ML021420059

Part a - Susquehanna Steam Electric Station, Supplement to Proposed Amendment No. 239 to License NPF-14 & Proposed Amendment No. 204 to License NPF-22, HPCI Automatic Transfer to Suppression Pool Logic Elimination

ML021420059
Person / Time
Site:	Susquehanna
Issue date:	02/04/2002
From:	Byram R G Susquehanna
To:	Document Control Desk, Office of Nuclear Reactor Regulation
References
-RFPFR, PLA-5425, TAC MB2190, TAC MB2191
Download: ML021420059 (244)

Text

Robert G. Byram Senior Vice President and Chief Nuclear Officer PPL Susquehanna, LLC Two North Ninth Street Allentown, PA 18101-1179 Tel. 610.774.7502 Fax 610.774.6092 rgbyram@pptweb.com b

• I flflm I Ur..am -m mm , TM FEB 0 4 2002 U.S. Nuclear Regulatory Commission Attn: Document Control Desk Mail Station OP 1-17 Washington, DC 20555 SUSQUEHANNA STEAM ELECTRIC STATION SUPPLEMENT TO PROPOSED AMENDMENT NO. 239 TO LICENSE NPF-14 AND PROPOSED AMENDMENT NO. 204 TO LICENSE NPF-22: IPCI AUTOMATIC TRANSFER TO SUPPRESSION POOL LOGIC ELIMINATION PLA-5425 Docket No. 50-387 and 50-388

Reference:

1)PLA-5322, R. G. Byram (PPL) to USNRC Document Control Desk, "Proposed Amendment No. 239 to License NPF-14 and Proposed Amendment No. 204 to License NPF-22: HPCl Automatic Transfer to Suppression Pool Logic Elimination

", dated June 8, 2001.2) Letter, NRC to R1 G. Byram (PPL), "Susquehanna Steam Electric Station, Units I and 2 Request for Additional Information Re: Elimination ofAutomatic Transfer offHigh Pressure Coolant Injection Pump Suction Source (TAC Nos. MB2190 and MB2191)", dated December 18, 2001. The purpose of this letter is to provide supplemental information necessary for the NRC staff to complete its review of the license amendment proposed in Reference

1. PLA-5322 proposed deletion from the Unit 1 and Unit 2 Technical Specification Table 3.3.5.1-1 the "High Pressure Coolant Injection (HPCI) System Suppression Pool Water Level -High" (Function 3e). Implementation of this proposed change eliminates automatic transfer of the HPCI pump suction source from the Condensate Storage Tank to the Suppression Pool for a high Suppression Pool level. Implementation of the proposed change and the associated plant modifications are essential to eliminate a vulnerability identified by the PPL Susquehanna (PPL) Individual Plant Evaluation (IPE). ApDJ Document Control Desk PLA-5425 The Nuclear Regulatory Commission staff has reviewed Reference 1 and has determined that additional information is required in order to complete the NRC review. The additional information requested is documented in a Request for Additional Information (RAI) dated December 18, 2001, (Reference 2). Attachment 1 to this letter contains responses to the NRC Request for Additional Information (Reference 2), and supplemental support documents referenced in the responses.

Both Attachments 2 and 3 contain information proprietary and/or confidential to PPL. Accordingly, it is respectfully requested that the information which is proprietary and/or confidential to PPL be withheld from public disclosure in accordance with 10 CFR 2.790 S.. of-the Commission"s-regulations.-tn-addition,-Attachment 3-contains information-which should be withheld from disclosure in the interest of national defense and to protect the security of the Susquehanna Steam Electric Station. This request is supported by an affidavit which sets forth the basis on which the information may be withheld from public disclosure by the Commission and addresses with specificity the considerations listed in paragraphs (a)(1) and (a)(4) of 10 CFR 2.790 of the Commission's regulations.

A non-proprietary/non-confidential version of calculation EC-RISK-1083, "Risk Associated with the Removal of the HPCI Automatic Suction Swap from the CST to the Suppression Pool" is included as Attachment

4. We trust that this information is sufficient for NRC to complete its review by May 1, 2002 to support implementation of the modifications, procedure changes, and operator training in September 2002. If you have any questions, please contact Mr. D. L. Filchner at (610) 774-7819.

Sincerely, cc: NRC Region I w/o attachment Mr. S. L. Hansell, NRC Sr. Resident Inspector w/o attachment Mr. D. S. Collins, NRC Project Manager wlo attachment BEFORE TILE BEFORE THE UNITED STATES NUCLEAR REGULATORY COMMISSION In the Matter of PPL Susquehanna, LLC: Docket No. 50-387 SUPPLEMENT TO PROPOSED AMENDMENT NO. 239 TO LICENSE NPF-14: IIPCI AUTOMATIC TRANSFER TO SUPPRESSION POOL LOGIC ELIMINATION UNIT NO. 1 Licensee, PPL Susquehanna, LLC, hereby files a supplement Proposed Amendment No. 239 in support of a revision to its Facility Operating License No. NPF-14 dated July 17, 1982. This amendment involves a revision to the Susquehanna SES Unit 1 Technical Specifications.

PPL Susquehanna, LLC "By: RG .Br Vice-P sident and Chief Nuclear Officer Sworn told subscribed before me This AF,"day 2002.I N Notarial Seal M -y J.mLanson, E Iresy Pub4, My Commission ExprsJurl U 2004 K4ZOtary Pblicýý BEFORE THE BEFORE THE UNITED STATES NUCLEAR REGULATORY COMMISSION In the Matter of PPL Susquehanna, LLC Docket No. 50-388 SUPPLEMENT TO PROPOSED AMENDMENT NO. 204 TO LICENSE PiP-221-HPCI-AUTOMATIC TRANSFER TO SUPPRESSION POOL LOGIC ELIMINATION UNIT NO. 2 Licensee, PPL Susquehanna, LLC, hereby files a supplement to Proposed Amendment No. 204 in support of a revision to its Facility Operating License No. NPF-22 dated March 23, 1984. This amendment involves a revision to the Susquehanna SES Unit 2 Technical Specifications.

PPL Susquehanna, LLC By: 2R.y am Sr. Vic President and Chief Nuclear Officer Swo and subscribed before me this iM 0me0 s- ar Public Notarial Seat W NancyJ. Lannen, Notaty Pubico I ntown, Lehigh County W Commissuon Expires June 4, 2004 rA.kn eq 111V Attachment 1 to PLA-5425 Page 1 of 7 Attachment 1 -Response to RAI Questions NRC Question la. lb, 1c, and ld The licensee stated that this change would address a potential vulnerability identified in the individual plant examination that is associated with a specific scenario involving an anticipated transient without scram. The licensee also indicates that this change would reduce operator burden during a station blackout event. However, if this automatic feature were removed, new operator actions may be necessary during a small-break loss of-coolant accident (LOCA) to manually transfer the suction from the CST to the SP on high SP level. Because there are as well -as- positiveiskTiipacts associated with the proposed change, the Nuclear Regulatory Commission (NRC) staff requires the following information to support its review: NRC Question la Provide the plant's current core damage frequency (CDF) and large early release frequency (LERF) and the plant's revised (i.e. assuming the proposed change is implemented)

CDF and LERF. In addition, the licensee should provide a breakdown of the current and revised CDF and LERF contribution by initiating event and needs to provide a discussion of the impacts of the proposed change on the individual event sequences/initiating events. PPL Response The current and revised CDF and LERF values are contained in calculation EC-RISK-1083, which is included as Attachment 4 to this correspondence.

The calculation determines that the CDF and LERF are reduced by 12% (from 4.45 E-7 to 3.92 E-7) and 85% (from 5.52 E-8 to 8.49 E-9) respectively if the automatic suction swap is changed to a manual suction swap for mean and the upper 95% confidence level operator error rates. The calculation determines that CDF reduction decreases from 12% to 7% and the LERF remains at 85% when the operator error associated with making the manual suction swap and controlling RPV water level with HPCI is assumed to be 100%. The breakdown of current and revised CDF and LERF initiating event contributions are provided in Attachment 4, Calculation EC-RISK-1083, "Risk Associated with the Removal of the HPCI Automatic Suction Swap from the CST to the Suppression Pool", Section 5.3.

Attachment 1 to PLA-5425 Page 2 of 7 NRC Question lb Provide a description of how the licensee assures that the current probabilistic risk analysis (PRA) models reflect the as-built, as-operated plant and if the current PRA has been through an industry peer review certification process. If there was a peer review, please provide the overall findings of the review (by element) and discuss any elements rated low (e.g., less than a 3 on a scale of 1 to 4) or any findings that potentially affect the sequences impacted by the licensee's proposed change. The licensee will need to address any identified weaknesses in the PRA models that might affect the results associated with this license amendment.

PPL Response The PPL approach to risk assessment isto develop an accurate and definitive description of each accident sequence and to realistically portray the role of human intervention as defined by Emergency Operating Procedures (EOPs) to terminate accident progression.

The original PRA model was developed under the auspices of the PPL Quality Assurance Program. The program requires each calculation to be prepared and reviewed by qualified engineers and approved by qualified supervision.

This process assured the original model was consistent with the as built and as operated plant. A discussion of how plant changes are incorporated in the model follows.

There are four ways plant changes can impact the model: design changes, procedure changes, configuration changes and reliability data changes. Each activity is captured and its impact reflected in the model. Risk Management is a design consideration for all plant modifications.

The design engineer is required to evaluate the impact of each design change on the plant risk model using a screening checklist.

The design engineer is required to contact the Risk Analysis Subgroup for any design changes that do not pass the screening.

The Risk Analysis Subgroup then analyzes the impact of the change on the model. This process ensures that the model will be consistent with design changes.

The Risk Analysis Subgroup is responsible for the generation of the Emergency Operating Procedure (EOP) basis, preparation of the 50.59 evaluations, and validation of the procedure changes in the plant simulator.

EOP changes are well understood by the Risk Analysis Group and have often been motivated by risk reduction initiatives.

This close coupling between the Risk Analysis Subgroup and the plant EOPs ensures that EOP procedure changes are reflected in the model.

Attachment 1 to PLA-5425 Page 3 of 7 Changes in plant configuration and reliability data are linked to the risk model through the Maintenance Rule (10 CFR 50.65). The plant configuration is controlled using an on line risk monitor (currently ORAM/SENTENAL but transitioning to EOOS). The monitor is based upon ensuring defense in depth for both Core Damage (CDF) and Large Early Release (LERF). The EOOS model under development will include a real time calculation of both CDF and LERF. Thus, plant configuration is controlled within the bounds of the risk model using an online monitor.

Finally the reliability and availability criteria used to evaluate the performance of plant Systems, Structures and Components (SSC) were derived from the PRA study. SSCs which do not satisfy the criteria are placed into Maintenance Rule Category al where a recovery plan is developed to restore the SSC to the value consistent with the PRA study. Thus,itheieliability-latabase used fortihePRA-models-validated-via-implementation of the Maintenance Rule. The PRA has not been certified through an industry PRA certification program. In accordance with Reg. Guide 1.174, Section 2.2.3.3, July 1998, "An Approach for Using Probabilistic Risk Assessment in Risk-Informed Decisions on Plant Specific Changes to the Licensing Basis", a peer review or certification process should be used as the basis to justify PRA adequacy in terms of scope and quality. The PPL IPE received an expert peer review by Dr. William Vesely. Additionally, the NRC issued a revised SER for the Susquehanna IPE. The review by Dr. Vesely and the SER are included within this Attachment.

NRC Question 1c Provide a description of the revised PRA modeling and/or assumptions used to reflect the proposed change. This description should address the specific thermal hydraulic conditions that are impacted and were analyzed and any changes in success criteria or sequence / timing events. PPL Response In the PRA model, based on current Susquehanna design, Manual Rod Insertion (MRI) is assumed failed and HPCI will automatically swap its suction source to the suppression pool on high pool level. MRI is a relatively slow process to shutdown the reactor and is only successful if HPCI is available for makeup. If an ATWS occurs, the suppression pool temperature will rise above 190TF, the short-term HPCI limit, before the reactor can be brought to Hot Shutdown by MRI. When the automatic suction transfer occurs due to high suppression pool level, there are not enough rods driven into the core to shut down the reactor. HPCI is assumed failed shortly after the automatic transfer occurs due to the Attachment I to PLA-5425 Page 4 of 7 high suction temperature.

Without HPCI adding water, the Reactor Pressure Vessel (RPV) level will drop and the RPV must be depressurized while critical, which will cause core damage. In the revised PRA model, applicable after the proposed change to remove the HPCI automatic suction swap for high suppression pool level, success of M!I is allowed. The manual suction swap imposes an operator action for the small liquid LOCAs. With a small liquid LOCA the suppression pool level rises due to the liquid from the break and the HPCI exhaust steam. The level will exceed the manual transfer point (25 feet suppression pool level) in a minimum of 21 minutes. However, the temperature of the pool does not exceed 140*F. A suppression pool level above 25 feet does not automatically fail HPCI. HPCI will continue to exhaust steam into the suppression pool and the__suppression-pool level will not exceed the suppression pool load limit. However, if the operator does not control RPV level with HPCI, level 8 will be reached and HPCI will trip. HPCI will automatically restart if the RPV water level reaches level 2. The restart of HPCI with high suppression pool level permits water to enter the horizontal portion of the HPCI exhaust line and could cause a waterhammer upon restart of the system. The current Emergency Operating Procedure directs the operator to ensure HPCI and Reactor Core Isolation Cooling (RCIC) are running when the suppression pool reaches 26'. (The 26' will be changed to 25' in the next revision of the subject procedure.)

This action is not modeled in PPL's Computerized Fault Tree Analysis (CAFTA) because there is complete dependence between the operator manually initiating the suction swap and also assuring that HPCI is running. Therefore, one operator action is sufficient for modeling this event. The HPCI level concerns outlined above are addressed in the changes to the fault tree model as described below: An AND Gate, 152-II-N-LVLFAJL was added to the HPCI OR Gate, 152. The AND Gate has four inputs: a small liquid LOCA initiator, a basic event of an operator error to control RPV water level, a basic event of an operator error to manually swap the HPCI suction from the CST to the suppression pool and a switch to defeat this logic for the case using the automatic transfer.

Both operator actions need to fail during a small liquid break LOCA to fail HPCI. If either RPV level is controlled or the operator transfers the HPCI suction to the suppression pool, HPCI will not fail due to water intrusion into the exhaust line. If the operator controls RPV level below level 8, HPCI will not trip and try to restart. Continued operation of HPCI when the suppression pool level is above the automatic suction swap level is not a problem, as the exhaust steam will maintain the turbine exhaust piping free of water.

Attachment 1 to PLA-5425 Page 5 of 7 A small liquid break is the only scenario when it is desirable to align the HPCI suction to the suppression pool. For other initiators, the suppression pool level does not reach the HPCI suction manual transfer point of 25 feet or the pool level is above the manual transfer point but the suppression pool temperature exceeds 190'F (the ATWS sequence using MRI). NRC Ouestion ld Provide a description of any new operator actions that are required as a result of the proposed change (e.g., manual swap-over from the CST to the SP for small-break LOCA events), including the associated human error probabilities (HEP's) and the human reliability analysis bases for the HEP's (e.g., cause-based versus time-based, time available, proceduralized, difficulty of diagnosis and implementation, etc.) based on NRC Information-Notice97-78,-"Crediting of-Operator Actions-in Place o Automatic Actions and Modifications of Operator Actions, including Response Times," the licensee should provide for each of these operator actions: (1) the specific operator actions required; (2) the potentially harsh or inhospitable environmental conditions expected; (3) a general discussion of the ingress/egress paths taken by the operators to accomplish functions; (4) the procedural guidance for required actions; (5) the specific operator training necessary to carry out actions, including any operator qualifications required to carry out actions; (6) any additional support personnel and/or equipment required by the operator to carry out actions; (7) description of information required by the control room staff to determine whether such operator action is required, including qualified instrumentation used to diagnose the situation and to verify that the required action has successfully been taken; (8) the ability to recover from credible errors in performance of manual actions, and the expected time required to make such a recovery; and (9) consideration of the risk significance of the proposed operator actions.

PPL Response (1) New Operator Actions -There are two operator actions credited in the removal of the HPCI automatic suction transfer modification:

(1) RPV level control after a transient and (2) manual suction transfer.

The manual suction transfer is a new action not currently part of the Emergency Operating Procedures (EOP). This new action will be added to the procedures in accordance with the administrative program that governs EOP changes. It is anticipated that the step will read as follows: When the suppression pool level reaches 25', ensure HPCI and RCIC are running. If HPCI is injecting into the RPV and Suppression Pool Temperature can be maintained less than 140TF, transfer HPCI suction from the CST to the Suppression Pool. Success of either the RPV level control action or manual suction transfer will prevent loss of HPCI.

Attachment 1 to PLA-5425 Page 6 of 7 (2) Harsh/Inhospitable Environments

-There are no harsh or inhospitable environments for the new operator action. The new operator action, manual suction transfer, is accomplished in the control room by turning an existing electrical switch on panel 1(2) C601, an inner ring panel. (3) Inress/Egress

-The new operator action is in the control room, therefore, there is no ingress/egress path. (4) Procedural Guidance -Guidance for the RPV level control currently exists in the Emergency Operating Procedures.

Also, HPCI manual suction transfer will be added to the Emergency Operating Procedures when the modification is installed.

(5) Operator Training -This modification will not be implemented prior to training the operators.

A Licensed Reactor Operator will perform the new operator action in the control room. The operators will require training on the manual suction transfer.

Simulator exercises will be used to validate that a Licensed Reactor Operator would perform this action properly.

(6) Additional Support -There are no additional support personnel or additional equipment required for these actions.

(7) Control Room Staff Information

-Operators control RPV level by any entry condition for the RPV Control or Level/Power Control emergency operating procedures, step RC/L-4. The cue for initiating the HPCI suction transfer is suppression pool high level. The suppression pool high level condition is alarmed in the control room. The level switches that actuate the alarm are safety related and powered by a Class 1E 125VDC power. A 1E battery powers the control room annunciator; however, it is via a non-lE electrical panel. There are also two safety related suppression pool level indicators in the control room on 1/2C601.

The operator will receive indication of a successful suction transfer by the valve position indicating lights in the control room. (8) Recovery from Credible Errors -A credible error in implementing the manual suction transfer is an error closing the suction source from the CST without the suppression pool suction valve being open. If this error occurs, HPCI will trip on low suction pressure.

If this valve is inadvertently closed and there still is a valid HPCI initiation signal, this valve will automatically reopen and HPCI will automatically restart when the low suction pressure condition clears. The hand switch, which initiated the close signal, is a spring return to "auto" so the close signal is not continuous.

Hence, the potential operator error of closing the CST Attachment 1 to PLA-5425 Page 7 of 7 suction source is of no consequence.

The duration of the less than full suction flow condition is expected to be approximately 27 seconds. Loss of flow for 27 seconds has no impact on core cooling during small break LOCAs where HPCI is required to operate.

It should be noted that if the HPCI suction valve from the suppression pool is opened, the 100% open limit switch on this valve will initiate a close signal to the HiPCI suction valve from the CST. (9) Risk Significance of Proposed Actions -The proposed operator action, manual IIPCI suction transfer, is not risk significant.

There is no change in CDF for the operator action always being successful and there is only a 5% increase in CDF if the operator actioan always LERF is ins-ensifive to the manual HPCL suction transfer.

NRC Ouestion 2 The SABRE code is relied upon to assess the impact of the proposed change on the intermediate

-and small-break LOCA and inadvertent main steam isolation valve closure sequences.

This code has not previously been reviewed or approved by the staff. In order for the staff to complete its review of the proposed change, please submit the code and associated documentation.

The submittal should include the source code, available user documentation, and input data used to evaluate the sequences noted above. PPL Response:

Included in Attachment 1 to PLA-5425 are PPL Calculations EC-ATWS-0505, Rev. 8; EC-052-0593, Rev. 0; EC-SIMU-0501, Rev. 0; and EC-SATH-1007, Rev. 1 which provide software QA documentation for the SABRE computer code. Also supplied are Calculations EC-052-1018, Rev. 2 and EC-052-1025, Rev. 2 which support the proposed Technical Specification change to delete automatic HPCI suction transfer on high suppression pool level. The SABRE source code and SABRE input files are considered proprietary and are provided on 4mm tape along with instructions for code installation and execution in Attachment

2.

UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 6555-O001 August 11, 1998 Mr. Robert G. Byram Senior Vice President-Generation and Chief Nuclear Officer Pennsylvania Power and Ught Company 2 North Ninth Street Allentown, PA.18101

SUBJECT:

PLA-5425 ATTACHMENT i AUG 1- 129 VP-NUC. ENGBRG. & REVIEW OF THE SUSQUEHANNA STEAM ELECTRIC STATION, UNITS I AND 2, INDIVIDUAL PLANT EXAMINATION SUBMITTAL

-INTERNAL EVENTS (TAC NOS. M74478 AND M74479)-Dar MFr~am: Enclosed is the NRC staff's supplement to the October 27, 1997, staff evaluation report (SER) of the Susquehanna Steam Electric Station (SSES), Units I and 2, Individual Plant Examination (IPE) submittal for internal events, including Internal flooding.

On October 27, 1997, the NRC forwarded to Pennsylvania Power & Ught Company (PP&L) an SER stating that the NRC staff could not conclude that the SSES IPE submittal met the intent of S Generic Letter (GL) 88-20. In response, you submitted additional information and on February 27, 1998, briefed the staff on revisions made to address the issues identified in the SER. On Aprl 1, 1998, the staff audited the SSES IPE at your offices in Allentown, Pennsylvania.

The enclosed SER supplement addresses the !IaI resolution of the issues raised by the staff in its original SER. In the revised IPE, the caJo..ated mean core damage frequency (CDF) is about 7E-7/reactor-cycle of 15 months (or about 5E-7/reactor-year), which is about a factor of seven higher than the original IPE CDF of 1 E-7lreactor-cycle, -Anticipated transient without scram contributes about 63% to the CDF, Iros of decay heat removal contributes about 23%, internal flooding contributes about 10%, station blackout contributes about 2%, and transients contribute about 2%. The contribution of loss-of-coolant-accident (LOCA) and interfacing systems LOCA is less than 1%. You have implemented all of the plant Improvements Identified in the original IPE, as well as additional improvements.

On the basis of the Information provided, the staff concludes that the revised SSES IPE process is adequate to meet the following four objectives of GL 88-20: (1) To develop an appreciation for severe accident behavior, (2) To understand the most likely severe accident sequences that could occur at the plant, (3) To gain a more quantitative understanding of the overall probabilities of core damage'and fission product releases, and (4) If necessary, to reduce the overall probabilities of core damage and fission product releases by-modifying, where appropriate, hardware and procedures that would help prevent or mitigate severe accidents.

Robert G. Byram Therefore, the staff concludes that the SSES IPE submittal, as supplemented, meets the intent of GL 88-20. The staff notes your commitment to identify instances of plant improvements in order to maintain a low CDF or further decrease the CDF, at SSES. The staff also notes PP&L's strong in-house PRA capability.

The SSES IPE was performed almost entirely In-house; also, according to PP&L, it is continually using and updating the SSES PRA. Although the NRC staff had several concerns about the original SSES IPE approach, because of the revisions performed in the front-end portion, the ongoing use of the PRA in conjunction with PP&L's defense-in-depth approach, and the ongoing identification and implementation of improvements, the staff believes that the current front-end analysis of the SSES IPE presents an exemplary analysis.

The staff encourages you to continually confirm the IPE's reliability of equipment and operator performance ensuring that it portrays SSES plant capability under severe accident conditions.

However,-some weaknesses-still remainlnih-e-IPE's ba-rnd ahna-lysis.

The staff believes that it is unlikely that these remaining weaknesses have affected the overall conclusion from the revised analysis or the capability of identifying vulnerabilities; it may, however, limit its usefulness in other regulatory applications, especially in applications related to containment performance.

The staff believes that PP&L can enhance the usefulness of the SSES IPE by addressing the weaknesses discussed in the enclosed SER supplement.

It should be noted, that the staff focused Its review primarily on your ability to examine SSES Units I and 2 for severe accident vulnerabilities.

Although certain aspects of the IPE were explored in more detail than others, the review is not intended to validate the accuracy of the detailed findings (or quantification estimates) that stemmed from the examination.

Therefore, this SER does not constitute NRC approval or endorsement of any IPE material for purposes other than those associated with meeting the intent of GL 88-20. If you have any questions regarding the enclosed SER supplement, please contact me at (301) 415-1484.

Sincerely, Victor Nerses, Senior Project Manager Project Directorate 1-2 Division of Reactor Projects -1111 Office of Nuclear Reactor Regulation Docket Nos. 50-387/50-388

Enclosure:

As stated ccw/encl:

See net page-2 "

Mr. Robert G. Byram Pennsylvania Power & Ught Company Jay Silberg, Esq. Shaw, Pittman, Potts & Trowbrdge 2300 N Street N.W. Washington, D.C. 20037. Bryan A. Snapp, Esq. Assistant Corporate Counsel Pennsylvania Power & Ught Company 2 North Ninth Street Allentown, Pennsylvania 18101 Licensing Group Supervisor Pennsylvania Power & Ught Company 2 North Ninth Street Allentown, Pennsylvania 18101 Senior Resident Inspector U. S. Nuclear Regulatory Commission P.O. Box 35 Berwick, Pennsylvania 18603-0035 Director-Bureau of Radiation Protection Pennsylvania Department of Environmental Resources P. O. Box 8469 Harrisburg.

Pennsylvania 17105-8469 Mr. Jesse C. Tihton, III Allegheny Elec. Cooperative, Inc. 212 Locust Street P.O. Box 1266 Harrisburg.

Pennsylvania 17108-1266 Susquehanna Steam Electric Station, Units 1 &2 Regional Administrator.

Region I U.S. Nuclear Regulatory Commission 475 Ailendale Road King of Prussia, Pennsylvania 19406 General Manager Susquehanna Steam Electric Station Pennsylvania Power and Ught Company Box 467 Berwick, Pennsylvania 18603 Mr. Herbert D. Woodeshick Special Office of the President Pennsylvania Power and Ught Company Rural Route 1. Box 1797 Berwick, Pennsylvania 18603 George T. Jones Vice President-Nuclear Operations Pennsylvania Power and Ught Company 2 North Ninth Street Allentown, Pennsylvania 18101 Dr. Judith Johnsrud National Energy Committee Sierra Club 433. Orlando Avenue State College, PA 16803 Chairman Board of Supervisors 738 East Third Street Berwick, PA 18603 T STATES ,9 REGULATORY COMMISSION WASHINGTON, D.C. 2055-0001 STAFF EVALUATION REPORT SUPPLEMENT SUSQUEHANNA STEAM ELECTRIC STATION, UNITS 1 AND 2 INDIVIDUAL PLANT EXAMINATION (INTERNAL EVENTS ONLY) I. INTRODUCTION On December 13, 1991, Pennsylvania Power & Ught Company (PP&L) (the licensee) submitted the Susquehanna Steam Electric Station (SSES), Units I and 2, Individual Plant Examination (IPE) in response to Generic Letter (GL) 88-20 and associated supplements.

On November 4, 1992, and on December 17, 1996, the staff met with the licensee to discuss the Nuclear Regulatory Commission's (NRC) concerns regarding the SSES IPE. On January 11. 1993, the licensee submitted Volume 6 of the IPE and on June 23, 1997, the licensee provided additional information regardi ng issues raised by-the-staff..

The staff performed a "Step 1" review of the SSES IPE submittal and was supported by the Brookhaven National Laboratory.

On October 27, 1997, the staff sent its evaluation report to the licensee in which it was stated that the staff could not conclude that the SSES IPE met the Intent of GL 88-20. In response to this staff evaluation report (SER), the licensee revised Its IPE. On February 27, 1998. the licensee briefed the staff on the revisions it had made and on April 1, 1998, the staff audited the SSES IPE at the licensee's headquarters in Allentown, Pennsylvania.

The staff's audit focused on whether the licensee addressed the concerns documented In the October 27, 1997, SER. This supplement, therefore, documents the staffs findings and conclusions regarding the licensee's resolution of Its concerns.

In accordance with GL 88-20, PP&L had proposed in Its original IPE to resolve Unresolved Safety Issue (USI) A-45, "Shutdown Decay Heat Removal Requirements." The licensee had also proposed to resolve USI A-17, "System Interactions," as part of Its IPE. No other specific USIs or generic safety issues were proposed for resolution as part of the IPE. 11. EVALUAION In the revised IPE, the licensee calculated a core damage frequency (CDF) of about 7E-7/reactor-cycle, which" Is about a factor of seven larger than the CDF of I E-7/reactor cycle of the original submittal.

Anticipated transients without scram (ATWS) contribute about 63% to the CDF, loss of decay heat removal (DHR) contributes about 23%, internal flooding contributes about 10%, station blackout (SBO) contributes about 2%, transients contribute about 2%. The contribution from loss-of-coolant-accident (LOCA) and interfacing systems LOCA (ISLOCA) is less than 1%. In the SER of October 27, 1997, the staff expressed concerns in several areas. In particular, It was noted that the licensee did not provide sufficient evidence for the staff to conclude that the following areas were appropriately treated: common-cause failures (CCFs), human reliability analysis (HRA), plant-specific failures, and back-end (i.e.; containment performance) analysis, including the lack of sensitivity analyses.

The licensee addressed these concerns by revising Its CCF, HRA, and plint-specific data analysis, and performing a sensitivity study for the back-end analysis.

.2-Regarding the CCF analysis, the staff found that the original submittal treated CCFs inadequately for active components (e.g., diesels, valves, pumps, and batteries);

did not examine single failures to identify those that have a potential for common coupling; did not treat cross-system CCFs, particularly between the high-pressure coolant injection (HPCI) and the reactor core isolation cooling (RCIC) pumps; and did not consider CCFs due to test and maintenance.

In response to these concerns, the licensee reviewed the SSES operational history and revised its approach to CCF by incorporating in the IPE model CCFs for active components of Important systems (residual heat removal (RHR), emergency service water (ESW), RHR service water (RHRSW), and diesel generators);

examining single failures for common doupling; and Including CCFs for RCIC and HPCI and CCFs due to test and maintenance. (The licensee Identified

a. single failure with common-coupling potential, an ESW pump failure due to end bel erosion; it inspected the other pumps and indeed identified end bell erosion in those pumps as well, although they were in operable condition.

The licensee accounted-this failure- as CCF-for.

ESW.) Overall, It appears that the licensee performed a reasonable search for CCFs. In order to address the concern regarding low CCF values, the licensee used generic data (NUREG-1 150) instead of plant-specific data (estimated on the basis of examining SSESs procedures and practices).

In a similar manner, in order to address the concern for the. low plant-specific failure rates, the licensee substituted them with generic values. The licensee.did not provide a justification as to why these values are appropriate for SSES; therefore, although the licensee demonstrated the impact of the use of higher values on the IPE's results,.they did not demonstrate their applicability to SSES. The staff believes that this is a weakness of the revised IPE approach.

The licensee, however, performed uncertainty analysis throughout the IPE. Therefore, the staff believes that It is unlikely that this weakness has affected the licensee's overall conclusion from its revised analysis or its capability for identifying vulnerabilitles.

It may, however, have limited its ability to gain Insights.

Regarding the IPE's HRA, the staff found that the revisions in the treatment of both routine human actions (pre-initiator human events) and actions in response to an initiating event (post initiator human events) are appropriate.

Pre-initiator human events were explicitly modeled in the revised IPE and were segregated from random equipment failures to allow a better assessment of the contribution of human reliability to CDF and, therefore, the development of a better understanding of the role of human reliability on plant safety. According to licensee document EC-RISK-1063, 'the maintenance records were re-, examined to identify specific instances of undetected system unavailabilities caused by pre initiator human errors' (PP&L Calculation, pg. 34) for the period from July 1987 to January 1990. This search uncovered three instances of post-maintenance restoration errors. As a result, the licensee revised its IPE model to include post-maintenance restoration human errors'In specific components of the systems: HPCI; RCIC, low-pressure, coolant injection (LPCQ, diesel generators, altemate control rod drive pump, and standby liquid control system. A human error probability (HEP) mean value was estimated on the basis of plant-specific, data.I. The licensee treated miscalibration errors, which have a common-cause potential, using, as mentioned above, generic, NUREG-1 150, data. The staff notes that even a generic treatment of miscalibration is better than no treatment at all because it allows the performance of sensitivity analyses for deriving insights regarding the importance of miscalibration.

But, as noted above, the generic treatment of miscalibration Is a weakness in the licensee's HRA. Regarding post-initiator human actions, the licensee revised its IPE model to explici'y Include them on the event trees. The licensee's document, EC-RISK-1063, gives a detailed description of the process used to identify and quantify these actions. Accordingly, the licensee identified post-initiator human actions through a review of emergency procedures and Its defense-In-depth criteria that provide a reliable and updated source of actions performed In response to an initiating event. The licensee used two different approaches to quantify these actions. For those actions that could be quantified using plant-specific data docurnented In "Susquehanna Operator Response Data for Actual Events," or in "Susquehanna Operator Response Data From Simulated Events,* an HEP was-estimatedon-the-basesof-these data. For the remaining-actions,-data from NUREG/CR-4835 were used "because the method generation and its application am generally consistent with the approach being pursued at Susquehanna.*

The staff finds the licensee's approach of using plant-specific data for estimating HEPs a strength of the licensee's HRA. In general, the staff found that the licensee appropriately considered critical factors, such as the layout and accessibility of manipulated components, operator training for a specific action, the potential for confusion and misinterpretation of an emergency operator procedure entry condition, and time needed versus time available to perform an action. Furthermore, the dependencies between human actions and the Influence of the accident progression on human performance appear to have been treated appropriately.

On the basis of these findings, the staff concludes that the front-end analysis of the revised SSES IPE is reasonable.

In the original submittal, the licensee presented an approach to resolve USI A-45, Decay Heat Removal Reliability.

Taking into consideration the changes in the licensee's front-end analysis and quantitative results, its review of SSES plant-specific features, and the strategy it developed and implemented regarding this issue, the staff concludes that the licensee's IPE process used to search for DHR vulnerabilities is reasonable.

The licensee also proposed to resolve USI A-17. "System Interactions." as part of the IPE. The licensee did not identify any vulnerabilities with respect to A-17. According to GL 88-20. If a licensee concludes

• that no vulnerability exists at its plant that is topically associated with any USI or generic safety Issue the staff will consider the USI or GSI resolved for a plant upon review and acceptance of the results of the IPE.. The staff concludes, therefore, that the licensee has resolved USIs A-45 and A-17. Regarding the back-end analysis, the licensee conducted limited sensitivity studies to Investigate the conditional probability of containment failure given conditions of vessel breach at high pressure.

With the combination of core damage and vessel failure not at high pressure, the licensee calculated.a conditional probability of containment failure of 9 percent. By contrast, the combination of core damage and vessel failure at high pressure resulted In a conditional probability of containment failure of 54 percent This result appears to be reasonable. One specific aspect of the SSES IPE is the credit taken for preventing vessel failure with the core damaged under station blackout conditions through local operator actions focusing on providing alternate power (ac) or restoring ac power. According to EC-RISK-1 063, the licensee relies on operators stationed locally for performing these actions and the actions are well proceduralized and practiced.

In estimating pertinent HEPs, plant conditions and time needed versus time available to perform these actions were taken into consideration.

The staff notes that It was the intent of GL 88-20 for licensees to identify all potential means of accident mitigation.

Therefore, the staff finds this aspect as a strength.of the SSES IPE. It is noted however, that these actions contribute to a high probability of vessel failure prevention.

Therefore, the staff encourages the licensee to continually confirm the reliability of operator performance used in the IPE, ensuring that the IPE portrays SSES performance under severe accident conditions.

In general, the licensee indicated that core debris is 14 times more likely to be quenched in Vessel if-core-damage-progresses in a-manner consistent with the core relocation model used in the BWRSAR code, which the licensee used in the IPE, compared to the core blockage model employed in the industry-deVeloped MAAP code. The staff believes that code input assumptions, such as success criteria, may play a role in the reduction of vessel breach likelihood at SSES compared to other Mark II plants. Regarding the containment performance improvement (CPI) program recommendations, the SSES design includes a 30-day supply of compressed nitrogen for safety-relief valve actuation.

The licensee has also installed a mobile diesel generator to recharge the 125-volt dc batteries.

These plant capabilities provide enhanced depressurization system reliability.

The licensee has also provided threaded connections on the RHR service water system, which allow for alignment of the diesel-driven fire protection system pumps to the RHR system, thus providing an alternate water source for injection.

The licensee examined the issue of venting using an existing soft vent (i.e., the heating, ventilation, and air conditioning (HVAC) ducts). The HVAC piping will fail at expected vent pressures, now estimated at approximately 60 psig (based on the revised venting procedure) instead of at the 15 psig vent pressure proposed In the original IPE. The licensee indicated that it has developed procedures to maintain core cooling in the event that most reactor building equipment is lost by aligning systems external to the reactor building.

In addition, the licensee is evaluating a conceptual venting strategy that will provide a framework, based upon input such as the estimated source term and combustible gas challenges, to help decide if venting is a viable. option. This appears reasonable.

On the basis of this review, the staff concludes that the licensee's response to the CPI program recommendations is reasonable and consistent with the intent of GL 88-20. Some weaknesses exist, however, in the licensee's back-end analysis:

1. In the licensee's analysis, the accident sequence progression was terminated If the containment failed prior to. core damage; all sequences were then assumed to go to core damage in the reported CDF. Radionuclide releases were not calculated for these containment failures nor was a detailed understanding of plant response obtained. 2. The impact on conditional containment failure probability of some severe accident phenomena and resulting containment failure modes appear to have been understated.

As a result, all early and late containment failures, other than the containment failures resulting from loss of DHR discussed in item I above, are reported by the licensee to occur In less than one percent of core damage events, including ATWS and station blackout.

