/97 Table 3.7-16 Sheet 1 of 4 STRESS
SUMMARY
FOR MAIN STEAM ISOLATION VALVES
Criteria
Method of Analysis Allowable Stress or Minimum Thickness Required Calculated Stress or Actual Thickness Minimum body well thickness Minimum wall thickness in cylindrical portions of the valve
shall be calculated using the following formula: t = 1.468 in t = 1.593 in at 18.155 in diameter Loads: Design pressure and temperature t Pd SP C=+15212.. Primary membrane stress limit S = 7000 lb/in 2 per ASA B16.5 where:
S = allowable stress of 7000 psi P = primary service pressure, 655 psi d = inside diameter of valve at section being considered, in. C = corrosion allowance of 0.12 in. Minimum cover thickness tG CP S Wh SG C G=++178 3 12 1./ t = 4.37 in t = 5.00 in. S allow = 17500 psi Loads Design pressure and temperature, design bolting load, gasket load where: t = minimum thickness, in. G = diameter or short span, in.
UFSAR/DAEC-1 T3.7-28 Revision 13 - 5/97 Table 3.7-16 Sheet 2 of 4 STRESS
SUMMARY
FOR MAIN STEAM ISOLATION VALVES
Criteria
Method of Analysis Allowable Stress or Minimum Thickness Required Calculated Stress or Actual Thickness Primary stress limit: Allowable working stress per
ASME Code, Section VIII UG-
32(c)(2) 1971 Edition C = attachment factor S = allowable stress, psi
W = total, bolt load, lb
h G = gaskets, moment arm, in. C 1 = corrosion allowance, in. Cover flange bolt area
Loads Design pressure and temperature, gasket load, stem operational load, seismic load (design-basis
earthquake)
Bolting stress limit
- Allowable working stress per
ASME Standard Code for Pumps and Valves for Nuclear
Power In accordance with paragraph F.105.2.6 of ANSI B31.7, the maximum value of bolt stress which results from preload
operating pressure and differential thermal expansion shall not exceed
twice the allowable value listed in Table A.1, except that the stem operational load and seismic loads
shall be included in the total load carried by bolts. The horizontal and vertical seismic forces shall be applied at the mass center of the valve operator, assuming that the valve body is rigid and anchored. S = 67,000 psi at 575°F A B = 16.735 in. 2 S b = 45,000 psi Body flange thickness and stress
Loads: Design pressure and temperature, gasket load, stem operational load, seismic load (design-basis
earthquake)
Flange stress limits (S H , S R , S T): The allowable flange stresses are obtained from paragraph UA-52 Flange thickness and stress shall be calculated in accordance with
"Rules for Bolted Flange
Connections," ASME Code, Section VIII, Appendix II, except that the stem operational load and seismic loads shall be included in the total load carried by the flange. The horizontal and vertical seismic force shall be applied at the mass
center of the valve operator, assuming that the valve body is
rigid and anchored. S H = 26,200 psi
S R = 17,500 psi
S T = 17,500 psi t = 4.12 in. S H = 20,300 psi S R = 7,100 psi S T = 8,400 psi
UFSAR/DAEC-1 T3.7-29 Revision 13 - 5/97 Table 3.7-16 Sheet 3 of 4 STRESS
SUMMARY
FOR MAIN STEAM ISOLATION VALVES
Criteria
Method of Analysis Allowable Stress or Minimum Thickness Required Calculated Stress or Actual Thickness Valve disk Thickness Loads: Design pressure and temperature
Primary handling stress limit
- ASME Section III 1986
subsection NG-3200 S m = 18,200 psi 575°F 1.5S m = 27,300 psi 575°F Cylinder Flange Interface
1,250psi
P L = 6,500 psi P L +P b = 11,800 psi Cylinder/Sphere
Interface 1,250
psi
P L = 9,000 psi P L +P b = 18,800 psi
Stem Disk
Seating Area
1,250 psi
P L = 7,000 psi P L +P b = 27,000 psi
Disk Piston
Main Seat 1,250
psi
P L = 5,600 psi P L +P b = 7,500 psi UFSAR/DAEC-1 T3.7-30 Revision 13 - 5/97 Table 3.7-16 Sheet 4 of 4 STRESS
SUMMARY
FOR MAIN STEAM ISOLATION VALVES
Criteria
Method of Analysis Allowable Stress or Minimum Thickness Required Calculated Stress or Actual Thickness Valve operator supports: The valve assembly shall be analyzed assuming that the valve body is an anchored, rigid mass
and that the specified vertical and horizontal seismic forces are applied at the mass center of the operator assembly simultaneously
with operating pressure plus
deadweight plus operational loads Loads: Design pressure and temperature, stem operational load, equipment deadweight, seismic load (design-basis earthquake)
90% Tensile ASTM minimum yield S = 90,000 psi S = 49,500 psi (combined
bending and
tensile stress)
UFSAR/DAEC-1 T3.7-31 Revision 13 - 5/97 Table 3.7-17 Sheet 1 of 5 STRESS
SUMMARY
FOR MAIN RELIEF VALVES a Criteria
Method of Analysis Allowable Stress or Minimum Thickness Required Calculated Stress or Actual Thickness Minimum body wall thickness t Pd SP C=+15212.. t = 1.253 in. t = 1.5 in. Loads: Design pressure and temperature where:
Primary membrane stress limit
- Allowable working stress as defined by USAS B16.5 (7000 psi at primary service pressure) T = minimum required thickness, in.
S = allowable stress 7000 psi P = primary service pressure, 655
psi d = inside diameter of valve at
section being considered, in.
C = corrosion allowance of 0.12 in. Top flange Discontinuity analysis based on ASME Code, Section VIII, Paragraph UA 99 S m = 17,800 psi r = 7,335 psi r = 727 psi Loads: Design pressure and temperature, gasket load
Primary stress limit
- Allowable stress intensity, S m , as defined by ASME Standard Code for Pumps and Valves for
Nuclear Power a This table applies to the original Dresser Industries, Inc., main steam relief valves which were replaced in 1977 with valves manufactured by the Target Rock Corporation. UFSAR/DAEC-1 T3.7-32 Revision 13 - 5/97 Table 3.7-17 Sheet 2 of 5 STRESS
SUMMARY
FOR MAIN RELIEF VALVES
Criteria
Method of Analysis Allowable Stress or Minimum Thickness Required Calculated Stress or Actual Thickness Flange bolt area inlet flange, outlet flange, body to top flange
Loads: Design pressure and temperature, gasket load, operational load, and design-basis earthquake
Bolting stress limit
- Allowable stress intensity, S m , as defined by ASME Standard Code for Pumps and Valves for
Nuclear Power Total bolting loads and stresses shall be calculated in accordance
with procedures of Paragraph 1-
704.5.1, Flange Joints, of ANSI
B31.7 Nuclear Piping Code
A b = total bolt area in. 2 Body to top flange A b = 10.41 in. 2 Inlet flange A b = 9.88 in. 2 Outlet flange A b = 5.88 in. 2 Body to top flange A b = 13.86 in. 2 Inlet flange A b = 13.86 in. 2 Outlet flange A b = 8.8 in. 2 Flange thickness - inlet, outlet
Loads: Design pressure and temperature, gasket load, operational loads, and design-basis earthquake Flange thickness and stress shall be calculated in accordance with
procedures of Paragraph 1-704.5.1, Flanged Joints, of ANSI B31.7
Nuclear Piping Code Flange stress limits (S H , S R , S T): 1.5S m per ASME Standard Code for Pumps and Valves for
Nuclear Power Inlet flange S H = 25,950 psi S R = 25,950 psi S T = 25,950 psi
Outlet flange S H = 25,950 psi S R = 25,950 psi S T = 25,950 psi S m = 32,000 psi Inlet flange S H = 17,404 psi S R = 6,217 psi S T = 9,478 psi
Outlet flange S H = 15,128 psi S R = 14,628 psi S T = 4,671 psi S = 26,207 psi UFSAR/DAEC-1 T3.7-33 Revision 13 - 5/97 Table 3.7-17 Sheet 3 of 5 STRESS
SUMMARY
FOR MAIN RELIEF VALVES
Criteria
Method of Analysis Allowable Stress or Minimum Thickness Required Calculated Stress or Actual Thickness Valve disk thickness and stress Maximum stress in disk shall be calculated by: Loads: Design pressure and temperature ()()()()()S P tabbmbm a babm am bm max log=++++++3 42141 1 1 1 22244 22 2 2 Primary stress limit
- S m per ASME Code, Section III, Appendix II where:
t = disk thickness
P = design pressure
a = outer radius of disk
b = radius of center guide m = reciprocal of Poisson's ratio
UFSAR/DAEC-1 T3.7-34 Revision 13 - 5/97 Table 3.7-17 Sheet 4 of 5 STRESS
SUMMARY
FOR MAIN RELIEF VALVES
Criteria
Method of Analysis Allowable Stress or Minimum Thickness Required Calculated Stress or Actual Thickness Inlet nozzle diameter thickness and stress S FF A MM PR t i=++++1212 22 S = 26,250 psi S = 6,168 psi Loads: Design pressure and temperature, operational load, and design-
basis earthquake
Primary stress limit
- 1.5 x allowable stress intensity, 1.5S m, as defined by ASME Standard Code for Pumps and
Valves for Nuclear Power where: S = combined bending and tensile stress, psi
F 1 = vertical loads due to design pressure, lb
F 2 = vertical component of reaction thrust, lb
A = cross section areas of nozzle, in.2 M 1 = moment resulting from horizontal reaction, in.-lb
M 2 = moment resulting from horizontal seismic force at mass
center of valve, in.-lb
P = design pressure, psi
R 1 = inside radius of nozzle T = wall thickness, in.
UFSAR/DAEC-1 T3.7-35 Revision 13 - 5/97 Table 3.7-17 Sheet 5 of 5 STRESS
SUMMARY
FOR MAIN RELIEF VALVES
Criteria
Method of Analysis Allowable Stress or Minimum Thickness Required Calculated Stress or Actual Thickness Body stresses at neck below top flange S FF A M z=++12 S = 26,250 psi S = 705 psi Loads: Design pressure and temperature, operational load, and design-
basis earthquake
Primary stress limit
- 1.5 x allowable stress intensity, 1.5S m, as defined by ASME Standard Code for Pumps and
Valves for Nuclear Power where = S = combined bending and tensile
strength, psi
F 1 = axial load due to design pressure
F 2 = axial load due to seismic acceleration on components
attached to neck
M = moment due to seismic acceleration on components
attached to neck
UFSAR/DAEC-1 T3.7-36 Revision 13 - 5/97 Table 3.7-18 Sheet 1 of 6 STRESS
SUMMARY
FOR MAIN STEAM SAFETY VALVES
Criteria
Method of Analysis Allowable Stress or Minimum Thickness Required Calculated Stress or Actual Thickness Inlet nozzle wall thickness t PRSEP C=-+06. t = 0.143 in. t = 0.62 in. Loads: 1.1 x design pressure at 600 °F Primary membrane stress limit
- Allowable stress intensity, as defined by ASME Standard Code for Pumps and Valves for
Nuclear Power where: t = minimum required thickness in.
S = allowable stress, psi
P = 1.1 x design pressure, psi
R = internal radius, in. E = joint efficiency
C = corrosion allowable, in.
Valve disk thickness S s W A PA A==1 S S = 20,190 psi S = 14,351 psi Loads: 1.1 x design pressure at 600 °F Diagonal shear stress limit: 0.6 x allowable stress intensity, as defined by ASME Standard Code for Pumps and Valves for
Nuclear Power where: W = shear load, lb
A = shear area, in. 2 P = 1.1 x design pressure, psi
A 1 = side area, in. 2 UFSAR/DAEC-1 T3.7-37 Revision 13 - 5/97 Table 3.7-18 Sheet 2 of 6 STRESS
SUMMARY
FOR MAIN STEAM SAFETY VALVES
Criteria
Method of Analysis Allowable Stress or Minimum Thickness Required Calculated Stress or Actual Thickness and: A = S (R+R 1) S = slope of frustrum of shear cone, in. R = radius at base of cone, in.
R 1 = radius at top of cone, in. Inlet flange bolt area Loads: Design pressure and temperature, gasket load, operational load, and design-
basis earthquake Total bolting loads and stresses shall be calculated in accordance
with procedures of Paragraph 1-
704.5.1, Flange Joints, of ANSI
B31.7 Nuclear Piping Code S b = 27,700 psi S b = 17,296 psi Bolting stress limit
- Allowable stress intensity, S m , as defined by ASME Standard Code for Pumps and Valves
for Nuclear Power Inlet flange thickness Loads: Design pressure and temperature, gasket load, and seismic load (design-basis
earthquake) Flange thickness and stress shall be calculated in accordance with
procedures of Paragraph 1-704.5.1, Flanged Joints, of ANSI B31.7, Nuclear Piping Code S H = 27,300 psi S R = 27,300 psi S T = 27,300 psi S H = 21,363 psi S R = 10,811 psi S T = 4,584 psi
UFSAR/DAEC-1 T3.7-38 Revision 13 - 5/97 Table 3.7-18 Sheet 3 of 6 STRESS
SUMMARY
FOR MAIN STEAM SAFETY VALVES
Criteria
Method of Analysis Allowable Stress or Minimum Thickness Required Calculated Stress or Actual Thickness Flange stress limits (S H , S R , S T): 1.5S m per ASME Standard Code for Pumps and Valves
for Nuclear Power Valve spring - torsional stress S PD d C CC max.=+8 3 41 440615 Setpoint S= 84,000 psi
Maximum lift S = 112,500 psi Setpoint S = 52,616 psi
Maximum lift S = 91,747 psi Loads: W 1 = setpoint load, lb W 2 = spring load at maximum lift, lb Torsional stress limit
- 0.67 x torsional elastic limit
when subject to a load of W 1; 0.90 x torsional elastic limit
when subjected to a load of W 2 where: Smax = torsional stress, psi P = W 1 or W 2 = spring load, lb D = mean diameter of coil, in. d = diameter of wire, in. C = D d = correction factor Yoke rod area A F S m=2 A = 0.575 in. 2 A = 1.402 in. 2 Loads: Spring loads at maximum lift UFSAR/DAEC-1 T3.7-39 Revision 13 - 5/97 Table 3.7-18 Sheet 4 of 6 STRESS
SUMMARY
FOR MAIN STEAM SAFETY VALVES
Criteria
Method of Analysis Allowable Stress or Minimum Thickness Required Calculated Stress or Actual Thickness Primary stress limit
- Allowable stress intensity, S m , as defined by ASME Standard Code for Pumps and Valves
for Nuclear Power where: A = required area per rod, in.2
F = total spring load, lb
S m = allowable stress, psi Yoke bending and shear stresses S b M Z S s V A== S b = 17,800 psi S s = 10,700 psi S b = 12,314 psi S s = 2,998 psi Loads: Spring load at maximum lift Bending and shear stress limits: Bending - allowable stress intensity, S m , per ASME Standard Code for Pumps and
Valves for Nuclear Power;
shear - 0.6 x allowable stress intensity, 0.6 S m , per ASME Standard Code for Pumps and
Valves for Nuclear Power where: S b = bending stress, psi S s = shear stress, psi M = bending moment, in.-lb Z = section modulus, in. 3 V = vertical shear, lb
A = shear area, in. 2 Minimum body wall thickness t Pd SP C=+15212.. t = 0.311 in Body bowl t = 0.562 in. Loads: Primary service pressure where: t = required thickness, in. t = 0.218 in Inlet nozzle t = 1.228 in.
UFSAR/DAEC-1 T3.7-40 Revision 13 - 5/97 Table 3.7-18 Sheet 5 of 6 STRESS
SUMMARY
FOR MAIN STEAM SAFETY VALVES
Criteria
Method of Analysis Allowable Stress or Minimum Thickness Required Calculated Stress or Actual Thickness Primary Stress limit
- Allowable stress, 7000 psi, in
accordance with ASA B16.5 S = allowable stress, 7000 psi P = primary service pressure, 150 psi d = inside diameter of valve at section being considered, in. t = 0.250 in. Outlet nozzle t = 0.5625 in. Inlet nozzle combined stress S FF A M M Z=+++1212 S = 27,300 psi S = 5,159 psi Loads: Spring load at maximum lift operational load seismic load -
DBE Combined stress limit
- 1.5 x allowable stress intensity, 1.5S m , per ASME Standard Code for Pumps and Valves
for Nuclear Power where: S = combined bending and tensile stress, psi F 1 = maximum spring load, psi F 2 = vertical component of reaction thrust, lb A = cross section area of nozzle, in.2 M 1 = moment resulting from horizontal component of
reaction, in.-lb M 2 = moment resulting from horizontal seismic force, in.-lb
UFSAR/DAEC-1 T3.7-41 Revision 13 - 5/97 Table 3.7-18 Sheet 6 of 6 STRESS
SUMMARY
FOR MAIN STEAM SAFETY VALVES
Criteria
Method of Analysis Allowable Stress or Minimum Thickness Required Calculated Stress or Actual Thickness Spindle diameter F c EI L=2 2 Actual load F = 34,490 lb Load limit (0.2F c) F = 85,900 lb Loads: Spring load at maximum lift where: Fc = critical buckling load, lb E = modulus of elasticity, psi I = moment of inertia, in. 4 L = length of spindle in. compression, in. Spring washer shear area S s F A= S s = 15,960 psi S s = 2,430 psi Loads: Spring load at maximum lift Shear stress limit 0.6 x allowable stress intensity, 0.6S m , per ASME Standard Code for Pumps and Valves
for Nuclear Power where: S s = shear stress, psi F = spring load, lb
A = shear area, in. 2 UFSAR/DAEC-1 T3.7-42 Revision 13 - 5/97 Table 3.7-19 Sheet 1 of 4 STRESS
SUMMARY
FOR RECIRCULATION PUMPS
Criteria
Method of Analysis
Analytical Results Allowable Stress or Actual Thickness Minimum casing wall thickness t PRSEP C=+06. s = 1,5075 psi at 575°F t = 2.12 in. t = 2.500 in. Loads: Normal and upset condition, design pressure and temperature Primary membrane stress limit
- Allowable working stress per
ASME Code, Section III,
Class C where: t = minimum required thickness, in.
P = design pressure, psig R = maximum internal radius, in.
S = allowable working stress, psi E = joint efficiency
C = corrosion allowance, in. Minimum casing cover thickness Loads: Normal condition, design pressure and temperature S s F A= S s (design) = 3531 psi S t = 15,075 psi at 575°F S t = 8,750 psi at 575°F Stress limit: Allowable working stress per
ASME Code, Section III,
Class C where: S s = shear stress F = total upward force, in.-lb
A = area S s S t=3 S t = allowable tensile strength
UFSAR/DAEC-1 T3.7-43 Revision 13 - 5/97 Table 3.7-19 Sheet 2 of 4 STRESS
SUMMARY
FOR RECIRCULATION PUMPS
Criteria
Method of Analysis
Analytical Results Allowable Stress or Actual Thickness Cover and seal flange bolt areas Loads: Normal and upset conditions, design pressure and temperature, design gasket
load Bolting loads, areas, and stresses shall be calculated in accordance
with "Rules for Bolted Flanged
Connections," ASME Code, Section VIII, Appendix II Cover flange bolts W mi = 1,431,000 lb A mi = 71.6 in. 2 S b = 20,000 psi
A m = 84.2 in. 2 S b = 20,000 psi Bolting stress limit
- Allowable working stress per
ASME Code, Section III,
Class C Seal flange bolts W mi = 141,000 lb A mi = 7.05 in. 2 S b = 20,000 psi
A m = 8.82 in. 2 S b = 20,000 psi Cover clamp flange thickness Loads: Normal and upset conditions, design pressure and temperature, design gasket
load, design bolting load Flange thickness and stress shall be calculated in accordance with
"Rules for Bolted Flange
Connections," ASME Code, Section VIII, Appendix II Flange thickness and stress M o = 6,240,750 in.-lb S t = 17,500 psi t = 6.36 in. t = 8-1/4 in.
S = 17,500 psi Tangential flange stress limit
- Allowable working stress per
ASME Code, Section III,
Class C UFSAR/DAEC-1 T3.7-44 Revision 13 - 5/97 Table 3.7-19 Sheet 3 of 4 STRESS
SUMMARY
FOR RECIRCULATION PUMPS
Criteria
Method of Analysis
Analytical Results Allowable Stress or Actual Thickness Pump nozzle membrane and bending stress S DP A M Z F A L=++2 4 S L = 7,002 psi S c = 17,200 psi S m = 15,075 psi 1.5S m = 22,112 psi Loads: Normal and upset condition, design pressure, and temperature, piping reactions during normal operation S c PD t S s tR o J==2 S s = 0.001 psi S = 12,100 psi Combined stress limits
- 1.5S m , per ASME Standard Code for Pumps and Valves
for Nuclear Power, Class 1 SSSSS SLCLC s=+++22 2 2 12/ where: S L = longitudinal stress, psi S C = circumferential stress, psi S s = torsional stress, psi D = nozzle internal diameter, in.
P = design pressure, psi A = nozzle cross-section metal
areas, in. 2 UFSAR/DAEC-1 T3.7-45 Revision 13 - 5/97 Table 3.7-19 Sheet 4 of 4 STRESS
SUMMARY
FOR RECIRCULATION PUMPS
Criteria
Method of Analysis
Analytical Results Allowable Stress or Actual Thickness Pump nozzle membrane and bending stress (continued) M = maximum bending moment, in.-lb F = maximum longitudinal, force, lb t = nozzle wall thickness, in. J = polar moment of inertia, in. 4 R o = nozzle outside radius, in Z = polar section modulus, in. 3 Mounting bracket combined stress Loads: Flooded weight, DBE horizontal seismic force - 1.5g; DBE vertical seismic force -
0.14g Combined stress limit
- Yield stress Bracket vertical loads shall be determined by summing the equipment and fluid weights and vertical seismic forces. Bracket
horizontal loads shall be determined by applying the specified seismic force at mass center of pump-motor assembly (flooded). Horizontal and vertical
loads shall be applied simultaneously to determine
tensile, shear, and bending stresses
in the brackets. Tensile, shear, and bending stresses shall be combined to determine maximum combined
stresses. Maximum combined stress
Shear stress (vertical)
Lug 1 S = 2180 psi
Lug 2 S = 2180 psi
Lug 3 S = 4000 psi
Shear stress (horizontal)
Lug 1 S = 2960 psi
Lug 2 S = 2960 psi
Lug 3 S = 2960 psi S s = 8,650 psi
UFSAR/DAEC-1 T3.7-46 Revision 13 - 5/97 Table 3.7-20 Sheet 1 of 3 STRESS
SUMMARY
FOR RECIRCULATION VALVES
Criteria
Method of Analysis
Calculated Stress or Thickness Allowable Stress or Minimum Wall Thickness Minimum body wall thickness t PdSPy=x+15221 01.(). 4-in.valve t = .900 in. 4-in.valve t = 0.405 in. Loads: Design pressure and temperature where: t = minimum wall thickness, in.
P = design pressure, psig
22x18x22 in. valve t = 1.630 in. (discharge and
suction) 22x18x22 in. valve t = 1.520 in. Primary membrane stress limit
- Allowable working stress per
ASME Code, Section VIII;
USAS B31.1.0, 1967; manufacturer's standards
MSS-Sp66 d = minimum diameter of flow passage, but not less than 90% of the inside diameter at
welding end, in. S = allowable working stress, psi y = plastic stress distribution factor, 0.4 Body-to-bonnet bolt area Loads: Design pressure and temperature, gasket load, stem operational load, seismic load (DBE) Bolting stress limit
- Allowable working stress per
ASME Code, Section VIII, 1968 Total bolting loads and stresses shall be calculated in accordance
with "Rules for Bolted Flange
Connections," ASME Code, Section VIII, load shall be applied at the mass center of the valve operator, assuming that the valve body is rigid and anchored 4-in. valve A b = 9.29 in. 2 S b = 10,613 psi
22x18x22 in. valve (suction and
discharge)
A b = 47.52 in. 2 S b = 15.777 psi 4-in.valve Actual bolt area =
10.24 in.2 S allow = 20,000 psi
22x18x2 in. valve
Actual bolt area
S allow = 20,000 psi
UFSAR/DAEC-1 T3.7-47 Revision 13 - 5/97 Table 3.7-20 Sheet 2 of 3 STRESS
SUMMARY
FOR RECIRCULATION VALVES
Criteria
Method of Analysis
Calculated Stress or Thickness Allowable Stress or Minimum Wall Thickness Flange thickness and stress Loads: Design pressure and temperature, gasket load , stem operational load, seismic load (DBE) Flange thickness and stress shall be calculated in accordance with
"Rules for Bolted Flange
Connections," ASME Code, Section VIII, Appendix II, except that the stem operational load and seismic load shall be carried by the
flange. The horizontal and vertical 22x18x22 in. valve (discharge and
suction)
S H = 15,641 psi S R = 10,997 psi S T= 7,821 psi
20,139 psi
13,426 psi
13,426 psi Flange stress limit (S H , S R , S T): 1.5S m , per ASME Standard Code for Pumps and Valves
for Nuclear Power, Class 1 seismic forces shall be applied at the mass center of the valve operator, assuming that the valve body is rigid 4-in. discharge bypass S H = 13,408 psi S R = 6,303 psi S T = 11,935 psi Valve disk thickness Loads: Design pressure and temperature t a R Pd S m S a P d R t A08160816 2.(.)=x 22x18x22 in. valve (discharge)
t A = 2.018 in. S A = 1,904 psi 22x18x22 in. valve t a = 2.018 in. S a = 1,904 psi Primary bending stress limit: 1.5S m , per ASME, Standard Code for Pumps and Valves
for Nuclear Power, Class 1 where: t a = calculated disk thickness S a = design stress allowable t A = actual disk thickness 22x18x22 in. valve (suction)
t A = 2.009 in. S A = 14,031 psi 22x18x22 in. valve t A = 1.871 in. S a = 15,904 psi 4-in. valve t A = 0.550 in. S A = 14,366 psi 4-in. valve t A = 0.523 in. S a = 15,904 psi UFSAR/DAEC-1 T3.7-48 Revision 13 - 5/97 Table 3.7-20 Sheet 3 of 3 STRESS
SUMMARY
FOR RECIRCULATION VALVES
Criteria
Method of Analysis
Calculated Stress or Thickness Allowable Stress or Minimum Wall Thickness Valve disk thickness (continued) S A = allowable stress R = Radius of disk
P d = design pressure Valve operator supports Loads: Design pressure and temperature, steam operational load, seismic load (DBE) Yoke and yoke bolt stress limits: Allowable working stress per
ASME Code, Section VIII The valve assembly is analyzed assuming that the valve body is an anchored, rigid mass and that
specified vertical and horizontal seismic forces area applied at the mass center of the operator assembly simultaneously with
operating pressure plus deadweight
plus operational loads. These loads are used to determine stresses and
deflections for the operator support components Operator support bolt stress S b = bolt stress (psi)
22-in. (suction and discharge)
S b = 20,602 psi
4-in. bypass S b = 10,622 psi S b allowable = 20,000 psi
UFSAR/DAEC-1 T3.7-49 Revision 13 - 5/97 Table 3.7-21 STRESS
SUMMARY
FOR MAIN STEAM LOOP "A" Results of Analysis a (psi) Criteria Method of Analysis Maximum Stress Stress Limit Stress for all normal and upset loading must not exceed the limits of ANSI B31.1.0. Effects from the following loading combinations determined in accordance with rules of ANSI B31.1.0. The sum of the longitudinal stresses due to pressure and deadweight must be less than hot allowable stress. 6,421 17,500 The sum of the thermal expansion stress intensity range plus anchor displacement stresses caused by the OBE must be less than the allowable stress range for
expansion stresses. 7,144 26,250 The sum of longitudinal stresses due to pressure, deadweight, inertia effects of the OBE, and external loads must be less than 1.2 times the hot allowable
stress. 16,183 21,000 The sum of the longitudinal stresses due to pressure and deadweight plus the thermal expansion stress intensity range must be less than the sum of the
allowable stress range for expansion stresses plus the
hot allowable stress. 11,524 43,750 Primary stress for all load combinations that have a very low probability of occurrences must not exceed 1.5 times the limits of ANSI B31.1.0. Effects from the following loading combinations determined in accordance with rules of ANSI B31.1.0
The sum of the longitudinal stresses due to pressure, deadweight, inertia effects of the DBE, and external forces must be less than 1.8 times the hot allowable
stress.
23,006
31,500 The sum of the longitudinal stresses due to minimum pressure, deadweight, inertia effects of the OBE, and external forces must be less than 1.8 times the hot
allowable stress. 16,546 31,500 a All limits were met. Highest stresses are given. UFSAR/DAEC-1 T3.7-50 Revision 13 - 5/97 Table 3.7-22 STRESS
SUMMARY
FOR MAIN STEAM LOOP "B" Results of Analysis a (psi) Criteria Method of Analysis Maximum Stress Stress Limit Stress for all normal and upset loadings must not exceed the limits of ANSI B31.1.0. Effects from the following loading combination determined in accordance with rules of ANSI B31.1.0. The sum of the longitudinal stresses due to pressure and deadweight must be less than the hot allowable
stress. 5,572 17,500 The sum of the thermal expansion stress intensity range plus anchor displacement stresses caused by the OBE must be less than the allowable stress range for
expansion stresses. 8,773 26,250 The sum of the longitudinal stresses due to pressure, deadweight, inertia effects of the OBE, and external loads must be less than 1.2 times the not allowable
stress. 15,382 21,000 The sum of the longitudinal stresses due to pressure and deadweight plus the thermal expansion stress intensity range must be less than the sum of the
allowable stress range for expansion stresses plus the
hot allowable stress. 10,581 43,750 Primary stress for all load combinations that have a very low probability of occurrences must not exceed 1.5 times the limits of ANSI B31.1.0. Effects from the following loading combinations determined in accordance with rules of ANSI B31.1.0.
The sum of the longitudinal stresses due to pressure, deadweight, inertia effects of DBE, and external forces must be less than 1.8 times the hot allowable stress.
20,575
31,500 The sum of the longitudinal stresses due to minimum pressure, deadweight, inertia effects of the OBE, and external forces must be less than 1.8 times the hot
allowable stress. 15,745 31,500
a All limits were met. Highest stresses are given. UFSAR/DAEC-1 T3.7-51 Revision 13 - 5/97 Table 3.7-23 STRESS
SUMMARY
FOR MAIN STEAM LOOP "C" Results of Analysis a (psi) Criteria Method of Analysis Maximum Stress Stress Limit Stress for all normal and upset loadings must not exceed the limits of ANSI B31.1.0. Effects from the following loading combinations determined in accordance with rules of ANSI B31.1.0.
The sum of the longitudinal stresses due to the pressure and deadweight must be less than the hot allowable
stress. 5,428 17,500 The sum of the thermal expansion stress intensity range plus anchor displacement stresses caused by the OBE must be less than the allowable stress range for
expansion stresses.
5,692 26,250 The sum of the longitudinal stresses due to pressure, deadweight, inertia effects of the OBE, and external loads must be less than 1.2 times the hot allowable
stress. 18,010 21,000 The sum of the longitudinal stresses due to pressure and deadweight plus the thermal expansion stress intensity range must be less than the sum of the
allowable stress range for expansion stresses plus the
hot allowable stress.
