ML20148E667

Design Basis Analysis of Postulated Steam Generator Tube Rupture Event for Comanche Peak Steam Electric Station,Unit 1

ML20148E667
Person / Time
Site:	Comanche Peak
Issue date:	03/31/1988
From:	Boatwright W, Lo S, Maier S TEXAS UTILITIES ELECTRIC CO. (TU ELECTRIC)
To:
Shared Package
ML19302D349	List: ML20148E535 ML20148E638 ML20148E667
References
RXE-88-101-NP, NUDOCS 8803250275
Download: ML20148E667 (162)

Text

_ _

I

s. .. ._ _

i

_naw. - - - -----~~---~.~,so,~,nen_.-- --- --, .- - . - - - - . - -- --

/

I

%..--.~-----__...,.-.._, _ _ _ _

l l REACTOR ENGINEERING l l l

i l

ammmmuummmme ammmme unmumme naamme ummumme e-meme ======

.=.-

TUELECTRIC 8803250275 880315 """"'"'"

PDR ADOCK 05000445 A DCD

l 1

i RXE-88-101-NP i i

1 1

l l

I DESIGN BASIS ANALYSIS OF A POSTULATED i STEAM GENERATOR TUBE RUPTURE EVENT FOR COMANCHE PEAK STEAM ELECTRIC STATION, UNIT 1 1

March, 1988 W. James Boatwright  ;

Stephen M. Maier Sui-Sang to l

)

l I

Reviewed: D2 N i Brent L. Rice Supervisor, Transient Analysis I

1 Approved: e A s

Ausaf Husain Director, Reactor Engineering l

l l"!1UE E RI l

DISCLAIMER The information contained in this report was prepared for the specific requirement of Texas Utilities Electric Company (TUEC), and may not be appropriate for use in situations other than those for which it was specifically prepared. TUEC PROVIDES NO WARRANTY HEREUNDER, EXPRESS OR '

IMPLIED, OR STATUTORY, OF ANY KIND OR NATURE WHATSOEVER, REGARDING THIS I

REPORT OR ITS USE, INCLUDING BY NOT LIMITED TO ANY WARRANTIES ON MERCHANTABILITY OF FITNESS FOR A PARTICULAR PURPOSE.

By making this report available, TUEC does not authorize its use by others, and any such use is forbidden except with the prior written approval of TUEC.

Any such written approval shall itself be deemed to incorporate the disclaimers of liability and disclaimers of warranties provided herein. In no event shall TUEC have any liability for any incidental or consequentisi damages of any type in conInection with the use, authorized or unauthorized, of this report or the information in it, iii

ABSTRACT l l

In response to NRC requests, TU Electric has re-examined the postulated, design-basis, Steam Generator Tube Rupture Event (SGTR).

h The SCTR event is unique among the design-basis accidenes in that timely and correct operator intervention is required to terminate the event. Hence, using the CPSES control room simulator, extensive evaluations of the responses and response times required for

} termination of che postulated event were performed using CPSES reactor operators trained to follow the CPSES Emergency Response Guidelines.

The equipment required for termination of the postulated event was identified, and the safety classification of this equipment was verified.

A transient-specific RETRAN02 model of CPSES was developed in accordance with an approved Quality Assurance program. A  ;

representative SGTR analysis was simulated for comparison with a similar analysis performed for the Westinghouse Owners' Group.

Conservative initial conditions were ;dentified and included in the model. The effects of a number of postulated active failures were evaluated to determine which failure would result in the most severe transient. The active failure analyses resulted in the formulation of two design-basis scenarios. One scenario included the failure of an auxiliary feedwater throttling valve and resulted in the least margin to overfill of the ruptured steam generator. The second scenario included the failure of an atmospheric relief valve on the ruptured steam generator to close and resulted in the most severe postulated radiological consequences. I The analyses of these two scenarios demonstrated that sufficient time was available for the reactor operators to terminate the primary-to-secondary leakage from a postulated SGTR accident prior to l I

overfill of the faulted steam generator. In addition, the offsite i l

doses received at the Exclusion Area Boundary and the Low Population '

I Zone boundary were calculated to be within the limits of 10CFR100.ll.  ;

iv j l

TABLE OF CONTENTS PAGE ABSTRACT....................................................... iv TABLE OF CONTENTS.................... ......................... v LIST OF TABLES................................................. vii I

LIST OF FIGURES....... ....................................... viii CHAPTER

1. INTRODUCTION................................................... 1-1

2. USE OF RETRAN02 MOD 004 FOR SGTR ANALYSES........... ............. 2-1 i

3. THE RETRAN BASE PLANT MODEL OF CPSESl . . . . . . . . . . . . . . . . . . . . . . . . . . 3- 1

4. STEAM GENERATOR TUBE RUPTURE SCENARIO DESCRIPTION.............. 4-1

5. STEAM GENERATOR TUBE RUPTURE BASE MODEL DESCRIPTION............ 51 5.1 Genera 1.................................................. 51 5.2 Physical Description of the CPSESl-SGTR Model............ 5-2 5.3 Single-Node Steam Generator Mode 1......................... 5-3 5.4 Steam Generator Narrow Range Level Indication............. 5-5 5.5 Break Flow Model......................................... 5-6 5.6 Model 0ptions............................................, 5-8 5.7 Reactor Trip System...................................... 58 5.8 Point Kinetics Reactivity Parameters...................... 5-8 f.9 Steam Generator ARV Control System....................... 5-9 5.10 Boron Transport........................................... 5-10 5.11 Summary of Conservatisms in the SGTR Nominal Base Model... 5-10

6. JUSTIFICATION OF THE ADEQUACY OF THE CPSESl-SGTR MODEL......... 6-1 6.1 Nominal Casa Initial Conditions / Equipment Availability... 62 6.2 Operator Response Times.................................. 63 6.3 Analysis Results.......................................... 6-5 6.4 S u mm a ry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-20

7. OPERATOR RESPONSE TIME EVALUATION............................... 71

8. EVALUATIONS OF CONSERVATIVE ASSUMPTIONS........................ 8-1 8.1 Offsite Power Availability............................... 8-1 8.2 Timing of the Reactor Trip............................... 8-2 8.3 Initial Thermal-Hydraulic conditions..................... 83 v

i l

l i

l TABLE OF CONTENTS (Cont.)  !

l 1

CRAFTER PAGE l

8.4 Protection System Setpoints and Delay Times.............. 8-9 l 8.5 System capacities........................................ 8-12 ,

8.6 Control Systems.......................................... 8-13 l 8.7 Decay Heat................................................ 8-14 )

8.8 Summary.................................................. 8-15 j 1

9. REQUIRED EQUIPMENT AND ACTIVE FAILURE EVALUATIONS............... 9-1 1

9.1 Auxilia ry Fe edva ter Throt tling Valve . . . . . . . . . . . . . . . . . . . . . 9 -1 l 9.2 Main Feedwater Isolation Valve........................... 9-2 9.3 Miscellaneous Secondary Steam Isolation Valves............ 9-3 9.4 Steam Generator Safety Valves............................ 9-3 9.5 Steam Generator Atmospheric Relief Valves (ARVs).......... 9-4 l 9.6 Main Steam Isolation Va1ves............................... 9-5 9.7 Pressurizer Power-operated Relief Valve (PORV)............ 9-5 9.8 Pressurizer Safety Valve................................. 9-6 I 9.9 ECCS Pump Stop Switches................................... 9-6 9.10 SI Reset Device........................................... 9-6 9.11 Electrical Failures..................... ................. 9-6 9.12 Instrumentation Failures................................. 97 9.13 Summary.................................................. 9-7

10. DESIGN-BASIS SGTR EVENT FOR THE WORST CASE FOR STEAM GENERATOR 0VERFILL................................... 10-1 10.1 Operator Actions and Response Times....................... 10-1 ,

10.2 Auxiliary Feedwater Flow Rates............................ 10-3 I 10.3 Results................................................... 10-4 l 10.4 Summary................................................... 10-15 l 1

11. DESIGN-BASIS SCTR FOR THE WORST CASE RADIOLOGICAL DOSES. . . . . . . . 11-1 11.1 Conservative Assumptions with Respect to Radiological Doses......................................... 11-1 j 11.2 Operator Actions and Response Times....................... 11-2 11.3 Results................................................... 11 6 l 11.4 Thermal H.ydraulic Input for Dose Consequences Calculations.............................................. 11-26 1

11.5 Summary................................................... 11-27 i 12. RADIOLOGICAL DOSE ASSESSMENT................................... 12-1

. 13. CONCLUSIONS.................................................... 13-1 1

4

14. REFERENCES..................................................... 14-1 vi 1

3

.. ~. . -. - .

I LIST OF TABLES TABLE PAGE 6-1 Nominal Base Case - Assumptions and Plant Differences Between the Reference Plant and CPSESl....................... 6-21 l 6-2 Nominal Base Case - Sequence of Events....................... 6-22 l 6-3 Nominal Base Case - Key to Figures 6-1 through 6-7........... 6-23 1 7-1 SGTR Operator Response Times from Simulator Exercises........ 7-5 7-2 Conservative Operator Response Times Assumed for the CPSES1 Design-Basis SGTR Analyses.................... 7-6 8-1 Summary of Conservative Assumptions.......................... 8-16 10 1 Design Basis SGTR - Steam Generator Overfill Event Timeline............................................... 10-17 1

10-2 Design-Basis SGTR - Steam Generator Overfill l Key to Figures 10-1 through 10-10............................ 10-18 l 11-1 Design Basis SGTR a Radiological Consequences Event Timeline................................................ 11-29 11-2 Design-Basis SGTR - Radiological Consequences Summary of Integral Mass Releases............................ 11-30 11-3 Design-Basis SGTR - Radiological Consequences Key To Figure 11-1 through 11-12............................. 11-31 12-1 Radiological Dose Consequences of Postulated SGTR............. 12-4 vii

LIST OF FIGURES FIGURE PAGE 1

5-1 CPSESl-SGTR Model Flow Chart................................. 5-12 52 CPSESl-SGTR Noding Diagram................................... 5-13 6-1 Nominal Base Case Comparison - Pressurizer Pressure..... ... 6-24 6-2 Nominal Base Case Comparison - Pressurizer Level............ 6-25 6-3 Nominal Base Case Comparison - Intact Loop Temperatures..... 6-26 6-4 Nominal Base Case Comparison - SGTR Break Flow Rate.......... 6-27 6-5 Uominal Base Case Comparison - Steam Generator Levels....... 6 28 6-6 Nominal Base Case Comparison - Steam Generator Pressures.... 6-29  !

6-7 Nominal Base Case Comparison - Ruptured SG Liquid Volume..... 6 30 10-1 CPSES1 Design-Basis SGTR, Overfill - Pressurizer Pressure... 10-19 10-2 CPSES1 Design-Basis SGTR, Overfill - Pressurizer Level..... 10-19 10-3 CPSES1 Design-Basis SGTR, Overfill - SG Pressures........... 10-20 10-4 CPSES1 Design-Basis SGTR, Overfill - SG Levels............. 10-20 10-5 CPSES1 Design-Basis SGTR, Overfill - SGTR Break Flow Rate... 10-21 10 6 CPSESl Design-Basis SGTR, Overfill - Przr and SG Pressures. 10-21 10-7 CPSES1 Design-Basis SGTR, Overfill - Aux FV Flow Rates..... 10-22 10-8 CPSESl Design-Basis SGTR, Overfill - RCS Temperatures...... 10-22 10-9 CPSES1 Design-Basis SGTR, Overfill - RCS Loop Flow Rates.... 10-23 10-10 CPSES1 Design-Basis SGTR, Overfill - Ruptured SG Volume..... 10-23 11-1 CPSESl Design-Basis SGTR, Doses - Pressurizer Pressure..... 11-32 11-2 CPSE91 Design-Basis SGTR, Doses - Pressurizer Level........ 11-32 11-3 CPSES1 Design-Basis SGTR, Doses - RCS Temperatures......... 11 33 11-4 CPSES1 Design-Basis SGTR, Doses - Steam Genrator Pressures., 11-33

\

viii i

LIST OF FIGURES (Cont.)

l FIGURE PAGE 11-5 CPSES1 Design Basis SGTR, Doses - Steam Genrator Levels.... 11 34 11-6 CPSES1 Design-Basis SGTR, Doses - SCTR Break Flow Race...... 11-34 11-/ CPSES1 Design-Basis SGTR, Doses - Przr and SG Pressures..... 11-35 11-8 CPSES1 Design-Basis SGTR, Doses - Ruptured SG Liquid Volume............................................... 11-35 11-9 CPSES1 Design-Basis SGTR, Doses - Integrated Steam Releases to Condenser................................ 11-36 11-10 CPSES1 Design-Basis SGTR. Doses - Integrated Steam Releases to Atmosphere................................ 11-36 11-11 CPSES1 Desi.n-Basis SGTR, Doses - Integrated Break Flow..... 11-37 11-12 CPSE31 Design-Basis SGTR, Doses - Hot Leg Flashing Fraction.......................................... 11 37 ix

I CHAPTER 1 i

INTRODUCTION l

i The evaluation of the safety of a nuclear power plant includes analyses of the plant response to postulated disturbances in process variables and to postulated malfunctions or failures of equipment. One such postulated failure which is included in the safety evaluation of Comanche Peak Steam Electric Station, Unit 1 (CPSES1) is the complete severance of a single steam generator U-tube. The postulated Steam I t

Generator Tube Rupture (SGTR) event is classified as an ANS '

Condition IV fault - an occurrence that is not expected to take place, but is postulated because of the potential release of significant amounts of radioactive material.

f The Steam Generator Tube Rupture analysis currently in the Final Safety Analysis Report (FSAR) of the Comanche Peak Steam Electric Station is based on the assumption that the primary to-secondary leakage to the d

ruptured steam generator can be terminated by appropriate operator actions within 30 minutos. However. SGTR events experienced by two operating plants (Ginna and Prairie Island) required a significantly  ;

longer period of time to terminate the break flow rate by equalizing

, the primary and secondary system pressures. If the

]

primary-to-secondary leakage from a SGTR event were to persist over a

' )

sufficiently long period of time, the liquid inventory in the ruptured 1 i

steam generator would also increase. The subsequent overfill of the j ruptured steam generator into the main steamlines could threaten the integrity of the steamlines and the associated supports and could j result in increased releases of radioactive material.

I j In 1984, the Nuclear Regulatory Commission (NRC) requested additional

information from TU Electric concerning the SGTR analysis [1]. The following information was requested

J l

j 1-1 i  ;

4

l a) an evaluation of operator actions necessary to effect pressure equalization, and a conservative time estimate for each action, including initial delay time; b) an evaluation of whether liquid can enter the main steamlines as a result of a tube rupture, and the consequential effects on the integrity of the steam piping and supports; and, c) verification that all components credited in the analysis to mitigate the consequences of the SGTR event are classified as safety related, or justification fcr the use of any non-safety related component credited in the analysis.

In response to these requests by the NRC, TU Electric elected to participate in the Westinghouse Owners Group (WOG), SGTR subgroup program which attempted to address these concerns on a generic basis.

Extensive analyses of the revised SGTR scenario, based on a "bounding" reference plant design, were submitted to the NRC [2, 3, 4). The NRC issued two Safety Evaluation Reports (SERs), accepting the analysis approach and generic conclusions (5, 6). However, because of the numerous differences between plant designs and equipment, operating procedures, and operator training, the NRC required each utility to perform plant-specific analyses of the postulated SGTR event.

TU Electric has opted to perform the additional analyses required for resolution of the SGTR issue in house. This report describes the plant specific, design-basis analyses of the postulated Steam Generator Tube Rupture Event performed by TU Electric for the Comanche Peak Steam Electric Station. Unit 1.

The RETRAN02 MOD 004 [7] computer code was used with a detailed model of CPSES1 to simulate the postulated SGTR event. With this model, a "Nominal" SGTR scenario was executed for comparison with a similar analysis performed by the WOG, SGTR subgroup, which was based on the reference plant design. This comparison provided sufficient assurance that all key parameters were addressed in the plant-specific analysis, 1-2

in addition to providing qualitative assurance of the adequacy of the CPSES1-SGTR model.

An evaluation of the CPSES1 operator responses to the postulated event and the timing of those responses was performed using the results of numerous simulator exercises. In addition, these simulator exercises facilitated the identification of the plant equipment required for the termination of the postulated event.

Using a plant specific, "Nominal" model of CPSES1 and the plant-specific operator response times, evaluations were performed to identify conservative assumptions concerning equipment availability and initial conditions. Af ter incorporating these conservative assumptions into a "Conservative" RETRAN02 model of CPSES1, evaluations were

, performed to identify the single active failure which results in the most severe event with respect to the possibility of overfilling the ruptured steam generator and the single active failure which results in the most severe offsite radiological dose consequences. These evaluations resulted in the identification of the design-basis event scenarios.

Detailed thermal-hydraulic descriptions of the two design basis steam generator tube rupture events are provided in this report as well as a discussion of the radiological consequence assessment Athodology and the calculated radiological results.

The discussion of the RETRAN02 MOD 004 computer code and its adequacy for 4

use in the analysis of SGTR events Lay be found in Chapter 2. The RETRAN02 base plant model of CPSESl is described in Chapter 3. The SGTR scenario, including a discussion of the operator responses required to terminate the postulated event, is contained in Chapter 4.

Chapter 5 contains a description of the SGTR-specific RETRAN02 model of CPSES1, followed by the justification of the adequacy of this model in Chapter 6. A detailed discussion of the operator responses to a postulated SGTR and the timing of those responses may be found in Chapter 7. Chapter 8 contains the evaluation of tha conservative 1-3 J

I 4

1 l

l l

, I initial conditions incorporated in the analyeis of the SGTR event. The  ;

equipment required to be available to the operators to terminate the postulated SGTR event is discussed in Chapter 9, as are the effects of potential failures of this equipment.

The analysis of the design-basis SGTR event which minimizes the margin to overfill of the ruptured steam generator is presented in Chapter 10. The thermal-hydraulic analysis of the design-basis S3TR event which results in the most severe radiological consequences is presented in Chapter 11, followed by an assessment of the radiological consequences in Chapter 12. Finally, the conclusions drawn from thes; analyses are described in Chapter 13.

i i

t I

i i

l 4

1 1

1-4 -

l I

t CHAPTER 2 USE OF RETRAN02 MOD 004 FOR SGTR ANALYSES The reanalysis of the design basis SCTR was performed with the )

RETRAN02 MOD 004 computer code. The RETR /.N code was developed by Energy, Inc. (ne" El Services) under the sponsorship of tha Electric Power Research Institute (P RI). RETRAN02 is a one-dimensional, best estimate, transient, thermal-hydraulic computer code, capable of simulating most boiling water and ptcssurised m ' ?setor transients, exclusive of those events where the primary cool < ' stem becomes heavily voided [7]. RETRAN02 is widely 2 sed the .it the nuclear industry by over 25 domestic and foreign licensee:.

l In 1981, the Utility Group for Regulatory Action (UCRA) submitted RETRAN02 MOD 002 to the NRC for review. The NRC issued a Safety Evaluation Report (SER) in 1984, conditionally approving l

RETRAN02 MOD 002 with error corrections, for use in performing transient

{

analyses of non-Appendix K (non-LOCA) events (8).

l The NRC approved RETRAN02 for use in the analysis of non-14CA Transients. Among the transients specifically approved was the steam generator tube rupture (SGTR) event. The capability of RETRAN02 to adequately simulate SGTR transients has been demonstrated by analyces performed by the staffs of the Institute of Nuclear Power Operations (INFO) [9] and the EPRI-sponsored Nuclear Safety Analysis Center (NSAC)

(10). The Prairie Icland and Ginna tube rupture events were extensively analyzed using RETRAN02. These analyses were "best-estimate" analyses, i.e., no conservative assumptions were deliberately used. Thess analyses demonstrated that RETRAN02 may be 4

used to predict transient system responses which compare favorably with recorded plant responses. As concluded in a NSAC report (11]

describing parametric studies performed with the RETRAN02 model used to benchmark the Prairie Island event [10), "... RETRAN is a capable thermal-hydraulic code for steam generator tube rupture analysis."

2-1

l l

l In 1985, RETRAN02 MOD 003 was released by EPRI. RETRAN02 MOD 003 contained the corrections to the errors identified subsequent to the release of. 4 RETRAN02 MOD 002, including those identified during the NRC review. The current version of RETRAN02 is RETRAN02 MOD 004. RETRAN02 MOD 004 contains error corrections to RETRAN02 MOD 003 plus two new models. These error corrections and model enhancements were made in accordance with the Quality Assurance programs of EPRI and EI Services. One of the two new models is an improved control rod model for use with 1-D kinetics. The 1-D kinetics option was not used by TU Electric in SGTR analyses. The second model, which was not used by TU Electric in SGTR analyses, is an option to allow the computer code to edit the initial values of the control blocks. Neither the use nor the omission of this option affects the SGTR analyses. .

RETRAN02 MOD 004 is the most recently released version of REIRANO2 and contains corrections to those code errors identified prior to the time -

of its release. It is appropriate that this version of the code be used for the CPSESl simulation of the postulated SGTR event.

2-2

t CHAPTER 3 THE RETRAN BASE PLANT MODEL OF CPSES1 The Comanche Peak Steam Electric Station consists of two Westinghouse Pressurized Water Reactors. The analyses presented in this report are

, applicable only to Unit 1 (CPSES1).

CPSES1 is a four-loop plant with a rated thermal power of 3411 MWt.

The steam generators are the Westinghouse Model D-4. With the exception of the Nitrogen-16 (N-16) temperature synthesis control system, CPSESl does not have any features not found in similar vintage Westinghouse four-loop PWRs. The N-16 system synthesizes an average Reactor Coolant System (RCS) temperature baaed on the measured cold leg temperature and a normalized N-16 power measurement. The N-16 system is used in lieu of the more standard Delta-Temperature systems employed at similar Westinghouse plants.

TU Electric has adopted a "Base Model" approach to control the computer l models used for transient analyses. The Base Model consists of a detailed, best-estimate, CPSES1 plant model, including all NSSS-related control systems and most of the balance-of-plant control systems. The Base Model was developed, approved, and controlled in accordance with the Quality Assurance program for the Reactor Engineering Department.

All transient-specific models, either licensing or best-estimate, are derived from the approved Base Model in accordance with the procedures outlined in the Quality Assurance Program. Hence, a discussion of the specific model for the SCTR analysis requires a discussion of the best-estimate Base Model.

A quality-assured, two-loop, best-estimate model of CPSES1 was developed and issued by TU Electric in early 1985. This model was developed from as-built drawings, plant-specific design documents, etc. The model has been used for a variety of analyses, both 3-1

best-e.timate and quasi-best-estimate, in support of operational concerns at the plant. These analyses provided opportunities to extensively examine the qualitative performance of all portions of the model.

Among the analyses performed were several Emergency Drill scenarios, which included SGTRs in conjunction with other severe events. Other activities using the CPSESl model included the performance of a long-term, natural circulation, plant cooldovr. analysis and analyses to l support the optimization of several non-safety-related, plant control i

system parameters. The RETRAN02 model of.CPSES1 was also used to l simulate eleven transients, including a Steam Generator Tube Rupture, for use as a basis for evaluating the adequacy of the CPSES1 plant I 1

simulator.

The RETRAN02, CPSESL model was the subject of a Technical Review performed in 1985 by the RETRAN code developer, EI Services. The purpose of the review was to ensure that the raw, plant design data had l been correctly interpreted and formulated for proper use in RETRAN02, and that appropriate code options and models were selected. In addition, four transients were simulated with this original base plant model to test the model as an integral system. These transients were also reviewed by EI Services to qualitatively evaluate the accuracy of the results. EI Services based their evaluation of the transient results on numerous internal analyses performed for similar plants.

The current, best-estimate, plant model was issued in mid-1987 and addressed all concerns identified in the Technical Review performed by EI Services. In addition, many of the component models were improved, and several of the control systems were enhanced or added to the base model. The quality of the documentation describing the model was also significantly upgraded. The Steam Generator Tube Rupture model was derived from this Base Model of CPSESl.

3-2

i l

l CHAPTER 4 l

l STEAM GENERATOR TUBE RUPTURE SCENARIO DESCRIPTION i

Prior to discussing the SGTR-specific model, it is important to describe the postulated SGTR scenario. The following discussion outlines the "routine" scenario and operator .esponses with no active failures assumed. l The postulated, design-basis Steam Generator Tube Rupture event is initiated by a complete severance of a single steam generator U-tube.

The initial indications of the event are decreasing pressurizer pressure and level and a steam flow - feed flow mismatch. Other indications include increased charging flow, slight transients in steam generator water level indications, and a turbine runback as the RCS pressure reduction forces the overtemperature N-16 setpoint to within 1 3% of the calculated Overtemperature N-16 signal. In addition, radiation monitors on the main steamlines, steam generator blowdown header, and condenser offgas system will detect increased radiation levels, and an alarm will sound, providing an early, unique indication of a steam generator tube rupture. In accordance with Abnormal Conditions Procedures (12], following receipt of an alarm on 5% pressurizer level deviation, the reactor operators will isolate the 45 gpm letdown orifice, and start the first and then the second centrifugal charging pump. If the pressurizer level continues to fall, the reactor operators will manually actuate Safety Injection, thus causing a reactor trip, turbine trip, and auxiliary feedwater initiation.

Immediately following reactor trip, the Shift Supervisor will open the Emergency Response Guidelines (ERGS) (13). These procedures provide specific guidance to the reactor operators for the diagnosis of the cause of the reactor trip and the stabilization of the plant. After the event is diagnosed, transient-specific Emergency Operating 4-1

Procedures (EOPs) are provided within the ERGS to mitigate the effects i

of the transient and to stabilize the plant. <

l Following reactor trip, the event would be diagnosed as a SGTR based on i a marked difference in narrow range level indications between the intact and ruptured steam generators. The narrow range level I indication is provided by a safety-grade system. Other non-safety systems available for use in diagnosing the tube rupture include the radiation monitors described above; however, because the radiation monitors are not safety-grade components, credit was not taken in licensing-basis analyses for the proper operation of these monitors.

Following the diagnosis of the event as a SGTR and the identification of the ruptured steam generator, the E0Ps direct the operators to isolate the ruptured steam generator. This isolation would be achieved by throttling auxiliary feedwater to the ruptured steam generator, closing the main steam isolation valve (MSIV) on the steamline from the ruptured steam _ generator, and raising the setpoint pressure of the power-operated, atmospheric relief valve (ARV) on the steamline from I the ruptured steam generator. The setpoint pressure of the ARV would be raised, but would remain below the lowest setpoint pressure of any of the main steam safety valves. Procedural contingency actions are provided if any of the above actions can not be completed.