Appendix I to GL 88-20 recommended that licensees consider a maximum coolable debris bed to be 25 cm. For depths in excess of that (as proposed in the SSES IPE) both coolable and noncoolable outcomes should be considered and documented, even in the presence of a water layer provided by the drywell sprays, because of the possibility of the formation of a noncoolable debris crust. Noncoolab!e outcomes may lead to the occurrence of phenonema such as containment overpressure failure from noncondensible gas generation due to core concrete interaction or containment failure from corium attack on the drywell liner/concrete containment boundary__-.

-. The licensee assumed, however, that core debris released from the vessel post-accident will always be quenched on the drywell floor and, consequently, core-concrete interactions with the drywell floor, steel liner, or concrete containment will be prevented, as long as the drywell sprays provide a water pool on the drywell floor. Similarly, core debris attack on other structures, such as the downcomer vents, resulting in suppression pool bypass or loss of pool scrubbing, would not be possible, according to the licensee, given spray operation.

Additionally, the licensee did not consider the possible negative effects of water on the drywell floor, such as containment pressurization due to ex-vessel steaming resulting from fuel-coolant interactions.

3. The treatment of ISLOCA was characterized as limited in the staff's October 27, 1997, SER. The licensee has not revised its ISLOCA analysis and, consequently, it remains a weakness.

Ill. OQWCLUSIO On the basis of the information submitted by the licensee through either direct discussion with the staff or in writing, the staff concludes that the licensee's IPE is complete with regard to the information requested by GL 88-20 (and associated NUREG-1335), and that the licensee's IPE process is adequate to meet the objectives of the IPE program as stated in GL 88-20: 1. To understand the most likely severe accident sequences that could occur at the plant. 2. To develop an appreciation for severe accident behavior.

3. To gain a more quantitative understanding of the overall probabilities of core damage and fission product releases.

4. If necessary, to reduce the overall probabilities of core damage and fission product releases by modifying, where appropriate, hardware and procedures that would help prevent or mitigate severe accidents.

-6 Therefore, the staff now concludes that the SSES IPE submittal meets the intent of GL 88-20. The staff notes PP&L's commitment to identify instances of plant improvements in order to maintain a low CDF or further decrease the CDF, at SSES. The staff also notes PP&L's strong in-house PRA capability.

The SSES IPE was performed almost entirely in-house; also, according to PP&L, its staff is continually using and updating the SSES PRA. Although the staff had several concerns about the original SSES IPE approach, because of the revisions performed in the front-end portion, the ongoing use of the PRA in conjunction with PP&L's defense-in-depth approach, and the ongoing identification and implementation of improvements, the staff believes that the current front-end analysis of the SSES IPE presents an exemplary analysis.

The staff encourages the licensee to continually confirm the IPE's reliability of equipment and operator performance ensuring that it portrays SSES plant capability under severe accident conditions.

However, some weaknesses still remain In the IPE's back-end analysis.

The staff believes that it is unikely that these remaining weaknesses-have-affected-the-licensee's-overall-condusion-from its revised analysis or its capability of identifying vulnerabilities; it may, however, limit Its usefulness in other regulatory applications, especially in applications related to containment performance.

The staff believes that the licensee can enhance the usefulness of its IPE by addressing these issues, discussed in this document.

It should be noted, that the staff focused its review primarily on the licensee's ability to examine SSES Units I and 2 for severe accident vulnerabilities.

Although certain aspects of the IPE were explored in more detail than others, the review is not intended to validate the accuracy of the licensee's detailed findings (or quantification estimates) that stemmed from the examination.

Therefore, this SER does not constitute NRC approval or endorsement of any IPE material for purposes other than those associated with meeting the intent of GL 88-20. Principal Contributors:

E. Lois J. Lane Date: July 15, 1998 PLA-5425 ATTACHMENT I REVIEW OF THE SUSQUEHANNA IPE W. E. VESELY May 29, 1997 REVIEW OF THE SUSQUEHANNA IPE W. E. Vesely My conclusion overall is that the SSES IPE is a very detailed and competent analysis.

Based on my review of the accident sequences and based on my discussions with the chief engineer responsible for the IPE, I am convinced that the IPE models are thorough, detailed, and accurate.

The accident sequences that are defined are developed to the detail required to include all the contributing component failures and human errors. The accident sequences are not binned as in the usual PRA, but instead each individual sequence is followed in the evaluations and quantifications to determine the consequences of the sequence.

The plant damage state associated with each sequence is determined from the thermal-hydraulics of the sequence.

Sequences resulting in similar consequences which result in the same damage state are combined to give the total frequency of the damage state. This detailed analysis was necessary because of the defense-in-depth criteria utilized in the SSES IPE which required that specific, multiple equipment be identified which could mitigate each accident sequence.

In carying out its detailed treatments, the SSES IPE uses modeling approaches and quantification approaches which are different from those used in the usual PRA. Instead of focusing on the physical consequences and events which are associated with the accident sequences, the focus is on the plant damage state associated with the sequence.

The comprehensive set of plant damage states defined in the SSES IPE are: (1) core damage from inadequate cooling, (2) mechanical cladding damage, (3) core melt with 1 reactor vessel breach, (4) containment failure with core damage in-vessel, (5) containment failure with mechanical cladding damage in-vessel, (6) core melt with reactor vessel breach and containment failure, (7) wetwell venting without core damage, and (8) containment overpressure failure without prior core damage, and (9) core damage from inadequate cooling with vessel failure and containment overtemperature failure. The IPE focus on the plant damage states is done at the expense of a de-emphasis of the physical event and consequence descriptions.

For example, direct containment heating (DCH) is not explicitly evaluated in the SSES IPE. This makes the review of the sequences difficult with regard to the physical variables and events associated with the sequences, however the event bases for the sequences can be found in the IPE. The results of the SSES IPE are different from other PRAs and IPEs in that very low frequencies are calculated for core damage (1.1 x 10-7 per year), for core damage plus vessel failure (3.7 x 10-9 per year) and for core damage, vessel failure, and containment breach (7.9 x 101" per year). These calculated values are of the order of a factor of 50 to 100 lower than the corresponding values calculated in other PRAs and IPEs. Specific accident sequence frequencies for transients, ATWS, and LOCAs calculated in the SSES IPE are also low, ranging from one to several orders of magnitude lower than for other PRAs and IPEs. The very low accident frequencies reflect the significant design and procedure improvements that have been incorporated in the SSES plant. These significant improvements include: 2

1. Controlling the water level in the RPV within a wider band (from -60 to -161) during an ATWS as opposed to the narrow band used in other BWRs. This more flexible water level control not only reduces operator error but frees one operator for manual insertion of the rods which can significantly reduce the contribution from ATWS. 2. Adding an RCS bypass switch to allow immediate manual rod insertion which significantly reduces the contribution from ATWS which otherwise would be significant as it is in other BWRs. 3. Installation of a fifth, self-contained swing diesel which has its own DC system and which can be linked to any of the four ESS buses when other power sources fail. This significantly reduces the frequencies of accidents from loss of emergency power, which are significant contributors in other BWRs. 4. Eliminating the requirement to depressurize RPV on HCTL or PCPL which allows the operator approximately forty minutes to initiate SLCS in a fAll ATWS and approximately sixty minutes in a partial ATWS. This large time window significantly reduces the probability of an operator error of failing to initiate SLCS. In other BWRs the time window is significantly shorter, on the order of a factor of ten shorter, resulting in a significantly higher operator error probability which importantly contributes to the accident frequencies in other BWRs..3

5. Using the RWCU in the blowdown mode to remove decay heat from the containment which significantly reduces the requirement for venting the containment and significantly reduces the probability of containment failure given core damage. 6. Early controlled RPV depressurization to allow operation of the fire main if HPCI and RCIC fail, and early connection of the fire main using installed threaded attachments for RPV/PC injection.

This significantly reduces the contribution from HPCIIRCIC failure which is an important contributor in other BWRs. 7. Controlling the RPV water level so that it is in the TAF +5 feet range: and hence above the ADS set point. This basically eliminates the human errors of failng to inhibit the ADS and failing to initiate the ADS when there is a loss of high pressure injection.

These human errors are significant contributors in other BWRs. In all, more than a dozen significant design and procedure improvements have been instituted at the SSES plant compared to other BWRs. The NRC has issued a number of letters recommending that the SSES improvements be instituted at other plants. As a part of this review, a separate evaluation was carried out to determine the benefits of the SSES improvements by re-evaluating the Peach Bottom 1150 PRA assuming the SSES improvements were in place. The Peach Bottom CDF was determined to be 4.51 x 10'6 per year in the 1150 PRA and hence was already low. For the re-evaluation, the Peach Bottom 1150 data were used where applicable and SSES data were used for those events not covered by the Peach Bottom analyses.

With the SSES improvements in place, the dominant accident sequence contributions to CDF in the Peach Bottom PRA were 4 reduced by a factor of 10 to a factor of greater than 1000. This again confirms that the improvements instituted at the SSES plant are significant from a CDF reduction and risk reduction standpoint.

The 1150 PRA re-evaluations that were performed are attached.

In spite of its many good points, however, the SSES IPE poses a quantification problem.

The difficulty lies with the low CDF and low accident frequency values calculated in the SSES IPE. These low'values have little credibility in the way they are presented because they seemingly don't incorporate present PRA experience.

The PRAs and IPEs which have been conducted to date have provided a data base of equipment failure rates, human error rates, ccf probabilities, and resulting system failure probabilities and accident sequence frequencies.

Because of the different modeling used in the SSES IPE, this experience data base has not been accessed to the degree it could have been. The extra time allowed for post-initiator operator actions and the less demanding, but different, responses required from the operator result in human error probabilities which are not directly relatable to human error probabilities used for other BWRs. It is thus difficult to differentiate design and procedure differences from modeling differences.

The SSES IPE presents best estimates of the human error probabilities based on knowledgeable assessments.

However, these estimates need to recognize the possibility for higher human error probabilities existing, based on the experience data base to date. This does not necessarily involve changing the point estimates but instead assigning ranges to cover the possibility of having higher human error probabilities.

These ranges will also account for quantification uncertainties in post-initiator scenarios.

These ranges can then be used to calculate adjusted results in sensitivity analyses and uncertainty analyses.5 The very low accident frequencies calculated in the SSES IPE also increase the potential dominance of ccf contributions.

The added equipment redundancies in the SSES plant result in low probabilities from multiple equipment failures.

However, because of the low failure probabilities, there is a greater sensitivity to ccf contributions and to failure dependencies.

The SSES IPE included particular ccf contributions when these contributions were assessed to be compatible with SSES plant data. However the SSES plant data which was used to show compatibility with independent failure assumptions can also be used to show compatibility with a potential ccf probability eisting. This potential for ecf probabilities, although it is low, will significantly increase the ranges the accident frequencies can have. The low failure rates used for certain equipment will also have wider ranges when data uncertainties are considered.

Based on reviews of the SSES IPE quantifications, I believe the above adjustments when incorporated in the evaluations will provide a basis for the quantifications and will significantly increase the credibility of the SSES IPE numerical values. As an additional consideration, in the SSES IPE modeling of the accident sequences to determine the plant damage state, the range of events considered is truncated since only the most likely events are selected.

The selection of the most likely events gives the most likely responses and most likely system success states. However, this most likely event analysis neglects other less likely events which can result in greater consequences and greater demands on the system and on the operator.

The accident sequences therefore need to be reviewed to check for these variations in events. For example, as a sensitivity 6

study the results from the MAAP code should be compared with the results of the SSES IPE which utilized the BWRSAR code. The above issues with the SSES IPE quantification should not be the cause for the IPE being redone. Rejection of the IPE models and engineering analyses would lose the valuable and valid information in the IPE. The SSES IPE represents a state of advancement of plant design and procedures to reduce core damage frequency, to reduce the frequency of vessel failure and to reduce the frequency of containment failure. The SSES IPE also rep-resents a valuable tool for risk management applications.

The SSES IPE should not be redone. To address the difficulties with the SSES IPE, the quantification needs to be extended to include uncertainties and to recognize the different possibilities in the present experience data base. It is not the point values in themselves which provide confidence in the quantification, but their bases and the assigned ranges which account for uncertainties in the values. These uncertainties can be used to define adjusted values and to determine the ranges and adjusted values for the results. The re-evaluation of the numerical values is not a major effort and should entail approximately six man months worth of effort. This work will be worthwhile since it will upgrade the quantifications performed in the SSES IPE so that they have increased credibility.

This will in turn upgrade the credibility of the SSES IPE.7 ANALYSIS OF THE PEACH BOTTOM NUREG-1150 PRA WITH THE SSES MODIFICATIONS INCORPORATED The following pages present a re-analysis of the Peach Bottom NUREG-1 150 PRAI with the SSES modifications incorporated.

The re-analysis was done in two ways 1) incorporating the SSES equipment and procedure modifications but using the 1150 models and data and 2) also incorporating and taking credit for the SSES IPE differences in models and data. These analyses were requested from the chief engineer responsible for the SSES IPE and have been reviewed.

The first two tables present the summary results from the re-analysis.

The accident sequence indices are. those__usedjin-theNUREG-1150.-.

_The-'Delta -for-Methods"--represents-he-factor difference due to differing methods and the 'Delta for Modifications" represents the factor difference due to SSES equipment and procedure modifications.

The subsequent pages give the analyses of the individual sequences.

Additional sensitivity analyses are also carried out for the probability of failing to actuate SLCS.il"Analysis of Core Damage Frequency:

Peach Bottom, Unit 2 Internal Events", NUREG/CR-4550, August 1989.

No Credit for Modeling Differences Accident NUREG- Delta New Delta With PP&L Sequence 1150 Methods Frequency Modifications Mods note 1 1.64E-06 I 1.64E-06 0.0001824 2.99E-10 See Sensitivity 2 2 1.40E-06 I 1.40E-06 0.001 1.40E-09 See Sensitivity 1 3 2.79E-07 1 2.79E-07 0.28 7.81E-08 See Sensitivity 3 4 2.12E-07 1 2.12B-07 0.01 2.12E-09 See Sensitivity 4 5 1.90E-07 1 1.90E-07 1 1.90E-07 , ,ibw, 6 1.30E-07 I 1.30E-07 0.0001824 2.37E-11 See Sensitivity 2 7 1.25E-07 1 1.25E-07 0.0001824 2.28E-1 1 See Sensitivity 5 8 1.14E-07 1 1.14E-07 0.001 1.14E-10 See Sensitivity 1 9 8.73E-08 1 8.7313-08 0.01. 8.73E-10 See Sensitivity 4 10 5.72E-08 1 5.72E-08 0.01 5.72E-10 See Sensitivity 4 11 6.41E-08 1 6.41E-O8 0.01 6.41E-10 See Sensitivity 4 12 4.63E-08 1 4.63E-08 0.01 4.63E-10 See Sensitivity 4 13 4.37E-08 1 4.37E-08 0.001 4.37E-11 See Sensitivity 1 14 3.29E-08 1 3.29E-08 0.001 3.29E-11 See Sensitivity 1 15 2.69E-08 1 2.69E-08 0.001 2.69E-11 See Sensitivity 1 16 2.45E-08 1 2.45E-08 0.01 2.45E-10 See Sensitivity 4 17 2.20E-08 1 2.20E-08 0.28 6.16E-09 See Sensitivity 3 18 1.70E-08 1 1.70E-08 1 1.70E-08 Ds M r, ui. Totals 4.51E-06 4.51E-06 2.98E-07 Delta due to modifications

= (2.98E-7)/(4.51E-6)

= 0.06608239 Credit for Modeling Differences and Modifications Delta Modifications 0.0001824 0.001 0.28 0.01 1 0.0001824 0.0001824 0.001 0.01 0.01 0.01 .0.01 0.001 0.001 0.001 0.01 0.28 1 With PP&L Mods 2.99E-10 1.40E-09 4.30E-09 2.12E-09 1.90E-08 2.37E-11 2.28E-11 1.14E-10 8.73E-1-0 5.72E-10 6.41E-10 4.63E-10 4.37E-11 3.29E-11 2.69E-11 2.45E-10 6.16E-09 1.70E-09 note See Sensitivity 2 See Sensitivity 1 See Sensitivity 3 & 7 See Sensitivity 4 See Sensitivity 6 See Sensitivity 2 See Sensitivity 5 See Sensitivity 1 -Se Seitvit-See Sensitivity 4 See Sensitivity 4 See Sensitivity 4 See Sensitivity 1 See Sensitivity 1 See Sensitivity 1 See Sensitivity 4 See Sensitivity 3 See Sensitivity 6 Totals 4.51E-06 4.06E-06 Delta due to modifications

= (3.80E-8)/(4.51E,6)

= 0.009363433 Accident Qp nh'pnep 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 NUREG 1150 1.64E-06 1.40E-06 2.79E-07 2.12E-07 1.90E-07 1.30E-07 1.25E-07 1.14E-07 8,73E-08 5.72E-08 6.4LE-08 4.63E-08 4.37E-08 3.29E-08 2.69E-08 2.45E-08 2.20E-08 1.70E-08 Delta Methods 1 1 0.055 1 0.1 1 1 1 1 1 1 1 1 1 1 1 0.1 New Freauencv 1.64E-06 1.40E-06 1.53E-08 2.12E-07 1.90E-08 1.30E-07 1.25E-07 1.14E-07 5.72E-08 6.41E-08 4.63E-08 4.37E-08 3.29E-08 2.69E-08 2.45E-08 2.20E-08 1.70E-09 3.80E-08 1150 Methods Sensitivity 1 Impact of Manual Rod Insertion and ATWS Procedure Changes on ATWS CDF NUREG 1150 Cut Set 1 2 6 9 13 NS-1 NUREG 1150 PE 8.OOE-07 5.OOE-07 8.50E-08 6.10E-08 3.80E-08 Point Estimate Frequency 7.98E-07 5.OOE-07 8.50E-08 6.06E-08 3.80E-08 6.46E-09 1.49E-06---

Cut Set T3A*RPSM*SLCS8 T3A*RPSM*SLCS9 T3A*RPSM*SLCS1O T3C*RPSM*SLCS8 T3C*RPSM*SLCS9 T3C*RPSM*SLCS10 SLCS sensitivity 0.005 0.001 7.98E-07 7.98E-07 1.25E-07 2.50E-08 8.50E-08 8.50E-08 6.06E-08 6.06E-08 9.50E-09 1.90E-09 6.46E-09 6.46E-09 -1.08E 9.76EA -0.....PP&L Modifications designed to mitigate these sequences

1. Add RSCS bypass switch to allow immediatemanual rod insertion

2. Change level control band to -60 to -161 -Frees operator to insert rods. BWROG procedures requires one operator to control pressure and one to control level 3. Eliminate requirement to depressurize on HCTL & PSP -avoids unnecessary and potentially core damaging depressurization during MRI. Operator has at least 40 minutes in full and 60 minutes in partial ATWS to start SLCS. These modifications to the equipment and procedures allow the operator to successfully complete MRL Sensitivity 2 Impact of Susquehanna Modes on Accident Sequence #1 NUREG Point 1150 Cut Set NUREG Estimate # 1150 PE Frequency 14 3.70E-08 3.70E-08 P(CD/SBO)

= 0.001832 1. Early controlled RPV depressurization to allow success of fire main should HPCI/RCIC fail to run S.... .(EO -1/200- 30) .... ....... .. ..........

.. ..... .. ....... ............ . 2. Early connection of the fire main, using installed threaded attachment for RPV/PC injection (ES-013 001). 3. Early connection of 100kw generator to supply DC power to DC busses (EO-1/200-030

& ES-002 001). 4. Elimination of HPCI suction swap from CST to the Suppression Pool on high pool water level, prevents failure of HPCI on high water temperature.

5. Early alignment of HPCI in pressure control mode to reduce chance of SORV & multipleistarts HPCI and RCIC (EO-1/200-030).

6. Installation of a maintenance swing diesel generator that can be substituted into any of the 4 ESS busses (OP-024-004).

E diesel has self-contained DC system. 7. Modified Emergency Service Water System so that 4 diesel must fail for SBO rather than two combinations of two diesel We have two self-contained diesel driven fire pumps and over 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> to connect fire main NUREG-1150 Cut Set 14 T1 ESW-XHE- ACP-DGE- ACP-D INJ-Fails GH LOSPNR18H FO-EWS FR-_EWB I AI-DE M[ .IGOHW1 -FR-EDGC R12HR R INJ-Fails -PCI and RCIC= 0.5, due to either harsh environment or loss of batteries.

Modifications and procedure changes 1, 2, 4, 5 significantly reduce the loss injectiom Based upon these mod NJ-Fails becomes: INJ-Fails (HPCI and RCIC fail to run 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />) and (Both diesel driven fire pumps fail to start and run 23 hours2.662037e-4 days <br />0.00639 hours <br />3.80291e-5 weeks <br />8.7515e-6 months <br />) INJ-Fails (0.1 x 0.1) x ((0.003 + 0.016) x (0.067 + 0.01 + 0.016)) = 1.77E-05 Sensitivity 2 Impact of Susquehanna Modes on Accident Sequence #1 (continued)

Modification 3 significantly reduces the chance of loss of DC power. Using NUREG 1150 numbers for the charger diesel we get 0.003 + 0.016 = 0.019 These modifications change the NUREG 1150 sequence by replacing the term INJ-Fails with failure: failure of either the charge diesel or the new injection capability, or (0.019 + 1.80E405)

= 0.019 Therefore the NUREG 1150 sequence becomes: P(cd cut set 14) = (3.70E-8/0.5) x 0.019 = 1.41E-09 Risk reduction from-these-models

= 2.0E-093.-70E-08

=0037959 Modification 6 allows onsite AC power to be recovered in less than 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />. lighting is available to perform the manipulation of these breakers P(swing diesel) = 0.3 Risk reduction from this diesel becomes a straight multiplier of 0.3 P(cd cut set 14) = (3.70E-08/0.5) x 0.019 x 0.3 = 4.22E-10:

Risk reduction from 1, 2, 3,4, 5 & 6 = 0.011388 Modification 7 result in the requirement that all four diesel fail for SBO. NUREG 1150 does not report common cause failure of 4 diesel to run. Therefore, Ill assume a common cause couple of 1.0 for the fourth given the third. Risk reduction for third and fourth diesel = 0.016 (NUREG-1 150 diesel fails to run) P(cd cut set 14) = 3.70E-08/0.5) x 0.028 x 0.3 x 0.016 = 6.75E-12 Risk reduction from 1, 2,3,4,5, 6 & 7 0.000182 Sensitivity 3 Impact of Susquehanna Design on Accident Sequence #3 Analysis of Susquehanna plant scram data shows that the loss of feedwater given plant trip with the MSIVs open is 0.28. In NUREG-1 150 a value of 1.0 was used since the water level is lowered below the MSIV isolation.

Susquehanna procedures keep the water level above the level I isolation; plus a bypass switch has been installed that bypasses the MSIV isolation on level 1. The MSIV isolation on high drywell pressure has also been removed. Therefore credit for feedwater is appropriate.

Modifications L. Level control band target above MSIV isolation set point. 2. Installation of a bypass switch that bypasses MSIV isolation on Level 1. 3. Removal of MSIV isolation on high drywell pressure.

Note: Susquehanna modeling would have all low pressure ATWS events with uncontrolled LPI proceed to core damage.Risk Reduction from modifications

= 0.28 Sensitivity 4 Impact of Modifications on Accident Sequence #4 Modifications:

1. Installation of a bypass switch that allows the operator to bypass the low pressure permissive

2. change to the Aux Load shed (LOCA load shed) scheme that allows the operator to reload the D condensate pump and inject water into the RPV. Failure of the low pressure permissive has no impact on the ability of the condensate pump to inject to the vesseL NOTE: The low pressure permissive circuit consists of two division which forms a one out of two taken twice logic. Division I uses Barksdale pressure sensors, which uses a bordon tube for pressure measurement.

Division II uses a Barton pressure sensor which uses a diaphragm for pressure measurement.

Since different instruments are used for pressure measurement, one would anticipate a less likely incident rate-for-CCF.---....

Medium break LOCA calculations demonstrate that the operator has at least 648 seconds after the RPV pressure decays to the HPCI low pressure trip. This provides the operator ample time to establish vessel injection from either condensate of 1 of 8 low pressure ECCS pumps. Two different operators control condensate and ECCS flow. With 10 minutes to establish flow the operator is at least as likely to initiate injection flow as to inject SLCS. Therefore a conservative value of 0.01 is applied.Risk reduction from Modifications

> 0.01 Sensitivity 5 Impact of modifications on Accident #7. This sensitivity is just like Sensitivity 2 except in this case HPCI is already failed. Therefore no credit is taken for the risk reduction associated with continued HPCI operation.

INJ-Fails (1.0 x 0.1) x ((0.003 + 0.016) x (0.067 + 0.01 + 0.016)) = 0.000177 These modifications change the NUREG 1150 sequence by replacing the term NJ-Fails with failure: failure of either the charger diesel or the new injection capability, or (0.019 + 1.70E-04)

= 0.019 Early failure of HPCI has little impact on the injection success rate due to the diversity on injection systems provided by the modifications.

Therefore the risk reduction is the same as Sensitivity

1. 2.

Sensitivity 6 Treatment of Battery Common Cause Failure Development of the battery failure rates used in the Susquehanna IPE are discussed in volume 3 Section C.7.2.3. The analysis was based upon a review of LER through 6131/87 and NPRDS data from 1/1/84 through 12131/89.

Based upon this data a battery failure rate on demand was estimated to be 2.40E-7/hr.

The authors of NUREG/CR-3831 report following values for battery failure from their investigations:

3.80E-08/hr (low), 6.40E-07 (recommended) and 3.OOE-06/hr (high). Clearly the Susquehanna value is within this range. A value of 3.OOE-06 was used in NUREG-1150.

Susquehanna did not include a common cause couple for batteries.

A common cause couple of 0.0023 was applied for failure of 4 batteries given failure of the first. This common cause couple is based upon work in NUREG-0666, and is largely attributed to common maintenance errors. The diesel generators at Susquehanna can utilize DC power from either unit. Maintenance is generally performed on the batteries during refueflng outages due to the 2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> AOT associated with a battery being inoperable.

Therefore this common cause couple should not apply across units. The batteries used to start the diesel are selected using a selector switch in the diesel bay. Upon LOOP with failure of the diesel to start a non-licensed operator (NPO) will be dispatched to the diesel bayes in alphabetical order to manually initiate the diesel EO-1/200/030).

Time studies show that no more than 10 minutes is required for the control room operator to observe the SBO, dispatch the NPO to the diesel bays and have the NPO at the A diesel paneL BWRSAR calculations show that given power uprate conditions, core damage will occur in 79 minutes following reactor trip and a high pressure boil off. Therefore the NPO has 69 minutes to identify the loss of DC on the engine panel, change the position of the DC power selector switch, place the engine in local and push the start button. The failure to recover offsite power in 68 minutes is estimated to be 0.1 in NUREG-1150.

This is considered a conservative estimate for diesel recovery since there are many causes of loop that require many hours to restore, where this recovery action requires positioning two selector switches and pushing the start button. Based upon this evaluation the NUREG-1150 common mode failure is applied to only one unit and an operator error of 0.1 is applied to the selection of the alternate battery supply.

Sensitivity 7 Impact of Modeling Differences and Low Pressure ECCS Control Logic on Accident Sequence 3 Susquehanna ATWS calculations, as well as BWROG EPG documentation, identify the potential for severe core damage if the reactor is depressurized while either unborated or slightly borated. In the case where SLCS is successful, but low pressure injection cannot be controlled, the injection flow may flush the boron out of the core and cause a power excursion.

For this reason without modification to the ECCS control logic, all low pressure ATWS events which rely on ECCS for core cooling result in core damage. PP&L has modified the ECCS control logic in a manner that allows the operator to ccmtrol the low pressure ECCS flow within the requirements of ATWS. Therefore with success of SLCS, the operator can feed the vessel with 2-3000 gpm using LPCL Prior to this modification the operator had to either lockout ECCS flow for five minutes or stop and start pumps. This modification was required to satisfy defense in depth. The operator has about 300 seconds in full ATWS and 460 seconds in partial ATWS to initiate a rapid depressurization to-avoid core damage.- An error rate of-0.001 was.assignedto failure to_ depressurize the reactor during ATWS based upon the time allowed. The operator can either feed with condensate or LPCI given installation of the control circuit modification.

The probability of failure to control low pressure injection is assigned at 0.01, given the ability to control flow and the 101 inch control band. Crediting the operator for initiating rapid blowdown has allowed us to uncover additional potential U significant operator errors during ATWS. Discovery of this error has led to a plant modification to allow the operator to control low pressure flow and a level control band that assures core cooling and reactivity control, while providing the operator with a procedure that can be implemented.

Replace NURBG-1 150 operator error -ESF-XHE-FO-DATWS

= 0.2 with [operator fails to depressurize or operator fails control low pressure injection)

= (0.001 + 0.01)= 0.011 Risk Reduction becomes: 0.011/0.2

= 0.055 This risk reduction is placed in the modeling column due to the different treatment in low pressure ATWS operaion.

Comparison of NUREG-1150 and Susquehanna Analysis Peach Item Description Bottom Susquehanna note T3A Transient with PCS initially available 2.5 2 T3C IORV 0.19 0.01 1 RPSM Mechanical ATWS 1.OOE-05 1.80E-05 2 SLCS-XIIIE-RE Driver Operator fails to restore SLCS after test 3.19E-02 2.OOE-03 3 SLCS-XHE-FO Driver Operator fails to initiate SLCS 2.OOE-02 2.OOE-03 3 SLC-SYS-TE SLC SLCS system unavailable due to test 3.40E-03 3 ESF-XHE-FO DATWS Operator fagils to d6pressuize during ATWS 0.2.. HCI-TDP-FS 20S37 Turbine Driven Pump Fails to start 3.00E-02 1.70E-02 T1 Loss of Offsite Power 0.079 0.071 ESW-XHE-FO EHS Operator fails to start Emergency heat Sink 0.9 na ACP-DGE-FR EDGB Emergency Diesel Generator falls to run 0.016 0.024 ACP-DGE-FR EDGC Emergency Diesel Generator fans to run 0.016 0.024 INJ-Fails Failure of Injection Systems 0.5 1 DGHWN R12HR Failure to recover Diesel at 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> 0.55 1 LOSPNR18HR Failure to recover offsite power at 18 hours2.083333e-4 days <br />0.005 hours <br />2.97619e-5 weeks <br />6.849e-6 months <br /> 0.0074 0.069 PLA-5425 ATTACHMENT I CONTENTS 1. SABRE Computer Code Documentation:

"* Calculation EC-ATWS-0505, Rev. 8 "* Calculation EC-052-0593, Rev. 0 "* Calculation EC-SIMU-0501, Rev. 0 "* Calculation EC-SATH-1 007, Rev. 1 2. SABRE Calculations Supporting Technical Specification S.... Chan-- D-eletiiiig-Automatic HPCI Suction Transfer on High Suppression Pool Level: "* Calculation EC-052-1018, Rev. 2 "* Calculation EC-052-1025, Rev. 2 CALCULATION

/ STUDY COVER SHEET and NUCLEAR RECORDS TRANSMITTAL SHEET 1. Page 1 of Total Pages 3&.>2. TYPE: 5. TRANSMITTAL#:

>9. DESCRIPTION:

Reactor Dynamics I Calc. >3. NUMBER: EC-ATWS-0505

>4. REVISION:

8 A >6. UNIT: 3 *>7 QUALITY CLASS: Q SABRE: A Computer Code for Simulation of Boiling Water *>8. DISCIPLINE:

T Under Failure-to-Scram Conditions

10. Alternate Number: RA-B-NA-0 12: Computer Code or Model used: 13. Application:

• > 14 Affected Systems:

• If N/A then line 15 is mandatory.

• • >15. NON-SYSTEM DESIGNATOR:

• • If N/A then line 14 is mandatory

16. Affected Documents:

45 SABRE, SIMULATE, SIMTRAN SUPERSEDED BY: 11. Cycle: Fichee x Dis 058 ATWS EOPC 0 SAR Change Req'd 17.

References:

PLA-4480, FSAR Section 15.8 18. Equipment

/ Component

1. 19. DBD Number: >20. PREPARED BY Mark A. Chaiko 14h,. Print Name Siwaure >21. REVIEWED BY Kevin W. Brinckman Print Name >21A. VERIFIED BY Kevin W. Brinckman Print Name -sgatr >22. APPROVED BY Casimir A. Kukielka Print Name 2 "/ v >23. ACCEPTED BY PP&L I DATE PrintLName Signature BDATE NELERVE TO BE COMPLETED BY NUCLEAR RECORDS NR-DCS SIGNAURE/ATE(4 4 , Cj SEP 2 700 ADD A NEW COVER 1'AUB ?UK PAAG-1 FEH VISIUN FORM NEPM-QA-0221-1, Revision 3, Page 1 of 2, ELECTRONIC FORM 0 N .Verified Fields EC (J Arn't q XK4 Thyvn rr II Zc ENGINEERING CALCULATION STUDY PPL, Inc REVISION DESCRIPTION SHEET REVISION NO: 8 CALCULATION NUMBER: EC-ATWS-0505 This form shall be used to record the purpose or reason for the revision, indicate the revised pages and / or affected sections and give a short description of the revision.

Check ( x ) the appropriate function to add, replace or remove the affected pages. Revised Affected A R R Description

/ Res Aected d p m Purpose of Revision Pages Sections d

• d I v 4 Contents x Added Section 5.15 6 Computer Case x Reran last SIMULATE case in Table (new Run# is 0002019) Summary 13 ......1 -............

....x- -Provided-overall-description-of code-changes......

21 2.1 x Fixed typos in Eq. (2.1-3) 83 2.10 x Modified equation (2.10-1) because of addition of"rainout" model. 84 2.10 x Changed Eq. (2.10-4).

Added definition of enthalpy of water vapor at surface of drywell pool. Also incorporated "rainout" model. 1 5 2.10 x Incorporated "rainout" model. 5a 2.10 x Incorporated "rainout" model. M 7,91 2.10 x Incorporated "rainout" model. 110 3.2 x Corrected typos in Eqs. (3.2-3), (3.2-4), and (3.2-6).

122,123,124, 5.1 x Reran SABRE Case 01. 125 128, 129, 130, 5.2 x Reran SABRE Case 02. 131 134, 135, 136 5.3 x Reran SABRE Case 03. 139, 140 5.4 x Reran SABRE Case 04. 141, 142, 143, 5.5 x Reran SABRE Case 05. Also modified Case 05 to get drywell conditions and 144, 145, 146, break conditions that would initiate rainout.

149, 150, 5.6 x Reran SABRE Case 06. 153, 155, 156, 5.7 x Reran SABRE Case 07. 157, 158 159, 160, 163- 5.8 x Reran SABRE Case 08. 172 REVISION TYPE: F1 SUPERSEDED BY CALCULATION NUMBER EC (check one) I] FULL REVISION 0 PAGE FOR PAGE FORM NEPM.QA-0221-2, Revision 2, ELECTRONIC FORM S ENGINEERING CALCULATION STUDY PPPL, Inc REVISION DESCRIPTION SHEET REVISION NO: 8 [CALCULATION NUMBER: EC-ATWS-0505 This form shall be used to record the purpose or reason for the revision, indicate the revised pages and / or affected sections and give a short description of the revision.

Check (x) the appropriate fiinction to add, replace or remove the affected pages. Reie fetd A R R DescriptionI Reied Afetd d p In Purpose of Revision Pages Sections d I 179,181-185 5.10 x Reran SABRE Case 10. Reduced max step size from 100 msec to 50 msec. 188,189 5.11 x Reran SABRE Case 11. 191, 194, 195- 5.12 x Reran SABRE Case 12. -199 ___ _ 203-212 5.13 x Reran SABRE Case 13. 213,214,216, 5.14 x Reran SABRE Case 14. 217,218,219, 220, 221 __________

___ 221a-2211 5.15 x Added SABRE Case 15 to check drywell temperature response to small steam I break. 324 F.69 x Corrected typs. N/A x Revised cover sheet. la N/A x Changed revision page l1b N/A x Added revision page REVISION TYPE: j) SUPERSEDED BY CALCULATION NUMBER EC (check one) 5 FULL REVISION 0 PAGE FOR PAGE FORM NEPM-QA-0221-2, Revision 2, ELECTRONIC FORM PAGE 2 CONTENTS COMPUTER CASE

SUMMARY

6. 1. INTRODUCTION 13 2. MODEL DESCRIPTION 19 2.1 Fluid Dynamics Equations for Jet Pump, Lower Plenum, Core, Bypass, 19 Upper Plenum, Riser, and Separator Regions 2.2 Fluid Dynamics Equations for Downcomer and Steam Dome Regions 26 2.3 Fuel and Cladding Heat Balance Equations 29 2.4 Neutron Kinetics 31 2.4.1 Neutron Kinetics Parameters 31 2.4.2 GovemingEquations 33 2.4.3 Methodology for. Initial Flux Distribution 34 2.4.4 Methodology for Transient Flux 44 2.4.5 Modification of Thermal Absorption Cross Section to Account 47 for Dissolved Boron 2.4.6 Boron Transport Model 54 2.4.7 Decay Heat Fraction 58 2.4.8 Control Rod Model 59 2.5 Recirculation Pump Model 74 2.6 Main Steam Line Model 76 2.7 Feedwater Model 80 2.8 Control Rod Drive and Core Spray Models 82 2.9 HPCI and RCIC Models 82 2.10 Containment Model 83 2.11 Heat Capacitance of Vessel and Internals 94 3. NUMERICAL SOLUTION METHOD 96 3.1 Control-Volume Formulation of Flow Equations.