9,925 43,750 Primary stress for all load combinations that have a very low probability of occurrences must not exceed 1.5 times the limits of ANSI B31.1.0. Effects from the following loading combinations determined in accordance with rules of ANSI B31.1.0. The sum of the longitudinal stresses due to pressure, deadweight, inertia effects of DBE, and external forces must be less than 1.8 times the hot allowable stress.
24,661 31,500 The sum of the longitudinal stresses due to minimum pressure, deadweight, inertia effects of the OBE, and external forces must be less than 1.8 times the hot
allowable stress 18,373 31,500
a All limits were met. Highest stresses are given. UFSAR/DAEC-1 T3.7-52 Revision 13 - 5/97 Table 3.7-24 STRESS
SUMMARY
FOR MAIN STEAM LOOP "D" Results of Analysis a (psi) Criteria Method of Analysis Maximum Stress Stress Limit Stress for all normal and upset loadings must not exceed the limits of ANSI B31.1.0. Effects from the following loading combinations determined in accordance with rules of ANSI B31.1.0. The sum of the longitudinal stresses due to pressure and deadweight must be less than the hot allowable
stress. 6,402 17,500 The sum of the thermal expansion stress intensity range plus anchor displacement stresses caused by the ODE must be less than the allowable stress range for
expansion stresses. 7,094 26,250 The sum of the longitudinal stresses due to pressure, deadweight, inertia effects of the ODE, and external loads must be less than 1.2 times the hot allowable
stress. 15,227 21,000 The sum of the longitudinal stresses due to pressure and deadweight plus the thermal expansion stress intensity range must be less than the sum of the
allowable stress range for expansion stresses plus the
hot allowable stress. 11,567 43,750 Primary stress for all load combinations that have a very low probability of occurrences must not exceed 1.5 times the limits of ANSI B31.1.0. Effects from the following loading combinations determined in accordance with rules of ANSI B31.1.0. The sum of the longitudinal stresses due to pressure, deadweight, inertia effects of the DBE, and external forces must be less than 1.8 times the hot allowable
stress. 20,217 31,500 The sum of the longitudinal stresses due to minimum pressure, deadweight, inertia effects of the OBE, and external forces must be less than 1.8 times the hot
allowable stress. 15,590 13,500 a All limits were met. Highest stresses are given. UFSAR/DAEC-1 T3.7-53 Revision 13 - 5/97 Table 3.7-25 STRESS
SUMMARY
FOR RECIRCULATION LOOP "A" Results of Analysis a (psi) Criteria Method of Analysis Maximum Stress Stress Limit Stress for all normal and upset loadings must not exceed the limits of ANSI B31.1.0. Effects from the following loading combinations determined in accordance with rules of ANSI B31.1.0. The sum of the longitudinal stresses due to pressure and deadweight must be less than the hot allowable
stress. 8,828 14,425 The sum of the thermal expansion stress intensity range plus anchor displacement stresses caused by the OBE must be less than the allowable stress range for
expansion stresses. 21,616 27,050 The sum of the longitudinal stresses due to pressure, deadweight, and inertia effects of the OBE must be less than 1.2 times the hot allowable stress. 15,519 17,310 The sum of the longitudinal stresses due to pressure and deadweight plus the thermal expansion stress intensity range must be less than the sum of the
allowable stress range for expansion stresses plus the
hot allowable stress. 25,078 41,470 Primary stress for all load combinations that have a very low probability of occurrences must not exceed 1.5 times the limits of ANSI B31.1.0. Effects from the following loading combinations determined in accordance with rules of ANSI B31.1.0. The summary of the longitudinal stresses due to pressure, deadweight, and inertia effects of the DBE must be less than 1.8 times the hot allowable stress. 22,828 25,965 The sum of the longitudinal stresses due to minimum pressure, deadweight, and inertia effects of the OBE must be less than 1.8 times the hot allowable stress. 16,623 25,965
a All limits were met. Highest stresses are given. UFSAR/DAEC-1 T3.7-54 Revision 13 - 5/97 Table 3.7-26 STRESS
SUMMARY
FOR RECIRCULATION LOOP "B" Results of Analysis a (psi) Criteria Method of Analysis Maximum Stress Stress Limit Stress for all normal and upset loadings must not exceed the limits of ANSI B31.1.0. Effects from the following loading combinations determined in accordance with rules of ANSI B31.1.0. The sum of the longitudinal stresses due to pressure and deadweight must be less than the hot allowable
stress. 8,581 14,425 The sum of the thermal expansion stress intensity range plus anchor displacement stresses caused by the OBE must be less than the allowable stress range for
expansion stresses. 25,298 27,050 The sum of the longitudinal stresses due to pressure, deadweight, and inertia effects of the OBE must be less than 1.2 times the hot allowable stress. 16,713 17,310 The sum of longitudinal stresses due to pressure and deadweight plus the thermal expansion stress intensity range must be less than the sum of the allowable stress
range for expansion stresses plus the hot allowable
stress. 28,673 41,470 Primary stress for all load combinations that have a very low probability of occurrences must not exceed 1.5 times the limits of ANSI B31.1.0. Effect from the following loading combinations determined in accordance with rules of ANSI B31.1.0. The sum of the longitudinal stresses due to pressure, deadweight, and inertia effects of the DBE must be less than 1.8 times the hot allowable stress. 24,418 25,965 The sum of the longitudinal stresses due to minimum pressure, deadweight, and inertia effects of the OBE must be less than 1.8 times the hot allowable stress. 17,816 25,965
a All limits were met. Highest stresses are given. UFSAR/DAEC-1 T3.7-55 Revision 20 - 8/09 Table 3.7-27 REACTION LOADS RECIRCULATION INLET NOZZLE
Forces (kips) Moments (in.-kips) Fx Fy Fz Mx My Mz Total thermal, weight, and seismic loads 5.1 13.8 3.9 543 265 292 External mechanical (weight and seismic) loads
only 1.1 7.3 2.3 250 96 47 Hydraulic loads applied to thermal sleeve 0 3.8 14.3 -44 0 0 Notes: 1. Forces Fx and Fy are located 118.5 in. from vessel axis.
- 2. Nozzle loads may be in either the positive or negative direction.
- 3. All values are at design temperature and pressure.
- 4. These nozzle loads are also listed in APED-B11-232, APED-B11-235, and APED-B11-001<7>.
UFSAR/DAEC-1 T3.7-56 Revision 14 - 11/98 Table 3.7-28 SAMPLE CHART FIRST MODE GREATER THAN SPECTRUM PEAK METHOD Pipe Size, Span, and Load a Insulation Class and Temperature (°F) No Insulation Ambient IV & V 125-150 III 251-350 II 351-500 I 501-750 1/2-in. pipe Span (ft-in.) 5-1 4-8 4-8 4-2 4-2 Load (lb) 7.0 8.5 8.5 11.5 11.5
3/4-in. pipe Span (ft-in.) 5-8 5-4 5-4 4-9 4-9 Load (lb) 12.0 14.0 14.0 17.0 17.0
1-in. pipe Span (ft-in.) 6-5 6-1 6-1 5-7 5-7 Load (lb) 20.0 22.0 22.0 26.0 26.0
1-1/2-in. pipe Span (ft-in.) 7-8 7-4 7-4 6-11 6-11 Load (lb) 42.0 45.0 45.0 51.0 51.0
2-in. pipe
Span (ft- in.)) 8-6 8-2 8-0 7-9 7-7 Load(lb) 72.0 75.0 78.0 81.0 85.0
a Load is for each support. UFSAR/DAEC-1 T3.7-57 Revision 14 - 11/98 Table 3.7-29 SAMPLE CHART MODIFIED SPECTRUM METHOD
Elevation (ft) Pipe Size, Span, and Load a Below 195 Below 165 Below 135 1/2-in. pipe Span (ft-in.) 4-6 4-9 5-0 Load (lb) 5.0 5.0 4.5
3/4-in. pipe Span (ft-in.) 5-0 5-6 5-6 Load (lb) 8.0 7.0 7.0
1-in. pipe Span (ft-in.) 6-0 6-6 6-9 Load (lb) 9.0 8.0 8.0
1-1/2-in. pipe Span (ft-in.) 7-0 7-3 7-6 Load (lb) 20.0 20.0 18.0
2-in. pipe Span (ft-in.) 9-0 9-6 10-0 Load (lb) 50.0 55.0 58.0
a Load is for each support. DAMES/MOORELJDIRECTIONOFINFORMATIONFLOW--LEGENDDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORT-,iIISPECIAL*IIIIIIIIIIIIiiII_Il-__zl01i=;<:1-,:2:.Uj'Zul<IL0:::2:00euz>'0---,II___.J...II>J°i=<uiL:U...P615lcnv*IIII>C)ziLI1iLIIIREACTIONSONSAFETYIEQUIP.NOZZLESIJ,PIPINGLAYOUIPIPINGLAYOUTIrGROUPISTRESSPIPINGSYSTEMSGROUPEMSIJI0Gz-clozZII0.J-lIIenOI'>zE;I,II-"-0-1111I!:!0_1-0::ullllIUa::2:-c.JII-1-1III.;:)>....a..Olu01:)eng:g>-0::101&1-cI-A.Uu>I:llu.J1-.enZ!Z...iiiELECTRICALen----'-GROUPIVENDORS.J<0::i!1&1III.oZ°....I-uZZIII-c:2:::::ija..u:2:°°0'III.>1iIIi1enZZ°0i=<Z-C.Jou.J...u<z....ou-00:2:ZOIl>D.A.E.e.PROJECTORGANIZATIONI'SITESEISMOLOGYII1-CONTAINMENTVESSELSTRESSANAL:(SISir-----------IFOUNDATIONaBUILDINGRESPONSECURVES-+I.SITESEISMOLOGYI,-.---.-----.....,..----..IBUILDINGLAYOUT8MASSESJOHNA.BLUME:18ASSOC.Iiz°su"-U...APPROVALOFVENDORCALCULATIONS*xDOCUMENTATIoN"I-IL_IIIL__I--IBECHTELCORPORATIONIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIORVESSELSEISMIC-RE*S-pcfN*SE-*----------*.--t-ON__UIPMENTNOZZLESr.....**.APPLICATIONSENGINEERINGIIIIIIIII______...1fI'---':---,III-,I__.JREACTORVESSELSEISMICRESPONSEVENDORSDESIGNENGINEERING,IISEISMICIItSYSTEMSDESIGNSII,_R.S_IICOMPONENTI/8D.He.DESIGNSPROJECTAND(INCLUDINGR.P,V.lFORC..w,SPECTRAIIPIPINGENDREACTIONONIREACTORVESSELNOZZLEStPIPINGENDREACTIONSONEQUIPNOZZLESRESPONSESPECTRAGE/APEOr-IL__I______z0--------_.1-1enluuli'0-en!a3tZ1&1__.>0---IIIIIIIIISeismicDesignInterrelationshipsFigure3.7-1 LDPSAR5FECTRUM------_____MODIFll:DTIMEHISTORY.SPECTRUMdecINI.ZOw(J)Z.....0";0-0Cf)Wer:::0.............0f-0:ier:::!wIwou'UOIT01WI*-+1-----,1,.1----.,....1----r-j-----,Ir---------,'00.000.250.500.751.001.251.501.752.00PERIODINSECONDSDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTGroundMotionEquationNormalizedto0.06GResponseSpectra,Damping=0.005.Figure3.7-2 l/)(Y"lozo_....CI:o0::W--lWU-.rU-WCDZ0-0CD'Wo0::oo_-----MODIFIEDTIMEHISTORYspeCTRUMINTERPOLATEDPSAR.SFec.TRUM0.250,500.751.001.25PERIODINSECONDS1.501.752.00DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTGroundMotionEquationNormalizedto0.06GResponseSpectra;Damping=0.010Figure3.7-3 MODIFIEDTIMEHISTORYSPECTR.UMPSAR/zo..........-4f-e-;a:a0:::::W-lWu-'a:C;;IW(f)ZCLaen.wa0:::::aa7-S-----'11-.00PERIODINSECONDSDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTGroundMotionEquationNormalizedto0.06GResponseSpectra,Damping=0.020Figure3.7-4 lJ)('I"')oOlDN.ZOZoPSAR/5PEc:rR.UM(MODIFIE.DTIMESPECTRUMoo--0:00::l.J.JWU-.:ru-WI'./)ZOr-I0...0(J)*WO0::0.250.500.751.001.25PERIODINSECONDS1.501.752.00DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTGroundMotionEquationNormalizedto0.06GResponseSpectra,Damping=0.050Figure3.7-5
o o. .-._------
._----------
we (J)CD z..:.om w'" z..: o.....a: O:::e WeD..J0 l&.JO U u a: 0.25 o.50 0.75 I.00 PERIOD IN SECONDS 1.25 1.50 1.75 2.00 DUANE ARNOLD ENERGY CE:NTER IES UTILITIES, INC.UPDATED FINAL SAFETY ANALYSIS REPORT Example of Building Analysis Horizontal aBE Response Spectra Reactor Building, Elev.Damping-0.005,0.010,0.020,0.050 Figure 3.7-7 Revision 18-10/05
- .
SaTYPICALRESPONSECURVEf2=FUNDAMENTALFREQUENCYOFSIMPLESUPPORTSPANPEAKOFRESPONSECURVEMETHODDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTPeakofResponseCurveMethodFigure3.7-10 o(J)zoexa:ILl...JILJUu<t...Jexa:uILl(/)FREQUENCY(f)TYPICALSPECTRUMCURVEzoi=exa:ILJ...JUJUU<<...J<<a:uUJ(/),II,/1II./FREQUENCY(f)MODIFIEDSPECTRUMCURVEMODIFIEDSPECTRUMCURVEDUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTModifiedSpectrumCurveFigure3.7-11 0Q0U)80dC)ZQ.:=0NQ0qU)Ns=0>-0ZI&J:JaI&JN0::\L.0-1000od0000000000CZ)eX)eX)reNeX)10V0000ddd000dI('.9)DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTAccelerationSpectrumvs.FrequencyFigure3.7-12 3.Thecurveusedfordynamicanalysisisshownin*Fl.gure10-Cl.16-4.2.areeithersocketweldedormitrebend(forelbow).ityfactorLSone.NOTE:1.Pipeweighttionincluded)=5.086lbs/ft.,bO\(.\),/00.,J)..Wi')1,0,,,0'\t;/'",4-0Ic,l'll°,..,\'1">.,,*v1:1......'\c:>-_f\..t'1../'0/I/o'l/---?,...'b./..IV1rJL,\01.1)0'),\\SSKo-'}O'1-5S'1h...:.........0,.....C\-"""-(...."'1-1\08s..........':....0."OJ,0...\'\C;/c>::0',S'-......c;'"r".....II0,/0/1,055"/'35,03D65?-,0A<'It;;JJ!J.I0r;c.......-00CJ::0:)::>)::>-ICJrrJrrJCCJr-)::>rrJZnrrJ.......-IVlz;0)::>llJ)::>.......;03-0r-n:z--'0CDVlr-r......)::>.......0lO-0GlC......rrJ:x:rrJ-s-0CD......-I-Iz::s-<rrJWlOQ<>;0.)::>Gl-.....Jr-z-0-<IllJ......'<)::>0w0r::E:nc-<rrJrrJc+Vl;0:z.......-IVlnrrJ0;0;03:rrJ-0-0)::>0z;0-<-I-...-----,,-'(Y;\.j.........//;< 7:.yxNOTE:1.Fittingsareeithersocketweldedormitrebend(forelbow).Flexibilityfactorisone.ThespectrumcurveusedfordynamicanalysisisshowninFigurelO-Cl.16-63.2.Pipesarenotinsulated.5'-0"lire;I:'v.,-;".,',,0c.....-..*J0-000)::>)::>-I0/TI/TIc:0r-)::>/TI:z"n/TIVl.....-IAI:z;0)::>"'"1'}.'>>>V:&J23>OIt3)::>.....;0.........Jo,"'Cr-n:zIIIJ,I-vSD*to-t>'"--'07,.UI.....ro/1(/lQVlr-r-II's::--0)::>.....0Or,,{)......"CDI.ro"'C/TI:r:/TI......-I-I:zw::s.to-<",......Qo;0Ir-)::>lJ).......SlJ:z-0-<+::>'<)::>00r-::0:ns::rt-</TI",Vl;0:z......-IVln0;0;03:/TI-0-0)::>0z;0-<-I UFSAR/DAEC - 1 3.8-1 Revision 23 - 5/15 3.8 DESIGN OF SEISMIC CATEGORY I STRUCTURES
All structures and equipment are designed in accordance with applicable codes for dead loads, live loads, seismic loads, and wind loads. The loading conditions and combinations are determined by the function of the structure and its importance in meeting the plant safety and
power generation objectives.
To protect vital equipment and systems, certain critical plant structures must remain functional both during and following the most severe natural phenomena that could affect the
site. These conditions are considered in the design and are investigated and defined in Chapter 2. Combinations of structural loads resulting from environmental events, normal operation, and
design accidents are given in this Section.
Dead load includes self-weight of structures and permanent equipment or any other permanent loads contributing stress, such as soil or hydrostatic loads. The live loads that have
been used in the design of structures are given in Table 3.8-1.
The following notations are used in this section:
U = required ultimate load capacity
D = dead load of structure, equipment, and other loads contributing permanent stress L = live load as indicated in Table 3.8-1
R = force on structure from the rupture of any one pipe
T o = thermal loads due to temperature gradient through wall under operating conditions
P = design-basis accident pressure load
H o = force on structure from the thermal expansion of pipes under operating conditions
T A = thermal loads due to temperature gradient through wall under accident conditions
H A = force on structure from the thermal expansion of pipes under accident conditions
E = OBE resulting from ground surface accelerations listed in Section 3.7
E' = DBE resulting from ground surface accelerations listed in Section 3.7 UFSAR/DAEC - 1 3.8-2 Revision 23 - 5/15 A = hydrostatic load due to high water level at elevation 767 ft
W = wind load
W' = tornado
= capacity reduction factor (defined in ACI 318-63, Section 1504)
f s = allowable stress for structural steel
F y = yield strength for steel
3.8.1 CONCRETE CONTAINMENT
Because the DAEC uses the Mark I steel containment (light bulb-torus), this section is
not applicable.
3.8.2 STEEL CONTAINMENT
3.8.2.1 Description of the Containment
The Mark I suppression type containment consists basically of two steel pressure vessels,
the drywell and the torus, joined by large vent pipes.
3.8.2.1.1 Drywell
The drywell is a steel pressure vessel with a spherical lower portion, 63 ft in diameter, and a cylindrical upper portion 32 ft in diameter. The overall height is approximately 108 ft 9 in. The design, fabrication, inspection, and testing of the drywell vessel complies with requirements of the ASME B&PV Code, Section III, Subsection B, "Requirements for Class B Vessels," Summer 1968 Addenda, and ASME Code Cases 1177, 1330, and 1413 which pertain to containment vessels for nuclear power stations. The primary containment is primarily
fabricated of SA-516 GR 70 plate.
The drywell was designed for an internal pressure of 56 psig coincident with a temperature of 281 °F with applicable dead, live, and seismic loads imposed on the shell. Thus, in accordance with the ASME Code, Section III, the maximum internal drywell pressure is 62 psig. Design external pressure is 2 psig at 281 °F. Thermal stresses in the steel shell due to temperature gradients were taken into account in the design. Containment stresses are within allowable stresses permitted by the ASME Code. Detailed investigation was performed in areas where local buckling could occur due to the effects of concentrated loads, thermal loads, and non-axisymmetric distribution loads. Where stress intensities in these areas were not covered by the ASME Code, allowable stresses were determined using recognized buckling formulae such UFSAR/DAEC - 1 3.8-3 Revision 23 - 5/15 as those used by the American Institute of Steel Construction (AISC) and recognized, reputable authors (Roark, Timoshenko, and Grintner).
Special precautions not required by codes were taken in the fabrication of the steel drywell shell. Charpy V -notch speciments were used for impact testing of plate and forging material to give assurance of correct material pr operties. Plates, forgings, and pipe associated with the drywell had an initial NDT temperature of 0 °F when tested in accordance with the appropriate code for the materials. It is not intended that the drywell will be pressurized or subjected to substantial stress at temperatures below 30 °F. The drywell is enclosed in a reinforced concrete structure for shielding purposes. In
areas where it backs up the drywell shell, this reinforced concrete provides additional resistance to deformation and buckling of the shell. Above the transition zone, and below the flange, the drywell is separated from the reinforced concrete by a gap of approximately 2 in. Shielding over the top of the drywell is provided by removable, segmented, reinforced concrete shield plugs.
In addition to the drywell head, one combination double-door air lock/equipment lock, one bolted equipment hatch, and one bolted personnel access hatch are provided for access into
the drywell.
3.8.2.1.2 Pressure Suppression Chamber
The pressure suppression chamber is a steel pressure vessel in the shape of a torus located below and encircling the drywell, with a major diameter of 98 ft 8 in. and a cross-sectional diameter of 25 ft 8 in. The pressure suppression chamber contains the suppression pool and the gas space above the pool. The suppression chamber will transmit seismic loading to the reinforced concrete foundation slab of the reactor building. Space is provided outside the chamber for inspection.
The toroidal suppression chamber was designed to the same material and code requirements as the steel drywell vessel. The material has an NDT temperature of 0 °F. 3.8.2.1.3 Vent System
Large vent pipes connect the drywell and the pressure suppression chamber. A total of
eight circular vent pipes are provided, each having a diameter of 4 ft 9 in. The vent pipes were designed for the same pressure and temperature conditions as the drywell and suppression chamber. Jet deflectors in the drywell at the entrance of each vent pipe prevent possible damage to the vent pipes from jet forces which might accompany a pipe break in the drywell. The vent
pipes are fabricated of SA-516 GR 70 steel, and comply with requirements of the ASME B&PV Code, Section III, Subsection B. The vent pipes are provided with two-ply expansion bellows to accommodate differential motion between the drywell and suppression chamber. The vent pipe bellows are designed and fabricated to the same criteria as the containment vessels (ASME Section III, Class B, Summer 1968) with complete radiograph, dye penetrant, or magnetic
particle inspection as required by the code. UFSAR/DAEC - 1 3.8-4 Revision 23 - 5/15 The drywell vents are connected to a 3 ft 6 in. diameter vent header in the form of a torus which is contained within the airspace of the suppression chamber. Projecting downward from the header are 48 downcomer pipes, 24 in. in diameter and terminating not less than 3 ft below the water surface of the pool. The vent header has the same temperature and pressure design requirements as the vent pipes.
3.8.2.2 Applicable Codes, Standard and Specifications
The design of all structures and facilities conforms to the applicable general codes and
specifications listed below, except where specifically stated otherwise:
- 1. Uniform Building Code (UBC) 1970
Portions that apply to seismic design of Nonseismic structures only.
- 2. American Institute of Steel Construction (AISC)
Specification for the Design, Fabrication, and Erection of Structural Steel for Buildings, 1963 and 1970.
- 3. American Concrete Institute (ACI)
Building Code Requirements for Reinforced Concrete (ACI 318-63) and Requirements for Reinforced Concrete Chimneys (ACI 307-69).
- 4. American Welding Society (AWS)
Standard Code for Arc and Gas Welding in Building Construction (AWS D.1.0-66 and AWS D.2.0-66).
- 5. API Specification 650 for Welded Steel Storage Tanks.
- 6. ASME B&PV Code, Section III, Cla ss B, governs the design and fabrication of the drywell and suppression chamber.
- 7. Official Linn County, Iowa,. Building Code.
- 8. American Society of Civil Engineers, Paper 3269, for wind design requirements.
- 9. American Iron and Steel Institute Specification for the Design of Light Gauge Cold-Formed Steel Structural Members, 1960.
UFSAR/DAEC - 1 3.8-5 Revision 23 - 5/15 Specifications and codes set down minimum requirements for the design and construction of structural elements of any structure.
Although some special structures involving unique problems are not explicitly covered by the codes, many provisions concerning the quality of materials and construction and design
principles are considered to be generally applicable. Sound engineering knowledge, experience, and judgment, which are so essential in engineering, rather than blind adherence to codes, must be used to adapt the various codes to structures that may not appear to be within the scope of the
codes.
3.8.2.3 Loads and Loading Combinations
3.8.2.3.1 Basis for Loading Combination
The load combination basis for Seismic Category I structures is summarized as follows:
Load Combination Minimum Requirements for Seismic Category I Structural Components
Normal loads + operating-basis earthquake Within code allowable stresses
Normal loads + maximum probable flood No functional failure
Normal loads + design-basis earthquake No functional failure
Normal loads + tornado loads No functional failure
Normal loads + design-basis LOCA No functional failure
In addition to the above, the primary containment (including penetrations) and the reactor vessel support pedestal will be designed for normal loads = design-basis LOCA loads +
earthquake loads for no functional failure.
3.8.2.3.2 Loading Using Normal Limits
Reinforced Concrete
Reinforced-concrete structures are designed for ductile behavior whenever possible, that
is, with steel stresses controlling. UFSAR/DAEC - 1 3.8-6 Revision 23 - 5/15 Concrete structures are designed to satisfy the most severe loading combinations, based
on the load factors shown below:
U = 1.5 D + 1.8 L + 1.0 T o + 1.25 H o U = 1.25(D + L + H o + E) + 1.0 T o U = 1.25(D + L + H o + W) + 1.0 T o U = 0.9 D + 1.25(H o + E) + 1.0 T o U = 0.9 D + 1.25(H o + W) + 1.0 T o In addition, for ductile moment-resisting concrete space frames and for shear walls,
U = 1.4(D + L + E) + 1.0 T o + 1.25 H o U = 0.9 D + 1.25 E + 1.0 T o + 1.25 H o For structural elements such as equipment supports carrying mainly earthquake forces,
U = 1.0 D + 1.0 L + 1.8 E + 1.0 T o + 1.25 H o Structural Steel
Steel structures are designed to satisfy the following loading combinations without
exceeding the specified stresses:
D + L + T o + H o - stress limit = f s D + L + T o + H o + E - stress limit = 1.25 f s D + L + T o + H o + W - stress limit = 1.33 f s In addition, for structural elements such as struts and bracings carrying mainly
earthquake forces,
D + L + T o + H o + E - stress limit = f s UFSAR/DAEC - 1 3.8-7 Revision 23 - 5/15 3.8.2.3.3 Loading Using Higher Limits
The Seismic Category I structures are in general proportioned to maintain elastic behavior when subjected to various combinations of dead loads, thermal loads, seismic loads, and accident loads. The upper limit of elastic behavior is considered to be the yield strength of the effective load-carrying structural materials. The yield strength, F y , for steel (including reinforcing steel) is considered to be the guaranteed minimum given in appropriate ASTM
specifications. The yield strength for reinforced-concrete structures is considered to be the ultimate resisting capacity as calculated from the ultimate strength design portion of ACI 318-63.
Concrete
Concrete structures are designed to satisfy the most severe of the following loading combinations:
U = 1.0 D + 1.0 L + 1.0 E' + 1.0 T A + 1.25 H A + 1.0 R
U = 1.0 D + 1.0 L + 1.0 E' + 1.0 T o + 1.0 H o + 1.0 R U = 1.0 D + 1.0 L + 1.0 A + 1.0 T o + 1.25 Ho U = 1.0 D + 1.0 L + 1.0 W' + 1.0 T o + 1.25 H o U = 0.95 D + 1.25 E + 1.0 T A + 1.0 H A + 1.0 R
U = 1.05 D + 1.05 L + 1.25 E + 1.0 T A + 1.0 H A + 1.0 R
Structural Steel
Steel structures are designed to satisfy the most severe of the following loading combinations without exceeding the specified stresses:
D + L + R + T o + H o + E' - stress limit* = 1.5 f s D + L + R + T A + H A + E' - stress limit* = 1.5 f s D + L + A + T o + H o -stress limit* = 1.5 f s D + L + T o + H o + W' - stress limi*t = 1.5 f s
- Maximum allowable stress in bending and tension is 0.9 F
- y. Maximum allowable stress in shear is 0.5F y 2013-018 2013-018 UFSAR/DAEC - 1 3.8-8 Revision 23 - 5/15 Concrete structures are designed using the ultimate strength design method and allowable
stresses in accordance with ACI 318-63. In no case did the actual design stresses for the DAEC exceed the ACI allowable stresses. The only modification to the ACI provision is in the assignment of load factors as indicated in Section 3.8.2.4. The conservative choice of loading conditions justifies the use of this modification.
Concentrated loads were provided for by the addition of special restraining systems.
3.8.2.3.4 Pipe Jet Effects
The primary containment system is designed to withstand forces imposed by an earthquake that occurs simultaneously with a LOCA. In addition to the pressure and the thermal loading condition described in Section 6.2.1, the primary containment is designed to withstand the jet forces shown below at the locations indicated from any direction within the drywell:
Location Jet Force (maximum) Interior Area Subjected to Jet Forces
Spherical part of drywell 393,000 lb 2.19 ft 2 Cylinder and sphere to
cylinder transition 325, 000 lb 1.80 ft 2 Suppression chamber 21,000 lb Each pipe These forces are described in Section 3.6.2
UFSAR/DAEC - 1 3.8-9 Revision 23 - 5/15 The capabilities of the primary containment with respect to stress levels and load combinations for the postulated events are noted below:
Primary Containment (including penetrations)
(1) D + L + T A +H A + P + E ASME, Section III, Class B, without the usual increase for seismic loading.
(2) D + L + H A + P + R + T A + E Same as (1) above, except local yielding is permitted in the area of the jet force where the
shell is backed up by concrete. In areas not backed up by concrete, primary local membrane stresses at the jet force do not exceed 0.90 times the yield point of the material at 300 °F. (3) D + L + H A + P + R + T A + E Primary membrane stressed in general do not exceed the yield point of the material. The same criteria as in (2) above are applied to the
effect of jet forces for this loading condition. The jet forces consist of steam and/or water at 300 °F maximum. The jet forces do not occur simultaneously. However, a jet force is considered to occur coincident with design internal pressure and a temperature of 150 °F. 3.8.2.3.5 Flooded Containment
For the postaccident internal flood condition, the containment is analyzed for the dead load of the vessel, all equipment and appurtenances, OBE, and hydrostatic water pressure to the
refueling floor level The maximum possible postaccident flood condition temperature is 212 °F; the containment is designed for a temperature of 281 °F. Thus, temperature stresses are not governing criteria. The allowable stresses for this condition are as follows:
ASME Section III B material - code permitted yield AISC material bending - 1.5 x yield bearing - 1.2 x yield shear - 0.8 x yield compression - 1.0 x yield All welds - 0.8 x yield Concrete bearing - 0.8 f 'c UFSAR/DAEC - 1 3.8-10 Revision 23 - 5/15 3.8.2.3.6 Forces Applied by Piping
Piping systems that impose forces on structures from thermal expansion are supported so as to reduce the effect both in the piping and on the structures. When properly supported, the loads on the structures do not substantially change from the normal operation to the postaccident
condition.
3.8.2.4 Design and Analysis Procedures
The drywell concrete reinforcement design was based on a combination of theories of shell design, chimney design, and ring design.
A finite-element analysis was used to evaluate local stresses caused by end moments of deep beams at the drywell interface.
Structural design and construction were performed in such a way as to prevent cracking by mix design, pour limitation, and curing precautions.
The effect of shear was evaluated, and the stresses are within the allowable limits for an
unreinforced web. Thus, no shear reinforcing was required.
All steel structures are designed by the elastic analysis method.
For all structures, minimum factors of safety of 1.5 against overturning and 1.2 against uplift (buoyancy) have been maintained.