Upon isolation of the ruptured steam generator, the RCS would be cooled l

such that a subcooling margin would be maintained when the RCS is l

depressurized. If the steam dump (turbine bypass) valves were available, they would be used to cool the RCS, If offsite power were unavailable, the air-operated steam dump valves would not be available; hence, the RCS would be cooled by relieving steam through the ARVs on the intact steam generators. The target temperature for the RCS cooldown corresponds to a specified subcooling margin at the pressure of the ruptured steam generator.

Once the RCS had been cooled, and the subcooling margin at the ruptured steam generator pressure had been obtained, the RCS would be 4-2

depressurized to terminate the break flow. If the reactor coolant pumps were operating, the pressurizer sprays vould be used to conserve RCS inventory. If offsite power were unavailable, the RCPs would also be unavailable; thus, one pressurizer power-operated relief valve (PORV) would be used to depressurize the system. The depressurization would be terminated when the RCS pressure was less than the pressure in the ruptured steam generator, when a maximum pressurizer level was exceeded, or when the RCS subcooling margin limit was approached.

The final significant operator action required to terminate the event is to stop the Emergency Core Cooling System (ECCS) pumps. When the RCS is at low pressures, the capacity of these pumps is relatively high; thus, the RCS would repressurize quickly if they were not stopped.

After the ECCS pumps are stopped, the RCS and ruptured steam generator pressures would become equal due to flow through the ruptured tube, thereby removing the driving force behind the leakage flow and terminating the break flow. Additional operator actions, discussed below, would be required to prevent the primary-to-secondary break flow from becoming reestablished while taking the plant to cold shutdown conditions.

1 Immediately following the termination of the ECCS flow, the RCS pressure would increase slowly due to the addition of decay heat and the lack of any active heat removal process (no steam relief from l intact steam generators); hence, the break flow rate would tend to be reestablished. Additional procedural guidance has been provided to the reactor operators to prevent the flow of a significant amount of RCS fluid into the ruptured steam generator.

Following the termination of the ECCS flow, the reactor operators would continuously maintain equal RCS and ruptured steam generator pressures, using the pressurizer sprays, a pressurizer PORV, or the auxiliary spray if letdown has been reestablished. Following the initiation of a controlled coold7vn to bring the plant to cold shutdown conditions, the 4-3

RCS would be maintained at a pressure lower than the ruptured steam generator pressure, thus backfilling the RCS with the ruptured steam generator fluid. This process would continue as the plant is cooled to cold shutdown conditions, thus preventing the liquid volume in the ruptured steam generator from increasing significantly at any time during the recovery following the event.

Because the steam generstors have a finite volume, the break flow must be terminated in a timely manner to prevent overfill of the ruptured steam generator. Overfill of the ruptured steam generator could lead to filling of the main steamlines with liquid, thereby increasing the likelihood of several consequential events. If the steamlines are struccurally adequate, the filling and pressurization of the steamline could result in opening of a main steam safety valve. Because these valves are not designed for water-relief, it may be postulated that these valves would fail to close, resulting in a prolonged release directly to the atmosphere. If the steamlines are not structurally adequate, a consequential main steamline break may also be postulated.

In either case, the overfill of the ruptured steam generator could lead to unacceptable consequences, and hence, must be avoided.

4-4

I

\

\

CHAPTER 5 STEAM GENERATOR TUBE RUPTURE BASE MODEL DESCRIPTION 5.1 General As indicated in Chapter 3, the RETRAN02 models used for the simulation of the design-basis ste- generatar tube rupture event were derived from the plant-specific, best-estimate model of CPSESl. During the development of the final SGTR model, several "intermediate" models were developed for use in the evaluation of conservative assumptions and effects from postulated active failures. The use of the intermediate  !

models in the development of the design-basis models is illustrated in j Figure 5-1. j l

In the following discussions, the "Nominal Base Case" is that model l derived directly from the best-estimate model. Few conservatisms were incorporated into this model. The Nominal Base Case was used to simulate the postulated transient using a similar set of initial conditions and assumptions as was used for the generic tube rupture analysis performed by the Westinghouse Owners Group [2); a comparison of the two analyses is presented in Chapter 6. In addition, this Nominal Base Case was used to identify critical parameters and to determine conservative values for these parameters to be used in the design-basis SGTR analyses.

The set of conservative assumptions was then incorporated into the Nominal Base Case to derive the "Conservative Base Model". The I Conservative Base Model contains all of the conservative assumptions to be included in the design-basis models, with the exception of the single active failures. The Conservative Base Model was then used to evaluate the effects of the postulated active failures to determine which single failure would result in the least margin to overfill of the ruptured steam generator and which single failure would result in the most severe radiological consequences.

5-1

5.2 Physical Descriotion of the CPSES1-SGTR Model The noding of the CPSES1 steam generator tube rupture model (Figure 5-2) is similar to the noding described in NSAC 47 -[10),

wherein the Prairie Island event was analyzed. The best-estimate models of the Prairie Island [11] and Ginna plants [9] used multi-node steam generator models. The use of a single node steam generator in the CPSES1 model will be addressed in Section 5.3.

In addition to a more detailed steam generator secondary model, the best-estimate model of Ginna employed greater detail in the upper head region of the reactor vessel. The detail in the reactor vessel upper head of the Ginna RETRAN model was considered necessary for the detection of suspected void formation in the upper head region.

Because anal;n12 *ndicated "la upper head voiding occurs during the coural of the CPSESl recovery simulations, a single volume representation of the upper head was determined to be adequate.

The Nominal Base Cass Steam Generator Tube Rupture model for CPSES1 contained 48 fluid volumes and 80 junctions. Plant loops designated as Loop 1, 2, and 3 were combined into model Loop 1. Plant Loop 4 was modeled separately and included a non-equilibrium pressurizer model.

The hot legs, intermediate legs, reactor coolant pumps, and cold legs were each represented as single volumes. In the reactor vessel, the upper head, upper plenum, downcomer, lower head, and core bypass regions were individually modeled as single volumes. The fluid region of the core was modeled as three equi-volume regions.

On the primary side of the steam generators, the inlet and outlet plena were each represented by single volumes. The U-tube region was modeled with six fluid volumes per modeled steam generator. The secondary sides of the steam generators were modeled with a single separated fluid volume. The main steamlines and the main steam header were explicitly modeled.

5-2

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

l l

l 1

1 The model included 38 heat conductors; three for the core region, six j in each steam generator U-tube region, and 21 to represent the thermal storage capacity of the thick metal masses of the RCS vessel and piping, the pressurizer walls, and the steam generator vessel walls.

l To maintain conservatively high primary temperatures and pressures, and j hence, a high primary-to-secondary break flow, no heat transfer to the containment though the RCS piping, ressel valls, pressurizer valls, or through the steam generators valls was modeled.

I 5.3 Single-Noda Steam Generator Model l

I A single-node steam generator model is adequate for relatively long, slow transients where the steam generator mass and pressure are the l important parameters and for transients where there is little time for i significant feedback from the secondary to the core. For these types of transients, the use of a single-node steam generator model can provide similar results to those obtained with a multi-node model and with a significant savings of computer resources. The main drawback to l using the single-node steam generator is its inability to allow direct prediction of the indicated water level.

The level indication in a U-tube steam generator is greatly influenced by the "manometer" effects between the tube bundle region and the downcomer. In a single-node steam generator model, these regions are lumped with other parts of the steam generator into a single region; thus, no detail is available from which the level indication may be directly predicted. However, there are parameters upon which a level prediction may be derived.

Derivation of a level indication from the steam generator mixture l volume is the principal mechanism used. This mechanism is particularly I useful when the void fraction is small, and the steam generatot circulation ratio is small. For these conditions, the manometer effect between the downcomer and tube bundle regions is small and the mixture 5-3

I volume allows a reasonable prediction of the steam generator level. l The level indication from the mixture volume is approximate when there I is a rapid change in the void fraction of the mixture due to a change in the heat removal or addition rates to the mixture. I In the single-node steam generator model, the preheater section is not explicitly modeled; therefore, the overall calculated heat transfer coefficients must be adjusted in order to transfer the correct amount of heat from the RGS to the steam generator. This adjustment was made in RETRAN by adjusting the heat transfer area of the U-tubes to remove the correct amount of heat. The adjustment is reasonable during steady or slowly changing flow conditions, but less reasonable during rapidly changing flow conditions when the preheater may be bypassed by a different flow fraction.

Use of a single-node steam generator for the SGTR analysis is conservative in that the pressure difference used to determine the break flow is based on the average steam generator pressure. For an l actual cold leg, tubesheet surface, SGTR, the local steam generator pressure would be greater than the average pressure. The higher local pressure is due in part to the geometry of the steam generator flow distribution plates and in part to the overall static head of the steam l 1

generator fluid. The fluid at the cold leg tubesheet surface consists primarily of relatively dense, main feedvater injected into the preheater box and slightly subcooled recirculation flow. After the main feedwater flow is isolated, the importance of this effect is reduced.

The transient thermal-hydraulic characteristics of the single-node steam generator following the injection of cold auxiliary feedwater also conservatively affect the postulated SGTR event. Due to the simplicity of the model, the auxiliary feedwater is injected directly into the steam generator boiling region, thus causing a rapid collapse of any voids. In turn, the collapse of the voids results in a lower steam generator pressure. Because the primary-to-secondary brenk flow 5-4

i increases as the primary-to-secondary pressure difference increases, the rapid void collapse is conservative for this event.

5.4 Steam Generator Narrow Rance Level Indication At CPSES1, the Narrow Range Level Indication is derived from a differential pressure transmitter; hence, the collapsed liquid volume is the most reasonable fluid volume parameter to be used for the level calculation. The appropriate collapsed liquid volume to use is that volume between the lower and upper level taps. However, this region is lumped with the rest of the steam generator, it is not possible to obtain only the liquid volume between the level taps and outside the primary separators.

The use of the mixture volume for the level indication is the other alternative. Use of the mixture volume provides a good indication of the steam generator fluid shrink and swell, if the change in the level er heat transfer rate is relatively slow. However, if there is a rapid change, e.g., a reactor trip, the use of the mixture volume does not compensate for the lack of detail in the single-node model; thus, the change in the level is overpredicted. Following reactor trip, when the void fraction in the steam generator is relatively small, use of the mixture volume and the collapsed liquid volume provide essentially the same results. The mixture volume reflects the shrink and swell effects due to pressure changes, and thus, given the coarseness of the steam generator model, is the preferred parameter upon which to base the single-node steam generator level calculation.

Based upon drawings of the Model D-4 steam generators at CPSES1, a table of the indicated steam generator water level as a function of mixture volume was developed for use in calculating an approximate narrow range level. Due to the coarseness of the steam generator model, the dynamic characteristics of the level indication are not expected to be extremely accurate; however, the trends and relative 5-5 l

e .

changes between steam generators are expected to be reasonably indicative of actual level behavior.

5.5 Break Flow Model As will be discussed more thoroughly below, the primary-to-secondary pressure difference and the fluid density have the most impact on the break flow rate. Even though the primary-to-secondary pressure difference is less on the cold leg side of the steam generator than on the hot leg side, the increased fluid density on the cold leg side results in the greater break flow rate. Thus, the SGTR break location which results in the greatest flow is on the cold leg side of the steam generator.

The RETRAN control systems and positive and negative fills (leakage paths) were used to calculate the flow rate through the ruptured U-tube. The mass flow rates from both sides of the tube were conservatively calculated based on the steam generator outlet plenum I enthalpy and pressure. The mass removed from the steam generator inlet i

and outlet plena was added to the steam generator fluid volume through l a positive fill. The energy removed from both sides of the ruptured tube was averaged (esing flow rate weighting) and added to the steam generator volume through the positive fill.

l Because flow out of the "long" tube will be resistance-limited, the shorter the tube, the greater the flow rate. Therefore, the shortest U-tube was assumed to break. In order to maximize flow out of the "short" tube, the break was assumed to occur at the surface of the steam generator tube sheet. Because the frictional resistance to flow l for the short tube is relatively small, flow out of the short tube may initially be choked.

Because actual choked flow conditions are not expected to exist for significant periods of time, the resistance-limited flow equation (Equation 5-1) was conservatively utilized to calculate the break 5-6

I l

flow rate. The friction factor and form loss coefficients which were used resulted in a very conservative calculation of the l resistance-limited flow.

l V =

([2ge.p.AT .144.dP] + (KT + fL/D])h (5-1) where: W = mass flow rate out one end of the. tube; ge - gravitational constant; p= fluid density; A

7

= cross-sectional flow area of the tube; dP -

primary-to-secondary pressure differential; K = total form loss coefficient for tube entrance T

and exit; and,  !

fL/D = frictional loss term.

l Use of only the resistance-limited flow equation simplified the model because the choked flow rate did not have to be calculated and compared i 1

with the resistance-limited flow rate in order to obtain the break flow i rate. In addition, because choked flow was not credited, use of the resistance-limited flow rate resulted in a conservative calculation of the break fivv rate. Note also that choking was not allowed to occur at the entrances to the U-tubes, nor at the break locations. In l addition, the tube ends were assumed to discharge fluid into an "infinite" plenum; thus the local pressure at the break location was minimized. As previously discussed, no credit was taken for flow distribution plates in the steam generator, or for the higher preheater pressure, both of which would have increased the local pressure.

From Equation (5-1), it is noted that smaller values of form losses and frictional losses result in larger flow rates. The form losses, K ,

7 consist of the entrance, exit, and direction change losses for the long tube section and entrance and exit losses for the short section. For conservatism, only entrance and exit form losses were modeled, using sharp-edged contraction / ext ansion geometries. To provide an added measure of conservatism, the exit loss coefficient was reduced by 15%

5-7

l l

to bound those postulated cases where the break geometry cocid be I slightly less restrictive to flow.

Because L/D is fixed by geometry considerations, the frictional losses were minimized by using a conservatively small friction factor, f. The friction factor was calculated for a smooth pipe and a conservatively large Reynolds number. The calculated value was then reduced for .

conservatism.

5.6 Model Ootions The Combined Forced / Natural Convection Heat Transfer Map Option selected in the base plant model was also used in the SGTR base model.

The slip model was not used for the SGTR licensing-basis analysis because the effects of a slip model in the steam generators would not be evident with a single-node steam generator model, i

5.7 Reactor Trio System A reactor trip signal was assumed to be generated only on low pressurizer pressure. The validity of this assumption will be discussed in Chapter 8.

5.8 Point Kinetics Reactivity Parameters Licensing-basis, End-of-Life (EOL) reactivity parameters were used as input to the point kinetics model used in the CPSESl-SGTR model. EOL i l'

reactivity parameters were chosen because the moderator density coefficient is most positive (moderator temperature coefficient is more negative) than at Beginning-of-Life. A "most positive" moderator density coefficient retards the cooldown of the primary system. ,

1 Because the SGTR is not an overpower . overpressurization , or a criticality-restricted event, this event is rather insensitive to the l

5-8 I

i 1

reactivity parameters selected. Furthermore, a multiplier of 1.2 was used to provide a conservative value for the decay heat assumed in the analysis (120% of 1971 ANS Decay Heat curves).

A relatively large (absolute) Doppler defect was conservatively utilized in the SGTR analysis in order to maximize the positive reactivity during the cooldown. I The time dependent scram reactivity worth was based on a 4% dk/k scram worth, a conservative 3.30 second rod drop time, EOL reactivity parameters, and a conservative value for rod worth as a function of rod position. The 4% dk/k scram worth is conservative with respect to the total rod worth less the most reactive rod, which was assumed to be l stuck out of the core. l The moderator density coefficient for CPSES1 will be positive at I operating conditions; hence, if the moderator density decreases, the reactivity will decrease. When selecting an appropriate MDC, the goal is to cause the primary pressure and temperature to be maximized during the depressurization, hence a MDC should be selected which results in the greatest addition of positive reactivity as the moderator density increases. Therefore, a conservative moderator density coefficient of 0.43 (delta-p)/(delta-gm/cc) was used.

5.9 Steam Generator ARV Control System The control system for the steam generator ARV was also modified for the SGTR model. It is conservative in the SGTR event for the steam generator ARV to function normally to maintain steam generator pressure below the setpoint pressure of the main steam safety valves following a reactor / turbine trip. Thus, the ARV control system for the ruptured steam generator was allowed to function normally.

In order to maximize the RCS temperature, the ARVs in the intact steam generators was placed in the manual mode of control, and the pressure 5-9

in the intact steam generators was allowed to rise toward the setpoint pressure of-the main steam safety valves. Later in the transient, the ARVs on the intact loops were used in the acnual mode to effect the RCS cooldown.

5.10 Boron Transoort Following the initiation of the ECCS flow, borated water'is injected into each cold leg. A simple, but conservative, method of modeling the reactivity effects of the injected boron was used for the SGTR model.

In this model, the RCS was considered to be a single volume, or "pot".

Following a delay time to allow for the borated we.wr to traverse through the cold leg to the top of the core, the ECCS flov was assumed to mix instantaneously with the entire primary volume. Mixing of this type is conservative, because the primary fluid would have a more dilute concentration of boron than if the ECCS fluid were to mix only with the cold leg fluid.

For the SGTR event, the precise time when the injected boron reaches the core is not of vital importance; therefore, the total delay time was estimated in a conservative manner.

A conservative boron worth of 9.0 pcm/ ppm was used to determine the total boron reactivity. The soron concentration in the ECCS fluid was 2000 ppm, which is the minimum concentration allowed by the Technical Specifications in the Refueling Water Storage Tank (RWST). For consistency with other EOL conditions, an initial boron concentration of 14 ppm was assumed.

5.11 Summary of Conservatisms in the SGTR Nominal Base Model The CPSESl-SGTR model described above is used as the Nominal Base Case for the comparative analysis discussed in Chapter 6. As previously stated, relatively few conservatisms have been incorporated into this 5-10

t l

l l

l l base model. Those conservative assumptions which are included in this I model are summarized below.

1) A sing 16-nodi sesam gsnsrstor model was used.

2) Choked flow through the ruptured tube, at either entrance to the  ;

ruptured tube, and at either exit of the ruptured tube was neglected. Only resistance limited (inertial) flow was considered. The density and pressure assumed for calculation of the resistance-limited break flow were those of the steam generator outlet plenum.

3) Conservative End-of-Life reactivity coefficients and point kinetics were used to provide the reactivity feedback. Boron from the ECCS was modeled in a conservative fashion. Reactivity control was not an immediate concern for the simulated portions of the SGTR event.

4) Only the Low Pressurizer Pressure signal was assumed to generate a reactor trip.

5) The pressure in the ruptured steam generator was limited to the setpoint pressure of the atmospheric relief valve, rather than that of the main steam safety valve. The pressures in the intact steam generators were allowed to rise to the setpoint pressure of the main stesm safety valves.

5-11

CPSES-1 BASE PLANT MODEL P

SGTR-SPECIFIC MODEL "NOMINAL" CASE i

EVALUATIONS OF CONSERVATIVE ASSUMPTIONS U -

CONSERVATIVE CASE i

l i

ACTIVE FAILURE ACTIVE FAILURE STUDIES STUDIES

-RADIOLOGICAL l cnNerotrFNcFc - S/G OVERFILL '

DESIGN-BASIS DESIGN-BASIS MODEL MODEL l

-RADIOLOGICAL - S/G OVERFILL CONSEOUENCES FIGURE 5-1. CPSESI-SGTR Model Flow Chart 5-12

l!

I e

i r

r.

r t

.'_-~._- .

v.'gA 4

2

^

a .

8g .

l P f _

~

~

2 Y _

OB i .

. D ui

. 1 S

. n -

- E

.. - e . S

. .. P

. C n .

r o

l .,  ?

. f m

• (, m b -

a r

g i .- .

. i a

. 3 D e

I

- [

y i

g n

" 4 d

o g . -

N

=

y 3_

a' k

c

' t_ e

,'~ D

~ .

1 a e s

a f B

- R

@ T

- G

- S I

.  ?

. 2 g

5

. m

=

, .. E

. . R U

G

{ I F

4

• J' 8e, H _

2m Y N p r

l

l l

l CHAPTER 6 l

JUSTIFICATION OF THE ADEQUACY OF THE CPSESl-SGTR MODEL l

l 1

The adequacy of the CPSESl-SGTR model was assessed by comparison with a l similar avslysis performed by the V0G, SGTR subgroup for a "Reference Plant" design. In this chapter, a comparison of the input assumptions used for the two analyses as well as a detailed discussion of the

)

analysis results will be presented. In addition, significant differences between the two analyses will be traced to specific plant I differences, specific analysis assumptions, or known modeling i differences.

During the development of their design-basis models, the V0G, SGTR ,

subgroup developed a "Nominal" base model for the SGTR event into which i

relatively few conservative assumptions were incorporated [2] . A fairly complete description of the analysis inputs and results was provided in the analysis reports submitted to the NRC by the subgroup, i l

This WOG analysis was selected for a comparison with the results of the i i

RETRAN02 model of CPSESl. A comparison between the WOG analysis and a similar, CPSESl-specific analysis was useful for providing qualitative assurance of the adequacy of the CPSESl-specific model for the analysis I of the SGTR event. I

(

Both of the previously discussed analyses were performed using nominal l

plant conditions; however, the generic SGTR analysis performed by the (

V0G,SGTRsubgroupwasbasedonaVestinghouse{ ]a,e plant with

( ]a,c steam generators. This design differs from the CPSES1 design I

t in many respects which affect the severity of the postulated SGTR l event. A discussion of many of the analysis assumptions follows, and a list of specific values used for the two analyses is shown in Table 6-1, 6-1

6.1 Nominal Case Initial Conditions /Eauiement Availability Assumptions concerning equipment availability and performance for the "Nominal" base case analyses by both the VOC and TU Electric are shown below:

a,c

1) Initialcorepowerwas[ ]ofnominalfullposer.

2) Loss-of-offsite power was assumed to occur at the time of the reactor trip,

3) Pressurizerheaterswere{ ]a,cfor consistency with the V0G analysis. -

4) Prior to the reactor trip, the steam generator water level was controlled by the steam generator water level control system.

5) For consistency with the WOG analysis, the main feedvater flow was isolated on the safety injection signal.

6) For consistency with the WOG analysis,[the turbine runback / rod controlsystemwasnotmodeled]a,cand the reactor trip on the Overtemperature N-16 reactor trip signal was not credited.

7) The CPSES1 auxiliary feedvater (AFW) system consists of two motor driven pumps and one turbine driven pump. For consistency with the WOG analysis, all auxiliary feedwater pumps were assumed to deliver auxiliary feedvater to all steam generators on the safety injection actuation signal.

8) The charging and letdown flow systems were not modeled.

6-2

9) The operator response times used[' ]a,c vere:

, .a,c a) Identify and isolate ruptured S/G b) Begin RCS cooldown c) Begin RCS depressurization d) Terminate ECCS flow These assumptions were used only for comparison of the "Nominal" analysis by TU Electric with the Reference Plant analysis. The applicability of these assumptions to the CPSES1, design-basis analyses are evaluated.in Chapter 8.

6.2 Operator Response Times The operator response times listed above were used only if plant conditions met a minimum set of conditions specified in the Emergency Response Guidelines (ERGS). Those minimum plant conditions are discussedbelow.(

la.c 6.2.1 Isolation of the Ruotured Steam Generator Consistant with the V0G analysis, the isolation of the ruptured steam generatorwasassumedtooccurat( ]a , c.or when the water level reached ( ]a,cof the narrow range span, whichever was longer. Because the Model D-4 steam generators at CPSES1 have an extended narrow range span, following a reactor trip, the steam generator narrow range level does not normally fall below approximately 30% of span. Thus, for the CPSES1 analysis,therupturedsteamgeneratorwasisolatedat(

]a cfollowing the initiation of the tube rupture. Isolation was accomplished by closing the main steam isolation valve on the ruptured steam generator, raising the setpoint pressure for opening of the atmosphere relief valve on the ruptured steam generator and terminating auxiliary feedvater to the ruptured steam generator. For the Reference Plant analysis (2), the time when the narrow range level 6-3

1 exceeded ( )a,cspan, 666 seconds, was used to isolate the ruptured steam generator.

6.2.1 Cooldown of the Reactor Coolant System Tha cooldown of the reactor coolant systemwasinitiated(

after the isolation of the steam generator. Steam was discharged

)'

through the atmospheric relief valves (ARVs) on the intact steam generators until a sufficient RCS subcooling margin was established at 3

the pressurs of the ruptured steam generator.

6.2.3 Deoressurization of Reactor Coolant System a,c

[ )afterthecompletionoftheRCScooldown,theprimary system was depressurized by opening a power-operated relief valve l (PORV) on the pressurizer to reduce the pressure differential between the primary and secondary systems.

In accordance with the CPSES1 ERGS.

the PORV was closed when any of the following criteria was satisfied:

a. The ruptured steam generator pressure was higher than that of the RCS, and the pressurizer level was greater than 20%; or. ,

l

b. Pressurizer level was greater than 78%; or,

c. RCS subcooling margin was less than 150F. i For the Reference Plant analysis, using Revision 1 of the generic Westinghouse ERGS, the criteria for the completion of the depressurization were: I a.

The RCS pressure was less than the ruptured steam generator pressure, and the pressurizer level was greater than 3%: or,

b. The pressurizer level was greater than 77%; or,

c. The RCS subcooling was less than 300F, 6-4

l l

l l

l a

For both analyses, thePORVwasclosedwhenthe( ]cr,citerion was satisfied.

l l

i 6.2.4 Termination of Safety Iniection 1 I

In accordance with the ERGS, the ECCS flow could not be terminated unless the RCS pressure was stable or increasing. For the CPSES1 analysis,thiscriterionwassatisfiedwithin{ ]a,c following the end of the depressurization. For the Reference Plant analysis, this  !

a,c jafterthe criterion was assumed to be satisfied more than[ I l

end of the depressurization when the RCS pressure had risen to 50 psi ,

j greater than the ruptured steam generator pressure.

6.3 Analysis Results The analysis results are discussed in this section. A description of j 1

the responses of each of the significant parameters, as calculated by the CPSESl-SGTR model, is followed by a discussion of the differences between the CPSES1 analysis and the Reference Plant analysis.

A sequence of events for the CPSES1 analysis is provided in Table 6-2.

The reactor was tripped on low pressurizer pressure, followed by turbine trip and a loss of offsite power. When the low pressurizer pressure setpoint for safety injection was exceeded, the safety injection actuation signal (SIAS) caused the isolation of the normal feedwater flow and the actuation of the auxiliary feedvater system, Therupturedsteamgeneratorwasisolatedat( )a,cfollowed by the initiation of the RCS cooldown at 900 seconds. The RCS depressurizationwasstarted( )a,cafter the termination of the RCS cooldown and was stopped when the RCS pressure was less than the ruptured steam generator pressure and the pressurizer level exceeded 20% span. ]a,cfollowing the end of TheECCSflowwasterminated(

the depressurization.