96 3.1.1 Nodalized Continuity Equation 96 3.1.2 Nodalized Energy Equation 98 3.1.3 Nodalized Momentum Equations 100 3.1.3.1 Nodal Representation of Inertial Term 100 3.1.3.2 Nodal Representation of Convective Term 103 3.1.3.3 Nodal Representation for Spatial Integral of 103 Pressure Gradient 3.1.3.4 Nodal Representation of Gravitational Term 104 3.1.3.5 Nodal Representation of Wall and Spacer Friction Terms 104 3.1.3.6 Nodal Representation of Channel Inlet and 105 Outlet Friction Terms 3.1.3.7 Overall Momentum Equations 105 3.1.4 Nodalized Fuel and Cladding Heat Balance Equations 107 ,3.2 Calculation of the Flow Regime 109 3.3 Calculation of Gr ,dcQ, and dP'/d 115 3.4 Initialization and Temporal Integration Procedures.

116 4. DISCUSSION OF CODE LIMITATIONS 118 4.1 Single-Channel Core Model 118 4.2 Steam Line Model and Feedwater Controller.

118 4.3 Separator Model: 118 4.4 Downcomer Water Level 118 5. CODE BENCHMARKING 119 5.1 MSIV Closure with Scram--Comparison to Plant Data 119 5.2 Containment Response for Unmitigated ATWS-Comparison 126 to CONTAIN Code 5.3 MSIV Closure ATWS-Comparison to PP&L RETRAN Calculations 132 5.4 Inventory Boildown with Scram Failure-Comparison to SIMUTALE-E 137 5.5 Small Break LOCA-Comparison to CONTAIN Results 141 5.6 Turbine Trip ATWS from MEOD Rod Line-Comparison to 147 GE TRAC Results 5.7 ATWS Initiated by Pressure Regulator Failure Open-Comparison to 152 GE Results 5.8 MSIV-Closure ATWS with Boron Injection-Comparison to GE Calculations 159 5.9 Calculation of Counter-Current Flow Limit-Comparison to 173 Kutateladze CCFL Correlation 5.10 Suppression Pool Heatup from Decay Heat-Comparison to 177 Heat Balance Calculation 5.11 Small Break LOCA -Comparison to GE SAFER/GESTR-LOCA Results 186 5.12 Check of SABRE Mass and Energy Balances with Core Spray Injection 190 5.13 Check of SABRE Steady-State Initialization and Fuel/Clad Temperature 200 Calculations with Partial Length Fuel Model 5.14 MSIV Closure ATWS with SLCS Failure-Shutdown with MRI 5.15 Drywell Temperature Response During Small Steam Break APPENDIX A DERIVATION OF DOWNCOMER AND STEAM DOME MASS AND ENERGY BALANCES A.1 Downcomer Region. A.2 Steam Dome Region. APPENDIX B DERIVATION OF FUEL AND CLADDING HEAT BALANCE EQUATIONS.

APPENDIX C DERIVATION OF ERROR TERM IN EQUATION (3.1.2-1)

APPENDIX D AUXILIARY CALCULATIONS D.1 Geometric and Hydraulic Parameters for Code Input. D. 1.1 Jet Pump Region. ..D-li2-Lower-PenumnRegion.- . D.1.3 Core Region. D.1.4 Bypass Region. D.1.5 Upper Plenum Region. D.1.6 Riser Region. D. 1.7 Separator Region. D.1.8 Steam Dome Region. D.1.9 Downcomer Region D.2 Data for Reactor Heat Structures D.3 Partial Derivatives of Fluid Density D.3.1 Subcooled Liquid D.3.2 Two-Phase Mixture D.3.3 Super-Heated Vapor D.4 Steam Condensation Rate D.5 Elevation of Feedwater Spargers D.6 SRV Flow Area and Critical Mass Flux Correction D.7 Time Constants for Coastdown of Recirculation Pump Flow and Feedwater Enthalpy D.8 Delayed Group Fractions and Decay Constants D.9 Valve Data for Turbine-Trip and MSIV-Closure Transients D. 10 Safety/Relief Valve Set Points D. 11 Control Rod Insertion Time D.12 Set Point for ATWS Recirculation Pump Trip D.13 Time Delay for Recirculation Pump Trip on Main Turbine Trip D.14 Pressure Regulator Gain 213 221a 222 222 224 225 230 231 231 231 -233 234 238 239 239 240 242 243 246 248 248 248 250 250 252 252 253 253 254 255 255 256 256 256 I D. 15 Feedwater Controller Gain and Time Constant 256 D. 16 Boron Transport Time 257 D. 17 Model Parameters for Boron Injection System 257 D. 18 Data for Containment Model 258 D.18.1 Initial Pool Temperature 258 D.18.2 Cross-Sectional Area of Suppression Pool 258 D,18.3 Initial Suppression Pool Water Level 258 D.18.4 Initial Suppression Chamber Air Space Temperature 258 D.18.5 Heat Load From Reactor Vessel 259 D. 18.6 Drywell Cooling Load 259 D.18.7 Drywell Free Volume 260 D.18.8 Initial Drywell Temperature 260 D.18.9 Initial Relative Humidity in Drywell 260 D.18.10 Initial Relative Humidity in Wetwell 260 D.18.11 Initial Drywell Pressure 260 D.18.12 Initial Wetwell Pressure 260 D. 18.13 Heat-Transfer-From SRV Tailpipe -.. ... ..260 D.18.14 Surface Area of Drywell Heat Structures 262 D.18.15 Volume of Drywell Steel Structures 264 D.18.16 Surface Area of Wetwell Heat Structures 264 D.18.17 Volume of Wetwell Steel Structures 264 D. 18.18 Characteristic Length of Containment Heat Structures 265 D.19 Model for Wide Range Level Indication 265 D.20 Core Spray Flow 266 APPENDIX E FORTRAN Program Used to Calculate Heated Channel 268 Response in Section 5.9 APPENDIX F Base SABRE Input Deck for Power Uprate 283 Conditions with 9x9 Core (U2C9) APPENDIX G Base SABRE Input Deck for Power Uprate 326 Conditions with 10x10 Core (U2C10)APPENDIX H Base CONTAIN Input Deck for SABRE Benchmark Studies 344

/r- -d COMJPUTER CASE SUW¶ARY {SDaULATE=E CASES)9904110 X U2C7 3-D Cross-section and Power distribution calculation.

Run includes 7 cross-section cases for the following rod patterns:

1.) 4 rods in + 4 rods at notch 10 2.) 8 rods in + 4 rods at notch 10 3.) 12 rodsin+4rodsat notch 10 4,) 20 rods in + 4 rods at notch 10 5.) 36 rods in + 4 rods at notch 10 6.) 68 rods in + 4 rods at notch 10 7.) all rods in Run #9904110 was restarted from data set r9507299.0012 Cycle exposure = 11.562 GWDIMTU 9904111 X U2C9 3-D Cross-section and Power distribution calculation.

Run includes 7 cross-section sets for the following rod patterns:

1.) all rods out 2.) 4 rods in 3,) 8 rods in 4.) 16 rods in 5.) 32 rods in 6.) 64 rods in 7.) all rods in. Run #9904111 was restarted from data set r9602506.0002 Cycle exposure = 15.700 GWD/MTU 9904112 U2CI 0 3-D Cross-section and Power distribution calculation.

Run includes 7 cross-section sets for the following rod patterns:

1.) all rods out 2.) 4 rods in 3.) 8 rods in 4.) 16 rods in 5.) 32 rods in 6.) 64 rods in 7.) all rods in. Run #9904112 was restarted from data set r9808566.0001 Cycle exposure = 15.2 GWD/MTU 9905326 X U1C12 3-D Cross-section and Power distribution calculation.

Run includes 7 cross-section sets for the following rod patterns:

1.) all rods out 2.) 4 rods in 3.) 8 rods in 4.) 16 rods in 5.) 32 rods in 6.) 64 rods in 7.) all rods in. Run #9904113 was restarted from data set r9903AS4.0001 Cycle exposure = 14.800 GWDIMTU 0002019 X U2C10 Core power calculation with 16 rods inserted.

Run #0002019 was restarted from data set r9808566.0001 Cycle exposure = 15.2 GWDIMTU Power result is compared against SABRE in Section 5.14.

COMPUTER CASE

SUMMARY

(SIMTRAN-E CASES) Ron ~ ei Mi es~slc POO 9900339 X U2C7 1-D Cross-section set calculation Run includes 7 cross-section sets for the following rod patterns:

1.) 4 rods in + 4 rods at notch 10 2.) 8 rods in + 4 rods at notch 10 3.) 12 rods in + 4 rods at notch 10 4.) 20 rods in + 4 rods at notch 10 5.) 36 rods in + 4 rods at notch 10 6.) 68 rods in + 4 rods at notch 10 7.) all rods in SIMTRAN Run #9900339 was run using SIMULATE restart file generated from SIMULATE Run#99041

10. Cycle exposure = 11.562 GWD/MTU 9900340 x U2C9 1-D Cross-section set calculation Run includes 7 cross-section sets for the following rod patterns:

1.)all rods out 2.) 4 rods in 3.) 8 rods in 4.) 16 rods in 5.) 32 rods in 6.) 64 rods in 7.) all rods in. SIMTRAN Run #9900340 was run using SIMULATE restart file generated from SIMULATE Run#9904111 Cycle exposure = 15.700 GWDIMTU 9900341 X U2C10 1-D Cross-section set calculation Run includes 7 cross-section sets for the following rod patterns:

1.) all rods out 2.) 4 rods in 3.) 8 rods in 4.) 16 rods in 5.) 32 rods in 6.) 64 rods in 7.) all rods in. SIMTRAN Run #9900342 was run using SIMULATE restart file generated from SIMULATE Run#9904112 Cycle exposure = 15.2 GWVD/MTU 9900369 x u1C12 1-D Cross-section set calculation Run includes 7 cross-section sets for the following rod patterns:

1.) all rods out 2.) 4 rods in 3.) 8 rods in 4.) 16 rods in 5.) 32 rods in 6.) 64 rods in 7.) all rods in. SIMTRAN Run #9900343 was run using SIMULATE restart file generated from SIMULATE Run#99041113 Cycle exposure = 14.800 GWD/MTU 9900370 X U2C7 1-D Cross-section set calculation.

Same as Run #9900339 except that input parameter limnft was set to 13 instead of 50 to eliminate warning message in output. Comparison of output files showed that there was no difference in kinetics parameters.

The cross-section sets for Run #9900339 and 9900370 are identical.

.-I PAGE 8 COMPUTER CASE

SUMMARY

(CASMO-3 CASES)[ Dvi Lattice calculations to determine dependency of macroscopic B-10 thermal absorption cross section on boron concentration and void fraction.

Calculations are for ATRIUM-10 fuel at hot zero power conditions and lattice type 92. Void fraction ranges from 0 to .80 and boron concentration ranges from 0 to 800 ppm. Cycle exposure ranges from 0 to 60 GWDIMTU.I I ii i i i I ii- ii i i i i -Ii r ; -r

• PAGE 9 COMPUTER CASE

SUMMARY

(SABRE CASES) 01 MS IV Closure with scram -comparison to plant data (Q 5.1) SABRE input file is cOl.dat kinetics file is u2c7.simtran.out (SIMTRAN Run#9900339) 02 X Containment response for unmitigated ATWS -comparison against CONTAIN code(Q 5.2) SABRE input file is c02.dat Kinetics file is u2clO.simtran.out (SIMTRAN Run#9900341) 03 X MSIV closure ATWS -Comparison to PP&L RETRAN Calculations

(§5.3) SABRE input file is c03.dat Kinetics file is u2c9.simtran.out (SIMTRAN Run#9900340) 04 X Inventory Boildown with Scram Failure -Comparison to SIMULATE-E static calculations(§ 5.4) SABRE input file is c04.dat Kinetics file is u2cg.simtran.out (SIMTRAN Run#9900340) 05 X Containment response to a Small break LOCA -Comparison to CONTAIN Results (§ 5.5) SABRE input file is c05.dat Kinetics file is u2c7.simtran.out (SIMTRAN Run# 9900339) 06 X Turbine trip ATWS from MEOD rod line -Comparison to GE TRAC Results (§ 5.6) SABRE input file is c06.dat Kinetics file is u2c7.simtran.out (SIMTRAN Run#9900339) 07 X ATWS Initiated by Pressure Regulator Failure Open -Comparison to GE Results (§ 5.7) SABRE input file is c07.dat Kinetics file is u2c7.simtran.out (SIMTRAN Run#9900339) 08 X MSIV closure ATWS with Boron Injection

-Comparison to GE Calculations.

(§ 5.8) SABRE input file is c08.dat Kinetics file is u2c7.simtran.out (SIMTRAN Run#9900339) 08a X Same as Case 08 but one SRV out of service.

08b X Same as Case 08 but hydraulic time step size changed from 25 msec to 20 msec for t>30sec.

08c X Same as Case 08 but hydraulic time step size changed from 25 msec to 30 msec for t>30sec.

08d X Same as Case 08 but hydraulic time step size changed from 25 msec to 35 msec for t>30sec.

08e X Same as Case 08 but hydraulic time step size changed from 5 msec to 10 msec for t<30sec.

08f X Same as Case 08 but hydraulic time step size changed from 5 msec to 2.5 msec for t<30sec.

08g X Same as Case 08 but number of boron mixing nodes in core changed from 10 to 9. 08h X Same as Case 08 but number of boron mixing nodes in core changed from 10 to 8. 08i X Same as Case 08 but hydraulic time step size changed from 5 msec to 1.25 msec for t<30sec.

PAGE 10 COMPUTER CASE

SUMMARY

(SABRE CASES) 1icqhe- .. 08j X Same as Case 08 but error control parameters, RTOL and ATOL, for kinetics solution are increased from 1 .E-05 to 5.E-05. 08k X Same as Case 08 but error control parameters for kinetics solution, RTOL and ATOL, are decreased from 1 .E-05 to 5.E-06. 081 X Same as Case 08 but boron entrainment exponent b in Eq. (2.4.6-1 ) is set to 1 A. 08m x Same as Case 08 but boron entrainment exponent b in Eq. (2.4.6-1) is set to 2. 10 X Suppression Pool heatup from decay heat -Comparison to Heat Balance Calculation. (Q 5.10) SABRE input file is cl 0.dat Kinetics file is u2clO.simtran.out (SIMTRAN Run#9900341) 11 X Small break LOCA compared to GE results (§ 5.11) SABRE input file is cl l.dat Kinetics file is. u2c7.simtran.out (SIMTRAN Run#9900339) 12 X Check of SABRE mass/energy balances with Core Spray injection (Q 5.12) SABRE input file is c1 2. dat Kinetics file is u2c9.simtran.out (SIMTRAN Run#9900340) 13 X Check of SABRE Fuel/Clad Temperature Calculations with partial length fuel rods (J 5.13) SABRE input file is c1 3. dat Kinetics file is u2clO.simtran.out (SIMTRAN Run#9900341) 14 X MSIV-Closure ATWS with SLCS failure-Shutdown with MRI. (§ 5.14) SABRE input file is c14.dat Kinetics file is u2clO.simtran.out (SIMTRAN Run#9900341) 14a x Same as Case 14 except that run was made using SABRE Version 2.4 (Point Kinetics model)

F- % -I I COMPUTER CASE

SUMMARY

(CONTAIN CASES) Diskett Micro-Description f ichie 02 x Containment response for unmitigated ATWS -comparison against SABRE code(§ 5.2) Mass and energy source tables are developed from SABRE output for SABRE Case 02. 05 x Containment response to a Small break LOCA -Comparison to SABRE Results (§ 5.5) Mass and energy source tables are developed from SABRE output for SABRE Case 05.

PAGE 12 COMPUTER CASE

SUMMARY

.(SABRE Source Code Listing and Input Files)Dskette ýMiCO De 0ipif -' fiche~~-;>.-~

X X filename=sabre_30.f

-FORTRAN source code listing of SABRE code Version 3.0 x X filename=common.txt

-Common blocks for SABRE code ____ __ _____Version 3.0 X ___ filename=cOl .dat -input file for SABRE Case 01 x ____filename=c02.dat

-input file for SABRE Case 02 X ___ filename=c03.dat

-input file for SABRE Case 03 X ___ filename=c04.dat

-input file for SABRE Case 04 X ___ filename=c05.dat

-input file for SABRE Case 05 X filename=c06.dat

-input-file for SABRE Case 06. x filename=c07.dat

-input file for SABRE Case 07 X filename=c08.dat

-input file for SABRE Case 08 X filename=cO~a.dat

-input file for SABRE Case 08a X filename=cO8b.dat

-input file for SABRE Case 08b X ___ filename=c08c~dat

-input file for SABRE Case 08c x filename=cO~d.dat

-input file for SABRE Case 08d X ___ filename=cOfe.dat

-input file for SABRE Case 08e X filename=cO8f.dat

-input file for SABRE Case 08f X filename=cO8g-dat

-input file for SABRE Case 08g X fllename=cO8h.dat

-input file for SABRE Case 08h X ___ filename=cO8i.dat

-input file for SABRE Case 08i X ___ filename=cO8j.dat

-input file for SABRE Case 08j X ___ filename=cO8k-dat

-input file for SABRE Case 08k X ___ filename=c08l.dat

-input file for SABRE Case 081 X ___ filename=cO8m.dat

-input file for SABRE Case 08m, x ____filename=cl O~dat -input file for SABRE Case 10 x filename=cl I .dat -input file for SABRE Case 11 X ___ filename=c12.dat

-input file for SABRE Case 12 X filename=cl 3.dat -input file for SABRE Case 13 X ___ filename=c1

4. dat -input file for SABRE Case 14 x ____filename=u2c9.base.dat

-base 9x9 core model x ___filename=u2cl0.

base.dat -base 10xl0 core model tt /3 1. INTRODUCTION Revision 8 (SABRE version 3.1) to this calculation incorporates a correction to the primary containment temperature/pressure calculation.

In previous versions of SABRE, the specific enthalpy of steam in the drywell and wetwell atmospheres was approximated by the specific enthalpy of saturated steam at the drywell/wetwell temperature.

Although this approximation was acceptable for analysis of transients involving liquid breaks and/or steam discharge to the suppression pool via SRVs, it was found to produce significant error in drywell temperature response in the case of a steam break. Therefore, a function routine for computing superheated enthalpy as a function of temperature and pressure was added to SABRE. Enthalpy is computed using the ASME FORTRAN function hss. When adding the function routine hss, the lower level routines hip and hgp (saturated liquid and vapor enthalpy as a function of pressure) also had to be included.

This generated a conflict with existing function routines of the same name. Consequently, hip and hgp were replaced with the corresponding ASME routines.

This correction involving the superheated steam enthalpy does not affect any calculation of record as no steam-break analyses have previously been performed with SABRE. As part of this revision, two additional corrections were made. It was found that the input data defining the parameters AP. and APF 0 in Eqs. (2.10-1l)-(2.10-13) were not passed correctly to subroutine concal.f because of incorrect naming of the variables in the main program. The variable names were changed in the main program. This error caused these parameters-to take the value 0.0 in previous versions of SABRE. As a result, the vacuum breakers opened somewhat earlier than they should have. From rerunning benchmark problems in §5, it was determined that the effect of this coding error is not significant.

The second correction involves adding a "rainout" model to prevent the primary containment atmospheres from becoming supersaturated with water. This model can be important when simulating containment respose to liquid breaks in which the break flow becomes subcooled relative to the drywell atmosphere conditions.

If this occurs, the drywell will start to cool down and atmosphere conditions may become saturated.

The rainout model removes excess water from the atmosphere and deposits it in the pool thus preventing the relative humidity from exceeding 100%. This calculation package provides documentation of the SABRE code and benchmarking of the SABRE reactor and primary containment models. In addition, two base-case SABRE input decks and four 1-D cross-section files are developed in this calculation package. The two base case SABRE input decks correspond to an ANF 9x9-2 core model and to a Siemens ATRIUM-10 core model. SABRE can only model the physical characteristics of one fuel type. Therefore, mixed cores of 9x9 and 10xl0 fuel are modeled by approximating the core as a full core of the dominant fuel type. An effective fuel-clad gap conductance is used to make the thermal response of the core model consisting of the dominant fuel type consistent with the actual mixed core. In addition, the cross-section data is developed from a SIMULATE model of the actual mixed core. Overview of SABRE Code In an ATWS (Anticipated Transient Without Scram) event, a BWR (boiling water reactor) may be operated over a wide range of conditions far removed from those encountered during normal operation.

In a non- " isolation ATWS (main steam isolation valves open) for instance, the reactor operates in natural circulation with large reductions in feedwater enthalpy caused by loss of turbine extraction steam flow to feedwater heaters. Decreased make-up-flow enthalpy results in increased core-inlet subcooling which can lead to unstable power oscillations.

1 If the ATWS event involves closure of the main steam isolation valves (MSIVs), a drop in water level occurs due to loss of feedwater flow. Vessel makeup is then provided by the HPCI (high pressure coolant injection) system which supplies highly subcooled water to the vessel at a reduced flow rate. At the low reactor water levels which result with HPCI injection, the feedwater spargers become uncovered and cold make-up coolant is injected directly into a region occupied by saturated steam. The development of condensation on the injection flow has a significant modulating influence on increasing core-inlet subcooling.

An ATWS may also involve depressurization of the reactor vessel as a mitigative response to failure of certain equipment.

For example, should I-PCI fail to function, the reactor could be depressurized below -600 psia to allow coolant injection by intermediate-pressure-range condensate pumps. In order to properly formulate a mitigative strategy for response to an ATWS event, an understanding of the reactor dynamic behavior over the wide range of conditions described above is required.

Consequently, the SABRE (Simulation of ATWS in Boiling-Water Reactors) computer code was developed by PP&L to simulate BWR transient behavior under natural circulation conditions with failure to scram. SABRE contains thermal-hydraulic models of the reactor jet pump, lower plenum, core, bypass, upper plenum, riser, separator, steam dome, and downcomer regions of the reactor vessel. A three-node, radially-lumped parameter model describes fuel-to-coolant heat transfer in each of the axial nodes within the core region. Nuclear heating effects are simulated using a two-group, one-dimensional kinetics model. l-D cross section files are developed for U2C7, U2C9, U2C10, and U1C12 cores. The SABRE code also includes a model of the primary containment.

In an isolation ATWS, reactor steam is discharged to the primary containment where it is condensed within the suppression pool. As pool temperature rises, the containment begins to pressurize, and alternate methods of reactor shutdown (boron injection or manual insertion of control rods) are required to maintain containment structural integrity.

It is important to emphasize, in the context of emergency operating procedure development, that SABRE results are not used to justify reactor operation under conditions where core dynamics are poorly behaved by demonstrating that fuel integrity can be maintained.

On the contrary, at PP&L, the SABRE calculations are used to identify operating regimes where core/containment integrity is likely to be threatened.

The mitigative strategy for ATWS is then constructed to avoid these severe operating 1 Wulff, W., Cheng, H.S., and Mallen, AXN., "Causes of Instability at LaSalle and Consequences from Postulated Scram Failure", in Proceedings of International Workshop on Boiling Water Reactor Stability, Holtsville, New York, October 17-19, 1990.

i1UL1 regimes whenever possible.

Consequently, the reactor is maintained within an operating domain where there is confidence that core and containment integrity will be preserved.

Although SABRE was developed to study reactor behavior under ATWS conditions, it can also be used to investigate reactor and containment response to small break LOCAs or anticipated reactor transients such as an MSIV closure with scram. The NRC has examined SABRE results for ATWS scenarios as part of their review of proposed changes to the BWR Owners' Group Emergency Procedure Guidelines.

2 ATWS simulations were performed by the NRC using the TRAC-BF1 and RAMONA 4B computer codes. When discussing their results in the Safety Evaluation Report, the NRC makes the following statement, "Results obtained with these two codes were reasonably consistent and were also comparable to PP&L findings for Susquehanna BWRs using SABRE, after making adjustments to compensate for different procedural assumptions." It is concluded, based on this Safety Evaluation, and subsequent conversations with the NRC 3 , that SABRE can be used for ATWS analysis.

The FORTRAN listing of the SABRE code and the base-case SABRE input files are included on diskette (see Computer Case Summary).

In addition, the SABRE neutron kinetics data bases which were developed with the SIMULATE-E code, are also included on microfiche as indicated in the Computer Case Summary.

The basic assumptions and modeling features used in the SABRE code are summarized below: Reactor Model 1. The flow is one-dimensional, and each flow region (core, bypass, upper plenum etc.) has an axially-uniform flow area. 2. Acoustic waves travel at infinite speed. 3. Potential and kinetic energy effects are neglected.

4. Slip between phases is governed by a Drift-Flux model. 5. Complete vapor-liquid separation occurs at exit of steam separator (steam dryer not explicitly modeled).

6. Flow regime can be co-current or counter-current.

2 Nuclear Regulatory Commission, "Safety Evaluation Report Modifications to the Boiling Water Reactor (BWR) Emergency Procedure Guidelines to Address Reactor Core Instabilities," June 6, 1996. 3 PLA-4480, "Unit 2 Cycle 9 ATWS Evaluation," File R41-2, Docket No. 50-388, July 23, 1996.

PAGE 16 7. The axial power shape is non-uniform and time varying.

8. The 764 fuel bundles are averaged into a single channel.

9. Gamma heating within fuel bundles is specified as part of input data. 10. Gamma heating in bypass channel is specified as part of input data. 11. Efficiency of steam condensation is 95% when downcomer level is a meter or more below the feedwater nozzles.

12. Twenty-seven axial hydraulic nodes are used for the core region. 13. Twenty-five axial and two radial nodes are used to model the fuel. The cladding is modeled with twenty-five axial nodes and one radial node. The number of fuel pins can vary axially along the core. 14. In the outer radial node of the fuel, the volumetric heat generation rate is 10% greater than the radially-averaged heat generation rate to account for self shielding effects.

15. Core power is computed using a two-group l-D kinetics model with six delayed neutron groups. 16. The kinetics model includes reactivity contributions from control rod insertion, boron injection, and variations in moderator density and fuel temperature,.

17. Axial power shape for fission power changes with core conditions based on the 1 D kinetics model, but the power shape for decay power remains equal to the initial axial power shape. 18. Injected boron stagnates in lower plenum if total core flow is less than 5 MLb/hr. 19. Stagnated boron re-mixes if total core flow exceeds 15 MLb/hr. 20. Perfect mixing occurs in each boron mixing node. 21. Dissolved boron does not affect coolant density. The density of injected boron solution is assumed to be the same as water at the same temperature.

22. Boron volatility is neglected.

23. Boron transport is calculated from coolant flow. 24. In general, a boron mixing node consists of several hydraulic nodes.

PAGE 17 25. The thermal absorption cross section for B-10 is assumed to be a linear function of the boron concentration in the fuel channels and the bypass region. 26. Model includes thermal capacitance of reactor vessel and vessel internal structures.

Sensible heat given off by these structures is superimposed on core power. 27. Pump power decays exponentially following a recirculation pump trip. 28. For a recirculation pump runback, pump power decays exponentially to a lower value which maintains the desired total core flow. 29. An inertial flow model describes steam flow through closing MSIVs. 30. Pressure wave phenomena in steam lines is neglected.-

31. HPCL/RCIC extraction steam flow is included in steam dome mass/energy balances.

32. SRV flow and break flow during a LOCA are computed with homogeneous critical flow model. 33. A proportional controller approximates the feedwater controller response.

34: Upon loss of feedwater heating, the decrease in feedwater temperature is approximated by a delayed exponential decay. 35. CRD flow reaches thermal equilibrium with coolant in lower plenum; flow enters reactor in lower node of bypass region. 36. RPV injection systems consist of feedwater, HPCI, RCIC, condensate, Core Spray, SLCS, and CRD. Containment Model 37. The drywell and wetwell are each modeled with a single control volume. 38. Model includes thermal capacitance of structural steel and liner plate, but thermal capacitance of concrete structures is neglected.

39. Pressure within drywell and wetwell is computed with Ideal Gas equation of state. 40. Vacuum breakers and downcomer vents are modeled.

41. Drywell coolers and suppression pool cooling system are modeled.

42. Suppression pool letdown can be described by specifying a constant letdown flow rate. 43. Model includes heat transfer from SRV tailpipes to suppression chamber atmosphere.

44; Heat structure model includes natural convection, condensation, and radiation heat transfer effects.

45. Drywell heat load includes dissipation from reactor vessel. 46. Wetwell air space varies with suppression pool level. 47. Pool layer can form on drywell floor. Depth of pool layer can reach-op ot downcomer pipes. 48. Condensation on liner plate and structural steel drains into drywell/wetwell pool. 49. Effect of pool formation on drywell free volume is negligible.

50. Model includes heat transfer/evaporation effects at surface of drywell/wetwell pool. 51. For LOCA, break flow comes to pressure equilibrium with drywell.

In this calculation references are generally provided in the form of footnotes.

Please be aware that all footnotes begin with the number 1 in each section of the calculation.

Within a section of the calculation, there is no cross reference to footnotes outside that particular section.

PAGE 19 2. MODEL DESCRIPTION Figure 2-1 shows a schematic diagram of the SABRE model of the SSES reactor. A single core channel is used to represent the average flow and power conditions within the 764 fuel bundles.

Models of the bypass, upper plenum, riser, separator, steam dome, downcomer, jet pump, and lower plenum regions are also included.

Vessel coolant sources consist of feedwater flow, high pressure injection, condensate, core spray, SLCS (Standby liquid control), and CRD (control rod drive) flow. In this Section, the governing thermal-hydraulic equations for the various regions of the reactor vessel are presented.

A description of the fuel and cladding heat balance equations and the nuclear heating model are then given. 2.1 Fluid Dynamics Equations for Jet Pump, Lower Plenum, Core, Bypass, Upper Plenum, Riser, and Separator Regions Within the scope of the present model, the fluid at a particular location within a region can consist of liquid (subcooled or saturated), vapor (saturated or superheated), or two-phase mixture.

The flow is one-dimensional and each region of the reactor is characterized by an axially uniform flow area. Over each cross section of a region, it is assumed that the fluid is in thermodynamic equilibrium which implies that subcooled boiling effects are neglected.

Locally, the fluid is treated as incompressible so that all acoustic phenomena are neglected (i.e., acoustic waves are assumed to travel at infinite speed). The fluid is, however, globally compressible in the sense that fluid properties vary with the overall system pressure.

This description allows modeling of transient reactor behavior involving overall pressure variations which occur on a time scale which is considerably slower than that associated with the propagation of pressure waves. Potential and kinetic energy effects are neglected in the present analysis.

The model considers slip between phases, a non-uniform, time-varying axial power profile, local pressure losses at the inlet and outlet of the various regions, and wall and fuel-spacer frictional effects.

With these assumptions the mixture continuity and energy equations, respectively, for a flow region of the reactor are given by 1 a = T,,(2.1-1) and i(l 9 +[(1 -aý,uth, + a u, hgh 1 dP" + --++ q"h -++yrh,. (2.1-2) a e ~z Jdt A ]Lahey, R.T. and Moody, F.J., The Thermal-Hydraulics of a Boiling Water Nuclear Reactor, Second Edition, American Nuclear Society, La Grange Park, Illinois, 1993, Equations 5.57 & 5.96, pp. 208-219.

Feedwater I HPCI, RCJC Condensate I SLCS Figure 2-1 SABRE Model of Susquehanna reactor.PAGEI v2O-j PAGE 20 Turbine Steam Flow The mixture momentum equation is given by 2 OGj (la)pU,u +apgu 9P GJGj Fj.p f -s + go_ a +' + ý-+ K,+K D.5(z -o)+ K 2 ,0(z -L)] = 0. C7 ý_2peL Dh Dhý (2.1-3) The subscripts g and e designate gas and liquid phases. In the momentum equation (2.1-3), the sign of the elevation term pgL is negative for the jet pump region since the normal flow direction for the jet pump region is downward.

Also, the elevation change along the lower plenum differs from the flow path length. Therefore, for the lower plenum region, the value of [elevation at core inlet]- [elevation at jet pump outlet] is used in place of the flow path length L in the term pgL of Eq. (2.1-3). In addition, the fuel spacer frictional pressure drop term fs(D/Dh appears only in the momentum equation for the core region. The Dirac delta function 8(z -ý), where ý = 0 and ý = L, designates the location of the local pressure losses at the region inlet and outlet. Additional variables in the above conservation equations are defined below: G(z,t) = fluid mass flux hg = gas-phase enthalpy (Btu/Lbm), h, = liquid-phase enthalpy (Btu/Lbh,), h(z, t) volume-weighted enthalpy (Btu/Lbm), h = enthalpy associated with mass source y. (Btu/Lbm), P(z,t) = local fluid pressure (Lb, /ft2), P*(t) = system pressure = steam dome pressure (Lbf /ft2), t time (sec), u(z,t) = fluid velocity (ft/sec), z = axial coordinate (ft) (0 < z < L), a = void fraction, P 8 = gas-phase density (Lbm /ft3), pA = liquid-phase density (Lb,, /ft3), and 7. = volumetric source of mass per unit time (Lb. /ft 3 sec). The mixture density p(z, t), the volume-weighted enthalpy -h(z, t), and the fluid mass flux G(z, t) are defined through 2 Lahey, R.T..and Moody, F.J., The Thermal-Hydraulics of a Boiling Water Nuclear Reactor, Second Edition, American Nuclear Society, La Grange Park, Illinois, 1993, Equation 5.65 p. 212.

p=p,+ (-4)p,, h = [apghg +(I -a)ph, ]p, and (2.1-5) G = p, (I- a)u, + pga u,. (2.1-6) The system pressure P*(t) is a known function oft from additional mass and energy balance equations describing the steam dome and downcomer regions.

Closure of the thermal-hydraulic system is accomplished through use of the drift-flux relation: (aC0 -1)u +Co(I-a)u, =-V,. (2.1-7) where C, is the radial bubble concentration parameter, and Vg, is the drift velocity.

In SABRE, Co and Vg, are computed from the Ohkawa-Lahey void model which accurately predicts the counter-current flooding limit.3 The unknowns in this system of 7 equations consist of p, G, h, ug, up, a, andP. The vapor and liquid phase densities and specific enthalpies are determined from the equations of state as described below: Subcooled Liquid Region This thermodynamic region is defined by h(z,t) < h, (Y) where h, (P*) is the specific enthalpy of saturated liquid at the system pressure P*. In this thermodynamic region, the liquid phase density and liquid-phase specific enthalpy are given by p 1 (z,t)= p[P (t4'h(z,t)]

and h,(z,t)= h(z,t4 (2.1-8) Saturated Region This thermodynamic region is defined by h, (Pr' )<i(z, t)< h, (P*) where h, (P*) is the specific enthalpy of saturated vapor at the system pressure P*. In this region, the liquid-phase and vapor phase densities and enthalpies are given by he(z,t)= h[P,(t)l p,(zt)= pf[P*(t)1 hg(z,t)= h,[P*(t)1 andpg(z,t)=

p,[P*(t)1 (2.1-9) 3 Ohkawa, K. and Lahey, KT., "The Analysis of CCFL Using Drift-Flux Models", Nuclear Engineering and Design, 61, 245-255, 1980.(2.1-4)

Superheated Vapor Region The superheated vapor region is defined by h,, [P (t)] < h. Here the gas-phase specific enthalpy and density are determined from h9(zt)= t andp(z,t)

=(2.1-10)

Appropriate boundary conditions for Equations (2.1-1)-(2.1-3) are formulated by specifying the local pressure on the boundaries of the region and the volume-weighted enthalpy on any boundary where fluid (liquid or vapor) enters the region. Under natural circulation conditions, the pressure in the downcomer at the elevation of the jet pump throat is calculated as the steam dome pressure plus the static head of the fluid above the jet pump throat. Obviously, the natural circulation flow rate depends strongly on the height of the water column above the jet pump throat. Therefore, in order to accurately model the downcomer water level, a level-versus-volume table (see Appendix D, Section D. 1.9) is used to determine level from the calculated fluid volume. The downcomer level-volume table consists of 18 data points which incorporate the cross-sectional flow area variation of the region. Conservation equations which yield the downcomer fluid volume are described in Section 2.2. In Eq. (2.1-2) the heat flux term q"(zt) is zero for all regions of the reactor except the core where it accounts for heat transfer from the fuel rods to the coolant. (Heat transfer to and from the vessel structures is also included in this term; see discussion in Section 2.4.7) In the single-phase region of the core (liquid or superheated vapor), the Dittus-Boelter 4 correlation is used to calculate the coolant-film heat transfer coefficient.

In the bulk boiling region of the core, where the coolant is saturated, the film coefficient is computed from the Thom correlation.

5 18.66 .90,720) where H= film coefficient for boiling region (Btu/sec.ft2.°F), Tsurf = clad surface temperature (OF), p * = system pressure (lbp/ft 2), and Tcool = coolant temperature (0 F). In the subcooled-boiling region, the film coefficient is calculated as a linear combination of the liquid-phase and bulk-boiling coefficients.

The onset of nucleation within the subcooled region 4 Burmeister, L.C., Convective Heat Transfer, Wiley, New York, p. 486, 1983. 5 Thom, J.R.S., Walker, W.M., Fallon, T.A., and Reising, G.F.S., "Boiling in Subcooled Water During Flow in Tubes and Annuli", Proc. Inst. Mech. Eng., 3C180, 1966.

is determined from the nucleation model of Davis and Anderson 6 as described by Lahey and Moody.7 The methodology for identifying departure from nucleate boiling is taken from the RAMONA-3B code.8 RAMONA uses a combination of two critical heat flux correlations 9 , 1 0 to cover the expected range of flow conditions within the core channels.

In the transition boiling region, where the cladding surface temperature is greater than the value corresponding to DNB and less than the rewetting temperature, the film coefficient is calculated as a linear combination of the nucleate-boiling and film-boiling coefficients.

The clad rewetting temperature TRw associated with the break down of stable film boiling is estimated from the following empirical correlation, 1 1 TRw =0.13P +0.86 (2.1-12) Tcf Pc t where TJ,, and P1cr are the critical temperature and pressure of water, and P is the pressure.