3.8.2.4.1 Limiting Stress and Strain for Concrete
At ultimate strength of concrete structural members, concrete stress is not proportional to strain although strains vary linearly with distance from the vertical axis. The actual geometric shape of concrete compression stress distribution varies considerably. To determine the ultimate capacity of the structural member, it is not necessary to know the exact shape of compression stress distribution; only the magnitude of the resultant of the compression stresses and its location need be known. The actual complex stress distribution may be replaced by a fictitious one of some simple geometric shape that results in the same total compression force applied at the same location in the member when it is on the point of failure.
ACI 318 has adopted a rectangular stress distribution with an average compressive stress intensity of 0.85 f 'c corresponding to a strain of 0.003 in the extreme concrete fiber. The average compressive stress of 0.85 f 'c is thought of as acting over part of the section with a depth of 0.85 c, where c is the distance to the actual neutral axis from the compression edge. It has been experimentally verified that ultimate strength of members calculated by means of this assumed equivalent stress block agrees with the actual ultimate strength. UFSAR/DAEC - 1 3.8-11 Revision 23 - 5/15 Columns with or without moments about one or both major axes may be thought of as uni-, bi- or tri-axial stress distributions. The ultimate load or carrying capacity of such members is calculated using the general propositions of ultimate strength design based on rectangular equivalent stress block and ultimate concrete strain of 0.003. Structural members with bi- and tri-axial stresses were designed such that the sums of the strains did not exceed the limiting
strain of 0.003.
The important point is that the maximum concrete strain should not exceed 0.003 under any state of stress because this figure conservatively represents the strain in compression at
which concrete will fail by crushing.
3.8.2.4.2 Use of Ultimate Strength Method
Ultimate strength design method is used to determine the ultimate carrying capacities of all reinforced-concrete structural members. The ultimate carrying capacities of structural members are calculated using f y for reinforcement and 0.85 f 'c for concrete.
3.8.2.4.3 Load Factors
Design loads are increased by overload factors to obtain the ultimate loads to be used for design. Overload factors are selected such that these ultimate loads have an acceptably small probability of ever being exceeded. The theoretical member capacities are lowered by the reduction factor to allow for variations in quality of materials and construction and the approximations and assumptions made in theoretical analysis. Considering the high standards of quality control enforced during the construction of nuclear power plants, the need for the factor in the design of these structures is debatable.
Load factors of 1.0 are used only for combinations of loading conditions that are extremely improbable (e.g., a combination of dead load + live load + DBE + thermal loads +
pipe rupture). It should be noted that the DBE itself is a hypothetical maximum that may never occur during the lifetime of the structure.
In view of the extreme improbability of the occurrence of the loading combinations for
which load factors of 1.0 are used and in view of the availability of some reserve in the ultimate capacity of a structural member corresponding to the factor , it is felt that the load factors of 1.0 are reasonable and justified. UFSAR/DAEC - 1 3.8-12 Revision 23 - 5/15 3.8.2.4.4 Design of Openings
As a minimum requirement, all openings in walls and floors were provided with additional reinforcements at sides and corners. Openings in shear walls were included in the plane stress finite-element analysis. Reinforcement was provided in accordance with the stress
concentration around the openings without exceeding the allowable stresses.
When openings were provided within an area of high-stress concentrations, the openings were analyzed by accepted engineering principles and completely framed as required to carry the
loads.
3.8.2.4.5 Piping Near Containment Penetration
The following design criteria were applied to piping systems in the design of the containment penetrations:
- 1. Nuclear Class I - Large-Diameter Piping
- a. The flued head is anchored to the reactor building concrete structure. The flued heads and associated structures are capable of sustaining all the postulated loads from the process piping. The stresses resulting from the application of these loads will remain below yield unless appropriate analyses can demonstrate that no gross loss of structural integrity is suffered. A metallic bellows-type expansion joint is provided between the flued head fitting and the containment penetration, thereby allowing essentially free thermal expansion of the drywell shell.
- b. The containment penetration is designed for the bellows spring rate forces resulting from thermal displacement of the containment during normal
operation and postaccident conditions.
- c. The stresses in the structural restraints for loadings resulting from a postulated break in the piping, postaccident conditions, or seismic-induced loads are limited to:
Steel 0.9 yield stress Concrete 0.8 f 'c
- d. With the exception of the main steam system, the stresses in the piping and flued head fittings will be within the allowable stresses permitted in ANSI B31.7, including Code Case 70. The stresses in the main steam piping and flued head fitting are within the allowable stresses permitted in
ANSI B31.1.0. UFSAR/DAEC - 1 3.8-13 Revision 23 - 5/15
- 2. Nuclear Class I - Small-Diameter Piping
- a. If the piping is connected directly to the containment nozzle at the flued head, all piping reactions will be transmitted to the containment. In general, however, the penetration is much larger than the piping, and therefore the piping has little influence on it.
- b. The flued head is designed to take the full moment carrying capability of the process pipe.
- c. The stresses in the piping and flued head fitting will be within the allowable stresses permitted in ANSI B31.7, including Code Case 70.
- d. The stresses in the containment vessel from piping loads caused by pipe rupture, postaccident, and seismic conditions will be within the allowable stresses permitted by the 1968 Edition of ASME Code, Section III.
- 3. Nuclear Class II Piping
- a. The piping is connected directly to the containment nozzle at the flued head, and all piping reactions will be transmitted to the containment.
- b. The flued head is designed to take the full moment carrying capability of the process pipe.
- c. The stresses in the flued head fittings and piping from pipe rupture, postaccident, and seismic loads will be within the allowable stresses permitted in ANSI B31.1.0.
The piping systems that could impose significant reactions to the containment penetrations are the Nuclear Class I, large-diameter, high-pressure piping systems. As described above, these piping systems are restrained and anchored in such a way as to preclude damage to the containment penetration. These piping systems are main steam, feedwater, core spray, residual heat removal, high-pressure coolant injection, and reactor water cleanup.
There are several small-diameter piping systems that contain high pressure. Although these are not expected to possess the energy to cause any significant damage, they were investigated and restraints or anchors were placed where required to keep containment penetration stresses to an acceptable limit. Among these are the main steam drain, recirculation loop sample, and the control rod drive return systems. UFSAR/DAEC - 1 3.8-14 Revision 23 - 5/15 3.8.2.5 Structural Acceptance Criteria
The acceptance criteria, generally as prescribed by appropriate codes, are stated in Section 3.8.2.3 along with the load combinations to which each criterion is applicable.
3.8.2.6 Materials, Quality Control, and Special Construction Techniques
3.8.2.6.l General
Detailed specifications and working drawings for the installation of all construction materials were of such scope as to ensure that the quality of work was commensurate with that necessary to preserve the integrity of the plant structures. Noncombustible and fire-resistant materials were used wherever necessary throughout the facility, particularly in areas containing systems or components that affect unit safety.
3.8.2.6.2 Concrete
All concrete work was in accordance with ACI 318-63 and technical specifications incorporating the latest engineering knowledge in quality construction. Admixtures were added to improve the quality and workability of the plastic concrete during placement and to retard the set of the concrete. Maximum practical size aggregate, water-reducing additives, and a low slump were used to minimize shrinkage and creep. Aggregates conformed to "Standard
Specifications for Concrete Aggregate," ASTM Designation C-33.
3.8.2.6.2.1 Portland Cement. Portland cement conformed to "Specifications for Portland Cement," ASTM C-150.
3.8.2.6.2.2 Aggregates. The acceptability of aggregates was based on the ASTM tests listed in Table 3.8-2. These tests were performed by a qualified commercial testing laboratory.
3.8.2.6.2.3 Water. Water used in mixing concrete was clean and free from injurious amounts of oils, acids, alkalis, salts, organic materials, or other substances that could be deleterious to cement, aggregate, or steel.
3.8.2.6.2.4 Mixing and Placing. Concrete mixes were designed in accordance with ACI 613, using materials qualified and accepted for this work. Trial mixes were tested in accordance
with applicable ASTM Codes as indicated in Table 3.8-3. UFSAR/DAEC - 1 3.8-15 Revision 23 - 5/15 Slump, air content, and temperature measurements were taken before cylinders were cast from the sample. Slump tests were performed in accordance with ASTM C-143, "Standard Method of Test for Slump of Portland Cement Concrete." Air content tests are performed in accordance with ASTM C-231, "Standard Method of Test for Air Content of Freshly Mixed Concrete by the Pressure method." Compressive strength tests were made in accordance with ASTM C-39, "Standard Method of Test for Compressive Strength of Molded Concrete
Cylinders."
The evaluation of compressive tests is in accordance with ACI 214-65.
3.8.2.6.2.5 Crack Control. Structural design and construction were performed in such a way as to prevent cracking by mix design, pour limitation, and curing precautions. The stress limits should result in very limited cracking on the order of a few hundredths of an inch. Such
cracking would not significantly affect th e leak resistance of the structure.
3.8.2.6.3 Reinforcing Steel
Reinforcing steel for concrete consists of deformed bars meeting the requirements of
ASTM A615-68 and appropriate ASTM specifications for the various grades and strengths of bars employed. Placing and splicing of bars are in accordance with the requirements of ACI
318-70.
Mill test results were obtained from the reinforcing steel supplier for each heat of steel to substantiate the required composition, strength, and ductility.
ASTM-A-615 states that the deformations shall be measured on one bar of each 10 tons for No. 3 to No. 11 bars and one bar of each 25 tons for No. 14 and No. 18 bars. Random testing by the supplier in this fashion verified that the deformations on the bars met the requirements of
Safety Guide 15.
In addition to this check by the supplier, a Bechtel inspector performed periodic random inspections at the fabricator's plant to verify the compliance of the rebar to ASTM A-615, detail drawings, and requirement specifications. UFSAR/DAEC - 1 3.8-16 Revision 23 - 5/15 3.8.2.6.4 Structural Steel
Structural steel was in conformance with one of the following specifications:
Structural Steel, ASTM A36-70
High-Strength Structural Steel, ASTM A440-66
High-Strength Low-Alloy Structural Manganese Vanadium Steel, ASTM A441-66a
Carbon Steel Plates with Improved Transition Properties, ASTM A442-69
High-Strength Steel Bolts for Structural Joints, ASTM A325
Certified mill test reports or certified reports of tests made by the fabricator or a testing
laboratory in accordance with ASTM A6-69a and the governing specification constituted evidence of conformity with one of the above AS TM specifications. In addition, the fabricator, when appropriate, provided an affidavit stating that the structural steel furnished met the requirements of the grade specified (ASTM C-150).
3.8.3 CONCRETE AND STEEL INTERNAL STRUCTURES OF STEEL CONTAINMENTS
3.8.3.1 General
The design of all steel floor framing systems including connections to vertical members is in accordance with AISC Manual of Steel Construction 1963 and 1970 editions.
3.8.3.2 Reactor Concrete Pedestal
The reactor vessel pedestal is a cylinder with concrete walls. The outside diameter of the pedestal is and the inside diameter was formed by the outside diameter reactor vessel er ection skirt. A 3-in. layer of shotcrete was applied to the inside face of the erection skirt to act as thermal insulation. Reactor vessel loads are transferred from the skirt to the pedestal concrete by means of shear rings welded circumferentially around the skirt.
The cylindrical pedestal cantilevers from a spherical-shaped base formed by the inside of the drywell. The shears and moments are transferred to the drywell foundation concrete by the
erection skirt acting as a shear ring and by friction between the drywell and the concrete.
UFSAR/DAEC - 1 3.8-17 Revision 23 - 5/15 The overall design was based on very conservative assumptions to allow for the complex interactions of the various loads. The loading combinations and allowable stresses are given in
Table 3.8-4. The design stresses are within the allowable stresses resulting in a high safety
factor.
As shown in Table 3.8-4, the accident + DBE + temperature was considered. Since, for the accident condition combined with the OBE, the OBE load factor for ultimate strength design is 1.25, this condition is not as severe as the accident condition + DBE; thus, the accident +
OBE condition need not be considered.
The reactor vessel supports were designed to withstand (1) forces imposed by operating loads, (2) seismic loads, and (3) jet forces resulting from the complete instantaneous severance of any one of the largest connecting pipes. The magnitudes of the seismic forces imposed on the vessel are determined by dynamic analysis as described in Sections 3.7.2 and 3.10.1. Both the
OBE and DBE were considered.
The jet forces considered in the design of the reactor pressure vessel (RPV) supports are those resulting from a complete severance of one of the main steam lines or one of the
recirculation outlet lines. Jet forces considered in the design, calculated as described in Section
3.6.2.2.4.1, were as follows:
RPV Nozzle Force (kips)
Main Steam 325 Reciruclation outlet 393
The transient and steady-state thermal gradients in the reactor pedestal were determined using a finite-element heat flow computer program based on the work of Wilson and Mitchell
- 2. The calculated stresses in the pedestal are within the allowances permitted by the ACI ultimate
strength design.
Since the reactor vessel for the DAEC was field fabricated, the reactor support skirt extends through the drywell vessel foundation, and a bolted support system is not used.
The concrete shear stresses at the RPV pedestal base for the condition of DL + LL + DBE + Jet are 33.5 psi with the drywell empty and 35.1 psi with the drywell flooded. The shear
stresses in the steel shear ring connecting the spherically shaped concrete base to the drywell for the condition of DL + LL + DBE + Jet are 3100 psi with the drywell empty and 3200 psi with the drywell flooded. This assumes no friction between the concrete and steel.
The factor of safety for shear stress at the connection between the bottom of the drywell and the concrete foundation for the DL + LL + DBE + Jet condition is 6.9 with the drywell empty and 4.3 with the drywell flooded. These safe ty factors are based on the yield strength of the material and neglect friction between the drywell steel plates and the concrete. UFSAR/DAEC - 1 3.8-18 Revision 23 - 5/15
3.8.4 OTHER SEISMIC CATEGORY I STRUCTURES
The criteria described here apply to Seismic Category I structures and are intended to supplement applicable industry design codes where necessary to provide design safety margins for rare events like postulated LOCAs, DBEs, tornados, and missiles.
3.8.4.1 Description of the Structures
This section describes the structures, the loads and load combinations, and the static and dynamic analyses used in the structural design of Seismic Category I structures and briefly discusses typical structural elements of the reactor building.
Design procedures used for the reactor building were also used for the other Seismic
Category I structures, such as the control building, offgas stack, turbine building, intake structure, and pump house.
3.8.4.1.1 Reactor Building Floor System
The reactor building has more than one type of floor system. Basically, the floor system is of composite construction that consists of cast-in-place reinforced concrete so interconnected with steel beams or precast concrete members that the component elements act together as a unit.
The floor system is designed to support dead load, equipment loads, laydown loads, piping loads, live loads, and vertical seismic loads based on dead load plus permanent equipment
loads.
The floor slabs are generally designed as one-way continuous slabs in accordance with
ACI 318-63.
Most of the steel beams and girders are designed as composite sections in accordance with AISC using 0.75-in.-diameter headed shear studs.
The refueling floor at is selected to illustrate the implementation of the design criteria because of its critical function and particularly heavy loading. In addition to
the 1000-lb/ft 2 floor live load, the floor slab area supporting the fuel cask and the reactor head is designed to sustain a load of 100 tons distributed over an area of 5 ft 0 in. in diameter or a load of 110 tons distributed over an area of 6 ft 7 in. in diameter.
Table 3.8-5 lists the allowable stresses and loading combinations as they apply to the particular component described. The design stresses are within the allowable stresses, and the system is structurally adequate.
UFSAR/DAEC - 1 3.8-19 Revision 23 - 5/15 3.8.4.1.2 Reactor Building Concrete Wall
The west wall of the reactor building is selected to illustrate the implementation of the
design criteria.
The wall is constructed in the following way:
Substructure - Superstructure - This wall is designed to resist several loading combinations. It acts as a shear wall for seismic forces in the north-south direction. The substructure wall experiences soils and hydrostatic loads, and the superstructure wall is designed to withstand normal wind loads (W) as well as tornado loads (W'). The wall was also investigated for the effect of tornado missiles, and the wall thickness was determined to be adequate.
The governing design conditions and allowable stresses are given in Table 3.8.6.
3.8.4.1.3 Reactor Building Steel Superstructure
The reactor building steel superstructure houses the 100-ton crane runway. It is a steel structure and consists of rigid portal frames. Each frame consists of flange columns and rafters.
Horizontal bracing has been provided in the plane of the roof to transmit lateral forces to vertical
bracing in walls.
All bracing has been designed to resist wind or earthquake loading. The rigid frames
have been designed for dead and live loads, wind loads, or earthquake loads. UFSAR/DAEC - 1 3.8-20 Revision 23 - 5/15 The superstructure has also been designed for tornado loads (W1) on the assumptions that all or part of the metal siding is blown away by the tornado and the basic superstructure frame is subjected to full tornado winds of 300 mph. Under these conditions, the frame will withstand the loading without failure. The stresses may exceed normal allowable stresses, but
will not exceed 90% of the yield stress of structural steel.
Table 3.8-7 summarizes the governing loading combinations, method of analysis and allowable stresses. The design stresses are well within the limits of the allowable stresses.
3.8.4.1.4 Precast T-beams
The T-beams act as supports for the poured-in-place concrete, and thus form a composite concrete floor system.
The concrete elements constructed in separate placements are so interconnected that the elements respond to loads as a unit.
The grade floor is designed to act as a diaphragm for the basement walls and does not experience bending moments and the reversal of stresses at the supports from horizontal seismic loads. However, vertical seismic loadings that are based on dead load and permanent equipment loads will cause tension in the top fibers of the slab system at the supports. Rebars have been
added to account for these additional stresses.
Details of the design connections are shown in Figure 3.8-1.
3.8.4.1.5 Precast Concrete Panels
The connections to the structural columns were designed to withstand a construction wind pressure of 40 psf or suction of 19 psf. They were also checked for the DBE seismic accelerations at the reactor building operating floor level that would be the worst-case seismic
condition.
The mode of failure for the connections under seismic loadings would be pull-out of the
concrete inserts in the panel.
The factor of safety against the failure of all these inserts and the individual panel coming
loose is 18.7.
UFSAR/DAEC - 1 3.8-21 Revision 23 - 5/15 Typical connection details for the reactor building precast panels are shown in
Figure 3.8-2.
3.8.4.1.6 Reactor Building Crane
The crane runway is an integral part of the reactor building superstructure. The anchorage
of the runway rails to the crane runway is designed to resist the horizontal and vertical forces transmitted by the crane during an earthquake. Safety clamps secure the bridge end trucks to the
runway rails.
Safety clamps also secure the trolley to the bridge rails. The crane is designed for the earthquake accelerations at the level of the crane runway. The anchorage clamps and the crane structure are designed for the most severe of the following:
Allowable Stress Increase
D + L + impact and crane forces (per AISC specification)
None D + L + OBE None D + L + DBE 0.9 F y bending, 0.5 F y shear D + tornado 0.9 F y bending, 0.5 F y shear 3.8.4.1.7 Masonry Block Walls
Masonry walls perform the functions of partitions, partition bearing, shielding, shielding bearing, missile shielding, missile protecti on, shielding blockout, and partition/fire walls. Table 3.8-8 lists the masonry walls and their functions.
The NRC IE Bulletin 80-11 required the DAEC to identify and reevaluate the design adequacy of all masonry walls which are in the proximity or have attachments from safety-related piping or equipment such that wall failure could affect a safety-related system. A review of all known masonry block walls that are in buildings containing safety-related components was completed to determine which are near to, or have attachments to, safety-related piping or equipment. There were 445 masonry walls reviewed, and of these, 150 were neither in the proximity of nor had attachments from safety-related piping or equipment. The remainder were considered to be in proximity of or have attachments from safety-related piping or equipment. Table 3.8-8 lists the walls and identifies the safety-related system or the components attached to the wall. Column 2 of the table provides the priority assigned to the wall for reevaluation. The
order of review was as follows: UFSAR/DAEC - 1 3.8-22 Revision 23 - 5/15 Priority 1 Masonry walls which support safety-related piping 2.5 in. or greater.
Priority 2 Masonry walls which support any safety-related item weighing 100 lb or more. Priority 3 Masonry walls which support any nonsafety-related item weighing 100 lb or more, but are in the proximity to or have attachments from safety-related items.
Priority 4 Masonry walls which support loads less than 100 lb, but are in the proximity to or have attachments from safety-related items.
Priority 5 Masonry walls which will not have a detailed survey or be reanalyzed because the wall is neither in proximity to nor attached to safety-related items. The criteria for the reevaluation of the masonry walls can be found in Attachment 3 of
Reference 3. The walls were reevaluated. Reference 3 is the DAEC response to IE Bulletin 80-11 concerning masonry wall design. Reference 4 transmitted DAEC's response to the NRC requests for additional information relative to masonry wall design.
The NRC accepted all but five of the walls as meeting the requirements of IE Bulletin 80-
11 (Reference 5). Three of these walls were reevaluated and found to be acceptable using elastic methods. This was reported to the NRC by Reference 6. The final two were reevaluated and found to be acceptable by using the results of new seismic analyses for the reactor and turbine
building. These new analyses incorporated radiation damping associated with soil-structure interaction and provide more realistic time histories and input acceleration. The input
accelerations for the two walls in question were thereby reduced significantly. This was reported to the NRC by Reference 7.
3.8.4.2 Applicable Codes, Standards, and Specifications
The documents referenced in Section 3.8.2.2 are also applicable to all Seismic Category I
structures.
3.8.4.3 Loads and Loading Combinations
The loading criteria of Section 3.8.2.3 are also applicable for all Seismic Category I structures. In addition, the effects of missiles and of the failure of Nonseismic structures is
considered. UFSAR/DAEC - 1 3.8-23 Revision 23 - 5/15 3.8.4.3.1 Turbine Missiles
All safety-related equipment was evaluated with respect to its shutdown capability in the unlikely event that a turbine- generator failure would result in a postulated turbine missile.
For a further discussion of Turbine Missiles see Section 3.5.1.3.
3.8.4.3.2 Tornado-Generated Missiles
Tornado-generated missiles are discussed in Section 3.5.1. To prevent a loss of function due to tornado-generated missiles, both structural stability and penetration have been
investigated. This concept is used throughout the plant wherever tornado-generated missiles could damage Seismic Category I equipment, unless other provisions are provided in the UFSAR, i.e. missile protection for the EDGs (see Section 8.3.1.3). As stated in Section 8.3.1.3, separation of the EDGs meet single failure criterion and components of the EDGs are located so as to minimize the possibility of damage due to explosions or missiles.
3.8.4.3.3 Interaction Between Seismic Category I and Nonseismic Structures
All Seismic Category I structures are protected from damage by Nonseismic structures
during an OBE or DBE earthquake by physically separating Nonseismic structures from Seismic Category I structures, with the exception of the emergency diesel compartment of the turbine building and the emergency pump house compartment.
A complete dynamic analysis has been conducted for the turbine building to ensure the integrity of Seismic Category I equipment within the building and Seismic Category I equipment and structures adjacent to it.
For structures defined as partially Seismic Category I and partially Nonseismic, those portions of Nonseismic structures housing Seismic Category I equipment are designed in accordance with Seismic Category I design criteria. In addition, the structure as a whole was investigated to ensure that damage to the Nonseismic part would not endanger the area housing Seismic Category I equipment. This generally involves a complete dynamic analysis of the entire building. The structures falling within this category are the turbine building and the pump
house. 2012-012 UFSAR/DAEC - 1 3.8-24 Revision 23 - 5/15 3.8.4.4 Design and Analysis Procedures In general, procedures for all Seismic Category I structures are as given in Section
3.8.2.4. However, a few additional considerations apply.
3.8.4.4.1 Criteria
The Seismic Category I concrete and steel structures are designed considering three interrelated primary functions for the design-loading combinations described in Section 3.8.2.3.
The first consideration is to provide structural st rength equal to or greater than that required to sustain the combination of design loads and provide protection to other Seismic Category I structures and components. The second consideration is to maintain structural deformations within such limits that Seismic Category I components and/or systems will not experience a loss of function. The third consideration is to limit excessive containment leakage by preventing excessive deformation and cracking where containment integrity is required.
In general, structures are analyzed by elastic methods, and structural design is performed using the ultimate strength method for reinforced concrete and working stress method for
structural steel as defined in ACI 318-63, a nd in AISC Manual of Steel Construction, 1970. Finite-element stress analysis and other techniques are also used where applicable or necessary.
3.8.4.4.2 Two- or Three-Dimensional Stresses
Structural steel members of Seismic Category I structures subjected to two- or three-dimensional stress conditions are typically beams and beam columns. Beams subjected to bending about both axes were designed so that the sum of the bending stresses about each axis did not exceed the stress limits recommended by the AISC. Beam columns subjected to bending about both axes and axial load were designed using the interaction formulas presented in the AISC specifications. The interaction formulas contain the allowable stress limits for bending and axial loads as parameters. The allowable stress limits used have appropriate margins of safety built into them.
3.8.4.4.3 Shear and Bond Stresses
The allowable ultimate diagonal tension shear stress permissible on an unreinforced web is f 'c. The biaxial state of shear stress may be thought of as that existing in a beam subjected to shear parallel to both major axes. In such cases, the allowable shear stress in each direction may be taken as half of the above value. A beam subjected to shear in both directions and torsional moments could be thought of as being in a state of triaxial shear stress. The allowable ultimate torsional shear stress is taken as 2.4 f 'c when torsion acts alone. The design for cases involving shear in both principal directions and torsion is based on interaction formulas as recommended
by Mattock. 8 The ultimate bond stress allowed is 6.7 f 'c/D for top bars and 9.5 f 'c/D for other bars. The main requirement for safety against bond failure is that the length of the bar from any
point of steel stress (f s or at most f v) to its nearby force end must be at least equal to its development length. If this requirement is satisfied, the magnitude of the nominal flexural bond stress along the beam is of only secondary importance. The integrity of the members is ensured UFSAR/DAEC - 1 3.8-25 Revision 23 - 5/15 even in the face of possible minor local bond failures. If the actual available length is inadequate for full development, special anchorage, such as hooks, must be provided.
3.8.4.5 Structural Acceptance Criteria
To prevent a loss of function for all structures required for safe shutdown, limiting values were established to control the maximum structural deformations to within defined limits and to provide strength equal to or greater than that required to sustain the loads and limit deformations. The limiting values for deformation and strength are normally set by the need to maintain
structural integrity, the need to prevent structural deformation from displacing the equipment to the extent that the equipment suffers a loss of function, and the need to prevent deformation that would inhibit the ability of the structures to control leakage. In the design of Seismic Category I structures, structural deformations were never the controlling criteria.
Limit design of concrete and plastic design of steel were never used. All steel structures were designed so that for any combination of loads the stresses did not exceed 0.9 F y in bending and tension and 0.5 F y in shear.
The yield strength for reinforced-concrete structures is considered to be the ultimate resisting capacity as calculated from the ultimate strength design provision of the ACI 318-63.
3.8.5 FOUNDATIONS
The overturning moments were calculated using estimated lateral acceleration coefficients exceeding those determined by the reactor building seismic analysis. These overturning moments were included in the finite-element analysis of the foundation mat and rock substructure. This analysis yielded vertical bearing values at the foundation mat and
substructure interface.
The safety factor against the overturning of the reactor building is 2.3 for the DBE combined horizontal and vertical accelerations. UFSAR/DAEC - 1 3.8-26 Revision 23 - 5/15 REFERENCES FOR SECTION 3.8
- 1. Broms, B. B., "Crack Width and Crack Spacing in Reinforced Concrete Members,"
Journal of the ACI, Vol. 62, No. 10, 1965.
- 2. Wilson, E. L., and Mitchell, R. E., Journal of Nuclear Engineering and Design, Vol. 4, 1966.
- 3. Letter from L. D. Root, Iowa Electric, to J. G. Keppler, NRC, Region III,
Subject:
Response to IE Bulletin 80-11 Concerning Masonry Wall Design, dated November 10, 1980 (With Attachment) (LDR-80-335).
- 4. Letter from L. D. Root, Iowa Electric, to H. R. Denton, NRC
Subject:
DAEC, IE Bulletin 80-11; Masonry Wall Design, dated October 6, 1982 (LDR 82-264).
- 5. Letter from D. Vassallo, NRC, to L. Liu,
Subject:
Masonry Wall Design, IE Bulletin 80-11, dated August 22, 1985.
- 6. Letter from R. W. McGaughy, Iowa Electric, to H. R. Denton, NRC,
Subject:
DAEC Masonry Wall Design, IE Bulletin 80-11, dated November 4, 1985.
- 7. Letter from R. W. McGaughy, Iowa Electric, to H. R. Denton, NRC,
Subject:
DAEC Masonry Wall Design, IE Bulletin 80-11, dated April 30, 1986.
- 8. Mattock, "How to Design for Torsion," ACI SP-18.
'
'
'
'fvfc=
' 'fvfc=
'
'fvfc=
'
'
'
'
fvfc=
'
'
'
UFSAR/DAEC - 1 3.9-1 Revision 22 - 5/13 3.9 MECHANICAL SYSTEMS AND COMPONENTS
3.9.1 SPECIAL TOPICS FOR MECHANICAL COMPONENTS
3.9.1.1 General
The introduction to Section 3.7 discusses the organizations responsible for the seismic design of mechanical components for the DAEC. Section 3.7.3 describes analysis methods, procedures, computer codes and criteria used in the seismic analysis of nuclear steam supply system and balance-of-plant mechanical systems and components. The dynamic testing and analysis of piping systems and mechanical components are discussed in Section 3.9.2. Loading conditions, design criteria, and loading combinations for structures and equipment are discussed in Section 3.9.3. Design loads, stress limits, and allowable deformations of the CRD system are discussed in Section 3.9.4. The response of the reactor vessel internals to loads imposed during normal, upset, emergency, and faulted conditions is discussed in Section 3.9.5. The reactor vessel
design is discussed in Section 5.3.3.
3.9.1.2 Reactor Internals Design Analysis
Both elastic and inelastic stress analysis techniques may be used in the design of
the reactor vessel core support and reactor internal structures to show that specified stress limits are not exceeded. When an inelastic stress analysis is performed on these components, the elastic (linear) system analysis is checked to see if the analysis requires modification. The procedure is to perform a linear analysis with the stiffness of the inelastic component reduced to the stiffness value corresponding to the inelastic displacement value. A nonlinear dynamic analysis is performed in lieu of a linear analysis if the natural frequencies of the system with reduced stiffness deviate significantly from those of the unreduced system. Tables 3.9-1 through 3.9-4 summarize the stress results obtained from certain key reactor vessel, core support structure, and reactor internal structure components.
In order to ensure that the reactor and internals are adequately designed to resist the oscillatory nature of blowdown forces, a dynamic analysis was made of the reactor internal components being acted upon by the applied forces. The component natural periods were determined from a comprehensive dynamic model of the reactor pressure vessel (RPV) and internals, including hydrodynamic mass effect of the water inside the
reactor pressure vessel.
The dynamic analysis was performed by determining the natural frequencies and mode shapes of the internal components under consideration. The oscillatory forces were then applied to these components to determine the dynamic load response. Negligible amplification of the applied forces has resulted.
In addition, extensive dynamic analyses of the reactor and internals were performed before the initiation of the preoperational test program as DAEC is considered UFSAR/DAEC - 1 3.9-2 Revision 22 - 5/13 to be the prototype plant for BWR/4-183 inch reactors. The results of these analyses were used to generate the allowable vibration levels during the preoperational test. The
vibration data obtained during the preoperational test were analyzed in detail, and the results of the data analysis, vibration amplitudes, natural frequencies, and mode shapes were then compared to those obtained from the theoretical analysis (Reference 3). Such comparisons, in addition to data from forced oscillation tests, provide the analysts with added insight into the dynamic behavior of the reactor internals. The additional knowledge gained is used in the generation of the dynamic models for vibration, seismic, and LOCA analysis.