6-5

Also shown in Table 6-2 are the times of the significant events calculated for the Reference Plant. As discussed below, the differences in the calculated times may be attributed either to specific design features (e.g., safety injection flow rate, auxiliary feedwater flow rate, pressurizer size, etc.) and/or specific modeling differences (e.g., stroke times for steam generator ARVs) . Also, the plant-specific ERGS have some impact on the calculated results. j Significant parameters from both the Reference Plant analysis and the CPSESl-specific analysis are plotted in Figures 6-1 through 6-7. A key to the symbols used in the figures is provided in Table 6-3.

Discussions of specific aspects of the analysis follow.

6.3.1 General Descriotion As may be observed in Figures 6-1 through 6-7, following initiation of i the SGTR, the RCS pressure and level decreased linearly until the low pressurizer pressure reactor trip setpoint was exceeded. A turbine trip and a loss of offsite power occurred simultaneously with the  ;

reactor trip. When the low pressurizer pressure setpoint for the SIAS was reached, ECCS flow was 'ritiated, and the main feedwater isolation valves were closed, followed by the initiation of auxiliary feedvator.

Following the turbine trip, the steam generator pressure rose rapidly.

The ruptured steam generator ARV opened to relieve the pressure. After auxiliary feedvater was initiated, the steam generator pressure was determined by the amount of cold auxiliary feedvater entering the steam generators.

Following the manual isolation of steam flow from the ruptured steam generator, the gressure in the intact steam generators was maintained at the ARV setpoint pressure. The RCS pressure fell off sharply when the RCS cooldown was initiated. The cooldown was initiated by opening the ARVs on the intact steam generators and was terminated when the required RCS subcooling margin at the ruptured steam generator pressure wasobtained.( )a.clater, the RCS was depressurized by opening the pressurizer PORV. In accordance with the CPSES1 ERGS, the 6-6

depressurization was terminated when the RCS pressure became lower than the ruptured steam generator pressure, and the pressurizer level was greater than 20% span. [ ]a,clater, the ECCS flow was secured, marking the termination of the event.

I 6.3.2 RCS (Pressurizer) Pressure CPSES1 Analysis l

As shown in Figure 6-1, the RCS pressure decreased linearly with time until the low pressurizer pressure reactor trip setpoint was exceeded.

The turbine was assumed to trip simultaneously with the reactor; however, since the turbine stop valves closed faster than the control l rods fell into the reactor core, there was a brief spike in the pressure prior to the sharp drop-off following reactor trip.

Following reactor trip, the pressure continued to fall toward the setpoint for the generation of a SIAS on low pressurizer pressure. The pressure decrease was momentarily stopped as the cold leg temperature, and therefore the hot leg temperature increased following the loss of main feedvater flow, but then continued to fall as the RCS was drained through the ruptured steam generator tube. The pressure stabilized after the ECCS vas initiaced, and then began to rise after the ruptured steam generator was isolated, and the ECCS flow rate adjusted to match the break flow rate. The pressure fell' rapidly during the initial portions of the cooldown, but stabilized as the cooldown rate decreased with the depressurization of the steam generators. The slight increase in pressure near 1150 seconds was caused by the throttling of the auxiliary feedvater to the intact steam generators. The pressure decrease stopped after the pressurizer emptied at 1240 seconds. This change was attributable to the use of a non-separated, homogeneous volume in the surge line, which acted as the liquid region of the pressurizer. 'The fluid in the surge line was drained from the pressurizer, and was, therefore, hotter than elsewhere in the RCS.

Thus, the RCS was maintained at the saturation pressure corresponding to the fluid temperature in the surge line. Because this effect 6-7 k

resulted in conservatively high RCS pressures, continued use of this model was deemed acceptable. l l

Af ter the cooldown was terminated, t'ie pressurizer %g : to refill as the RCS fluid volume addition, due to the ECCS flow, exceeded the redaction due to RCS shrinkage and flow through the ruptured tube. The I pressure increased more rapidly as the pressurizer was refilled and the steam bubble was compressed. When the PORV was opened, the pressure fell sharply until the RCS pressure became less than the ruptured steam generator pressure. At this point, fluid from the ruptured steam generator backfilled into the RCS. Because the fluid in the ruptured steam generator was hotter than that in the RCS, a portion of the steam generator fluid flashed in the U-tubes and collected in the upper tube regions. This collection of fluid acted as a homogeneous pressurizer volume, much as the surge line had behaved earlier in the event. Thus, the rate of RCS pressure decrease was significantly reduced when backfill occurred.

I After the RCS depressurization was completed, the RCS pressure increased, and the voids in the U-tubes v re collapsed due to the pressurization and temperature reduction attributable to the high ECGS flow. After the voids were collapsed, the pressurizer pressure  ;

increased more rapidly as the pressurizer level increased and the steam bubble was compressed.

Following the termination of the ECCS flow, the pressure continued to kucrease, but at a slower rate, toward an "equilibrium" condition where {

l the decrxace in fluid volume through the break was compensated by the expansion of the fluid as decay heat was deposited in the RCS. The pressure remained at a slightly higher pressure than the ruptured steam generator pressure until operator actions uere taken to reduce the j pressures in both systems. As vill be discussed in Chapter 7, these

{

operator actions!:ould have taken the form of either initiating a controlled cooldown of the RCS or using t'ra pressurizer PORV to reduce t

che RCS pressure, l i

l 6-0 l

)

l

i Reference Plant Differences I

1 Because of the larger break flow rate and the smaller RCS and pressurizer volumes, the Reference Plant initially depressurized at a faster rate than CPSESL. Following reactor trip, the Reference Plant continued to depressurize until the equilibrium pressure was reached.

Again, due to the larger break size, this pressure was significantly lower than for CPSESl. Because the Reference Plant had a smaller pressurizer, the pressurizer emptied significantly sooner. The effects of isolating the ruptured steam generator, as seen in the plot of the Reference Plant analysis (Figure 6-1), were masked in the CPSES1 analysis due to the presence of a water level in the pressurizer. The increaseinthepressuretapediatelygriortotheRCSde{ressurization was not as sharp as in the CDSKil ar..alysis. Thery ere two primary contributing factors to the slower rate of repressu?ization: 1) the p.*essurizir in the Ref At{n e P'lant an.lysth was still empty; and, 2) the break fiov in the Reference Plahn was larger, indicating that the g

repressurization attributrble to the E^bS 'fiov was less than for the CPSES1 analysis.

i a s r \

Following the termination of th \ ECCS,flov,wthe pre ln ucq in the

' s

, Reference Plant analysis stasilized almost i=mediad',5y, whereas the

, , pressure in the CPSES1 italysis tidressed to a new, stable pressure.

i$%

. \ Again, this differesse !.n("equilibdium" pressures was prim rily attributable to the larger tuba diameter (larger break flow rate) in the Reference Ptait.

1 s

x \

6 A.3 Er;1pturizer 1 m

)

\ , 5

, \ CPSES1 Ar.QJ,. s; ,

, k The pressurizer level, shown in Figure 6-2, decreased linearly dith time following the initiation of the SCTR. At the time of reactor trip, the level fell sharply,and then stabilized as the ECCS acted to 5

maintain level. When the auxilicry fe.idvater to the intact steam generators was throttled after 900 seconds, the level increased as the

\

6-9 t i', s - , ' '

t s e

\ \

\

.>s t  ? .

RCS temperatures rose. Shortly after the start of the RCS cooldown, the level fell offacc.le low. The pressurizer actually emptied 150 seconds later. The pressurizer level returned on-scale during the depressurisation of the RCS. Following the end of the RCS depressurization, the level remained steady while the voids formed in the U-tube region of.the ruptured steam generator during the f

depressurization were collapsed. The level then increased more rapidly until the ECCS flow was terminated.

As the RCS heated up, the RCS inventory expanded, further increasing both pressure and level. The volumetric expansion of the RCS fluid forced subcooled fluid into the pressurizer. The resultant reduction in steam volume forced the steam to become superheated. Thus, a condition existed where there was superheated steam over a subcooled fluid. Hence, even though the pressurizer level decreased when the break flow exceeded the volumetric expansion caused by the addition of decay heat, the interphase mass transfet in the pressurizer acted to maintain and even increase the pressurizer pressure. This effect continued until the temperatures of the two phases in the pressurizer equalized.

Following operator actions to recover from the event (discussed in Chapter 7), the level would be maintained by occasional depressurization of the RCS.

Reference Plant Differences The pressurizer level plots for the Reference Plant analysis show similar trends. The most significant difference is due to the smaller pressurizer volume and larger break flow rate which caused the pressurizer level to fall off-scale low following the reactor trip. In addition, because the break flow rate was greater than the ECCS capacity, the'1evel did not return on-scale until the RCS was depressurized to refill the pressurizer.

6-10

i i

l It is suspected that the ratio of decay heat to the RCS volume was significantly less for the Reference Plant than for CPSESl. This assumption is consistent with the lack of pressure increase in the Reference Plant analysis following the termination of the ECCS flow.

In addition, the larger diameter U-tubes may have allowed sufficient primary-to-secondary flow-such that the primary pressure did not significantly increase as the RCS expanded due to the decay heat addition.

6.3.4 RCS Temeeratures CPSES1 Analysis As shown in Figure 6-3, following the initiation of the SGTR, the RCS temperatures decreased very slightly as the reactor power dscreased.

The power reduction was due to the effects of the moderator density coefficient. As the RCS depressurized, the fluid density, and hence, the power generated in the core also decreased, resulting in a lower hot leg temperature and a corresponding reduction in the cold leg temperature. The density decrease due to the depressurization was l partially offset by the increase in density due to the lower temperatures. Following the reactor trip, turbine trip, and the subsequent loss of main feedvater flow, the cold leg temperatures I increased rapidly toward the steam generator saturation temperature.

The cold leg temperature increase was terminated by the opening of the ARVs and the addition of large amounts of AFW to the steam generators.

)

Following reactor trip, the hot leg temperature initially fell, but then increased slightly as the RCS flow coasted down and the cold leg temperatures increased. A hot-to-cold leg temperature difference corresponding to natural circulation quickly developed, thus maintaining the hot leg temperature approximately 300F above the core inlet average temperature. This natural circulation flow was maintained by' steam release through the ARV. Due to the relatively long loop transit times in natural circulation, the hot leg temperature response lagged behind the cold leg response by at least 60 seconds.

Thus, the increased cold leg temperature due to the throttling of 6-11

auxiliary feedwater to the intact loops was not seen in the hot leg  !

temperature response until after the RCS cooldown had been initiated.

As the steam generators were depressurized, the cold leg temperatures, and hence, the hot leg temperatures, also decreased; however, the {

l transient hot-cold leg temperatura difference was slightly larger than before due to the long loop transit times. Following the completion of the RCS cooldown, the cold leg temperature began to rise; however, the hot leg temperature continued to decrease for a longer period of tima than'could be accounted for by the loop transit times. The hot leg temperature was based on the core inlet temperature, which is an average of the cold leg temperatures. At this point in the transient, the cold leg temperature of the ruptured loop was relatively cool due to the large proportion of ECCS fluid mixed with the almost stagnant RCS fluid. Thus, even though the intact loop cold leg temperature was increasing, the core inlet temperature and the corresponding hot leg temperature, continued to fall for several seconds.

Following the termination of the ECCS flow, the cold leg temperature j increased. Because the plot shown in Figure 6-3 actually shows the

)

steam generator outlet temperature rather than the cold leg temperature, the temperature increase, due to the termination of the ECCS flow, was delayed by one loop transit time.

Reference Plant Analysis l

With few e'aceptions, the plots of the intact loop temperatures follow 1 j

essentially the same trend as those in the CPSES1 analysis. Following the initiation of the event, the reactor power was held constant. The RCS temperatures increased slightly as the core flow rate was reduced  !

and the heat transfer in the steam generators decreased. After the i reactor trip, the hotter pressurizer fluid flowed into the hot leg as l the RCS inventory was drained through the break, and thus caused an l increase in the hot leg temperatures on the pressurizer loop. The i slight increase in hot leg temperature caused the steam generator pressure to increase, which acted to stabilize the cold leg i temperature, a

6-12 l

1

In addition, as described in Section 6.3.3, it is surmised that the ratio of decay heat to RCS volume was significantly less for the Reference Flant than for CPSESl. This assumption is consistent with the decrease in RCS temperatures seen in the Reference Plant analysis PVior to the initiation of the RCS cooldown. The effect of the reduced decay heat /RCS volume ratio was most obvious after the pressurizer emptied, thus ending the heat addition due to the relatively hot pressurizer fluid.

It may also be postulated that the primary-to-secondary heat transfer was greater in the Reference Plant analysis. The increased heat transfer . rate is consistent with the increased primary cooldown rate and the relatively high steam generator pressures (Section 6.3.7).

Another significant difference was the sharp increase in hot leg temperature during the RCS depressurization. The obvious reason for the temperature increase was the mixing of the influx of saturated fluid from the ruptured steam generator with the RCS fluid. However, because of the timing of the temperature increase and the consideration of the loop transit times, the backfill of the hotter steam generator fluid did not appear to be the correct reason.

6.3.5 Break Flow Rate CPSES1 Analysis Following the initiation of the event, the break flow rate (Figure 6-4) decreased as the RCS pressure decreased. After the reactor and turbine were tripped, and the ensuing RCS pressure decrease and steam generator pressure increase, the break flow rate dropped sharply. As the ECCS flow rate caused the RCS pressure to rise, and the auxiliary feedvater flow decreased the steam generator pressure, the break flow rate gradually increased.

The initiation of auxiliary feedvater caused the break flow rate to increase briefly as the steam generator pressures were reduced. The rate of increase fell sharply as the ruptured steam generator pressure increased following the isolation of auxiliary 6-13

h .

feedwater. When auxiliary feedwater to the intact steam generators was throttled,- the resultant RCS pressure increase caused the break flow rate to increase. The flow rate followed the RCS pressure trend curing the cooldown of the RCS; however, the magnitudes of changes in the break flow rate due to changes in RCS pressure were partially offset by the increased break fluid density. Following the initiation of the RCS depressuriaation, the break flow again fell sharply. The break flow rate momentarily increased when a slug of cool ECCS fluid entered the steam generator outlet plenum. The cooler ECCS fluid entered the outlet plenum when the flow in the ruptured loop briefly reversed during the depressurization (recall that a higher density. fluid results in a higher break ficw rate).

When the RCS was depressurized with the pressurizer PORV, the break flow rate feil sharply until the RCS pressure became less than the ruptured steam generator pressure. Due to the lower RCS pressure, the break flow became negative, i.e., the flow direction changed due to backfill from the ruptured steam generator into the RCS. Following the end of the depressurization and the termination of ECCS flow, the break flow rate increased toward a stabilized value of approximately 30 lbm/sec. This relatively high break flow rate was due to the increases in RCS pressure discussed in Section 6.3.2. Additional operator actions, described in Chapter 7, would be required to continue the RCS cooldown or to reduce the RCS pressure to terminate the break flow.

Reference Plant Analysis Prior to the termination of the ECCS flow, the break flow rate followed the RCS pressure predictions for the Reference Plant. The ruptured steam generator pressure in the Reference Plant analysis was held essentially constant throughout the event, thus only the RCS pressure and the RCS average density affected the break flow rate. Following the termination of ECCS flow, the break flow rate gradually decreased as the RCS and secondary pressures tended to equalize. The pressure equalization was hastened by operator actions taken to initiate a 6-14

l controlled RCS cooldown in accordance with recovery guidelines.

6.3.6 Steam Generator Narrow Ranne Level Response CPSES1 Analysis The transient response of the steam generator narrow range level indication is shown in Figure 6-5. After the SGTR begins, the void fraction in the ruptured steam generator was increased due to flashing of the RCS coolant; hence, there was an increase in the level indication. In addition, the increased steam generator mass, due to the break flow, caused the level to increase before the Steam Generator Vater Level Control System could act to maintain the indicated level at its setpoint. The steam generator water level controller acted to maintain the level near the nominal value. Following the reactor and turbine trips and the initiation of auxiliary feedwater, the level fell off sharply as the heat from the primary required for void generation was lost. Because of the extended narrow range span on the Model D 4 steam generators, the level did not fall off-scale following a reactor trip. The level began to rise when the ARVs were opened, but fell again as the reactor heat generation rate fell and the cold auxiliary feedwater collapsed the voids in the steam generator mixture.

Throughout the remainder of the event, there was relatively little voiding in the ruptured steam generator, thus the level increase was primarily due to mass addition into the steam generator through the ruptured tube. Inflection points in the plot of the ruptured steam generator level are due to the isolation of the ruptured steam l

generator (including auxiliary feedwater isolation) and the opening of the ARV on the ruptured steam generator as the pressure rose following the isolation. The level continued to rise until the RCS was depressurized, at which point, the level stabilized as the RCS was backfilled from the ruptured steam generator. Following the pressure increase, due to the RCS heatup after the termination of the ECCS flow, I the level again began to increase. Additional reactor operator 1

6-15

l I

recovery actions would be required to initiate a controlled cooldown and take the plant to cold shutdown conditions.

The indicated levels in the intact steam generators romained constant until the time of reactor and turbine trip. As in the ruptured steam generator plot (Figure 6-5), the levels increased as the ARVs were opened allowing increased void formation and then fell as the cold auxiliary feedwater collapsed the voids in the steam generators.

Operator actions were modeled to throttle the auxiliary feedwater to maintain the intact steam generator levels to between 10% and 50% span; orgo, the levels were maintained near a value of 45% span. When the RCS cooldown began, the levels in the intact steam generators rose due to the rapid void formation resulting from the sharp pressure reduction. The levels then began to fall as the inventory was depleted, but were prevented from falling much below 45% span by the addition of auxiliary feedwater. There was a sharp drop in the levels when the ARVs were closed, resulting from an increase in the steam generator pressure, a rapid void collapse, and a corresponding reduction in the mixture volume. The levels vero maintained near 45%

span for the remainder of the event.

Reference Plant Analysis The trends shown in the plots of narrow range level are similar to those of the CPSES1 analysis. Following the initiation of the event, the level began to decrease in all steam generators. This decrease is believed to be due to the slight increase in RCS temperature and corresponding increase in steam generator pressu es discussed in Section 6.3.4. The level in the ruptured steam generator was slightly greater than that in the intact generators.

The level fell offscale following the reactor trip. As expected, because of the increased mass addition, the level returned on-scale in the ruptured steam generator before the intact steam generators. The inflection points identified in the CPSES1 analysis are also noticeable l in the plot of the ruptured steam generator level. As will be ,

1 6-16 i

discussed in Section 6.3.7, no inflection point is observed when the ruptured steam generator pressure rose to the ARV setpoint pressure as in the CPSES1 model.

The plot of the level in the intact steam generator is similar to the corresponding plot from the CPSES1 analysis. Because the n itrov range spanfortheReferencePlant's( ]a,c steam generator was less than the span for CPSES1, the magnitudes of the level variations were magnified. The effects of initiating a controlled cooldown to again I

terminate the break flow were seen in the level response near 2200 seconds. This specific operator action to recover from the event i was not explicitly modeled in the CPSES1 analysis.

1 1

l i

1 6.3.7 Steam Generator Pressure Response

)

I CPSES1 Analysis. Common Resoorses The steam generator pressure response is shown in Figure 6-6. Because they are connected through the main steam header, the pressure responses of the steam generators were approximately the same prior to isolation of the ruptured steam generator. Following the initiation of the SCTR, the pressures in the steam generator decreased slightly due to the corresponding reduction in core power. Because the steam dumps were not modeled, the steam generator pressures rose sharply following \

turoine trip. The pressure leveled off as the ARVs were opened, then fell following the initiation of auxiliary feedwater flow.

CPSES1 Analysis. Ruotured Steam Generator Resoonse Following the closure of the MSIV on the ruptured steam generator and the termination of AFW to that steam generator, the ruptured steam generator pressure increased toward the ARV setpoint. Because this pressure increase was driven primarily by the compression of the steam bubble as the steam generator liquid volume was increased due to the break flow, the rate of increase was relatively slow. The pressure was maintained at the ARV set pressure until the RCS cooldown was 6 17 ,

1

initiated, at which time, the steam generator pressure very slowly drifted down.ard. Some of the pressure decrease, caused by the RCS cooldown, was counteracted by the pressure increase caused by the steam compression in tha steam doma. During the RCS depre:surization, the RCS pressure became less than that in the ruptured steam generator pressure, thus reducing the ruptured steam generator pressure. The pressure reduction was terminated when the pressurizer PORV was closed, ending the RCS depressurization.

Even though the RCS pressure increased following the termination of the ECCS flow, the steam generator pressure decreased at a very slow rate.

This pressure decrease was due to the addition of cooler RCS fluid and reverse heat transfer from the steam generator to the cooler RCS fluid. The degree of the pressure decrease was offset somewhat by the compression of the steam bubble caused by the addition of the RCS fluid through the ruptured tube. The ruptured steam generator would be depressurized through the RCS as the reactor operators recovered from the SGTR event.

g&fj31 Analysis. Intact Steam Generator Resoonse After the AFW to the intact steam generators was throttled, the intact steam generator pressure also increased as the cold auxiliary feedvater in the generators was heated by the hotter RCS fluid. The RCS ,

cooldown, effected by opening the ARVs on the intact steam generators, caused a rapid decrease in the steam generator pressure. As expected,  !

as the steam generator pressure decreased, the rate of change decreased; thus, the temperature reduction was more rapid at the start than toward the end cf the cooldown period. Subsequent to the l

termination of the cooldown, the pressure began to recover, increasing toward the saturation pressure corresponding to the RCS temperature.

Reference Plant Analysis I

1 Prior to the reactor trip, the steam generator pressure increased due l to the increase in the RCS temperatures and the injection of hotter xCS 6-18

i

. l fluid into the secondary. Following reactor trip, the pressure in the Reference Plant analysis quickly rose to the setpoint pressure of the ARVs. The differences in the predicted pressure between the CPSE31  ;

i analysis and the Reference Plant analysis are attributed to differences ,

in the valve and steam generator models used. In the CPSES1 model, the )

valve was modeled to modulate open in response to a pressure error signal. The maximum stroke time was used to prevent the valves from "popping" open or closed in an unrealistic manner. In the Reference Plant analysis, however, the amount of inventory required to maintain the volume pressure at the valve setpoint pressure was instantaneously removed from the system (up to the maximum rated flow through the valve). Thus, because the "relief valves" responded instantaneously, the pressure response from the Reference Plant is expected to be  !

smoother than the CPSES1 response when the steam pretsure was being maintained at the ARV setpoint. In addition, when the RCS cooldown was I initiated, the initial rate of depressurization was expected to be somewhat greater.

The effects of initiating the controlled cooldown near 2200 seconds can be seen in the plot (Figure 6-6) of the intact steam generator pressure. The remaining portions of the pressure plots are similar to those of the CPSES1 analy' sis, i

l 6.3.8 Ruotured Steam Generator Liouid Volume 1

l CPSES1 Analysis The collapsed liquid volume (Figure 6-7) sas used to determine the margin to overfill of the ruptured steam generator. Due to the action of the Steam Generator Water Level Control System, the liquid volume in the ruptured steam generator remained approximately constant until the reactor and turbine were tripped. At that titc, the volume increased sharply until the main feedwater pumps were tripped. The volume fell slightly as some inventory was released through the ARV to maintain pressure. The addition of large amounts of auxiliary feedwater 6-19 l

continued to cause the steam generator liquid volume to increase, but at a much slower rate. Following the isolation of auxiliary feedwater to the ruptured steam generator, the rate of volume increase, nt.w a function of the SCTR break flow rate, was again reduced. The volume in the ruptured steam generator fell alightly when the RCS was depressurized to a lower pressure. Following the end of the depressurization and the termination of the ECCS flow, the liquid volume in the ruptured steam generator rose until the recovery actions were taken to initiate the controlled RCS cooldown.

Reference Plant Analysis The trends displayed in the plot of ruptured steam generator water volume (Figure 6-7) are similar to those predicted by the CPSESL analysis.

Due to the magnitude of both thc break flow and the auxiliary feedwater flow of the Reference Plant, the volume increased more rapidly in the Reference Plant analysit than that for the CPSES1 analysis.

Notice that the auxiliary feedwater capacity of the Reference Plant

( )a,cvas roughly the same as the capacity of CPSES1 (four steam generators). The stable water volume following 2200 seconds was due to the controlled cooldown. initiated in the recovery phase of the event.

6.4 Summary The results of the analysis of the "Nominal" postulated steam generator tube rupture event performed using RETRAN02 MOD 004 and the CPSESL-specific model have been compared with a similar analysis performed for the V0G, SCTR subgroup and approved by the NRC (6). With the exception of an unexpected increase in the hot leg temperature discussed in Section 6.3.5, all significant differences between the two analyses have been attributed to known modeling differences or to differences between the designs of the Reference Plant and CPSESl. The one unexplained difference does not affect the conclusion that the CPSES1,RETRAN02 MOD 004 model is well-suited for simulating a postulated SCTR event.

6-20

TABLE 6-1 Nominal Base Case -

Assumptions and Plant Differences Between the Reference Plant and CPSES1 Reference Plant CPSES1

. .a.c Initial RCS pressure, psia 2250 Initial pressurizer water volume, ft**3 1080 Steam Generator Tube Inner Diameter, inches 0.664 Pressurizer Pressure for Reactor Trip, psia 1925 Pressurizer Pressure for SI, psia 1844 SG ARV Set Pressure, psia 1140 1

SIS Pump Delay, sec 22 j AFW Delay, sec 60 l

AFW Flow Rate (total gpm) 2218 AFW Temperature (0F) 120 Decay Heat (1971 ANS Decay Heat Curve) 100%

I I

l I

i

! 6-21

TABLE 6-2 Nominal Base Case -

Sequence of Events Time (see)

Evang Reference Plant CPSES1 Tube Failure 0 0 Reactor Trip 259 357 Turbine Trip 259 358 Safety Injection Signal 285 372 Safety Injection ---

394 AFV Actuation 346 432 Isolation of Ruptured SG - ' '

( .

Start Cooldown 901 Pressurizer Emptied 1240 Complete Cooldown 1326 Pressurizer Refilled 1400 Start Depressurization

)a,c Complete Depressurization 1563

, . a.c Terminate SI t

6 22

l TABLE 6-3 1 i

Nominal Base Case -  !

Key to Figures 6-1 through 6-7 l Point Event A Reactor Trip  !