The film boiling heat transfer coefficient is obtained from the Dittus-Boelter correlation using the mass flux and fluid properties associated with the steam flow. q". (zt) in the energy equation [see Eq. (2.1-2)] represents the local volumetric heat generation rate within the coolant. For the present model, this quantity is zero for all regions except for the core and bypass where it accounts for gamma heating of the coolant. The heat generation rate within each of these two regions is specified as part of the code input. The magnitude of gamma heating is assumed to be a constant fraction of the total core power. In the momentum equation, the two-phase friction multiplier (p is calculated from the Martinelli Nelson correlation which includes the flow correction developed by Jones 1 2.cp is correlated as ,p = F-(G,P*) [.2 (p,/ -1.0) x0"24]+ 1.0, (2.1-13) where 6 Davis, E.J. and Anderson, G.H., "The Incipience of Nucleate Boiling in Forced Convection," AIChE J., 12, 1966. 7 Lahey, R.T. and Moody, F.J., The Thermal-Hydraulics of a Boiling Water Nuclear Reactor, Second Edition, p. 253, American Nuclear Society, La Grange Park, Illinois, 1993. 8 Wulff, W., Cheng, H.S., Diamond, D.J., and Khatib-Rahbar, M., "A Description and Assessment of RAMONA 3B Mod. 0 Cycle 4: A Computer Code with Three-Dimensional Neutron Kinetics for BWR Systems Transients," NUREG/CR-3664, p. 100, January 1984. 9 Condie, K.G. and Bengston, S.G., "Development of the Mod 7 Correlation," EG&G Report, Attachment PN-181 78. 1 0 Smith, R.A. and Griffith, P., "A Simple Model for Estimating Time to CHF in a PWR LOCA," Natl. Heat Transfer Conference, ASME paper 76-HT-9.

1 1 Hsu, Y., Graham, R.W., Transport Processes in Boiling and Two-Phase Systems, p. 118, Hemisphere, Washington, 1976. 1 2 Lahey, R.T. and Moody, F.J., The Thermal-Hydraulics of a Boiling Water Nuclear Reactor, Second Edition, p. 270, American Nuclear Society, La Grange Park, Illinois, 1993.

AAUE 25 F(G,P*)=1.36+3.47xlO6P*+3.60xlO-4G-1.79xlO-P*G ifG< 194 (2.1-14) and F(G,P*)=1.26-2.78xlOC 6 P +3.31/G+5.40xlO-1P*/G ifG >194 (2.1-15) For local pressure losses such as those occurring across a fuel spacer and the inlet or outlet of a flow region, the homogeneous two-phase multiplier I is used in computing the pressure drop, where D = I + x(v, -V01 Vt (2.1-16) In (2.1-16) vg is the vapor-phase specific volume and vt is the liquid-phase specific volume. The flow quality x is defined by Wt + Wg where w. and w, are the mass flow rates of vapor and liquid respectively.

Under counter-current flow regimes (wg > 0 and w, < 0), x can be greater than 1 or less than 0. In the SABRE calculation, x is set to I if it becomes greater than 1, and x is set to 0 if it becomes less than 0. Pressure changes associated with flow area differences between adjoining flow regions are neglected in SABRE. This is not a serious code limitation because SABRE employs momentum equations which are integrated over the entire natural circulation loop. If large axial gradients in mass flow do not occur, then the reversible losses tend to cancel in the integrated momentum equations.

This simplifying approximation of neglecting reversible pressure losses is also used in the RAMONA-3B code. 13 13 Wulff, W., Cheng, H.S., Diamond, D.J., and Khatib-Rahbar, M., "A Description and Assessment of RAMONA-3B Mod. 0 Cycle 4: A Computer Code with Three-Dimensional Neutron Kinetics for BWR Systems Transients," NUREG/CR-3664, p. 84, January 1984.

2.2 Fluid

Dynamics Equations for Downcomer and Steam Dome Regions Both the downcomer and steam dome are variable-volume regions. The combined volume of these two regions is, however, fixed and is denoted as V 0.In the following discussion, the subscript "DC" refers to the downcomer region and the subscript "SD" refers to the steam dome region. In the downcomer, coolant can be added by means of injection flow and by means of the liquid exiting the steam separator region. Furthermore, when the downcomer water level drops below the feedwater spargers, coolant is injected directly into a region occupied by saturated steam. In this case the steam condensing on the cold make-up flow provides an additional source of mass to the downcomer region. Fluid exits the downcomer through the entrance to the jet pump region. In depressurization transients, where flashing occurs in the downcomer, vapor separation at the fluid surface also contributes to mass loss from this region. The mass and energy balances for the downcomer region can be written as dVSD( DCdhDC + (V Vv PDCl]dP* Pdt dr dt LT dt (2.2-1) Winj + W,,l+ W1 410-W1 (o1 0)- Wgjo. 0- W _,-W6,. and -d VSD& ýD dkhc L V ( g&PDC I}d dt dt L0 J SD W.. .1 + ,,,h -W,h.j+Wo~hs

+ Wts (Ls, t)hIs (Ls, t)_-W, (0, t)h ,(0, t)_-W, (0, t)h, (0, t) (2.2-2) where poc is the downcomer fluid density, VsD is the steam dome fluid volume, W,,,j is the injection flow rate, Wvs is the rate of vapor loss from the downcomer region due to buoyancy effectsWeald is the mass addition rate due to steam condensation, and Wbrk is the break flow rate which is used in the simulation of small break LOCAs. W. and Ws refer to the mass flow rates within the jet pump and separator regions respectively.

The relation VD, = V, -VSD, where VDc is the downcomer fluid volume and VSD is the steam dome fluid volume, has been used in obtaining (2.2-1) and (2.2-2). In (2.2-2) hDc is the downcomer volume-weighted enthalpy, hinj is the injection-flow enthalpy, hts and h, are the liquid-phase enthalpies in the separator and jet pump regions, hgwc and hg are the gas-phase enthalpies in the downcomer and jet pump regions. Potential and kinetic energy effects are neglected in (2.2-2). Also, it is assumed that the latent heat associated with condensation is completely absorbed by the injection flow. Similarly, the mass and energy balance equations, respectively, for the steam dome region are given by dV'SD DP .l 4D 1 SD S+I apD, dP dt ahSD dt .aP* jt (2.2-3) and -dV F a _ 1d ( a- idP* d- + L I d- OP p J)J &i (2.2-4) Wgs (Ls, t )hg (Ls, t ) + W,,, h g- (WI. + AD -W..hgm where p and hSD are the density and volume-weighted enthalpy of the fluid in the steam dome region. is the rate at which steam exits the vessel through the steam lines, and hg,. denotes the enthalpy of vapor in the steam dome region. Note that Wm consists of the steam flow turbine stop valve, the bypass valve, the safety relief valves, and the HPCI/RCIC steam admission valves. Derivations of (2.2-1)-(2.2-4) are given in Appendix A. The homogeneous, equilibrium critical flow model is used to compute W, under isolation conditions where steam is discharged through SRVs. For the unisolated case, the steam flow rate is calculated from a pressure regulator model (see Section 2.6). The steam condensation rate Wcod is defined in terms of a condensation efficiency qi. If ,I=O, no condensation occurs on the subcooled injection flow; if TI=1, the injection flow is heated to the saturation state by the condensing steam. Also, accounts for steam condensation on the water exiting the steam separators when this liquid is subcooled, and the downcomer level is below the separator outlet (this effect is only important under conditions where the reactor is shutdown and operated under forced-flow conditions).

Accounting for condensation on subcooled liquid exiting the separators, and using the above definition of il, the steam condensation rate W".d is given by Wr W(h,- -+WS(LSthf (2.2-5) where = efficiency of condensation on subcooled liquid exiting separators.

For purposes of computational simplicity, ri is taken to be a linear function of the water level in the downcomer region. tl=O if level is greater than or equal to the elevation of the feedwater spargers, and Tl=rmax (Tq., is typically 0.95; See Section D.4) if water level is less than or equal to a specified elevation (which is below the spargers) denoted rMU r_ Z I rAUE 28 by Lmax. The parameter Lmax is measured in inches above TAF and is supplied as part of the input data. A similar approach is used to calculate q.,* Under conditions where flashing occurs in the bulk downcomer region, the vapor separation rate Wvs is computed from W. = Vb. ADzXW~ Pg: (2.2-6) where Vbr = bubble rise velocity (ftlsec) ADC = downcomer flow area at interface between downcomer and steam dome regions (ft 2) oaDC = downcomer void fraction PgDC = gas phase density in downcomer (Ibm/ft 3) In Eq. (2.2-6), the bubble rise velocity is computed from Equations (2.1-6) and (2.1-7) using the Ohkawa-Lahey void model. At the surface of the vapor-liquid mixture in the downcomer, ut = 0 and G = pgoc ug. Using these relations, and the fact that w = pg(X ugA, Equations (2.1-6) and (2.1-7) lead to the following expression for the bubble-rise velocity:

Vb V = j-,DC (2.2-7) 2.3 Fuel and Cladding Heat Balance Equations The heat transfer model presented here describes the transient temperature behavior within an average power fuel rod within the core. Axial conduction within the fuel rod is neglected since the dominant temperature gradient occurs in the radial direction.

A lumped-parameter model is used to describe the radial flow of heat from the fuel to the coolant. The overall resistance to heat transfer from fuel to coolant consists of the thermal resistance of the fuel, the fuel-cladding gap, the cladding, and the coolant convective film at the outer surface of the cladding.

The radially-lumped parameter model employed here is a generalization of the model formulated by Tong. 1 With two equal-volume radial fuel nodes, the governing fuel heat transfer equations are given by PfýaT fa 2t (Ttnr-+r ' * ;, and (2.3-1) P at -" tInr + rba-fa Pf~C', ) T -z(~ -T) at (rb + 2rI + r. (2.3-2) where pf = fuel density (lbmnft 3), Cpf = fuel specific heat (Btu/lbm.0 F), ra = radius of inner fuel node (ft), rb = fuel pellet radius (ft), Tf(zc,t) = fuel temperature (0 F), kf = fuel thermal conductivity (Btu/sec.ft.°F), iy (zct) = volumetric heat generation rate within inner radial fuel node (Btu/sec.ft 3), ;y; (z,,,t)= volumetric heat generation rate within outer fuel node (Btu/sec.ft 3), 'Tong, .L.S., "Simplified Calculation of Thermal Transient of a Uranium Dioxide Fuel Rod," Nuclear Science and Engineering, Vol 11, pp. 340-343, 1961.

PAGE 30 rci, rco = Tc(zc,t) =Hg kcl cladding inside and outside radii respectively (ft), cladding temperature (OF),-gap conductance (Btu/ft2.sec.°F), and = cladding thermal conductivity (Btu/ft.sec.°F).

In Equations (2.3-1) and (2.3-2), the subscript "a" refers to the inner fuel node, and the subscript "b" refers to the outer fuel node. In the outer fuel node, the volumetric heat generation rate is specified as 10% greater than the radially-averaged volumetric heat generation rate to account for self-shielding effects within the fuel (Cs = 0.9 in §3.1.4).

Note that the volumetric heat generation rates "f.(zc,t) and i;(z 0 ,t) include the effects of the axial power shape. Similarly, the heat balance equation for the bulk temperature of the cladding can be written as P (r2 r-(__ _r,) 1 (t -_- r 2 2r., H9 4r~,,ka.( S3- T-,) (2.3-3)where PcI TPzt Tcool(Zo0 Td'(.A,0 Himm= cladding density (lbm/ft 3) = cladding specific heat (Btu/lbm.°F)

= coolant temperature within core region (OF) = clad surface temperature (OF), and = coolant film heat transfer coefficient (Btu/ft2-sec-.F).

The governing equation for the surface temperature of the cladding To, is obtained from (2.3-3) with an appropriate adjustment of the resistance terms so that the rate of energy flow into and out of the cladding is referenced to the outer surface of the cladding.

The resulting expression is P , (r2 -",' _ -_r_)_ [- 2 1 ] rbkfl, 2 r.Hg .k, 2r.Hi 8 (2.3-4)Derivations of Equations.

(2.3-1)--(2.3-4) are given in Appendix B.

PAGE 31 2.4. Neutron Kinetics A one-dimensional neutron kinetics model based on the two-group, one-dimensional neutron diffusion equations is used in SABRE to compute core power. The kinetics equations are solved using a finite-difference procedure.

The model includes lower and upper reflector regions in addition to the active core region. Within the active core region, there are twenty-five material regions having spatially-independent neutronics parameters.

The kinetics model can simulate the effects of boron addition by reactor scram, manual control rod insertion (MRI), and Standby Liquid Control System injection.

2.4.1 Neutron

Kinetics Parameters Neutron kinetics parameters are obtained from output of the SIMTRAN-E 1 computer code. S1MTRAN-E collapses the 3-D cross-sections calculated by the SIMULATE-E 2 code to a 1-D (axial) cross section set. In the active core region, 1-D parameters are constant over 6" axial segments referred to as material regions. The core is designed such that the radially-averaged fuel composition does not vary axially within these 6" segments.

The neutronics parameters obtained from the SIMTRAN-E code are in the form of polynomial curve fits. The polynomials give the cross sections, diffusion coefficients, and delayed neutron fraction as functions of perturbations in fuel temperature and moderator density. Curve fit equations are provided for the following kinetics parameters:

D, = diffusion coefficient for fast neutrons (cm), = macroscopic absorption cross section for fast neutrons (crn 1), = macroscopic down-scattering cross section for fast neutrons (cm'1), v1 ED q = [average number of neutrons produced in a fission induced by a neutron in fast group] x [macroscopic fission cross section for fast group] (cnf 1), K E P = [energy per fission conversion factor] x [macroscopic fission cross section for fast group] (cm'1), D2 = diffusion coefficient for thermal group (cm), 'McClure, L.A, and Gose, G.C., "SLMTRAN-E:

A SIMULATE-E to RETRAN-02 Datalink," NP-5509 CCM, Electric Power Research Institute, Palo Alto, California (1987). 2 Ver Planck, D.M., Cobb, W.Rt, Borland, RKS., and Versteegen, P.L., "SIMULATE-E:

A Nodal Core Analysis Program for Light Water Reactors," NP-2792-CCM, Electric Power Research Institute, Palo Alto, California (1983).

PAGE 32 Ea2 = macroscopic absorption cross section for thermal group (cm'1), v 2 E12 =[average number of neutrons produced in a fission induced by a thermal neutron] x [macroscopic fission cross section for thermal group] (cmql), 1C Zf2 = [energy per fission conversion factor] x [macroscopic fission cross section for thermal group] (cm'), 01, = average speed of fast neutrons (cm/see), 02 =average speed of thermal neutrons (cm/sec), and P= delayed neutron fraction.

In the Susquehanna SIMTRAN model, the reactor core is divided axially into 27 material regions. In each of these material regions, the 12 parameters listed above are :functions of time only. The material regions consist of the lower and upper reflectors and twenty-five 6" axial segments within the active core (region containing fuel). For each of the 27 material regions, the kinetics parameters are approximated by polynomials in fuel temperature and moderator density fluctuations.

For example, the polynomial representation of D in neutronic regionj, where] j {l,...,27}, is D 1', (t) = + c, A+ j -- + c, 1 2 AU, + Ci 3 AU 1 iA A , +c 1 4 A14ujuJ+c, 5 AT/ j 7AuAuj (2.4.1-1) where = 7>(t)- T 1 (O)(2.4.1-2) and Auj = & (t) -pj (0) (2.4.1-3) p 1 (0) Here Tf~j (K) is the average fuel temperature in material regionj, and pj is the average moderator density in material regionj (Lbrn/ft 3). The constants cj, in (2.4.1-1) are computed by the SIMTRAN-E code and given in the SIMTRAN output file. The other eleven neutronics parameters are represented by expressions identical to (2.4.1-1).

Within the present calculation, four cycle-specific (EOC) cross-section sets are developed:

U2C7, U2C9, U2C1O, and U1C12. These are discussed in detail in §2.4.8.

PAGE 33 2.4.2 Governing Equations With the assumption that all neutrons are born in the fast group, the two-group, time dependent, one-dimensional neutron diffusion equations are 3 Fast Group (Group 1) 1 G-0~t I a[ zt)O3/4(Z' 0tjD ý (Z oB02(Z)g I(Z't)..xZ)(t u 1)(z't) at az [I('t ,('1 R('t P('t + (1-ffJYfl(z,t) p 1 (z,t)+(l-j)v 2 yf 2 (Zt) q, 2 (z,t) 6 (Z X(zt) (p, (Z't) + Z X,. C. (Z't0 (2.4.2-1) m=I and Thermal Group (Group 2) 12 ( 2 (zt 0 a4 &P 2 (Z, ) 14 u 2 (z't) 't aZ[D 2 (Z, t) & j -, 2 (Z,t) p 2 (zIt) +ZJZI(zt9 1 zIt)(24-)

where (p = group-1 (fast) neutron flux (cm-2 sece), 92 = group-2 (thermal) neutron flux (crn 2 sec%'), and B2 (z) = radial buckling (cm-2). The radial buckling B.2 (z) is time-independent and is listed in the SIMTRAN output file. The neutron flux is assumed to go to zero at the inlet of the lower reflector and the outlet of the upper reflector.

Therefore, boundary conditions for the fast and thermal flux consist of 9p, (0 ,t) = cpl (L,t) = 0 (2.4.2-3) and (P 2 (0,t) = 92 (L,t)= 0. (2.4.2-4)

Note that z-- corresponds to bottom of lower reflector region, z=L corresponds to top of upper reflector region. 3 Stacey, W.M., Space-Time Nuclear Reactor Kinetics, Academic Press, New York, 1969.

Conservation equations for delayed neutron precursors consist of aCm(z,0t)

-1 0[V (z,t) (p,(Z,t) +V 2 f 2 (z,t), 2 (zt)]- X, C(z,) where 6 P3 = O M = delay fraction, m=1 m {1,...,6}

.(2.4.2-5)k= decay constant for delayed neutron precursor group m (sec 1), and Cm = concentration of group m delayed neutron precursors, Initial conditions for the fast and thermal flux, and for the delayed precursor concentrations are obtained by the methodology discussed in §2.4.3. The reactor thermal power is computed from 4 L Q(t) =fdz [I c Yfl(z,t) (p(z,t) +rK Xf 2 (z,t) P 2 (z,t) ] 0 where K = energy per fission conversion factor.(2.4.2-6)Since the initial flux is computed as the solution of an eigenvalue problem, the amplitude of the initial flux is arbitrary (only the flux shape is relevant), and therefore, a scaling factor is included in the SABRE code so that the power computed from the integral in (2.4.2-6) is equal to the initial specified core power. 2.4.3 Methodology for Initial Flux Distribution The initial flux distribution, initial precursor concentrations, and the initial neutron multiplication factor keff are obtained by solving the eigenvalue problem associated with the steady-state, l-D, two-group diffusion equations.

The eigenvalue problem is expressed as 4 J.A. McClure, and G.C. Gose, SIMTRAN-E:

A SIMULATE-E to RETRAN-02 Datalink, NP-5509-CCM, p. 11-25, Electric Power Research Institute, Palo Alto, CA, 1987.

d [D (z) dT(z__)] +DA(z) B2(z)(p,(z)+/-+

1.(z) 9 1 (z) dz dzR + -., (z)(z)= [ v 1 fl(z)1 (z)+ v2Ff.2 (z) 9 2 (Z)]lkeff, (2.4.3-1) d D:!(Z) d92(Z)- + E "2(Z) T 2 (Z) =- (Z()(PI(Z), (2.4.3-2) dz az C,,(z) = F [, (z) p 1 (z)+v 2 If 2 (z) P 2 (z)], (2.4.3-3)

X m and PI (0) = (PI (L) = 92 (0) = 2(L) = 0. (2.4.3-4)

The terms on the left hand side of (2.4.3-1) represent neutron loss due to axial and radial leakage, absorption and scattering.

The two terms in the numerator on the right hand side of (2.4.3-1) represent neutron production due to fast and thermal fission. Thus the eigenvalue keff in (2.4.3-1) is the ratio of neutron production to neutron destruction which is the definition of the effective neutron multiplication factor.5 The factor 1/kff has been introduced into the steady-state neutron diffusion equations in order to facilitate calculation of a steady-state flux distribution.

Eqs. (2.4.3-1), (2.4.3-2), and (2.4.3-4) are homogeneous in the fast and thermal flux, and therefore, a non-trivial solution to these equations will exist only for a specific value of the neutron multiplication factor kf. In the SABRE code, the power method 6 is used to calculate the initial flux shape and the corresponding neutron multiplication factor. These results provide the necessary initial conditions for the transient kinetics problem.

Power Method for Solution of Eigenvalue Problem The kinetics parameters in (2.4.3-1) and (2.4.3-2) are defined on 27 material regions within the core. These material regions consist of the lower and upper reflectors and the 25 axial cells within the active core region. The equations (2.4.3-1) and (2.4.3-2) are solved, using a finite-difference procedure, on each of these 27 material regions. The solutions are coupled by requiring continuity of the fast and thermal neutron flux and the neutron current density at the interface of any two material regions. These interfacial coupling conditions consist of PIA() = 91i-1(41) j r #2,3,...,N}, (2.4.3-5) 5 j. j. Duderstadt and L.J. Hamilton, Nuclear Reactor Analysis, p. 216, Wiley, New York, 1976. 6 j. J. Duderstadt and L.J. Hamilton, Nuclear Reactor Analysis, pp. 216-218, Wiley, New York, 1976.

(P2j(O)dz. JO z _D1 ~j dlj(z J) d(12j~ (Z -01) ' z -"zj=o = -' j _Zd J 1 zj -I=Lj-I and (2.4.3-6)(2.4.3-7)(2.4.3-8)where the subscriptj denotes the material region within the core (j c {1,2,...,N}

where N = number of material regions = 27), and Lj is the length of each material region such that zj _ E(O, Ls). Within a material region, the diffusion coefficients and cross sections in (2.4.3-1) and (2.4.3-2) are constant.

Therefore, within material regionj, (2.4.3-1) and (2.4.3-2) become _-D -d 2 PIj(zj) 2 2j , Rj (d2R Zj)-aj 1 j(Zj) +1 ,Jpij(zj)

Z.1 and-vlj F, q 1 , 1 P'(zj) + 2 ,.j f,2,j (P 2 ,j(Zj)]Ikeff, dp 2 ( "Z) +aZj (P2,j(zj) dzj (2.4.3-9)

(2.4.3-10)

The second-order, central-difference formula, d 2 (pk ___________

2 Dk 2 = .kj 2T' +0 D -j2 (Az ) dj Aj i G {23,...,Mj

-1)is used to approximate the diffusion terms in (2.4.3-9) and (2.4.3-10).

In (2.4.3-11), the subscript k defines the flux group (1--fast group and 2--thermal group),j denotes the material region, and i specifies the mesh point within the material region. Each material region can have a different number of mesh points. That is, Mj can be different for each of the 27 material regions and is really a function ofj. At the interior mesh points of a material region, the finite-difference approximation to Eqs. (2.4.3-9) and (2.4.3-10) are (2.4.3-11)

[I BI- AzX + fVlj k~ff i+ Di+1 DW A 2 ' i +D 1 ,j ]=-Azjv 2 ,j (P2j and D i 1 p'+[- 2D 2 1-Az2 1:'a,,biPj

+D 2 gp2 i i+1 AZ2 yi + 22,,Az ja, z, "I-2,(2, j -i. Z j(P Iij i 4 2,...,4.j

-1) . (2.4.3-13)

Equations (2.4.3-12) and (2.4.3-13) provide relations for determining the fast and thermal neutron flux at the interior grid points of each material region. The flux on the boundary points of each material region is determined from boundary conditions (2.4.3-4) and matching conditions (2.4.3-5) through-(2.-4.3-8).

A 3-point sloping difference formula is used to obtain the following finite-difference approximations to (2.4.3-7) and (2.4.3-8):

and Di, j [_3(l .+,492j , *-I _ m ,2 2Azj _.P 3-" j] -j_" " -"( PJ 2-1 + 4P,. iJ-, ""]D j-1 1- j- Ij 2A, z z 22Azj1 2J *- 2, j-1 2, j-1 (2.4.3-14)

(2.4.3-15) wherej e {2,3,...,N}.

In terms of the values of the flux at the boundary grid points, Eqs. (2.4.3-5) and (2.4.3-6) become (2.4.3-16)

(2.4.3-17)

Substituting (2.4.3-16) into (2.4.3-14) and (2.4.3-17) into (2.4.3-15) leads to the following expressions for the fast and thermal flux at the interface of the material regions in terms of the values at interior mesh points: 4 (p 1 , i -13 ,+- ,_ MJ-p 2 -I , 9 J-2 3(1 + , -jiE {2,3,,..,N}

(2.4.3-18)

(2.4.3-12) i E- {2,3,..,M.

-1)(p1 -mMjI -- zI -1_Z'I'j IýDi i , j PAGE 38 and (9 2 , f (P 2 ,j 1 3(1 + 92, j)(2.4.3-19) plAj- C DI, j~and [LZ Az. .D- ] i (2.4.3-20)

The additional relations, 2 3 4 M -1 _ j-2 Mi = 441, j+] (pl.j+l--IIJ,j+ (pj+ -4p 1 -( Ij .3-=(2.4.3-21) and (p2 _3 M -Mj-2 4 pt z .j 2 ,s + 2-, 2+, j+ 24, = P j q)2,j (2.4.3-22) result from incrementingj by I in equations (2.4.3-18) and (2.4.3-19).

It is convenient to rewrite Eqs. (2.4.3-12) and (2.4.3-13) as C-1' 4 +a 1 1 3 (fQ, +b 1 2 3 p],.1 C I, j (P ],j + , p i j + b ' ( l, i, d ' j i iI i i i+1 =~ c2, j (P 2. +a2,i , p2,j +2i (p2,j=d, and where a _ =_2D 1 AZ2D 1 ," 1 BR A 2 , 2_ Vjj ljA 2 ,j k'i F"I sLj, where (i=-2,3, ...,rM-1) (=-2, 3, ...,Mj-1)(2.4.3-23)

(2.4.3-24) j c= 2,3,..., NJ j i t1, 2,..., N, -1)E- 0, 2,...2 N -1 PAGE 39 di = 2 V, .F"f2,j i ' keff c2, j 2,j b2', j D2,,, and dj= -Azs" F, 9 1 The power iteration method for computing the neutron flux shape and the associated eigenvalue keff proceeds as follows: Stepmi An initial guess, consistent with the boundary conditions, = 0, q. -0, IMW =0, and MN = 0, is made for the fast and thermal flux(p1 -Also, an initial guess is made ¢2,N ' for the eigenvalue kg Step 2 The neutron generation rate is integrated numerically, using Simpson's Rule, over the length of the core region to obtain the source integral S, S =jd [v, ,,,(zq),(z

+v 2 f 2 (Z) 9p 2 (4) 0 Step 3 The Thomas algorithm 7 is used to solve for the fast flux at the interior mesh points of the N material regions of the core. For each material region (j E {1,...,N ), the solution proceeds as follows: Define 5 () =1 , j and F(1)= 0 7 Anderson, D.A, Tannehill, J.C., and Pletcher, R.H., Computational Fluid Mechanics and Heat Transfer, p. 99, Hemisphere, New York, 1984.

b .i+1 Compute F(i+1)= -for ic-, 2 ,...,M,-2}

ai +ci+l F(i) Compute 5 (i + 1) = -i -6(i) for.., -2} a;,+ +cj F(i) 4,+j ,1 Compute ýpq =F(i)pl+4

+6(i) for ie{Vf-A,M 1-2,...,2}

T c n s are computed from the fast flux calculated in Step 3. An improved estimate for the thermal flux is obtained by solving the tri-diagonal system (2.4.3.-24) using the Thomas Algorithm.

st_e_ The source integral S in Step 2 is recomputed with the improved estimates of the fast and thermal flux. An improved estimate for kffis calculated from the iteration formula 8 k(,+') _S(n+l) off 1 S(n) where n denotes the iteration index. Step 6 A check is made for convergence of the multiplication factor keff If Ik,(;+') k- k I < , the calculation process is stopped. Otherwise, the calculation proceeds to Step 7. Step 7 S(n) takes the value of rS(+) and k,(;) takes the value of k.,(7 .The calculation then transfers back to Step 3 and proceeds until the acceptance criterion in Step 6 is satisfied.

After the initial neutron flux is computed as described above, a second level of flux initialization is performed to remove any initial perturbations from the numerical solution.

This is accomplished by performing internal time steps within the SABRE code prior to initiating the transient simulation.

The transient solution is started at t = -40 8 J. i. Duderstadt and L.J. Hamilton, Nuclear Reactor AnaWysis, p. 217, Wiley, New York, 1976.

seconds and all of the governing kinetic and thermal-hydraulic equations are numerically integrated up to t---0. During the time that t<0, small adjustments are made to the value of keff (typically a few tenths of a mk) and the pressure regulator setpoint so that when t=0 the power and reactor pressure are acceptably close to the values specified in the SABRE input file. When the transient simulation is initiated at t=-0, all initial perturbations have been removed through the dynamic initialization process.

Table 2.4.3-1 and Figure 2.4.3-1 compare the initial axial power shape calculated by SABRE for Case 03 (see Computer Case summary) against the SLMTRAN-calculated power shape. SABRE Case 03 was calculated using the U2C9 kinetics file (u2c9.simtran.out) which is the output file for SIMTRAN Run#9900340.

The normalized power shape calculated by SIMTRAN was obtained from this output file. The SABRE-calculated power shape was obtained from the output file for SABRE Case 03. The mesh spacing used in the SABRE finite-difference solution of the two-group diffusion equations is essentially the same as that used in the PP&L SLMTRAN model. In SABRE, 33 grid points are used in each of the two reflector regions and 13 grid points are used in each of the active core material regions. In the SIMTRAN model, 12 grid points are used in each- of the 25 activ-ec-re aat eial regions. Thirteen points were used m SABRE because the number of grid points must be odd since spatial integration of the flux is performed using Simpson's rule. The results in Figure 2.4.3-1 show reasonably close agreement.

Sensitivity studies were performed on the grid point spacing in SABRE, and little change in the solution was observed upon increasing the number of grid points. The RMS difference between SABRE and SIMTRAN is computed from the values in Table 2.4.3-1 as follows: RMS = (100%)x 1,/ (SABRE, -S!MTRAN)2 = 4.86% Comparison of PP&L Nuclear Fuels' predictions for core axial power shape against TIP (Traversing In-Core Probe) measurements show that the RMS uncertainty associated with the predictions ranges from 5% to 12%.9 Therefore, the differences between SABRE and SIMTRAN are well within the uncertainty of the core-physics methodology.

9 PP&L Report PL-NF-90-001A, "Application of Reactor Analysis Methods for BWR Design and Analysis," pp. 85-90, July, 1992.rý%3r -+ I PAGE 42 Table 2.4.3-1 Normalized Axial Power Predicted by SABRE and SIMTRAN 25 0.1509 0.163 0.000146 24 0.4473 0.482 0.001204 23 0.7748 0.834 0.003505 22 0.9415 1.01 0.004692 21 1.0693 1.14 0.004998 20 1.163 1.23 0.004489 19 1.2283 1.29 0.003807 18 1.2741 1.33 0.003125 17 1.2987 1.35 0.002632 16 1.3276 1.36 0.00105 15 1.3141 1.34 0.000671 14 1.288 1.30 0.000144 13 1.2512 1.25 1.44E-06 12 1.2055 -14.19 -0.00024 11 1.1535 1.13 0.000552 10 1.1007 1.07 0.000942 9 1.05 1.01 0.0016 8 1.005 0.955 0.0025 7 0.9708 0.914 0.003226 6 0.9506 0.892 0.003434 5 0.946 0.885 0.003721 4 0.9444 0.883 0.00377 3 0.9231 0.863 0.003612 2 0.8195 0.761 0.003422 1 0.4022 0.362 0.001616 1.6 -...-. 1.4 4 S1.2 0 ,.;, i9'~~ 0.8 -0.6 0.82L 0 0g a...,9 __ Fiue24 4.C mprsno nta core-P-.~9 J- -axa power Shape' fo A RECs 3 gis SJMTRAN-calculated power shape.

r-Lt- 44 2.4.4 Methodology for Transient Flux Since the kinetics parameters are constant over each of the N material regions comprising the core, the time-dependent, two-group diffusion equations are written for each material region denoted by the subscriptj.

The regional fast and thermal group equations, respectively, are .1 ( j(z)t) =D 1 1 ,(t) j ,'(z,'2t)

D 1 j(t) BR 'j 3 (z 3 ,t) Ol~~t GI' f1'j2 t9J( 0+ 0 j(Zj't) + (1-J keff 1 (t)cp, (z 1 ,t) +(1- 1 3)VZJ(t) j2jt ,ff k keff (t)' %(zj, t) + X,, Cm.j (zj, t) (2.4.4-1) m=1 and 1 & ji(zit)_ 22 , 1 (2 ,t) -227, t) 922j(zj,t)+ (t i(, (2.4.4-2)

Similarly, the delayed-neutron precursor equations for each neutronic region consist of acm~j(Zi,)

-f ([ t) I f 1 Q~)(p 1 1 j zj, t) +vz2ijt) Eff~iQ)P 2.i(Zi) P 0 m C.'j (Zj,t) . m e{1,2,...,6)

(2.4.4-3)

The indexj defining the material regions ranges from 1 to N where N=27 for the Susquehanna core model. Region I is the lower reflector, region 27 is the upper reflector, and regions 2-26 constitute the active core. The length of each region is denoted as Lj. The subscript m ranges from 1 to 6, since there are six delayed neutron groups. Inclusion of the factor l/k,, in (2.4.4-1) and (2.4.4-3) ensures that the reactor simulation starts from a critical condition.

Fast and thermal flux boundary conditions consist of p(, (Ot)=O and 9p,,v(LN,t)=O (2.4.4-4)and p 2 , 1 (O,t)=O and (P 2 ,1(LN,t)=O.

At each interface between neutronic regions the flux and the normal component of the neutron current density are continuous.

Therefore, the interfacial conditions consist of ij(O,t) = qi.1 (L 1 , 0 (Pj(,t) = P2,j1(Lj-1, t).(2.4.4-6)

(2.4.4-7)(2.4.4-8)and ,2,j1_1*zJ(=

-L)2 j 0 a i =0 = -2j .&j-1 z-j.- =L j -j {2,3,...,N}

.As shown for the steady-state problem in §2.4.3, these relations lead to the matching conditions (2.4.3-18), (2.4.3-19), (2.4.3-21), and (2.4.3-22) which express the interfacial flux in terms of the flux at interior grid points. Initial conditions for the neutron flux and precursor concentrations are obtained as described in Section 2.4.3. A method of lines (MOL) approach is used to solve the transient equations (2.4.4-1)(2.4.4-3).

In the MOL, finite-difference approximations are applied only to the spatial derivatives in equations (2.4.4-1)-(2.4.4-3) thus converting PDEs (2.4.4-1)-(2.4.4-3) into a coupled set of ODEs with time as the independent variable.

At the interior grid points of each material region, the spatial derivatives are approximated with a second-order, central-difference formula. Application of the MOL yields the following set of ODEs for the neutron flux and precursor concentrations at the interior grid points of each material region: 1 0 Lamarsh, J.Rt, Introduction to Nuclear Reactor Theory, pp. 13 5-136, Addison-Wesley, Reading, MA, 1966.(2.4.4-9)(2.4.4-5)-Dij a 'j~z;'t) -ljlGP'- z-,t ' :zj 1j-= .0 = & j-1 PAGE 46 I dD -R2p, +(t)+D 1 ,(t) D -(I) -al j(t) +Q0 k) 2 2 1 it) AjpiB) (I-.)~j~ ý j(t)-.,tttj) ke+ ) fIJ(tfJJ t)+(1-3)V2.(t)2j 6 -j (t) (P],i(t)+ExCm(t)

(2.4.4-10) m=-I and d1 ' t)) 2"(t)+/-qP2'(t) u 1 (t) ,t =i DZ.y~t Az~~t) ' t +)P Q)+/-'-t)q',( D2, (t) dt (2.4.4-11)

Similarly, the delayed-neutron precursor equations for each neutronic region consist of dCý j(t) dt [I_ I _3. Ej~(Q~ P,~)v"~) "(t) ' C~ylt _flj~)p ,J-t I2J (t I t 92 dt k~ff 21 C 1. J (t mrE={1,2,..., 6} (2.4.4-12)

Herej denotes the particular material region, and i specifies the grid point within the material region. In the transient solution, the fast and thermal flux at the interface between two adjacent material regions is computed from (2.4.3-18), (2.4.3-19), (2.4.3 21), and (2.4.3-22).

The flux on the lower boundary of the lower reflector and the upper boundary of the upper reflector is specified in accordance with boundary conditions (2.4.4-4) and (2.4.4-5).

The ODEs approximating the transient reaction-diffusion system are integrated with the ODE solver LSODES (Livermore Solver for Ordinary Differential Equations with General Sparse Jacobian Matrices)." A relative error criterion of 1.E-05 is specified in the SABRE code. LSODES automatically selects time-step size and method order as the computation proceeds in order to optimize the computational efficiency.

For temporal integration of the ODEs, an implicit solution option based on the backward differentiation formulas proposed by Gear was chosen (LSODES input parameter mf=-222).

Thirty-three grid points were used in the upper and lower reflector regions, and 13 grid points were used in each of the 25 axial nodes within the active fuel region. This mesh spacing is consistent with that used to solve the I-D kinetics equations in the PP&L RETRAN model of Susquehanna. " A.C. Hindmarsh, ODEPACK, A Systematized Collection of ODE Solvers. In Scientific Computing (RS. Stepleman et al eds.), VoL 1, pp. 55-64. LMACS Trans. on Scientific Computation, North-Holland Publ., Amsterdam, 1983. 12 Gear C.W., Numerical Initial Value Problem in Ordinary Differential Equations, Prentice-Hall, Englewood Cliffs, NJ (1971).