The components that were identified as being susceptible to flow- induced vibrations (FIV) in the original design basis for the BWR/4-183 size plants, and the components which have encountered problems in actual operation (such as steam dryers and jet pump sensing lines) were re-evaluated for Extended Power Uprate (EPU). Although operation at EPU conditions reduces the margin to FIV with respect to allowable stress, the resulting maximum stresses remain within the acceptance criteria.
As part of the Steam Dryer Tie Bar Replacement project that occurred in 2010, a comparative analysis was used to evaluate the design change. This type of analysis is described in EPRI report BWRVIP-181-A. BWRVIP-181-A has been reviewed by the NRC and approved through a Safety Evaluation Report. This analysis compared the welded connections and natural frequencies of the new design, which was machined from a solid piece, to the original design, which was an L-bracket.
3.9.2 DYNAMIC TESTING AND ANALYSIS
3.9.2.1 General Requirements
The dynamic analysis of the steam and recirculation piping and restraints with
forcing functions other than earthquake is not routinely performed. The vibration of the recirculation system caused by flow and/or the recirculation pump is measured in the field, and the actual displacements are compared with allowable displacements. Allowable displacements are defined as those displacements that produce stresses whose amplitudes are less than one-half the endurance limit, as defined in the ASME Code, Section III. The dynamic loads acting on the main steam line include those associated with turbine stop valve closure, relief/safety valve lifting, and steam flow. Turbine stop valve closure has been evaluated using plants with a similar geometry and operating conditions as DAEC and has shown to be a small contributor of overall stress (~2%).
The effects of relief and safety valve lifting have been evaluated by an analysis that properly and conservatively accounts for the dynamic loads. Flow-induced vibration in the main steam lines has been shown to be insignificant by actual measurements with remote instrumentation (Section 14.2.14).
Dynamic testing is not required in ASME Code, Section III, for DAEC Class 1 mechanical equipment. However, the seismic stress (obtained from criteria specified in 2011-014 UFSAR/DAEC - 1 3.9-3 Revision 22 - 5/13 Section 5.2.4) of the Class 1 equipment is added to those from other loading conditions to ensure that the equipment will function as required.
Seismic Category I piping is designed to the maximum relative differential movement that could occur at the support points. Since the stresses resulting from this maximum relative differential movement are not likely to occur in phase with the maximum stresses due to dynamic response of the pipe, if any, the two were combined on a root-mean-square basis.
The seismic stresses are combined in accordance with the design rules of
Section 5.2.
The dynamic analysis of the recirculation piping is discussed in Section
3.7.3.8.3.1
3.9.2.2 Piping and Pipe Restraint Structural Integrity Acceptance Criteria
The acceptance criteria that were implemented to confirm the structural integrity
of the piping and pipe restraints in the event of vibratory responses were vibration measurements taken of the final installation of the recirculation piping system during preoperational tests. The actual displacements were compared with allowable displacements. Allowable displacements were defined as those displacements that produce stresses whose amplitudes are less than half of the endurance limit, as defined in
the ASME Code, Section III. The ASME Code, Section III, page 99, defines the endurance limit as "...two times the S a value at 10 6 cycles in the applicable fatigue curve...."
A portable Bentley-Nevada Vibration Meter was used for measuring the vibration amplitudes on the recirculation piping system.
During the testing of main steam turbine stop valve closure and relief valve opening, instrumentation with remote readout capability was installed to provide actual data that could be compared with calculations to ensure that displacements were within allowable limits.
As part of the preoperational test specification for the recirculation piping system, the measurement of the vibration frequencies and amplitudes was required. Frequencies and amplitudes were taken in the horizontal di rection (X and Z directions) at the five locations listed below. When the total deflections (peak to peak) due to vibration were less than those shown in the right-hand column, stresses were less than half of the endurance limit, and therefore acceptable. When deflections exceeded those in the table below, a detailed analysis was made from the actual measured deflections to determine
their acceptability. UFSAR/DAEC - 1 3.9-4 Revision 22 - 5/13 Measurement Location Maximum Deflection (peak to peak)
On suction piping near pump 0.994 in.
Midspan on suction piping 0.043 in.
Midspan on discharge between pump and circular header 0.041 in.
Circular header (any convenient
location) 0.096 in.
Midspan of jet pump riser 0.018 in.
During the preoperational and startup testing program at the DAEC, all other Seismic Category I piping was observed for vibratory responses. When significant vibratory responses were observed, measurements were taken to determine actual displacements, and those measurements were compared with allowable displacements to
ensure that excessive vibration did not occur.
Dynamic testing was also used to assist in the design of Class 1 equipment. The dynamic testing employed was one of the two types described below:
- 1. Free Vibration Test
This test was performed on equipment whose response is dominated by the fundamental mode. The critical damping ratio and fundamental frequency were determined from this test and were used to verify or supplement calculated values used in the dynamic analysis of this equipment. This test was not used alone to demonstrate dynamic
capability.
In this test, an initial displacement or initial velocity is imparted to the equipment. The initial displacement is introduced by forcibly displacing the equipment and then suddenly releasing the force. The initial velocity is obtained by applying an impulse. Accelerometers or strain gauges are mounted on the equipment. After first ensuring that the equipment is vibrating in its primary mode, the critical damping ratio is calculated from the logarithmic decrement.
- 2. Forced Vibration Tests
The equipment is mounted on a shake table or driven by an eccentric UFSAR/DAEC - 1 3.9-5 Revision 22 - 5/13 shaker. The critical damping ratios, resonant frequencies, and the equipment's functional capability are determined.
The critical damping ratio of the equipment is determined by applying a sinusoidal acceleration and measuring the forced-response curve (amplitude versus forcing frequency). The critical damping ratio is then calculated by using the half-power method, fitting a theoretical forced-
response curve through the data points, or direct reading of the resonant amplification. The vibratory motion used is such that the vibratory loads equal or exceed the seismic loads represented by the applicable floor spectra. When testing is the only method used to demonstrate functional capability of equipment, the mounting conditions are simulated and the equipment is operating during and after the tests.
When the seismic testing is supplemented by analysis, the seismic stresses are added to those from normal and accident conditions in the appropriate loading combinations as described in Section 3.7 in order to ensure that the equipment will perform its required safety functions.
As an example, the tests and analysis performed on the HPCI turbine are summarized below.
The major structure of the turbine was qualified by dynamic analysis. The turbine-control-unit components were qualified by dynamic testing on a shake table with electrical and hydraulic systems functional. The-actual mounting brackets were simulated in the test mounting. Vibration in all three perpendicular axes (two horizontal and one vertical) was accomplished by orienting the equipment in three directions on a horizontal shake table. A resonant search was made from 1 to 200 Hz, and the components with substantial resonances below 33 Hz were modified before performing the functional qualification test. These modifications were applied to the standard design. This equipment was then tested with a sinusoidal input of 1.5g and then 3.0g for at least 30
sec at each of the arbitrary frequencies of 10, 15, and 23 Hz in each of the three perpendicular directions, with all systems operational. Since there were no functional failures, the equipment was deemed qualified for up to 3.0g horizontal or vertical maximum floor acceleration for all frequencies 33 Hz and below.
All tests conducted used methods and procedures delineated in the example above. In addition, the amplitudes supplied at the support brackets were equal to or greater than the levels predicted by system dynamic analyses.
All Seismic Category I equipment has been either tested or analyzed to ensure its ability to withstand the design loading requirements.
UFSAR/DAEC - 1 3.9-6 Revision 22 - 5/13 3.9.3 ASME CODE CLASS 1, 2, AND 3 COMPONENTS, COMPONENT SUPPORTS, AND CORE SUPPORT STRUCTURES
3.9.3.1 Loading Combinations, Design Transients, and Stress Limits
3.9.3.1.1 Loading Conditions
The loading conditions, which are to be considered in addition to loads from normal conditions, are divided into three categories: upset conditions, emergency
conditions, and faulted conditions. The conditions are defined as follows:
- 1. Normal Conditions
Any condition in the course of operation of the plant under planned, anticipated conditions, in the absence of upset, emergency, or faulted
conditions.
- 2. Upset Conditions
Any deviations from normal conditions anticipated to occur often enough that the design should include a capability to withstand the conditions without operational impairment. The upset conditions include abnormal operational transients caused by a fault in a system component requiring its isolation from the system, transients due to the loss of load or power, and any system upset not resulting in a forced outage. The upset
conditions include the effect of the specified earthquake for which the system must remain operational or must regain its operational status.
- 3. Emergency Conditions
Any deviations from normal conditions that require shutdown for the correction of the conditions or repair of damage in the system. The conditions have a low probability of occurrence but are included to provide assurance that no gross loss of structural integrity will result as a result of any damage.
- 4. Faulted Conditions
Those combinations of conditions associated with extremely low-probability postulated events whose consequences are such that the integrity and operability of the nuclear system may be impaired to the
extent where considerations of public health and safety are involved. Such considerations require compliance with safety criteria as may be specified by jurisdictional authorities. Among the faulted conditions may
be a specified earthquake for which safe shutdown is required. UFSAR/DAEC - 1 3.9-7 Revision 22 - 5/13 3.9.3.1.2 Safety Margins
In addition to the definitions above, the meaning of these terms is expanded in
quantitative probabilistic language. The purpose of this expansion is to clarify the
classification of any hypothesized accident or sequence of loading events so that the appropriate structural safety margins are applied. Knowledge of the event probability is necessary to establish meaningful and adequate safety factors for structural design. The following summary illustrates the quantitative event classifications:
Generic Definition P 40 , 40-Year Interval Event Encounter Probability Upset (likely) 1.0 > P 40 10-1 Emergency (low probability) 10-1 > P 40 10-3 Faulted (extremely low probability) 10-3 > P 40 10-6 The event probabilities currently in use as governing accident or faulted
conditions are the following:
P 40 (N and U and A o) = 10-1 to 10-2 P 40 (N and A d) = 10-3 P 40 (N and R) =10 -3 P 40 (N and A d and R) 1.5 x 10-6 where
N= normal loads U= upset loads excluding earthquake
A o = OBE A d = DBE R= any pipe rupture
The minimum safety factor decreases as the event probability diminishes; and if the event is too improbable (incredible: P 40 10-6), then no safety factor is appropriate or required. The minimum safety factor used in design is that associated with an event probability of 10 -5. The term SFmin , which appears in Table 3.9-5, is identical with the classical definition of a minimum safety factor on load or deflection. SFmin is related to the event probability by the following equation:
SFmin = 9 3 - log 10 P 40 where 10-1 > P 40 10-6 UFSAR/DAEC - 1 3.9-8 Revision 22 - 5/13 These expressions show the probabilistic significance of the classical safety factor concept as applied to reactor safety. The SFmin values corresponding to the current governing accident event probabilities are summarized as follows:
P 40 SFmin 10-1 2.25 10-2 1.80 10-3 1.50 10-5 1.125
3.9.3.1.3 Governing Loading Conditions and Criteria
The governing loading conditions are summarized as follows:
- 1. N and U and A o
- 2. N and A d
- 3. N and R
- 4. N and A d and R The loading criteria are covered by the ASME Code, Section III, other industrial
codes, or special criteria where no applicable standards exist. These special criteria are classified into four categories: deformation limits, primary stress limits, buckling stability limits, and fatigue limits.
Table 3.9-6 summarizes the loading conditions and criteria. Table 3.9-5 lists the
special loading criteria.
3.9.3.1.4 Design Criteria
These general design criteria are intended to apply to those ductile metallic structures or components that are normally designed using rational stress analysis techniques, such as pressure vessels and core support structures. The criteria may also be applied to those components or structures whose ultimate loading capability is determined by tests. These criteria are intended to supplement applicable industry design codes where necessary. Compliance with these criteria provides design safety margins that are appropriate to extremely reliable structural components when account is taken of rare event potentialities such as might be associated with a DBE or primary pressure boundary coolant pipe rupture, or a combination of events. UFSAR/DAEC - 1 3.9-9 Revision 22 - 5/13 There are many important Seismic Category I components or pieces of equipment that are not normally designed or sized directly by stress analysis techniques. Simple stress analyses are sometimes used to augment the design of these components, but the primary design does not depend on detailed stress analysis. These components are usually designed by tests and empirical experience. Complete detailed stress analysis is currently neither meaningful nor practical for these components. Examples of such components are valves, pumps, electrical equipment, and mechanisms. Field experience and testing are used to support the design. Where the structural or mechanical integrity of components is essential to safety, the components referred to in these criteria must be designed to accommodate the DBE, the OBE, a design-basis pipe rupture, or an appropriate combination. The reliability requirements of such components cannot be
quantitatively described in a general criterion because of the varied nature of each component and its specific function in the system.
3.9.3.1.5 Justification of Square Root of the Sum of the Squares (SRSS) Load Combination
Nuclear power plant structures and equipment are designed to accommodate many operational and transient-tape loads and load combinations. Included are such
postulated events as the loss-of-coolant accident (LOCA) and high-intensity earthquakes. In most cases for dynamic loadings, the peak structural dynamic response is calculated using linear elastic multidegree of freedom system analysis. When two or more structural dynamic responses are considered, the peak response from each dynamic loading event is combined using the square root of the sum of the squares (SRSS) rule. Subsequently, this combined peak dynamic response is added absolutely to the calculated static or slowly varying response. The use of the SRSS method for combining dynamic
responses has been restricted to separate physical events that are postulated to occur together but for which the precise timing (phasing) is unknown. Furthermore, the use of the SRSS combinations of peak responses has also been limited to physical loading events and structural systems where the dynamic responses have rapidly varying amplitudes and short duration responses. The application of SRSS combination is limited to combining earthquake, LOCA, and safety relief valve (SRV) actuation dynamic
responses.
To date, technical justification for the use of the SRSS rule has centered around
the fact that
- 1. The maximum peaks of individual responses are highly unlikely to coincide in time.
- 2. The large conservatism in the total design process ensures sufficient structural design margin to protect against failure.
The dynamic design margin inherent in nuclear plant structures designed to meet ASME Code (or equivalent) stress limits is related to the energy absorption capability of the component. For structures exhibiting even moderately ductile behavior, this margin is
significantly greater than the ratio of static failure stress to code allowable, and therefore UFSAR/DAEC - 1 3.9-10 Revision 22 - 5/13 ensures enough additional margin to protect against structural failure even in the unlikely event that the peak combined dynamic response exceeds the SRSS value.
The use of the SRSS method is technically justifiable and represents a prudent and practical engineering approach for the combination of dynamic responses originating from earthquake, LOCA transients, and SRV discharge loadings.
Because the phasing between simultaneous dynamic responses such as earthquake
and LOCA, is unknown, the SRSS rule provides a reasonable representation of resultant responses assuming their simultaneous occurrence. Combining the individual responses by the direct addition of the individual response peaks by the so-called absolute sum (ABS) method is unnecessarily conservative because of the low probability of the peaks being coincident. Hence, this "probable" response combination, accounting for the low probability of coincidence, is appropriate and thus has been the most common practice. Such a probable response combination is provided by the SRSS rule.
The SRSS rule has been used in the following three distinct engineering
applications:
- 1. Combination of seismic modal responses.
- 2. Combination of three responses along each of the three earthquake directions.
- 3. Combination of two or more dynamic responses in a given direction.
The justification for the use of the SRSS method in these applications is based on random vibrations and probability theory, and the observation of the following:
- 1. Amplitude variations in time and uncertain phasing of the responses.
- 2. Rapid variation and the short duration of the peak responses.
Thus, the maximum peaks of individual responses are unlikely to coincide. Furthermore, the probability of the actual response combination significantly exceeding the SRSS value is exceedingly small.
In addition to the low probability of significantly exceeding the SRSS value, the
consequences of exceedance are negligible because of the following:
- 1. The existence of margins at all steps of the design process, for example, dynamic load definition, structural models, damping parameters, and
allowable stress values in industry codes.
- 2. The additional inherent structural dynamic reserve margins due to the short duration of the peak responses and energy absorption capability of
the structure. UFSAR/DAEC - 1 3.9-11 Revision 22 - 5/13 3.9.3.2 Pipe Support Attachment Review
In response to IE Bulletin 79-02, the design and installation of pipe support
structural baseplates, the concrete expansion bolts attaching the baseplates to the concrete structure, and the masonry block walls were reviewed. By a combination of analysis and test, it was found that a small percentage of seismic supports had a smaller safety margin than had been intended. Appropriate modifications were made. A complete report of the review was submitted to the NRC by Iowa Electric letter LDR-80-283, dated October 17, 1980.
3.9.3.3 Critical Elements of the Station Piping
Critical elements of the station piping, including the connections of piping to the drywell and pressure suppression chamber of the primary containment, are designed to withstand, without overstress, the maximum forces resulting from the DBE. This has been accomplished by appropriate analyses of the important piping in systems critical to
reactor safety or to safe shutdown of the station. The stresses resulting from these earthquake forces have been calculated to be within the acceptable limits for the piping materials and other associated components involved, according to appropriate ANSI and
ASME Codes.
3.9.3.4 Review of Design Stress Calculations
An independent review of design stress calculations as well as assumptions used in the calculations has been performed within Bechtel and GE. In addition, the displacement of recirculation, steam, and feedwater system piping was observed in the
hot condition during startup testing of the DAEC to confirm design calculations.
3.9.4 CONTROL ROD DRIVE SYSTEM
Safety Objective
The safety objective of the control rod drive (CRD) mechanical design is to insert the control rods with sufficient speed to limit fuel barrier damage.
UFSAR/DAEC - 1 3.9-12 Revision 22 - 5/13 Power Generation Objective The power generation objective of the CRD mechanical design is to position the
control rod within the core to control power generation.
Safety Design Bases
The reactivity control mechanical design meets the following safety design bases:
- 1. Design provides for a sufficiently rapid control rod insertion so that no fuel damage results from any abnormal operating transient.
- 2. Design includes positioning devices, each of which individually supports and positions a control rod.
- 3. Each positioning device
- a. Prevents its control rod from withdrawing as a result of a single malfunction.
- b. Is designed so that no single failure of a component will prevent its control rod from being inserted.
- c. Is individually operated so that a failure in one positioning device does not affect the operation of any other positioning device.
- d. Is individually energized when rapid control rod insertion (scram) is signaled so that a failure of power sources external to the positioning device does not prevent other positioning devices' control rods from being inserted.
- e. Is locked to its control rod to prevent undesirable separation.
Power Generation Design Basis
The reactivity control mechanical design provides for positioning the control rods
to control power generation in the core.
3.9.4.1 Descriptive Information of Control Rod Drive System
The CRD mechanisms are part of the CRD system, which hydraulically operates the CRD mechanisms using processed condensate water as hydraulic fluid. The CRD mechanisms operate manually to position the control rods but act automatically to rapidly insert the control rods during abnormal conditions requiring rapid shutdown (scram).
The control rods, CRD mechanisms, and that part of the CRD hydraulic system necessary for scram are designed as Seismic Category I equipment. The piping and UFSAR/DAEC - 1 3.9-13 Revision 22 - 5/13 valves in the CRD system that are required to effect a scram and serve as part of the reactor coolant pressure boundary are designed and fabricated to high-quality levels as
defined in Section 17.1.
3.9.4.1.1 Control Rod Drive Mechanisms
The CRD mechanism (drive) used for positioning the control rod in the reactor core is a double-acting, mechanically latched, hydraulic cylinder using water from the
condensate storage tank as its operating fl uid (see Figures 3.9-1, 3.9-2, 3.9-3, and 3.9-4). The individual drives The drives do not interfere with refueling and are operative even when the head is removed from the reactor vessel. The drives are also readily accessible for inspection and servicing. The location makes maximum use of the water in the reactor as a neutron shield and gives the least possible neutron exposure to the drive components. Using reactor water from the condensate storage tank as the operating fluid eliminates the need for special hydraulic fluid. Drives can use simple piston seals whose leakage does not contaminate the reactor water and does cool the drive mechanisms and their seals.
The drives can insert or withdraw a control rod at a slow, controlled rate, as well as provide rapid insertion when required. A mechanism on the drive locks the control rod in 6-in. increments of stroke over the 12-ft length of the core.
A coupling spud at the top end of the drive index tube (piston rod) engages and locks into a mating socket at the base of the control rod. The weight of the control rod is sufficient to engage and lock this coupling. Once locked, the drive and rod form an integral unit that must be manually unlocked by specific procedures before components
can be separated.
The drive holds its control rod in distinct latch positions until the hydraulic system actuates movement to a new position. An alarm annunciates if the withdraw overtravel limit on the drive is reached. Normally, the seating of the control rod at the
lower end of its stroke prevents reaching the drive withdraw overtravel limit. If the drive can reach the withdrawal overtravel limit, this means the control rod is uncoupled from
its drive.
The individual rod indicators, grouped in one control panel display, correspond to
relative rod locations in the core. A separate, four-rod display is located just below the
large display on the vertical part of the benchboard. This display presents the positions of the control rod selected for movement and of the other rods in the affected rod group. For
display purposes, the control rods are considered in groups of four adjacent rods centered around a common core volume. Each group is monitored by four LPRM strings (see Section 7.6). Rod groups at the periphery of the core may have less than four rods. The four-rod display shows the positions, in digital form, of the rods in the group to which the
selected rod belongs. A white light indicates whic h of the four rods is the one selected for movement.
UFSAR/DAEC - 1 3.9-14 Revision 22 - 5/13 Drive Components. Figure 3.9-2 illustrates the operating principle of a drive, and Figures 3.9-3 and 3.9-4 illustrate the drive in more detail. The main components of the drive and their functions are described below.
The drive piston is mounted at the lower end of the index tube. This tube functions as a piston rod. The drive piston and index tube make up the main moving assembly in the drive. The drive piston operates between positive end stops, with a
hydraulic cushion provided at the upper end onl
- y. The piston has both inside and outside seal rings and operates in an annular space between an inner cylinder (fixed piston tube)
and an outer cylinder (drive cylinder). Because the type of inner seal used is effective in only one direction, the lower sets of seal rings are mounted with one set sealing in each
direction.
A pair of nonmetallic bushings prevents metal-to-metal contact between the piston assembly and the inner cylinder surface. The outer piston rings are segmented step-cut seals with expander springs holding the segments against the cylinder wall. A pair of split bushings on the outside of the piston prevents contact with the cylinder wall. The effective piston area for downtravel, or withdrawal, is approximately 1.2 in. 2 versus 4.0 in.2 for uptravel, or insertion. This difference in driving area tends to balance the control rod weight and ensures a higher force for insertion than for withdrawal.
The index tube is a long hollow shaft made of nitrided Type 304 stainless steel. Circumferential locking grooves, spaced every 6 in. along the outer surface, transmit the weight of the control rod to the collet assembly.
The collet assembly serves as the index tube locking mechanism. It is located in the upper part of the drive unit. This assembly prevents the index tube from accidentally moving downward. The assembly consists of the collet fingers, a return spring, a guide
cap, a collet housing (part of the cylinder, tube, and flange), and the collet piston seals.
Locking is accomplished by six fingers mounted on the collet piston at the top of
the drive cylinder. In the locked or latched position, the fingers engage a locking groove
in the index tube.
The collet piston is normally held in the latched position by a force of approximately 150 lb supplied by a spring. Metal piston rings are used to seal the collet piston from reactor vessel pressure. The collet assembly will not unlatch until the collet fingers are unloaded by a short, automatically sequenced, drive-in signal. A pressure, approximately 180 psi above reactor vessel pressure, must then be applied to the collet piston to overcome spring force, slide the collet up against the conical surface in the
guide cap, and spread the fingers out so they do not engage a locking groove. The collet piston is nitrided to minimize wear. UFSAR/DAEC - 1 3.9-15 Revision 22 - 5/13 A guide cap is fixed in the upper end of the drive assembly. This member provides the unlocking cam surface for the collet fingers and serves as the upper bushing for the index tube. It is nitrided to provide a compatible bearing surface for the index tube.
The center tube of the drive mechanism forms a well to contain the position indicator probe. This probe is an aluminum extrusion attached to a cast aluminum housing. Mounted on the extrusion are hermetically sealed, magnetically operated,
position indicator switches. Each switch is sheathed in a braided glass sleeve, and the entire probe assembly is protected by a thin-walled stainless steel tube. The switches are actuated by a ring magnet located at the bottom of the drive piston. The drive piston, piston tube, and indicator tube are all of nonmagnetic stainless steel, allowing the individual switches to be operated by the magnet as the piston passes. One switch is
located at each position corresponding to an i ndex tube groove, thus allowing indication at each latching point. An additional switch is located at each midpoint between latching points to indicate the intermediate positions during drive motion. Thus, indication is
provided for each 3 in. of travel. Duplicate switches are provided for the full-in and full-
out positions. One additional switch (an overtravel switch) is located at a position below the normal full-out position. Because the limit of downtravel is normally provided by the
control rod itself as it reaches the backseat position, the drive can pass this position and actuate the overtravel switch only if it is uncoupled from its control rod. A convenient means is thus provided to verify that the drive and control rod are coupled after installation of a drive or at any time during plant operation.
A flange and cylinder assembly is made up of a heavy flange welded to the drive cylinder. A sealing surface on the upper face of this flange makes the seal to the drive
housing flange. Teflon-coated, stainless steel ri ngs are used for these seals. In addition to the reactor vessel seal, the two hydraulic control lines to the drive are sealed at this face. A drive can thus be replaced without removing the control lines, which are permanently welded into the housing flange. The drive flange contains the integral ball, or two-way, check (ball-shuttle) valve. This valve directs either the reactor vessel
pressure or the driving pressure, whichever is higher, to the underside of the drive piston. Reactor vessel pressure is admitted to the valve from the annular space between the drive
and drive housing through passages in the flange. A strainer is provided to intercept foreign material at this point.
Water used to operate the collet piston passes between the outer tube and the
cylinder tube. The inside of the cylinder tube is honed to provide the surface required for
the drive piston seals.
Both the cylinder tube and outer tube are welded to the drive flange. The tops of
these tubes have a sliding fit to allow for differential expansion.
The upper end of the index tube is threaded to receive a coupling spud. The coupling (see Figure 3.9-1) accommodates a small amount of angular misalignment
between the drive and the control rod. Six spring fingers allow the coupling spud to enter UFSAR/DAEC - 1 3.9-16 Revision 22 - 5/13 the mating socket on the control rod. A plug then enters the spud and prevents uncoupling.
Two means of uncoupling are provided. With the reactor vessel head removed, the lock plug can be raised against the spring force of approximately 50 lb by a rod
extending up through the center of the control rod to an unlocking handle located above the control rod velocity limiter. The control rod, with the lock plug raised, can then be lifted from the drive.
The lock plug can also be pushed up from below, if it is desired to uncouple a drive without removing the reactor pressure vessel head for access. In this case, the central portion of the drive mechanism is pushed up against the uncoupling rod assembly, which raises the lock plug and allows the coupling spud to disengage the socket as the
drive piston and index tube are driven down.
The control rod is heavy enough to force the spud fingers to enter the socket and push the lock plug up, allowing the spud to enter the socket completely and at the plug to
snap back into place. Therefore, the drive can be coupled to the control rod using only
the weight of the control rod. However, with the lock plug in place, a force in excess of
50,000 lb is required to pull the coupling apart.
Materials of Construction. Factors that determine the choice of construction materials are discussed below.
The index tube must withstand the locking and unlocking action of the collet fingers. A compatible bearing combination must be provided that is able to withstand moderate misalignment forces. The reactor environment limits the choice of materials suitable for corrosion resistance. The column and tensile loads can be satisfied by an
annealed 300 series stainless steel. The wear and bearing requirements are provided by Malcomizing the completed tube. To obtain suitable corrosion resistance, a carefully
controlled process of surface preparation is used.
The coupling spud is made of Inconel 750 that is aged for maximum physical strength and the required corrosion resistance. Because misalignment tends to cause chafing in the semispherical contact area, the entire part is protected by a thin vapor-deposited chromium plating (electrolized). This plating also prevents the galling of the
threads attaching the coupling spud to the index tube.
Inconel 750 is used for the collet fingers, which must function as leaf springs when cammed open to the unlocked position. Colmonoy 6 hard facing provides a long-
wearing surface, adequate for design life, to the area contacting the index tube and unlocking cam surface of the guide cap.
Graphitar 14 is selected for seals and bushings on the drive piston and stop piston. The material is inert and has a low friction coefficient when water lubricated. UFSAR/DAEC - 1 3.9-17 Revision 22 - 5/13 Because a loss of seal strength is experienced at higher temperatures, the drive is supplied with cooling water to hold the temperature below 250°F. The Graphitar is relatively soft, which is advantageous when an occasional particle of foreign matter
reaches a seal. The resulting scratches in the seal reduce sealing efficiency until worked smooth, but the drive design can tolerate considerable water leakage past the seals into
the reactor vessel.
All drive components exposed to reactor vessel water are made of AISI 300 series
stainless steel except for the following:
- 1. Seals and bushings on the drive piston and stop piston are Graphitar 14.
- 2. All springs and members requiring spring action (collet fingers, coupling spud, and spring washers) are made of Inconel 750.
- 3. The ball check valve is a Haynes Stellite cobalt-base alloy.
- 4. Elastomeric O-ring seals are ethylene propylene.
- 5. Collet piston rings are Haynes 25 alloy.
- 6. Certain wear surfaces are hard faced with Colmonoy 6.
- 7. Nitriding by a proprietary new Malcomizing process, electrolizing (a vapor deposition of chromium), and chromium plating are used in
certain areas where resistance to abrasion is necessary.
- 8. The drive piston head is made of Armco 17-4PH.
Pressure-containing portions of the drives are designed and fabricated in accordance with requirements of the ASME Code, Section III.
3.9.4.1.2 Control Rod Drive Hydraulic System
The CRD hydraulic system (Figure 3.9-5) supplies and controls the pressure and flow to and from the drives. One supply subsystem supplies water to all hydraulic
control units (HCU) at the correct flow. Each hydraulic control unit controls the flow to and from one drive. The water discharged from the drives during a scram flows through the hydraulic control units to the scram discharge volume. The water discharged from a drive during a normal control rod positioning operation returns to the reactor vessel
through a reverse flow path involving the insert exhaust directional control valves of
nonactuated CRD hydraulic control units.
Originally, the water discharged from a drive during normal rod positioning
operation flowed back to the reactor vessel through a control rod drive return line. UFSAR/DAEC - 1 3.9-18 Revision 22 - 5/13 However, in order to eliminate the potential for thermal fatigue cracking of the return line nozzle at the reactor vessel, a blind (spectacle) flange was installed on the return line in an area of the reactor building that is accessible under most accident conditions. If emergency conditions render it desirable to provide water to the reactor vessel via the return line, the blind flange can be removed.
CRD Supply and Discharge Subsystems. The CRD hydraulic supply and discharge subsystems (Figures 3.9-5, 3.9-6 and 3.9-7) control the pressure and flow required to operate the CRD mechanisms. These hydraulic requirements, identified by the function they perform, are as follows:
- 1. An accumulator charging pressure of approximately 1400 to 1500 psig is required. Flow to the accumulators is required only during scram reset or system startup.
- 2. Drive pressure of approximately 260 psi above reactor vessel pressure is required. A flow rate of approximately 4 gpm to insert a control rod and 2 gpm to withdraw a control rod is required.