B Safety Injection Actuation Signal C AFW Actuation D Isolation of Ruptured Steam Generator E

Throttle ATV to Intact Steam Generators  :

F Start Cooldown G Complete Cooldown H Start Depressurization '

J Complete Depressurization 1 i K Terminate Safety Injection l

4 l

1 1

5 l

1 i

l

6-23

CPSES SGTR - NOMINAL BASE CASE m ssum m m ssunc 2.4 2.2 -

r I l 2- ", e 8 1J -

t.s - a 1-

a. i.4 "

l w

1.2 - #

l Uw 1-t 0.s -

0.s -

0.4 -

0.2 -

0 . . . . . . 6 i . . . i 0 0.4 0.8 1.2 1.5 2 2.4 (Theveense)

TWC (SECONOS)

REFERENCE PLANT

_ _ a.c .

i l

1 4

j _ -

FIGURE 6-1. Nominal Base Case Comparison - Pressurizer Pressure 6-24  !

1

CPSES SGTR - NOMINAL BASE CASE too mssumzm uvn.uncane so -

en -

70 -

so -

E so -

h m-so - I A J  !

r 20 -

8 3 C g to - O L o , , , , , ,

o c.4 o.s 1.2 1.s 2 2.4

%)

T1ut(stcQNOS) i REFERENCE PLANT

_ac FIGURE 6-2. Nominal Base Case Comparison - Pressurizer Level 6-25

i l

CPSES SGTR - NOMINAL BASE CASE wru:T ux, memmmes A

O ec 000 - 1 1

r "J A .

Soo -

E m- C 400 -

330 -

300 -

250 -

200 , , , , , , , , , , , ,

o 0.4 0.8 1.2 1.4 2 2.4

%)

T1WC ($CCONO3) 7-Hof -

T-CotD REFERENCE PLANT

, _ a,c 4

l i

l l

1 l

I l l 4

I FIGURE 6-3, Nominal Base Case Comparison - Intact Loop Temperatu e 6-26

\

i

l CPSES SGTR - NOMINAL BASE CASE i

, samemu n.ow mt ao -

l n-n

• u m. 4 k 1 l ""

• f5

so -

A I 20 -

f to -

o  !

-10 +

kV l i

2

-30 . . . . . . , , , , ,  !

0 0.4 o.a 1.2 1.s 2 l 2.4 (Thousands) l TNC (SECOND5)

_ REFERENCE PLANT

- a,c l

1 1

l l

l i

i i

I FIGURE 6-4. Nominal Base Case Comparison - SGTR Break Flow Rate 6-27 l

CPSES SGTR - NOMINAL BASE CASE

, s/o NAnnoW RANot I.DtL WOcADON so -

80 -

A M

70 - r_ _

4, a .

I h So -

w

, so . p --

40 -

E c so -

L 20 -

10 -

0 i -i i i . i i i

i i i 0 0.4 0.8 i

1.2 1.5 2 1.4 (Theveende) nWC (SCCONOS)

M Act -.- nuprunto REFERENCE PLANT

-a,c 4

l l

l FIGURE 6-5. Nominal Base case comparison - Steam Generator 1.evels 6-28 l -. _ - _ _ _ -

CPSES SGTR - NOMINAL BASE CASE mm cimmon mssuncs i 1.2 - H J 1_[ ]

1.1 - a  !

4 0

1- _1 E

0.9 - 1 2

am 0.8 -

Wl w; 0.7 - y , i e'

@ o.g . a g I-U g

d0.5 -

0.4 -

)

0J -

0.2 -

0.1 -

0 , , , ,

0 0.4 0.3 12 1.6 2 2.4 (Thewsonds)

Ttut (sCCONOS)

INTACT -

RUPTUnG l

1 RETERENCE PLANT  ;

a,c 1

1 l

I 1

I FIGURE 6-6. Nominal Base Case Comparison - Steam Generator Pressure 6-29

CPSES SGTR - NOMINAL BASE CASE  ;

MUPTUMD 5/C U0UW VOLUWC

. s-TOP or SIFAN CodRATOft j l

s- I i

m '

9 4-an b

w7 il o E ,

l 3

a

1. h 9 A h

g N I g 2- C 1-f 0 , , , , , , , , , , ,

0 0.4 0.8 1.2 1.6 2 2.4 (M)

Ttut (stCONOS)

REFERENCE PLANT i

-aC i

l I

l I

l l

i 1

l l

FIGURE 6-7. Nominal Base Case Comparison - Rupture S/G Liquid Volum 6-30 I

CHAPTER 7 i OPERATOR RESPONSE TIME EVALUATION i

The steam generator tube rupture event must be terminated by actions of the reactor operators. The plant-specific analysis required the use of operator response times which were consistent with plant-specific equipment, procedures, and training. In order to determine these t response times for CPSES1, a SCTR event using design-basis assumptions regarding equipment availability, but without any active failures, was included as one of the simulator exercises during the Licensed Operators Requalification Training in August and September of 1987.

The scenario used for the simulator exercises included an initial primary-to-secondary leak rate representartve of a complete severance of a U-tube. The radiation monitors were assumed to be inoperable. -

Other failures, designed to simulate the design-basis event, included the failure of all automatic reactor trips and a loss of all offsite power coincident with the manual reactor trip. The turbine and control rods were placed in manual control. In addition, the N-16 related i alarms and turbine runback alarms were defeated.

The timing of the first few operator responses following the initiation of the event have the most impact on the final severity of the postulated SGTR. These responses concern the identification and isolation of the ruptured steam generator and the initiation of the cooldown of the RCS. The operator response times for these actions were measured directly from obeervations of several simulator exercises. The operator response times for the remaining actions were estimated from examination of the CPSESl procedural guidelines and discussions with CPSES1 reactor operators concerning the difficulty in carrying out these actions.

1 1

As discussed in Chapters 4 and 6, the first significant operator action following the postulated SGTR is the isolation of the ruptured steam 7-1

generator. In accordance with the CPSES1 ERCS, the initial operator responses following the generation of a reactor trip signal are intended to ensure that the control rods had entered the core and that  ;

the required safety systems were operating. Subsequent actions are l directed toward the diagnosis of the event and then the identification of the ruptured steam generator. Following the isolation of the I ruptured steam generator, addittonal operator actions are required to ensure that the required equipment is available and to prepare to initiate a cooldown of the RCS. The initiation of a maximum rate cooldown is the second significant operator response.

A summary of the response times for these two significant responses, as recorded during the simulator exercises, is provided in Table 7-1.

Each exercise used one-half the personnel on a regular control room shift. Use of the "half-shift" was consistent with the minimum number of licensed operators required to be in the control room at any given time. Although ten exercises were executed, only nine were used to

~

determine the plant specific operator response times. ,

l Insufficient data was collected during Run 9 to allow determination of ,

the timing of the operator responses. Also, no confirmatory plots of critical parameters were available for Run 7 due to equipment problems; l however, the manually recorded operator rcsponse times were included. l In addition, that portion of the exercise following the isolation of the ruptured steam generator from the Run 4 exercise was not used. The simulator model of the instrument air reset sequence did not accurately j reflect the plant hardware. An action by the simulator instruccor in response to directions from the shift supervisor was required to 1

re-establish instrument air; the instructor was occupied elsewhere at j j the critical time, and the initiation of the RCS cooldown was delayed

by several minutes. )

l As discussed previously, the response of the simulator made the direct measurement of the operator response times following the initiation of I the RCS cooldown difficult. Thus, the rcaponse times required for the operators to prepare for and to initia'es the depressurization of the  !

7-2 4

I

RCS were derived from limited observations and discussions with operators. The procedural steps and the ease of performing these steps were considered. The required response time for the termination of the ECCS flow following the end of the RCS depressurization was similarly estimated. Through these discussions and observations, it was determined that a two minute time period was a conservative estimate of the time required to initiate the depressurization of the RCS following the end of the cooldown. Similarly, the time required to terminate the ECCS flow following the end of the RCS depressurization was

}

conservatively estimated as one minute.

For conservatism, the maximum operator response time that was required to identify and isolate the ruptured steam generator was used as the plant-specific identification and isolation time. Similarly, the f maximum time between the steam generator isolation and the beginning of the cooldown was used as the plant specific tiae for the beginning of the cooldown. These maximum times as well as the estimated times for the initiation of the depressurization of the RCS and the termination of the ECCS flow, are summarized in Table 7-2.

In summary, based on observations of several "half-shif ts" of licensed reactor operators responding to a severe Steam Generator Tube Rupture event, the operator action times presented below were determined to be applicable to CPSESl. These response times were based on procedures which have since been updated to clarify guidance for decisions based l on the judgement of the Shift Supervisor. More revisions, based on a generic revision of the Westinghouse ERCS, are forthcoming. Because )

most of the procedural steps which call for operator judgement will be clarified, it is anticipated that these procedural changes will reduce the overall time required for termination of a SGTR event as well as make the response times from different shif ts more uniform. Continuous training emphasis on the need to prevent steam generator overfill during a SGTR also will assure appropriate and timely operator responses to a Steam Generator Tube Rupture event.

7-3 l

i l

1 TABLE 7-1 )

i SGTR Operator Response Times from Simulator Exercises DELTA-TIME DELTA-TIME (minutes)

SGTR -

(minutes)

SG IS01ATION -

l SG IS01ATION START COOLDOWN l Run 1 10.84 4.66

Run 2 4.20 3.80 Run 3 12.00 3.00 Run 4 8.25 ----

Run 5 12.84 1.83 Run 6 10.00 2.84 i .

• Run 7 9.50 3.83 Run 8 7.50 4.75 Run 10 12.75 4.00 l l

l 1

• --No confirmatory plots available l l

i j l 1

i 74 4

i

l ',

I a

s R

/

TABLE 7-2 ,

s Conservative Operator Responsa Times Assumed for the CPSES1 Design-Basis SGTR Analyses ,

is SGTR initiation - SG isolation 13.0 minutes 1

)

SG isolation - initiation of RCS cooldown 5.0 minutes

~

Termination of RCS cooldown '

i

- Initiation of RCS depressurization 2.0 minutes ,

f l

) Termination of RCS depressurization I Termination of ECCS flow 1.0 minute N

l 1

1

(

1 f

I 1

l a

i ,

l 7-5

t i

1 I

/s i

/ ,

),

7 CHAPTER 8 i i

EVALUATIONS OF CONSERVATIVE ASSUMPTIONS

\j

{

Following the creati a f the Nominal Base Case, preliminary andfses

,and evaluations of the SGTR scenario were performed to identify c'ritical parameters and to determine the conservative direction with respect to overfill of the ruptured steam generator for each of these i

parameters. Conservative assumptions for this set of critical parameters were then incorporated into the Nominal Base Case to create the "Conservative" base nodel. As stated previously, this Conservative

, ' base model is the same as the design-basis models with the exception of the postulated single active failures.

For the least margin to overfill of the ruptured steam generator, it is I conservative to maximize the liquid mass in the ruptured steam j generator; therefore, the initial conditions and other assumptions are selected to maximize this mass. ]

l l

1 8.1 Offsite Power Availability 1 j

1 An evaluation of the effects of the loss of offsite power and of the timing of a postulated loss of offsite powcr was performed. For the design basis analysis, it vss determined that offsite power should be as9"medtobeavailablepriortoreactortripandlostasaresultof tha turbine trip coincident with reactor trip.

In the unlikely event that offsite power were to be lost coiwidently with the initiation of the SGTR, the ensuing reactor trip would cause an immediate decraase in RCS pressure, thes reducing the break flow rate. In addition, no turbins runback would occur; hence, there would be no increase in the steam' generator mass prior to reactor trip.

Assuming the. reactor operators require a constant amount of time to complete the actions outlined in the ERGS, the identification and 0-1

c._, - __ , . . - . ._ _. __ . ...

.,:( '

~ s 1 ',

Q'l 1

4N  ;

T 7

'q

• 16

,, (

i x ( 'g / '

)

.\ s,_,, ,

)

(

/ isolation of the ruptured steam generator and the initiation of the RCS I N cooldown would occur over a time period shorter than that assumed for cf

(, the design-basis s,:'enario. This relatively early isolation time would q

[

result in essentiall'y 'the same amount of auxiliary feedwater injection,

$, no increase in taass due to a possid turbine runback, a shorter event ts s p', m duration, and thus, a lower integweed break flow into the ruptured

-$ enerator.

Furtheragte, there is no credible event which' would steam g\ ,,

'\g 3 cause e nimultaneous steam generator tube rupture and a loss otNoffsite

.k # power. A  !

L *

<, . 't .

The primu f effect of assuming no loss of offsite power following '

\

coactor trip ycktid be the 1opraved '3rimary-to-secondary heat transfer

^

g' -.%

during th AC9.cooldown due to ti.4 continued operation of the reactor coolait pumps. Assuming of f sito, power was available af ter re:tetor trip, the pressurizer pressure decrease would be mitigeted by the s ,

pressurizer'htaters, thus deln ing the initiation of ?4CS Clow and

^ \

auxiliary'feedwater. Delaying the initiation of these: 3ys:.ecis would be

\

f non-cont.ervative with respect to overfill. However, the pressure , utd therefore, the break flow rate would be maintained at a slight 1'y higher val e nntil the heaters shut off on low pressurizer leN 1. Even'.LougC a the R4Ps(and pressurizer heaters) would be adding heat to the sy2 tem, '

\ - y tN immoved heat transfer 'due to the higher flow rate would allow a '

t '\

very npid cooldown co ne performed. The net result would be a quicker i

termination of the hrsak ff M , resulting in a greatet margin to )

overfill. ,

~

In addition, if offsite s.oddr were available, the preferred equipment- 1

\

for cooling and depressurizing the RCS would also be available. Thas, the operator response tites m uld be slightly less, t!

', i ,

J y ,

8.2 Timine of..thstifatsigr Trio -

Many of the conservative assumptions are concerned with the timing of l s ,o the rec.ctor tripjand its effect on the rupturedssteam generator mass. t 1

Thelevelin$herupturedsteamgeneratorvculdnotsignificantiv, i 1 x s i

4

v. .  ; l g L I 8-2 -

s s  ;

\

s .

.\ is  !

_ h' .. _s 3 i -

increase significantly prior to reactor trip due to the action of the main feedwater controller. Following reactor trip, the auxiliary feedwater, which does not have any steam generator level control system, would be initiated, and the steam generator level would rise I unchecked by any automatic control.

The auxiliary feedwater to the ruptured steam generator would be isolated by the first credited operator response which would occur at a fixed time following the initiation of the event. Therefore, it is conservative for the auxiliary feedvater to be initiated earlier, resulting in more auxiliary feedwater injected into the ruptured steam generator prior to steam generator isolation.

In addition, for the evaluation of the operator response times, the CPSESL plant simulator was configured to maximize the time of reactor trip after the SGTR was initiated; however, the initial operator response time was measured from the time of SGTR initiation. Thus, the initial operator response time represents a minimum time between reactor trip and steam generator isolation. However, in the SGTR analyses, the initial operator response time was assumed to occur at a j fixed time following the initiation of the tube rupture. In general, j bor,euse of the auxiliary feedwater flow, maximizing the time between the reactor trip and the initial operator response is conservative.

Therefore, it is conservative for the other analysis assumptions to result in an early reactor trip, thus maximizing the time between reactor trip and steam generator isolation.

8.3 ID itial Thermal-Hydraulic Conditions 8.3.1 Egweg

, , Due to an increased void fraction in the steam generators, a higher core power would result in a lower steam generator mass. Because the i cov ervative assumption should result in a higher steam generator mass, a "lower" core power should be used. A competing effect is the fact 8-3

that the decay heat would be greater if the reactor was tripped from a higher power. A higher decay heat would result in a higher RCS pressure (and thus greater break flow), as well as an extended cooldown period.

Both of the effects of a change in the initial power were accounted for in the model. The initial steam generator mass was increased to bound the cases where a lower reactor power would result (e.g., due to a turbine runbuck), and 100% full power was used to maximize the decay heat following the reactor trip. As an additional conservatism, a 20%

uncertainty was added to the decay heat (discussed in Chapter 7); thus, no additional penalties on the core power were included in the model, and nominal full power (3411 MWt) was used.

Zero Power Considerations As indicated above, a lower reactor power would result in a greater steam generator masa. Thus, the limiting case, zero power, must also be considered. Assuming the plant was at hot zero power conditions, auxiliary feedwater (under manual control) would be used to maintain l

the steam generator levels.

i Following the reactor trip alarm, most of the immediate operator actions would not be applicable because the plant would already be tripped. Because the steam generator level would be controlled manually, the increase in level resulting from a tube rupture would be noticed relatively quickly and compensated by the throttling of the auxiliary feedwater to that steam generator. Thus, the time required for identification and isolation of a ruptured steam generator would be greatly reduced from that required at power. At hot zero power, the steam generator would be maintained at a higher pressure than at full power, thus the initial break flow would be less. In addition, the detrimental effects of a postulated turbine runback would not have to be factored into the analysis. Furthermore, at hot zero power, the average RCS temperature would be significantly less, resulting in a reduced time requirement for the RCS cooldown.

8-4

1 l

The pr*ssurizer level would be lower at zero power; thus, the pressurizer would empty sooner. However, because the initial steam l

volume would be larger and the RCS shrinkage following the reactor trip l would not be as severe as for the full power case, the rate of depressurization would be slower, thus delaying the initiation of ECCS flow. Because the ECCS flow would cause ths RCS pressure to increase, delaying the initiation of the ECCS flow would result in a lower break flow rate.

As will be discussed in Section 8.3.7, the initial steam generator mass for the full power case was increased to the mass corresponding to approximately 40% power. In addition, the initial conditions used in the RCS and core corresponded to 100% power. This combination of conditions resulted in a greater potential for overfilling the ruptured steam generator than an analysis initiated at zero power with I consistent initial conditions, event sequences, and operator response times.

8.3.2 EGE Pressure 1

A reduced initial pressurizer pressure would result in an earlier  !

reactor trip (on low pressurizer pressure), thereby causing the auxiliary feedwater to be initiated at an earlier time. For accident analysis, the error band on the initial pressurizer pressure is 30 psi. Considering this uncertainty, the initial pressurizer pressure was assumed to be 2220 psia.

l 8.3.3 Initial RCS Temperature J

In order to maximize the integral break flow, the primary-to-secondary l pressure differential and the cold leg (steam generator outlet plenum) fluid density should be large, and the pressure differential should be maintained at a high value for as long as possible. With respect to the RCS temperatures, the break flow rate is maximized by a low T-cold in conjunction with a high T-avg. The low T-cold increases the fluid density while the high T-avg prolongs the cooldown period.

8-5

m l

)

Assuming the reactor power and RCS flow rate are constant, an increase

~

in T cold must be accompanied by a corresponding increase in T-avg.

Thus, changing the initial RCS temperature (typically changed 1 5.50F) would result in competing effects on the margin to overfill. Lower RCS l temperatures would result in a higher break flow rate, but would shorten the time required for cooldown of the RCS. Conversely, higher RCS temperatures would result in a lower break flow rate, but would '

lengthen the time required for the cooldown of the RCS. Because the I

difference in density for a 5.50F temperature difference is relatively small, and the break flow rate is proportional to the square root of the density, the effect of the density on the break flow rate is relatively small. However, because the proposed change in RCS temperature is a relatively large fraction of the total temperature l

difference during cooldown (~700F), the higher T-avg has a much greater effect on the margin to overfill than a reduced T-cold.

The nominal cold leg temperature was used for conservatism with respect to the fluid density. The cooldown time was prolonged by cooling to a temperature a,t least 50F less than required by the ERCS to conservatively account for the RCS average temperature uncertainties.

Further conservatism was achieved because the cooldown rate at the end of the cooldown period is less than that at the start of the cooldown period, thus prolonging the cooldown period. The smaller cooldown rate was due to a lower steam generator pressure, and thus, a reduced steam flow rate from the intact steam generators.

8.3.4 RCS Flow Rate As discussed above, the RCS average temperature should be maximized in order to prolong the required time for the RCS cooldown. This temperature increase was effected through the use of the thermal design flow rate rather than the best-estimate RCS loop flow rate. Use of the thermal design flow rate resulted in a higher RCS average temperature for a given cold leg temperature.

8-6

I l

l l

8.3.5 Steam Generator Pressure l l

l The nominal steam generator pressure corresponding to the nominal cold leg temperature was used in order to maintain thermal-hydraulic consistency within the model.

8.3.6 Steam Generator Mass l

l An increase in steam generator mass would result in a decrease in the margin to overfill of the ruptured steam generator. Thus, the initial mass was increased to account for any mass additions to the system which were not explicitly modeled. The most significant mass additions are due to the postulated turbine runback, either manually or initiated by the Overtemperature N-16 (OTN-16) control system, and to an early )

OTN-16 reactor trip.

From the Nominal Base Case SGTR simulation (Chapter 6), it may be seen that an automatic turbine runback would be initiated at approximately 138 seconds due to the approach of the OTN-16 signal to within 3% of the OTN-16 setpoint. Assuming that the turbine runback prevented the reactor from tripping on the OTE-16 system, the reactor would trip on low pressurizer pressure at 357 seconds. This time difference corresponds to 219 seconds of turbine runback. At CPSES1, the turbine would be runback at an average rate of 10%/ minute. Using the average runback rate, the turbine (and reactor) power at the time of reactor trip would be transitioning toward 63% of nominal. It was also assumed that the level was maintained in the ruptured steam generator as the power was reduced, thus causing an increase in the steam generator mass.

The turbine runback, with automatic rod control action, was designed to prevent a reactor trip on OTN-16; therefore, with a depressurization rate similar to a SGTR event and automatic rod control action to reduce reactor power and maintain the appropriate average temperature, the OTN-16 reactor trip setpoint would not be reached, and the reactor would not trip on OTN-16 prior to the low pressurizer pressure trip.

8-7

Assuming no' turbine runback occurs, the reactor would be. tripped on OTN-16 (at 233 seconds for the Nominal Base Case). Using conservative assumptions discussed elsewhere, auxiliary feedvater would be. initiated immediately. Thus, the additional auxiliary feedwater injected in the ruptured steam generator due to an early OTN-16 reactor trip must be included in the initial steam generator mass. <

l For conservatism, the additions to the ruptured steam generator initial i mass due to both the potential early reactor trip scenario and the l turbine runback case were used. Therefore, the initial mass was increased over the nominal full power mass. If turbine runback were to )

occur, the steam generator mass would be 10,500 lbm greater than nominal at the time of reactor trip calculated for the Nominal Case.

If no turbine runback were to occur, and the reactor were to trip on OTN-16, the steam generator mass would be 12,000 lbm greater than l i

nominal at the time of the reactor trip calculated for the Nominal l

Case. To conservatively bound both postulated scenarios, a total of 22,500 lba was added to the nominal initial steam generator mass. Note that increasing the initial steam generator mass by 22,500 lbm results in an initial steam generator mass corresponding to a steady-state power of approximately 40%. I An early, manual reactor trip would not result in a more severe event with respect to the margin to overfill. Due to a lack of information (alarms, etc.), the reactor operators would certainly not trip the reactor prior to the OTN-16 turbine reactor trip signal. The initial steam generator mass has conservatively been increased to bound postulated scenarios which result in an early reactor trip and the associated increase in integrated auxiliary feedvater flow.

8-8

=. . ._

8.4 Protection System Setooints and Delav Times 8.4.1 Reactor Trio Delav Because the time for the initial operator action following the start of the SGTR was fixed through the simulator exercises, the earlier the reactor is tripped, the greater the integrated auxiliary feedwater flow into the ruptured steam generator. Therefore, no delay time was used between the time the trip signal was generated and the time the rods began to fall into the core.

8.4.2 Turbine Trio Delav Since continued steam flow would decrease the steam generator mass and cool the RCS, the turbine is assumed to trip simultaneously with the reactor, with no time delay.

8.4.3 Steam Generator Atmospheric Relief Valve Ooerability The break flow would be maximized if the rupt.ured steam generator pressure is minimized. A minimum steam generator pressure would be attained if the steam generator ARV vere used to maintain pressure in i the ruptured steam generator rather than the safety relief valves. The minimum pressure would result in a greater primary-to-secondary pressure differential, and hence, a greater breek flow rate.

Therefore, the ARV on the ruptured steam generator was used to control the steam generator pressure rather than the main steam safety valves.

In order to maximize the RCS temperature and pressure, the ARVs were not used to control the pressure in the intact steam generators. The intact ARVs were used to "cooldown" the RCS later in the transient.

8.4.4 Low Pressurizer Pressure for Reactor Trio As noted abore, the earlier the reactor trips, the less the margin to overfill. The .Tominal setpoint for a reactor trip on low pressurizer 8-9

l l

pressure is 1910 psig. The uncertainties associated with the pressure measurement;are 4.5% of the 800 psi span. For conservatism, an additional 1% of span was used to determine the low pressurizer pressure setpoint to be used in the SGTR analyses. Thus, a maximum setpoint pressure of 1969 psia was used. l 8.4.5 Reactor Trio gn Overtemoerature N-16 As discussed in Section 8.3.6, the effects of an early reactor trip on Overtemperature N-16 were bounded by the increase in initial steam generator mass. Therefore, the reactor trip on OTN-16 was not modeled.

t 8.4.6 Low Pressurizer Pressure for ECCS Actuation An earlier actuation of the ECCS would result in the least depressurization of the primary, and hence, the greatest break flow rate. The nominal setpoint for generation of a SIAS on Low Pressurizer Pressure is 1829 psig, or 1844 psia. Because no adverse containment conditions were anticipated during the SGTR event, the total uncertainties used for the low pressurizer pressure reactor trip setpoint were used for the low pressurizer pressure SIAS setpoint.

Therefore, the low pressurizer pressure setpoint for generation of an SIAS on low pressurizer pressure was 1888 psia.

8.4.7 Main Feedvater Isolation Main feedvater flow would be lost following the loss of offsite power due to a variety of reasons. The feedwater regulating valves are air-operated and fail closed when the control solenoids are de-energized. Following the loss of offsite power, the feedvater regulating valves would fail closed when the control solenoids were de-energized. In addition, the valves would drift closed as the instrument air pressure decayed following the loss of offsite power.

During normal operation, the main feedvater pumps draw steam from the moisture separator-reheater (MSR), which in turn draws steam from the high pressure turbines. After the turbine stop valves close following 8-10

reactor trip, there would be no steam supply to the MSR, and thus, the )

main feedvater pumps would begin to coast down.  !

Furthermore, there is a check valve located downstream of the feedpumps, and upstream of the feedwater isolation valve; thus, as soon as the feedvater pump discharge pressure fell below the pressure of the steam generator (anticipated to occur very quickly, on the order of a few seconds), the check valve would prevent fluid from flowing out of the steam generator. Thus, the main feedwater flow would effectively be isolated from the ruptured steam generator.

To conservatively model the loss of main feedwater, the main feedwater flow rate was assumed to be ratped down to zero flow linearly over a five second period, beginning five seconds following the loss of offsite power.