PAGE 47 2.4.5 Modification of Thermal Absorption Cross Section to Account for Dissolved Boron The 1-D thermal absorption cross section 7 a2 obtained from the output file of the SIMTRAN code depends only on changes in moderator density and fuel temperature; there is no boron dependency built into this cross section. Since SABRE is used primarily to simulate reactor response during ATWS events in which the reactor is shutdown by injecting borated water via the SLCS, it is necessary to incorporate boron dependency into the SIMTRAN 1-D cross-section set. Rather than modifying the underlying lattice physics calculations to build boron dependency into the cross-section set, a much simpler approach is taken in this calculation.

This approach is described as follows.

For each fuel cycle, the PP&L Nuclear Fuels' Group supplies a Hot Shutdown Boron Concentration (HSBC) for use in cycle-specific ATWS analyses.

The HSBC is the boron concentration measured in ppm required-to bring the reactor from-hot full -power to-hot . zero power (no voids) with the Xenon concentration at the hot full-power condition.

The cycle-specific value of the HSBC obtained from the Nuclear Fuels' Group is used to empirically modify the boron-free thermal absorption cross section generated by SIMTRAN. In SABRE, a modified cross section of the form Za2,j Zaj "+-P CBBi; ,Cqj; (XBj) (2.4.5-1) is used where 3F j C Bj; O C Bj ) CB~C~(I Ack +AB + cB ,B.(1 A c +AB) A (2.4.5-2)

In (2.4.5-1), aZj represents the boron-free, 1-D thermal absorption cross-section computed by SIMTRAN for material regionj, and Z'2. , is the modified thermal absorption cross-section which includes dependency on dissolved boron. CBcj is the concentration of boron (ppm) in the coolant within the fuel bundle, CBBj is the concentration of boron in the bypass coolant, acj is the void fraction within fuel channel, otBj is the bypass void fraction, Ac is the fuel channel flow area, AB is the bypass channel flow area, and the subscriptj denotes a particular axial material region j r (1, 2,...,27}.

The parameter g is determined from the requirement that the core is critical at hot zero power conditions with the boron concentration equal to HSBC, i.e., when CB,cj=CB,Bj=HSBC, acj= cxBj-=O, and the fuel and moderator temperatures are equal to the saturation temperature corresponding to rated reactor pressure.

The parameter g is computed in an iterative fashion by SABRE during the initialization phase of the calculation.

This is accomplished by varying pi in accordance with a bisection procedure and repeatedly solving the eigenvalue problem defined by (2.4.3-1), (2.4.3-2), and .(2.4.3-3) at hot zero power conditions with the boron concentration equal to the HSBC until the eigenvalue k1ff converges to the hot full power eigenvalue (target keff) with no boron. The value of p computed in this manner is used in the transient calculation to compute the thermal absorption cross section from (2.4.5-1) and (2.4.5-2).

This approach forces the reactor core to shutdown in a manner consistent with the value of the HSBC. The functional form of (2.4.5-1) is verified by performing lattice physics calculations using the CASMO-3 computer code 1 3 with various void fractions and boron concentrations (CASMO Run#9900755 in Computer Case Summary).

The input data file and mem file (job control file) for this CASMO case are documented in Tables 2.4.5-1 and 2.4.5-2. Lattice calculation results for the macroscopic boron thermal absorption cross section for ATRIUM-10 fuel with lattice type 9214 are presented in Table 2.4.5-3.

Also contained within this table is the value ofF' calculated from (2.4.5-2).

In computing I', the values of the core and bypass flow areas Ac and AB are taken from the base ATRIUM-10 input deck for SABRE which is listed in Appendix G (Ac=78.19 ft 2 and AB=66.04 ft 2). In Figure 2.4.5-1, the B-I1 thermal absorption cross sections computed with CASMO-3 are plotted against the function :IF for bundle exposures of 10 and 20 GWD/MTU. A straight line connecting the largest and smallest CASMO-calculated cross sections is included in each plot in order to assess the linearity of the relationship between the B-10 thermal cross section and the function MF. It can be seen from Figure 2.4.5-1 that this relationship is indeed very nearly linear, and therefore, Eqs. (2.4.5-1) and (2.4.5-2) represent a valid model of the boron dependency in the 1-D thermal absorption cross section.

13 Malte Edenius and Bengt H. Forssen, "CASMO-3 A Fuel Assembly Bumup Program," User's Manual, STUDSVIK/NFA-89/3, Studsvik of America, Inc., 1087 Beacon St., Suit 301, Newton MA. 14 See Nuclear Fuels' Calculation NFE-2-10.-01, Rev. 0, p. C.6 for description of assembly and lattice type.

Table 2.4.5-1 CASMO Input File for Run#9900755 DIN, 10/ TITWI=40 TRJ-=840 0 N35622,IDE='9240'

• Lt92 -1Ib0x1--e19 13G66 -.40 VB RE, 1, 9.91010/2.556 RAE, 2, 9.91010/3.050 RLE, 3, 9.91010/3-400 RJE, 4, 9.91010/3.800 RLE, 5, 9.91010/4.150 RE, 6, 9.70209/4.150,7301=

6.00 RE: 7, 9.91010/4.450 RE, 8, 9.91010/4.680 BwR,10,1.29540,13.40612

.20320, .71374, .71374,1.1684 P ~ -7M71/;1!&

9~3 92238, %~239, 9440, 9 4 241,942 THE,0 R14,0 2 FIN, 1, .44196, .50254, LPI 11112 1112 11122 511111 5688 0 38788876 176867687 1357777321 SPA,6.970 6'6.900/302=8.46,718--17.54 irxre spacer ca, 576:. PDE,9.513 "DP0 .51 51152,2.53 3.54 45,5 55,6,6.57 75 88.599510 18.t,lý,19.A,26, -40:42'5,45,47.5,50,52.5,55,57.5,60.0 STA TIT,IDE='924A'

• +HIot 000.0 PP.: 0 Voids Brarch CAN,,# 293.15 OR, 600.0 STA TIT,JDE='92481

-+Hlot 000.0 PPM 40 Voids Branch RES, '9240- 0,5,10,15,20,25,30,40,50,60

%01,40 TMI) 51.22 TRi,561.22 PDE,0 CAN, 293 B5 STA TIT,IDE=1924C'

~*+Hot 000.0 PFF 80 Voids Bra-di RES,19240',0,5,10 15,2D 25 30,40,50,60

%01I,80 TM) 561.22 TRJ,561.12 PDE,0 CAN,, 293A5s XEN,0 SrA T1T,I0E--1924D'

• +Hot 200.0 PIM 0 Voids Bi-ardi RES, '9240' 0 5 10 15,20 25 30 40 50,60 vw1,o Tmo,!i.;ý2 TfU,56Mi.2 RS,6 CAN 293.15 BOR:190.0 XEN,0 SrA TJT,IDE='924E' -Mfot 200.0 PPM 40 Voids Bradih RES 192401 05:1015 20 25 30,40,50,60 rMUt 4U PAGE 50 )EN,0 STA TIT,IDE=-924F'

• +HIot 200.0 PPMN 80 Voids Braxh RES 19240',0,5 10 15 2 25 30,40,50,60 VDI:80 TMO,561.22'TRJ,561.02 PDE,0 CAN," 293.15 XEN,0 SEA TIT:!DE='924G'

• +Hot 400.0 PPM 0 Voids Branch RES,'9240-,0,5,10, 15,20,25,30,40,50,60 V.O1,0 TMO,561.22 TRJ,561.22 PDE,0 CAN, ,293.15 OR,"460.0 SEA TIT IDE=1924H'

• +Hlot 400.0 PPM 40 Voids Brarch RESS-9240-,0,5,10,15 2D 25 30 40,50,60 M3 , 40 7M) 561 .22 TRJ,561 .2 OE,0-: CAN, 293.15 XEN,0 STA TIT,113E=192411

• +Haot 100.0 PFM 80 Voids Bradic RES,'9240-,0,5810

'15 20 25 30,40,50,60

%OFBO8 TM),561.22 T~if,51.E2 PVE,O cAN 293.15 XEN,0 STA TIT,IDE=1924J'

• +Hot 800.0 PFM 0 Voids Branch RES, '9240' 0 5 10 15 20 25 304050 '60 %I01 : V, 6i.L iFU,56U.2 Fk~ CAN, 293.15 )GEN,0 STA TIT IIDE=1924K'

• +Ho0t 8M.0 PPM 40 Voids Bra-di REs,'9240',0 5,10 15 25 230,40,50,60

%0I,40 TM) 561 22 Tei56.W.2 PDE,0 CAN:, & 5 SEA TITIDE='924L'

• +if0t 8W0.0 PPM 80 Voids Brwdi RES:'92401,0,5,10

'15 20 25 30 40,50,60 ~Vol3 M 561 22 TFU,ig61.F2 iDE,0 CAN:,9 i5 BOR 761. XEN,0 STA BO3 Table 2.4.5-2 CASMO AMM File for Run49900755 1 It92.4.oA 2 Lt92.4.e-rw 3 *nme 4 *mW 5 *rwnLO 63 7 VCI=40,TR)=M.0,TK)=-ý.2,IDE=492401I

• Lt92--10x1O--e-4.19 13GcI5 .40 W a *nalp-* 9 txne* 11 1 taig 2 *nob& 3 *ncre 4 *aU* 7 /usý--m/NF/scripts/Casmo-icb.ksh 8 /users/NF/birVcasTa3N4P7-" 9 10 11 12 5 1 2 *nms* 3 Asers/chaiko/cmm 5 Lt92.4 6 *nme 7 At92.4.ci 8 /Lwm/NF/midxwT5/Lt92.g@dfite 9 *recýfi redk 10 *reqjire&

11 *rvm-*

w 10 10 10 u.u 0.4 0.8 0.0 0 0 200 I .4~I -+- I IU U.4 4 4 ---..- -..I U J A -I 10 10 0.0 400 U.4 400 3.9572E-03 4n0(s .JU f "tQ ---t . 3.0077E-22 l 2.9146E-22 2.4945E-03 1 .9878E-03 1.51 96E--03! 4-9665E-03 0 0 200 156.630382 113.2607641 43 0 313.2607641

-. 40 ,a * --r ,lr n/J --2 U-Ji -" r.Qd ." I QO-"O 151 10 0.0 800 9.8461 E-03 800 161 10 0.4 800 7.8428E-03 626.5215281 171 10 0.8 800 5.9967E-033.0430562 63 20 0.0 0 3.0696E-22 o_ 0 73 20 0.4 0 310119E-22 0 83 20 0.8 0 2.9293E-22 0 93 20 0.0 200 2.4903E-03 200 103 20 0.4 200 1.9750E-03 156.630382 113 20 0.8 200 1.4962E-03 113.2607641 123 20 0.0 400 4.9589E-03 400 133 20 0.4 400 3.9323E-03 313.2607641 143 20 0.8 400 2.9793E-03 226.5215281 153 20 0.0 800 9.8331E-03 800 163 20 0.4 800 7.7960E-03 626.5215281 173 20 0.8 800 5.9078E-03 453.0430562 0 Table 2.4.5-3 CASMO Results for Run#9900755 0 71 81 91 101 I2 121 131 141 t Void history is 0.4. ý Bypass void is zero in all cases.1.9878E-03

! 1.5196E-03,I

• 4.9665E-03

'0.8 I I 200 U.U10 400 PAGE 53 0.010 0.008 0.006 ,, (CA V -3) 0.004 0.002 0.0004 0 --200. 400 F'I(CBC*j; CBBj; cXCj;600 800 OLBj)0.01 0.008 0.006 XD.2 (RP -3) 0.004 0.002 01 0 200 400 600 800]LF(CB~Ci, CBBj, (tcj, 0XB'j) Figure 2.4.5-1 CASMO-3 calculated macroscopic thermal absorption cross section for B-10 plotted against the function IF given in Eq. (2.4.5-2).

Line segments included to illustrate the nearly linear relationship between B-10 cross section and Mr. Calculations are for ATRIUM- 10 fuel with lattice type 92. (CASMO Run#9900755)

PAGE 54 2.4.6 Boron Transport Model In the SABRE model, boron is transported throughout the reactor by the coolant flow. A control-volume approach, with up to eighteen control volumes (see Table 2.4.6-1), is used in calculating the local boron concentration.

Boron in each control volume is assumed to be perfectly mixed. In the SABRE boron mixing model, boron is only transported by liquid flow; the volatility of boron is neglected.

For boron mixing and transport, fine noding is used within the core region, because within this region, there is generally a large gradient in coolant void fraction due to heat addition from the fuel. Since boron is only contained within the liquid phase of the coolant, variations in void fraction are accompanied by variations in boron concentration.

In the other regions of the reactor vessel such as the upper plenum, riser, and separator, large spatial variations in void fraction do not typically occur and therefore coarse nodalization is used. In the Susquehanna reactors, boron is injected by the SLCS into the lower plenum region of the vessel. Since the-borated water-i-rjected by the SLCS is more dense than the reactor coolant, there is concern that under low-flow conditions boron may stagnate in the lower head of the vessel. Experimental studies on boron mixing efficiency, carried out by Dias, Yan, and Theofanous, 15,16 indicate that. complete upward entrainment of boron into the core "is 100% efficient and complete, down to flows of at least 6% (of rated), perhaps even as low as 4%." The discussion in the Technical Evaluation Report17 supporting the NRC's Safety Evaluation Report18 on ATWS mitigation strategies corroborates this view on boron mixing efficiency.

When referring to the boron mixing data of Dias et al, the TER states: "The new data suggests that mixing is almost perfect (i.e., 100% efficiency) even for flows as low as 4% to 5%." Here 100% core flow corresponds to 100 MLb/hr. SABRE simulates boron injection by adding boron to the upper node of the lower plenum region since this corresponds to the location of the injection standpipe within the reactor.

If the total core flow is greater than W 2 , which is specified as part of the SABRE input data, then the injected boron becomes completely mixed within this upper control volume. On the other hand, if the total core flow is less than a specified value W 1 , then the injected boron is immediately transferred to the bottom node of the lower plenum to simulate boron stagnation under low-flow conditions.

Thus, boron entrainment is 100% complete if the total core flow is greater than W 2 , and there is no entrainment of boron if the total core flow drops below W 1. " 5 Dias, M.P., Yan, H., and Theofanous, T.G., 'The Management of ATWS by Boron Injection," Proceedings, NURETH-5, Sept. 21-24, 1992, Salt Lake City, Utah. 16Dias, M.P., Yan, H., and Theofanous, T.G., "The Management of ATWS by Boron Injection," NUREGICR-5951, March 1993. 17 March-Leuba, J., "Technical Evaluation Report ORNL/NRC/LTR-94/28, Revision 2, November 1995. " 8 "Safety Evaluation Report: Modifications to the Boiling Water Reactor (BWR) Emergency Procedure Guidelines to Address Reactor Core Instabilities," Letter from D.B. Matthews (NRC) to KP. Donovan (Centerior Energy), June 6, 1996.

rIAUr 00 Based on the boron mixing studies referred to above, the core flow W 2 has a value of 6 Mlbm/hr. The value of W, is conservatively taken as 4 MLb/hr since there is no available information on boron entrainment for core flows less than 4 MLb/hr. For flows between W 1 and W 2 , the fraction XB of injected boron that is entrained upward into the core is computed from W_ ) <W <W2) (2.4.6-1) where Wrepresents the total core flow rate (fuel channel and bypass flow). Note that XB=I for W>W 2 and XB=O for W<WI. When b=1 in Eq. (2.4.6-1), the fraction of the injected boron which is entrained upward into the core decreases linearly from 1.0 to 0.0 as the total core flow decreases from W7 (6 MLb/hr) to W 1 (4 MLb/hr). Since the boron mixing experiments of Dias et al. suggest that 100% of the injected boron may be entrained at core flows as low as 4 MLb/hr, it is conservative to use the linear entrainment model represented by Eq. (2.4.6-1) with b=l. Unless otherwise stated, all ATWS simulations in this calculation package are performed with b=l. A sensitivity study with respect to the exponent b is performed for an MSIV closure ATWS in §5.8. Figure 2.4.6-1 shows the boron entrainment fraction XB for b=l/2, b=l, and b=2. Remixing of stagnated boron is also an important phenomenon considered in the SABRE model. If the total core flow is less than the remixing threshold Wo (a value of 15 Mlbm/hr'9 is used, but as with W 1 and W 2 , any value can be specified as part of the code input data), the flow exiting the jet pumps is assumed to pass directly into the upper node of the lower plenum so that the bottom node of the lower plenum is completely bypassed.

Thus, there is no remixing of stagnated boron if the total core flow is less than Wo. The experimental studies on boron mixing indicate that little flow reaches the bottom of the lower head if the total core flow is less than the threshold value Wo (15 Mlbm/hr).

For flow rates greater than W 0 it is not clear what fraction of flow extends into the region near the bottom of the lower head. Therefore, in the SABRE boron mixing model, the following assumptions are made with regard to the flow distribution in the lower plenum: "* No recirculation flow passes through the bottom node of the lower plenum if the total core flow Wis less than Wo. "* The flow through the bottom node of the lower plenum is W- Wo if the total core flow is greater than Wo.19 See Section F.29.8.

PAGE 56 Table 2.4.6-1 Description of Control Volumes Used in Calculating Local Boron Concentration in Reactor Coolant 3/4 R~eactor Floyw R6iou y Nuxbt' ofkd se niortiM t~ -M I!)e O o Jet Pumps 1 L6wer Plenum 2 Lower Reflector 1 Active Core 2 1 1 to 10 Upper Reflector 1 Bypass 1 Upper Plenum, Riser, & Separator 1 Downcomer I 2 0 Number of nodes is specified as part of the input data.

1 0.9 0.8 0.7 0.6. 0.5 0.4 0.3. 0.2 0.1 O0 Figure 2.4.6-1 Boron entrainment factor , as a function of the exponent b in Eq. 2.4.6-1.S 0 U 0 C C Lii C 0 0 m 0 V S 0 U IL 3.5 4 45 5 5.5 6 6.5 7 TotaI Core Flow (MLb/hr)3 PAGE 58 2.4.7 Decay Heat Fraction In addition to the fission power, heat is generated within the fuel as a result of fission product decay. In SABRE, decay power is computed from the eleven-equation decay model used in the RETRAN code: QdyW(t)= Qfi (O)Z'",()

(1 (2.4.7-1) where the decay heat group "concentrations" yi(t) are obtained from the solution of dy + 1) -0 Yi Q ,(0) (2.4.7-2) with (2.4.7-3)

In the above decay heat equations, Qdecay() = decay power (Btu/sec), Qfiss(t) = total fission power = qi,, (t) Vf (Btu/sec), Vf = volume of fuel (fM 3), yi(t) = "concentration" of i-th decay heat group (sec), = decay constant for the i-th group (sec-1), and Yi = yield fraction for the i-th group (dimensionless).

Values for the X, and Yi are given in the RETRAN code description manual. Summation of the fission power, decay power, and heat dissipation from the reactor vessel Q_, and vessel internals Qv7 (see Section 2.11), gives the total heat generation rate Q(t) within the reactor, Q0)= Qf (t)+ Qd..y(0)+

Qf 1 (t) + Qw (t) (2.4.7-4) 2 1 McFadden, J.H., "RETRAN-02

-A Program for Transient Thermal-Hydraulic Analysis of Complex Fluid Flow Systems," Volume 1, NP-1850-CCMA, Electric Power Research Institute, Palo Alto, California, November 1984.

VAUE 59 2.4.8 Control Rod Model The kinetics file for SABRE consists of a SIMTRAN output file containing seven 1-D cross-section sets with each set corresponding to a different control rod pattern. The differences among the seven cross-section sets are solely due to the presence of control rods. Void and fuel temperature feedback effects are maintained constant across the seven cross-section sets. That is, control rod position is assumed to change instantaneously relative to the initial or base rod configuration which typically corresponds to an all-rods-out configuration.

In each cross-section set, dependency on moderator density and fuel temperature is explicitly included through polynomial expressions of the form shown in Eq. (2.4.1-1).

The effect of control rod movement is incorporated into the SABRE 1-D diffusion theory model by interpolating between appropriate cross-section sets within the SIMTRAN output. The seven cross-section sets in the SIMTRAN output consist of the following:

1. Initial control rod configuration (typically all rods out or a small number of fully or partially inserted rods) 2. Set 1 with four additional rods completely inserted (The four additional rods are quarter-core symmetric)

3. Set 1 with eight additional quarter-core-symmetric rods completely inserted 4. Set 1 with sixteen additional quarter-core-symmetric rods completely inserted 5. Set 1 with thirty-two additional quarter-core-symmetric rods completely inserted 6. Set 1 with sixty-four additional quarter-core-symmetric rods completely inserted 7. All rods in Generation of the SLMTRAN output with these seven cross-section sets is a two-step process. First, a SIMULATE restart file is obtained from the PP&L Nuclear Fuels' Group for the desired cycle, initial rod pattern, and cycle exposure.

A SIMULATE run specifying the six additional rod patterns is then made using this restart file. The SIMULATE calculation generates a second restart file which is then used by SIMTRAN, along with a SLMTRAN input file, to generate the 1-D cross-section file used by SABRE. As part of the present calculation, four 1-D cross-section sets are generated.

These sets are described in Table 2.4.8-1.

SABRE includes two different control rod models. The first is a scram model in which all control rods move simultaneously into the reactor core over a specified time period. The scram time is specified as part of the SABRE input data. In the case of a reactor scram, only Sets 1 and 7 out of the seven cross-section sets contained within the SIMTRAN output file are used. As the bank of rods moves into the core, interpolation PAGE WU between cross-section sets I and 7 is carried out on a nodal basis. That is, a control rod fraction is computed for each axial node (material region) of the core. If the control rod bank has not reached a particular axial node, then the control rod fraction for that node is 0, and cross-section Set 1 is used for that material region. If, on the other hand, the rods have passed through a particular material region, then the control rod fraction for that region is 1 and cross-section Set 7 is used. Note that the I-D cross-section sets generated by SIMTRAN are defined on a nodal basis. If the control rod bank is passing through an axial node of the core, then the control rod fraction for that material region is between 0 and 1 depending on how far the rods have moved through the node. In this case, the cross-section set for that material region is computed by interpolating between cross section Sets I and 7 using the control rod fraction as the interpolation parameter.

The second control rod model contained in SABRE is used to describe reactor shutdown by manual control rod insertion (MRI) using the Control Rod Drive (CRD) system. In an ATWS event in which the SLCS fails to initiate and the CRD pumps are available, the control room operator can individually select and insert control rods from the Control Room bench board using the Reactor Manual Control System. The acceptance criteria for CRD Stroke Time Test (TP-055-001,Re,.

7) indicates that the-insertion time for a single control rod is 60 seconds. In the SABRE code, the rate of control rod insertion for MRI is specified as part of the input data. It is assumed that the operator will drive rods so that they are distributed fairly uniformly about the core. Since MRI is a relatively slow process, the important changes in core flux occur on a global basis. That is, it is not necessary to describe detailed changes in neutron flux as individual rods move through the core. The only relevant effect which needs to be described is the gradual decrease in core average power as rods are individually inserted as this is the only factor which influences suppression pool heat up during an ATWS with SLCS failure. During MRI, the effect of gradually increasing control rod density within the core is captured by interpolating, on a core wide basis, between the appropriate SIMTRAN cross-section sets. Seven different cross-section sets with increasing control rod fraction are generated in order to obtain a fairly detailed description of overall core response during MRI.

PAGE 61 Table 2.4.8-1 Description of 1-D Cross Section Sets for SABRE

.fa -U.Table 2.4.8-2 SIMULATE Input File (u2c7.simulate.in)Used in Generating U2C7 Cross-Section File 15 15 25 5 0 0 0 0 0 15 16216 43172 24 72100 10 80 13 0 0 0 amcas2920 202002222222222 disk L2C Al,,l (01 BE) BEGD(P EO DEE p, P" O 'T SWI PRFS 1 11.562 11.562 00.0D0 (1) Execute INWUT & MLic 2 0 (2) 1=Initialize Wilire to present distrib & reset to 2 (7) DWirpt vaLue of stbfcot Lsed to calc density a0 51 1 20 S4. 0 (3) 0=Use restart Xe -4Jse equit Xe 21 S2 0 (5) Search Option O=No Search 2=Pwr Search 7zuse fi Le power distr 21 1 0 (9) 0=No DP Catc 3=0etL FIB 4=Apprx FIB6 21 93 3 (27) =C0mtrot bLade deptetion has no effect 21 S17 0 99 ddisk L2C7 ATUS (02 4 rods in) (25) Limit on nodat power distribution peak 2 S24 100. (2) 2=Carry existing HaiIing fiLes from prev case Wo charge (2) 1=Initialize Hailing to present distrib & reset to 2 20 S1 1 (3) OtUserestartXe 24keeqiLt Xe 21 S2 0 (5) Search option O=o Search 2=Pwr Search 7=use file power distr 21 S1 7 30 8 ST --.------------

30 9 S7 --00-----------

30 10 S7----------------

30 11 S7 ----10---------

30 12 S7 -------------

30 13 S7------------

30 14 S7 ---------30 15 S7--------

99 ddisk 127 ATmS (MS 8 rods in) (25) Limit an nodal power distribution peak 2 S24 100. (2) 2=Carry existing Hailing fites frao prev case Wo dcrge 20 S1 2 (3) 0=UserestartXe 24seeqtiLXe 21 S2 0 (5) Search Option O=No Search 2=*r Search 7=Tse fiLe power distr 21 S1 7 30 8 S7 --.------------

30 9 S7 --00-----------

30 10 S7 T-----------

.... 30 11 S7 --0010 ---------30 12 S7 -------------

30 13 S7------30 14 S7----------

30 15 S7--------

99 ddsk U2C7 ATWS (04 16 rods in) (25) Limit on nodaL power distribticrn peak 2 S24 100. (2) 2=Carry existing HaiLing files fran prey case W/o chýa 20 sl 2 (3) 04ieurestartXe 2--seec#z(lXe 21 S2 0 (5) Search Option 044a Search 2*Pa Search 7-se fiLe power distr 21 1 7 30 8 S7 .. .. .. .. .. .-... . 30 9 S7 --00---00 30 10 S7 -- ... 30 11 S7 --001000 30 12 S7 --. ..... 30 13 S7 . 30 14 S7 . 30 15 S7 -------99 ddisk L12C7 ATlS (05 32 rods in) (25) Limit on ncdat power distribution peak 2 S24 100. (2) 2--Carry existing HaiLirg files from prey case w/o dhange 20 S1 2 (3) O=LerestartXe 2--Jse ejiLXe 21 S2 0 (5) Search Option O=No Search 2=Pwr Search 7=use file power distr 21 51 7 30 8 S7 --.------------

30 9 S7 --00--00--0

..O . 30 10 S7 ----------.----

30 11 S7 --001000---00-.

30 12 S7 ------------

30 13 S7 --00--00 ... 30 14 S7 --.--- ---30 15 S7 -------99 duisk L2C7 AllS (06 64 rods in) (25) Limit on noda[ pier distributian pea* 2 S24 100. (2) 2=Carry existing Hailing fites fronm pre case w/o charge 20 SI 2 (3) O=UserestartXe 2-Use ei Xe 21 S2 0 (5) Search qption O=No Search 2Pr Search 7=Ae fiLe power distr 21 91 7 30 8 S7 --.-- 0 ------ 00-30 9 S7 --00--00--00

.-30 109S7 00--00--00- 30 11 S7 --001000--00

--. 30 12 S7 -- 00--00 ---30 13 S7 --00-- 0---30 14 S7 00 -- 00-- -30 15 S7 -------99 ddisk L2C7 ATJS (07 ALL Rock in) (25) Limit on rnoal power distribution peak 2 S24 100. (2) 2=Carry existing Hailing files from prey case W/o dange 20 Sl 2 (3) 0=UserestartXe 2-use ect Xe 21 S2 0 (5) Search Option 0=#o Search 2=Pwr Search 7=use fite power distr 21 Sl 7 30 8 S7 00 00000000000000 30 9 S7 0000000000000000 30 10 S7 00000000 0000000 30 11 S7 0000000000000000 30 12 S7 00000000000000 30 13 S7 000000000000 30 14 S7 0000000000 30 15 S7 00000000 99 ]LAST t-/AUL D0%Table 2.4.8-3 SIMULATE Input File (u2c9.simulate~in)

Used in Generating U2C9 Cross-Section File 15 15 25 5 0 0 0"0 015 16216 43172 24200100o10o 0 0 0 0 Dkcs2920 2020022.22222222 ddisk tLM A11S (01 ALL Rods Out) BEGE)(PE)D MPDELEG POWE WT SUB PIES 1 15.700 15.700 00.000 (1) Execute INlT & UCZLE 2 0 (2) 1=Initialize Hailing to present distrib & reset to 2 (7) O=irptA aLue of stkcooL used to catc density 20 Si 1 20 S4 0 (3) 04=serestart Xe 2'.seaejit Xe 21 S2 0 (5) Search Option O=No Search 2=wu Search 7=wse file power distr 21 Sl (9) O=No DP CaLc 34)etL FIBWR 4-pprx FIBSR 21 S3 3 .(27) O=Control blade depletion has no effect 21 S17 0 30 8 S7 --......

30 9 S7--------------

30 10 S7................

30-11--S-----------

---

30 12 S7 30 13 S7------------

30 14 S7----------

30 15 S7--------

99 dcisk LM9 ATS (02 4 rods in) (25) Limit on nodal power distribution peak 2 S24 100. (2) 2=Carry existing IaiLing files fran p-et case w/ocadge (2) 1=Initialize HaiLing to present distrib & reset to 2 20 S1 1 (3) 0tkUerestartXe 2-Uzse equi Xe 21 S2 0 (5) Search Option O=lo Search 2*>w Search 7=use file power distr 21 S1 7 30 8 S7 30 9 S7 --00............

30 10 S7................

30 11 S7................

30 12 S7..............

30 13 S7------------

30 14 S7----------

30 15 S7--------

99 dJisk LMC ANS (3 8 rods in) (25) Limit on nodal er distribution peak 2 S24 100. (2) 2=Carry existing Hailing files from prey case w/o charge 20 S1 2 (3) 0=tserestartXe 2--Useeq*a Xe 21 S2 0 (5) Search option O=%o Search 2=Pwr Search 7=use fiLe power distr 21 S1 7 30 8 S7 ................

30 9 S7 --cO-----------

30 10 S7 ---------------

"30 11 S7 --00-----------

30 12 S7 -------------

30 13 S7-----------

30 14 S7----------

30 15 ST -------99 ddisk L12C9 ATW (04 16 rods in) (25) Limit on rodaL power distribution peak 2 S24 100.

(2) 2=Carry existing Hailing files from prey case W/o dage 20 Si 2 (3) 0UwserestartXe 2-4Jeqoil Xe 21 52 0 (5) Search Option 0=No Search 2=Pwr Search 7=use fi te pow distr 21 S1 7 30 8 S7 .--.-----------

30 9 S7 --00--00-------

30 10 S7 T-----------....

30 11 S7 --00--0 .O-.....

30 12 S7 ---------30 13 S7T ...... .... .. 30 14 S7 ---------30 15 S7 -------99 ddisk L2C9 ATWS (05 32 rods in) (25) Limit on nxbl. power distributicn peak 2 S24 100. (2) 2=Carry existing Hailing files from pew case w/o chane 20 S 2 (3) 0=UserestartXe 2ise- 21 S2 0 (5) Search Optio O=No Search 2=Pwr Search 7ue file power distr 21 SI 7 30 8 ST --.------------

30 9 ST --00--00--00--. . 30 10 S7 ...--------.--

30 11 S7 ----0 0--00 ... 30 12 S7T..............

30 13 S7 --00-- 00 .. 30 14 S7 ---------30 15 S7--------

99 ckldsk UW.9 AIUS (06 64 rods in) (25) Limit on nodal power distribition peek 2 S24 100. (2) 2-Carry existing Hailing files from prey case wo chare 2D sl 2 (3) ")se restar-tXe AI~se e~ I Xe 21 S2 0 (5) search option 0=No Search 2=lwr Search 7=use file power distr 21 Si 7 30 8 S7 --.-- 0 .-- 00-30 9 S7 --00--00--00

... 30 10 ST 00--00--00- 30 11 S7 --00--00 --00-30 12 S7 --. 00 --000 30 13 S7 --00--00 .-30 14 S7 00 -- 00 --.. 30 15 S7 T------ . 99 Wdisk 12C9 ATWS (07 ALl Rods in) (25) Limit on rndal pwer distribition peak 2 S24 100. (2) 2=Carry existing Railing files fran prey case w/o chanee 20 Si 2 (3) 0=userestartXe 2MUse ecotXe 21 S2 0 (5) Search option O=Mo Search 2*ir Search 7Tuse file porw distr 21 S T 7 30 8 S7 0000000000000000 30 9 S7 0000000000000000 30 10 S7 0000000000000000 30 11 S7 0000000000000000 30 12 S7 00000000000000 30 13 S7 000000 0000 30 14 S7 0000000000 30 15 S7 0000000 99 r,%ur- no Table 2.4.8-4 SIMULATE Input File (u 2 cI0.simulate~in)

Used in Generating U2Cl0 Cross-Section File 15 15 25 5 0 0 00 0 0 0 0 0 Onas230 20]2002222222222 duisk t2C10 ATWS (01 All Rods Ohtj) BEGEMPEMD PDELG'POER t" SUB PRES 1 15.290 15.200 00.000 (1) Execute IWUT & M 2 0 (2) 1=Initialize HaiLirg to present distrib & reset to 2 (7) O=irp.t value of sucoot used to catc density 20 SI 1 23 ) 4 0 (3) OtUse restart Xe 24-se eqayt Xe 21 52 0 (5) Search Option O=No Search 2*r Search 7=use file poer distr 21 S1 0 (9) 044o DP Calc 3=OetL FIBIR 4Apprx FIBWR 21 S3 3 (27) O=Ccntrot blade depletion has ro effect 21 S17 0 30 8 S7 30 9 S7................

30 10 S7 30 11 S7 30 12 S7..............

30 13 S7-30 14 S7----------

30 15 S ----.--.-99 ddisk LaiO ANlS (02 4 rods in) (25) Limit on nodat pwoer distribation peak 2 S24 100. (2) 2=Carry existing Hailing files from prey case '/o charge (2) 1=Initialize Hailing to present distrib & reset to 2 20 S1 1 (3) O=tserestartXe 2-4seequil Xe 21 S2 0 (5) Search Option 0=No Search 2=Pwr Search 7se file pier distr 21 S1 7 30 8 S7 --T-30 9 S7 --00-----------

30 10 S7 30 11 S7................

30 12 S7-------------

30 13 S7............

30 14 ST ---------30 15 S7 -------99 ddisk 12C1O AM"S (03 8 rods in) (25) Limit on nodat power distribtion peak 2 £24 100. (2) 2=Carry existing Hailing files from prev case w/o dage 20 S1 2 (3) 0=lserestartXe 2-Useeauit Xe 21 S2 0 (5) Search Option D=No Search 2=Pwr Search 7=use file per distr 21 S1 7 30 8 S7 ................

30 9 S7 --00............

30 10 S7 ................

30 11 S7 --O............

30 12 S7 30 13 S7............

30 14 S7 31 15 S7 -------99 duisk L2C10 ADS (04 16 rods in) (25) Limit on nodal power distribition.

peak 2 £24 100. (2) 2-Carry existing Haitire files fron prey case w/o charge 20 21 21 30 30 30 30 30 30 30 3D 99 ddisk 12C10 ATI 2 20 21 21 30 301 30 1 30 1 30-1; 3D 14 30 1 99 ddisk U2C10 Ana 2 2O 21 21 30 8 30 9 30 10 30 11 30 12 30 13 30 14 3D 15 99 ddisk 2C10 ATm 2 20 21 21 30 8 3D 9 30 10 30 11 30 12 30 13 30 14 30 15 99 ILAST.Sl 2 (3) O=4serestartXe 2-tseeqit Xe S2 0 (5) Search Option 0=4o Search 21Pwr Search 7=use file power distr 51 7 8 S7.. .... 9 S7 --00---00-......

10 57 -- ... 11 S7 - 0 12 S7 -- ... 13 S7 14 S7..........

15 S7 ........

S(05 32 rofs in) (25) Limit on nodat power distribution pek S24. 100. (2) 2=Carry existing Hailing files frm pev case w/o chare S1 2 S(3) 04be restart Xe Z-Jse ec iL Xe S20 (5) Search Optio O=nNo Search 2=Pwr Search 7=use f ie powr distr S 7 8 S7 -- ... 9 ST --00-- 00 00-0 .. 0 ST ........ 1 ST --

- S7---------

4 ST --. .---7. . 5 S7-- ---- --- 5S7..........

(06 64 rock in) (25) Limit on rndal power distribution peak S24 1(0. (2) 2-Carry existing Hailing files fram prev case w/o charge S1 2 (3) O r r 2se ec.iLtXe S2 0 (5) Search Option O=No Search 2 r Search i7use fi le power distr 51 7 ST ---- 00 -----00 ST -- 00 .--00--. 0-S7 00 -- 00 -- 00 -- 00 .. S7 -- OD -- 00 -- 00 -S7 -- 00 -- 00 --0. S7 00-- O0 -S7 OD -- OD --. S7 -------(07 All Rods in) (25) Limit on nodal power distribution peek S24 100. (2) 2=Carry existing Hailing files fran prev case w/o change S1 2 (3) O=UserestartXe 2--Uweecpjit Xe S2 0 (5) Search Option O=No Search 2*ig- Search 7wuse file power distr S1 7 ST 00 00 00 000 0000 00 S7 0000 00 00 00 00 00 00 S7 00 O0 O 0 O00 00 O0 S7 00 O00 O00 00 00 00 S7 00 00 O0 00 0 O 00 S7 00 O0 00 0 0000 S7 00 00 O0 G00 0 S7 OD 00 O0 O0 PAGE 68 Table 2.4.8-5 SLMIULATE Input File (ulcl2.simulatedn)

Used in Generating U1C12 Cross-Section File 15 15 25 5 0 0 0 0 0 15 16216 43172 24200100 10 8D 0 0 0 0 202002222222222 ddisk UlC12 ATWS (01 AiL Rods Out) EWE DELEa' KRw Wr SUB PRES 1 14.800 14.800 00.000 (1) Execute INPUT & ILJQE 2 0 (2) l=MnitiaLize Hailing to present distrib & reset to 2 (7) (=irput vaLue of stbc~oo used to caLc density 2O Si 1 20 S4 0 (3) O=0serestartXe 24UseeqaiL Xe 21 S2 0 (5) Search Option 0=4O Search 2*w Search 7=use file power distr 21 Si C (9) O=No DP Caic 3=DetL FIsa 4#gn FIBWR 21 S3 3 (27) O=ControL blade depletion has no effect 21 S17 0 30 8 S7................