- 3. Cooling water to the drives is supplied at approximately 20 psi above reactor vessel pressure. (Cooling water can be interrupted for short periods without damaging the drive.)
- 4. The scram discharge header pipe is sized to receive and contain all the water discharged by the drives during a scram. A scram discharge volume of at least 3.34 gal per drive is required. The scram discharge volume normally contains air at atmospheric pressure, except during scram when it is filled with water and air and until the scram signal is cleared. The scram discharge volume will reach reactor pressure following a scram.
- 5. The scram valve pilot air header supplies compressed air to the hydraulic control units for operation of the scram inlet valves and scram exhaust valves through the scram pilot valve. In the event of a scram, the header is
depressurized by action of which are operated from the reactor protection system trip system and provide a backup to the scram pilot valves. In the event of an ATWS-ARI actuation, the header
would also be depressurized by action of which are operated from the ATWS-ARI system as described in section 7.2.3.
The CRD hydraulic supply and discharge systems provide the required functions with the pumps, filters, valves, instrumentation, and piping shown in Figure 3.9-5 and
described in the following paragraphs.
Duplicate components are included where necessary, to ensure continuous system operation if an inservice component requires maintenance.
UFSAR/DAEC - 1 3.9-19 Revision 22 - 5/13 One supply pump pressurizes the system with deaerated, low conductivity water from the condensate reject line.
One parallel spare pump is provided for standby. Each pump is installed with a
suction strainer. A discharge stop-check valve prevents bypassing flow through the nonoperating pump. Flow is continuously diverted through a minimum-flow bypass connection into the condensate storage tank. Consequently, pump overheating is avoided if the pump discharge is inadvertently closed.
Two parallel filters remove foreign material larger than 50 m absolute (25 m nominal) from the hydraulic supply subsystem water. Only one filter is in operation at any given time. The filter installation allows the addition of 2 ft minimum thickness of temporary external radiation shielding for personnel protection. A local differential-pressure indicator and control room alarm monitor the filter element as it collects foreign material. A strainer in each filter discharge line protects the hydraulic system in the event of filter-element failure.
Accumulator charging pressure is established by the discharge pressure of the system supply pump and is maintained by the system flow control valve during normal operation. During scram, the scram inlet (and outlet) valves open and permit the stored energy in the accumulators to discharge into the drives. The resulting pressure decrease in the charging water header allows the CRD supply pump to "run out" (i.e., flow rate to
increase substantially) into the control rod drives via the charging water header. The flow sensing system upstream of the accumulator charging header detects high flow and closes the flow control valve. This action maintains increased flow through the charging
water header.
Pressure in the charging header is monitored in the control room with a pressure indicator and high pressure alarm. Individual accumulators have low-pressure alarms.
During normal operation, the flow control valve maintains a constant system flow rate. This flow is used for drive flow, drive cooling, and system stability.
Drive water pressure required in the drive header is maintained by the drive pressure control valve, which is manually adjusted from the control room. A flow rate of approximately 6 gpm (the sum of the flow rates required to insert and withdraw a control rod) normally passes from the drive water pressure stage through two solenoid-operated stabilizing valves (arranged in parallel) and then goes into the return line downstream of
the cooling pressure control valve. The flow through one stabilizing valve equals the drive insert flow; that of the other stabilizing valve equals the drive withdrawal flow. When operating a drive, the required flow is diverted to that drive by closing the
appropriate stabilizing valve. Thus, flow through the drive pressure control valve is
always constant.
Flow indicators in the drive water header and in the line downstream of the
stabilizing valves allow the flow rate through the stabilizing valves to be adjusted when
necessary. UFSAR/DAEC - 1 3.9-20 Revision 22 - 5/13 Differential pressure between the reactor vessel and the drive pressure stage is indicated in the control room.
A small amount (nominally .012 gpm) of CRD drive water is diverted to each of the two ambient reference columns for the reactor water level instrumentation to purge the columns of non-condensible gases (see Section 7.6.4.6).
The cooling water header is located upstream from the cooling pressure control valve. Water not required for drive cooling passes through the cooling pressure control valve to the reactor vessel. The cooling pressure control valve is manually adjusted from the control room to produce the required cooling water pressure.
The flow through the flow control valve is virtually constant. Therefore, once
adjusted, the drive pressure control valve and the cooling pressure control valve can maintain their required pressure independent of reactor pressure. Changes in the setting
of the pressure control valves are required only to adjust for the changes in the cooling requirements of the drives, as their seal characteristics change with time. A flow indicator in the control room monitors cooling water fl ow. A differential-pressure indicator in the control room indicates the difference between reactor vessel pressure and drive cooling
water pressure. Although the drives can function without cooling water, seal life is shortened by long-term exposure to reactor temperatures. The temperature of each drive is recorded in the control room, and excessive temperatures are annunciated.
The scram discharge volume is used to limit the loss of reactor water discharged from all the drives during a scram. It is also used to contain the reactor water that leaks past the drives following a scram. This volume is provided in the scram discharge header.
During normal plant operation, each scram discharge header is empty, and the
drain and vent valves are open. Position indicator switches on the drain and vent valves actuate valve lights in the control room.
During a scram, the scram discharge volume partly fills with water discharged from above the drive pistons. While scrammed, the CRD seal leakage from the reactor continues to flow into the scram discharge volume until the discharge volume pressure
equals the reactor vessel pressure. A check valve in each hydraulic control unit prevents reverse flow from the scram discharge header volume to the drive.
A test pilot valve allows the discharge volume vent and drain valves to be tested
without disturbing the reactor protection system or the ATWS-ARI system. Closing the discharge volume valves allows the outlet scram valve seats to be leak tested by timing the accumulation of leakage inside the scram discharge instrument volume. UFSAR/DAEC - 1 3.9-21 Revision 22 - 5/13 The initial design of the scram discharge volume included six level-measuring magnetrol float switches to prevent operating the reactor without sufficient free volume present to accommodate scram discharge. In response to IE Bulletin 80-17 (Reference 1), the scram discharge volume was modified by adding a redundant and diverse set of six thermally-actuated liquid level switches and a redundant set of vent and drain valves. The modifications were implemented to ensure that there will be no accumulation of water in the scram discharge volume and that there will be no blowdown if a single vent or drain valve does not close on a scram signal. Both sets of level switches are set at the same setpoint levels and provide identical functions (scram, rod withdrawal block, and alarm). At the first (lowest) level, one level switch initiates an alarm. At the second level, one level switch initiates a rod withdrawal
block to prevent further withdrawal of any control rod. At the third (highest) level, the four level switches (two for each reactor protection system trip system) initiate a scram to shut down the reactor while sufficient free volume is still present in the scram discharge volume to receive the scram discharge water. After a scram and before reactor operation can be resumed, the level-measuring switches must be cleared by draining the scram discharge volume. When the initial scram signal is cleared from the reactor protection system, the scram discharge instrument volume signal is overridden with a key-lock override switch, and the scram discharge volume is drained. Additional control room information regarding the sequence of events during and after a scram is obtained by computer logging of changes in the status of the scram discharge volume vent and drain valves (two vent valves and two drain valves) and 10 of the 12 scram discharge volume level measuring switches (the two "alarm" switches are not monitored). Piping and equipment pressure parts in the CRD hydraulic supply and discharge subsystems are designed and fabricated in conformance with the guidance as described in
Sections 3.1 and 3.2. Hydraulic Control Units. Each hydraulic control unit furnishes pressurized water, on signal, to a drive unit. The drive then positions its control rod as required. The operation of the electrical system that supplies scram and normal control rod positioning
signals to the hydraulic control unit is described in Section 7.7.3.
The basic components in each hydraulic control unit are manual, pneumatic, and electric valves; an accumulator; related piping; electric connections; filter; and instrumentation (see Figures 3.9-5, 3.9-6, and 3.9-8). These components and their
functions are described below.
The insert drive valve is solenoid operated and opens on an insert signal. The valve supplies drive water to the bottom side of the main drive piston.
The insert exhaust valve also opens by solenoid on an insert signal. The valve discharges water from above the drive piston to the exhaust water header. UFSAR/DAEC - 1 3.9-22 Revision 22 - 5/13 The withdraw drive valve is solenoid-operated and opens on a withdraw signal. The valve supplies drive water to the top of the drive piston.
The solenoid-operated withdraw exhaust valve opens on a withdraw signal and discharges water from below the main drive piston to the exhaust header. It also serves as the settle valve. The valve opens following any normal drive movement (insert or withdraw) to allow the control rod and its drive to settle back more quickly into the
nearest latch position.
The speed control valves regulate the control rod insertion and withdrawal rates during normal operation. They are manually adjustable flow control valves used to regulate the water flow to and from the volume beneath the main drive piston. A correctly adjusted valve does not require readjustment except to compensate for changes
in piston seal leakage.
The scram pilot valve is operated from the reactor protection system trip system. A single scram pilot valve controls both the scram inlet valve and the scram exhaust valve. The scram pilot valve is a three-way, dual-solenoid-operated, in-line, normally
energized valve. On a loss of electric signal to the solenoid coils, such as the loss of
external ac power, both coils deenergize and the exhaust port opens on the pilot valve. The pilot valve (Figures 3.9-5 and 3.9-6) is designed so that the trip system signal must be removed from both solenoids before air pressure can be discharged from the scram valve operators. This prevents the inadvertent scram of a single drive in the event of a
failure of one of the solenoid pilot valves.
The scram inlet valve opens to supply pressurized water to the bottom of the drive piston. This quick-opening globe valve is operated by an internal spring and system pressure. It is closed by air pressure applied to the top of its diaphragm operator. A position indicator switch on this valve energizes a light in the control room as soon as the
valve starts to open.
The scram exhaust valve opens slightly before the scram inlet valve, exhausting water from above the drive piston. The exhaust valve opens faster than the inlet valve because of a large spring in the valve operator. Otherwise the valves are similar.
The scram accumulator stores sufficient energy to fully insert a control rod independent of any other source of energy. The accumulator consists of a water volume pressurized by nitrogen. A piston separates the water on top from the nitrogen below. A check valve in the accumulator charging line prevents a loss of water pressure in the
event supply pressure is lost.
During normal plant operation, the accumulator piston is seated at the bottom of
its cylinder. The loss of nitrogen decreases the nitrogen pressure, which actuates a pressure switch and sounds an alarm in the control room. UFSAR/DAEC - 1 3.9-23 Revision 22 - 5/13 To ensure that the accumulator is always able to produce a scram, it is continuously monitored for water leakage. A float-type level switch actuates an alarm if water leaks past the piston barrier and collects in the accumulator instrumentation block.
3.9.4.1.3 Control Rod Drive System Operation
The CRD system performs rod insertion, rod withdrawal, and scram. These
operational functions are described below.
Rod insertion is initiated by a signal from the operator to the insert valve solenoids. This signal causes both insert valves to open. The insert drive valve applies reactor pressure plus approximately 100 psi to the bottom of the drive piston. The insert exhaust valve allows water from above the drive piston to discharge to the exhaust
header.
As is illustrated in Figure 3.9-2, the locking mechanism is a ratchet-type device and does not interfere with rod insertion. The speed at which the drive moves is determined by the pressure drop through the insert speed control valve, which is set for approximately 4 gpm for a shim speed (nonscram operation) of 3 in./sec. During normal insertion, the pressure on the downstream side of the speed control valve is 90 to 100 psi
above reactor vessel pressure. However, if the drive slows for any reason, the flow
through, and pressure drop across, the insert speed control valve will decrease; the full differential pressure (260 psi) will then be available to cause continued insertion. With
260-psi differential pressure acting on the drive piston, the piston exerts an upward force
of 1040 lb.
Rod withdrawal is, by design, more involved than insertion. The collet finger (latch) must be raised to reach the unlocked position (see Figure 3.9-2). The notches in
the index tube and the collet fingers are shaped so that the downward force on the index tube holds the collet fingers in place. The index tube must be lifted before the collet fingers can be released. This is done by opening the drive insert valves (in the manner described in the preceding paragraph) for approximately 1 sec. The withdraw valves are
then opened, applying driving pressure above the drive piston and opening the area below the piston to the exhaust header. Pressure is simultaneously applied to the collet piston. As the piston raises, the collet fingers are cammed outward, away from the index tube, by
the guide cap.
The pressure required to release the latch is set and maintained at a level high enough to overcome the force of the latch return spring plus the force of reactor pressure opposing movement of the collet piston. When this occurs, the index tube is unlatched and free to move in the withdraw direction. Water displaced by the drive piston flows out through the withdraw speed control valve, which is set to give the control rod a shim speed of 3 in./sec. The entire valving sequence is automatically controlled and is initiated by a single operation of the rod withdraw switch. UFSAR/DAEC - 1 3.9-24 Revision 22 - 5/13 During a scram, the scram pilot valve and scram valves are operated as previously described. With the scram valves open, accumulator pressure is admitted under the drive piston, and the area over the drive piston is vented to the scram discharge volume. The large differential pressure (initially approximately 1500 psi and always
several hundred psi, depending on reactor vessel pressure) produces a large upward force
on the index tube and control rod. This force gives the rod a high initial acceleration and provides a large margin of' force to overcome any possible friction. After the initial
acceleration is achieved, the drive continues at a nearly constant velocity. This
characteristic provides a high initial rod insertion rate. As the drive piston nears the top
of its stroke, the piston seals close off the la rge passage (buffer orifices) in the stop piston tube, and the drive slows. Each drive requires approximately 2.5 gal of water during the scram stroke. The water capacity in each drive accumulator is adequate to complete a scram in the required time at low reactor vessel pressure. At higher reactor vessel pressures, the accumulator is
assisted on the upper end of the stroke by reactor vessel pressure acting on the drive via the ball check (shuttle) valve. As water is forced from the accumulator, the accumulator
discharge pressure falls below reactor vessel pressure. This causes a check valve, located in the drive flange, to shift its position to admit reactor pressure under the drive piston. Thus, reactor vessel pressure furnishes the force needed to complete the scram stroke at higher reactor vessel pressures. When the reactor vessel reaches full operating pressure, the accumulator is actually not needed to meet scram time requirements. With the reactor at 1000 psig and the scram discharge volume at atmospheric pressure, the scram force without an accumulator exceeds 1000 lb. Allowable scram insertion times are given in the Technical Specifications. 3.9.4.1.4 Control Rod Drive Housing Supports
The CRD housing supports (Figure 3.9-9) protect against additional damage to the nuclear system process barrier or damage to the fuel barrier by preventing any significant nuclear transient in the event a drive housing breaks or separates from the bottom of the reactor vessel.
Safety Design Bases
- 1. Control rod downward motion is limited, following a postulated CRD housing failure, so that any resulting nuclear transient could not be sufficient to cause fuel damage.
- 2. Clearance is provided between the housings and the supports to prevent vertical contact stresses due to their respective thermal expansion during
plant operation.
Description UFSAR/DAEC - 1 3.9-25 Revision 22 - 5/13 The control rod housing supports are illustrated in Figure 3.9-9. Hanger rods, about 10 ft long by 1.75 in. in diameter, are supported from the beams on stacks of disk springs that compress about 2 in. under the design load.
The support bars are bolted between the bottom ends of the hanger rods. The
spring pivots at the top and the beveled loose fitting ends on the support bars prevent substantial bending movement in the hanger rods.
Individual grids rest on the support bars between adjacent beams. Because a single-piece grid would be difficult to handle in the limited work space and because it is necessary that control rod drives, position indicators, and incore instrumentation components be accessible for inspection and maintenance, each grid is designed to be assembled or disassembled in place. Each grid assembly is made from two grid plates, a clamp, and a bolt. The top part of the clamp acts as a guide to ensure that each grid is
correctly positioned directly below the respective CRD housing that it would support in
the postulated accident.
When the support bars and grids are installed, a gap of about 1 in. at room temperature (approximately 70ºF) is provided between the grid and the bottom contact surface of the CRD flange. During system heatup, this gap is reduced by a net downward
expansion of the housings with respect to the supports. In the hot operating condition, the gap is approximately 0.25 in.
In the postulated CRD housing failure, the CRD housing supports are loaded
when the lower contact surface of the CRD flange contacts the grid. The resulting load is
then carried by two grid plates, two support bars, four hanger rods, their disk springs, and two adjacent beams.
To provide a housing support structure that absorbs as much energy as practical without yielding, the allowable tension and bending stresses were taken as 90% of yield, and the shear stress as 60% of yield. These are 1.5 times the corresponding AISC
allowable stresses of 60% and 40% of yield. This stress criterion is considered desirable for this application and adequate for the "once in a lifetime" loading condition.
For mechanical design purposes, the postulated failure resulting in the highest forces is an instantaneous circumferential separation of the CRD housing from the reactor
vessel, with an internal pressure of 1250 psig (reactor vessel design pressure) acting on
the area of the separated housing. The weight of the separated housing, control rod drive, and blade, plus the pressure force, is approximately 35,000 lb. This force is multiplied by a factor of three for impact, conservatively assuming the housing travels through a 1-UFSAR/DAEC - 1 3.9-26 Revision 22 - 5/13 in. gap before contacting the supports. The total force (10 5 lb) is then treated as a static load in design formulas.
Safety Evaluation
The downward travel of CRD housing and its control rod following the postulated housing failure is the sum of the compression of the disk springs under dynamic loading and the initial gap between the grid and the bottom contact surface of the CRD flange. If the reactor were cold and pressurized, the downward motion of the control rod would be limited to the approximate 2-in. spring compression plus approximately a 1-in. gap. If the reactor were hot and pressurized, the gap would be approximately 0.25 in. and the spring compression slightly less than in the cold condition. In either case, the control rod movement following a housing failure is limited substantially below one drive "notch" movement (6 in.). The nuclear transient from sudden withdrawal of any control rod through a distance of one drive notch at any position in the core does not result in a transient sufficient to cause damage to any radioactive material barrier.
The CRD housing supports are in place any time the reactor is to be operated. The housing supports may be removed when the reactor is in the shutdown condition
even when the reactor is pressurized, because all control rods are then inserted. Even if a control rod is ejected under the shutdown condition, the reactor remains subcritical, because it is designed to remain subcritical with any one control rod fully withdrawn at any time.
At plant operating temperature, a gap of approximately 0.25 in. is maintained between the CRD housing and the supports; at lower temperatures the gap is greater. Because the supports do not come in contact with any of the CRD housing, except during
the postulated accident condition, vertical contact stresses are prevented.
Inspection and Testing
When the reactor is in the shutdown mode, the CRD housing supports may be removed to permit inspection and maintenance of the control rod drives. When the support structure is reinstalled, it is inspected for proper assembly, particular attention
being given to ensure that the correct gap between the CRD flange lower contact surface and the grid is maintained.
3.9.4.2 Applicable Control Rod Drive System Design Specifications
As discussed in Section 3.9.5.1.3, the guide tubes are designed as lateral guides
for the control rods and as vertical support for a fuel support casting and four fuel assemblies. The 89 guide tubes are made of Type 304 stainless steel. The guide tubes are straight cylindrical tubes whose nominal dimensions are as follows: Inside diameter, 10.420 in. Wall thickness, 0.165 in. UFSAR/DAEC - 1 3.9-27 Revision 22 - 5/13 Length, 159 in.
Significant limits are as follows:
Minimum wall thickness, 0.144 in. Circular within, 0.030 in.
The guide tube can be subjected to any or all of the following loads:
- 1. Inward load due to pressure differential.
- 2. Lateral loads due to flow across the guide tube.
- 3. Deadweight.
- 4. Seismic.
3.9.4.3 Design Loads, Stress Limits, and Allowable Deformations
3.9.4.3.1 Pressure Differential Loading
The pressure differential loading on the guide tube is evaluated under normal, emergency, upset and faulted conditions, such as recirculation and steamline break accidents. See Section 3.9.5.2 for a more detailed discussion of this evaluation.
3.9.4.3.2 Lateral Loading
During normal operation, the 16 jet pumps spaced around the circumference of the shroud support ring will be discharging the recirculation flow into the lower plenum.
This flow will pass through the forest of guide tubes and enter the core. The lateral loads on the guide tubes will be very small. Even if it is assumed that all the recirculation flow
passes through the outer ring of the guide tubes over only a 6 ft vertical height (the guide tubes are 159 in. long), and if it is further assumed that all the velocity head of the water
flowing between the tubes is lost, the lateral load would be less than 1 psi. This is negligible in terms of guide tube bending stress (approximately 230 psi).
This analytical conclusion is supported by observations in the field. With the
reactor head off but with rated core volum etric flow, the flow distribution in the core was measured by placing a flow measuring device on top of selected fuel assemblies
- 2. No significant maldistribution of flow was observed. From these observations, it can be concluded that no signi ficant radical pressure distribution exists in the lower plenum since any such distribution would have resulted in higher flow in the fuel assemblies at the core periphery than those in the central regions of the core. Since there was full-rated volumetric flow during th e tests, it can be concluded that during reactor power operation there will not be a significant pressure drop across the guide
tubes. UFSAR/DAEC - 1 3.9-28 Revision 22 - 5/13 The lateral loading that could occur during LOCAs has been examined for the entire spectrum of credible accidents, and it is concluded that there is no accident that can
generate significant lateral guide tube loads.
- 1. Steam-Line Breaks
Figure 3.9-14 shows the flow patterns in the lower plenum during the course of a steam-line break. Early in the transient, the recirculation system will continue to provide flow into the lower plenum but at a rate less than normal flow. The lateral load on the guide tubes will thus be less than that occurring during normal operation.
When the vessel pressure has dropped below the saturation value of the fluid in the lower plenum, steam voids will start to form and fluid will leave the region via the core and the jet pump diffusers. (Cavitation effects will have caused the recirculation system drive pumps to become inoperative, thus allowing reverse flow in the jet pump diffusers.) The flow in the lower plenum is essentially axial with respect to the guide tubes, and no significant lateral loads will be generated.
- 2. Recirculation-Line Break
The recirculation line of a jet-pump-equipped BWR does not connect directly to the lower plenum of the reactor, and it is because of this geometry that the rupture of a recirculation line does not produce
significant lateral loads on the guide tubes. Figure 3.9-15 shows the flow patterns in the-lower plenum during a recirculation-line break. Immediately following the accident, the flow in the jet pump diffusers associated with the broken loop will reverse. The flow from the jet pumps of the unbroken loop will increase to 120% of rated flow; this flow will go partly to providing the reverse flow in the 8 jet pumps associated with the broken loop and partly to core flow. The maximum lateral loadings on the guide tubes will occur as a result of the 120% recirculation flow from the one pump in the unbroken side. As discussed above, no significant lateral loads exist during normal operation when 100% of rated flow
- is being generated by the recirculation system; thus, it can be concluded that, during the early phases of a recirculation-line break when 120% of rated flow from one pump will be crossing some guide tubes, there will be no significant lateral loads.
When all the liquid outside the core shroud has been discharged through the break, the blowdown flow will change to steam. There will be a rapid increase in the vessel depressurization rate at this time, which will cause steam voids to be generated throughout the lower plenum inventory. This void creation will force flow out of the lower plenum via both the core and UFSAR/DAEC - 1 3.9-29 Revision 22 - 5/13 the jet pump diffusers. This situation is very similar to the conditions in the lower plenum during the latter stages of the steam-line break. As discussed above, the flow in the lower plenum is essentially axial with respect to the guide tubes, and no significant lateral loads will be
generated. Thus, the recirculation-line break will not result in large lateral
guide tube loads.
- The impact of Increased Core Flow (105% of rated) is judged to be minimal also.
3.9.4.3.3 Deadweight Loading
The column load on the guide tube changes only slightly with time. During normal operating conditions, the column load is the deadweight of the fuel and fuel support casting minus the force due to the core pressure drop acting on the fuel bundles. The column load for a main steam line break decreases further due to the increase in
pressure differential across the fuel bundles. The pressure differential for a recirculation-line break remains essentially the same as normal operating conditions. To simplify the analysis, it was assumed that the column load is 2762 lb and is never decreased by any pressure differential operating on this area. The column load versus time curve is then a
straight line at 2762 lb.
The 2762 lb is derived as follows:
Fuel bundles = 675 lb (4) = 2700 lb Fuel casting = 62 lb (1) = 62 lb Total 2762 lb
3.9.4.3.4 Control Rod Displacement
As mentioned in Chapter 15, the reactor is shut down by void formation in the
core in the event of a recirculation-line break. However, the control rod insertion versus time will meet the Technical Specifications limits.
For the recirculation-line break, the drywell pressure would reach the 2-psig scram setpoint in less than 0.1 sec; thus, control rod movement would start at this time.
3.9.4.3.5 Summation of Maximum Applied Loads
There are two primary modes of failure to be considered in the guide tube analysis: (1) excessive stress and (2) excessive elastic deformation. First, it will be shown that the allowable stress limit will not be exceeded and, second, that the elastic deformations of the guide tube never are great enough to cause the free movement of the
control rod to be jeopardized.
The first mode of failure is evaluated for the faulted condition. UFSAR/DAEC - 1 3.9-30 Revision 22 - 5/13 The evaluation of the second mode of failure is based on clearance reduction between the guide tube and the control rod. The minimum allowable clearance is 0.099 inches. This assumes maximum ovality and minimum diameter of the guide tube, and the maximum control rod dimension. Referring to Figure 3.9-18, it can be seen that, the
clearance between the control rod and the guide tube would only be reduced by approximately 0.016 inches for a 40 psi pressure differential (which is greater than the pressure differential under faulted conditions). Since the calculated maximum displacement does not exceed the minimum allowable clearance, failure due to excessive elastic deformation will not occur.
Two types of instability also were considered in the analysis of guide tube design. The first was the classic instability associated with vertically loaded columns (i.e. buckling). The second was the diametral collapse of a circular tube under external to internal differential pressure.
These evaluations concluded that there is significant margin between the
calculated loads due to either buckling or differential pressure and the allowable stresses. Thus, the guide tube is not an unstable column.
To demonstrate the adequacy of the guide tube design, a series of sensitivity studies have been conducted. Parameters evaluated are guide tube ovality and wall
thickness and their effect on yield pressure and radial deflection. The results of the
sensitivity study are shown in Figures 3.9-16, 3.9-17, and 3.9-18.
These figures demonstrate the sensitivity of a BWR guide tube to changes in the manufacturing tolerances. Figure 3.9-17 shows the sensitivity of the yield pressure calculation to initial ovality of the guide tube assuming minimum wall thickness. Figure
3.9-16 shows the sensitivity of the yield pressure calculation to guide tube wall thickness assuming the maximum initial ovality. Figure 3.9-18 shows the radial deflection for a range of differential pressures for two given initial ovalities and minimum wall thickness.
3.9.4.4 Control Rod Drive Performance Assurance Program
3.9.4.4.1 Development and Design Conformation Tests
The development drive (one prototype) testing to date included more than 5000 scrams and approximately 100,000 latching cycles. One prototype was exposed to simulated operating conditions for 5000 hr. These tests demonstrated the following:
- 1. The drive easily withstands the forces, pressures, and temperatures imposed.
- 2. Wear, abrasion, and corrosion of the nitrided Type 304 stainless parts are negligible. Mechanical performance of the nitrided surface is superior to that of materials used in earlier operating reactors.
- 3. The basic scram speed of the drive has a satisfactory margin above UFSAR/DAEC - 1 3.9-31 Revision 22 - 5/13 minimum plant requirements at any reactor vessel pressure.
- 4. Usable seal lifetimes in excess of 1000 scram cycles can be expected. Operating experience on BWR plants to date confirms the above development test findings.
3.9.4.4.2 Factory Quality Control Tests
Quality control of welding, heat treatment, dimensional tolerances, material verification, and similar factors is maintained throughout the manufacturing process to ensure reliable performance of the mechanical reactivity control components. Some of the quality control tests performed on the control rods, CRD mechanisms, and hydraulic
control units are listed below:
- 1. Control rod absorber tube tests:
- a. Material integrity of the tubing and end plug is verified by ultrasonic inspection.
- b. The boron-10 fraction of the boron content of each lot of boron carbide is verified.
- c. Weld integrity of the finished absorber tubes is verified by helium leak testing.
- 2. CRD mechanism tests:
- a. Pressure welds on the drives are hydrostatically tested in accordance with the ASME Code, Section III, Class A vessels.
- b. Electric components are checked for electrical continuity and resistance to ground.
- c. Drive parts that cannot be visually inspected for dirt are flushed with filtered water at high velocity. No significant foreign material is permitted in effluent water.
- d. Seals are tested for leakage to demonstrate correct seal operation.
- e. Each drive is tested for shim motion, latching, and control rod position indication.
- f. Each drive is subjected to cold scram tests at various reactor pressures to verify correct scram performance.
- 3. Hydraulic control unit tests:
UFSAR/DAEC - 1 3.9-32 Revision 22 - 5/13 a. Hydraulic systems are hydrostatically tested in accordance with Section 5.2.
- b. Electric components and systems are tested for electrical continuity and resistance to ground.
- c. Correct operation of the accumulator pressure and level switches is verified.
- d. The unit's ability to perform its part of a scram is demonstrated.
- e. Correct operation and adjustment of the insert and withdrawal valves are demonstrated.
3.9.4.4.3 Operational Tests
After installation, all rods, hydraulic control units, and drive mechanisms are
tested through their full range for operability.
Each time a control rod is withdrawn a notch during normal operation, the operator can observe incore monitor indications to verify that the control rod is following the drive mechanism. All control rods that are partially withdrawn from the core can be
tested for rod-following by inserting or withdr awing the rod one notch and returning it to its original position, while the operator observes incore monitor indications.
To make a positive test of the control-rod-to-control-rod-drive coupling integrity, the operator can withdraw a control rod to the end of its travel and then attempt to
withdraw the drive to the overtravel position. The failure of the drive to reach the overtravel position demonstrates rod-to-drive coupling integrity.
Hydraulic supply subsystem pressures can be observed from instrumentation in the control room. Scram accumulator pressures can be observed on the nitrogen pressure gauges. Surveillance requirements for the control rod drive system are included in the
Technical Specifications.
3.9.5 REACTOR PRESSURE VESSEL INTERNALS
3.9.5.1 Design Arrangements
The reactor vessel internals include the following components:
Startup neutron sources Core shroud UFSAR/DAEC - 1 3.9-33 Revision 22 - 5/13 Shroud head and steam separator assembly Core support (core plate) Top guide Fuel support pieces Control rod guide tubes Jet pump assemblies Steam dryers Feedwater spargers Differential pressure and liquid control line Incore flux monitor guide tubes Surveillance sample holders
The overall arrangement of the internals within the reactor vessel is shown in
Figures 3.9-19 and 3.9-20.
Although it was not mandatory, the manufacturer of the reactor vessel internals
used weld procedures and welders qualified in accordance with the intent of ASME Code, Section IX. This means that welding techniques, procedures, methods, qualifications, and tests were used that ensured that the design integrity and design requirements were maintained on the equipment items to which each level of welding control was applied. All facets of reactor internals that form a primary pressure boundary
were welded per ASME Code, Sections III and IX. Much of the non-pressure boundary reactor internals were welded under ASME Code, Section IX, requirements also. The jet pump instrumentation lines were welded per Mil-Specification MIL-T-5021C, which meets or exceeds ASME Code, Section IX, requirements. In other cases, proprietary welding documents containing requirements excerpted from ASME Code, Section IX; AWS Standards; and GE processes and procedures were used. Examples of ASME Code, Section IX, variables that are excepted by some GE welding documentation are for
procedures described in ASME Code, S ection IX, paragraphs V-2b-1, V-2b-2, V-2b-3, V-2d-1, V-2d-2, V-2e-1, V-3, V-7a, and V-7b, where some of these are not applicable due to the process that GE used, or they were not required to ensure the design intent of the components to which these weld documents were applied. Also excepted from ASME Code, Section IX, were some welder qualification paragraphs such as those that called for weld specimens to be radiographed or cross sectioned, whereas GE required cross sectioning only. The quality control aspects of all welding specifications used on
reactor internals are available for quality assurance audit as required.