8.4.8 Auxiliary Feedwater System Initiation Delay l

In order to maximize the ruptured steam generator mass, the ,

l motor-driven and turbine-driven auxiliary feedwater pumps were assumed '

to deliver fluid to the steam generators simultaneously with the loss of offsite power. The time required to start and load the diesel generators was conservatively neglected.

1 8.4.9 Safety Iniection Initiation Delav I

In order to maximize the RCS pressure, the centrifugal charging pumps 1 were assumed to start immediately following the generation of a SIAS.

The time required to start and load the diesel generators was l neglected.

8-11

l l

\

8.5 System capacities 8.5.1 ECCS Flow Rates The ECCS flow rates were conservatively based on the design pump output plus 10% and included best-estimate frictional loss, elevation heads, etc.

8.5.2 Auxiliary Feedwater Flow Rates The turbine-driven and motor-driven auxiliary feedwater pumps were automatically started following the loss-of-offsite-power. The auxiliary feedwater flow rates from these pumps to the steam generators .

were maximized, based on design pump characteristics and the most conservative control valve configuration. Because the reactor operators would throttle the auxiliary feedwater flow into the intact steam generators to maintain level, the flow rate into these steam generators is not especially critical, as long as sufficient flow was available to cool the primary and not enough flow was added to initiate an overcooling event. The ERGS require that the narrow range level be maintained between 10% and 50%. In addition, the flow to the intact steam generators would be manually throttled during much of the SGTR analysis.

The flow rate to the ruptured steam generator, on the other hand, is more crucial. For the conservative model, in which there are no single active failures, the auxiliary feedwater flow to the ruptured steam generator was essentially the same as that to the intact steam generators prior to the isolation of the ruptured steam generator.

8-12 L.__

l l

l l 1

8.6 Control Systems 8.6.1 Pressurizer Pressure and Level Control Use of the pressurizer heaters prior to reactor trip would delay the reactor trip on low pressurizer pressure. Therefore, the pressurizer 1

heaters were not modeled. The sprays would not be used during a depressurization event. Therefore, the pressure control system was not used in the conservative model.

The operation of the level control system prior to reactor trip would increase the time until the reactor would trip on low pressure. These systems would not be available after the ss of offsite power and initiation of the SIAS (Phase A Isolation). Because an delay in the time to reactor trip is non-conservative, the pressurizer level control system is not modeled.

8.6.2 Steam Cenerator Water Level Control l

The steam generator water level control system was assumed to be in the '

automatic mode. The assumption of automatic feedvater control prior to reactor trip is conservative because a constant steam generator level l until the time of reactor trip would increase the time required for identification of the ruptured steam generator. If the feedwater controller were in manual, the steam generator water level would be '

under close observation, particularly during transient conditions (such as a sharp decrease in pressurizer pressure and level). The effect on the steam generator mass prior to reactor trip would be slight; however, the ruptured steam generator would be identified early in the transient, conceivably prior to reactor trip.

8.6.3 Turbine Runback / Automatic Rod Control Because of the depressurization of the RCS, the 0 rertemperature N-16 control systems would act to reduce the setpoint of the Overtemperature N-16 reactor trip. A turbine runback would be automatically initiated 8-13

when the setpoint neared the N 16 power. In addition, the automatic rod control would reduce reactor power (and consequently, the N-16 power signal) to maintain the appropriate average temperature. If offsite power were available, the steam dumps could open to further reduce the average RCS temperature. The feedwater control system would adjust the main feedwater flow rate to maintain a constant steam generator level. The net result would be that the mass in the steam generator at the time of reactor trip would be increased. Therefore, the assumption that turbine runback occurs is conservative with respect to the margin to overfill.

The turbine runback, and ensuing reactor power reduction, have a negligible effect on the primary parameters used to maximize break flow. The pressure in the RCS could vary somewhat due to the reduced steam flow followed by actuation of the automatic rod control to maintain RCS average temperature; however, the magnitude of the fluctuations would be relatively small. The decrease in RCS average temperature due to the lower core power would have only a marginal effect on the break flow rate and the cooldown time. No other significant effects were anticipated.

A turbine runback was not explicitly simulated in the SGTR analysis; however, the effects of the turbine runback were conservatively accounted for as described in the previous discussions concerning the initial steam generator mass (Section 8.3.7).

8.7 Deesy Heat Since continued heat addition from the core retards the i depressurization of the RCS, the inclusion of a 1.20 multiplier on the 1971 ANS decay heat curves is conservative. l l

l l

i 8-14 1

i

8.8 Summary Evaluations of those parameters other than active equipment failures l have been performed in order to determine conservative values of the parameters for the postulated SGTR event. The conservative values will be used in the evaluations of the postulated active failures (Chapter 9) and in the design-basis analyses described in Chapters 10 and 11.

i l

i 8-15 i

TABLE 8-1 Summary of Conservative Assumptions Critical Parameter Desinn-Basis Assumptiori Offsite Power Lost on Reactor Trip Initial Power 100% RTP ,

Initial RCS Pressure 2220 psia Initial Pressurizer Water Level 70% span Initial Cold Leg Temperature Nominal RCS Flow Rate Thermal Design Initial Steam Generator Pressure Nominal Initial Steam Generator Mass + 22,500 lbm Reactor Trip Delay None Turbine Trip Delay None Steam Generator Atmospheric Relief Valve Ruptured Loop Only Low Pressurizer Pressure for Reactor Trip 1969 psia Reactor Trip On OTN-16 Bounded Low Pressurizer Pressure for ECCS Actuation 1888 psia Main Feedwater Isolation On Reactor Trip Auxiliary Feedwater System Initiation Delay None Safety Injection Initiation Delay None ECCS Flow Rates Maximum Auxiliar* Feedvater Flow Rates Maximum Pressurizer Pressure and Level Control None Steam Generator Water Level Centrol Prior to Rx Trip Turbine Runback Bounded Decay Heat 120% 1971 ANS 8-16 1

CHAPTER 9 REQUIRED EQUIPMENT AND ACTIVE FAILURE EVALUATIONS 1

The following discussions concern the equipment and instrumentation which has been identified as being required for the successful l termination of a postulated design-basis steam generator tube rupture.

Because this equipment is required for termination of the tube rupture event, the effects of the postulated failure of this equipment must be considered. In the discussions that follow, the required equipment is I

described, followed by a evaluation of the effects of an active failure 1 of that component or system. In general, the required equipment (components and systems) is required to control the auxiliary feedwater flow, isolate the ruptured steam generator, and cool and depressurize the RCS. The postslated failures of these components were evaluated to determine the effect on the margin to overfill of the ruptured steam generator, the effect on the radiological dose consequences, and the effect on the operator response times described in Chapter 8. ,

l As described in Chapter 4,,the SGTR scenario includes a reactor trip,  ;

turbine trip, and safety injection actuation. A loss of offsite power )

is assumed to occur simultaneously with the turbine trip. Automatic signals to start all of the auxiliary feedwater pumps and the safety injection pumps will occur following the station blackout signal due to the loss of offsite power. By design, these trip and actuation systems are fully safety-grade systems and single failure proof; thus, no single active failure will prevent the reactor and turbine from tripping, nor the auxiliary feedwater and safety injection actuation signals from being generated.

9.1 Auxiliary Feedwater Throttline Valve The throttling valves between the turbine-driven auxiliary feedwater pump and the steam generator are locked fully open; thus, a fully open 9-1

failure of any of these valves is of no consequence. The throttling valves between a motor-driven auxiliary feedwater pump and its associated steam generators are normally at some position less than fully open. In addition, these valves automatically throttle to maintain a set discharge pressure from the motor-driven auxiliary feedvater pumps. If one valve were to fail fully open, the other would throttle closed to maintain the discharge pressure, thus further increasing the amount of auxiliary feedwater flow through the fully open valve. The maximum flow is obtained when the throttling (control) valve to the ruptured steam generator from the motor-driven pump fails fully open. In addition, based on the proximity of the isolation valve controls and the auxiliary feedwater pump controls to the throttling valve control and the valve position indication, it has been conservatively estimated that two minutes are required to identify that the throttling valve has not closed and to close the isolation valve or to stop the appropriate auxiliary feedvater pump.

The failure of this throttling valve results in a reduction in the margin to overfill of the ruptured steam generator. An analysis was performed to quantify the effect on the margin to overfill of the ruptured steam generator for this active failure. This failure has 1 only a negligible effect on the radiological consequences of a postulated tube rupture.

l l

9.2 Main Feedwater Isolation Valve '

During normal power operation, the steam supply to the main feedwater pump turbines comes from the moisture separator-reheater, which obtains its steam supply from the main turbines. Assuming that the turbine trips following reactor trip, the steam supply for the main feedwater pump turbines would be lost and the feedwater pumps would quickly begin to coastdown. As soon as the main feedvater pumps begin to coastdown, the discharge pressure of the pumps falls below the pressure of the steam generator; thus, significant quantities of main feedwater will not enter the steam generators following the reactor and turbine trips.

9-2 l

l

l \

l In addition, the feedvater regulating valve is a fail-closed, air-operated valve. This non-safety grade valve will close due to de-energizing of the control circuit due to a loss of offsite power.

Following the loss of instrument air, this valve will drift closed.

This valve vill also close on low T-avg coincident with reactor trip.

Therefore, the postulated failure to close of the feedwater isolation valve following the generation of a SIAS has no significant effect on the tube rupture event.

9.3 Miecellaneous secondary Steam Isolation valves The blowdown isolation valves and the process sampling valves are protected by redundant valves. The main steam drippet isolation valves, and some process sampling valves are downstream of the MSIVs.

The drippot isolation valves are not qualified as active valves and no credit is taken for their proper operation. The turbine-driven auxiliary feedvater pump steam admission valves are qualified as active valves. The steam flow rate through any of these valves is bounded by the steam flow rate through a stuck open ARV and would not have any significant impact on the operator response times assumed in the SGTR analyses.

9.4 Steam Generator Safety Valves Analysis indicates that these valves do not open during a steam generator tube rupture event; thus failure of the code safety valves is not a Concern.

9-3

l 9.5 Steam Generator Atmospheric Relief Valves (ARVs)

The steam generator ARVs are used for two distinct purposes in the termination of the SGTR event - 1) Isolation of the ruptured steam generator, and 2) cooldown of the RCS, It was assumed that all ARV block valves are open at the time of the event in accordance with CPSES1 administrative controls. The ARVs are classified as active valves and may be manually operated from the control room. The ARVs are air-operated valves and are equipped with air accumulators should offsite power be lost. The block valves are not currently classified as active and are only operable locally with a handwheel.

For Case 1), the ARV on the ruptured steam generator is postulated to fail open, thus maximizing the amount of radioactive steam released to the atmosphere. The block valves for the ARVs have manual operators which must be operated locally. Due to the postulated environment in the room where the block valves are located, it cannot be assured that these valves would be closed should an ARV fail to close. Thus, the failure of an ARV to close would result in the complete blowdown of that steam generator. A significant number of operator responses is required to terminate the primary-to-secondary break flow; thus, the recovery from this event scenario will require a considerable period of time. Based on the amount of radioactive steam released, this failure would result in a significant increase in the severity of the radiological consequences. Due to the long times required for termination of the event, it may be postulated that this failure could l adversely affect the margin to overfill as well.

For case 2), an ARV on an intact steam generator was assumed to fail to open. This fault will prolong the time required to cool the RCS, thus decreasing the margin to overfill. An analysis was performed to quantify the effect on the margin to overfill of the ruptured steam ,

8enerator for this active failure. l l

9-4 l

l t  !

9.6 Main Steam Isolation Valves The failure of a main steam isolation valve (MSIV) on an intact loop to close does not pose any problems, because the valves are not required to close to terminate the event. On the other hand, the failure of the MSIV on the ruptured loop to close will make termination of event more difficult. Downstream of the MSIVs and upstream of the turbine stop valves are a number of relatively small branch steam lines with non-nuclear safety-related isolation valves. Most of the isolation valves fail open following the loss of instrument air, thus providing a relatively large steam flow path from the ruptured steam generator.

The steam flow path allows the ruptured steam generator to depressurize in an uncontrolled manner. However, because the total steam leakage is less than the flow rate through a steam generator ARV, this potential failure is bounded by the failure of a steam generator ARV.

9.7 Pressurizer Power-Operated Relief Valve (PORV)

One pressurizer PORV is required to depressurize the primary to terminate the SGTR break flow. Should the PORV fail to close, the FORV block valve may be closed to terminate the primary depressurization with little impact on the operator response time. In addition, since only one FORV (and open block valve) is required for depressurization, the failure of one PORV (or block valve) has no effect on the margin to overfill or on the offsite dose. Also, these is only a negligible effect on the operator response times.

The pressurizer PORVs and block valves are active valves and may be operated manually from the control room in the event of a loss of offsite power. In addition, the PORVs are equipped with nitrogen accumulators to provide a motive force should offsite power not be available.

9-5

1 s

i 9.8 Pressurizer Safety Valve j These valves never open during the postulated SGTR event.

9.9 ECCS Pumo Stoo Switches Af ter the RCS has been depressurized to terminate the SGTR break flow, timely operator action is required to stop the ECCS pumps in order to prevent the break flow from restarting. To stop the ECCS pumps, the operator normally turns the appropriate pump stop switch to "0FF".

Should this fail to stop the pumps, the switch may be overridden by the "Pull-to-Lock" feature to manually stop the ECCS pumps. A negligible amount of additional time is required to terminate ECCS flow shon1d the pump stop switches fail.

9.10 SI Reset Device This device resets all components and systems "initiated" by the SIAS.

Because the operators manually reset those components and systems required for termination of the SGTR event, no significant effects or time delays are anticipated.

1 9.11 Electrical Failures A loss of offsite power is assumed to occur coincidently with the reactor trip. If one of the diesel generators were to fail to start, one train of ESF were unavailable, and the total ECCS flow rate would be signiticantly reduced. The reduced ECCS flow rate would allow the RCS to depressurize to a lower equilibrium pressure where ECCS matches break flow and result in a lower break flow rate.

9-6

l l

l 9.12 Instrumentation Failurgi Sufficient redundancy (number of channels) exists such that the single failure of an instrument channel has a negligible effect on the recovery from a postulated tube rupture event.

9.13 Summary 9.13.1 Marnin to Overfill Using the CPSESl-SGTR model described in Chapter 5 and the conservative assumptions identified in Chapter 8, analyses were performed to quantify the effect on the margin to overfill of two of the postulated single active failure 3. Ths failure of the throttling valve between a motor-driven auxiliary feedwater pump and the ruptured steam generator was determined to have a greater impact on the margin to overfill than the postulated failure to open of an ARV on an intact steam generator.

Therefore, the fully open failure of the auxiliary feedwater throttling 1 valve was the worst, single active failure with respect to overfill of 1 I

the ruptured steam generator.

9.13.2 Radiological Consecuences The failure of the ARV on the ruptured steam generator to close resulted in the most steam released directly to the atmosphere from the ruptured steam generator; thus, this failure was the worst, single active failure with respect to the offsite radiological doss.

9-7

l l

I l l

CHAPTER 10 DESIGN-BASIS SGTR EVENT l l

FOR THE WORST CASE FOR STEAM GENERATOR OVERFILL l

. As discussed in Chapter 9, it was determined that the failure of the chrottling valve between the motor-driven auxiliary feedwater pump and the ruptured steam generator would result in the least margin to overfill of the ruptured steam generator at the time of event termination. The Nominal Bete Case discussed in Chapter 5 was modified to include the conservative assumptions discussed in Chapter 8 and the worst single active failure identified in Chapter 9 to develop the overfill design-basis model. A complete description of the SGTR analysis inputs and the results using the overfill design-basis model is provided in this chapter, j A description of the effects of the single active failure on the operator actions and response times discussed in Chapter 7 is found in i i

Section 10.1. Section 10.2 contains a description of the modifications i to the conservative base model required to incorporate the single active failure. The analysis results are described in the remaining sections. )

i 10.1 Operator Actions and Response Times I

i The failure resulting in the fully open throttling valve between the motor-driven auxiliary feedwater pump and the ruptured steam generator would most likely not be noticed until the reactor operator was directed by the Emergency Response Guidelines (ERGS) to throttle the auxiliary feedvater to the ruptured steam generator in order to maintain the level to between 10% and 50% of the narrow range span. If the level exceeded 50% span, the operators would completely close the throttling valve, thus effectively terminating the auxiliary feedwater I flow to the ruptured steam generator. It was assumtd in the 10-1

design-basis analysis that the operator would notice that the throttling valve had failed fully open when, in accordance with the  !

ERGS, he was directed to throttle auxiliary feedvater to the ruptured j steam generator. In addition, it was conservatively assumed that the operator would require two minutes to realize what had happened and to terminate auxiliary feedwater to the ruptured steam generator. To terminate auxiliary feedvater to the ruptured steam generator, the operators could close the appropriate isolation valve, er should those i

valves also fail, the operators could stop the motor-driven auxiliary feedvater pump. The controls for these actions are all grouped together on the same control panel, such that the time required to terminate auxiliary feedvater to the ruptured steam generator is minimal.

The ERCS direct the reactor operttors to first isolate steam flow from the ruptured steam generator and then control / isolate the auxiliary feedwater to the ruptured steam generator. Therefore, it was assumed that the MSIV on the ruptured steam generator was closed at the same time as in the conservative base model, i.e., at thirteen minutes. In j addition, no credit was taken for the operator action to raise the setpoint pressure of the ARV on the ruptured steam generator. By I

neglecting this action, the ruptured steam generator was maintained at a lower value, thus increasing the primary-to-secondary break flow rate. Tvc additional minutes war. ~1sumed to allow time for the operator to close the appropriate auxiliary feedwater isolation valve; hence, thi auxiliary feedwater was assumed to be isolated at fifteen minutes following the start of the SGTR event. As in the conservative 1

base model, five minutes were allowed between the time the ruptured j steam generator was isolated and the start of the RCS cooldown.

Therefore, the RCS cooldown was assumed to start twenty minutes following the start of the SGTR. l l

l Two minutes following the termination of the RCS cooldown, the RCS was I depressurized using one pressurizer PORV. One minute after the RCS depressurization had been completed, the ECCS pumps were stopped and placed in standby, marking the end of the event.  ;

l 10-2 .

1

l Post-Event Ooerator Responses Immediately following the termination of the event, the RCS pressure would increase slowly due to the addition of decay heat and the lack of any active heat removal process (no steam relief from intact steam generators); hence, the break flow rate would tend to be l l

re-established. Therefore, additional procedural guidance has been l provided to the reactor operators to prevent the flow of a significant amount of RCS fluid into the ruptured steam generator. Following the termination of the ECCS flow, the reactor operators would continuously maintain equal pressures in the RCS and in the ruptured steam generator ,

l using the auxiliary spray, if letdown had been re-established by this time, or the pressurizer PORV. Following the initiation of a  !

controlled cooldown to bring the plant to cold shutdown conditions, the RCS pressure would be lower than the ruptured steam generator pressure, i thus again backfilling the RCS with the ruptured steam generator fluid. This process would be continued as the plant was cooled to cold shutdown conditions, thus preventing the liquid volume in the ruptured steam genereLor from increasing significantly at any time during th'e post-Steam Generator Tube Rupture event recovery.

1 10.2 Auxiliary Feedwater Flow Ratet j In order tu maximize the amount of auxiliary feedwater injected into the ruptured steam generator, an analysis was performed using the i auxiliary feedvater valve arrangement which results in the maximum auxiliary feedwater flow rate to the ruptured steam generator. For this '

valve arrangement, the throttling valves from the turbine-driven auxiliary feedwater pump to the steam generators were fully open. The throttling valve from the motor-driven auxiliary feedwater pump to the ruptured steam generator was assumed to fail fully open. The throttling valve to the other steam generator which is served by tne motor-driven auxiliary feedvater pump, partially closed automatically in order to prevent pump runout. The auxiliary feedvater flow rates were thus calculated to be a total of 1337 gpm at a back pressure of 10-3 w

l l

l l

1200 psia to the intact steam generator and 614 gpm at a-back pressure I of 1200 psia to the ruptured steam generator. The auxiliary feedwater model was pressure-dependent; hence, as the steam generator pressures decrease, the auxiliary feedvater flow rates increase and vice. versa.

10.3 Lesults Using the auxiliary feedwater flow rates and operator response times discussed above, a simulation of the postulated tube rupture event was performed using the RETRAN02 code. A sequence of events timeline is provided in Table 10-1. Time-dependent plots of the critical parametere are shown in Figures 10-1 through 10-10. A key to the symbols used in the figures is provided as Table 10 2. Detailed discussions of the plotted parameters are provided below.

10.3.1 General Descriotion As may b6 observed in Figures 10-1 through 10-10, following initiation of the SGTR, the RCS pressure and level decreased linearly until the low pressurizer pressure reactor trip setpoint was reached 5.4 minutes later. Turbine trip, loss of'offsite power, and the initiation of auxiliary feedwater flow all occurred simultaneously with the reactor trip. The main feedwater flow was isolated 10.0 seconds later (5.0 second delay with a 5.0 second closure time). The SIAS setpoint on low pressurizer pressure was exceeded 1.9 minutes after the reactor trip.

Following the turbine trip, the steam generator pressures rose rapidly, 1

and the ruptured steam generator ARV opened to relieve the pressur-s. l After auxiliary feedwater flow was initiated, the steam generator l

pressure was determined by the amount of cold auxiliary feedwater flow entering the steam generators.

Following the isolation of steam flow from the ruptured steam generator at 13.0 minutes, the pressure in the intact steam generators rose somewhat due to the loss of the steam relief path, although the magnitude of the increase was small due to the depressurisation caused 10 4

i l > N l

l 1

by the cold auxiliary feedvater. The pressure in the ruptured steam I generator continued to fall until the auxiliary feedvater flow to the >' l

~

ruptured steam generator was isolated at 15.1 minutes. The pressure ,

l then rose to and was controlled at the ret pressure of the ARV. The \ s RCS pressure began to rise as the ECCS flow initially matched and then  !

exceeded the break flow. This pressure increas'e was terminated when the RCS cooldown began at 20.1 minutes. The cooldown was initiated by opening the ARVs on the intact steam generators and was terminated when i a RCS subcooling margin of at least 150F at the ruptured steam generator pressure was obtained. Two minutes later, the RCS was depressurized by opening the pressurizer FORV. In accordance with the ERGS, the depressurization was terminated when:

1) the RCS pressure was lower than the rupy"red steam generator pressure,.and the pressurizer level was greater than 20% span; or,  ;

2) when the subcooling margin was less than 15(F; or, I

3) when the pressurizer level was greater than 78% span.

For this analysis, the reduced RCS pressure criterion was modeled to be satisfied when the RCS pressure was 100 psi less than the ruptured steam generator pressure. Because the ruptured steam generator pressure began to decrease when the RCS pressure was less the steam generator pressure, this 100 psi pressure difference was not attained, ;i and the depressurization was terminated on high pressurizer level. One minute later, the ECCS pumps were secured, marking the termination of the event. The analysis was extended, with no additional operatar responses modeled, until 46.7 minutes, i 10.3.2 RCS (Pressurizer) Pressure As shown in Figure 10-1, the RCS pressure decreased linearly with time until the low pressurizer pressure reactor trip setpoint was exceeded.

\

The turbine was assumed to trip sim'titane.ously with the reactor; however, since the turbine stop valves closed faster than the control 10 5 I

i

x . 1

/g rods fell into the reactor core, there was a brief spike in the pressure prior to the sharp drop-off following reactor trip. 3 o

Following reactor trip', the pressure briefly leveled off and even I

'/

(

increased as the cold leg, and therefore, the hot leg temperatures, I I( c

' increase following:'the turbine trip"and loss of main feedvater. The pressure then rehumed its decrease as the break flow drained the RCS 1

, s l untiltheSIAS[satpoint'onlowprassuriserpressurewasreached. The pressure stabilised as the ECCS wes inititted and then began to rise j

\ s i. i, 1 after the ruptured steam generator was isolated, and the ECCS flow rate ,

adjusted to match the break flow rate. The pressure fell rapidly

, during the 1nitial portions of the cooldown due to the RCS inventory 1

, shrinkage , but stabilized when the cooldown rate decreased as the\rpteam

! + 8 ,

generators depressurized. After the cooldown was terminated, the ,

pressure agald in' creased'.ba,cause the ECCS flow rate was greater than the break [Jow rara. When the PORV was\openad, the pressure fell l sharply until the RCS pressure became lent than the ruptured steam '

generator pressure. Until the PORV was closed, the pressure continued j to dscrease, but at a much slower rats., 'f;is change was caused by the l backfill and depressurization of the ruptured steam generator through the RCS. The stall pressure increase following the end of I.

depressurizationivas ,again caused by the ECCS flew into the RCS, a

Because the volumetric expansion of the RCS due to deposition of decay 1

heat compensated for the RCS, mass lost through the break, the prossure 1

l

(

was relatively s't4ble following the termination of the ECCS flow. The 1

/ l pressure would rerain at a slightly higher pressure than the ruptured j 3 s .s i

steam generator vntti operspor actions nere taken to reduce the l

, pressures in both systems.  !

(

+' k 10.3.3 Pressurizer Level As shown in Figure 10-2, the pressurizer level decreased linearly with time following the initiation of,tha SGTR. At the time of reactor trip, the level fell sharply, beganito recover from the drop until the heat removeJ by the secondary exceeded the heat addition in the primary s ,  !

)

10 6 L

\

f

s- .

' j j

is \

, j

\

N-) l and then continued to decrease as the break flow r$ntli-t.60.d. After the x >

iniciar.lon of the ECCS flow, the level stabilized, and then, following the icolation of th'e ruptured steam generkt.dr, N gru' t o yino os the I l

ECCS flow rate exceeded the break flow rate. ha level fell due' to shrinkage du?ing the RCS cor?.down and went off-scale (ielvhtowardthe end of nhe cooldown period. When the pressuriger Pnv was opened, the j level rose very rapidly one in part a.o the swll "ef f ect caused by the depressurization;hr.vever,amoreitapkttantcauseway the sharp rise in ECCS flow as the RCS was depres:,drized. After the RCS depressurization was terminated, the level briefly stabilized. Afted the ECCS flow was terminated, the level fell again, %c at a signific Ant 3y slower rate.