30 9 S7 3D 10 S7 30 11 S7 30 ¶2 S7 30 13 S7. 30 14 S7 30 15 S7 99 cUisk ULC12 ATlS (02 4 reds in) (25) Limit on nodal power distribution peak 2 S24 100. (2) 2="Carr existing Haiting files fron prey case Wo dirge (2) 1=InitiaLize Hailing to present distrib & reset to 2 20 81 1 (3) O=UserestartXe 2-ise qudt Xe 21 S2 0 (5) Search Option O=Co Search 2=Pwr Search 7=use fite power distr 21 Si 7 30 8 S7 ................

30 9 S7 --00-30 10 S7 -- ... 30 11 S7................

30 12 S7...............

30 13 S7------------

30 14 S7 --.--.---

---30 15 S7 -99 duisk LUC12 AilS (03 8 rods in) (25) Limit on nodal powe distributian peak 2 S24 100. (2) 2=Cary existing Hailing files from prey case go dcage 20 Si 2 (3) OAlserestartXe 24Jseecuit Xe 21 S2 0 (5) Search Option 0=No Search 2=*. Search 7Nuee file powe distr 21 S1 7 30 8 S7 -....... 30 9 S7 --0 . 30 10 S7 -- ... 30 11 S7 --00 30 12 97 -30 13 S7 30 14 S7..........

30 15 S7--------

99 cdisk ULC12 ATWS (04 16 rock in) (25) Limit on nodal power distribution peak 2 824 100.

20 21 21 30 8 30 9 30 10 30 11 30 12 30 13 30 14 30 15 99 ckisk U1C12 ATWS 2 20 21 21 30 8 30 9 30 10 30 11 30 12 30 13 30 14 30 15 99 ciisk L1C12 ATWI 2 20 21 21 30 8 30 9 30 10 30 11 30 12 30 13 3014 30 15 99 ddisk U1C12 ATWS 2 20 21 21 30 8 30 9 30 10 30 11 30 12 30 13 30 14 30 15 99 PAGE 69 (2) 2=Carry eisting Hailing files fran prey case w/o change S1 2 (3) 0=Use restartXe 2-=Use eaitXe S2 0 (5) Search Option 040 Search 2iwr Search 7=use file powr distr 51 7 S7 --.------------

S7 --00-- O-------S7 ---------------

S7 --00 0-- O -- ---S7 -------------

S7 ...... S7----------

S7--------

(05 32 rods in) (25) Limit on nodal power distribution peak 524 100. (2) 2-Carry existing Hailing files fram prey case w/o dmhne SI 2 (3) 0=JserestartXe 24JseequiL Xe S2 0 (5) Search Option 0=1o Search Z*ti- Search 7=use file power distr Si 7 S7 -- -------------

ST -OD 00--S7 --00--00--0

.. S7 00.. .... ... ST -- -- O0- .--S7 -.00-... -. S7 -------(06 64 rods in) (25) Limit on nodat power distribution peak S24 100. (2) 2-Carry existing Hailing files from prev case Wo change S1 2 (3) O=UserestartXe 2iL eqcl Xe S2 0 (5) Search option O=Wo Search 2*Pw Search 7=use file power distr 51 7 S7 --.-- 0 .-- 00-S7 -- 00-- O0 -- 0 --S7 00 -- 00 -- 00 -- O0 -S7 -- 00-- 0 -- 0--S7 -- OD0--0O-S7 -- 00 -- O0 --.. S7 GO-- 0 --S7 .... (07 ALL Rods in) (25) Limit on nodal power distribution peak S24 100. (2) 2=Carry existing Hailing files from prey case w/o dchage Sl 2 (3) 0'=(. restartXe 2=Jseecuit Xe S2 0 (5) Search Option O=No Search 2WPr Search 7=ue file poe distr S5 7 S7 00 00 00 00 O0 00 00 00 S7 00 0 00 0000 00 00 00 S7 00 00 00 00 00 00 00 00 S7 00 00 00 00 00 00 00 00 S7 O0 00 00 00 O0 0 00 S7 00 00 00 00 00 00 S7 00 00 00 0 000 S7 00 00 O00 PAGE 70 Table 2.4.8-6 SIMTRAN Input File (u 2 c7.simtran.in)

Used in Generating U2C7 Cross-Section File /useschaiko/SIJATE/u2cT.simulate.restart 1 2 3 4 5 6 7 9999 Le7 ATIS SJMTRAN EC-A1W-0505 Rev 7 1 1 1 0 6-- A -.- -.dsk, 2 3 4 5 6 7ddi~r~ku~~jj 10 80 0 0 14100 dimmion 0 0 0 0 0 0 0 0 edit fLags 24 0 1 1 2 2 0 iter,kdfLag, I# Lagksolv, krwtfktwf, ,+,, 64 rdirdtf 45.8244 rhdef 1 11 Lstsau / ncnnm 0.60 1.20 2.40 3.60 4.80 6.00 7.20 8.40 9.60 10.80 12.00 33 12 12 12 12 12 1212 1212 12 12 12 12 12 12 12 12 12 12 12 12 1212 12 12 33 1.057 1.131 1 25 0.00 -0.10 .100 .200 .300 .400 1 25 0.0 -100. 60. 225. 312001 5 1 0 312001 .2 .4 .6 .8 1.0 0. .17150 .21132 .25230 .2939 312001 -33458 312002 5 1 0 3120(2 .2 .4 .6 .8 1.0 0. .0043117 .0043622 .0044160 .0044720 312002 .0045297 SIW1 312003 5 1 0 312003 .2 .4 .6 .8 1.0 0. .00490 .010767 .016879 .02354 3120a3 .029769 SIGTR2 312004 5 1 0 312004 .2 .4 .6 .8 1.0 0. .46027 .6878 .91722 1.14790 312004 1.3788 SIIM2 312005 5 1 0 312005 .2 .4 .6 .8 1.0 0. .067142 .0686B .07"216 .071771 312005 .-073 WJ.2 312006 3 1 0 31200 .33478 .61W 1.0 0. 3.771n0.S 3.694E40 3.63%E.5 WEL1 312007 3 1 0 312007 .33478 .6197 1.0 0. 1.527E.07 1.4492E#07 Simhi 312008 5 1 0 312008 .2 .4 .6 .8 1.0 0. .085254 .11156 .13784 .16418 312008 .19060 SilifI 312009 5 1 0 312009 .2 .4 .6 .8 1.0 0. .00048065

.00051302

.00054752 312009 .00058543

.00062D40 SjIR1 312010 5 1 0 312010 .2 .4 .6 .8 1.0 0. .0030222 .0067297 .010649 .014724 312010 .01M15 SIGTI2 312011 5 1 0 312011 .2 .4 .6 .8 1.0 0. .209o -34945 .49690 .6442 312011 .7929 SIGf2 312012 5 1 0 312012 .2 .4 .6 .8 1.0 0. .0053686 .0063474 .0073415 .0083368 312012 .0097332 WEL2 312013 31 0 312013 .33478 .61987 1.0 0. 3.7710E#(5 3.639E(-5 VEL1 312014 3 1 0 312014 .61W7 1.0 0. 1..5278E-07 1.4839E+07 1.4492E#07 BLAW 550101 1 4 1 0 2 2001 ( GROP 1 TOTAL REmA -REFLECTlR1) 550201 2 4 1 0 2 2002 ( MROP 1 ABCRPTION

-REFLECTCR1) 550301 3 4 1 0 2 2003 ( 0JP 1 DlS('TTrl

-REFLECKORI) 55040 6 4 1 0 2 23 4 ( W0JP 2 TOTAL RIWMAL -REFLEClOR1) 550501 7 4 1 0 2 2005 ( mP 2 ABSORTION

-REFLECTOR1) 550601 21 4 1 0 2 2306 ( ROUP 2 WJMITY -RF.LECTCR1) 550701 20 4 1 0 2 2O07 ( GUP 1 ,LMCITY -REFLECTCR1) 550102 1 4 1 0 2 2008 ( GRCP I TOTAL RBS JAL -REFLECTn ) 550202 2 4 1 0 2 2009 ( GaP 1 /SPBWTJIO

-REFLECTI) 5503M 3 4 1 0 2 2010 ( GIP 1 oamNSrrTM

-REFLECRtR2) 5504(2 6 41 022011 ( RUP 2 TOTAL RBDaAL -REFLECMR2) 550502 7 4 1 0 2 2012 ( iMP 2 ABSWTIOI -REFLECICR2) 550602 21 4 1 0 2 2013 ( MP 2 Wl=iTY -IEFLEC1OR2) 550702 20 4 1 0 2 2014 ( Gim 1 ELOrITY -REFLECTIR2)

W rAUt /1 Table 2.4.8-7 SIMTRAN Input File (u2c9.simtran.in)

Used in Generating U2C9 Cross-Section File /users/chaiko/SI#4JLATE/u2c9.simruLate.restart 1 2 3 4 5 6 7 9999 tE9 ATWI SITRfd EC-ATWS-0505 Rev 7 1 1 1 0 6 jdisk, itr,krs,krutakin 234567 1080 0 0 14100 dienuium 0 0 0 0 0 0 0 0 editfLes 24 0 1 1 2 2 0 iter,kdfLag,kpfLqg,ksolv,knutf,ktwtf~kphiwt 6 4 nd_,rdtf 45.8244 rhoref 1 11 Lstsai / rcrm 0.60 1.2D 2.40 3.60 4.8D 6.00 7.20 8.40 9.60 10.80 12.00 33 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 33 1.057 1.131 125 0.00 -0.10 .100 .200 .300 .400 125 0.0 -100. 60. 225. SiGTR1 312001 5 1 0 312001 .2 .4 .6 .8 1.0 0. .17030 .2112 .25230 .29339 312001 .33458 SIGAl 312M2 5 1 0 3120D2 .2 .4 .6 .8 1.0 0. .0043117 .0043622 .0044160 .0044720 312002 .0045297 SIGRI 312005 5 1 0 312003 .2 .4 .6 .8 1.0 0. .0049850 .010767 .0168M9 .023234 312005 .029769 SiGTR2 312004 5 1 0 312004 .2 .4 .6 .8 1.0 0. .46027 .689M .91722 1.14790 312004 1.3788 S1GA2 3120D5 5 1 0 3120(15 .2 .4 .6 .8 1.0 0. .067142 .(66B .070216 .0717T1 312005 .0733a6 VE12 3120D6 3 1 0 31206 .33478 .619W7 1.0 0. 3.771(E.05 3.69,4E+05 3.6396E(5 YELl 31207 3 1 0 312007 '.33478 .6197 1.0 0. 1.527&07 1.4839E+(07 1.4492E+07 SIEml 3120O8 5 1 0 312008 .2 .4 .6 .8 1.0 0. .08254 .1115 .13784 .16418 312008 .19060 SI(Al 3120095 1 0 312009 .2 .4 .6 .8 1.0 0. .OODW .00051302

.0054752 312009 .00058343

.000620 SIGR1 312010 5 1 0 312010 .2 .4 .6 .8 1.0 0. .00(222 .0067297 .010W49 .014724 312010 .018015 SIGTR2 312011 5 1 0 312011 .2 .4 .6 .8 1.0 0. .2agO .34945 .49690 .6448 312011 .7928 SIU2 312012 5 1 0 312012 .2 .4 .6 .8 1.0 0. .0053686 .0063474 .0083568 312012 .0093342 VSl2 312013 3 1 0 312013 .33478 .61967 1.0 0. 3.7710E-05 3.6944-k05 3.63%E5 VLl 312014 3 1 0 312014 .33478 .61987 1.0 0. 1.5Z7a.07 1.4883.07 1.4492E-07 550101 1 4 1 022001 ( GIOUP 1 TOTAL RSOALW -REFLECItR1) 55C201 2 4 1 0 2 2002 ( WOP 1 ASORPTION

-REFLECTOR1) 550301 3 4 1 0 2 2003 ( ROU.P 1 DOGNSfTlfl

-REFLECTOR1) 550401 6 4 1 0 2 2004 ( GRUP 2 TOTAL X -REFLECrOR1) 550501 7 4 1 0 2 2005 C(URIP 2 ABSORPTION

-REFLECITRI) 550601 21 4 1 0 2 2006 C ILP 2 VELOCITY -REFLECTOR1) 550701 20 4 1 0 2 2007 (GROP 1 VELOCITY -REFLECTOR1) 550102 1 4 1 0 2 2008 GRO0P 1 TOTAL RBGWL -REFLECTOR2) 550202 2 4 1 0 2 2009 C G P 1 ABSORTION

-REFLECTIR2) 55002 3 4 1 0 2 2010 ( MOJP 1 0 lTSTM -REFLECTOR2) 5504026 4 1 0 2 2011 ( CDP 2 TOTAL ROX. -REFLECITR2) 550502 7 4 1 0 2 2012 G CORP 2 ABSORPTION

-REFLECTOR2) 55060 21 4 1 022013 (RP 2 VELOCITY -REFLECTOR2) 55079 20 4 1 0 2 2014 C WP 1 VELOCITY -REFLECT(R2) 99 rsuAkt /Z Table 2.4.8-8 SIMTRAN Input File (u 2 cI0.simtran.in)

Used in Generating U2C10 Cross-Section File /users/chaiko/SIRLATE/tc10.sirruiate.restart 1 2 3 4 5 6 7 9999 U210 ATWS SIMNRAN EC-ATIS-0505 Rev 7 1 1 1 0 6 Jdisk, ibw,krs,kroAkrin 234567 10 80 0 0 14100 dirrmions 0 00 00 0 00 edfitf Lag 24 0 1 1 2 2 0 itrkfLgkflg sovkwfkwfkhw 6 4 rdardtf 45.8&44 rharef 1 11 [stsai I nscm 0.60 1.20 2.40 3.60 4.80 6.00 7.20 8.40 9.60 10.80 12.00 33 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 33 1.057 1.131 1 25 0.00 -0.10 .100 .2 .300 .40D 1 25 0.0 -100. 60. 22. SIGR 312001 5 1 0 312D01 .2 .4 .6 .8 1.0 0. .17U30 .21132 .25230 .29339 312001 .33458 Sji 31202 5 1 0 312002 .2 .4.6 .8 1.0 0. .0043117 .0043622 .0044160 .0044720 312DO2 .0045297 SIGR1 3120(1 5 1 0 3120W3 .2 .4 .6 .8 1.0 0. .oom50 .010767 .016sN9 .023234 312003 .029769 SiGTR2 312004 5 1 0 312004 .2 .4 .6 .8 1.0 0. .46027 .68728 .91722 1.14790 312004 1.3788 S1Iw2 3120O5 5 1 0 3120W5 .2 .4 .6 .8 1.0 0. .067142 .068668 .070216 .071771 3120X5 .073336 VB.2 312006 3 1 0 312006 .33478 .61987 1.0 0. 3.T10E.05 3.6944E-05 3.6396EgS aEL1 312007 3 1 0 312007 .33478 .61997 1.0 0. 1.5278p8.7 1.48399E07 1.4492E607 SIGTRi 312008 5 1 0 312008 .2 .4 .6 .8 1.0 0. .0E254 .11156 .13784 .16418 312008 .1060 siGAI 3129 5 1 0 312009 .2 .4 .6 .8 1.0 0. .000Q8M6 .00051302

.00054752 312009 .00058343

.00063)40 siGRI 312010 5 1 0 312010 .2 .4 .6 .8 1.0 0. 03B(222 .006797.01oo649

.014724 312010 .01M15 SiGTR2 312011 5 1 0 312011 .2 .4 .6 .8 1.0 0. .20390 .34945 .4960 .644a 312011 .79M Si(2 312012 5 1 0 312012 .2 .4 .6 .8 1.0 0. .0053686 .0063474 .00734c .0083368 312012 .0093342 VEL2 31201 3 1 0 312013 .33478 .61W 1.0 0. 3.7710E+05 3.69446-05 3.6396E+05 wiLi 312014 3 1 0 312014 .33478 .61987 1.0 0. 1.5Z78E.07 1.4839E+07 1.44926407 BLANK 550101 1 "4 1 0 2 2001 ( GRUF 1 TOTAL mom -REFLECTIOR1) 55(201 2 4 1 0 2 2002 ( GRoUP 1 AB/RPTlON

-REFLECTOR1) 55W1 3 4 1 0 2 2003 ( foup 1 DONMTTIt -REFLECITR1) 550401 6 4 1 0 2 2004 ( GOP 2 TOTAL RBOW -REFLECrOR1) 550501 7 4 1 0 2 2005 ( GUP 2 ABSWTIGN -REFLECTOR1) 550601 21 4 1 0 2 2006 ( GROUP 2 VEI.COITY

-IEFLECrCl1) 550701 20 4 1 0 2 2007 ( GUP 1 vELCCITY -REFLECrCR1) 1 4 1 0 2 2008 ( GUP 1 TOTAL MUM. -REFLEC102 550202 2 4 1 0 2 2009 ( GaRP 1 ABRTION -REFLECTW2) 55032 3 4 1 0 22010 (OUP 1 DilsCT -REFLEC ) 55040 6 4 1 022011 ( UP 2 TOTAL R -REFLECraR) 550502 7 4 1 0 2 2012 ( GSP 2

-REFLECTCR2) 550602 21 4 1 0 2 2013 ( GUP 2 VELOCITY -REFLECMT2) 550702 20 4 1 0 2 2D14 ( GUP 1 WI.OCITy -REFLECKN2) 99 Table 2.4.8-9 SLMTRAN Input File (u0cI2.simtramin)

Used in Generating UlC12 Cross-Section File /userVchaiko/SIHLA7hJu1c12.sinu1ate.restart 1 2 3 4 5 6 7 9999 UI102 AMS SINTRAN EC-ATWS45-Rev 7 1 1 1 0 6 jdisk, iblkrs,krout,krin 234567 1080 0 0 14100 dimmsions 0 0 0 0 0 0 0 0 ecdtftas 24 0 1 1 2 2 0 iter,kdfLag, k9fLag,ksoLv, kwf , kltf, kqhiwt 6 4 rdiJndtf 45.8244 rhoref 1 11 tstson / nrcm 0.60 1.20 2.40 3.60 4.80 6.00 7.20 8.40 9.60 10.80 12.00 33 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 33 1.057 1.131 1 25 0.00 -0.10 .100 .200 300 .400 125 0.0 -100. 60. 225. SIGTR1 312D01 5 1 0 312001 .2 .4 .6 .8 1.0 0. .17030 .2113 .25 .29339 312D01 .33458 SIAl 3120025 1 0 312002 .2 .4 .6 .8 1.0 0. .0043117 .0043622 .00441 .oW47 312D2 .0045297 SIR1 3120 5 1 0 3120M .2 .4 .6 .8 1.0 0. .00985o .010767 .01679 .o23234 31205 .029769 -.. SJGTR2 312004 5 1 0 3120M4 .2 .4 .6 .8 1.0 0. .46027 .68M .9122 1.14790 312004 1.3788 SIMR2 3120(5 5 1 0 312005 .2 .4 .6 .8 1.00. .067142 .06868 .070216 .071771 3120(5 .0732 VEL2 312006 3 1 0 312006 .33478 .61W 1.0 0. 3.7710E-15 3.6944E.04 3.6396E+5 VEL1 312007 3 1 0 312007 .3347B .61967 1.0 0. 1.5Z7Eio7 1.4492E#07 SIGTRI 312O(3 5 1 0 3120D3 .2 .4 .6 .8 1.0 0. .05254 .11156 .13784 .16418 3120D3 .19060 SIGht 312OO9 5 1 0 312009 .2 .4 .6 .8 1.0 0. .0008005 .00051302

.0054752 312009 .00058343

.0006200 SIGR1 312010 5 1 0 312010 .2 .4 .6 .8 1.0 0. .0(30222 .0067297 .010649 .014724 312010 .0115 SiGTR2 312011 5 1 0 312011 .2 .4 .6 .8 1.0 0. .2M30 .34945 .49690 .644a2 313)11 .79289 SIM2 312012 5 1 0 312012 .2 .4 .6 .8 1.0 0. .0053686 .0063474 .0073403 .0053368 312012 .0095342 VEL2 312013 3 1 0 312013 .3M478 .61987 1.0 0. 3.771E+05 3.69V&Eo5 3.63ME+5 Va.l 31201431 0 312D14 .33478 .61W7 1.0 0. 1.5M27 1.483*407 1.4492E#07 550101 1 4 1 0 2 2001 ( MUP 1 TOTAL. RBOA. -REFLECCRI1) 55W0 2 4 1 0 2 200 ( GQUIP 1

-REFLECrOcI) 550301 3 4 1 0 2 203 ( GRJP 1 DOWSCATT -REFLECIOR1) 55040 6 4 1 0 2 2004 ( GMP 2 T0TALR B 1 -REFLEC"R1) 550501 7 4 1 0 2 2005 ( GRJP 2 ABSORPTION

-REFLECTOCR) 550601 21 4 1 0 2 2006 ( GRP 2 VELOCITY -REFLECTORl) 5507oo 20 4 1 0 2 2007 ( RUP 1 VELOCITY -REFLECICRI) 550102 1 4 1 0 2 2008 ( GRP 1 TOTAL RBL -REFLECITR2) 55(2(2 2 4 1 0 2 2009 ( GROUP 1 ABSORTIOl

-REFLECT(32) 550302 3 4 1 0 2 2010 ( GOUP 1 DCISCATra

-IEFLECToR2) 5502 6 4 1 0 2 2011 ( GOP 2 TOTAL RDDOW -REFLECTOR2) 550502 7 4 1 0 2 2012 ( GUP 2 ABSORPTION

-IEFLECTMR2) 5506(2 21 4 1 0 2 2013 ( GaCLP 2 VELOCITY -REFLECTCR2) 550702 2D 4 1 0 2 2014 ( GROUP 1 VELOCITY -REFLECTCR) 99 rAUE 74 2.5 Recirculation Pump Model A simplified pump model is used in SABRE to maintain initial rated flow conditions and to simulate the subsequent flow coast-down which results from a recirculation pump trip or a pump runback. The pressure in the downcomer region at the jet pump throat is calculated from pO = p, + g-gpDCHth IDc dWj +. AP vPD / WJO (2.5-1) g g. d& where P°(t) = downcomer fluid pressure at jet pump throat (lbf/ft 2), P*(t) steam dome pressure (lbf/ft 2), PDC (t) downcomer fluid density (Ibm/Wf 3), Hth(t) height of downcomer fluid above jet pump throat (ft), IDC = downcomer fluid inertia. (ft-1), WI = jet pump flow rate = total core flow rate (Ibm/sec), and =p(t) rate of mechanical energy supplied to fluid by the recirculation pump (ft-lbpsec).

If a recirculation pump trip occurs, the pump power 1, (t) decays exponentially according to dQP ) (2.5-2) dt Under normal operation, or in the case of a single pump trip or a pump runback, the pump power response is approximated with a simple first-order lag function coupled with a proportional control model, RP & -0E W- ( W) (2.5-3) dt where Ep,(1) A (0) + G,, [W,., -Wj(t)]. (2.5-4) In (2.5-4), GQ is a gain factor, and W, is the target jet pump flow. W, for various recirculation pump operating conditions is supplied as part of the input data. The values specified in the base SABRE input deck (Section F.23) are listed in the following table. If there are no pump trips or runbacks, W., is equal to the initial total core flow.

1 pump tripped and 1 pump at normal operating 65 2 pumps run back to 30% speed -142 1 pump tripped and 1 pump at 30% speed 36 The values listed in the above table are generally known from plant data or from the operating cycle power flow map. That is, the'total core flow is considered known under the various recirculation pump operating conditions.

Hence, the purpose of the SABRE recirculation pump model is not to predict the core flow rate under various recirculation pump operating configurations, but rather to set up reactor flow conditions that are consistent with recirculation pump operation.

The recirculation pump model defined by (2.5-3) and (2.5-4) is a simple controller which forces the total core flow to the target value shown in the above table based on the recirculation pump operating conditions.

The gain factor in (2.5-4) is specified within the source code and was chosen so that the calculated total core flow rate gave good agreement with the values in the above table under the indicated recirculation pump operating conditions.

2.6 Main Steam Line Model The SABRE main steam line model computes the steam line pressure, the flow through the MSIVs, the steam flow to the main turbine, the steam flow through the bypass valves, and the steam extraction by the feedwater turbines.

The mass balance for the steam line is dt sV =_ r~b _ W -WFWT (2.6-1) where VsL = Steam line volume (ft 3)1 (see §F.26. 11), PsL = Density of steam in steam line (Lbm/ft 3), WWrw = Flow rate through MSIV (Lbm/sec), WTb = Steam flow rate to main turbine (Lbm/sec), W3,,. = Flow rate through turbine bypass valves (Lbm/sec), and WFwr = Steam extraction rate by feedwater turbines (Lbm/sec).

It is assumed that the steam density changes are isentropic, and that the steam behaves as an ideal gas so that 0- (P-J and '-- -(2.6-2) °S dt y Pa, where PsL = Steam line pressure (psia), Y = Cp!Cv = 1.26 for steam at conditions of interest, and SL = Reference density and pressure.

Substituting (2.6-2) into (2.6-1) leads to z " =.F 'kz W -W w -W -W,) (2.6-3) dt yV sz XI From Calc. SA-MAC-003 (EC-SJMU-0501), Rev. 0, p. 3, the steam line pressure decay rate (psi/sec) due to steam extraction by the feedwater turbines and leakage/thermal losses is I The steam line volume is decreased by the amount specified in §F.26.12 in the event of a turbine trip.

X-w = 2.21+1.8 2 4 9 WI, + 5.292li,2 (2.6-4) where the normalized feedwater flow rate WF. = (WFw / 13.3 Mlb/hr). Using this result in (2,6-3) leads to the following relation for the steam line pressure ;__PS )(w -urb i (2.6-5) The steam flow through the MSIV is calculated from the following momentum equation:

di 2pW Awv. where IsL = Steam line fluid inertia (fr 1) Wmm = Steam flow through the MSIV (Lbm/sec), PSD = Steam dome pressure (psia), Pwro = Density of steam at MSIV (Lbm/fte), and AmrV = Full Open MS1V flow area (ft) (see §F.26.14).

The loss coefficient K in (2.6-6) is expressed as K = f= KMw + K', (2.6-7) where KMsrv is the loss coefficient for the MSIV, and K' is chosen so that PsL is equivalent to the turbine inlet pressure.

That is, given the initial steam dome pressure (§F.8) and the initial turbine-inlet pressure (§F.26.5), the value of K' is back calculated so that the initial value of P. is equal to the initial turbine inlet pressure.

The loss coefficient multiplier

f. accounts for the increased flow resistance during closure of the MSIV. The multiplier

f. is a function of valve stem position and is specified as part of the input data (see §F.25.6).

Equations (2.6-5) and (2.6-6) make up the SABRE main steam line model. In order to solve (2.6-5) and (2.6-6), expressions for the turbine steam flow WT,, and turbine bypass flow WB need to be specified.

Expressions for these two variables are provided by the pressure regulator model. A block diagram of the pressure regulator model used in SABRE is presented in Figure 2.6-1. This model describes the essential features of the Susquehanna pressure regulator.

The SABRE model of the pressure regulator was developed from descriptions provided in Calc. EC-FUEL-0969, "SSES RETRAN Controller Model," and in the training manual for SSES Systems Course SY028. In Figure 2.6-1, The Lead-Lag block is defined by the differential

equation, T-t~-+y= {+Idc 2.6-8) where x is the input to the block, y is the output, and G is the gain. The Lag block is defined by "c +y = Gx. (2.6-9) As steam line pressure changes, the turbine steam flow and the bypass steam flow will change even if there is no change in turbine control valve position and turbine bypass valve position.

A decrease in steam line pressure will cause the turbine and bypass flow to decrease.

The opposite effect will occur if steam line pressure increases.

The actual change in flow due to a change in steam line pressure would require a detailed model of the flow resistances within the turbine and bypass system which is beyond the scope of this work. An approximate correction to the flow can be derived, however, by considering the turbine as a lumped resistance so that the turbine flow is governed by Pa , Pc K W7ýb ý .- 0 2 (144) g. p The steam density can be expressed as p. =const pJT. Substituting this relation into the above expression, neglecting the condenser pressure compared to the steam line pressure, and absorbing all constants into the loss factor leads to PI+I/7 SL (2.6-11) K The flow correction factor is obtained by dividing (2.6-11) by the initial turbine steam flow WLb, Flow Correction Factor (2.6-12) ..WoL PmoSL This pressure ratio is also used to modify the bypass flow.

(Turb. Irdet Press.) -(Press.Regulator Set PoinO Flow Correction Factor Accounting for Steam Une Pressure Change Flow Correction Factor Accounting for Steam Une Pressure Change Figure 2.6-1 Block diagram of pressure regulator model used in SABRE.

I _%r QVJ 2.7 Feedwater Model A proportional control model with a first-order lag is used in SABRE to regulate feedwater flow and maintain downcomer water level near its initial value. This is a simplified approximation to the actual feedwater control system. In MSIV-closure and turbine-trip transients, where feedwater heating is lost, the decrease in feedwater enthalpy is approximated as an exponential decay following a period where there is no loss in feedwater temperature.

The time constant rh (sec) defining the enthalpy decay is supplied as part of the code input data (see §F.24.9).

Following an MSIV closure, feedwater can continue to inject to the reactor until main steam line pressure decays to the point where the feedwater turbines can no longer generate sufficient power to overcome the vessel pressure.

The feedwater turbine design specification sheet indicates that feedwater flow to the vessel can be maintained with a steam line pressure of 175 psia. 1 The rate of steam line pressure decay following an MSIV closure is computed from Eq. (2.6-5).

In the previous versions of SABRE, the rate of main steam line pressure decay following MSIV closure was computed directly from the decay constant (2.6-4). This gave correct results for MSIV closure events in which the pressure regulator operated as designed, but in the PREGO (pressure regulator failure open -see §5.7) event, this led to feedwater being available for a longer period of time than it should have been. (In the PREGO event, steam continues to be extracted by the turbine and the turbine bypass system even after the MSIVs close.) As a result, suppression pool temperatures for the PREGO ATWS event were calculated to be higher than they should have been. Since the modeling error led to conservatively high suppression pool temperatures for the PREGO event, none of the ATWS analyses performed previously with SABRE are invalid. In the present version of SABRE, this modeling error is corrected, and following MSIV closure, the main steam line pressure is obtained from the solution of (2.6-5). With the corrected model, feedwater flow is lost almost immediately on low steam line pressure following MSIV closure in the PREGO ATWS event. Figures 2.7-1 and 2.7-2 show plant data for feedwater flow and feedwater temperature following an MSIV closure. The plant data indicates that there is no drop in temperature during the time that injection is taking place. Once injection stops, there is a very slow decay in coolant temperature presumably due to heat losses to the surroundings.

Figures 2.7-1 and 2.7-2 indicate that approximately 0.1 MLbm of feedwater is injected to the vessel, following initiation of MSIV closure, without a drop in feedwater temperature.

In the SABRE code, following an MSIV closure or turbine trip, the feedwater temperature-is held constant until 0.1 MLbm of coolant are injected to the vessel. The temperature of any additional coolant injected with the feedwater system is decayed away with a time constant "mh which is supplied as part of the SABRE code input data.GEK-38479 (1OM 42) 14 13 ,12 c 10 8,7, ,-J 07 LL .6. S4. U- 2.1 0 -Figure 2.7-1 MSIV Closure on High MSLR (Plant Data) 06/14/83 31-83070615023101 70 90 110 130 150 170 190 Time (sec) Figure 2.7-2 MSIV Closure on High MSLR (Plant Data) 06/14/83 31-83070615023101 383.4 " 383.2 E 383 I= 382.8 0 382.6. 382.4 U. < 382.2 382 PAGE 81 70 90 110 130 150 170 190 Time (sec)

2.8 Control

Rod Drive and Core Spray Models Coolant addition to the vessel by the Control Rod Drive (CRD) system is modeled as a mass and energy source in the bypass region. The mass and energy associated with the CRD flow is deposited in the first cell (bottom cell) of the bypass region. It is assumed that the CRD flow reaches thermal equilibrium with the fluid in the lower plenum before it reaches the bypass region. This has a slight cooling effect upon the fluid in the lower plenum. The Core Spray system is also modeled as a mass and energy source in the bottom cell of the upper plenum. Core Spray flow as a function of the pressure difference between the reactor and the suppression chamber atmosphere is specified as part of the input data. Core Spray suction is taken from the suppression pool. 2.9 HPCI and RCIC Models SABRE includes models which describe injection by the LPCI and RCIC systems. The injection rates for these systems can be specified as constants, or the code will calculate the injection rates if the systems are operated in level control mode where the desired RPV water level is specified as part of the SABRE input data. Pump affinity equations are used in conjunction with polynomial curve fits of the pump and turbine performance curves to calculate the HPCI and RCIC turbine steam flow rates as a function of pump injection rate and RPV pressure.

The enthalpy of the fluid discharged from the turbines (and deposited in the suppression pool) and the turbine speeds are also computed.

Details of the [PCI and RCIC system models are described in PP&L calculation SA MAC-002 (EC-052-0593), December, 1992.

2.10 Containment Model The SABRE primary containment model can describe the suppression chamber and drywell pressure/temperature response to an ATWS, a small break LOCA, or a normal plant transient.

SABRE models the drywell as a single control volume. The suppression chamber free volume is also described with a single temperature and pressure.

Pressures within the drywell and suppression chamber are computed from the ideal gas equation of state. Heat sinks within the containment consist of the suppression pool along with the containment liner plate and structural steel. Thermal capacitance of the concrete is neglected because the heat transfer to concrete is expected to be small during the relatively short duration events (ATWS, LOCA, or plant transient) of interest.

In any event, neglecting the thermal capacitance of the concrete is conservative since it leads to somewhat higher containment temperatures and pressures.

The components within the drywell and suppression chamber free volume consist of N 2 and water vapor. During an ATWS, vacuum breakers prevent the buildup of large differential pressures across the diaphragm slab-by allowing steam and N 2 to flow from the suppression chamber to the drywell. In a LOCA, steam discharged into the drywell is directed through downcomer vents and condensed in the suppression pool. The suppression pool temperature calculation accounts for mass and energy from steam discharged through the downcomer vents, SRVs, and HPCI & RCIC turbines.

Effects of the suppression pool cooling system and heat transfer between the containment air space and structural steel are also included in the model. Mass balance equations for water vapor and N 2 within the drywell free volume are given by dMD, =w dt = WevapDP + YWScIvm -W ..a'D + XWb,.k -Wgve, -W,=.Dw (2.10-1) and dM NDW dM = (I -YWSc)W. -*(2.10-2) dt Drywell average temperature is computed from the following energy balance which is taken from the COTTAP code: 1 1 Chaiko, MA., and Murphy, M.J., "COTTAP: A Computer Code for Simulation of Thermal Transients in Secondary Containments of Boiling Water Reactors," Nuclear Technology, Vol. 94, pp. 44-55, 1991.

4r---CDW dTDW = QRr2 -Qcooier -QsfeeI,DW

-[c DTWdMNW-4DW dMwgD dTddt +TDw R W +WRN ycrc + WJYyShC+4[S]

dm, -w h(r'w) Tat -_ [c,, L, + Wi,.g -W DWhcond DW -rc LpwTww,,nI

-hDW Wgw,ie, + XWbr,,,hw 2 (PDW) (2.10-3) where CDW is defined by C P h M9 ,DW_ _ Mg wRw -M~vwR CDW [ dT ] N, DWTDW + [c' NDW MNDW +MDW ,DW W NDW N (2.0-DW (2.10-4)In these equations the following variable definitions are used:[CP]Nw [CP]N,Sc [dC/dT]fDw hC 0 d,DW h (PDW) hwg,DW hsD hWDP MW9DW= Specific heat of N2 in the drywell (Btu/Lbm-°R), = Specific heat of N 2 in the suppression chamber (Btu/Lbm.°R), = Derivative of N 2 specific heat with respect to temperature in drywell (Btu/Lbm-°R-OR), = Enthalpy of condensate on drywell structures (Btu/Lbm), = Enthalpy of saturated steam at drywell pressure (BtuALbm), = Enthalpy of water vapor within suppression chamber atmosphere (Btu/Lbm), Enthalpy of water vapor in drywell atmosphere (Btu/Lbm), = Enthalpy of water vapor at surface of DW pool (Btu/Lbm), = Enthalpy of saturated water at drywell atmosphere temperature (Btu/Lbm)

Mass of water vapor in drywell atmosphere (Lbm), Mass of N 2 in drywell atmosphere (Lbm), M N,Dw 9 a YNVDW Mass fraction of N 2 in drywell atmosphere, YNsc -Mass fraction of N 2 in suppression chamber atmosphere (Lbm), YW'sc Mass fraction of water vapor in suppression chamber atmosphere, W. mass flow rate from suppression chamber to drywell through vacuum breakers (Lbmlsec), Wai,.DW rainout rate in drywell atmosphere (Lbm/sec), RW Ideal gas constant for water vapor = 0.1104 Btu/Lbm-°R, RN -Ideal gas constant for nitrogen = 0. 0710 Btu/Lbm-°R, TDW Drywell atmosphere temperature

('R), W..dDw Steam condensation rate in drywell (Lbm/sec), W.,Výw Evaporation rate from water accumulated on drywell floor (Ibm/sec), WN,vt -Flow rate of nitrogen through downcomer vent pipes (lbm/sec), W. Flow rate from suppression chamber to drywell through vacuum breakers (Lbm/sec), W, Flow rate of water vapor through downcomer vent pipes (bin/sec), Qv-2 Drywell heat load due to dissipation from reactor vessel (Btu/sec), Q.111 Energy removed by drywell coolers (Btu/sec), Qselw = Heat transfer from drywell air to steel structures in drywell (Btu/sec), and V Fraction of break flow which flashes to steam.