3.9.5.1.1 Core Structure
The core structure surrounds the active core of the reactor and consists of the core shroud, shroud head and steam separator assembly, core support, and top guide. This structure is used to form partitions within the reactor vessel, to sustain pressure differentials across the partitions, to direct the flow of the coolant water, and to locate laterally and support the fuel assemblies, control rod guide tubes, and steam separators.
Figure 3.9-20 shows the reactor vessel internal flow paths.
Core Shroud UFSAR/DAEC - 1 3.9-34 Revision 22 - 5/13 The core shroud is a stainless steel cylindrical assembly that provides a partition to separate the upward flow of coolant through the core from the downward recirculation flow. This partition separates the core region from the downcomer annulus, thus
providing a floodable region following a recirculation-line break. The volume enclosed by the core shroud is characterized by three regions, each with a different shroud diameter. The upper shroud has the largest diameter and surrounds the core discharge plenum, which is bounded by the shroud head on top and the top guide below. The central portion of the shroud surrounds the active fuel and forms the longest section of the shroud. This section has an intermediate diameter and is bounded at the bottom by the core support. The lower shroud, surrounding part of the lower plenum, has the smallest diameter and, at the bottom, is welded to the reactor vessel shroud support ring (see Section 5.3).
Shroud Head and Steam Separator Assembly
The shroud head and steam separator assembly is bolted to the top of the upper shroud to form the top of the core discharge plenum. This plenum provides a mixing chamber for the steam-water mixture before it enters the steam separators. Individual stainless steel axial flow steam separators, shown in Figure 3.9-21, are attached to the top of standpipes that are welded into the shroud head. The steam separators have no moving parts. In each separator, the steam-water mixture rising through the standpipe passes over vanes that impart spin to establish a vortex separating the water from the steam. The separated water flows from the lower portion of the steam separator into the recirculation flow downcomer annulus.
Core Support
The core support (core plate) consists of a circular stainless steel plate stiffened with a rim and beam structure. Perforations in the plate provide lateral support and guidance for the control rod guide tubes, incore flux monitor guide tubes, peripheral fuel support pieces, and startup neutron sources. The last two items are also supported
vertically.
The entire assembly is bolted to a support ledge between the central and lower portions of the core shroud. Alignment pins that bear against the shroud are used to correctly position the assembly before it is secured.
Top Guide
The top guide is formed by a series of stainless steel beams joined at right angles to form square openings. Each opening provides lateral support and guidance for four fuel assemblies. Holes are provided in the bottom of the beams to anchor the incore flux monitor guide tubes and startup neutron sources. The top guide is positioned with alignment pins that bear against the shroud.
3.9.5.1.2 Fuel Support Pieces UFSAR/DAEC - 1 3.9-35 Revision 22 - 5/13 The fuel support pieces, shown in Figure 3.9-22, are of two basic types, namely
peripheral and four-lobed. The peripheral fuel support pieces, which are welded to the core support assembly, are located at the outer edge of the active core and are not
adjacent to control rods. Each peripheral fuel support piece will support one fuel assembly and contains an orifice assembly designed to ensure proper coolant flow to the fuel assembly. Each four-lobed fuel support piece will support four fuel assemblies and
is provided with orifice plates to ensure proper coolant flow distribution to each fuel assembly. The four-lobed fuel support pieces rest in the top of-the control rod guide
tubes and are supported laterally by the core support. The control rods pass through slots
in the center of the four-lobed fuel support pieces. A control rod and the four adjacent fuel assemblies represent a core cell (see Section 4.2).
3.9.5.1.3 Control Rod Guide Tubes
The control rod guide tubes, located inside the vessel, extend from the top of the
CRD housings up through holes in the core support. Each tube is designed as the lateral
guide for a control rod and as the vertical support for a four-lobed fuel support piece and the four fuel assemblies surrounding the control rod. The bottom of the guide tube is supported by the CRD housing (see Section 5.3) which in turn transmits the weight of the guide tube, fuel support piece, and fuel assemblies to the reactor vessel bottom head. A thermal sleeve is inserted into the CRD housing from below and is rotated to lock the control rod guide tube in place. A key is inserted into a locking slot in the bottom of the CRD housing to hold the thermal sleeve in position.
3.9.5.1.4 Jet Pump Assemblies
The jet pump assemblies are located in two semicircular groups in the downcomer
annulus between the core shroud and the reactor vessel wall. Each stainless steel jet pump consists of a driving nozzle, suction inlet, throat or mixing section, and diffuser (see Figure 3.9-23). The driving nozzle, suction inlet, and throat are joined together as a removable unit, and the diffuser is permanently installed. High-pressure water from the recirculation pumps (see Section 5.4) is supplied to each pair of jet pumps through a riser pipe welded to the recirculation inlet nozzle thermal sleeve. A horseshoe-shaped riser
brace, is welded to the jet pump riser at the inside "U" portion of the brace, and the cantilever portions of the brace are welded to pads extending from the reactor vessel wall.
The nozzle entry section is connected to the riser by a metal to-metal, spherical-to-conical seal joint. Firm contact is maintained by a holddown clamp. The throat
section is supported laterally by a bracket attached to the riser. There is a slip-fit joint
between the throat and diffuser. The diffuser is a gradual conical section changing to a
straight cylindrical section at the lower end.
To monitor its flow, each jet pump has a sensing line which is welded to the diffuser at two points. To ensure against fatigue failure of the sensing lines, beam-and-clamp assemblies have been installed on certain jet pumps. This modification reinforces UFSAR/DAEC - 1 3.9-36 Revision 22 - 5/13 the welds and changes the natural frequency of the sensing lines to avoid resonance at excitation frequencies which were measured during plant startup. The beam-and-clamp assemblies are described in reference 5.
3.9.5.1.5 Steam Dryers
The steam dryers remove moisture from the wet steam leaving the steam separators. The extracted moisture drips down the dryer vanes to the collecting troughs, then flows through tubes into the downcomer annulus (see-Figure 3.9-24). A skirt extends from the top of the steam dryers to the steam separator standpipe, below the water level. This skirt forms a seal between the wet steam plenum and the dry steam flowing from the top of the dryers to the steam outlet nozzles.
The steam dryer and shroud head are positioned in the vessel with the aid of vertical guide rods. The dryer assembly rests on steam dryer support brackets attached to the reactor vessel wall. Upward movement of the dryer assembly is restricted by steam
dryer holddown brackets attached to the reactor vessel top head.
3.9.5.1.6 Feedwater Spargers
The feedwater spargers are perforated stainless steel headers located in the mixing plenum above the downcomer annulus. A separate sparger is fitted to each feedwater nozzle and is shaped to conform to the curve of the vessel wall. Sparger end brackets are
attached to vessel brackets to support the weight of the spargers and position the spargers away from the vessel wall. Feedwater flow enters the center of the spargers and is discharged radially inward and downward to mix the cooler feedwater with the downcomer flow from the steam separators and dryers before it contacts the vessel wall. The feedwater also serves to collapse the steam voids in the mixing plenum and to subcool the water flowing to the jet pumps and recirculation pumps.
3.9.5.1.7 Core Spray Lines
Two 100%-capacity core spray lines enter the reactor vessel through the two core spray nozzles (see Section 5.3). The lines divide immediately inside the reactor vessel.
The two halves are routed to opposite sides of the reactor vessel and are supported by clamps attached to the vessel wall. The lines are then routed downward into the downcomer annulus and pass through the upper shroud immediately below the flange. The flow divides again as it enters the center of the semicircular header, which is routed
halfway around the inside of the upper shroud. The ends of the two headers are supported by brackets designed to accommodate thermal expansion. The line routing and supports are designed to accommodate differential movement between the shroud and vessel. The
other core spray line is identical except that it enters the opposite side of the vessel and
the headers are at a slightly different elevation in the shroud. The correct spray distribution pattern is provided by a combinati on of distribution nozzles pointed radially inward and downward from the headers.
3.9.5.1.8 Differential Pressure and Standby Liquid Control Line UFSAR/DAEC - 1 3.9-37 Revision 22 - 5/13 The differential pressure and standby liquid control line serves a dual function within the reactor vessel--to inject liquid control solution into the coolant stream (see Section 9.3.4) and to sense the differential pressure across the core support assembly (described in Section 5.3). This line enters the reactor vessel at a point below the core shroud as two concentric pipes. In the lower plenum, the two pipes separate. The inner pipe terminates near the lower shroud with a perforated length below the core support assembly. It is used to sense the pressure below the core support during normal operation and to inject liquid control solution when required. This location facilitates good mixing and dispersion. The inner pipe also reduces thermal shock to the vessel nozzle should the standby liquid control system be actuated. The outer pipe terminates immediately above the core support and senses the pressure in the region outside the fuel assembly channels.
3.9.5.1.9 Incore Flux Monitor Guide Tubes
The incore flux monitor guide tubes extend from the top of the incore flux monitor housings (see Section 5.3) in the lower plenum to the top guide. The power range detectors for the power range monitoring units and the dry tubes for the source range monitoring and intermediate range monitoring (SRM/IRM) detectors are inserted through
the guide tubes. The guide tubes are held in place below the top guide by spring tension. A latticework of clamps, tie bars, and spacers gives lateral support and rigidity to the guide tubes. The bolts and clamps are welded, after assembly, to prevent loosening
during reactor operation.
3.9.5.1.10 Initial Startup Neutron Sources
Each initial startup source consists of two irradiated antimony rods within a single beryllium cylinder. The antimony-beryllium cylinder assemblies are further encased in
stainless steel tubes. These tubes have fitted nosepieces on one end and axial spring-loaded detent pins on the other end. The nosepieces and detent pins mate, respectively, with notches in the top of the core support plate and the bottom of the top guide to
position the startup sources securely in the vertical position. The design provides a sufficient source of neutrons in the core to ensure that the core neutron flux monitors are
operating and that significant changes in core reactivity can be readily detected by the installed neutron flux instrumentation (see Section 7.6.1).
3.9.5.1.11 Surveillance Sample Holders
The surveillance sample holders are welded baskets containing impact and tensile specimen capsules (see Section 5.3). The baskets hang from brackets that are attached to the inside wall of the reactor vessel and extend to mid-height of the active core. The radial positions are chosen to expose the specimens to the same environment and maximum neutron fluxes experienced by the reactor vessel itself while avoiding jet pump removal interference or damage.
3.9.5.2 Loading Conditions
UFSAR/DAEC - 1 3.9-38 Revision 22 - 5/13 3.9.5.2.1 Evaluation Methods
To determine that the safety design bases are satisfied, responses of the reactor vessel internals to loads imposed during normal, upset, emergency, and faulted conditions were examined. The effects on the ability to insert control rods, cool the core, and flood the inner volume of the reactor vessel were determined.
The ASME Code, Section III, for Class A vessels was used as a guide to determine limiting stress intensities and cyclic loadings for the reactor vessel internals. When buckling was not a possible failure mode and stresses were within those stated in the ASME Code, either the elastic stability of the structure or the resulting deformation was examined to determine whether the safety design bases were satisfied. Events To Be Evaluated
The examination of the spectrum of conditions for which the safety design bases must be
satisfied reveals three significant events:
- 1. LOCA: A break in a recirculation line. The accident results in pressure differentials, across some of the reactor vessel internals, that exceed normal loads.
- 2. Steam-line break accident: A break in one main steam line between the reactor vessel and the flow restrictor. The accident results in significant pressure differentials across some of the reactor vessel internals.
- 3. Earthquake: This condition subjects the reactor vessel internals to significant forces as a result of ground motion.
The analysis of other conditions existing during normal operation, abnormal
operational transients, and accidents shows that the loads affecting the reactor vessel
internals are less severe than the design-basis postulated events.
3.9.5.2.2 Recirculation-Line and Steam-Line Break
Accident Definition
The recirculation-line break is the same as the design-basis LOCA described in Section 6.3 and Chapter 15. A sudden, complete circumferential break is assumed to
occur in one recirculation loop. UFSAR/DAEC - 1 3.9-39 Revision 22 - 5/13 The analysis of the steam-line break assumes a sudden, complete circumferential break of one main steam line between the reactor vessel and the main steam line restrictor. A steam-line break upstream of the flow restrictors produces a larger blowdown area than a break downstream of the restrictors. The larger blowdown area results in greater pressure differentials across the reactor assembly internal structures.
Both the recirculation-line break and steam-line break have been examined for a spectrum of initial reactor operating conditions. These studies show that a recirculation-line break at any initial reactor condition would be very mild with regard to resultant
pressure differentials.
The steam-line break accident produces significantly higher pressure differentials across the reactor assembly internal structures than does the recirculation-line break. This results from the higher reactor depressurization rate associated with the steam-line break. The depressurization rate is less for mixed flow than for steam flow. Therefore, the steamline break is the design-basis accident for internal pressure differentials.
For a steam-line break accident, a low initial reactor power level results in a more severe pressure transient across some components than would be the case at maximum power. This is because the difference between energy removal through the break and
energy addition to the reactor vessel inventory increases as the reactor power decreases. Thus, the depressurization rate following a steam-line break would increase with
decreasing initial reactor power level. Consequently, both a high power and a low power case are examined.
The maximum differential pressures across the reactor assembly internals resulting from the postulated accidents are discussed in Section 15.3.5. Figure 15.3-1
shows the differential pressures for various internals.
Response of Reactor Internals to Pressure Differences
The maximum differential pressures are used, in combination with other structural loads, to determine the total loading on the various reactor vessel internals. The internals are then evaluated to assess the extent of deformation and collapse, if any. Of particular interest are (1) the responses of the guide tubes and the metal channels around the fuel bundles and (2) the potential leakage around the jet pump joints.
The guide tube was evaluated for collapse caused by externally applied pressure, as discussed in Section 3.9.4.3.5.
Channel Response with Respect to Structural Integrity
The channel wall P at which a DAEC channel assembly will fail due to yielding is 16.0 psi. The calculated maximum P for a main steam line break is 14.8 psi (Section 15.3.5.4). Since this is below the yield limit, the channel would maintain its integrity due
to the increased P experienced during a main steam line break. UFSAR/DAEC - 1 3.9-40 Revision 22 - 5/13 Channel Response with Respect to Channel/Control Blade Interaction The clearance to the control rod is affected by the channel wall P due to the added elastic channel wall bulge. This effectively reduces the gap between channels which may lead to interaction with the control blade roller. The nominal gap between fuel channels is 700 mils. If the components of the gap are assumed to be a normal distribution with a standard deviation equal to 1/3 the tolerance range, the minimum gap is 632 mils. The channel bulge due to the maximum P is 30 mils. Subtracting this from the gap standard deviation produces a minimum gap between channels at the roller location of 602 mils. Assuming the control blade is centered in the channel gap and the control rod roller diameter is at its maximum of 525 mils, then a net
one-sided clearance to the control rod roller of 39 mils exists. Therefore, if the main steam line break were to occur near the beginning of life, no channel/control blade
interaction would be expected. Channel Response with Respect to Channel/Control Blade Interactions Towards the End of Life The channel wall will experience irradiation induced creep causing more bulge throughout its life as a function of exposure. For conservatism, an exposure of 50 GWD/MTU is assumed to quantify the irradiation induced creep permanent deflection. Where there is interference, the channel and control rod will interact. If the control rod is assumed to be rigid, then it must displace the channel equal to the amount of the interference. This displacement will produce a normal force at the roller channel contact.
Since the roller is essentially pinched between the channel gap, it will slide as opposed to rolling and the normal force produced by the displacement will contribute to a friction force that the control rod drives must overcome to insert the control rod. The calculated
frictional force is 330 lbs. This calculated force is very conservative since the actual
friction loading could not exceed the weight of the control rod or a "no settle" condition would have occurred and the rod would have been fully inserted. The maximum normal hydraulic drag as specified by the control rod is 230 lbs. Adding the maximum weight of
a control rod (250 lbs) to the frictional force and the hydraulic drag results in a total resistive force of 810 lbs. With an available driving force of 1040 lbs, there remains a margin of 230 lbs to insert the control rod. It is concluded that the main steam line break accident can pose no significant interference to the movement of control rods.
Additional analysis indicates that no fuel pins will come in contact with the fuel element channels as the result of the DBE c oncurrent with rapid depressurization of the reactor core.
Jet Pump Joints: An analysis was originally performed to evaluate the potential leakage from within the floodable inner volume of the reactor vessel during the
recirculation-line break and subsequent LP CI reflooding. The two possible sources of leakage are the following: UFSAR/DAEC - 1 3.9-41 Revision 22 - 5/13 1. Jet pump throat to diffuser joint.
- 2. Jet pump nozzle to riser joint.
The jet pump to shroud support joint is welded and therefore is not a potential source of leakage. The slip joints for all jet pumps leak no more than a total of 225 gpm. The jet pump bolted joint, by analysis, is shown to leak no more than 542 gpm for the pumps through which the vessel is being flooded.
The summary of maximum leakage is as follows:
Total leakage through all throat to diffuser
joints 225 gpm Total leakage through all nozzle to riser
joints 542 gpm Total maximum rate 767 gpm
The original sizing calculation procedure for LPCI capacity included a total leakage of 3000-gpm from the core shroud. The ECCS performance analysis in Section
15.2 accounts for all known leakage paths in the core shroud and ECCS flow paths. It is
concluded that the reactor vessel internals retain sufficient integrity during the
recirculation-line break accident to allow reflooding of the inner volume of the reactor vessel.
3.9.5.2.3 Seismic Analysis of the RPV and Internals
The seismic loads on the reactor pressure vessel (RPV) and internals are based on a dynamic analysis of the RPV and internals shown in Figure 3.9-28. The dynamic model of the RPV and internals is briefly described below.
The presence of a fluid and structural components (e.g., fuel within the RPV) introduces a dynamic coupling effect. Dynamic effects of water enclosed by the RPV are accounted for by the introduction of a hydrodynamic mass matrix.
The seismic model of the RPV and internals analyzed has one horizontal
translation coordinate for each node point considered in the analysis. Due to the approximate coincidence of the mass center and elastic axis of the building, and the symmetry of the RPV and internals, one horizontal coordinate was excluded. The remaining translational coordinate for the vari ous node points was the vertical coordinate. This coordinate (vertical) was excluded because the frequency content of earthquakes is
such that the vertical frequencies of the RPV and internals are well above those of earthquakes. Dynamic loads due to vertical motion were added to or subtracted from the Subsequent evaluations determined that this leakage path has negligible imp act. The current analysis (Section 15.2) uses a bounding generic value of 600 gpm. UFSAR/DAEC - 1 3.9-42 Revision 22 - 5/13 static loads of the components, whichever was the more conservative. The two rotational coordinates about each node point were excluded because the moment contribution of rotary inertia is negligible. The remaining rotational coordinate is neglected since the building, and hence the RPV, has negligible torsional motion.
Seismic analysis was performed by coupling the lumped mass model of the RPV and internals with the building model to determine the system natural frequencies and mode shapes. The load response of the RPV and internals was then determined by the response spectrum method. The spectral accelerations, velocities, and displacements for the OBE and DBE are taken from figures in Section 2.5.4 for the modes of interest. The root mean square of these individual modal responses was then used for design
calculation.
The natural frequencies of the reactor internals, reactor vessel, and pedestal system in the vertical direction have been found to be 19 Hz or higher. The examination of the response spectra shows no significant amplification at this frequency. Hence, omitting the vertical motion from seismic analysis to reduce the analytical complexities is
acceptable. The effects of vertical excitations are accounted for by increasing or decreasing (whichever causes higher stress) the weight of the various components by a
percentage equal to the vertical acceleration expressed in percent "g".
The basis for the derivation of the LOCA excitation input design is contained in the response to Question A.9 of Amendment 6 to the Browns Ferry Nuclear Power Station Units 1 and 2 Design and Analysis Report. The peak values of dynamic pressure
differences calculated for the LOCA were then used in the design of the reactor
internals.
To ensure that no significant dynamic amplification of loading occurs as a result of the oscillatory nature of the blowdown forces, a comparison was made of the periods of the applied forces and the natural periods of the reactor internal components being acted upon by the applied forces. These periods were determined from a dynamic model
of the RPV and internals shown in Figure 3.9-28.
Besides the real masses of the RPV and internals, the hydrodynamic mass effects
of the water inside the RPV were also accounted for.
The natural frequencies of the first five modes for the DAEC RPV and internals
are tabulated below: Mode Frequency Hz 1 18.8 (shroud) 2 25.5 (fuel-guide tube) 3 33.3 (RPV head) 4 46.6 (RPV head) 5 51.2 (RPV head)
UFSAR/DAEC - 1 3.9-43 Revision 22 - 5/13 All other reactor internal components have natural frequencies higher than those shown above. The smallest period of the applied force (approximately 0.7 sec), , is more than 10 times the largest period of the component upon which the force acted (i.e., natural frequency of component is more than 10 times greater than the frequency of the applied load). It is evident that this conclusion would apply for the higher modes, since they would have shorter periods. It is a well-known fact that for a damped single degree of freedom system subjected to a sinusoidal forcing function, the amplification factor is
essentially equal to unity. Therefore, it was concluded that no significant load amplification occurred because of the "slowly" changing nature of the applied load and that a statically applied load equal in value to the peak transient load could be used for
design purposes.
Rather than measuring input forcing functions during normal operation, the vibration response of the RPV internals was actually measured.
3.9.5.2.4 Conclusions
Response analyses of the reactor vessel internals show that deformations are sufficiently limited to allow both adequate control rod insertion and proper operation of the core standby cooling systems. Sufficient integrity of the internals is retained during accident conditions to allow successful reflooding of the reactor vessel inner volume. The analyses considered various loading combinations, including loads imposed by
external forces.
3.9.5.2.5 Inspection and Testing
Quality control methods were used during the fabrication and assembly of reactor vessel internals to ensure that the design specifications were met.
The reactor coolant system, which includes the reactor vessel internals, was
thoroughly cleaned and flushed before fuel was loaded initially.
During the preoperational test program, operational readiness tests were performed on various systems. In the course of these tests, such reactor vessel internals as the feedwater spargers, the core spray lines, and the standby liquid control system line
were functionally tested.
Before the startup of the unit, steam separator-dryer performance tests were conducted on several plants using the DAEC separator-dryer design concept to determine
the carryunder and carryover characteristics. It was not planned, therefore, to test the DAEC steam separators and dryers to determine these characteristics since they would have been already demonstrated on other plants. Moisture carryover is determined from sodium-24 activity in samples from main steam lines and samples from primary containment during normal radiochemical quality surveillance tests.
Vibration analysis of reactor vessel internals was included in the design to eliminate failures caused by vibration. UFSAR/DAEC - 1 3.9-44 Revision 22 - 5/13 The nature of the tests and the components tested were dictated largely by the
results of vibrations testing conducted earlier on the other plants. The DAEC vibration test plan was completed during the first quarter of 1972.
The vibration analyses and tests were designed to determine any potential hydraulically induced equipment vibrations and to verify that the structures do not fail
because of fatigue. The structures were analyzed for natural frequencies, node shapes, and vibrational magnitudes that could lead to fatigue at these frequencies. With this analysis as a guide, the reactor internals were instrumented and tested to ascertain that
there are no gross instabilities. The cyclic loadings were evaluated using, as a guide, the
cyclic stress criteria of the ASME Code, Section III. Field test data were correlated with
the analyses to ensure the validity of the analytical techniques on a continuing basis.
For Extended Power Uprate (EPU), the original test data was reviewed to determine which internal components were likely to experience significant vibration at the new operating conditions. The actual measured frequencies and vibration amplitudes
were linearly extrapolated to the uprated cond itions. The effects of forced vibration and resonance on the extrapolation methodology were considered and found to be inconsequential. The extrapolated vibration amplitude response at EPU conditions was compared with the acceptance criteria to obtain the percent criteria for each mode. The percent criteria for all modes were absolute summed. This total percent criteria was
shown to be less than 100%.
In addition, vessel internal components that operating experience have shown to be susceptible to flow-induced vibration problems (such as steam dryers and jet pump sensing lines) were included in the EPU evaluation. While no new vulnerabilities were identified, it was recommended that continued inspections, per BWRVIP-06, for dryer drain channel cracking be performed. The critical reactor internals were tested up to 51 Mlbm/hr core flow and at 50%, 75%, and 100% load line during the original startup tests (Reference 3). This data was
used in the evaluation of Increased Core Flow (ICF) (105% of rated). The expected vibration levels for ICF were estimated by extrapolating the vibration data recorded
during startup. The extrapolation was from 51 Mlbm/hr to 51.5 Mlbm/hr (105% of rated core flow). The vibration was extrapolated by using the square relationship: vibration varies as the square of flow. The design c onfiguration and basis used for rated flow is also used for increased core flow.
The results of the vibration of evalua tion (Reference 7) show that continuous operation at 1,912 MWt and 105% of the rated core flow (51.5 Mlbm/hr) does not result in any detrimental effects on the reactor internal components. There is no concern for vane passing frequency (VPF) resonance up to the maximum design pump speed VPF of 142.5 Hz. The maximum design recirculation pump speed at DAEC is 1,710 rpm. This
corresponds to a VPF of 142.5 Hz. The calculations for the ICF operating condition indicate that the vibration of the components evaluated are within the acceptance criteria. UFSAR/DAEC - 1 3.9-45 Revision 22 - 5/13 The acceptance criterion of 10,000 psi peak stress intensity is less than the ASME Code criteria of 13,600 psi.
The reactor vessel and internals were designed to ensure adequate working space and access for the inspection of selected components and locations. Criteria for selecting the components and locations to be inspected were based on the probability of a defect
occurring or enlarging at a given location and include areas of known stress concentrations and locations where cyclic strain or thermal stress might occur.
3.9.5.3 Design Bases
3.9.5.3.1 Power Generation Objectives
Reactor vessel internals (exclusive of fuel, control rods, and incore nuclear instrumentation) are provided to achieve the following power generation objectives:
- 1. Maintain partitions between regions within the reactor vessel to provide correct coolant distribution, thereby allowing power operation without fuel damage.
- 2. Provide positioning and support for the fuel assemblies, control rods, incore flux monitors, and other vessel internals to ensure that control rod movement is not impaired.
3.9.5.3.2 Safety Design Bases
The reactor vessel internals meet the following safety design bases:
- 1. The internals are arranged to provide a floodable volume in which the core can be adequately cooled in the event of a breach in the nuclear system process barrier external to the reactor vessel.
- 2. The deformation of internals is limited to ensure that the control rods and emergency core cooling systems can perform their safety functions.
- 3. The mechanical design of applicable internals ensures that safety design bases 1 and 2 are satisfied so that the safe shutdown of the plant and the removal of decay heat are not impaired.
3.9.5.3.3 Power Generation Design Bases
The reactor vessel internals are designed to meet the following power generation
design bases:
- 1. They provide the proper coolant distribution during all anticipated normal operating conditions to allow power operation of the core without fuel damage.
UFSAR/DAEC - 1 3.9-46 Revision 22 - 5/13
- 2. They are arranged to facilitate refueling operations.
- 3. They are designed to facilitate inspection.
3.9.6 INSERVICE TESTING OF PUMPS AND VALVES
An inservice testing program for pumps and valves has been prepared. This program is revised as required for each 120-month inspection interval to incorporate the
latest applicable addenda to ASME Code, Section XI.
Inservice testing of pumps and valves complies with the requirements of Subsections IWP and IWV of ASME Code, Section XI, respectively.
3.9.6.1 Relief Requests
When compliance with ASME Code, Section XI, is impractical for specific items, relief is requested from the NRC in compliance with 10CFR50.55a(g)(5).
3.9.6.2 Inservice Testing Program
The inservice testing program for the DAEC fourth 10-year Inservice Testing interval commenced February 1, 2006. The current revised program has been prepared and implemented according to the 2001 Edition of Section XI of the ASME Code, through the 2003 Addenda.
In response to NRC Generic Letter 87-06, a list of all pressure isolation valves along with information on periodic tests was submitted in Reference 4. UFSAR/DAEC - 1 3.9-47 Revision 22 - 5/13 REFERENCES FOR SECTION 3.9
- 1. U.S. Nuclear Regulatory Commission, Failure of Control Rods to Insert During a Scram at a BWR, IE Bulletin 80-17 (series), 1980-1981.
- 2. H. T. Kim, Core Flow Distribution in a Modern BWR as Measured at Monticello, NEDO-10299, 1971.
- 3. General Electric Co., "Duane Arnold Reactor Internals Vibration Measurements", NEDE-23736, Oct. 1977.
- 4. Letter from R. McGaughy (Iowa Electric) to T. Murley (NRC),
Subject:
Periodic Testing of Leak Tight Integrity of Pressure Isolation Valves (Generic Letter 87-
06), dated June 11, 1987 (NG-87-1881).
- 5. J. G. Erbes, Safety Evaluation of the Jet Pump Sensing Line Clamp and Beam Assembly Duane Arnold, RDE #04-387, General Electric Company, March 1987.
- 6. General Electric Report, Safety Analys is Report for Duane Arnold Energy Center Extended Power Uprate, NEDC-32980P, May 2001.
- 7. GE Hitachi Nuclear Energy, Safety Analysis Report for Duane Arnold Energy Center Increased Core Flow, NEDC-33439P, Revision 3, August, 2009.
UFSAR/DAEC-1 T3.9-1 Revision 20 - 8/09 Table 3.9-1 STRESS
SUMMARY
FOR SHROUD SUPPORT LEGS
Primary Stress (psi) Criteria Loading Stress Type Allowable Calculated ASME Code, Section III Primary Stress Limit
for SB-168 For normal and upset
condition Normal and upset condition loads General membrane 23,300 (a ) S m = 23,300 psi Dead weight Operating-basis earthquake For emergency
condition Emergency condition loads General membrane 34,950 16,748 1.5S m = 34,950 psi Dead weight Design-basis earthquake
For faulted condition Faulted condition loads General membrane 46,600 42,300 2.0S m = 46,600 psi Dead weight Design-basis earthquake
Jet reaction forces
Pressure drop across shroud support and shroud during faulted condition a Since the calculated stress for the emergency condition is less than the allowable stress for this loading, this stress has not been listed.
Note: The shroud support legs are stiff enough to prevent buckling.