The level vonid . stabilize as the break flow was terminited and could rise slightly due to the deposition of decay heat into the RCS fluid from the reactor cers. l 1

1 10.3.4 Reactor ?p_'g;_ *.espons.1 Prior to reactor trip, the core poker decreased slightly as the RCS pressure decreased. The power reduction was due to the me,derator density coefficient, which intriduced negative reactivity into the core l as the RCS became slightly lest, dense. Subsequent to reactor trip, the reactor power was 120% ofthe '1971 ANS decay heat curve and was essentially unaffected by tas SGT1 transient. Boron injected though '

the ECCS prevented the core from becoming critical as the RCS was cooled, j

10.3.5 SteamCenerayjfyessure,Rosconse Common Resoorutts l l

i

/ r. shown in Figure 10-3,'the pressure responses of the steam generator vere approximately the tame prio'; to isolation of the ruptured steam generator. This effe.cc if dura tc the fact that the steam generatora l

are connected througt the preasure-equalizing, main steam hetader. i l

10-7

-l,

f ^, aN fi s 4

[ l

< , l Following the initiation of the tube rupture, the pressures in the  ;

steam generators decreased slightly due to the corresponding reduction

, .in core power. The pressure increased' sharply following the turbine trip; however, due to the opening of the ARV and the immediate injection of large amounts of cold auxiliary feedvater flow into the steam generators (Figure 10-7), the steam generator pressure did not reach the actuation pressure of any of the main steam safety valves.

Primarily because of the cold auxiliary feedwater flow, the-steam generator pressure decreased until the MSIV on the ruptured steam generator was closed. After the ruptured steam generator was isolated from the intact steam generators, the pressure responses of the steam generators diverged.

Ryotured Steam Generator Resoonse ,

Following the termination of auxiliary feedwater to the ruptured steam generator, the ruptured steam generator pressure increased toward the ARV setpoint. Because this pressure increase was driven primarily by the compression of the steam bubble as the steam generator liquid volume was increased due to the break flow, the rate of increase was ,

relatively slow. The pressure was maintained at the ARV set pressure until the RCS cooldown was initiated, at which time, the steam generator pressure very slowly drifted downward. Some of the pressure decrease due to the RCS cooldown was counteracted by the pressure increase caused by the steam compression in the steam dome. During the RCS depressurization, the RCS pressure became lower than the ruptured steam generator pressure, thus reducing the ruptured steam generator pressure. The pressure reduction was terminated when the pressurizer ,

I PORV was closed, ending the RCS depressurization. At that time, the 1

steam generator pressure stabilized at the RCS pressure. The ruptured l steam generator would be depressurized through the RCS as the reactor operators recover from the postulated tube rupture event.

\

Intact Steam Generator Responga

]

l After the auxiliary feedwater to the intact steam generators was 10-8

I 1

l l

throttled, the intact steam generator pressure stabilized. The RCS

\

cooldown, effected by opening the ARVs on the intact steam generators, j caused a rapid decrease in the steam generator pressure. As expected, l as the pressure decreased, the rate of change decreased; thus, the l temperature reduction was more rapid at the start than toward the end of the cooldown period. Subsequent to the termination of the cooldown, l l

the pressure began to recover, increasing toward the saturation i pressure corresponding to the RCS temperature throughout the remainder l of the event.

10.3.6 Break Flow Rate i

As shown in Figure 10-5, following the initiation of the event, the break flow decreased as the RCS pressure decreased. After the reactor 1

and turbine were tript and the ensuing RCS pressure decrease and steam generator pressure increase, the break flow rate dropped sharply. As  ;

the ECCS flow rate caused an increase in the RCS pressure and the cold 1 1

auxiliary feedvater flow caused a decrease in the steam generator pressure, the break flow rate gradually increased. The RCS cooldown 1

and resultant depressurization then caused the break flow rate to decrease; however, the magnitude of the decrease was reduced due to the increase in the dentity of the break fluid.

When the RCS was depressurized with the pressurizer PORV, the break flow rate fell sharply until the RCS pressure became less than the ruptured steam generator pressure (Figure 10-6). Due to the lower RCS l pressure, the break flow became negative, i.e., the flow direction 1 changed as tha ruptured steam generator backfilled into the RCS. Note that at this time, the flow rate in the RCS loop with the ruptured steam generator was essentially stagnant, with occasional periods of forward or reverse flow. During the periods of reverse flow, the fluid in the steam generator outlet plenum contained a high percentage of cold ECCS fluid, thus the density was relacively high, causing erratic changes in the SGTR break flow rate. Following the end of the depressurization and the termination of the ECCS flow, the break flow rate increased toward a stable flow rate of approximately 7 lbm/sec.

10-9

This positive flow was due to the slightly higher RCS pressure (primarily due to decay heat). l 4

1 Operator actions in accordance with recovery procedures would be required to continue the RCS cooldown to prevent the break flow from l l

becoming re-established. I l

l 10.3.7 Auxiliary Feedwater Flow Rates l l

The time-dependent auxiliary feedwater flow rates are shown in ,

Figure 10-7. Prior to isolation or throttling of the auxiliary feedwater, the flow rates increased as the respective steam generator pressures decreased, and vice versa. Because auxiliary feedvater was assumed to start instantaneously following reactor trip, the initial flow rate was high due to the relatively low steam generator pressure. l However, because the turbine trip occurred sLmultaneously with the initiation of auxiliary feedwater, the duration of the high flow rate j wac very short as the steam generator pressure rapidly increased. The flow rate decreased and then began to increase again as the pressures in the steam generators fell due to the opening of the ARV on the ruptured steam generator and the cooling effect of the auxiliary feedwater.

The auxiliary feedwater to the intact steam generators was throttled 1

beginning at 13.0 minutes. Thereafter, the flow rates were manually I 1

controlled to maintain a level of approximately 35% of span, i

l As described above, no operator responses to control auxiliary feedvater to the ruptured steam generator were credited until two additional minutes had elapsed. Thus, the auxiliary feedwater to the ruptured steam generator continued unthrottled until 15.08 minutes following the initiation of the tube ruptura.

10-10

l 10.3.8 RCS Temoeratures The temperature res'ponses at tha core exit and in the RCS cold legs following the postulated steam generator tube rupture event are shown in Figure 10-8.

The core exit temperature is a function of the average core flow rate, the average core inlet temperature, and the power generation rate in the core. Note that due to the higher flow rates and greater number of intact loops, the impact of the ruptured loop on the average cote inlet temperature is significantly less than that of the intact loops.

Core Exit Temnerature Following the start of.the SGTR event, the core exit temperature decreased slightly due to the reduction in core power caused by the positive moderator density coefficient. Because the turbine trip was conservatively assumed to occur simultaneously with the reactor trip, the temperature briefly spiked prior to the insertion of a significant amount of negative reactivity into the core from the reactor trip.

Following rea: tor trip, the core exit temperature fell as the power generation rate was sharply reduced, but began to rise again as the core flow rate dropped after offsite power was lost and the core inlet temperature increased following the loss of main feedvater. This increase was reversed after the injection of cold ECCS fluid, which reduced the core inlet temperature somewhat. The core exit temperature then began to drift slowly downward as the natural circulation flow was established. The temperature increased slightly following the isolation of auxiliary feedvater to the ruptured steam generator.

As ruptured loop flow rate decreased and eventually reversed, the temperature difference between the hot and cold legs increased due to the lower core flow rate. Following the termination of the ECCS flow, the core exit temperature rose fairly rapidly as the core inlet temperature increased.

Once natural circulation flow conditions were restored, the core exit temperatures tracked the cold leg temperatures.

10-11

l Cold Len Temoeratures l

Following the start of the tube rupture event, the cold leg temperatures decreased slightly as the hot leg temperatures decreased due to the reduced core power. After the reactor trip, turbine trip, loss of main feedwater and coolant pump coastdown, the cold leg temperatures increased. The injection of auxiliary feedwater and the opening of the ARV stopped the cold leg temperature increase.

Following the initiation of the ECCS flow, the cold leg temperatures decreased as the cold ECCS fluid mixed with the warmer RCS fluid.

Because the RCS flow rates were small, the ECCS flow had a more pronounced effect on cold leg temperatures than if the RCPs were operating. The cold leg temperature in the ruptured loop dive.rged from , i those in the intact loops after the auxiliary feedwater to the intact steam generators was throttled.

Following the complete isolation of the ruptured steam generator, the cold leg temperatures of the intact loops were stable due to the stable pressure in the intact steam generators. The cold leg temperature in the ruptured loop fell as the pressure in the ruptured steam generator continued to decrease due to the blowdown through the open ARV. The temperature then increased following the termination of auxiliary feedwater to that steam generator. After the RCS cooldown was I initiated, the cold leg temperature in the intact loop fell as expected. The temperature rose again following the end of the cooldown as the steam generator pressure increased. Following the termination of the ECCS flow and hence the loss of relatively large amounts of cold ECCS fluid, the calculated cold leg temperatures rose sharply toward the saturation temperature of their respective steam generators. The

. temperature then began to rise slowly as the decay heat from the core was a6ded to the RCS.

l Following the termination of auxiliary feedwater to the ruptured steam generator, the cold leg temperature of the ruptured loop rose as the ruptured steam generator pressure increased toward the setpoint i 10-12 1

1

l I

l pressure of the ARV. After the RCS cooldown began, the flow rate in  !

the ruptured loop fell. As the loop flow rate fell, the relative contribution of the ECCS fluid to the average cold leg temperature increased, thus the ruptured loop cold leg temperature continued to decrease as the loop flow rate decreased.

Near the end of the RCS cooldown, reverse flow was established in the ruptured loop. Due to the influx of a greater amount of RCS fluid relative to ECCS fluid, the cold leg temperature began to rise; however, this increase was again reversed when the RCS cooldown was terminated and the flow rate into the ruptured loop agaXn approached zero. During the depressurization, the cold leg temperature increased or dropped in accordance with the relative amount of RCS flow in the ruptured loop. Following the termination of the ECCS flow, the cold leg temperature increased toward the cold leg temperature of the intact loops.

10.3.10 RCS Loop Flow Rates As shown in Figure 10-9, the loss of offsite power causes the RCS flow rates to decrease toward flow rates consistent with natural circulation  :

conditions. After the MSIV on the ruptured steam generator was closed and auxiliary feedvater to the intact steam generators was throttled,  !

the flow rate in the intact loop stabilized at approximately 5% of full flow. The flow rates in the ruptured loop remained higher than those l j

in the intact loops due to the continued addition of auxiliary feedvater. Following the termination of auxiliary feedvater to the the ruptured steam generator, the ruptured loop flow rate decreased slightly to a stable value. This flow was held constant by the small steam relief through the ARV on the ruptured steam generator. The flow rate in t a intact loops was maintained by the addition of auxiliary feedwater to the intact steam generators. l Following the initiation of the RCS cooldown, the flow rate in the intact loop increased slightly as the natural circulation driving head 10-13

was increased due to the increased steaming in the intact steam generators.

Because of the asymmetric cooldown, the flow rate in the ruptured loops fell toward zero and eventually reversed direction. This flow reversal was stopped following the end of the RCS cooldown. The RCS depressurization resulted in a slight disruption of the natural circulation flow due to the effects on the driving pressure head, which included voiding in the U-tubes of the ruptured steam generator'.

The RCS depressurization also caused several erratic changes in the flow direction of the ruptured loop flow. The direction changes corresponded to the backfill of saturated fluid from the ruptured steam generator. Because this fluid was warmer than the RCS fluid, the RCS fluid density decreased, resulting in the formation of voids which collected in the U-tubes. The formation of the voids changed the pressure distributions and the loop flow rate. The new flow rate allowed colder ECCS fluid into the steam generator outlet plenum, which lowered the pressure, and increased the break flow rate. This cycle was then repeated.

Following the termination of the ECCS flow, natural circulation flow continued in the intact loops, with a small reverse flow rate in the ruptured loop.

10.3.10 Ruotured Steam Generator Liould Volume The collapsed liquid volume (Figure 10 10) was used to determine the margin to overfill of the ruptured steam generator. Due to the action of the Steam Generator Water Level Control System, the liquid volume in the ruptured steam generator remained approximately constant until the reactor and turbine were tripped. At that time, the volume increased sharply until main feedwater was lost. The addition of large amounts of auxiliary feedwater continued the volume increase, but at a much slower rate. Following the isolation of the auxiliary feedvater to the ruptured steam generator, the rate of volume increase was again reduced and became a function of the break flow rate. A slight perturbation in 10-14 l

the level increase was caused when the ARV opened slightly beginning at 14 minutes and continued to relieve small amounts of steam for approximately 10 minutes. .

The volume in the ruptured steam generator decreased briefly when the RCS was depressurized to a lower pressure. Following the end of the depressurization and the termination of the ECCS flow, the liquid volume in the ruptured steam generator stabilized and the event was terminated.

1 The analysis was stopped after the ECCS flow was terminated. The j additional operator responses, discussed in Chapter 7, ensure that the liquid volume would not have increased significantly, and would in fact decrease as the plant was brought to cold shutdown conditions.

10.3.12 Avtilable Marrin to overfill The total steam generator volume is 5954 ft3 The maximun liquid volume in the rupture steam generator was 5634 ft3, resulting in a margin to overfill of 320 ft3 I

l 10.4 Summary The postulated design-basis steam generator tube rupture event was analyzed using conservative operator response times, a conservative set of initial conditions, conservative assumptions concerning equipment availability, and the single active failure which resulted in the least margin to overfill of the ruptured steam generator.

10-15 l

\

I l

Despite the consideration of all of these conservative assumptions, the analysis results indicate that approximately 320 ft3 of steam space is available prior to the overfill of the ruptured steam generator.

'l l

4 l

e l

i I

l i

1 10-16

l TABLE 10-1 Design-Basis SGTR - Steam Generator Overfill Event Timeline Time Event 5.00 sec. Begin SGTR 5.05 min. Reactor trip on low pressurizer pressure, turbine trip, AFV initiation, loss of offsite power 5.13 min. Begin Main FW isolation 6.96 min. Low pressurizer pressure - SIAS 13.08 min. Operator Action - Close MSIV, Loop 4 Throttle AFW, Loop 1 15.08 min. Isolate AFW, Loop 4 20.08 min. Pagin Maximum rate RCS cooldown 31.38 min. End Maximum rate RCS cooldown 33.38 min. Begin RCS depressurization to refill pressurizer 36.67 min. End RCS depressurization to refill pressurizer 37.67 min. Terminate ECCS flow 40.00 min. End transient simulation 1

i 10-17

TABLE 10-2  !

Design-Basis SGTR - Steam Generator Overfill i Key to Figures 10-1 through 10-10 l 1

Foint Event j l

l A Reactor trip on low pressurizer pressure, j turbine trip, AFV initiation, loss of offsite power I

B Low pressurizer pressure - SIAS i

C Operator Action - Close MSIV, Loop 4, Throttle AFW, Loop 1 D Isolate AFW, Loop 4 )

)

Begin maximum rate RCS cooldown E

F End maximum rate RCS cooldown C Begin RCS depressurization to refill pressurizer 11 End RCS depressurization to refill l pressurizer J Terminate ECCS flow K End transient simulation i

10-18

I PRZ35UR!ZDt PRt3SURE 2.4

8 2.2 - A D

a c

" I 2- l 1J -

1.5 -

I E7 1.4 -

"I a EI 1.2 -

RI O$ 1-5 * .

o.a -

c.s -

c.4 -

c.2 -

I o, . . s 4 .

o 20 do TIME ( m )

FIGURE 10-1. CPSES1 Design-Basis SGTR, Overfill - Pressurizer Pressure PRE 35UR:ZDt LEVII. INDICATION too 1 1

90 -

1 ao - l l

70 -

$ so -

li 5 so -

.a

$.J 40 -

4

" l 20 - , g I

C D 20 - N( I I T

10,-

o s s s e o 20 4o

- T1MC (Wp0 FIGURE 10-2. CPSES1 Design-Basis SGTR, Overfill - Pressurizar 1.evel 10-19

,, sTDM coecuTat PRtssuRts j s

1.2 - C u

1 1.1 -

O w L.---

6

/

, 1-%

6- \ "

b5 c.9 -

w!

s.

Ni o4 -

E#

A c.7 -

c.s - J c.5 - l o.4 , , , ,

o 20 4o TWC(wen l

FIGURE 10-3. CPSES1 Design-Basis SGTR. Overfill - SG Pressures l l

3g s/C NARROW RANGC L.CVD. NDtCADON so - l 1

I so - 1 I

ro - o l m '

I I li So -

s

)

5 so -

1 e

3 ao -

E F so - ,

t 20 -

, to -

0 a s , ,

0 m 4 TWC (um) mAct - m FIGURE 10-4. CPSES1 Design-Basis SGTR Overfill - SG Levels 10-20

SSTR BREAK FUlM fuTE So

.j .

4 70 =

4 sO '- A I D

m so - \'* I Y F 8 S l a

s a- o i 4

so -

=

E 20 -

J d to -

o M

-20 . i i i o 2 #

TWC haN) q FIGURE 10-5. CPSES1 Design-Basis FGTR, Overfill - SGTR Break Flow Rate i

PRZR & 3/G PREISURES 2'

s ,

i 22" A g g

2- r

l 1 0

1A-N '

l 1.s -

I ET 1.4 - I

.> w* j E

\q E

M-i l o.s -

o.4 -

I o -

3 5 h E h i o 20 40

) ist (use

- Pan nuPfunn as FIGURE 10-6. CPSES1 Design-Basis SGTR, Overfill - Przr and SG Pressures 10-21 4

I l i

4 i

l

AlbuuMtf FYZDWATUt FLDW Raft 3 Put 3/C A 0 if it 700 = j l

soo -

E c M

Soo - ,

a.

S 4oo -

tf f soo -

y 20o -

10o -

o , a 4 .

0 2o ao TtWC (WW)

FIGURE 10-7. CPSES1 Design-Basis SGTR, Overfill - Aux W Flow Rate core DUT AND COLD LEC TDPC%TURE3 sso i, e

'." - Yw , I ,

sso - [

r soo - D r c b l N*

W 4so -

E f

, soo -

3 sso -

soo -

250 -

l 200 i , , ,

J o N g TWC (um)

t-oci - Tc-mTAct re-auPfumEn FIGURE 10-8. CPSES1 Design-Basis SGTR, Overfill - RCS Temperatures 10 22

NORWUZID RCS'LDOP FLDW Raft 3 25 I

20 -

1 15 - 1 5

S l

b 10 - i b I C C o g .

W I =3 M

5- ,

o d

!a V F

-10 . , . .

0 20 40 TWC(ust)

FIGURE 10-9. CPSES1 Design-Basis SGTR, Overfill - RCS Loop Flow Rates Ruptunto s/c Ltnao vetuut Tor or sTEAW ctHERATon EE _

o ,

4-f E;

• l w

= =s 3-

33 oe 1

I h8  %,

o i j 2-1 1

1-0 i i i i 0 20 40 TWC (uui)

FIGURE 10-10. CPSES1 Design-Basis SGTR. Overfill - Ruptured SG Volume

~10-23

CHAPTER 11 DESIGN-BASIS SGTR FOR THE WORST CASE FOR RADIOLOGICAL DOSES As discussed in Chapter 9, it was determined that the failure of the atmospheric relief valve on the ruptured steam generator results in the worst offsite radiological dose consequences. The thermal-hydraulic aspects of the postulated event are discussed below. The radiological dose consequences from this event are presented in Chapter 12.

A discussion of the conservative assumptions, as applicable to the dose consequences analysis, is presented in Section 11.1. Section 11.2 describes the effects of the single active failure on the operator actions and response times discussed in Chapter 7. The thermal-hydraulic analysis results are described in Section 11.3.

11.1 Conservative Assumotions with Resoect to Radiological Doses ,

i Because the objective of this analysis was to maximize the radioactive i release to the at;osphere, the conservative assumptions incorporated into the conservative model, were reevaluated to ensure that those assumptions remained valid when the radiological consequences were to be maximized.

As stated previously, the failure to close of the atmospheric relief valve on the ruptured steam generator has been identified as the single active failure which resulta in the greatest amount of steam release to the atmosphere from the ruptured steam generator. The ARV is assumed l co remain open until the plant is brought to cold shutdown conditions.

Thus, the conservative directions for plant parameters should ensure that the primary-to-secondary break flow is maximized and that the maximum amount of steam is relieved to the atmosphere. An implication of this latter goal is that the ruptured steam generator mass should 11-1

be such that long periods of steam relief may be sustained. Thus, the criteria used to determine the conservative directions of initial conditions for the' design-basis case for radiological doses are the same as those used for the design-basis case for margin to overfill.

Hence, the assumptions concerning initial conditions and equipment availability in the conservative base model (which minimized the margin to overfill) are applicable for the design-basis case for radiological dose, 11.2 Operator Actions and Response Times 11.2.1 Actions Standard SGTR Recovery Responses (0 to 40 minutes)

Due to the failure of the ARV to close, the operator actions required to terminate this tube rupture event would be considerably different from the standard event termination procedure discussed in Chapter 4.

The ARV failure to close would not be noticed by the reactor operators until after the ruptured steam generator had been identified and isolated; however, recognition at this stage would be contingent upon the steam generator pressure being less than the setpofat pressure of the ARV. Because the ruptured steam generator would be isolated from the other steam generators, and the hot RCS fluid would continue to leak into this steam generator, it can not be assured that the pressure would be less than the ARV setpoint pressure at the time of isolation; however, because the ARV vould be fully open, the pressure could not be maintained for long solely by the break flow. Thus, the failure of the ARV to close should be readily identifiable by the reactor operators shortly after the ruptured steam generator is isolated.

To conservatively prolong the assumed response times, the assumption was made that the valve position indications were inaccurate or went unnoticed by the operators. The CPSES1 Emergency Response Guidelines (ERGS) [13] specifically direct the operators to look for a faulted 11-2

l steam generator by verifying that the steam generator pressure in any steam generator is not decreasing in an uncontrolled manner.

Because the pressure in the ruptured steam generator would be decreasing, ostensibly without reason, the Shift Supervisor would transition to the FAULTED STEAM GENERATOR ISOLATION procedure. The first steps either require a cursory review of selected plant parameters or would have been performed in earlier procedures. The "Response Not Obtained" action is to dispatch an operator tc locally close valves or block valves. Personnel from CPSES Operations have indicated that because of the potentially adverse environment created by the steam release through the ARVs, operators would not be sent into the steam tunnels to locally isolate the valves; hence, no action would be taken to isolate the failed ARV. Because the event had already been identified as a SGTR, the operators would then return to the STEAM GENERATOR TUBE RUPTURE procedure. ,

1 In the SGTR Emergency Operating Procedure (EOP), the operators check the ruptured steam generator pressure. If the pressure in the ruptured steam generator is less than 605 psig, the transition is made to a contingency action procedure. Analyses have indicated that at this '

time (approximately 23 minutes), the ruptured steam generator pressure would be significantly greater than 605 psig; therefore, no transition would be made, and the reactor operators would initiate a cooldown of the RCS at the maximum rate using the ARVs on the intact steam generators. This RCS cooldown would be terminated when the core exit temperature reached was approximately 4600F. Based on the continuing pressure decrease in the ruptured steam generator, the operators would l l

then transition to an alternate reccvery procedure,' SGTR VITH LOSS OF '

REACTOR COOLANT - SUBCOOLED RECOVERY REQUIRED. j l

Subcooled Recovery Continnency Responses (40 - 90 minutes) i i

After the transition to the subcooled recovery E0P, the operators would begin a controlled cooldown of the RCS at a maximum rate of 1000F/hr in the intact loop cold legs. The ARVs on the intact steam generators 11-3 i

i l

l 1

would again be used for the cooldown. After the controlled cooldown  !

was initiated, the operators would depressurize the RCS to refill the pressurizer to at l' east 20% span. The E0P then instructs the operators to determine if a centrifugal charging pump (CCP) and a safety injection pump (SIP) could be stopped.

Following the rapid repressurization of the RCS due to the ECCS flow, the subcooling requirements for stopping one CCP and then one SIP would be satisfied shortly after the RCS depressurizaticn. The proced'res u require that one CCP be stopped, the pressure allowed to stabilize, and then the SIP be stopped. In order to maximize the break flow, it was assumed that both pumps were stopped simultaneously at the estimated time the SIP would have been stopped. The operators would then cycle back through this procedure, maintaining a minimum pressurizer level, until a sufficient subcooling margin had been established, which would allow an additional SIP to be stopped. The controlled RCS cooldown and depressurization would be continued until the conditions required for placing the Residual Heat Removal System (RHRS) in service were established or until the plant parameters became such that other contingency procedures were required.

Saturated Recovery Contingency Responses (after 90 minutes)

The Shift Supervisor would transition from the Subcooled Recovery E0P when the ruptured steam generator narrow range level exceeded 98% span. In accordance with the procedure, when the narrow range level in the ruptured steam generator exceeded 98% span, the Shift Supervisor should consult with the Technical Support Center and transition to another alternate recovery procedure SGTR VITH- LOSS OF REACTOR COOLANT - SATURATED RECOVERY REQUIRED. This contingency procedure is designed to reduce the SGTR break flow rate by reducing the number of CCPs and SIPS in operation and by intentionally depressurizing the primary to the saturation pressure at the core exit

• emperature.

I 11-4 i

In accordance with this procedute, it would then be determined that the subcooling requirements for stopping the second SIP vere satisfied at the time of the transition. After the'second SIP was stopped, the pressurizer PORV would then be opened to depressurize the RCS to the saturation pressure at the core exit temperature. The depressurization would be stopped if the pressurizer level exceeded 78% of span, or when the saturation pressure was attained. The remaining procedural steps require the operators to continually assess the plant conditions as the cooldown continues. The pressurizer PORV would be opened at intervals to maintain the RCS st the saturation pressure at the core exit temperature.

When the RCS conditions were suitable, the RHRS could be placed in service; however, because the RERS would allow cooldown of the RCS at the maximum allowable rate and because steam relief from the intact steam generator.s was not in itself sufficient to maintain the 1000F/hr cooldown rate, no credit was taken for the initiation of the RHRS until two hours following the initiation of the tube rupture. Operation of the RHRS was credited for maintaining the maximum cooldown rate after two hours, 11.2.2 Response Times The major operator responses are included with the event timeline presented in Table 11-1.

To ensure that reasonably conservative operator response times were used, it was assumed that following the isolation of the ruptured steam generator, fivu minutes were required for the Shift Supervisor to transition to the FAULTED STEAM GENERATOR procedure, return to the STEAM GENERATOR TUBE RUPTURE procedure, and then to begin the maximum rate RCS cooldown. This five minutes was added to the five minutes shown in Chapter 7 to be required following the isolation of the ruptured steam generator and prior to the initiation of the RCS cooldown. Following the completion of the RCS cooldown, it was conservatively assumed that an additional five minutes would be 11-5 1

i I

required for the transition to the first cont i ngency recovary procedure j and for the initiation of the controlled cooldown. Two minutes were 1 assumed to elapse prior to the initial RCS depressurization, and then, l a minimum of two minutes between consecutive depressurizations was assumed. If the pressurizer level had not fallen below 10% span, no l depressurization would be modeled. Finally, it was assumed that one CCP and one SIP were stopped five minutes after the start of the i initial depressurization.