SABRE computes the relative humidity in the drywell from the relation.RHDI = (100) Pj(TD) (2.10-4a) where RI-DW = Relative humidity in drywell (%), P',DV = partial pressure of water vapor in the drywell atmosphere, and P,, (Tow) = saturation pressure of water at the drywell temperature.

If the relative humidity exceeds 100%, wet steam is beginning to form in the drywell atmosphere.

When this occurs, the code removes the liquid from the atmosphere by means of a "rainout" model. The rainout rate within the drywell atmosphere is computed from the simple control model 10.0 if RHDW < 100 e 2O(RffD W-1 00) iV RH DW > 100 The rainout rate computed from (2.10-4b) is very sensitive to the relative humidity; this prevents the RHDw from exceeding 100% by any appreciable amount. Water that is removed from the atmosphere is added to the drywell pool. The same model is used to prevent relative humidity from exceeding 100% in the wetwell atmosphere.

When the pressure differential between the drywell and suppression chamber is large enough to clear the downcomer vents, the flow rate, W,,t (Ibm/sec), through the vents is calculated from =~e (D "ýSC A'Pp)2ýgjl 4ý4)p (2.10-5)where drywell pressure (psia), P = suppression chamber pressure (psia), APO = differential pressure (psia) required to displace water column from vent line, g. = 32.2 ft-lbm/lbf-sec 2 , PDW gas density in drywell (lbm/ftl), At= total flow area of downcomer vents (f1 2), and t= loss coefficient for downcomer vent. In order to stabilize the numerical calculation, the vent flow area A,. is linearly increased over a small pressure range (2 psi): A~t = 0 if PDW-Psc-APopen

< 0 (2.10-6) A -A if 0:P pDw-Psc-APo, <2 psi (2.10-7) A,vjt = ALnt if PDW-PSC-APOPen>

2 psi (2.10-8) In Eqs. (2.10-7) and (2.10-8), AL, is the total downcomer vent flow area when the vents are completely cleared of water. If the suppression chamber pressure exceeds the drywell pressure by AP,,. (see §F.63.3 for specification of this parameter), vacuum breakers, which are connected to 5 of the downcomer vents, will open and allow the flow of steam and nitrogen to the drywell.

The loss coefficient K, for the flow path from the suppression chamber to the drywell is specified as part of the SABRE input data (see §F.63.2).

This loss coefficient is based on the flow area of the 24" downcomer pipe. In SABRE, the flow rate through the vacuum breakers (Ibm/sec) is computed from 0 if PSC'PDW < APoM (2.10-9) V, =PD ',I44),p if PSC-PDW->

APOP, (2.10-10)

= ~g(P K PDW The vacuum breakers are full open when the differential pressure reaches APFO (see §F.63.4).

Therefore, the vacuum breaker flow area A, (ft 2) is calculated from if PscPDow < Apope= PSC -PD W -1PP IAPFO -Apope if APO < PSC-PDW < APFo if PSC-PDW> APFO (2.10-11)

(2.10-12)

(2.10-13)The total full-open flow area, AL (fR 2), associated with flow through the vacuum breakers is 5 times the flow area of a 24" vent pipe or 10.25 W (see §F.63.1).

The mass and energy balance equations for the suppression chamber atmosphere are d~wgscW _ d t = W ,sc dMNY+ dt = YNS cW. + WN,,nt '(2.10-14)

(2.10-15)and dT- =Q7P + Q,. -Q.,,lx -[CP L, cTsc dMNd -dMwg+ + LdMPCc dMdctt T Rw + RN d -1]c WcYLcNxTsc

-wyg.. + dt-dt -TC JXvW.Y Wh -Wx + WN,. [cpL, Tv -WC 4' (T.1) (2.10-16)where -= dC sc L dT JN'sc MN' T p+ L, ]N + MNg+M dhg,sc R wsc T+IdTsc WMSCRw -MSCR. (2.10-17)Variables not previously defined consist of Specific heat of N 2 in the suppression chamber atmosphere (Btu/Lbm-°R),[CPINSC

= Specific heat of N 2 evaluated at the suppression pool temperature (Btu/Lbm-°R), [dCWdT]NSC

= Derivative of N 2 specific heat with respect to temperature in suppression chamber (Btu/Lbm.OR.OR), = Enthalpy of water vapor at surface of suppression pool (Btu/Lbm), h4c Enthalpy of water vapor within the suppression chamber atmosphere (Btu/Lbm), h~d=sc Enthalpy of condensate on suppression chamber structures (Btu/Lbm), MNSC = Mass of N 2 in suppression chamber atmosphere (Lbm), M= Mass of water vapor in suppression chamber atmosphere (Lbm), QP-= Rate of heat transfer from suppression pool to suppression chamber atmosphere (Btu/sec), Rate of heat transfer from suppression chamber atmosphere to steel structures (Btu/sec), QTP Rate of heat transfer from SRV tail pipes to suppression chamber atmosphere (Btu/sec), T= Suppression pool temperature (OR), Tsc Suppression chamber atmosphere temperature (0 R), and = Evaporation rate from suppression pool (lbm/sec).

In the drywell, the structural steel is divided into 4 groups: ceiling, wall, floor, and internal structures.

The ceiling, wall, and floor are all parts of the drywell liner. The drywell liner is divided into these three parts because the heat transfer coefficients between the atmosphere and the steel structures depends on the direction of heat flow. Heat transfer between the drywell atmosphere and the drywell heat sinks includes natural convection, radiation, and steam condensation effects. Forced convection heat transfer may be significant under conditions where drywell coolers are operating because of the induced circulation caused by the cooler fans. However, in most accident sequences of interest, the drywell coolers are trippedbecause of a LOCA signal (see §F.58 & F.59). Therefore forced convection heat transfer is neglected.

Even if the drywell coolers are operable under some conditions, it is acceptable to neglect forced convection heat transfer because it will lead to conservatively high drywell temperatures.

The thermal response of steel structures is described using a lumped-parameter model. Heat transfer within the suppression chamber is modeled in the same manner as in the drywell, except that there are no steel structures corresponding to a ceiling and a floor. In place of heat transfer to a floor, the suppression chamber atmosphere exchanges mass and energy with the suppression pool surface. There is no liner plate on the ceiling of the suppression chamber.

Natural convection and radiation heat transfer correlations for the drywell and suppression chamber atmospheres are taken from the COTTAP code. 2 The Uchida correlation is used to model condensation heat transfer when the steel temperature drops below the dew point of the atmosphere (this is also the correlation used in the COTTAP code). The mass transfer rate between the suppression pool and the suppression chamber atmosphere is calculated from the following expression 3 S(2.10-18) where A = Surface area of suppression pool (ft 2), Coefficient of mass transfer for water vapor at suppression pool surface (fi/sec), pWI = density of water vapor at pool surface (lbd/ft), P = average pressure of nitrogen in convective boundary layer (psia), Pwsu= Partial pressure of water vapor at the surface of the suppression pool (Lb /.ft2), and P = Partial pressure of water vapor in suppression chamber atmosphere (L bf /ft2). 2 Chaiko, M.A., and Murphy, MJ., "COTTAP: A Computer Code for Simulation of Thermal Transients in Secondary Containments of Boiling Water Reactors," Nuclear Technology, Vol 94, pp. 44-55, 1991. 3 Collier, J.G., Convective Boiling and Condensation, 2nd Edition, p. 324, McGraw-Hill, New York, 1972.

The mass transfer coefficient in (2.10-18) is calculated from the mass-transfer/heat transfer analogy (Collier, p. 326) hc.,sp (2.10-19) where hc.,SP Natural convection heat transfer coefficient at pool surface (Btu/f?-sec-oF), = Density of water vapor at pool surface (Lbl/f), and P [CP , = Specific heat of water vapor at pool surface In (2.10-19), it has been assumed that the Lewis number raised to 2/3 power is close to 1 Which is generally the case.4 In calculating the heat transfer between the suppression pool and the suppression chamber atmosphere, both natural convection and radiation effects are considered.

The heat transfer coefficient for a horizontal slab is used. The rate of heat transfer from the pool to the suppression chamber atmosphere is calculated from Qa.. = hp,,(Ts, -Tsc (2.10-20)

If a large amount of energy is rapidly added to the suppression pool, then the vapor pressure of the pool can reach the total pressure of the suppression chamber atmosphere (see benchmark study in §5.2). This situation corresponds to the onset of suppression pool boiling. Under pool boiling conditions, the pressure and temperature of the suppression chamber atmosphere are no longer unknown quantities.

The atmosphere pressure is equal to the vapor pressure of the pool, and the atmosphere temperature should quickly approach the pool temperature.

Therefore, rather than using complicated correlations to predict the pool heat transfer and mass transfer coefficients under boiling conditions, the two coefficients are treated as parameters and are adjusted so that the heat and mass transfer rates are sufficient to maintain pressure and temperature equilibrium within the suppression chamber.

The drywell heat load from the reactor vessel Oa, and the cooling load of the drywell coolers Q., are calculated from the following expressions:

4 McQuiston, F.C., Parker, J.D., Heating Ventilating, and Air Conditioning Analysis and Design, 4th Edition, pp. 547-550, John Wiley, New York, 1994.

a 6 Qx 2 = (hA)R2 [TR% -TD] (2.10-21) and Q.1 = (u), [TD -T 1,] (2.10-22) where Tw is the reactor vessel temperature, which is calculated from Eq. (2.11-1) and T.. is the cooling water temperature for the drywell coolers. (hA), 2 is a constant which is calculated from the initial reactor heat load and the initial drywell and reactor coolant temperatures. (UA)L, is calculated from the initial drywell cooling load, the initial drywell temperature and the cooling water temperature; this coefficient is also maintained constant throughout the entire transient.

Values of (hA)R, and (UA), 1 are calculated in Sections D.18.5 and D.18.6. SABRE solves a mass and energy balance for the suppression pool in order to obtain an estimate for the pool temperature during a transient simulation.

Steam discharged-through the SRVs and steam exhausted from the HPCI and RCIC turbines are sources of mass and energy. Under LOCA conditions, water (liquid and vapor) can also be added to the pool by means of flow through the downcomer pipes. Liquid will flow through the downcomers when the depth of the water layer on the drywell floor reaches 18". The model used in SABRE considers energy extraction by the suppression pool cooling (SPC) system, but neglects suppression pool energy losses to the surrounding walls. Mass extraction from the pool consists of HPCI, RCIC, and Core Spray suction as well as suppression pool let down. RCIC is initially aligned to the CST, but transfers to the suppression pool on low CST level. The mass and energy balance equations for the suppression pooi are given by d W = r + W9 + W c + + + W. _ WL" dt R PI CC -WRC -WCs -w v -wD + w. ] (2.10-23) and t hgw MvIIS dt = hsDWSRV + + + -hS (W I, + W RCC + W S + W.)+ h14DWW,. + hwDPWW,., +hcondscW,,sc

-s-Wevh + W ,sc a(T -d)h W td (2.10-24) where MI mass of water in suppression pool (Lbm), h,. -specific enthalpy of saturated vapor at suppression pool temperature (Btu/Lbm), -specific enthalpy of suppression pool water (Btu/lbm), S = Condensation rate on suppression chamber steel structures (lbm/sec), WM -flow rate of steam through SRVs (Ibm/see), WAa flow rate of steam through HPCI turbine (Ibm/see), -B HPCI injection rate if -PCI suction is aligned to the suppression pool (Ibm/sec), WRc RCIC injection rate if RCIC suction is aligned to the suppression pool (lbm/sec), Wcs = Core spray injection rate (lbm/sec), W= suppression pool letdown flow (Lbm/sec), WRac -flow rate of steam through RCIC turbine (Ibm/sec), W~w,,, -flow rate of steam through downcomer vents (Ibm/sec), flow rate of liquid water through downcomer vents (ibm/sec), enthalpy of water in drywell pool (Btu/lbm), h = volume-weighted enthalpy in reactor steam dome region (Btu/lbm), h~ = enthalpy of steam discharged from HPCI turbine (Btu/lbm), hjget = enthalpy of steam discharged from RCIC turbine (Btu/lbm), QP = rate of heat loss from surface of pool (Btu/sec), and QSlC rate of energy loss from suppression pool due to SPC system (Btu/sec).

The heat removal rate by the suppression pool cooling system is calculated from Osc = NL,,,, 3 15(Ts,-TP (2.10-25) where NLOOP = number of SPC loops in operation (0, 1, or 2), Tsp = suppression pool temperature (OF), and Tsw = service water temperature (IF). The service water temperature is specified as part of the input data (§F.62). The suppression pool temperature is computed as the saturation temperature of liquid water with enthalpy h4,. The factor, 315, in Equation (2.10-25) was determined from the containment heat removal capability of 134 x 106 Btu/hr per loop for 90'F service water and 208'F pool temperature (Table 6.2-2 of SSES FSAR). During a LOCA event, water will accumulate on the drywell floor. When level reaches 18", liquid will overflow to the suppression pool through the downcomer vents. The governing equations for the drywell pool level and enthalpy are similar to (2.10-23) and (2.10-24).

PAGE 94 2.11 Heat Capacitance of Vessel and Internals In a reactor depressurization transient, such as an ADS blowdown, the vessel and internal structures will transfer heat to the coolant. These steel structures will effectively behave as a heat source. Vaporization occurring on the surface of the submerged parts of these structures will reduce the rate of vessel depressurization and increase the suppression pool temperature rise associated with the blowdown.

In SABRE, heat dissipation by the reactor vessel and internal steel structures is described with simple lumped-parameter models. For the reactor vessel, the governing equation is MiC,< dTa T (2.11-1) di where x = mass of reactor vessel (Ibm), CP = specific heat of steel (Btu/lbm-°F), TR = reactor metal temperature (OF), hR 1 = coefficient for heat transfer between coolant and reactor vessel lR. = coefficient for heat transfer between reactor vessel and drywell atmosphere (Btu./sec-j2-°OF), A = surface area of reactor vessel (ft 2), T = saturation temperature of water at reactor pressure (CF), and TDW = drywell atmosphere temperature (OF). Heat transfer between the vessel internals and the coolant is described by d( (2.11-2) where Mv = mass of vessel internals (excluding the fuel) (Lbm), Tv = temperature of vessel internals (OF), hv = coefficient for heat transfer between vessel internals and reactor coolant (Btu/sec-ff-°F), and PAGE 95 A = heat transfer area for vessel internals (ft 2). The initial temperatures, T,(0) and T40(), are specified as the coolant saturation temperature, TIa. The heat transfer coefficients h_, and hw depend on the reactor water level. Calculation of the heat transfer coefficients hl 1 and hvl is discussed in Appendix D, §D.2. The rate of heat transfer from the steel structures to the reactor coolant is given by =:pAp T -T,(2.11-3) and QI = hn A (TU -T) (2.11-4) In order to simplify the heat transfer model the heat transfer rate, Qjt + Qv is added to the decay heat generation rate within the core [see Eq. (2.4.7-4)].

Although this is a very simplistic model, it does balance the steel sensible heat loss from the vessel and vessel internal structures with vapor generation during an RPV depressurization event so that the suppression pool temperature response during an RPV blowdown will be modeled correctly.

3. NUMERICAL SOLUTION OF THERMAL-HYDRAULIC EQUATIONS

3.1 Control

Volume Formulation of Flow Equations In this section, the partial diferential equations governing the reactor thermal-hydraulics are spatially integrated to obtain conservation equations in control-volume form. A diagram of a nodalized flow region is shown in Figure 3.1-1.Junctionjj+1 Junction j Junctionj-1 zj +A zj = z j+l z J j iz j Figure 3.1-1 Nodalization Scheme fbr Flow Equations.

3.1.1 Nodalized

Continufty Equaftih For the jet pump, lower plenum, core, bypass, upper plenum, riser, and separator regions, the governing thermal-hydraulic equations are integrated as follows. Each of the terms in the continuity Equation (2.1-1) are integrated over the control volume length Az. Integration of the first two terms in (2.1-1) is carried out as follows: *'f -ap(z'td = d Z J Z dp= ) (3.1.1-1) a t _p(z,4= dt-J (3.1.1-2)-,J+A z 1 d' =~ ,')- G (z j, t) = ,)- t)_ C (,) = z Zi-and where t j+JAZ Zj PAGE 97 a (t)-G(z, t), and 6 1+, t G(z 1 , = G(z, + Az, 1) Integrating the right-hand-side of Equation (2-1) gives zj ,4z zj4& A=A~. (3.1.1-3)

Zj In Equation (3.1.1-3), rF (t) (Lb/sec) is the mass source within control volume j, and A is the flow area of the control volume which is assumed to be uniform throughout the entire flow region. Combining Equations (3.1.1-1), (3.1.1-2), and (3.1.1-3) gives the continuity equation for control volume j: dt 6,() ,()+r t jE C{ 2,..., N (3.1.1-4)

A where N is the number of control volumes within the flow region. Substituting the equation of state, P 1 0t P O'W 01 where Zj into Equation (3.1.1-4), leads to the following form of the nodal continuity equation:

SdP* d ,P" " (3.1.1.) For certain flow regions (lower reflector region and first node in upper plenum; see Figure 2-1) it is necessary to generalize the nodal continuity equation to allow for the situation where there are multiple flow streams into or out of the control volume. The generalized continuity equation is given by PAGE 98 pI. a_ _ dP PI-ej+(1-cj+,)z2If, +Pg j+, ij. j+1 A dt (3.1.1-7) where j e C,2,...,N and Mj number of flow paths on control volume boundary defined by z = zj (z is the axial coordinate for the flow region), and Mj+1 = number of flow paths on the control volume boundary defined by Z=zj+1 =zj +Az. 3.1.2 Nodalized Energy Equation Each term in the energy equation (2.1-2) is integrated over the control volume length Az in order to obtain the control-volume energy balance. For the first term the result is a [z=. .t)=hz, f)(zdj.(z, t)dz =Az p,(t) & +____t (3.1.2-1)

Derivation of the error term in Equation (3.1.2-1) is given in Appendix C. Integration of the second, third, and fourth terms leads to dz a [p,(_-a)u, h. +p.auhu ] a z(3.12-2) d" Az1(3.1. -3) Jad Jdt' and 1 dz q'(zt)-A-A fdz q(z, (t)_A (3.1.2-4) where Qj(t) is the heat generation rate (Btu/sec) within control volume j. Integration of the heat-flux and mass-source terms results in J dz Ah AJt :- q;(t) (3.1.2-5)

Zj+A where q(t) is the control-volume average heat flux (Btu/sec ft 2), and zj+L%2 hoyt rj ) fdz (z, t) h, (t): ='A(3.1.2-6)

A Zj The following control-volume energy balance is obtained by substituting the results in (3.1.2-1)

-(3.1.2-6) and (3.1.1-5) into Equation (2.1-2): J-h .-----I---~j+ L 'p dt A A A (3.1.2-7)

As in the case of the continuity equation, the energy balance can be generalized to allow for multiple flow paths into and out of the control volume. The generalized nodal energy equation is given by AZ j+ Jzh -h ~ ji [Pjk (I -4i j) i~ j + 0 9gJ C~ 9j [01j+ -(I y 1+lu)i +1 An j+1 + 1,9 j+1 d 1 j 1 ig j+1 hg j+1 + p=1 F aPj1dP* +.Qi AZPh h Az[ -j -qj +AzP =i Sap dt A A ' A (3.1.2-8)

PAGE 100 3.1.3 Nodalized Momentum Equations In this section, each term in the momentum equation (2.1-3) is integrated over the flow region length L in order to arrive at the overall momentum equation for the flow region. 3.1.3.1 Nodal Representation of Inertial Term Since the length of the flow region is constant, L iG(z, t) d L fa dGz t) f G(z,t) (3.1.3.1-1) 0 0 Integration of the continuity equation (2-1) from z=O to z=z leads to z a G ~ ,t -= 0( O t) +" & -y ( t-- " f d a Substitution of (3.1.3.1-2) into (3.1.3.1-1) gives LL L z Ll idz G(z, t) = d[Got+dj t(,t)-Jdzd ap4,) S bt s to th d i i ( Applying integration by parts to the double integrals in (3.1.3.1-3) gives L z L 0dz da yo ,t)=fdz(L-zoy(z,t) 0 0 0 (3.1.3.1-3)

(3.1.3.1-4) and L-5)L ,L dp 0 0 0 Substitution of (3.1.3.1-4) and (3.1.3.1-5) into (3.1.3.1-3) leads to fdaG(z,t)

= G(O,t) fdz (L-z)y(z,t)-_dz (L- z) ap(z,t) (3.131 0 at & 0& We now want to develop a control-volume representation for the right-hand-side of Equation (3.1.3.1-6).

Each of the terms on the RHS of (3.1.3.1-6) are evaluated below: 1-6)

PAGE 101 Nodal Approximation to 1st Term on RHS of (3.1.3.1-6)

L G(O,t) = L 61 (t). [Note that L=O(1)] Nodal Approximation to 2nd Term on RHS of (3.1.3.1-6)

LNN Ld I¢ If+A N ..j Z =(L _ -zh (Zt) =L zy(Z't) 0 j1- Z,_ Zi (3.1.3.1-7)

(3.1.3.1-8) where t, E (z + +/- Az). The mean value theorem 1 has been used in obtaining the result in Equation (3.1.3.1-8).

Expanding ) about z; (the inlet of cell j) gives Since Ej E (Zzz+ ) and zj = O(Az), Equation (3.1.3.1-9) can be rewritten as L = L-zj + O(Az). (3.1.3.1-10)

Substitution of (3.1.3.1-10) into (3.1.3.1-8) leads to L t I N { f ( ]t ) + O(AZ2) f (3.1. +(z -.1-11)fzyz~

= N L1

[Note that N =O(Az-I I j._lL 3) Nodal Approximation to 3rd term on RHS of (3.1.3.1-6)

The following result is obtained by replacing the integral over the flow region with a summation integrals over the individual nodes: L L -N "y+Az dz(Lo -z z(L z) P(zt)=d dz(L-z)p(zjt)

(311.3.1-12)

Kom, GA., and Kom, T.M., Mathematical Handbook for Scientists and Engineers, 2nd Edition, pp. 4.7-1 & 4.7-2, McGraw-Hill, NY, 1968.

From the definition of the nodal density pj(1), and from the mean-value theorem, zj+tlZ where tj E(zj~zj +Az).For the integrals in Equation (3-27), p (z, t) is expanded about the point z = ý, to obtain P(z,x)= tt+ pzt (3.1.3.1-14)

With the use of Equation (3.1.3.1-13), (3.1.3.1-14) can be expressed as p(z,,)= p(tj,t)+O(Az)=

pj(t)+O(Az) where z c(zj, z+l) and j=1,...,N.

(3.1.3.1-15)

Substituting (3.1.3.1-15) into (3.1.3.1-12) gives the following result: zLIzap (zt) d N "JdZL-) , :yJ .- ZA(Lz)[p,(,)+

o(AZ)] D a t dt J_1 Xj =N {[dpj(t)].

ZJ+4A -z) o(2 N ZJ+t.Z ={(L-z)dz}+O(Az)

"= ,l'j AZN [dp 1 (t)] [L z ] ,&z2 + 0f(A&z) 77-1 2 "N Fs '(t)_ =Az'ý Fap j A=7-0 [VL Lzj+OAh ) N ~~O~t)ap 1&ii()(3.1.3.1-16)

(3.1.3.1-13)(z-ý,)+...wherez ZE(zj,zj+,)

I,' IIJ, Substituting (3.1.3.1-7), (3.1.3.1-11), and (3.1.3.1-16) into (3.1.3.1-6) gives the following nodal approximation to f dz , t." 0o at 0 at -d +(t ~mv7 where I = Fluid Inertia , (3.1.3.1-18) and (3.1.3.1-19)

The mass sources considered in SABRE consist of the CRD flow which is deposited in the bottom cell of the bypass channel, borated water which is deposited in the top cell of the lower plenum, and Core Spray flow which is added to the bottom cell of the upper plenum region (feedwater, HPCI, and RCIC injection are included in the steam dome/downcomer model). 3.1.3.2 Nodal Representation of Convective Term The integral of the convective term in Equation (2.1-3) can be evaluated in terms of variables defined on the region boundaries:

j [01xpu+ Pg U [ ) g ~ 12b,-k-is~

2-dz j (-tp t o .u] (_7)5u2 +07i 1' uL, -[(l-j)O, u2 -.< 0 f12. u 0 z (3.1.3.2-1) 3.1.3.3 Nodal Representation for Spatial Integral of Pressure Gradient The integral of the pressure gradient is evaluated in terms of the channel pressure drop, Ldz g. aP(zýt) = & [P(L ,t)- P(Ot)] = g& K.0+,-)-f(t)I (3.1.3.3-1) f rn.5. .n 3.1.3.4 Nodal Representation of Gravitational term The total elevation head is evaluated in terms of the nodal densities as follows: L N IL N Jdz p(z,t) g =g&~4 Z!Jdz p(z,t)= g Azr N )+(A (3.1.3.4-1) 1 AZ 0 0 0 j1 3.1.3.5 Nodal Representation of Wall and Spacer Friction Terms The wall friction integral is replaced by integrals over individual nodes: _ G(z, t) I G(zt0)1p(P(Z) f N Zj+b G(zt)1G(zt)1 Ip(z,t) (3.1.3.5-1) 2Dh 0 P-zt) 2 h ._(1 z Following the approach used to derive Equation -(3.1.3.1-15), we can obtain the following relations: (p(z,t) =pj(t)+O(Az) where z E (z/,zj+1) and j=l,...,N, (3.1.3.5-2) and p,(z,t)=p, 1 j(t)+O(Az) where zcE(z,+,)

and j=1,...,N (3.1.3.5-3)

The mass flux G(zt), where z c (zj,zj+1), can be approximated by the average of the flow on the control volume boundaries:

G(z, t) = 2j (t)+ +;, (t)]+ O(Az2) for z E (zj,zj,).

(3.1.3.5-4)

Combining Equations (3.1.3.5-1), (3.1.3.5-2), (3.1.3.5-3), and (3.1.3.5-4) leads to the following result: f G(zt)L(zt)p(zt)

Azfw N K(t)+G+jI(t)] (t)+GQ+1 (to Pi(t) (z fw dz -~z t)[~,t p(,t , + O(z 2Dh Rep 1 (Zt) 8Dh j..1 j 0 (3.1.3.5-5)

Similarly, the fuel-spacer friction term in Equation (2.1-3) can be integrated over the channel length to obtain fS dG(z,t)rY(z,tj(D(z,t)

Azfs , Dh 0 p(z,t) 8Dh p(t) (3.1-3.5-6) 3.1.3.6 Nodal Representation of Channel Inlet and Outlet Friction Terns From the definition of the Delta function, K 4 1(zi) ( (z- o) K, G(o, 1) P(o, 1ý 0(o,t) fpd( , , -2 p ,(o ,t) 20K 2,, P1 (0, t,, 2 0 (3.1.3.6-1)

Similarly, for the channel outlet friction term we have K 2 jL G(z, t) IG(z, t) 1I(z, t) 6 (z -L) K 2 G(L, t) K;(L, t, c(L, t) 2- pt(,t) =2 p (L,t) _K2 (v+,(t)fv+, (tI (+1 (t) (3.1.3.6-2) 2 P N+1,(t) 3.1.3.7 Overall Momentum Equations Combining the results in (3.1.3.1-17), (3.1.3.2-1), (3.1.3.3-1), (3.1.3.4-1), (3.1.3.5-5), (3.1.3.5-6), (3.1.3.6-1), and (3.1.3.6-2) yields the overall momentum equation for the flow channel, I d+/-) + M(l)g~()P~~)

31371 where F(t) is defined by (3.1.3.1-19), and M(t) is given by.-- Iwo

• - "V.J V m0=111-0Q)]

Mt~)r1;(t)+/-ciQ)

^g(t) ^90 Lj + ~ [~, (t) + 2(I)+ IQ(t)+ Q+' ý(t~~()j , p(t) AfN KQ)+/-+ ^(t)] PJQt).+-6~

1+Q D i (t) + ,DE j+ 8Dh kiPti(t) K 1+ (t)NGN+0 + K (tyi (tIi + K 2 (v+ I (() +/-0(Az). 2 W:Q 2 II0 (3.1.3.7-2)

An overall momentum equation for the entire circulation path, from the jet pump inlet to the separator outlet, is obtained by summing the regional momentum equations (Equation 3.1.3.7-1) for the jet pump, lower plenum, mixing region, core, upper plenum, riser, and separator regions: + M,, = g, (P -)+() (3.1.3.7-3) where FL JIkJ and MLF =Mk (3.1.3.7-4) k-B k*B The subscript k is used here to identify the various flow regions. k can take the following values J, L, C, B, U, R, or S which refer to the jet pump, lower plenum, core, bypass, upper plenum, riser, and separator regions, respectively.

In (3.1.3.7-3), PO is the pressure within the downcomer region at the elevation of the jet pump throat, and P* is the steam dome pressure.

In formulating the momentum equation for the natural circulation loop, the inertia associated with the upper and lower reflectors are neglected.

An additional momentum equation, independent of the unknown regional boundary pressures, can be obtained by subtracting the individual momentum equations for the core and bypass regions. The resultant expression is d[IcF & (t)- IEFE (t)J = M,(t)-Mc (t)+ O(Az) (3.1.3.7-5)

'ft PAGE 107 Equations (3.1.3.7-3) and (3.1.3.7-5) are the momentum relations which are solved in SABRE to obtain the reactor flow behavior.

3.1.4 Nodalized

Fuel and Cladding Heat Balance Equations The regions occupied by fuel and cladding are also divided into control volumes of length Az. Integration of the heat balance equations (2.3-1), (2.3-2) and (2.3-3) over the control volume length of the core Az, leads to a set of ordinary differential equations for the fuel and cladding nodal temperatures ToJ, T 1 j and T,, respectively: ( t-f dT 1 2 kf(7> ~T -j)s 2 )j r..2 , j r+rba + r qj (3.1.4-1) frý.2b~ di __ (p(Cp,)j(r -T J) F (rb1-r) I___ (rn 41rbk , 2r.H .4rok,,j_ (3.1.4-2) and (p~ r _~~ r (Tj,., -Qd (Ta.,. -Tf- , -r-- E =(rb-,re)

+ 1 (r~rci F.rl-.,.'

m 4 rkk, + 2.dH-- 4-,k. .. (3.1.4-3) where q,. [ Qf-Q)W SjW + Qd ( Si(O) j(o-fq_ -fq 2), q7, j () =, q(I), qj() = (2-c, ) q-(t),

Vf,j = volume of fuel in axial control volume j of the active core region (ft), fql = fraction of total core power deposited in active core region as direct moderator heating, fq 2 = fraction of total core power deposited in bypass region as direct moderator heating, Nc = number of control volumes in active core region, Qf (t) = total core fission power (Btu/sec), Qd (t) =total decay heat generation rate (Btu/sec), Sj (t) = axial power shape function = (power in node j)/(average nodal power), and Cs = fraction of pin power generated in inner radial fuel node. Note that the subscript j defines the axial node within the active fuel region (j=l, 2, ..., 25). All other variables have been defined previously (see §2.3). Calculation of the heat flux to the coolant within the reactor core requires determination of the clad surface temperature.

The surface temperatures are calculated from (see 2.3-4) r, ) cwf.-j (T-, dT -__1.-T.. ) di + ( -r] 2r ] S 4 rbkf zj 2 ~H' 2r,,k,.j L O -...J (3-1-4-4)

...3.2 Calculation of the Flow Regime In Section 3.1.3, two spatially-integrated momentum equations were derived (Equations 3.1.3.7 3 and 3.1.3.7-5).

These equations define two flow rates within the reactor circulation path, namely the jet pump inlet flow rate GQ and the core inlet flow rate Gc. Once these two flow rates are specified, the flow behavior within the entire loop is determined by continuity and energy constraints in conjunction with the flow-regime model, i.e., the Drift-Flux relation (2.1-7). The present section describes the calculational method used to determine the local reactor flow rates given the two inlet flows, Gn and ci. Given the jet pump inlet flow, the outlet flow conditions for the jet pump region are computed from the nodal continuity and energy equations along with the Drift-Flux relation.

Once the jet pump outlet flow is computed it defines the inlet flow to the lower plenum region. The same computation is then carried out for the lower plenum region: namely, the continuity, energy, and Drift-Flux equations are used to compute the lower-plenum outlet flow. Thus, the scheme provides a method of computing the flows sequentially as it "marches" through the circulation loop. In order to compute the flows at the outlet of the lower reflector where the flow branches into the active core and the core bypass, additional information is required; this is where the value of GC;1 is used. Details of the computational scheme will now be given. First, use of the Drift-Flux relation (2.1-7) will be discussed.

At the j'th junction between two control volumes, this relation becomes Similarly, the mass flux atjunctionj is given by (see Equation 2.1-6) Gj= 1o (I -dj) f)J + 15+ 6 U^ gJ (3.2-2) If the value of the junction mass flux is specified, then Equations (3.1-1) and (3.2-2) can be solved for the vapor and liquid velocities.

Values of oi, C., V.j , 1gfj, and O5j are determined from the adjacent volumetric values. The particular volumetric value used, i.e., the upwind or downwind value, depends on the direction of the phasic velocities.

This is discussed in more detail below. For determining junction variables from volumetric quantities, it is computationally convenient to introduce two limiting values of the coolant mass flux, Guj and Gfj.. If Gj > Gfj, then the flow regime is co-current upward. If Gj then the flow is co-current downward.

In order to obtain defining expressions for (fj and Equations (3.2-1) and (3.2-2) are solved for the vapor and liquid velocities, a 1 9 kgj 60'1+ti-a6, , j 6-.j )_From the definition of CO [see Lahey & Moody, The Thermal-Hydraulics of a Boiling Water Nuclear Reactor (1976), p. 206], it can be seen that the quantity (1-aC 0) is positive or zero. Then from (3.2-4), the flow is co-current upward (z<j > 0) if the numerator in (3.2-4) is greater than zero. This leads to the following flow-regime criterion expressed in terms of the known volumetric quantities:

j f -j- -1 Pg-i- Cojli I-a 1.1 C 0 ,-1 1 (Co-current upward flow).Similarly, the flow is co-current downward if the numerator in (3.2-3) is negative, 5j < Vj.j ,W P C0.J (Co-current downward flow).(3.2-6)Note the upwind/downwind scheme used in determining these flow-regime limits. It follows that counter-current flow is defined by (Counter-current flow).(3.2-7)In the counter-current flow regime, junction variables are calculated as a weighted average of the adjacent volumetric values. Observe that when G = Gj, the liquid velocity is zero because this is the dividing point between positive and negative liquid flow. Also, when Gj = Gtj, the gas velocity is zero since this is the dividing point between positive and negative gas flow. Therefore, 1gj =0 if Gj = G j iV' J 2 if (3.2-8) (3.2-9)and I/ /(3.2-3)(3.2-4)(3.2-5)and where Gljj < Gj < G~j I-L;-A I W.-UOU* PAGE 111 wg = mass flow rate of vapor at junction j (Lbm/sec), and 4 = flow area ofjunction j (ft 2). Given the junction mass flux, Equations (3.2-8) and (3.2-9) define the mass flow of vapor at the boundaries of the counter-current flow regime. In SABRE, the vapor flow rate within the counter-current flow regime is computed from linear interpolation, and the liquid flow rate is calculated from conservation of mass. That is Ag A; for Gj< <G (3.2-10) and w G 1 A 1-1 2 vg 1 j for (iu,< dj <( 6ft" (3.2-11) Table 3.2-1 summarizes the calculational scheme used to obtain junction variables from the known control volume quantities.

With the above results, description of the computational scheme for the local flow rates within the circulation loop can now be completed.

In the calculation of local flows, Qn, %CJ, and dP*/dt are given (from momentum conservation and steam-dome/downcomer conservation equations).

All of the remaining junction flows are then computed from continuity and energy constraints along with the Drift-Flux relation.

The calculation scheme begins at the jet pump inlet. Since Gd, is specified, the junction variables at the inlet boundary are computed according to Table 3.2-1. The flow of mass and energy across the jet pump inlet is now defined. In order to calculate the flow of mass and energy across the outlet boundary, Equations (3.1.1-6) and (3.1.2-7) are solved for the temporal pressure derivative dP*/dt, dP* d 2 a, -d 1 a 2 1 (3.2-12) dt a 1 a2 -a 2 1 a 1 2 where apj all= AzAP* ap, a 2 l &1 + /i. ~ ~F -

d-d -t- +iv, +Fj, and dvg, hg, tj -j~ l hj , h fj+, +QJ +Az ph q'h, Ff. Since ýv,, jl, h, j, and L are all functions of the outlet boundary mass flux (as described in Table 3.2-1), the only unknown on the RHS of Equation 3-63 is Gj,-. As stated earlier, dP*/dt is known at this stage of the calculation (calculation of dP*/dt, G41, and Gce is discussed in the next Section).

The known value of the pressure derivative is denoted as (dP'*/dt)0.Subtracting (dP*/dt)o from (3.2-12) results in the following algebraic equation for Gj÷,, d 2 a 1 1 -d, a 2 1 --(3.2-13) a,, a22-a 2 1 a, 2 a Newton's method is used to solve Equation (3.2-13) for the outlet flow, GJ+1 , of node j. A bisection procedure is used in the event the Newton's iteration fails to converge.