UFSAR/DAEC-1 T3.9-2 Revision 12 - 10/95 Table 3.9-2 STRESS
SUMMARY
FOR VESSEL SUPPORT SKIRT
Primary Stress (psi) Criteria Loading Stress Type Allowable Calculated ASME Code, Section III Primary Stress Limit for SA-516 Grade 70
For normal and upset
condition Normal and upset condition loads General membrane 19,600 (a ) S m = 19,600 psi Dead weight Operating-basis earthquake For emergency
condition Emergency condition loads General membrane 29,400 5,400 1.5S m = 29,400 psi Dead weight Design-basis earthquake
For faulted condition Faulted condition loads General membrane 39,200 7,700 2.0S m = 39,200 psi Dead weight Design-basis earthquake
Jet reaction forces a Since the calculated stress for the emergency condition is less than the allowable stress for this loading, this stress has not been listed. UFSAR/DAEC-1 T3.9-3 Revision 12 - 10/95 Table 3.9-3 STRESS
SUMMARY
FOR STABILIZER BRACKET - ADJACENT SHELL
Primary Stress (psi) Criteria Loading Stress Type Allowable Calculated ASME Code, Section III Primary Stress Limit for SA-533 Grade B, Class I
For normal and upset Normal and upset Pure shear 16,000 (a ) condition condition loads
0.6 x 26,700
= 16,000 psi Operating-basis earthquake For emergency
condition Emergency condition loads Pure shear 24,000 (a) 1.5 x 16,000
= 24,000 psi Design-basis earthquake For faulted condition Faulted condition loads Pure shear 32,000 13,000 2.0 x 16,000 Design-basis earthquake = 32,000 psi Jet reaction forces a Since the calculated stress for the faulted condition less than the allowable stress for this loading, this stress has not been listed.
UFSAR/DAEC-1 T3.9-4 Revision 20 - 8/09 Table 3.9-4 STRESS
SUMMARY
FOR REACTOR VESSEL INTERNALS AND ASSOCIATED EQUIPMENT Sheet 1 of 9 Criteria Loading Primary Stress Type or Location Stress Shroud Allowable Calculated Primary Stress Limit The allowable primary membrane and membrane plus bending stresses are based on ASME Code, Section III, for Type 304 stainless steel plate. For normal and upset condition Stress intensity S A = 1.0 S m Normal and upset condition loads General membrane 16,950 psi 4,259 psi* S m = 16,950 psi Upset condition pressure drop Operating-basis earthquake Weight of structure For Faulted condition Faulted condition loads General membrane 33,900 psi 8,782 psi Slimit =2.0 S m = 2.0 x 16,950 Faulted condition pressure drop
= 33,900 psi Design-basis earthquake Weight of structure For faulted condition Slimit = 2.0 S m = 2 x 16, 950 Faulted condition loads General membrane 33,900 psi 22,686 psi = 33,900 psi Pressure drop after main plus bending steamline rupture Acoustic Weight of structure Design Basis Earthquake Note: 2 S m is conservatively used as the Faulted condition allowable stress.
- Bending stress was conservatively considered as the primary membrane stress
UFSAR/DAEC-1 T3.9-5 Revision 20 - 8/09 STRESS
SUMMARY
FOR REACTOR VESSEL INTERNALS AND ASSOCIATED EQUIPMENT Sheet 2 of 9 Criteria Loading Primary Stress Type or Location Stress Top Guide - Longest Beam Primary Stress Limit The allowable primary membrane stress plus bending stress is based on ASME Code, Section III, for Type 304 stainless steel plate.
For normal and upset condition Stress intensity S A = 1.5 Normal and upset condition loads General membrane plus bending 25,388 psi 15,258 psi S m = 1.5 x 16,925 psi = 25,388 psi Operating-basis earthquake Weight of structure For emergency condition Emergency condition loads General membrane 38,081 psi 26,900 psi Slimit = 1.5S A = 1.5 x 25,388 Design-basis earthquake plus bending = 38,081 psi Weight of structure
For faulted condition Faulted condition loads General membrane 50,775 psi 26,900 psi Slimit = 2S A = 2 x 25,388 (same as emergency plus bending = 50,775 psi condition)
Top Guide Beam End Connections Primary Stress Limit ASME Code, Section III, defines material stress limit for Type
304 stainless steel. For normal and upset condition Stress intensity S A = 0.6 Normal and upset condition loads Pure shear 10,155 psi 6,509 psi S m = 0.6 x 16,925 psi = 10,155 psi Operating-basis earthquake Weight of structure For emergency condition Emergency condition loads Pure shear 15,232 psi 11,800 psi Slimit = 1.5 Design-basis earthquake Weight of structure
S A = 1.5 x 10,155 psi = 15,232 psi UFSAR/DAEC-1 T3.9-6 Revision 20 - 8/09 Table 3.9-4 STRESS
SUMMARY
FOR REACTOR VESSEL INTERNALS AND ASSOCIATED EQUIPMENT (Continued) Sheet 3 of 9 Primary Stress Type Stress Criteria Loading or Location Allowable Calculated Top Guide Beam End Connections (Continued) For faulted condition
Slimit = 2S A = 2 x 10,155 psi
= 20,310 psi Faulted condition loads (same as emergency
condition) Pure shear 20,310 psi 11,800 psi Top Guide Seismic Restraint Blocks Primary Stress Limit ASME Code, Section III, defines material stress limit for Type 304
stainless steel plate. For normal and upset condition
Stress intensity S A = 0.6 Normal and upset condition loads Operating-basis earthquake
Weight of structure Pure shear 10,155 psi 7,600 psi S m = 0.6 x 16,925 = 10,155 psi For emergency condition
Slimit = 1.5 S A = 1.5 x 10,155 = 15,232 psi Emergency condition loads Design-basis earthquake
Weight of structure Pure shear 15,232 psi 15,100 psi For faulted condition
Slimit = 2S A = 2 x 10,155
= 20,310 psi Faulted condition loads (same as emergency
condition) Pure shear 20,310 psi 15,100 psi Core Support Primary Stress Limit The allowable primary membrane
stress plus bending stress is
based on ASME Code, Section III, for Type 304 stainless steel plate. Normal and upset condition loads Normal operation pressure
drop Operating-basis earthquake General membrane plus bending 25,388 psi 16,982 psi UFSAR/DAEC-1 T3.9-7 Revision 20 - 8/09 Table 3.9-4 STRESS
SUMMARY
FOR REACTOR VESSEL INTERNALS AND ASSOCIATED EQUIPMENT (Continued) Sheet 4 of 9 Primary Stress Type Stress Criteria Loading or Location Allowable Calculated Core Support (continued) For allowable stresses, see Top Guide - Longest Beam, above. Emergency condition loads Normal operation pressure
drop Design-basis earthquake General membrane plus bending 38,081 psi 26,196 psi Faulted condition loads Pressure drop after recirculation line rupture Design-basis earthquake General membrane plus bending 50,775 psi 29,889 psi Core Support Aligners Primary Stress Limit ASME Code, Section III, defines material stress limit for Type 304
stainless steel. Normal and upset condition load Operating-basis earthquake Pure shear 10,155 psi 0 a For allowable shear stresses, see Top Guide Beam and Connections, above. Emergency condition load Design-basis earthquake Pure shear 15,232 psi 0 a Faulted condition load Design-basis earthquake Pure shear 20,310 psi 0 a Reactor Pressure Vessel Stabilizer Primary Stress Limit AISC Specification for the instruction, fabrication, and erection of structural steel for buildings. For normal and upset condition
AISC allowable stresses, but without the usual increase for earthquake loads Upset condition loads Spring preload Operating-basis earthquake Rod Bracket 63,000 psi 22,000 psi 14, 000 psi f t = 62,500 psi b f b = 12,200 psi f v = 3,900 psi a The friction force between core support and core support flange due to the preload of 9,000 lb per stud is greater than the shear load on the core support induced by design-basis earthquake. Therefore, all
aligners will not see any shear load during a design-basis earthquake. These studs have the capability, without exceeding allowable stresses, of being preloa ded to even higher levels per the newest produced ASME Section III, Subsection NG, for reactor internals criteria. b The ratio maximum stress/stress limit is highest for upset loading conditions.
UFSAR/DAEC-1 T3.9-8 Revision 20 - 8/09 Table 3.9-4 STRESS
SUMMARY
FOR REACTOR VESSEL INTERNALS Sheet 5 of 9 AND ASSOCIATED EQUIPMENT (Continued) Primary Stress Type Stress Criteria Loading or Location Allowable Calculated Reactor Pressure Vessel Stabilizer (Continued) For emergency conditions
1.5 x AISC allowable stresses Emergency condition loads Spring preload
Design-basis earthquake Bracket 33,000 psi
21,000 psi f b = 16,100 psi f v = 5,100 psi For faulted conditions Faulted condition loads Spring preload
Design-basis earthquake
Jet reaction load Bracket 36,000 psi
21,500 psi f b = 12,200 psi
f v = 5,600 psi Control Rod Drive Housing Support Primary Stress Limit AISC Specification for the design, fabrication, and erection of
structural steel for buildings. For normal and upset condition
F a = 0.60 Fy (tension) F b = 0.60 Fy (bending) F v = 0.40 Fy (shear) For faulted conditions
F a limit = 1.5F a (tension) F b limit = 1.5f b (bending) F v limit = 1.5F v (shear) F y limit = material yield strength Faulted condition loads Dead weight Impact force from
failure of a control
rod drive
housing
(dead weights and earthquake loads are very small as compared to jet
force) Beams (top cord) Beams (bottom cord) Grid structure 33,000 psi 33,000 psi
33,000 psi
33,000 psi 41,500 psi
27,500 psi f a = 11,800 psi f b = 19,800 psi f a = 9,900 psi f b = 13,800 psi f a = 40,000 psi f b = 11,100 psi Control Rod Drive Housing Primary Stress Limit The allowable primary membrane
stress is based on the ASME Code, Section III, for Class A vessels for Type 304 stainless steel. For normal and upset conditions
S m = 15,800 psi at 575°F Normal and upset condition
loads Design pressure Stuck rod scram loads
Operating-basis
earthquake Maximum membrane stress intensity occurs
at the tube-to-tube
weld near the center
of the housing for normal, upset, and emergency
conditions 15,800 psi 14,480 psi UFSAR/DAEC-1 T3.9-9 Revision 20 - 8/09 Table 3.9-4 STRESS
SUMMARY
FOR REACTOR VESSEL INTERNALS Sheet 6 of 9 AND ASSOCIATED EQUIPMENT (Continued) Primary Stress Type Stress Criteria Loading or Location Allowable Calculated Control Rod Drive Housing (Continued) For emergency conditions
Slimit = 1.5 S m = 1.5 x 15,800 = 23,700 psi Emergency condition loads Design pressure Stuck rod scram loads
Design-basis earthquake 23,700 psi 22,030 psi Control Rod Drive Primary Stress Limit The allowable primary membrane
stress plus bending stress is based
on ASME Code, Section III, for
SA-212 TP 316 tubing. For normal and upset condition S A = 1.5 S m = 1.5 x 17,375 = 26,060 psi Normal and upset condition loads c Maximum hydraulic pressure from the control rod drive supply pump Maximum stress intensity occurs at a
point on the Y-Y
axis of the indicator
tube 26,060 psi 20,790 psi Control Rod Guide Tube Primary Stress Limit d The allowable primary membrane
stresses plus bending stress is
based on the ASME Code, Section III, for Type 304 stainless steel
tubing. For normal and upset condition e S m = 15,800 15,800 psi 5,049 psi For faulted condition Slimit = 2.0 S A = 2.0 x 23,700 = 47,400 psi Faulted condition loads Dead weight
Pressure drop across
guide tube due to failure of main steamline Design-basis earthquake The maximum bending stress under
faulted loading
conditions occurs at
the center of the
guide tube 15,800 psi 9,175 psi c Accident conditions do not increase this loading. Earthquake loads are negligible. d. The Guide Tube was also evaluated and qualified for its ability to resist buckling under the external pressure differentials and the axial compressive loads for the Normal, Upset, Emergency, and Faulted conditions (See Reference 3.9 - 7).
- e. The Normal/Upset condition allowable stress is conservatively used for Faulted condition allowable stress. Table 3.9-4
UFSAR/DAEC-1 T3.9-10 Revision 20 - 8/09 STRESS
SUMMARY
FOR REACTOR VESSEL INTERNALS Sheet 7 of 9 AND ASSOCIATED EQUIPMENT (Continued) Primary Stress Type Stress Criteria Loading or Location Allowable Calculated Orificed Fuel Support Primary Stress Limit The allowable primary membrane stress plus bending stress is based on ASME Code, Section III, for Type 304 stainless steel. For normal and upset condition Stress intensity S A = 15,580 psi Normal and upset condition loads General membrane plus bending 15,580 psi 12,657 psi Upset condition pressure drop Operating-basis earthquake Weight of structure For faulted condition Faulted condition loads General membrane 35,440 psi 23,413 psi Slimi t =35,440 psi Pressure drop after main Plus bending steamline rupture Design-basis earthquake Note: Quality factor (0.65) was
considered in P m + P b allowable stress. Weight of structure Incore Housing Primary Stress Limit The allowable primary membrane
stress is based on ASME Code, Section III, for Class A vessels, for Type 304 stainless steel. For normal and upset condition
- e. S m = 15,800 psi at 575ºF 15,800 psi 15,290 psi For emergency condition (H+E)
Emergency condition
loads Design pressure
Design-basis earthquake Maximum membrane stress
intensity occurs at
the outer surface
of the vessel
penetration 15,800 psi 15,290 psi
- e. The calculated stresses are based on the Emergency and Faulted loads. This stress is conservatively used for the Normal and Upset conditions also since it is less than the membrane stress allowable of Sm.
UFSAR/DAEC-1 T3.9-11 Revision 20 - 8/09 STRESS
SUMMARY
FOR REACTOR VESSEL INTERNALS Sheet 8 of 9 AND ASSOCIATED EQUIPMENT (Continued) Primary Stress Type Stress Criteria Loading or Location Allowable Calculated Hydraulic Control Unit Piping Primary Stress Limit For ANSI B31.1.0 for power
pressure piping For normal condition
S h = 15,000 psi Normal condition load Maximum normal hydraulic system pump
pressure 3/4 in. drive withdraw piping 15,000 psi 14,596 psi For upset and emergency
condition When upset or emergency
condition exists for less than 1% of the time, the code allows 20%
increase in stress.
S a = 1.2 S b = 18,000 psi Upset condition load Shut off pump pressure
Operating-basis
earthquake
Emergency condition Shut off pump pressure
Design-basis earthquake 3/4 in. drive withdraw piping
3/4 in. drive
withdraw piping 18,000 psi
18,000 psi 16,950 psi
16,950 psi UFSAR/DAEC-1 T3.9-12 Revision 20 - 8/09 Table 3.9-4 STRESS
SUMMARY
FOR REACTOR VESSEL INTERNALS Sheet 9 of 9 AND ASSOCIATED EQUIPMENT (Continued) Primary Stress Type Stress Criteria Loading or Location Allowable Calculated Fuel Channels Primary Stress Limit Allowable stress S m for zircaloy determined according to methods recommended by
ASME Code, Section III. Normal and upset condition loads Operating-basis
earthquake Normal pressure load Membrane and bending 20,793in-lb f 6,850 in-lb g Emergency limit load 1.5 x normal limit load
calculated, using 1.5 S m = yield Emergency condition load Design-basis
earthquake Normal pressure load Membrane and bending Faulted condition load Design-basis earthquake
Loss-of-coolant accident
pressure Membrane and bending 31,190 in-1b f 13,700 in-lb g Recirculating Pipe and Pump Restraints Primary Stress Limit Structural steel: AISC
specification for the design, fabrication, and erection of
structural steel for buildings. For normal or upset condition F a = 0.60 F y (tension) For faulted condition F a limit = 1.5 F a (tension)
F y = yield strength cable (wire rope) Faulted condition loads Jet force from a complete circumferential
failure (break) of
recirculation line Brackets on 22-in. pipe Cable on pump
restraints 33,000 psi 99,000 psi 29,300 psi 61,200 psi For faulted condition
F a = 0.80 F u (tension) F u = ultimate strength f Maximum limit accounting for pressure loads. g Maximum moment. UFSAR/DAEC-1 T3.9-13 Revision 20 - 8/09 Table 3.9-5 Sheet 1 of 5 SPECIAL LOADING CRITERIA a 1. Criteria F - Deformation Limit
Any one of General Limit
- a. (permissible deformation) 0.9 (analyzed deformation causing loss of function) 1 SFmin b. (permissible deformation) 1.0 (experimental deformation causing loss of function) SFmin 2. Criteria F - Primary Stress Limit
Any one of a. (elastic evaluated primary
stresses) 2.25 (ASME III normal event permissible primary
stresses) SFmin b. (permissible load) 1.5 (largest lower bound limit load with y.p. = 150% S m ASME III) 2 SFmin c. (permissible load) 0.9 (elastic-plastic 3 calculated load causing loss of
function) SFmin a Superscript numbers are keyed to the notes at the end of this table. UFSAR/DAEC-1 T3.9-14 Revision 20 - 8/09 Table 3.9-5 Sheet 2 of 5 SPECIAL LOADING CRITERIA a
- 2. Criteria F - Primary Stress Limit General Limit (Continued)
d.. (elastic evaluated nominal primary stress) 4 0.75 (conventional ultimate strength at temperature) SFmin e. (elastic-plastic evaluated nominal primary stress) 5 0.9 (conventional ultimate strength at temperature) SFmin f. (permissible load) 0.9 (plastic instability load)6 SFmin g. (permissible load) 0.9 (ultimate load from fracture analysis) 7 SFmin h. (permissible load) 1.0 (ultimate load or loss of function load from test) SFmin a Superscript numbers are keyed to the notes at the end of this table. UFSAR/DAEC-1 T3.9-15 Revision 20 - 8/09 Table 3.9-5 Sheet 3 of 5 SPECIAL LOADING CRITERIA a 3. Criteria F - Buckling Stability Limit General Limit
Any one of a. (permissible load) 2.25 (ASME III normal event permissible load) SFmin b. (permissible load) 0.9 (stability analysis load) 8 SFmin c. (permissible load) 1.0 (instability load from test) SFmin 4. Criteria F - Fatigue Limit
Summation of fatigue damage usage with operation loads following
Miner hypotheses, either one
- a. Mean fatigue cycle usage from analysis 9 0.05 b. Mean fatigue cycle usage from test 9 0.33 c. Design fatigue cycle usage from analysis per ASME III or ANSI B31.7 1.0 a Superscript numbers are keyed to the notes at the end of this table.
UFSAR/DAEC-1 T3.9-16 Revision 20 - 8/09 Table 3.9-5 Sheet 4 of 5 NOTES
- 1. "Loss of function" can only be defined quite generally until attention is focused on the component of interest. In cases of interest where deformation limits can affect the function of Seismic Category I structures, they will be specifically delineated. Examples where deformation limits apply are: control rod drive alignment and clearances for proper insertion, core support deformation causing fuel disarrangement, excess leakage of any component, or required pumps or valves failing to operate.
- 2. The "lower bound limit load" is here defined as that produced from the analysis of an ideally plastic (nonstrain hardening) material where deformations increase with no further increase in applied load. The lower bound load is one in which the material everywhere satisfies equilibrium and nowhere exceeds the defined material yield strength using either a shear theory or a strain energy of distortion theory to relate multiaxial yielding to the uniaxial case.
- 3. It is permissible to take credit for material strain hardening in computing the load. A shear or strain energy of deformation theory may be used with a monotonic stress curve at the temperature of load application. Any approximation to the actual stress strain curve which everywhere has a lower stress for the same strain as the actual monotonic curve may be used.
- 4. The effective membrane stresses are to be averaged through the load carrying section of interest. The simplest average bending, shear, or torsion stress distribution which will support the external loading will be added to the membrane stresses at the section of interest.
- 5. Strain hardening of the material may be used for the actual monotonic stress strain curve at the temperature of loading or any approximation to the actual stress strain curve which everywhere has a lower stress for the same strain as the actual montonic curve may be used. Either the shear or strain energy of distortion flow rule may be used.
- 6. The "plastic instability load" is defined here as the load at which any load bearing section begins to diminish its cross-sectional area at a faster rate that the strain hardening can accommodate the loss in area. This type analysis requires a true stress-true strain curve or a close approximation based on monotonic loading at the temperature of loading.
- 7. For components which involve sharp discontinuities (local theoretical stress concentration >3) the use of a "fracture mechanics" analysis, where applicable, utilizing measurements of plane strain fracture toughness may be applied to compute fracture loads. Correction for finite plastic zones and thickness effects as well as gross yielding may be necessary. The methods of linear elastic stress analysis my be used in the fracture analysis where its use is supported by experimental evidence. Examples where fracture mechanics may be applied are for fillet welds or
end of fatigue life crack propagation.
UFSAR/DAEC-1 T3.9-17 Revision 20 - 8/09 Table 3.9-5 Sheet 5 of 5 NOTES (Continued)
- 8. The ideal buckling analysis is often sensitive to otherwise minor deviations from ideal geometry and boundary conditions. These effects shall be accounted for in the analysis of the buckling stability loads. Examples of this are ovality in externally pressurized shells or eccentricity of column members.
- 9. Fatigue failure is defined here as a 25% area reduction for a load carrying member which is required to function or excess leakage causing loss of function, whichever is more limiting. In the fatigue evaluation the methods of linear elastic stress analysis may
be used when the 3S m range limit of ASME III has been met. If 35 m is not met, account will be taken of (a) increases in local strain concentration, (b) strain ratcheting, (c)
redistribution of strain due to elastic-plastic effects. ANSI B31.7 piping code may be used where applicable or detailed elastic-plastic methods, strain hardening my be used, not to exceed in stress for the same strain, the steady-state cyclic strain hardening measured in a smooth low cycle fatigue specimen at the average temperature of interest. UFSAR/DAEC-1 T3.9-18 Revision 20 - 8/09 Table 3.9-6
SUMMARY
OF LOADING CONDITIONS AND CRITERIA Loading Conditions Criteria Reactor pressure vessel 1 C 1 , F 2 C 1 , F 3 C 1 , F 4 C 1 , F Internals 1 F 2 F 3 F 4 F
Piping 1 C 2 , F 2 F 3 F 4 F
Equipment and valves 1 C 2 , F 2 F 3 F 4 F
Supports and restraints 1 F 2 F 3 F 4 F where the criteria are
C 1 = ASME Code, Section III
C 2 = Industry codes
F = Special loading criteria (see Table 3.9-5)
UFSAR/DAEC-1 T3.9-19 Revision 20 - 8/09 Table 3.9-7 Deleted
.o7.------ ....--------------------_--. .06.05.04.=->-......J<<>0.03.02-.01 liP YIELD (psid)-DUANE ARNOLD ENERGY CENTER IOWA ELECTRIC LIGHT&POWER COMPANY UPDATED FINAL SAFETY ANALYSIS REPORT Guide Tube Ovality , I Figure 3.9-17 120r-------------------------------....., 100.062 OVALITYC-UI 60 u z UJ 0: UI"-"-5 UJ 0::::>Vl Vl UJ 0: 40 a.20 ROLLER MINIMUM CLEARANCE 0.1 0.09 0.08 0.07 0.06 0.05 0.03 0.02 0.01 OL.--_-.L__....L..__..l-__L-_-L__-L__..l-_---JL-_-1. __..J o RADIAL DEFLECTION (;n)DUANE ARNOLD ENERGY CENTER IES UTILITIES, INC.UPDATED FINAL SAFETY ANALYSIS REPORT Guide Tube Deflection Figure 3.9-18 Revision 17 10/03
UFSAR/DAEC - 1 3.10-1 Revision 13 - 5/97 3.10 SEISMIC QUALIFICATION OF SEISMIC CATEGORY I INSTRUMENTATION AND ELECTRICAL EQUIPMENT
3.10.1 SEISMIC QUALIFICATION CRITERIA
3.10.1.1 General Electric-Supplied Instrumentation and Control Equipment
All instrumentation required for nuclear safety is capable of performing all functions important to safety during normal reactor operation, design-basis accidents and postaccident operation while being subjected to accelerations that are in excess of those calculated for the DBE at the point of attachment of the instrument (or module) to the
building structure. Qualification is achieved by test and/or analysis at acceleration values of 1.5g horizontal and 0.5g vertical over a frequency range of 0.25 to 33 Hz. The seismic analysis of balance-of-plant (BOP) instrumentation is discussed in Sections 3.10.1.2 and 3.7.2, and Table 3.10-1 lists the criteria used to determine seismic classification of BOP instrumentation.
Acceleration at the point of attachment for a specific instrument is related to the floor acceleration by the transmissibility of the supporting structure (panel or rack). The racks and panels are designed to have low amplification (close to 1 at frequencies below 10 Hz and not to exceed 2.5 at frequencies above 10 Hz up to 33 Hz). The amplification characteristics of each general type of rack or panel design are demonstrated by a vibration test supplemented by analysis of the low end of the frequency spectrum (outside the capability of the test equipment). Instrumentation device types are individually qualified by vibration test for 1.5g (or more) horizontal and 0.5g (or more) vertical.
A panel or rack assembly is thus conservatively qualified for use where the actual
floor acceleration does not exceed the value obtained by dividing the lowest instrument qualification value by 2.5. However, where the response spectrum (acceleration versus frequency) of the floor at the location of the panel is know and the amplification spectrum (amplification versus frequency) of the panel (or rack) is known, a more accurate qualification limit may be established.
Seismic qualification at 1.5g horizontal and 0.5g vertical at point of attachment is sufficient to ensure operability of instrumentation in a worst-case loading situation at any location of essential instrumentation in the plant.
The small incremental loading contributed by the connecting wiring (given appropriate cable support) is considered to be adequately provided for by the margin contained in the general seismic qualification requirement.
The effect of electric conduit connections to instruments and of instrument piping connections to instrument racks (again given appropriate conduit and pipe support) is generally to increase the stiffness of the instrument or rack support system and thus reduce rather than increase the maximum loading on individual instruments. UFSAR/DAEC - 1 3.10-2 Revision 13 - 5/97 Attachment systems (bolts, clamps, etc.) are demonstrated to be capable of supporting operating instrumentation that they are designed to support during seismic testing without the benefit of additional support normally offered by connections to cables, conduits, and instrument piping.
Condensing chambers, temperature reference columns, and SRM/IRM dry tubes
are designed and fabricated in accordance with the ADME Code, Section III (Class 1 equipment), and are required to be inspected by a third party and appropriately code stamped as certification of their compliance. They are also required to be dynamically analyzed with seismic forces superimposed on normal operating loads from system pressure and temperature for purposes of qualification.
Table 3.10-2 lists maximum usable g levels for which the various types of instrumentation devices or module have been seismically qualified by actual vibration
testing.
3.10.1.2 Bechtel-Supplied Instrumentation and Control Equipment
The purchase specifications for the instrumentation and control equipment
supplied by Bechtel required that each type of Class 1 device by individually qualified for Seismic Category I service by vibration test or suitable analysis.
The methods of test or analysis used for seismic qualification of Class 1 electrical instrumentation and control equipment met the general requirements of IEEE Standard
344-1971.
The maximum acceleration levels (vertical and horizontal) that each device can endure without failure while performing its function were determined by the test and/or
analysis.
The device was accepted for application in DAEC only if the failure threshold
acceleration level was greater than the calcu lated DAEC DBE acceleration at the point of its attachment to the supporting structures (appropriate panel, rack, or individual mounting), under all modes of operation including the design-basis accident and
postaccident operation.
The supporting structure was analyzed to show that, during a DBE, none of the instruments or electrical devices mounted th ereon would be subjected to acceleration levels equal to or greater than their failure threshold acceleration.
Table 3.10-3 lists the essential (Class 1) generic types of instrumentation and
control devices supplied by Bechtel for the DAEC. UFSAR/DAEC - 1 3.10-3 Revision 13 - 5/97 3.10.2 METHODS AND PROCEDURES FOR QUALIFYING ELECTRICAL EQUIPMENT AND INSTRUMENTATION
All types of instrumentation and control devices used in reactor protection system and safeguard systems were tested in an operational condition. The equipment, for instance, was supplied with appropriate input signals and/or trip inputs and monitored
with the trips set within 2% (upscale and downscale) of the input signal value. Relays were monitored in energized and deenergized conditions for both normally open and normally closed contacts. Pressure, level, and flow switches were vibrated while provided with simulated input signals that a pproached setpoints within 2% of setting and switch contacts monitored for false closure or opening (spurious trips). The instrument or device was mounted the way it is mounted in its actual application in the plant.
During seismic scans, the devices were monitored for resonant frequencies using either accelerometers, strobe lights or both. The accelerometers were connected to charge amplifiers which were used to drive a recorder for permanent recording of data. A meter output strobe light aided in detecti on of the resonant frequencies and response modes of the devices. The detection and expl oration of the resonant frequencies were first made over the criteria frequency range to detect possible weak points that could
result in failure during subsequent endurance and higher acceleration runs.
Vibration endurance and maximum acceleration scans over the frequency range were then conducted (after the resonant search) to subject the hardware to the maximum
specified accelerations of 1.5g horizontal in two perpendicular axes and 0.5g vertical. Also each instrument or device was tested at 33 Hz at increasing amplitudes until the maximum acceleration without malfunction was determined.
Table 3.10-2 lists maximum usable g levels for which the various types of instrumentation devices or modules have been seismically qualified by actual vibration
testing.
3.10.3 METHODS AND PROCEDURES OF ANALYSIS OR TESTING AND
SUPPORTS OF ELECTRICAL EQUIPMENT AND INSTRUMENTATION
In general, the method selected to satisfy seismic design criteria for cable tray supports, battery and instrument racks, and control consoles consists of the following:
- 1. Calculation of the fundamental frequency.
- 2. Determination of the acceleration response from the building floor spectrum curves.
- 3. Stress calculations to verify structural adequacy of cabinets and support bolts and brackets.
UFSAR/DAEC - 1 3.10-4 Revision 13 - 5/97 Some units, like battery chargers, were tested by subjecting them to vibrations simulating the seismic disturbance.
The means used to verify the adequacy of the seismic design are discussed in
Section 3.7.2. UFSAR/DAEC - 1 3.10-5 Revision 13 - 5/97 REFERENCES FOR SECTION 3.10
- 1. General Electric Service Advice Letter (SAL) 721-PSM-174.1, "PVD and HGA Relay Seismic Data," dated May 12, 1983.
UFSAR/DAEC-1 T3.10-1 Revision 13 - 5/97 Table 3.10-1 CRITERIA FOR SEISMIC CLASSIFICATION OF INSTRUMENTATION Seismic Requirements of Process Line to Which Instrument is Attached Function of Instrument
Safety-Related a Not Safety Related Seismic Category 1 Instrument process lines: Seismic Category 1 criteria up to the instrument itself
Instrument itself: Must remain functional during and after the seismic
event Instrument process lines: Seismic Category 1 criteria up to instrument itself; if an auto/remote isolation valve (EFCV) exists, Seismic
Category 1 criteria need
only extend through the
isolation valve. No seismic requirements None should exist Instrument process lines: No requirement except that instrument lines that
penetrate the drywell b will have EFCVs and will meet Seismic Category 1 criteria from the drywell
penetration through the
EFCV.
Instrument itself: No requirements aInstruments are classified as safety related if they Initiate reactor shutdown (scram). Initiate reactor vessel-drywell isolation; this includes instruments that will isolate individual process lines upon indication of rupture or leakage. Initiate reactor building isolation Activate emergency core cooling systems and their support systems for initial and long-term core cooling. Two additional classifications are implied by the above and are listed for completeness. Monitor/indicate possible radioactive process line leakage. Monitor for possible radioactive releases. b Reactor coolant pressure boundary process lines will have EFCVs.