The transition into the second contingency recovery procedure was modeled by stopping the second SIP at ninety minutes into the event.

Analysis indicated that the ruptured steam generator level vent offscale high at approximate).y 80 minutes; however, for conservatism, it was assumed that the transition va.: not made until 90 minutes. The initial depressurization to saturate the primary was conservatively assumed to occur five minutes later, at 95 minutes.

l 11.3 Results j i

I A time sequence of events is shown in Table 11-1. Plots of critical parameters are shown in Figure 11-1 through Figure 11-8, and a key to the symbols used in these figures is shown in Table 11-3.

I As discussed in Section 11.1, because of the myriad of operator l responses and the operational goals for the particular phases of the j event, the discussion of the critical transient parameters is divided into three phases:

1) Standard SGTR Termination Responses (0 to 40 minutes); i

2) Subcooled Recovery Contingency Responses (40 to 90 minutes);

and,

3) Saturated Recovery Contingency Responses'(after 90 minutes).

11-6

11.3.1 General Descriotion Standard SGTR Termination Responses (0 to 40 minutes)

I Following the reactor trip and ensuing turbine trip, the ARV on the i ruptured steam generator was modulcted fully open on a pressure error signal. The steam relief and the large amounts of relatively cold auxiliary feedvater acted to quickJy reduce the secondary pressure.

Normally, the ARV vould close as the pressure fell below the setpoint pressure of the ARV; however, for this case,it was assumed to remain open. Thus the secondary ptessure continued to fall. After the reactor operator actions to close the MSIV on the ruptured steam generator and throttle auxiliary feedvater to the i.. tact steam generator are completed, the intact steam generator pressure began to rise as the cold auxiliary feedvater was varmed (level swell), thus compressing steam bubble in the steam generator. After the MSIV on the ruptured loop was closed, only the ruptured steam generator continued to depressurize and at a slightly faster rate than before. The narrow range level in the ruptured steam generator, which had been rising due to the overfeeding of the steam generator by the auxiliary feedvater,  ;

fell off sharply as the inventory being steamed away exceeded that being added through the ruptured steam generator tube. The level in the intact steam generator was maintained near 35% of span (the EOPs '

direct the operators to maintain the level between 10% and 50% narrow range span).

The transition from Phase 1 to Phase 2 was considered to be made during the five minutes between the end of the maximum rate RCS cooldown and the beginning of the controlled rate cooldown.

Subcooled Recovery Contintency Responses (40 to 90 minutes)

This phase of the SGTR termination began when the Shift Supervisor transitioned from the STEAM GENERATOR TUBE RUPTURE procedure to the SGTR VITH LOSS OF REACTOR COOLANT - SUBC00 LED RECOVERY REQUIRED procedure. This transition was the first significant deviation from 11-7

the standard SGTR termination responses and occurred between the end of the maximum rate RCS cooldown at 39.97 minutes and the start of the controlled rate RCS cooldown at 44.97 minutes. As described previously, the first operator response was to initiate a controlled cooldown of the RCS using the intact ARVs. Ideally, the ARVs would have been modeled to modulate to the position which would relieve sufficient steam to maintain a 1000F/hr cooldown rate. However, because of the long lag between the time a modification of the ARV open area was made and the time a change in the cold leg temperature cooldown rate could be observed, the ARV control model developed for this event modulated the ARV from essentially full open to full closed. Because the continuous modulation is not realistic from an operations viewpoint, a minimum value of 65% open was arbitrarily set beginning at 60 minutes. The ARV control model then modulated the valve as necessary to maintain the cooldown rate.

Once the controlled cooldown was initiated, the procedures direct the operators to depressurize the RCS to refill the pressurizer to greater than 201 span. After the pressurizer was refilled, the one CCP and then one SIP was stopped. The reduced ECCS flow resulted in a lower equilibrium pressure; hence, the RCS pressure trended toward this new pressure. Because the steam generator pressures were relatively low, the 1000F/hr cooldown rate could not be sustained. As the ruptured steam generator pressure continued to decrease (but at a much slower rate), the equilibrium pressure described above was easily maintained.

Due to the relatively low steaming rate from the ruptured steam generator, and the higher SGTR break flow rate, the steam generator level indication increased significantly during this period, eventually going off-scale (high) at approximately 80 minutes. Once the level had exceeded 98% span, the contingency procedures directed the Shift Supervisor to transition to the Saturated Recovery procedure, thus ending this phase of the transient.

11-8

1 l

1 l

I l

Saturated Recovery Continnancy Resoonses (after 90 minuttgl l l

1 The final phase of the SGTR termination began when the Shif t Supervisor transitioned from the subcooled recovery procedure to the saturated i recovery procedure. The transition was made when the narrow range level indication in the ruptured steam generator became greater than l 98% span. Even though the analysis indicated that the level exceeded 98% span at approximately 80 minutes, for conservatism, the transition was not made until 90 minutes. The operational goal in the saturated I

recovery procedure was to minimize the break flow rate by minimizing the primary-secondary pressure differential. The pressure minimization was accomplished by reducing the ECCS capacity and depressurizing the RCS to near saturation. The subcooling criteria provided in this procedure for stopping ECCS pumps less stringent than in the subcooled recovery procedure, thus an additional SIP was stopped at 90 minutes.

After the second SIP was stopped, the break flow rate was greater than the capacity of the single operating CCP, thus the pressure and level immediately began to fall as a new equilibrium pressure was sought.

Five minutes after the SIP was stopped, the pressuriser PORV was re-opened to depressurize the primary to the saturation pressure at the temperature indicated by the core exit thermocouples. The depressurization was halted when the desired pressure was attained.

This procedure was repeated until RRRS entry conditions were satisfied. Even though the RRRS entry conditions were satisfied at approximately 110 minutes, for conservatism, no credit for !cs operation was taken until afte- 120 minutes.

11.3.2 RCS (Pressurizer) Pressure l

Standard SCTR Termination Resoonses (0 to 40 minutes)

Following the initiation of the tube rupture, the pressurizer pressure (Figure 11-1) decreased linearly due to the steam bubble expansion as the RCS was drained through the steam generator tube. The pressure dropped sharply following the reactor trip on low pressure; however, there was a brief pressure spike prior to'the decrease. This spike was 11-9

4 5

due to the quicker response of the turbine stop valve closure relative to control rod motion followins the simultaneous reactor and turbine trips. Following the post trip drop, the' pressure rose slightly as the cold leg temperature rose, thus raising the hot leg temperature and pressurizer pressure. The continued SGTR break flow caused the pressurizer pressure to continue.to fall, but at a slower rate. The break flow rate decreased following the post trip RCS pressure drop and the steam generator pressurization, but as the steam generators ,

depressurized due to the open ARV and the, auxiliary feedwater, the ,

break flow increased, thus increasing theirate of RCS depressurization. With the inittacion of the ECCS flow, the not rate of inventory loss from the RCS was greatly reduced, resulting in a reduced rate of depressurization. One additional cause for the change in the depressurization rate was the effect of the growth of the pressurizer steam bubble, which caused a given change in the RCS inventory to become less apparent.

1 Following the closure of the MSIV on the ruptured loop and the '

4 throttling of the auxiliary feedwater flow to the intact steam generators, the RCS pressure decrease was halted. The cold leg 't temperat-re increase in the intact icopa, due to increases in the' intact steam generator saturation temperature, counteracted the continued temperature decrease in the ruptured loop cold leg temperature; hence, the hot leg temperature stabilized as the core inlet temperature stabilized. In addition, the average RCS temperature and density were also stabilized. s3ecausa less ECCS tluid was required to make up for RCS shrinkage and thersevas more ECCS fluid injected than was lost through the ruptured steam generator tube, the pressurizer pressure tended to increase toward the pressure at which .

the ECCS flow rate equaled the SCTR break flow rate.

This pressure increase continued until the RCS cooldown was initiated at 23 minutes, At the beginning of the cooldown, the pass flow rate required to compensate for the RCS shrinkage plus the mass lost through the ruptured steam generator tube was greater than that being added through ECCS injection; however, when the RCS pressure had fallen to 11-10 ,

\

J l

4

- - . c is - - - ' - > - -

[r _

(

[

a Y below 1400 psia, the RCS backpressure had'hecome low enough that the 3 higher capacity safety injection pus.ps; (SIPS) could contribute significant amounts of fluid to the total ECCS flow. This additional

\'

fluid caused the pressurizer- which had completely emptied at l

\

\

,, approximately 30 minutes, to refill. The compression of the steam .

, bubbleinthepressurizercausedaktharppressureincrease. The '

pressure continued to rise toward that pressure (slightly less than 1600 psia) where the ECCS fluid compensated for the RCS shrinkage and  ;

the mass lost through the bre,ak. Following the end of the maximum rate cooldownat39.97 minutes,theECCShapacitywasnolongerrequiredto r

make up for the RCS shrinkage; therefore, the pressure climbed toward a nsv "equilibrium" pressure.

l I Subcooled Recovery Continnency Resoonses (40 to 90 minutes)

I i l

i Near the end of the maximum rate RCS cooldown, the pressuriser pressure (Figure 11-1) began trending toward an equilibrium pressure where the ,

ECCS flow rate was approximately the same as the SGTR break flow rate plus that mass required to compensate for shrinkage of the RCS during l

1 the cooldown. Immediately following the start of the controlled cooldown, and with the continuance of the RCS shrinkage, the rate of pressure increase slowed significantly. The operator action to ref,111 the pressurizer by opening one PORV resulted in a rapid depressurization. Because the pressure at the end of the depressurization was low (approximately 700 naia), the ECCS flow rate was relatively large, which resulted in a rap?d repressurization of the RCS. Following the end of the depressurization, one CCP and one SIP were stopped, thus reducing the ECCS flow rate by one-half. Because l the ECCS flow rate was lover, a lower equilibrium pressure was l established. This pressure was maintained throughout the remainder of I this phase of ths transient. <

Notice the small indentations on the plot of pressurizer pressure (Figure 11-1) after sixty minutes. These indentations represent brief periods o,f minor depressurizations which followed the brief injections of auxiliary feedwater into the intact steam generator as the nominal 11-11 i

,, - n, . - - , ,

. , 1 steam generator level was naintained. When the auxiliary feedwater injections occurrod, ths ECS temperatures became slightly lower, resul ing in the depressions in the pressura plots.

Saturate 1 1ecovery Cortinsenty Resoonses (af tg 90141MM Pcllowing the termf uation of the F.TP, leaving only a ' single CCP averattig', the preasurizer pressure began to fall. Because the SGTR break flow was greater than the capacity of the sinel.a CC7,'the pressure decretsed as the pressurizer uns drained. A new,. equilibrium pzassure would< eventually be re. ached at which the break flow rate equaled the C P flovtrate. However, prior to the arrival at the new equilibri'.;m pre raura , the pressurizer PO!? was opened at 95 minutes, resulting in a rapid decreasa in pressure.

The RCS depressurir.ation was procedurally alleved to continue until the RCS vas saturated or unti: the pressurizer level exceeded 78% span.

For this analysis, the core exit saturation pressuro was not attained in,the pressurizer before the PORV was closed due to high pressurizer level at 115.28 minutea. Pecause of the high level, the'RCS could not be further depressurized. In addition, because of the low F.C3 pressure, the capacity of the single CCP was such that the pressure began to increase as ite pressurizer was filled. This presrure increate would htve continued until the reactor operators-re-established chargtng and letdown to control level The addition of the RHRS would also be used to control the pressurizer level (and hence, the pressure as well).

i Because the pressurizer pressure sac lower than elsewhere in the RCS, no voiding in the RCS was predicted. During the depressurization, when p the saturation oversure of the pressurizer was approached, the rate of depressurization changed, following the saturation pressure of the pressurizes- tnroughout the remainder of the event.

h 11-12 g rw - , - + , - --

11.3.3 Pressurizer Level Standard SGTR Termination Resoonses (O'to 40 minutes)

Following the start of the SGTR, the pressurizer level (Figure 11-2) fell linearly until the reactor trip. Subsequent to the expected post-trip drop, the level briefly stabilized as the hot and cold leg temperatures increased. The level then continued to fall due to the RCS shrinkage and the continued break flow. The pressurizer level plot mirrors the pressure plot (Figure 11-1) until the RCS cooldown was initiated. The level then fell offscale low because the ECCS flow could not make up for both the RCS shrinkage and the SGTR break flow.

The pressurizer emptied at approximately 30 minutes and began to refill at approximately 32 minutes.

Subcooled Recovery Continnency Responses (40 to 90 minutes)

Beginning at 46.97 minutes, one pressurizer PORV was opened in order to hasten the refilling of the pressurizer. The PORV was closed when the level was greater than 20% span. However, because the RCS had been depressurized to a telatively low pressure, the ECCS flow rate was j quite large, thus the pressurizer refilled to a level greater than 20% j span. The large ECCS flow rates continued until an equilibrium pressure was approached; thus, the pressurizer level similarly continued to increase until the equilibrium pressure was approached.

l l

The bumps apparent in the level plot (Figure 11-2) are due to the sporadic addition of auxiliary feedwater to the intact steam generators, which was discussed in Section 11.3.2.

Saturated Recovery Continnency Resoonses (after 90 minutes)

Following the stoppage of the SIP, the level began to fall as the RCS was drained through the ruptured tube. When the FORV was opened at 95 minutes, the level initially rose as the pressurizer and RCS fluid expanded due to the lower pressure. However, because the break flow was greater than the ECCS flow, the level began to decrease. The 11-13

l effects of primary coolant expansion at lower pressures, decreased the I break flow rates and increased ECCS flow rates at the lower pressures, were sufficient to overcome the reduction in level due to the break l 1

flow. Thus, the pressurizer level increased slowly throughout the I remainder of the simulation, 11.3.4 Reactor Power Response Prior to reactor trip, the core power decreased slightly as the RCS pressure decreased. The power reduction was due to the moderator density coefficient, which introduced negative reactivity into the core as the RCS became slightly less dense. Subseqi' .t to reactor trip, the reactor power was 120% of the 1971 ANS decay heat curve, and was essentially unaffected by the SCTR transient. Boron injected though <

the ECCS prevented the core from becoming critical as the RCS was cooled.

11.3.5 RCS Temneratures Standard SGTR Termination Responses (0 to 40 minutes)

Foll> wing the initiation of the SGTR, the RCS temperatures (Figure 11-3) decreased very slightly as the reactor power decreased.

The power reduction was due to the effects of the moderator density coefficient described in Section 11.3.4. As the RCS depressurized, the fluid density in the core also decreased, resulting in a lower hot leg temperature and an ensuing reduction in the cold leg temperature. The pressure spike discussed previously caused a similar spike in the hot leg temperature. This spike was not seen in the cold leg temperature due to the masking effect of the temperature increase following the loss of main fesdwater. Following reactor trip, turbine trip and the subsequent loss of main feedwater, and the loss of forced circulation in the RCS, the cold leg temperatures increased rapidly. The cold leg temperature increase was termint.ted by the opening of the ARV on the ruptured steam generator and the addition of large amounts of auxiliary feedwater to the steam generators.

l 11-14 l

. )

l l

l Following reactor trip, the hot leg temperature initially fell, but then increased as the RCS flow coasted'down due to the loss of offsite l power. A hot-to-cold leg temperature difference corresponding to natural circulation quickly developed, thus maintaining the hot leg temperature approximately 300F above the core inlet average temperature. This natural circulation flow was maintained in all loops by steam release through the ARV on the ruptured steam generator. As the steam generators vera depressurized, the cold leg temperatures, and thus the hot leg temperatures, also decreased. When the MSIV was  ;

1 closed, the heat removal capacity from the intact loops was sharply l reduced. The cold leg temperatures in the intact loops then rose I

toward the saturation pressures of the intact steam generators.

Because steam relief from the ruptured steam generator untinued, the  !

cold leg temperature in the ruptured loop continued to decline. The cold leg temperature was expected to fall at a faster rate following the isolation of the ruptured steam generator; however, the increase and stabilization of the hot leg temperatures counteracted much of this effect.

Once the maximum rate cooldown had been initiated, the cold leg temperature in the ruptured loop began to decrease at a significantly l sharper rate. Note that the inflection point on the ruptured loop cold leg temperature plot lagged behind the start of the RCS cooldown by approximately one loop transit time (on the order of 300 seconds).

This delay was due to the time required for the reduced intact loop cold leg temperatures to be transmitted through the RCS loop to cause a lower hot leg temperature, and assuming the hot-to-cold leg temperature difference was approximately the same, to cause a lower cold leg temperature in the rupture loop.

As the steam generators depressurized, the rate of temperature decrease was reduced. Another system action which affected the RCS temperature was the control system which simulated the operator actions to maintain level at approximately 35% span in the intact steam generator and caused the auxiliary feedwater to be sporadically throttled. The 11-15

flatter portions of the temperature plot (Figure 11-3) correspond to throttling of the auxiliary feedwater, whereas the continuation of the temperature decreas's corresponds to periods of injection of the cold l auxiliary feedwater.

The RCS cooldown was terminated at approximately 41 minutes. Because there was no further steaming, the steam generator level remained near 35%, hence no additional auxiliary feedwater was added. The combined l effects of the termination of the steam flow and auxiliary feedvater flow caused the RCS temperature cooldown to cease, and the temperatures l began to increase as the steam generator saturation temperature began )

to rise. In addition, because the RCS flow rate was also reduced, the l

hot leg temperature began to increase.

Subcooled Recovery Continnency Responses (40 to 90 minutes)

Following the end of the maximum rate cooldown, the ARVs on the intact steam generators were closed and the auxiliary feedwater to the intact l steam generators was drastically throttled; therefore, the cold leg l temperatures on the intact loops began to rise. The rise in the cold l leg temperature resulted in an increase in the hot leg temperature as well.

l The controlled rate RCS cooldown was initiated at 45.23 minutes. As expected, the cold leg temperatures, followed by the hot leg temperatures, began to fall. The decrease was interrupted by the depressurization of the RCS, which temporarily impededt'he natural circulation flow. Following the end of the depressurization, one SIP and one CCP were stopped.

The stoppage of the pumps is readily distinguished in Figure 11-3 by the sharp increases in both the cold and hot leg temperatures. The cold leg volumes used in the RETRAN model are homogeneous volumes, thus the temperature of the entire volume is influenced by the mixture of the cold ECCS fluid. When the ECCS flow was reduced by half, the average enthalpy in the cold legs became more influenced by the RCS 11-16

1 fluid and less by the ECCS fluid; hence, the cold leg temperatures were l expected to rise. The hot leg temperature also rose as the hot-to-cold !

leg temperature difference was maintained. l Following the reduction of the ECCS flow by one-half, the cooldown of l both the cold legs continued. The hot leg temperatures followed the cold leg temperatures, maintaining a nearly constant temperature difference. As before, the cold leg temperature in the loops with the

)

intact steam generators varied slightly as auxiliary feedvater was injected and then throttled. Because a minimum valve open area was l

specified for the ARVs after 60 minutes, the effect of the cycling of the ARVs became less pronounced. The throttling of the auxiliary feedvater flow was also evident in the hot leg temperature response; however, because the hot leg temperature depended on the average core inlet temperature, and perfect mixing in the lower plenum was assumed, the variations in the temperature were less pronounced.

Saturated Recovery Continnency Responses (after 90 minutes)

When the SIP was stopped, the cold leg temperatures rose immediately.

This temperature rise was due to the reduction of flow into the cold l leg volume which was attributable to the cold ECCS fluid. Again, because the core exit temperature reacted to a change in the core inlet average temperature, the core exit temperature also increased. Recall that the controlled cooldown was still in progress; Dwever, the reduction in the pressure in the intact steam generators was such that the target cooldown rate of 1000F/hr could not be maintained. Thus, the cooldown rate was slowed to approximately 300F/hr. The effects of the periodic injection of auxiliary feedvater continued to be noticeable in the intact cold leg and core exit temperature responses.

Note that the ARVs on the intact steam generators were not constantly fully open, as would be expected. This inconsistency was due to the control system being used to control the cooldown. An instantaneous 10-second average was used to determine the cooldown rate. When auxiliary feedvater was injected into the intact steam generators, the cold leg temperature momentarily decreased at a rate greater than 11-17

1 1000F/hr; hence, the ARVs were modeled to partially close. When the auxiliary feedvater was throttled, the RCS cooldown rate became much less than 1000F/hr; thus, the ARVs rapidly opened fully.

I 11.3.6 Steam Generator Pressure Response I i

Standard SGTR Termination Resoonses (0 to 40 minutes) l I

Because the steam generators are connected through the main steam headers, the pressure responses of the steam generators were very nearly the same prior to isolation of the ruptured steam generator at 13.08 minutes.

I Following the initiation of the SGTR, the pressures in the steam generators decreased slightly due ;o the corresponding reduction in 1 core power and RCS temperate.res. Because the steam dumps were not ,

modeled and only the ARV on the ruptured steam generator was assumed available, the steam generator pressure rose sharply following turbine trip. Due to the opening of the ARV and the immediate injection of 1

large amounts of cold auxiliary feedwater, the steam generator I l

pressures did not reach the accumulation pressure of any of the main '

steam safety valves. Primarily due to the cold auxiliary feedwater, the steam generator pressures decreased until the ruptured steam generator was isolated, and the auxiliary feeovater flow to the intact steam generators was throttled. After isolation, the pressure  !

responses of the intact and ruptured steam generators diverged.

After the auxiliary feedvater to the intact steam generators was throttled, the intact steam generator pressure rose toward the typical no-load pressure (~1100 psia). The RCS cooldown, effected by opening the ARVs on the intact steam generators, caused a rapid decrease in the steam generator pressure. As would be expected, as the pressure decreased, the rate of pressure decreased; thus the pressure reduction was more rapid at the start than toward the end of the cooldown period. The pressure was noticeably affected by the periods of throttling auxiliary feedvater flow; the pressure increased when the 11-18

auxiliary feedvater was throttled and decreased when injection continued.

Following the closure of the MSIV on the ruptured steam generator and the isolation of ruxiliary feedvater flow to this steam generator, the pressure in this steam generator continued to fall. As the hot leg temperature sta'ilized o (see Section 11.3.4), the rate of pressure decrease in the ruptured steam generator became significantly slower.

An equilibrium state was reached where the pressure rise due to the SGTR break flev was offset by the release of steam through the ARV.

Following the start of the RCS cooldown, and the ensuing decrease in the hot leg temperatures, the ruptured steam generator pressure again fell off smoothly as the the steam generator depressurized. Because of the buffering effect of the RCS, the ruptured steam generator was essentially unaffected by the cycling of auxiliary feedvater to the intact steam generators.

Subcooled Recovery Contin 2ency Responses (40 to 90 minutes)

Following the termination of the maximum rate cooldown at approximately 40.23 minutes, the pressure in the intact steam generators began to rise. This pressure increase was primarily due to the level swell as the cold auxiliary feedvater varmed and compressed the steam bubble.

Once the controlled RCS cooldevn was initiated by periodically opening the ARVs at 44.97 minutes, the intact steam generator pressure again began to fall. Throughout the remainder of this phase of the transient, the intact steam generator pressures increased when the ARVs were fully or partially closed or when the auxiliary feedvater was throttled, and decreased when the ARVs were opened or when the auxiliary feedvater flow rate was increased. As the pressure fell, decreasing the differential pressure between the steam generators and the atmosphere, the steam flow rate through the ARVs also decreased.

Therefore, the steam pressure tended to level out as the atmospheric pressure was asymtotically approached.

11-19

! I l

l Similarly, the rate of depressurization in the ruptured steam generator also decreased as the atmospheric pressure was approached. Another factor, which acted'to maintain the pressure, was the addition of the .

varmer RCS fluid through the ruptured tube. Some of the fluid flashed to steam immediately, while the remainder acted to maintain the steam pressure by compressing the steam bubble as the steam generator liquid volume increased.

Saturated Recovery Continnency Resoonses (after 90 minutes)

The steam generators had virtually depressurized by the start of this l

phase of the transient. There was less than 150 psi of pressure i difference driving the steam relief rate; hence, the complete depressurization of the steam generators was approached very slowly. I After the RCS had been depressurized to near saturation pressure, the break flow had become relatively small. The added pressure in the ruptured steam generator due to steam bubble compression and primary coolant flashing was compensated by steam relief through the ARVs. The effects of the periodic injections of auxiliary feedwater into the intact steam generators can be seen in the plot of intact rteam generator pressure (Figure 11-4).

11.3.7 Steam Generator Narrow Ranne Level Response The steam generator level response is shown in Figure 11-5. As discussed in Section 5.3, only a qualitative discussion the steam generator level response is possible. Thus, while the absolute level predicted during the SGTR simulation may not be exactly accurate, the indicated trends are correct. In addition, for this conservative case, the differences in the indicated levels may be misleading. The void fraction in the ruptured steam generator was reduced in order to obtain a desired initial steam generator mass. When the reactor and turbine tripped, the voids were collapsed, but because there initially were fewer voids in the ruptured steam generator, the mixture volume did not decrease as much as in the intact steam generators.

~

11-20

l l

l l

Ruotured Steam Generator Standard SGTR Termination Resoonses (0 to 40 minutes)

After the SGTR began, liquid from the break flow was added to the steam 1 generator, and the void fraction in the mixture region of the ruptured steam generator was increased due to flashing of the RCS coolant. The increased void fraction resulted in an increase in the steam generator level indication; however, the steam generator water level controller acted to maintain the level near the nominal value.

s Following the reactor and turbine trips and the initiation of auxiliary feedwater, the level rose linearly until the MSIV was closed and the i auxiliary feedwater was isolated. The isolation of the ruptured steam generator from the intact steam generators caused an increase in the )

amount of flow through the ARV on the ruptured steam generator- 1 I

therefore, the mixture void fraction was increased and the level indication responded appropriately. However, the loss of auxiliary feedwater and the continued steaming caused the ruptured steam generator inventory :c decrease, thus the indicated level decreased.

As expected, there was a slight change in the rate of decrease during the period where the RCS hot leg temperatures stabilized, and the equilibrium between the SGTR break flow and the ARV break flow was attained. During this period, the SGTR break flow rate increased, and the RCS flow rate in the ruptured loop decreased, resulting in an increase in the mixture void fraction and a change in the rate of level decrease.

Following the start of the RCS cooldown, the rate of level decrease again rose and then fell until the cooldown was completed. Note that the rate of depressurization had dropped sharply from the original rate of depressurization, thus the mass release rate through the ARV had become significantly less. The increase in the RCS pressure caused the SGTR break flow rate to increase. The increase in the break flow rate coupled with the decrease in mass release through the ARV caused the 11-21

{

j level indication to increase. Because nothing occurred to halt the process until the break flow was terminated, the indicated level in the ruptured steam generator continued to increase throughout the remainder of the event.