UFSAR/DAEC-1 T3.10-2 Revision 13 - 5/97 Table 3.10-2 Sheet 1 of 3
SUMMARY
OF SEISMIC QUALIFICATION OF INSTRUMENTATION
Maximum Usable Level (g) Horizontal Vertical Equipment Description Remarks
- 1. Voltage preamplifier 8.5 8.5
- 2. TIP ball valve 25 25
- 3. IRM detector >1.5 >1.5 (maximum not determined)
- 4. Reactor level switch (Yarway) snap acting 10 10 5. Temperature control switch 12 12 6. Contractor (GECr 105) 12 12
- 7. Indicator trip unit (GE/MAC) 15 15 8. LPRM fixed incore detectors >1.5 >1.5 (maximum not determined) 9. TIP shear valve assembly 10 10 10. Timer 9 9
- 11. Temperature switch 4 4
- 12. Pressure transmitter 10 12
- 13. Flow switch 4 10 HPCI and RHR minimum flow bypass 14. Pressure switch 11 11
- 15. Flow switch (standby liquid flow) 15 15 16. Flow converter 15 15 17. Flow auxiliary unit 11 11
- 18. Source-range monitor 3 15
- 19. Intermediate-range monitor (dc) 1.5 0.5 20. Power supply (20 vdc) 1.5 0.5
UFSAR/DAEC-1 T3.10-3 Revision 13 - 5/97 Table 3.10-2 Sheet 2 of 3
SUMMARY
OF SEISMIC QUALIFICATION OF INSTRUMENTATION
Maximum Usable Level (g) Horizontal Vertical Equipment Description Remarks
- 21. Intermediate-range monitor 3 15 22. Sensor converter 15 15 23. Pressure switches 15 15 Scram and low pressure permissive
- 24. Temperature element 15 15
- 25. Pressure switch (drywell) 15 15 Drywell pressure scram 26. Pressure switch 15 15 27. Pressure switch 10 10
- 28. Pressure switch (drywell) 15 15 29. Pressure switch 2 2 30. Relay (CR120A) 12 12
- 31. Relay (HFA) 4 10
- 32. Relay (HGA) a 1.1 5 33. Relay (CR2820) 25 25 Time delay
- 34. Relay (CR120K) 25 25
- 35. Relay (CR120KT) 12 12 Time delay
- 36. Switch (SBM) 25 25
- 37. IRM range switch 8.5 8.5
- 38. T/C selection switch 25 25
- 39. Switch oil-tight (CR2940) 20 20 aApplication of this relay, where opening of a normally closed contact on a deengergized HGA relay can defeat a safety action, is unacceptable (see Reference 1)
UFSAR/DAEC-1 T3.10-4 Revision 13 - 5/97 Table 3.10-2 Sheet 3 of 3
SUMMARY
OF SEISMIC QUALIFICATION OF INSTRUMENTATION
Maximum Usable Level (g) Horizontal Vertical Equipment Description Remarks
- 40. IRM trip auxiliary 12 12
- 41. Scram solenoid fuse panel 10 10 42. Fuse 15 15 43. Gamma chamber 1.5 0.5
- 44. Controller 5 5
- 45. Manual loading station 2 2
- 46. Millivolt converter 3 3
- 47. Pressure transmitter 2 2
- 48. Flow transmitter 2 2
- 49. Pressure transmitter 12 12
- 50. Dual alarm unit 5 5
- 51. Proportional amplifier (flow summer) 3 3 52. Square-root converter 11 11 53. GE/MAC power supply 11 11
- 54. LPRM "page" 1.5 0.5
- 55. APRM "page" 1.5 0.5
- 56. ICPS "page" 1.5 0.5
- 57. RBM "page" 1.5 0.5
- 58. PRM system 1.5 0.5
- 59. Agastat TR relay (GE No. 164C5257) 4.6 4.6
UFSAR/DAEC-1 T3.10-5 Revision 18 - 10/05 Table 3.10-3 BECHTEL GENERIC INSTRUMENT LIST
Switch, pressure, bourdon tube Switch, pressure, diaphragm
Switch, differential pressure
Switch, level, electronic
Switch, hand Indicator, ammeter Transmitter, pressure, force balance Transmitter, differential pressure, force balance Timer, electric, pneumatic rc
Controller, electronic input and output
Converter, signal, square root
Recorder, electronic
Power supply, solid state, dc, low voltage
Valve, control, air-operated Valve, control, motor-operated
Valve, excess flow check
Valve, solenoid Analyzer, hydrogen Analyzer, oxygen
- Electrical control relays Original design classification. In September 2003, NRC revised 10 CFR 50.44, "Standards for Combustible Gas Control for Nuclear Power Reactors," to downgrade these instruments to non-safety grade.
UFSAR/DAEC - 1 3.11-1 Revision 22 - 5/13 3.11 ENVIRONMENTAL DESIGN OF ELECTRICAL EQUIPMENT
3.11.1 EQUIPMENT IDENTIFICATION AND ENVIRONMENTAL CONDITIONS
3.11.1.1 Equipment Identification
Electrical equipment required to function under postulated harsh environment accident conditions is identified in the DAEC Equipment Data Base. The NRC Safety Evaluation Report (Reference 1) concludes that the DAEC Equipment Qualification Program is in compliance with the requirements of 10 CFR 50.49.
3.11.1.2 Environmental Service Conditions
3.11.1.2.1 Harsh and Mild Service Conditions
Environmental service conditions are categorized as harsh or mild. These categories describe the type of postaccident environment that the equipment is subjected to before or during the time that the equipment performs its safety function. To be classified as a harsh environment, a location must meet either of the following two criteria:
- 1. The total integrated radiation dose over the installed life of the component plus the required operating post accident time exceeds 1 x 10 5 rads for all equipment in the equipment environmental qualification program except for certain electronic devices identified in DAEC Design Control Procedures. The total integrated radiation dose limit over the required operating time for those electronic devices is 1 x 10 3 rads. DAEC Design Control Procedures address the radiation sensitivity of electronic devices such as silicon-controlled rectifiers, fiber optic cables, and metal oxide silicon field effect transistors and require design evaluations for their placement or relocation.
- 2. The nonradiation parameters as a result of a high energy line break (HELB) (temperature, pressure, humidity, etc.) would be significantly more severe than the non-radiation parameters that would occur during normal plant operation, including anticipated
operational transients.
All other equipment areas are classified as mild environments. The environmental service conditions for the site are described in DAEC controlled document QUAL-SC101 (Reference 9) which references calculations and analyses to support the values assigned to each area. The environmental service conditions analysis QUAL-SC100 (Reference 2) contains more detailed "working" guidance on harsh environment criteria.
In assessing radiological conditions for accident doses to equipment, DAEC methods are those based on RG 1.89. No change was made to adopt Alternative Source Term methodology (NUREG-1465) to the EQ Program as allowed by RG 1.183. UFSAR/DAEC - 1 3.11-2 Revision 22 - 5/13 3.11.1.2.2 Environmental Conditions Inside the Drywell
The environmental conditions inside the drywell on the basis of postulated LOCA and HELB conditions are described in Reference 2 and summarized in Table 3.11-1. The drywell pressure and temperature response following the postulated LOCA are provided in Chapter 15. Relative humidity is taken as 100% based on practical knowledge of a steam-water environment. Containment spray is limited to suppression pool water. The maximum flood level in the
drywell has been calculated to be at
3.11.1.2.3 Environmental Conditions Outside the Drywell, Subject to HELB
HELB is discussed in detail in Section 3.6. The environmental conditions (temperature, pressure, radiation and humidity) due to a pipe break in the steam tunnel, HPCI room, RCIC room, RWCU Heat Exchanger room, TIP room, TIP room mezzanine, torus room, and turbine building are listed in Reference 9. Analysis of submergence of safety-related electrical equipment outside the containment is not required but has been performed and all safety-related electrical equipment that is required to function to mitigate the consequences of the HELB in an area is located above the maximum flood elevation. Chemical and demineralized water sprays other than fire protection systems do not exist outside the drywell.
All electrical and control panels are located outside the drywell. Several panels are
located in areas subject to HELB. Each of these panels has been reviewed to determine what components are located on each panel and the function of each component. It has been determined that no component on an electrical panel or instrument rack located in a HELB environment is required to function to mitigate the consequences of the HELB in that area. Therefore, HELB environmental conditions are not applicable for electrical and control panels or for components mounted on these panels or racks.
3.11.1.2.4 Environmental Conditions Outside the Drywell Where the Recirculation of Post-LOCA Fluid Would Occur
The 30-day dose levels outside the primary containment are listed in Reference 9. Detailed radiation dose calculations have been performed for plant areas in which electrical panels and instrumentation are located. Those instruments located in radiation harsh environments (per the definition in Reference 2) and which meet the qualification requirements
in 10CFR50.49 or the Division of Operating R eactor (DOR) Guidelines, are evaluated for inclusion in the DAEC EQ Program based on the guidelines given in controlled administrative procedures and instructions. Those instruments which meet the program requirements for inclusion in the EQ Program are listed on the EQ Program master list (Reference 11). A list of equipment excluded from the EQ Program is contained in Reference 12. None of the areas in
which electrical panels are located are identified as radiation harsh.
UFSAR/DAEC - 1 3.11-3 Revision 22 - 5/13 3.11.1.2.5 Mild Environments
The plant areas that are not covered under Sections 3.11.1.2.2, 3.11.1.2.3, and 3.11.1.2.4 are considered mild environmental areas with respect to HELB and post-LOCA radiation. Equipment located in these areas is protected and maintained in a suitable environmental condition by the heating and ventilation systems. Heating and ventilation systems employing redundant components and powered by essential power buses are provided for the following
areas:
- 1. Control building, including control room, cable spreading room, battery rooms, and essential switchgear rooms.
- 2. Standby diesel-generator rooms.
- 3. Intake structure.
- 4. Emergency service water/RHR service water pump rooms.
- 5. Reactor building via standby gas treatment system and engineered safeguards area HVAC.
3.11.1.2.6 Evaluation of Service Conditions Inside Containment for a LOCA
As described in Reference 2, a combination of accidents establish the environmental
conditions inside the drywell. The following discussion of service conditions is based on these
accidents.
- 1. Temperature Conditions
The drywell temperature conditions are established by an intermediate (1.0 sq. ft to 0.25 sq. ft) or a small (0.1 sq. ft to 0.01 sq. ft) pipe break accident.
The drywell gas temperature reaches a maximum of 331°F which is above the containment design temperature of 281°F; however, the containment shell temperature does not reach that temperature.
DAEC has established the qualification temperature for the drywell as 340°F for equipment which has been qualified by the DOR Guidelines as shown in the BWR Environmental Design Specification (Reference 10). New and replacement equipment is qualified to the current requirements of 10 CFR 50.49 using the temperature from the
current analysis of record. UFSAR/DAEC - 1 3.11-4 Revision 22 - 5/13
- 2. Pressure Conditions The drywell pressure conditions are established by a recirculation line break. The maximum calculated DAEC drywell pressure is 45.7 psig.
- 3. Radiation
A 60-yr normal integrated dose plus a 30-day postaccident dose have been used to evaluate the adequacy of equipment qualification inside the drywell. Reference 9 gives the integrated dose for inside the drywell 30 days following a LOCA and the normal 60-year integrated dose. This plant-specific analysis establishes the maximum total
integrated radiation doses inside the drywell.
Only gamma radiation has been considered if the component is enclosed in a nonorganic material (e.g., valve operators, splices, and terminal boards in junction boxes). For components in organic material (e.g., cable), it has been shown that 70 mils of jacket insulation reduces the beta dose to less than 10% of the total gamma dose. Therefore, beta radiation is only considered in the total integrated dose for unjacketed cable.
- 4. Submergence
The maximum possible flood level in the drywell is approximately which corresponds to the entrance to the vent pipes from the drywell to the suppression pool. When the water level in the drywell reaches this elevation, it will flow into the torus via the vent pipes. All electrical equipment required to function under harsh environment accident conditions that is inside the primary containment is located above and therefore submergence is not considered in the evaluation of environmental qualification adequacy.
The maximum possible flood level is based on the following fluid volumes being discharged to the primary containment:
- a. Reactor vessel.
- b. Limited condensate storage tank discharge (limited by automatic transfer of ECCS pumps to the torus on high torus level or by manual operator action in accordance with emergency operating procedures).
- c. Recirculation system piping.
- d. Feedwater system piping.
- e. Main steam piping.
UFSAR/DAEC - 1 3.11-5 Revision 22 - 5/13 These volumes, combined with the maximum normal volume of water in the suppression pool, is less than the total volume of the suppression pool up to the suppression chamber/drywell vacuum breaker. Therefore, the maximum water level in the drywell corresponds to the entrance to the vent pipes from the
drywell to the suppression pool The drywell can be flooded with river water from the RHR service water system. While this mode of operation is
possible, it would be deliberate and the consequences of flooding the electrical equipment in the drywell would have to be considered before flooding. For the purpose of environmental qualification, the maximum flood elevation is
calculated to be
- 5. Containment Spray
The use of a chemical spray at the DAEC is limited to the potential use of the low-pressure injection system to inject suppression pool water into the drywell via spray headers. This suppression pool water is expected to maintain a relatively neutral pH even
following the postulated LOCA. Hence, the corrosive reactions with suppression pool
water are expected to be negligible because of its neutral condition.
3.11.2 QUALIFICATION TESTS AND ANALYSES
3.11.2.1 Qualification Test Requirements
Equipment and components requiring environmental qualification have been tested or
analyzed in accordance with 10 CFR 50.49 or the NRC Division of Operating Reactors (DOR) guidelines. A record of the demonstrated level of qualification is contained in the DAEC controlled documents (Qual series) for each equipment and component item.
3.11.2.2 Qualification Test Results
Qualification test results are recorded in the DAEC Equipment Data Base and Environmental Qualification files for each piece of equipment in the Environmental Qualification Program.
3.11.2.3 Qualification Methods
Equipment qualification methods are consistent with 10 CFR 50.49 requirements or DOR
guidelines, as applicable.
3.11.3 LOSS OF VENTILATION
The criteria governing the design of the air conditioning and ventilation for the control room and other safety-related equipment rooms require that maximum temperatures for normal operating conditions not be exceeded assuming the failure of any one active component. UFSAR/DAEC - 1 3.11-6 Revision 22 - 5/13 Reference 9 compares the maximum temperatures for normal equipment operation and the calculated maximum room temperature under accident conditions. The calculated maximum room temperatures are based on outside ambient temperatures of 90°F dry bulb and 76°F wet bulb. Local outside ambient is expected to exceed these values approximately 2.5% of the time.
If abnormal outside ambient temperatures occur, equipment operation will not be impaired. As an example, outside ambient temperature of 105°F maintained for 24 hr would cause the control room temperature to increase to 95°F if only one cooling system were in operation. This maximum room temperature is well below the maximum operating temperature and will not make the room uninhabitable.
To achieve the design objectives, the listed areas are cooled and/or ventilated by systems that provide for 100% redundancy. The calculated temperatures are based on only one of these units operating. Furthermore, the units are housed in Seismic Category I structures or, in the case of the diesel-generators, in a structure whose failure will not cause the failure of the contained equipment. Where cooling water is required for chillers or cooling units, it is supplied by the emergency service water system, which is also designed using safety system criteria. The
cooling and/or ventilating units are powered from redundant portions of the standby power system.
Based on this design, there is no postulated instance wherein a complete loss of air
conditioning or ventilation can occur. The redundancy ensures that one unit will provide the
designed cooling or ventilation, thereby ensuring the operability of safety-related control and electrical equipment.
The inspection and testing of the heating, ventilation and air conditioning (HVAC) systems is discussed in Section 9.4.
3.11.4 ESTIMATED CHEMICAL AND RADIATION ENVIRONMENT
3.11.4.1 Chemical Environment
The chemical environment for design-basis accidents is described in Section 3.11.1.2.
3.11.4.2 Radiation Environment
The radiation environment for design-basis accidents is described in Section 3.11.1.2.
An evaluation of ESF systems is described in the following sections.
In addition to the analysis presented in Chapter 15, an evaluation was made of the adequacy of the containment and engineered safety features using the assumptions described in Section
3.11.4.2.1.
UFSAR/DAEC - 1 3.11-7 Revision 22 - 5/13 3.11.4.2.1 Source Terms Assumptions
For the purposes of calculating the dose, heat loading, airborne, or waterborne activity, the following assumptions were made:
- 1. The halogen and noble gas initial sources were taken from data on fission product loading for BWR fuel. This data contains a more extensive list of isotopes, and results in a larger (conservative) radiological source term than would be obtained using only the isotopes listed in Table IV, External Gamma Dose Rates of TID-14844.
6
- 2. The core particulate activity was taken from the ANS standard afterheat curve.
7 The activity at any time was obtained by dividing the afterheat curve at that particular time by
an average energy of 0.7 MeV.
- 3. The charcoal adsorber iodine loading includes iodine-129 and iodine-127. The amount of each of these isotopes in the core was determined from Blomeke and Todd.
8
- 4. The activity in the suppression pool was assumed to be 50% of the core halogen inventory and 1% of the core particulate activity, which are instantaneously released to
the suppression pool.
- 5. The airborne activity in the primary containment was assumed to consist of 100% of the core noble gas activity, 25% of the core halogen activity, and 1% of the core particulate activity, which are instantaneously released to the primary containment.
- 6. The airborne activity noted in item 5 is released at a constant leak rate of 2.0% per day to the secondary containment, uniformly mixed in the secondary containment and released to the standby gas treatment system at the rate of 1.0 air changes per day.
- 7. For the determination of the activity and heat loading on the charcoal adsorbers and the HEPA filters in the standby gas treatment system, the primary containment activity noted in item 5 above was assumed to be released at a constant leak rate of 2.0% per day directly to the standby gas treatment system where the filter and adsorber efficiency was assumed to be 100%.
A historic table of the activities in the various systems at various times after the
TID-14844 release accident is shown in Table 3.11-3. The values in Table 3.11-3 are based on a primary containment leak rate of 2.0% per day. Reference 9 contains references to current analyses and source terms. These source terms contain a larger set of fission products and result in conservative dose rates compared to TID-14844.
UFSAR/DAEC - 1 3.11-8 Revision 22 - 5/13 3.11.4.2.2 Standby Gas Treatment System
The standby gas treatment system can contribute as a significant postaccident radiation source. The primary contribution is from the accumulation of radioactive iodine in the system's filter units. This source term is evaluated in calculations referred to in Reference 2.
The charcoal adsorber in each train contains approximately 1224 lb of net effective
charcoal. Charcoal radioiodine loading calculati ons for the DAEC were originally derived using the TID-14844 Source Term methodology. The total iodine loading at the end of 30 days of a design basis LOCA is 1900g. This is predominan tly iodine 127 and iodine-129. The resultant specific loading is 3.4 mg of iodine per gram of charcoal. Work performed at Oak Ridge National Laboratory has shown that removal efficiencies over 99% for both elemental and organic iodine can be achieved with charcoal loadings as high as 4.4 mg/gm.
DAEC analyses of dose consequences have been revised to use the Alternative Source Term of NUREG-1465. Most fission product iodine is now assumed to be released as a particulate form that will collect on the HEPA filters, reducing specific iodine loading 30 days after a LOCA to 0.003 mg of iodine per gram of carbon. This value is well below the 2.5 mg/gm iodine loading stated in Regulatory Guide 1.52. The larger values for iodine filter
loading are still conservatively used for charcoal filter radiation heating.
3.11.4.2.3 Emergency Core Cooling System Components
Postaccident recirculation of radioactive contaminated fluid is a potentially significant source term. Such source terms are evaluated in calculations referred to in Reference 9.
The emergency core cooling system components are located in compartments shielded from the torus and the drywell. All ECCS components that handle torus water after the LOCA are located inside the secondary containment. There are no ECCS components located inside the primary containment that would be required to function following the postulated LOCA and that could suffer significant radiation damage. The ECCS components have been examined for materials that are subject to radiation damage. The principal dose to the ECCS components located in the shielded compartments is from the radioactive torus water being circulated through the ECCS components. The residual
heat removal (RHR) pump suction is 14 in.; the discharge is 12 in. The dose rate on the surface of a 14-in. standard weight pipe is listed in Table 3.11-2.
The integrated dose is for various times after the accident based on the source terms listed in Table 3.11-3. The doses to various ECCS components in the compartments are
evaluated relative to the reference dose where the reference dose is equated to the 30-day dose
on the surface of the 14-in. standard weight pipe. This reference dose is 5.6 x 10 5 rad. The RHR and core spray pumps must operate after the LOCA. These pumps contain seals made of a carbon washer moving relative to a metal seat. Both the washer and seat UFSAR/DAEC - 1 3.11-9 Revision 22 - 5/13 materials have thresholds for damage on the order of 10 11 rad (see Reference 13). This threshold for damage is orders of magnitude above the reference dose of 5.6 x 10 5 rad. The seals also contain elastomer O-rings made of buna-N and a special elastomer material. One of the O-rings has a Teflon retainer ring. All of the O-rings and seals are stationary. All of these materials suffer 25% damage from doses on the order of 4 x 10 6 rad, which is less than the reference dose. The radiation dose to the Teflon would be an even higher percentage of damage. However, if
the O-rings and the Teflon retainer failed, there would be only a slight increase in leakage and the pump would continue to run and provide the essential ECCS cooling as all of the O-rings are
stationary seals.
The RHR system and core spray pump motors contain reservoirs of lubricating oil. The lubricating oil will stand a dose on the order of 10 8 rad, which is well above the reference dose of 5.6 x 10 5 rad. The RHR and core spray pump motor insulation is a proprietary material with a radiation damage tolerance for 25% damage to the dielectric properties, which is essentially the same as the reference dose. However, it should be noted that the motors are located several feet from the pipes and therefore should receive less than the reference dose.
3.11.4.2.4 Materials Within Primary Containment
- 1. The Company is not committed to Regulatory Guide 1.54, June 1973. See Chapter 1, Section 1.8.30 for details.
- 2. Drywell Coating - The doses at the interior surface of the drywell are listed in Table 3.11-2. These doses are based on the source strength listed in Table 3.11-3.
Service Level I coating materials on the interior surface of the drywell will be subjected to a 30-day dose of 2.3 x 10 7 rad at the surface of the drywell, which is within their radiation capability.
- 3. Electrical Penetrations - The primary containment electrical penetrations contain double seals. One seal is inside the primary shield and is subjected to a 30-day
integrated dose of 2.3 x 10 7 rad. The other seal is outside the primary shield and would be subjected to a dose that is orders of magnitude less than at the interior surface of the drywell. Both seals are made of a vacuum-cast epoxy resin. Some
low voltage power and control penetrations have been replaced. Seals on each end of these replacement penetrations are manufactured of polysulfone. The conductors passing through the replacement low voltage power and control penetrations are insulated with polyimide (Kapton).
3.11.4.2.5 Control Room
See Section 6.4.4
2011-021 2011-021 UFSAR/DAEC - 1 3.11-10 Revision 22 - 5/13 REFERENCES FOR SECTION 3.11
- 1. Letter from D. B. Vassallo, NRC, to L. Liu, Iowa Electric,
Subject:
Safety Evaluation Report, Environmental Qualification of Electric Equipment Important to Safety, Duane Arnold Energy Center, dated January 10, 1985.
- 2. QUAL-SC100 "Duane Arnold Energy Center Environmental Service Conditions Analysis".
- 3. Deleted.
- 4. Letter from T. A. Ippolito, NRC, to Duane Arnold, Iowa Electric,
Subject:
Safety Evaluation Report for the Environmental Qua lification of Safety-Related Electrical Equipment at the Duane Arnold Energy Center, dated June 3, 1981.
- 5. Letter from L. D. Root, Iowa Electric, to H. R. Denton, NRC,
Subject:
Response to NRC Staff's Safety Evaluation Report on Environmental Qualification of Safety-Related Electrical Equipment at the Duane Arnold Energy Center, dated September 8, 1981 (LDR-81-257).
- 6. J. J. DiNunno, et al., "Calculation of Distance Factors for Power and Test Reactor Sites," TID-14844, March 23, 1962.
- 7. "Proposed Standard - Energy Release Following Shutdown of Uranium Fueled, Thermal Reactors," American Nuclear Society (approved by Subcommittee ANS-5 on June 11, 1968).
- 8. J. O. Blomeke and M. F. Todd, "Uranium-235 Fission Product Production as a Function of Thermal Neutron Flux, Irradiation Time and Decay Time," ORNL-2127, November 12, 1958.
- 9. QUAL-SC101, Duane Arnold Energy Center Environmental and Seismic Service Conditions.
- 10. GE Specification 22A3018, BWR Environmental Requirements.
- 11. QUAL-E001, "Environmental Qualification Master List".
- 12. DBD-A64-001, "Environmental Qualification Topical Design Bases Document".
- 13. Rockwell, III, Theodore, TID-7004 - Reactor Shielding Design Manual, March 1956.
UFSAR/DAEC-1 T3.11-1 Revision 22 - 5/13 Table 3.11-1 MAXIMUM ENVIRONMENTAL CONDITIONS INSIDE THE DRYWELL
FOLLOWING THE POSTULATED LOCA/HELB Temperature Pressure a Relative Humidity b Containment
Spray c Gamma Radiationd,e,g Submergence Elevation f 340°F
45.7 psig 100% Suppression pool water 30-day 2.0 x 10 7 rad 60-year 3.2 x 10 7 rad 60-year plus 30-day 5.2 x 10 7 rad a Section 6.2 b Postulated.
c Section 1.8.
dGE Specification 22A3018, BWR Environmental Requirements. eThe beta radiation dose mentioned in paragraph 3, of section 3.11.1.2.6 is not included in the 30-day accident dose shown in the table. The beta dose requirement is shown in reference 9. f By evaluation. g Reference 9 to Section 3.11. 2012-001 UFSAR/DAEC-1 T3.11-2 Revision 22 - 5/13 Table 3.11-2 APPROXIMATE DOSE RATES AND INTEGRATED DOSES FOR VARIOUS EQUIPMENT OR LOCATIONS Loss-of-Coolant Fission Product Release Based on TID-14844 Assumptions, Integrated Dose (rad) 60-Year Normal Integrated Location or Equipment Maximum Dose Rate (rad/hr) 12 hr 2 days 30 days 180 days Doses (rad) (100% load factor at rated power) Surface 14 in.
standard weight pipe
1.4 x 10 4 7.5 x 10 4 2.5 x 10 5 5.6 x 10 5 7.9 x 10 5 0.0 Interior surface drywell (core
region)
1.0 x 10 6
4.8 x 10 6
1.2 x 10 7
2.3 x 10 7
3.3 x 10 7 3.0 x 10 7 Floor or corner compartment
containing core spray pump seals
3.3 x 10 1
1.2 x 10 2
1.3 x 10 3
3.8 x 10 3
9.1 x 10 3 7.5 x 10 3 Pump seals (ECCS) 1.4 x 10 4 7.5 x 10 4 2.5 x 10 5 5.6 x 10 5 7.9 x 10 5 0.0 Secondary containment
ground floor
area
1.3 x 10 2
5.3 x 10 2
4.8 x 10 3
1.4 x 10 4
3.3 x 10 4 4.5 x 10 2 Refueling floor 5.3 x 10 2 2.2 x 10 3 2.0 x 10 4 5.8 x 10 4 1.4 x 10 5 4.5 x 10 2 2012-001 UFSAR/DAEC-1 T3.11-3 Revision 22 - 5/13 Table 3.11-3 ACTIVITY, MASS LOADING, AND HEAT L OADING AT VARIOUS LOCATIONS FOR TID-14844 RELEASE ASSUMPTIONS, PRIMARY CONTAINMENT LEAK RATE OF 2.0% PER DAY Parameter 1 Hr 8 Hr 1 Day 10 Days 30 Days Peak Value and Time Activity in SP, Ci 2.1 x 10 8 9.6 x 10 7 5.6 x 10 7 1.8 x 10 7 8.0 x 10 6 2.1 x 10 8 at T = 0 hr Heat load in SP, kW 2.4 x 10 3 6.1 x 10 2 2.9 x 10 2 7.5 x 10 1 4.3 x 10 1 2.4 x 10 3 at T = 0 hr Activity airborne in PC, Ci 3.0 x 10 8 1.8 x 10 8 1.2 x 10 8 3.8 x 10 7 8.8 x 10 6 3.0 x 10 8 at T = 0 hr Heat load in PC, kW 1.6 x 10 3 4.7 x 10 2 2.4 x 10 2 6.6 x 10 1 2.3 x 10 1 1.6 x 10 3 at T = 0 hr Activity air in SC, Ci 2.0 x 10 5 8.4 x 10 5 1.4 x 10 6 6.6 x 10 5 1.1 x 10 5 1.6 x 10 6 at T = 50 hr Heat load in SC, W 1.4 x 10 3 2.8 x 10 3 2.9 x 10 3 1.1 x 10 3 2.5 x 10 2 2.9 x 10 3 at T = 24 hr Activity on HEPA, Ci 1.3 x 10 4 7.2 x 10 4 1.8 x 10 5 9.0 x 10 5 1.8 x 10 6 1.8 x 10 6 at T = 30 days Heat load of HEPA, W 4.5 x 10 1 1.1 x 10 2 3.9 x 10 2 1.7 x 10 3 3.0 x 10 3 3.0 x 10 3 at T = 30 days Activity on CF, Ci 5.7 x 10 4 2.3 x 10 5 4.5 x 10 5 9.0 x 10 5 4.8 x 10 5 9.0 x 10 5 at T = 10 days Heat of CF, W 2.6 x 10 2 8.4 x 10 2 1.2 x 10 3 1.8 x 10 3 9.6 x 10 2 1.8 x 10 3 at T = 10 days Iodine load on CF, g 1.8 x 10 0 2.1 x 10 1 6.6 x 10 1 6.6 x 10 2 1.9 x 10 3 1.9 x 10 3 at T = 30 days Key: SP = suppression pool PC = primary containment SC = secondary containment HEPA = high-efficiency particulate air filter CF = charcoal filter
TIME-r_J5SecRISETIME-BROADLINESINDICATEPERIODSWHEN104psi.,/.---104psiII1346°FMEASUREMENTSOFINSULATIONRESISTANCEI-I/3400f'CDANDCHARGINGCURRENTWEREMADE,-I-4hr.*27mln.,--75palll/3200F,-0RISETIMEJ-I-)-IJ-3hr.,38mIn.I-25pII1l/272°F--0\1-15psl1l/256OFI-V*....-24Ilr.,53mln.I--55mln-+3hr.,22min6hr.0110310029012080CIliii6701&.1a::)5enC/)1&.18:4-2palllc:......'"U00::E::x:-:x:--i0fT1fT1c:CJr-:x:-fT1:z"TlnfT1......-iz;0:x:-:x:-......;0r-n::z"Tl-i0-'.(1)Vlr-r-UJ1II:x:-......CJCc+"TlG)-SfT1:I:fT1(1)'"U-S-j-j:z:w0-<fT1.-t,;0t-'-'.):>0G)..-.--'2:-U-<I.(1):x:-0.....rn-<fT1fT1Vl;0:2:......-jVlnn10;0:::03:n1'"U'"U:x:-C>2:;0-<-j1II-'o:::s\D:::0(1)<.....en\D...... ,UJ*....IU>-UIX0*UJQ.0UJ>....Iex::>UJ*ZI0,4/I*III*I---=:.,.Tc-ue:t:>-,-Q.IXV10co1-.('0'1V1>-ex::cIUJC\.I-.--MC\.I..,..<I')IX;::)('0'1aILU-N.-r-or-ocoo......o000It)-=-MN9ISd-.3C1nSS3C1d,.----------------DUANEARNOLDENERGYCENTERIOWAELECTRICLIGHT&POWERCOMPANYUPDATEDFINALSAFETYANALYSISREPORTScheduleofTestEnvironmentsSaturatedSteamPressureCycleandLimitorqueOperationCycleFigure3.11-2.Q_}}