Intact Steam Generators Standard SGTR Termination Resoonses (0 to 40 minutesl The indicated level in the intact steam generators remained constant until the time of reactor and turbine trip. The post-trip dropoff was turned around by the addition of large amounts of auxiliary feedwater from both the turbine- and motor-driven auxiliary feedwater pumps.

Vhen operator action was modeled at 13 minutes to throttle the auxiliary feedwater to maintain level between 10% and 50% of span, the indicated level was greater than the average value of 35% span used by the auxiliary feedvater controller to simulate operator control of the auxiliary feedwater. Thus, auxiliary feedwater to the intact steam generator was essentially stopped until the level indication fell to below 35% span. The reduction of the auxiliary feedvater flow caused the indicated level to increase slowly as the cold auxiliary feedvater

~

warmed to the saturation temperature of the intact steam generator (level swell). When the RCS cooldown was initiated, the level spiked briefly due to the increased voids at the lower pressure, then fell rapidly as the steam generator inventory was depleted. The effects of the throttling of the auxiliary feedwater to these steam generators is seen in Figure 11-5 between 25 and 40 minutes. The indicated level increased when the auxiliary feedwater was injected and fell when the auxiliary feedwater was tlrottled. Note that the level swell was not enough to overcome the effects of a decrease in the steam generator inventory, thus the level fell when the auxiliary feedwater was throttled.

(

11-22 l . .

Subcooled Recovery Continnency Responses (40 to 90 minutes)

The indicated level's increased immediately when the ARVs were opened furth'er. The level increase was due to an increase in the amount of voiding, which in turn affected the mixture volume upon which the level i indication was based. Following the initial level increase when the ARVs were opened, the level fell as the steam generator inventory was depleted. The reduction in the ARV area caused an immediate decrease in level due to the void collapse as the pressure increased. Overlaid onto this pattern were the effects of the throttled auxiliary feedvater. As auxiliary feedwater was added to the intact steam generators, the level began to rise. The rate of level increase due to an injection of auxiliary feedwater was essentially the same every time the feedwater flow rate was increased, facilitating the identification of periods of auxiliary feedwater injection.

The level in the ruptured steam generator is a function of the break flow rate through the ruptured U-tube. When the RCS was depressurized near 47 minutes, the break flow rate was decreased, and the rate of level increase was lessened. This change in the level increase can be seen in the plot (Figure 11-5) of the ruptured steam generator narrow range level near 50 minutes. Following the repressurization of the RCS due to the ECCS flow, the rate of level increase continued until it was predicted to go offscale (high) at approximately 80 minutes.

Saturated Recovery Continnency Responses (after 90 minutes)

The level indications in the intact steam generators during this phase of the transient behaved as in the previous phases, with the effects of the ARV modulations and auxiliary feedwater flow rate adjustments obvious in the level plot.

11-23

11.3.8 Break Flow Rate

'tandard SGTR Termination Resoonses (0'to 40 minutes)

S Following the initiation of the event, the break flow rate (Figure 11-6) decreased as the RCS pressure decreased. After the reactor and turbine tripped and the ensuing RCS pressure decrease and steam generator pressure increase occurred, the break flow rete droppe/.

sharply. As the ECCS flow caused tha RCS pressure to increase, and the auxiliary feedwater flow caused the ruptured steam generator pressure to decrease (Figure 11-7), the break flow rate gradually increased.

After the ruptured steam generator was isolated, the tuptured steam generator continued to depressurize while the RCS began to repressurize; thus, the SCTR break flow rate increased. The rate of increase in the flow rate changed as the-RCS temperatures and pressure approached equilibrium values. The RCS cooldown and resultant depressurization caused the break flow rate to decrease; however, the magnitude of the decrease was reduced due to the increase in the density of the break fluid.

Near 30 minutes, when the ECCS flow rate was greater than the break flow rate and the RCS began to repressurize, the break flow rate again began to increase. This increase in flow rate continued as the pressuriier repressurized and the ruptured steam generator depressurized through the stuck open ARV.

Subcooled Recovery Continnency Responses (40 to 90 minutes)

The break flow rate (Figure 11 6) was primarily a function of the primary-to-secondary pressure differential. By the start of this phase of the transient, the secondary pressure had essentially leveled out and was decreasing fairly slowly; thus, the break flow rate followed the same general trend as the RCS pressure. However, because the fluid density also affected the flow rate, the magnitude of any change was dampened relative to the pressure change.

11-24

Saturated Recovery Continnency Responses (after 90 minutes)

Following the termination of the SIP flow, the break flow rate decreased as the pressurizer pressure fell off. When the pressurizer PORV was opened to depressurize the RCS to saturation, the break flow >

rate fell sharply as the primary-secondary pressure difference decreased (see Figure 11-7). Thereafter, the break flow rate trended in the same manner as the RCS pressure. Note that the RCS continued to be at a slightly greater pressure than the ruptured steam generator; however, the differential pressure was relatively small. Because the RCS density was high, the mass flow rate through the break continued to be significant, even though the driving pressure differential was small.

Note that after the final RCS depressurization, the break flow began to increase. This increase was caused by the continued influx of ECCS flow from one CCP and would continue until the reactor operators aligned the charging and letdown system and initiated RHR cooling, 11.3.9 Ruotured Steam Generator Liould Volume Standard SGTR Termination Responses (0 to 40 minutes)

~

The collapsed liquid volume (Figure 11-8) is used to determine the margin to overfill of the ruptured steam generator. Due to the action of the Steam Generator Water Level Control System, the liquid volume in the ruptured steam generator remained approximately constant until the reactor and turbine were tripped. At that time, the volume increased sharply until the main feedwater flow was lost. The addition of large amounts of auxiliary feedvater caused the liquid volume to increase, but at a much slower rate. Following the isolation of auxiliary feedwater to the ruptured steam generator, the rate of volume change became a function of the SGTR break flow rate and the ARV steam flow rate. Because the steam flow rate out the stuck open ARV was significantly greater than the break flow rate, the liquid volume decreased. As the ruptured steam generator pressure decreased and the 11-25

break flow rate increased, particularly near the end of the RCS cooldown, the flow rate out the ARV became less than the incoming fluid. Thus the liquid volume began to increase and continued to increase throughout the remainder of the event.

Subcooled Recovery Contineenev Responses (40 to 90 minutes) i Due to the steady break flow rate throughout this phase of the transient, the ruptured stear. generator liquid volume steadily increased. At the time oht RCS was depressurized to refill the pressurizer, reducing the break flow rate, the rate of volume increase was changed; however, the rapid repressurization renewed the volume increase.

Saturated Recovery Contingency Resconses (after 90 minutes) f f

The liquid volume in the ruptured steam generator continued to increase until the primary was depressurized until the saturation pressure corresponding to the core exit temperature was reached in the pressurizer. At this point, the primary-to-secondary leakage had become relatively small and was essentially offset by the steam flow through the ARV, thus effectively terminating the SGTR event. The final volume was approximately 5140 ft3 , well below maximum steam generator volume of 5954 ft3, 11.4 Thermal-Hydraulic Inout for the Dose Consecuences Calculation In order to calculate the radiological doses received at the exclusion area boundary and low population zone boundary following a postulated SGTR event, the following parameters were required for two distinct time periods, 0 - 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> and 2 - 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> following the initiation of the event:

1) Time dependent primary coolant flashing fraction; 11-26

2) Time dependent break flow;

3) Total mass released from the ruptured steam generator (to atmosphere);

4) Total mass released from the intact steam generators (to atmosphere); and,

5) Total mass released to the condenser.

Because the steam generator secondary was at a lover pressure than the primary fluid, a fraction of the relatively hot primary fluid would flash into steam as the primary coolant leaked into the secondary.

This fraction is referred to as the flashing fraction.

Time dependent values of the above parameters for the 0 - 2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> period were calculated explicitly as discussed in Section 11.3. The integral steam releases and integral break flow (Figure 11-9 through Figure 11-11) were calculated from the time-dependent values calculated during the simulation of the event. The time dependent flashing fraction for the 0 - 2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> period is shown in Figure 11-12.

It was conservatively assumed that the entire 6 hour6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> period from 2 - 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> was required to terminate the offsite release. The values of the required parameters at two hours were then extrapolated in a conservative manner to determine the integral releases for the 2 - 8 hour9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> period. Furthermore, it was assumed that the steam flow rates decreased linearly with time. The break flow rate was conservatively assumed to remain constant throughout the 2 - 8 hour9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> period. In reality, the mass flows would approach zero asymtotically. The flashing fraction was also assumed to remain constant throughout the 2 - 8 hour9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> period.

The integral mass releases shown in Table 11-2 were calculated for use in assessing the radiological dose consequences following the postulated steam generator tube rupture event.

11-27

I 1

t 11.5 aur w The design-basis scenario of the postulated steam generator tube rupture event has been analyzed in order to obtain steam release rates to be used to assess the radiological dose consequences of this event.

In the scenario, it was assumed that the atmospheric relief valve on the ruptured steam generator failed to close. Because of a potentially adverse environment, the reactor operators could not assure that this valve would be closed; hence, it was necessary to simulate the SGTR transient until the Residual Heat Removal System could be placed in service at two hours. The required thermal-hydraulic parameters for the dose assessment calculation were then derived from this transient simulation.

l I

11-28

TABLE 11-1 Design-Basis SGTR - Radiological Consequences Event Timeline Time Event 0.08 min Begin SGTR 5.05 min Reactor trip on Low Pressurizer Pressure, Turbine Trip, Aux FW initiation, Loss of Offsite Power 5.13 min Begin Main FW Isolation 6.77 min Low Pressurizer Pressure - SIAS 13.08 min Operator Action - Close MSIV, Loop 4, Isolate AFV, Loop 4, throttle AFW, Loop 1 23.08 min Begin Maximum rate RCS cooldown 39.97 min End Maxi =um rate RCS cooldown 44.97 min Begin Controlled RCS cooldown 46.97 min Begin RCS depressurization to refill Przr 49.92 min End RCS depressurization to refill Przr 51.94 min Stop one CCP and one SI pump 90.00 min Stop one additional SI pump 95.00 min Open Przr PORV to saturate RCS 115.80 min Close Przr PORV on high Przr Level 120.00 min End transient simulation 11-29 l

TABLE 11-2

) Design-Basis SGTR - Radiological Consequences t

l Summary of Integral Mass Releases I

Steam Released from Intact Steam Generators l

f 0-2 hours Condenser 992,000 lba Atmosphere 370,000 lba 2-8 hours Atmosphere 423,000 lbm Steam Released from Ruotured Steam Generator 0-2 hours Condenser 331,000 lba Atmosphere 259,000 lba 2-8 hours Atmosphere 144,000 lbm Leakane from RCS to Ruotured Steam Generator through the SGTR 0-2 hours 352,000 lba 2-8 hours 584,000 lba 11-30

\

TABLE 11-3 Design-Basis SGTR - Radiological Consequences Key To Figures 11-1 through 11-12 Point Event A Reactor trip on Low Pressurizer Pressure, Turbine Trip, Aux FW initiation, Loss of Offsite Power B

Low Pressurizer Pressure - SIAS C Operator Action - Close MSIV4, Isolate AFW4, throttle AFW1 D Begin Maximum rate RCS cooldown E

End Maximum rate RCS cooldown F Begin Controlled RCS cooldown G Begin RCS depressurization to refill Przr H End RCS depressurization to refill Przr J Stop one CCP and one SI pump K

Stop one additional SI pump L Open Przr PORV to saturate RCS M Close Przr PORV on high Przr Level 11-31

PREssuRIIDt PRtssuRE 2.4 2.2 - 0 A,

2- C

i,3 -

of.

1.s - V

$ l '

64c y 4

-Y t.2 -

{ '

'J 8 1-E o.s -

c.s - l M

o.4 -

o.2 - /

o . . . . . . . . . . 4 o 20 40 so so 10o 1to Ttut (utN)

FIGURE 11-1. CPSES1 Design-Basis SGT2, Doses - Pressurizer Pressure PREssumzzR LatL woicutoN too to - M so - l 7o -

n x

5 so -

L L M

J '

P 5 so -

N j

a

^

l 4o - ,

. so - 's 2o - ,h r o e o 9 F to -

\

o s

/\ s s s , , , . s , ,

o 20 40 so ao too 120 Tius (un)

FIGURE 11-2. CPSES1 Design-Basis SGTR, Doses - Pressurizar Level 11-32 l

CORE DOT AND COLD t.go TDAPDIATURES ase ,

o B l

g, D SSO -

M- c u 'r s 450 - N 400 - 6 A E

350 - J -

300 -

250 -

200 i i i . 6 i i a i i i 0 20 40 40 80 100 120 M (W84)

T-D07 70-4NTACT  : TO-RUPRNtED FIGURE 11-3. CPSES1 Design-Basis SGTR, Doses - RCS Temperatures STEAM ctNDLATOR PRES 5URES IJ .a O

t.2 -

o 1.1 -

1--

c.s -

n

{m 03 - c l D 6= ,

"I 0.7 -

W!

a 0-a6

{* 0.5 - 8 r e, J

0.4 - ,

1 q+yNr QJ -  %

0.2 -

0.1 -

0 i . . . . . . i i i .

0 20 40 50 80 100 120 TlWC (MDf) errAcr - Rurnstsn FIGURE 11-4. CPSES1 Design-Basis SGTR, Doses - Steam Generator Pressures 11-33

t,

, , _ S/o PDAW MNGC t.EVD. WDtCADoN 90 -  !

e

~

A ir yo - D 0"

i, 7

a l2ll I

s so .

r a w'@p

• 0

m. e ,

I if g q 40 -

so -

20 -

10 -

0 4 . . . . . . . , , ,

C 20 40 so so 100 12o TIME (WN) wrAct - mynstm FIGURE 11-5. CPSES1 Design-Basis SGTR, Doses - Steam Generator Levels

,,3 SGTR BREAK ncw mit 90 -

so -

ro E

- 70 - D g 8 l "2 L R ^ 't so . o i ,,_

N J!

s

[c /~ "

, 40 -

b so - ,

s to -

10 -

n 0 , , , , , , , , ,

o 20 40 so ao too 33, NE (W8)

TIGURE 11-6. CPSES1 Design-Basis SCTR, Doses - SGTR Break Flow Rate i

11-34 1

g_ PRZn 0 5/o PRESSURts 2.2 - ^ o 2- \ u t.s -

ou t.s - / ,

1d * ,,'

i g a -- -

R3 ,

06, t- -

E o.s -

a.s .

N M c.4 -

0.2 -

o , , , , , , , , , , ,

0 20 40 a0 80 100 12e NC(WIN)

Pnzn RurfuRED s/c FIGURE 11-7. CPSES1 Design-Basis SGTR. Doses - Przr and SG Pressures RUPTunto s/c uouto vowut s

Top or sitm otNERAron g ,

z o s- "

V-C E 4-l g g NJ h

• r 4

• WD A N HJ }u

$,! s- '

V ol Y

>g -

$~

j 2-t.

0 , a s a > a , , , , ,

o 20 40 so ao too tan 1WCOmN)

FIGURE 11-8. CPSES1 Design-Basis SGTR Doses - Ruptured SG Liquid Volume 11-35

1 INTECMATED STEAM RO.EASC TO CONDDfSDt 1.1 - .

1-I OJ -

DJ -

d, 0.7 -

"E 11 M* OJ -

S jo 0.4 -

g OJ -

0.1 -!

0 , , , , , , , , , , , , , , , , , , , 3 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1J 2 NC (HRS) mAcT so nupnntEn s0 FIGURE 11-9. CPSESI Design-Basis SGTR, Doses - Integrated Steam Releaces to Condenser INTEGRATED STEAM REZASE TO ANOSPHUtE 1

0A -

p OJ -

S 0.7 -

h s

OJ -

nw 0.4 -

h OJ -

s*

u-0.1 -

0 . i i i . . ,

0 2 4 s a TIME 08ts) mAct so - - nuruten so FIGURE 11-10. CPSES1 Design-Basis SGTR, Doses - Integrated Steam Releases to Atmosphere 11-36

WTEGRA113 scm SRfM FLOW 1

OJ -

u-I

$ c.7 -

l_ u -

4I Qj u-Ea f

a a.4 -

g u-m 0.2 -

c.1 -

0 . . . s . . . .

o 2 4 s a Ttut (MRs)

FIGURE 11-11. CPSESI Design-Basis SGTR, Doses - Integrated Break Flow HOT LIG RCS - SC FLASHINo FRACT!aN o.16 g o.15 -d 0.14 -

c.13 -

c.12 - D o.11 -

z

$ o.1 -

u g,gg . C

' a y O.os -

0.o7 - o

{

r o.04 - MJ s <

l o.os - 7 l x '

o.o4 - L/ ,, 1 o.as - V L.,.

o.o2 -

0.01 -

0 . . . . . . . . . . .

o 20 40 60 to 100 120 wt (um)

FIGURE 11-12. CPSES1 Design-tasis SGTR, Doses - Not Leg Flashing Fraction 11-37

CHAPTER 12 RADIOLOGICAL DOSE ASSESSMENT For the design-basis SGTR event for the Vorst Case, Rad;ological Consequences, a calculation was performed to evaluate the whole body and thyroid inhalation doses to an individual located at the exclusion area boundary (EAB, 1544 m) and the low population zone boundary (LPZ, 4 miles). The results were used to show that the distances to the EAB and the LPZ were sufficient to provide reasonable assurance that the calculated radiological consequences of a pontulated steam generator

, tube rupture did not exceed: (a) the exposure guidelines set forth in 10CFR100, Chapter 11 for the accident with an assumed preaccident iodine spike and (b) ten percent of the 10CFR100.11 exposure guidelines for the accident with an assumed concurrent iodine spike.

Two separate iodine spikes were considered:

Case I: A reactor transient had occurred prior to the tube rupture which raised the primary coolant iodine concentration to 60 uC1/gm dose equivalent iodine-131 (DEQ I-131).

Case II: The reactor trip or primary system depressurization associated with the postulated accident created an iodine spike in the primary system. The spike vas assumed to increase the iodine appearance race (inleakage from the defective fuel rods to the primary coolant) to 500 times the equilibrium appearance rate.

\

The assu=ptions listed belov vere used to dettecine the initial primary and secondary activities and to calculate the activity released and the offsite doses for the postulated steam generator tube rupture. The dose calculational method and dose conversion tasumptions are consistent with those contained in Regulatory Guide 1.109 [14).

I 12-1

5 '

. 1 c.

(-

L k

1) .The initial primary coolant iodine activity was as1umed to be at the Stan/ -hnical' Specification limit (STS 3/4.4.8 Snecific Activity f 1. sci /gm DEQ I-131 .

2) The primary coolant activity had been leaking into the secondary b'!

[

side for a period of time long enough to establish equilibrium activity concentrations in the steam geners. tors. The magnitude of the leakage, 1 gpa, was the STS limit (3/4.4.6.2 Reactor Coolant System Leakane) for primary to secondary leakage.

3) All noble gas activity was transported from the primary system to s the secondary system and the noble gas activity in the steam region of the steam generators was immediately released to the environment. d

4) Due to the pressure differential between the primary and secondary sides, a fraction of the primary coolant leaked to the defective steam generator was assumed to flash to steam. This flashed

(

fraction was assumed not to mix with the steam generator water, and therefore, was not subjected to any iodine removal process in the steam generator.

5) No credit was taken for decay of the radionuclides prior to their release to the environment.

6) A ground level release was assumed. No credit was taken for radioactive decay or cloud depletion due to ground deposition during the plume transport.

7) Vorst case, five percentile atmospheric dispersion factors vera, assumed.

3

8) A breathing rata cf 3.47E-04 m /s was assumed.

9) Conservative iodine partition factors of 0.01 were used in the steam generator and condenser to account for that fraction of 7

, 12-2

.icdino present in the steam generteor and condenser fluids which. l N in corried with the fluid as it is converted to steam.

I The integs al mass releases shown in Table 11-2 were calculated from the  !

thermal-Lydraulie onalysis of the Desir,t-basis Eve.te - Worst case 1 Radiological'.Consequen:es, described 11 Chapter 11, and ware used as' iq.ut fan the radiological dose assecament calculation. Th", integral 1

-mass ralsase.s for the 2 - 8 hour9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> period were based on extri.polations frez the 2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> values. The steam release races were conse rvatively sssumod to linearly de. crease to sero.st 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />. The primary to secondary leshage and the flashing fraction were conservatively assumed to be constat.t at thuit 2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> value throughout the 2 - 8 hour9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> period.

The calculated radiological doses are shown in Table 12-1. All of the calculated doses are well within the'10CFR100.11 limits of 300 Rem to the thyroid ar.d 2.1 rem to the whole body.

1 I

l l

l 12 3 l

l

.l l

TABLE 12-1 I

Radiological Dose Consequences of a Postulated SGTR

]

Minimum EAR (0-2 hrs) LPZ (0-8 hrs) i Whole Body Thyroid Whole Body Thyroid (Rem) (Rem) (Rem) (Rem)

J Case I - Preaccident Soike

- j Calculated Dose: 5.37x10-2 68.7 1.26x10-2 11,4 l

10CFR100.11 Limits: 25.0 300.0 25.0 300.0 '

ll 1

Case II - Accident-Initiated Soike l I

G Calculated Dose: 5.37x10-2 27.7 1.26x10-2 13.5 l

10CFR100.11 Limits: 2.5 30.0 2.5 30.0 P

1 12-4

l CHAPTER 13 CONCLUSIONS- i In response to NRC requests, TU Electric has re-examined the postulated, design-basis, Steam Generator Tube Rupture Event. The SGTR l event is unique among the design-basis accidents in that timely and correct operator intervention is required to terminate the event and prevent filling the ruptured steam generator with liquid. The overfill of the ruptured steam generator could threaten the integrity of the main steamlines and associated supports and could result in increased releases of radioactive material. Hence, using the CPSES control room simulator, extensive evaluations of the reactor operator responses and response times required for termination of the postulated event were performed using CPSES reactor operators trained to follow the CPSES Emergency Response Guidelines. The equipment required for termination of the postulated event was identified, and the safety classification of this equipment was verified.

A transient-specific RETRAN02 model of CPSES1 was developed in accordance with the TU Electric, Reactor Engineering Quality Assurance program and a representative SGTR analysis was simulated for comparison with a similar analysis performed for the Westinghouse Owners Group.

This comparison provided assurance of the adequacy of the CPSES1 Steam Generator Tube Rupture model. Conservative initial conditions were identified and included in the model. The effects of a number of postulated active failures were evaluated to determine which failure would result in the most severe transient. The active failure analyses resulted in the development of two design-basis scenarios. One scenario included the failure of an auxiliary feedvater throttling valve and resulted in the least margin to overfill of the ruptured steam generator.

The second scenario included the failure to close of an atmospheric relief valve on the ruptured steam generator, which resulted in the most severe postulated radiological consequences.

13-1

j For the design-basis Steam Generator Tube Rupture Event - Margin to Overfill, the margin to overfill of the ruptured steam generator was calculated to be 320 ft3 For the design-basis Steam Generator Tube Rupture Event - Worst Case Radiological Consequences, the offsite doses received at the exclusion area boundary and the low population zone boundary were calculated to be well within the limits of 10CFR100.11.

1 13-2 l  :

CHAPTER 14 REFERENCES

1. Nuclear Regulatory Commission letter to Texas Utilities Generating l Company, Docket Nos. 50-445 and 50-446, "Request for Additional l Information Regarding the Steam Generator Tube Rupture Accident Event for Comanche Peak Steam Electric Station (Units 1 and 2)",

April 9, 1984.

2. R. N. Lewis, et al. , "SGTR Analysis Methodology to Determine the Margin to Steam Generator Overfill", WCAP-10698-P-A, Westinghouse Electric Corporstion.

3. R. N. Lewis, et al. , "Evaluation of Offsite Radiation Doses for a Steam I Generator Tube Rupture Event", Supplement 1 to WCAP-10698-P-A, l Westinghouse Electric Corporation.

4. R. N. Lewis . r al., "Evaluation of Steam Generator Overfill due to a ,

Steam Generator Tube Rupture Event", WCAP-11002, Westinghouse Electric Corporation.

)

l

5. Nuclear Regulatory Commission letter to Westinghouse Owners Group, SGTR Subgroup, "Acceptance for Referencing of Licensing Topical Report l WCAP-10698, Supplement 1, ' Evaluation of Offsite Radiation Doses for a Steam Generator Tube Rupture Accident'", December 17, 1985.

l

6. Nuclear Regulatory Commission letter to Westinghouse Owners Group, SGTR l Subgroup, "Acceptance for Referencing of Licensing Topical Report WCAP-10698 "SGTR Analysis Methodology to Determine the Margin to Steam Generator Overfill", December, 1984", March 30, 1987.

7. J. H. McFadden, et al. , "RETRAN 02, A Program for Transient Thermal-Hydraulic Analysis of Complex Fluid Flow Systems",

NP-1850-CCM-A, Revision 3, Electric Power Research Institute, June 1987.

8. J. H. McFadden, et al., "RETRAN-02, A Program for Transient Thermal-Hydraulic Analysis of Complex Fluid Flow Systems",

NP-1850-CCM-A, Revision 3, Appendix A Electric Power Research l Institute, June 1987.

I

9. "A Thermal-Hydraulic Analysis of the Ginna Steam Generator Tube Rupture Event of January 25, 1982", INFO 84-009, Institute of Nuclear Power Operations, February 1984.

10. J. Chao, et al., "Using RETRAN-02 and DYNODE-P to Analyze Steam i Generator Tube Breaks", NSAC-47, I;uclear Safety Analysis Center. l May, 1982.  ;

1 I

14-1 j i

11. C. E. Peterson, et al., "Response of a E Two-Loop Plant to Steam Generator Tube Ruptures", NSAC-77 Nuclear Safety Analysis Center, July, 1984.

12. "Abnormal Conditions Procedures Manuals", Revision 37, Comanche Peak Steam Electric Station.

13. "Emergency Response Guidelines Manual. E Series", Revision 3, comanche Peak Steam Electric Station.

14. "Calculation of Annular Doses to Man from Routine Releases of Reactor Effluents for the Purpose of Evaluating Compliance with 10 CFR Part 50, Appendix I", Revision 1, Regulatory Guide 1.109, Nuclear Regulatory Commission, October, 1977.

14-2 I

_4t,-,A4,>A6 emA6-,A4a m--a L R mom 2m - mc1M-"- - E' ^*"-' =w " - ' 4- aa* 4 '=+ M==- 4-"- 4"M--- -

I

?

l I

iM N

l l

l 1

l l

l l

l I

_---.. - , _ _ _ . . . . , , - - ~ ~ - - . - - - - _ _ _ ---.._.___ _ --