Once Equation (3.2-13) is solved for the jet pump region, the flow of mass and energy across the inlet boundary of the lower plenum region is determined.

The solution process is then continued in a step-wise manner: equation (3.2-13) is solved for the lower plenum region to obtain the flow of mass and energy across the inlet boundary of the lower reflector region. At the outlet of the lower reflector region, the flow branches into the core and bypass channels.

Here the generalized continuity and energy equations (3.1.1-7)

& (3.1.2-8) are solved. Since the inlet flow to the core Gc, is specified, the convective terms in (3.1.1-7)

& (3.1.2-8) associated with mass and energy transport across the core inlet can be evaluated according to the scheme summarized in Table 3.2-1. The flow of mass and energy across the bypass inlet boundary can then be determined by solving the generalized form of (3.2-13).

Local flows along the core and bypass channels are then computed sequentially through solution of (3.2-13).

For the first node of the upper plenum, the generalized mass and energy balance equations, (3.1.1-7)

& (3.1.2-8), are solved because there are two flow streams at the inlet of the upper plenum. For the riser and separator regions. there is only one flow stream on the inlet and outlet boundaries, therefore (3.1.1-6) and (3.1.2-7) are used to obtain the flow of mass and energy across the region boundaries.

rrnji. , , IV ri I 1f 0 eQ5--", 5-. + 5-.St L 30Vd g) -o g) SQ ,0 5-. 5-. 5-. + 5-p)9) 5-. SQ 5-. 5 5-.I p) S.0 5-. 5-.p) 5-. -o 5-. e SQ 5-. 5-.0 0 0 0 a a a 0 a 'r1 0 0 It

3.3 Calculation

of GJ 1 ,Gc,,anddP*/dt In §3.2, calculation of the nodal boundary flows assumed that the jet pump inlet flow G, 1 , core-inlet flow Gk, 1 , and the rate of change of the system pressure dP*/dt are all given. This section discusses the calculation of these three quantities. 1 and Gc.1 are determined from solution of the spatially-integrated momentum equations (3.1.3.7-3) and (3.1.3.7-5).

Temporal integration of these two equations is accomplished using the implicit Euler's method. The finite-difference approximations for Equations (3.1.3.7-3) and (3.1.3.7-5) are FL,, (t)- FL..P (t -At)+ At {MP (t)- g.O (t)- P* (t)] }= 0 (3.3-1) and [e FC (t)- 'BFB(t)]-

[VC F (t- FB(t -A&)]-At [VB(t)-M O(1)1=0. (3.3-2) In (3.3-1) and (3.3-2), FLoP (t), MLoP (t), and [Ic FC (t)- I, FB(t)] are functions of all the nodal flows throughout the reactor system as well as the rate of change of the system pressure dP*/dt. However, as discussed in Section 3.2, all of the local flows are determined, through continuity and energy constraints, once the jet-pump-inlet flow, core inlet flow, and rate of change of system pressure are specified. (At this stage of the calculation process, the fluid enthalpy, density, and system pressure are all known because the equations governing these variables are integrated explicitly.

This is discussed further in the next section.)

Consequently, there are three basic unknowns, G;l, elI and dPP d/, associated with the two relations defimed by (3.1-1) and (3.3-2). The additional constraint required for closure of the system is provided by the downcomer and steam dome mass and energy relations.

Solution of the full system of equations is accomplished as described below: 1. An initial estimate of G, 1 and Q 1 is provided by the values of these variables at the previous time step. 2. dP*/dt is computed from Equations (2.2-1)-(2.2-4).

In calculating dP*/dt, the value of Gs, 2 (mass flux at the separator outlet) from the previous time step is used. 3. The nodal flowrates within the reactor are computed by the scheme described in §3.2. 4. Newton's method is used to obtain an improved estimate of G_,, and Gca, (a nested bisection procedure is used in the event the Newton's iteration fails to converge).

5. Steps 2-4 are repeated until convergence is obtained.

.Initialization and Temporal Integration Procedures All transient simulations are initiated at or near normal reactor operating conditions.

In the initialization process, the total reactor power, total core flow (active-core flow plus core-bypass flow), reactor pressure, downcomer water level, and downcomer subcooling are specified.

A SIMTRAN output file provides cross-section data for the 1 -D kinetics solution.

The code then calculates the neutron flux distribution, delayed-neutron precursors, and the axial power shape by solving the eigenvalue problem associated with the steady, two-group diffusion equations (see §2.4.3). It is important that the base cross-section set in the SIMTRAN output (Cross section set I of 7 described in §2.4.8) correspond as close as possible to the initial conditions specified in the SABRE input deck as this will effect the accuracy of the initial flux and power distribution.

This is due to the fact that the moderator density and fuel temperature perturbations Au 1 and A Tfi in Eq. (2.4.1-1), and in the analogous expressions for the other neutronic parameters, are all assumed to be zero at the initial reactor state. Once the initial power distribution is determined, the initial local flow conditions are computed along with the fuel and cladding temperatures and the pump power required to maintain the specified core flow. It is implicitly assumed that the reactor is critical at the initial specified operating condition.

Since the steady-state equations are solved sequentially to obtain first the initial neutron flux distribution and then the initial flow and fuel temperature distributions, it is difficult to obtain an initial reactor state which is completely absent of small initial disturbances.

Therefore, once the initial power and flow conditions are computed through sequential solution of the steady equations, a second level initialization process is initiated to remove any small perturbations which may otherwise appear at the start of the transient simulation.

This second level initialization process consists of running the code in the transient mode starting at a negative time (-40 seconds) and then integrating up to t=-O. The solution results for t<O are not printed to any of the SABRE output files so that this process is transparent to the code user. During this phase of the initialization process, small adjustments are made to the value of k.0- in Eqs. (2.4.4.10) and (2.4.4-12), and to the pressure regulator setpoint (see Figure 2.6-1), so that when the governing equations are integrated through negative time and reach zero, the total power and reactor pressure are as close as possible to the values specified in the SABRE input file. Typically, the adjustment to k, is on the order of a few ink. Printing to the SABRE output files begins when The transient kinetics equations are implicitly integrated as described in §2.4.4. All of the governing thermal-hydraulic equations are integrated explicitly except for the two momentum equations (3.1.3.7-3) and (3.1.3.7-5), and the equations governing the cladding surface temperature (3.1.4-4).

The explicit temporal integration is carried out with Eulers method. An outline of the overall numerical solution procedure is summarized below: 3.4

1. Using the thermal-hydraulic time step size specified as part of the SABRE input data, hkj, PP 0 , Wm VSJhD,hs, &) D , TsI2Tj,,, d,, Ci, Qd, and all containment parameters are obtained by explicit temporal integration.

2. Vapor and liquid-phase density and enthalpy are computed according to (2.1-8) to (2.1-10).

3. Coolant void fraction and mixture density are calculated from (2.1-4) and (2.1-5) since pt, p g,h, heg, and /i are known at this point. ap/6h and ap/.aP are then computed for each control volume. 4. Clad surface temperature is calculated by integrating (3.1.4-4) using the implicit Euler's method. The heat flux to the coolant is then computed.

In carrying out the implicit integration of (3.1.4-4), an initial estimate for Tf.j is obtained from the result of the previous step. 5. The two overall momentum equations (3.1.3.7-3) and (3.1.3.7-5) are integrated implicitly (see §3.3). The temporal integration involves the solution of two coupled transcendental equations

[(3.3-1) and (3.3-2)] for the jet pump and core inlet flows. These two equations are solved using a Newton's iteration scheme. Each iteration involves calculation of all the nodal flows in the reactor system. The local flows are computed according to the procedure outlined in §3.2. When convergence of the jet pump and core inlet flows is obtained, the thermal-hydraulic time step is complete.

6. After completion of a thermal-hydraulic time step, the LSODES ODE solver is called, and the 1I-D kinetics equations are integrated to obtain a solution at the time corresponding to the current hydraulic solution.

Once the kinetics solution is obtained at the new time level, the process begins again at Step 1.

4. DISCUSSION OF CODE LIMITATIONS In this section, the major SABRE modeling deficiencies are identified, and the effects of these code limitations on calculational results are discussed.

4.1 Single-Channel Core Model SABRE models the 764 fuel channels contained within the SSES reactor core as a single average-power channel with a flow area equal to the combined area of the individual channels.

This modeling approach causes the code to over-predict the period of nuclear-coupled density wave oscillations.

With a single-channel model, any core flow oscillations are accompanied by oscillations of similar magnitude in the rest of the flow path inside the shroud. Thus, inertial and frictional effects in the entire flow path from the jet pump inlet to the separator outlet have a strong influence on the oscillation dynamics.

In the case of a multiple core-channel model, however, not all of the individual core channels oscillate in-phase.

This out-of-phase mode allows significant flow oscillations in the core channels with relatively small flow oscillations in the regions external to the-core.

since the-fluid inertia is smaller in this latter case, -and -the influence of fluid friction is more spatially concentrated, a shorter oscillation period is anticipated.

A single core channel model will always over-predict the power/flow oscillation period because it is simulates all core channels oscillating in-phase.

4.2 Steam

Line Model and Feedwater Controller The SABRE code contains a single-volume steam line model; thus, the details of acoustic wave propagation within the steam lines is neglected.

In addition, SABRE employs a proportional control model to simulate the effect of the feedwater control system on the reactor dynamic behavior.

Although this model accounts for the basic function of the controller, which is to maintain downcomer water level in the neighborhood of the set point value, it neglects dynamics associated with the various time constants which characterize controller operation.

4.3 Separator

Model The separator model used in SABRE effects complete vapor/liquid separation at the exit of the separator region (see Figure 2-1) which is above normal water level. In actuality, the steam separators discharge liquid through a path in the side of the separator at a location below normal water level. Thus, at normal water level conditions the SABRE model under-predicts the flow path pressure drop for the liquid exiting the separator since it does not account for the hydrodynamic head associated with the liquid above the exit of the liquid discharge path. The SABRE code also neglects carry-over and carry-under effects associated with non-ideal separator operation.

4.4 Downcomer

Water Level Generally, SABRE has numerical difficulties if downcomer mixture level drops below the jet pump throat.

5. CODE BENCHMARKING 5.1 MSIV Closure with Scram--Comparison to Plant Data In this Section, SABRE results are compared to plant data for an MSIV closure event which occurred at Susquehanna Unit I on 6/14/83 at 0947. The scram was the result of Main Steam Line High Radiation levels which occurred because of operational problems with the condensate demineralizers.

The GETARS data for this event are logged on the General Office computer system on Tape #31; the case number is 83070615023101.

Data is extracted from PP&L mainframe computer storage by typing 'cndas50 extract' in a TSO session and then following the prompts.

A summary of event-specific code input parameters used in the SABRE calculation is given in Table 5.1-1. The sequence of events calculated by SABRE is presented in Table 5.1-2. This SABRE calculation corresponds to Case 01 in the Computer Case Summary.

Comparison of SABRE results with plant data is given in Figures 5.1-1 and 5.1-2. Figure 5.1-1 shows the fission power decay as a result of the scram. With a scram time of 2.0 seconds, the SABRE-calculated core power decay shows good agreement with the plant data. Scram time in SABRE was reduced from 2.8 sec to 2.0 sec because this scram occurred early in the initial reactor fuel cycle and the neutron kinetics data used in the SABRE simulation of this case corresponds to an EOC condition.

Unit I began commercial operation on 6-8-83, and the scram occurred on 6-14-83. At BOC some control rods are already inserted and therefore the core would shutdown considerably faster during a scram than it would at EOC. Also, at beginning of cycle, the core power is bottom peaked, and therefore, the core is more easily shutdown during a scram. This transient was simulated using a U2C7 EOC core model with the code initialized at pre-power uprate conditions.

The difference in UICI and U2C7 neutronic response is accounted for by reducing the scram time in the SABRE input from 2.8 sec to 2.0 sec to get reasonable agreement between the predicted power response and plant data (see Table 5.1-1). The total core flow response due to the scram and the recirculation pump runback is also displayed in Figure 5.1-1. A recirculation pump runback is initiated when level drops to +13" which leads to a slow coast down of the core flow. SABRE shows an earlier peak in core flow than the plant data. The minimum water level predicted by SABRE (Figure 5.1-1) shows excellent agreement with the plant data (less than 1" difference) although the SABRE calculated water level tends to drop somewhat slower than the actual level. Plant data for steam dome pressure response (Figure 5.1-1) shows an initial drop in pressure of about 25 psi caused by the rapid decrease in core void fraction associated with the scram. The SABRE calculation also shows an initial drop in reactor pressure, but the magnitude of the drop is larger. Following the initial drop in pressure, there is a rapid pressure increase due to MSIV closure. The pressure rise which is terminated before any SRVs open is halted by the onset of steam condensation as level drops below the feedwater nozzles [-24" (see Section D.5)], and subcooled feedwater is injected directly into a region occupied by saturated steam. When the sparger nozzles are re-submerged (at about 13 seconds) as level recovers, pressure begins a second increase.

The SABRE model indicates, however, that this pressure rise is also modulated by steam condensation effects. In this situation the steam condensation is a result of subcooled water exiting the steam separators.

Because of the high core flow rate and low power generation (decay heat) there is no boiling within the core, and subcooled liquid fills the entire region within the shroud. Although the details of the pressure trace calculated using SABRE differ somewhat with the plant data, the overall response over the first two minutes of the transient shows good agreement.

Figure 5.1-2 presents plots of total steam flow, feedwater flow, HPCI pump flow, and HPCI turbine speed. Note that the quasi-static HPCI model cannot predict the initial spikes in HPCI flow and turbine speed exhibited by the plant data; however, the predicted response of the system does show good overall agreement with the trend of the data. In this calculation, a value of -37" (rather than -38") was used for the HPCI and RCIC initiation setpoints.

The slightly higher value was used because SABRE slightly underpredicts the drop in water level due to the scram, and HPCIIRCIC would otherwise not have been initiated.

Figure 5.1-3 presents plots of the normalized thermal neutron flux during the scram. The flux is shown in the active core region (upper and lower reflector regions excluded).

The change in flux during the scram is consistent with the expected response.

Table 5.1-1 Changes Made to Base 9x9 Input Deck in Appendix F End time (F.2.1) 120 seconds Time steps (F.3) Max = Min = 5 msec for t<1 0 sec Max= Min = 30 msec for t>10 sec Initial power (F.6.1) 3293 MWth (pre-power uprate) Rated core power (F.6.2) 3293 MWth (pre-power uprate) Initial steam dome pressure (F.8) 1005 psia (to agree with plant data) Scram on specified time (F.19.6) 0.0 sec (scram on high MSL rad.) Scram time (F.19.10) 2.0 sec (decreased from 2.8 sec to 2.0 sec to account for fact that plant transient occurred at BOL and kinetics data corresponds to EOC). Scram time was adjusted until SABRE power decay matched reasonably close to data. HPCI initiation on low level (F.20.2) -37 inches RCIC initiation on low level (F.21.2) -37 inches Maximum feedwater flow (F.24.7) 14.0 Mlb/hr (to agree with plant data) Initiate MSIV closure on specified time 0.0 seconds (MSIV closure on high MSL rad.) F.25.1) Initial turbine inlet pressure (F.26.5) 950 psia (Dome pressure -55 psi)1'NEDC-32161P, "Power Uprate Engineering Report for Susquehanna Steam Electric Station Units I and 2," p. A.7-3.

IRA.Table 5.1-2 Sequence of Events Calculated by SABRE *** Kinetics file is /dOO/appL/sabre3vO/data/u2c7.simtran.out

• • • SABRE data file is /home/eamac/sabre_31/input/ec-atws-0505/cOl.dat

• • • This is not a restart case 1 S A B R E -Version 3.1 (01) MSIV Closure with scram -Compare to plant data t(sec)= .000 Low-Pres Condensate Injection Inop. t(sec)= .000 MSIV closure on specified time t(sec)= .000 Scram initiated on specified time Scram time (sec) = 2.00 t(sec)= 2.005 AlL Control Rods Inserted t(sec)= 3.915 Level Setpoint Setdown Setdown occurs when Level drops to Delay for setpoint setdown = .11E+02 t(sec)= 3.915 Recirc purp-A runback on Low Rx Lvt Setpoint for Runback = 13.00 in. Trip delay = .300E+01 sec t(sec)= 3.915 Recirc pump-B runback on Low Rx lvi Setpoint for Runback = 13.00 in. Trip delay = .300E+01 sec t(sec)= 4.005 NSIVs are closed t(sec)= 8.405 HPCI initiation on Low water Level Setpoint for initiation

-37.00 in. t(sec)= 8.405 RCIC initiation on Low water level Setpoint for initiation

= -37.00 in. t(sec)= 48.003 Main Turb Trip on high water Level Setpoint(inches)

54.000 t(sec)= 48.003 ROIC Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 48.003 HPCI Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 62.793 Feedwater Trip on high level Trip Setpoint = .54E+02 in.13.00 in. sec 90 -e----- Plantr SABRI 80 70 50 40 30 20 10 0 2 4 Time (sec).2 i LL V C C'ata 6 8 10 Tkie (sec) Time (sec)Figure 5.1-1 Comparison of SABRE results against plant data for MSIV-closure transient with scram. (SABRE Case 01)I. IL U 0 U I 0 I-Time (sec)8 6 4)ate

-r'114ý / ce -.2 U. E C' I I E. .2 0._.2 U. SL Time (sec)E EL) V Time (sec)Time (sec)40 60 Time (sac)Figure 5.1-2 Comparison of SABRE results against plant data for MSIV-closure transient with scram. (SABRE Case 01) 25 2C X 1. I-Ir X LL Axial Power Factor x .2 IL I-.IL 1 Axial Power Factor X I-.Axial Power Factor 1 Axial Power Factor 2 X .2 U.25 20 i-151 101 5 0 Axial Power Factor t-1.75 sec 1 P Power Factor Figure 5.1-3 Change In normalized thermal neutron flux during scram. Control rods are fully Inserted at t=2 seconds. (SABRE Case 01)X IL

5.2 Containment

Response for Unmitigated ATWS-Comparison to CONTAIN Code This section presents calculation results for an MSIV-closure ATWS with no boron injection or manual control rod insertion.

Makeup flow is provided by full IPCI and RCIC injection.

HIPCI suction is aligned to the CST for the entire event. The ATWS is initiated from 100% uprated power (3441 Mwt) and 100 M1bJhr total core flow. The core kinetics data correspond to U2C1O. Changes made to the base SABRE input deck are documented in Table 5.2-1. The sequence of events for the SABRE calculation are given in Table 5.2-2. A complete listing of SABRE output is provided on microfiche (see Computer Case Summary).

This SABRE calculation corresponds to SABRE Case 02 in the Computer Case Summary.

Figure 5.2-1 shows the reactor response calculated by SABRE. Plots of the containment variables are presented in Figures 5.2-2 & 5.2-3. The SABRE containment calculations are compared against results obtained with the CONTAIN code 1.The CONTAIN input and output files for this case are given on microfiche (see CONTAIN Case 02 in Computer Case Summary).

The base CONTAIN input deck is given in Appendix H. Source tables within the CONTAIN input file account for the steam addition to the suppression pool from SRV discharge as well as the heat addition/removal associated with the reactor vessel and the drywell coolers. These source tables are quite lengthy, and therefore they are not included in the base CONTAIN input deck which is presented in Appendix H (only abbreviated tables are included in the base CONTAIN input deck). In Figure 5.2-2, the rise in drywell temperature at about 800 seconds is due to loss of drywell cooling. The rapid rise in wetwell temperature and pressure at about 1000 seconds is caused by the onset of suppression pool boiling. Once the suppression pool begins to boil, the containment pressure follows the vapor pressure of the suppression pool. 'Murata, K. K., Carroll, D. E., Washington, K. E., Gelbard, F., Valdez, G. D., Williams, D. C., and Bergeron, K. D., "Ulser's Manual for CONTAIN 1.1: A Computer Code for Severe Nuclear Reactor Accident Containment Analysis", NUREG/CR-5026, November 1989.

Table 5.2-1 Changes Made to Base SABRE lOD Input Deck in Appendix G Table 5.2-2 Sequence of Events Calculated by SABRE

• Kinetics file is /dO0/appL/sabre3vO/datafu2cl0.

simtran.out SSABRE data file is /home/eamac/sabre_31/input/ec-atws-O5O5/cO2.dat

• • • This is not a restart case S $.A B R E -Version 3.1 (02) MSIVC ATWS -- NO SLC/MRI/SPC (U2C1O) t(sec)= .000 Scram is Failed t(sec)= .000 Low-Pres Condensate Injection mhop. t(sec)= .000 ARI is Failed t(sec)= .000 NSIV closure on specified time t(sec)= 4.005 MSIVs are cLosed t(sec)= 4.345 Recirc pump-A trip on hi Rx press. Setpoint for trip = 1135.00 psig Trip delay = .230E+00 sec. t(sec)= 4.345 Recirc pump-B trip on hi Rx press. Setpoint for trip = 1135.00 psig Trip delay = .230E+00 sec. t(sec)= 110.953 Feedwater Trip on Low Stm Line Press FLow stops when press < 175.00 psia t(sec)= 126.313 Level Setpoint Setdown Setdown occurs when level drops to 13.00 in. Delay for setpoint setdown .11E+02 sec t(sec)= 160.723 HPCI initiation on Low water Level Setpoint for initiation

-38.00 in. t(sec)= 160.723 RCIC initiation on low water Level Setpoint for initiation

-38.00 in. t(sec)= 884.533 DW Cooler Trip on Hi DW Press Trip Setpoint 1.720 psig I d T I S Sr 0 0 S C0 a. 0 2 S a. S 2. 0 E 2 Sn Time (see)I, L U,. cc Time (sec)Figure 5.2-1 SABRE calculation of reactor response for unmitigated ATWS evenL Time (see) Time (sec)

-714ý f 5 .0 S--e--- CONTAIN SABRE "225 30 a &200

• E =175 10' 1 150" 125 -10)5 000150 2 0 1000 ..t , , , , I , , , , ' .. 050 1D 150 2000500 1000 1500 2000 Time (sec) Time (sec) 60 300 275 50 -250 SABRE0S 40 -225 ,E C200 2 175 E 20- M 150, 10 125 Iioo 0 -e---CONTAIN 075 SABRE -0 500 1000 150 2000 500o 1000 1500 2000 Time (sac) Time (sac) Figure 5.2-2 Comparison of SABRE and CO .NTAIN predictions for unmitigated MSIVC ATWS. (SABRE Case 02) 300 275 0 2 %..0 o a E 6 .0 C 0 0. CL CL C*250 225 200 175 150 125 100 75 50.0 500 1000 1500 2000 Time (sec) Figure 5.2-3 Comparison of SABRE and CONTAIN suppression pool temperature calculations for unmitigated MSIVC ATWS.

5.3 MSIV Closure ATWS-Comparison to PP&L RETRAN Calculations.

SABRE results for an MSIV-closure ATWS, with U2C9 kinetics data, are compared with results obtained with the PP&L RETRAN model. The RETRAN simulation is documented in Calc. NFE-2-09-003 "Unit 2 Cycle 9 Nuclear Fuels Engineering ATWS Analysis." Table 5.3-1 lists the changes made to the base SABRE model. The changes in Table 5.3-1 were made to achieve agreement with the modeling assumptions used in the RETRAN calculation.

The sequence of events generated by SABRE is given in Table 5.3-2. This SABRE calculation corresponds to SABRE Case 03 in the Computer Case Summary.

Figure 5.3-1 shows a comparison of the SABRE and RETRAN calculated core power response to closure of the MSIVs. RETRAN calculates a peak power of 341% (of rated power) at 4.4 seconds, and SABRE predicts a peak power of 421% at 4.2 seconds. The SABRE calculation shows a secondary power spike of 2091/6 which is not present in the RETRAN calculation.

The secondary power spike is most likely caused by the surge in core flow predicted by SABRE following closure of the MSIVs (see Figure 5.3-2). In the SABRE model, the fluid density changes with the overall reactor pressure rather than with the local fluid pressure.

The RETRAN model on the other hand contains a fully-compressible flow model, and therefore, some difference in the core flow response to a rapid pressure change is expected.

Figure 5.3-2 indicates that the total mass of steam exiting the vessel in the 30 second simulation time is somewhat larger for the SABRE calculation than it is for the RETRAN calculation.

The larger mass of steam produced is consistent with the higher and more broad power spike predicted by SABRE for the transient.

The peak dome pressure calculated by SABRE is 1354 psia which occurs at 11.1 seconds.

RETRAN predicts a peak pressure of 1316 psia at 7.8 seconds. This comparison shows that SABRE is conservative with regard to predicting peak vessel pressure in the MSIV closure ATWS. Included in Figure 5.3-1 is a comparison of core average heat flux and Wide Range Indicated Level. Feedwater flow is maintained constant during the transient.

SABRE underpredicts the peak core-average heat flux compared to RETRAN, but the heat flux spike obtained from SABRE is much broader than that calculated by RETRAN.

PAGE 133 Table 5.3-1 Changes Made to Base SABRE 9x9 Input Deck in Appendix F 1 Calc. EC-ATWS-1O01.

Table 5.3-2 Sequence of Events Calculated by SABRE *** Kinetics file is /dOO/appl/sabre3vOIdata/u2c 9.simtran.out

• • • SABRE data file is /home/eamac/sabre_31/inpit/ec-atws-O505/cO3.dat

• This is not a restart case I S A B R E -Version 3.1 (03) SSES Power Uprate MSIV-CLosure ATWS U2C9 -Compare to RETRAN t(sec)= .000 Scram is Failed t(sec)= .000 Low-Pres Condensate Injection mnop. t(sec)= .000 ARI is Failed t(sec)= .000 NSIV closure on specified time t(sec)= .000 Feedwater Flow determined from Table Data in Table ends at t = .IOE+10 sec t(sec)= 4.005 MSIVs are closed t(sec)= 4.890 Recirc pump-A trip on hi RX press. Setpoint for trip = 1200.00 psig Trip delay = .230E+O0 sec. t(sec)= 4.890 Recirc pump-B trip on hi RX press. Setpoint for trip = 1200.00 psig Trip delay = .230E+00 sec. t(sec)= 19.925 Feedwater Trip on high Level Trip Setpoint = .54E+02 in. t(sec)= 21.325 Main Turb Trip on high water Level Setpoint(inches)

= 54.000

.1250 300. S1200 C~ 201150. SE 11oo0 2 7 1110 1050 --e-- RETRAN SABRE 0 5 10 1 20 25 30 0 5 10 15 20 25 Time (sec) Time (sec) 200 80. I S70' --e--- RETRAN 5SABRE 1500 .2 125 so.1" 40 10D 7 -S 13 25 10 00 10 15 20 25 30 0 5 10 15 20 25 30 Time (see) Time (sec) Figure 5.3-1 Comparison of SABRE and PP&L RETRAN predictions for U2C9 MSIV Closure ATWS event with one SRV out of service. (SABRE Case 03)

.f l3 47 2 VL E W a 'I) ap oa...,

• l. ...I. .....i 5 0 5" -a-R- RETRAN 0 SABRE -5;-T0 5 1 1 Time (sec)30 Figure 5.3-2 Comparison of SABRE and PP&L RETRAN predictions for U2C9 MSIV Closure ATWS event with one SRV out of service. (SABRE Case 03)II L 0 Time (see)20 2 PAGt -141 5.4 Inventory Boildown with Scram Failure-Comparison to SIMULATE-E As part of the U2C9 cycle-specific ATWS analysis for Susquehanna (Calculation EC-ATWS 1001), a data base was developed for core power as a function of core channel flow, reactor pressure, and core-inlet subcooling.

This data base was developed using the 3-D core simulator code SIMULATE-E.

The FORTRAN program listed in Appendix M of Rev. 6 of EC-ATWS 0505 interpolates within the data base to obtain core power as a function of the three operating parameters mentioned above. Under quasi-static conditions, SABRE should compute a critical core power consistent with that predicted by SIMULATE-E.

The purpose of this bencimiarking study is to verify that the 1-D kinetics calculation in SABRE for core power agrees with SIMULATE-E under quasi-static conditions.

In this scenario, the scram and ARI systems are failed; recirculation pumps and feedwater are tripped at t-O. Water level beigns to fall rapidly as inventory is boiled off. RCIC is allowed to initiate on low reactor level in order to slow the rate of level decrease and thus maintain quasi static conditions.

The MSIVs remain open throughout the transient.

The SABRE nuetronics model for this transient corresponds to U2C9 which is a mixed core of 9x9 and 10xlO fuel with the 9x9 fuel being the dominant fuel type. Changes made to the base 9x9 SABRE input deck are summarized in Table 5.4-1. This SABRE calculation corresponds to SABRE Case 04 in the Computer Case Summary.

The sequence of events calculated by SABRE is given in Table 5.4-2. Figure 5.4-1 shows SABRE results for core power, core channel inlet flow, downcomer level, and steam dome pressure.

For this relatively "slow" transient, the SABRE calculation for core power is consistent with the power predicted by the steady-state 3-D neutronics code, SIMULATE.

PAGE 138 Table 5.4-1 Changes Made to Base SABRE 9x9 Input Deck in Appendix F I GENE-637-024-0893, Table 2.1. 2 Calo. EC-ATWS-1001.

End time (F.2.1) 140 seconds Time step data (F.3) Max = 30 msec Min = 15 msec (O<t<10 sec) Max = 50 msec Min = 25 msec (10<t) Initial total core flow (F.7.1) 87 Mlb/hr (This is the initial core flow used in the SSES ATWS analysis).

1 Initial guess for core channel flow (F.7.2) 80 MLb/hr Initial downcomer subcooling (F.10) 29.49 Btu/2 Status of scram system (F.19.1) -1 (scram and ARI failed) HPCI operability flag (F.20.1) 0 (HPCI inoperable)

Time at which recirc. pump A' is tripped (F.23.1) 0 seconds.

Time at which recirc. pump 'B' is tripped (F.23.2) 0 seconds.

Time at which feedwater pumps are tripped (F.24.1) 0 seconds.

Set point for MSIV closure on low water level -129.D+09 inches (prevents MSIVs from closing on (F.25.2) low level) Time delay for ADS actuation (F.27.4) 102.D+09 seconds (prevent actuation of ADS)

Table 5.4-2 Sequence of Events Calculated by SABRE

• Kinetics file is /dOOlappLisabre3VM/data/u2cI.simtran.

out *** SABRE data file is /home/earac/sabre_31/irput/ec-atws-O505/cO4.dat

• • • This is not a restart case I S A B R E -Version 3.1 (04) SSES Power Uprate (U2C9)-Boitdown with scram failure t(sec)= .000 Scram is Failed t(sec)= .000 HPCI is inoperable t(sec)= .000 Low-Pres Condensate Injection mnop. t(sec)= .000 AR! is FaiLed t(sec)= .000 Feedwater Trip on specified time t(sec)= .000 Recirc pump-A trip on specified time Trip deLay .O00E+00 sec t(sec)= .000 Recirc pump-B trip on specified time Trip delay = .O00E+00 sec t(sec)= 33.309 RCIC initiation on Low water Level Setpoint for initiation

= -38.00 in. t(sec)= 93.209 ADS timer initiation Timer starts when Level drops to -129.00 in. Timer runs out in .102E+12 sec. t(sec)= 93.209 DI CooLer Trip on Low Rx Level Trip Setpoint = -129.000 inches 10 Time (sec)16 U9 0 n 0 S"SABRE 20 40 60 Time (seo)120"Time (eec) Time (eec)Figure 5.4-1 Comparison of SABRE core power calculation against SIMULATE-E for inventory boildown transient with scram failure (U2C9). (SABRE Case 04)70 60 50 40 30 20 IL C C In l0 IL E E ,O E CA 80- 100 120I i I ...........

~ /L-11=

5.5 Small

Break LOCA -Comparison to CONTAIN Results In this section, the SABRE containment model is compared with CONTAIN' results for a small liquid break. In the SABRE code, the break flow and enthalpy can either be specified in an input table or they can be calculated (if they are calculated the break location and area are specified).

A 0.02 if liquid break is specified in this benchmark study. The break is located in the downcomer region. The reactor is initially operating at 102% power (3510 Mwt) and 100 Mlb/hr total core flow. Kinetics data corresponds to U2C10. The break occurs at t=0. Changes made to the base SABRE U2C10 input deck are identified in Table 5.5-1. The sequence of events generated by SABRE is shown in Table 5.5-2. This SABRE calculation corresponds to SABRE Case 05 in the Computer Summary. The CONTAIN calculation corresponds to CONTAIN Case 05 in the Computer Case Summary.

The CONTAIN model consists of two control volumes which represent the drywell and wetwell.

The model accounts for the thermal capacitance of the steel structures within the containment.

Thermal capacitance of the concrete walls is neglected in the CONTAIN model used in this benchmark study because it is also neglected in the SABRE model. As in the case of the SABRE model, the CONTAIN drywell model allows for the formation of a pool layer as water discharged from the break accumulates on the drywell floor. When the height of the pool layer reaches 18", water will overflow into the downcomer pipes and enter the suppression pool. Much of the input data for this model is taken from PP&L calculation EC-THYD-001, "CONTAIN Model for Primary Containment".

In the CONTAIN calculation, the break flow and enthalpy calculated by SABRE were used as input data. That is, the mass/enthalpy source in the CONTAIN model corresponds to the break flow/enthalpy in SABRE. Figures 5.5-1 & 5.5-2 show a comparison of SABRE and CONTAIN results for the small break LOCA. Overall, the calculation results show good agreement for the containment temperatures, pressures, and water levels. The initial rise in suppression chamber air space temperature is caused by compression of gas in the wetwell atmosphere as nitrogen flows from the drywell to the suppression chamber. As the flow of nitrogen to the wetwell decays, the air space temperature begins to decrease as a result of heat transfer from the atmosphere to the pool. In the long term, the atmosphere temperature begins to lag the suppression pool temperature.

In Figure 5.5-2 the slightly higher drywell pool level predicted by SABRE is most likely due to the fact that SABRE calculates a somewhat higher drywell pool temperature during the first half of the transient.

The higher temperature results in a decreased water density and a greater water 1 Murata, K. K., Carroll, D. E., Washington, K. E., Gelbard, F., Valdez, G. D., Williams, D. C., and Bergeron, K. D., "User's Manual for CONTAIN 1.1: A Computer Code for Severe Nuclear Reactor Accident Containment Analysis", NUREG/CR-5026, November 1989.t L / "-V1-volume. The difference in suppression pool level predictions is a very small fraction of the total pool level. At t;-1000 sec, SABRE predicts that the break enthalpy begins to drop below the saturated liquid enthalpy corresponding to drywell temperature, i.e., the break is subcooled relative to drywell atmosphere conditions.

This begins to cause a drop in drywell temperature because heat from the drywell atmosphere is transferred to the subcooled break flow. At t-1200 see, the drywell atmosphere becomes saturated with water (relative humidity reaches 100%) because of the decreasing atmosphere temperature.

At this time "rainout" begins to occur (liquid is removed from the atmosphere) which slows the drop in drywell temperature.

As the drywell temperature and pressure continue to drop from the effects of the subcooled break flow, drywell pressure finally drops below wetwell pressure at about 1600 seconds. At this time the vacuum breakers open and there is a sudden drop in the wetwell atmosphere temperature.

Both codes predict vacuum breaker opening at about the same time.

Table 5.5-1 Changes Made to Base WlO0 Input Deck in Appendix G Problem endtime, (F.2.1) 3000 seconds Hyrali tmestep size (F.3) Max = 30, Min= 15 mnsec for t<20 sec Max 50, Min = 25 msec for t>20 sec Initial Power (F.6.1) 3510 MWth Initial CST water volume (F.14) 300,000 gal LOCA data (F.18) Liquid break flow area = 0.02 Break flow multiplier

= 1.25. Scram on time (F.1 9.6) 2.0 sec Suppression pool level at which HPCI suction 1.E+09 ft (no transfer-keep suction on CST) transfers to the SP (F.20.7) Time at which operator takes control of HPCI 600 sec injection (F.20.14)

RCIC operability flag (F.21.1) 0 (RCIC inoperable)

Time that recirc pump 'A' is tripped (F.23.1) 0.0 sec Time that recirc pump '6' is tripped (F.23.2) 0.0 sec Time at which FW is tripped (F.24.1) 4.0 sec Time at which MSIV closure is initiated (F.25.1) 2.0 sec Time at which controlled cooldown of RPV is 600 sec initiated (F.30.1) Cooldown rate (F.30.2) 9 0°F/hr SRV set points (F.31.4) Values from p. 4-16 of GENE-187-22-0992, 1993 (SAFER/GESTR-LOCA Analysis)

Initial suppression pool level (F.48) 24 ft (to agree with CONTAIN model) SP free area vs. elevation table (F.54) use free area of 5277 ft 2 for all elevations to agree with CONTAIN model.

/L-'-I Table 5.5-2 Sequence of Events Calculated by SABRE *** Kinetics file is /d00/appl/sabre3v0/data/u2clO.simtran.out

' SABRE data file is /home/eamaclsabre_31/input/ec-atws-0505/cO5.dat

• • • This is not a restart case 1 S A B R E -Version 3.1 (02) U2C10 -- 0.02 ft2 Liq break -- HPC[ aligned to CST t(sec)= .000 Liquid Break Break Area ..020 ft2 MultipLier on break flow = 1.250 t(sec)= .000 RCIC is InoperabLe t(sec)= .000 Low-Pres Condensate Injection Inop. t(sec)= .000 Recirc pump-A trip on specified time Trip delay = .000E+O0 sec t(sec)= .000 Recirc pump-B trip on specified time Trip delay = .OOOE+00 sec t(sec)= 2.003 MSIV cLosure on specified time t(sec)= 2.003 Scram initiated on specified time Scram time (sec) = 2.80 t(sec) 4.013 Feedwater Trip on specified time t(sec)= 4.493 HPCI initiation on Hi Drywelt Press. Setpoint for initiation

= .17E+01 psig t(sec)= 4.523 DW Cooler Trip on Hi DW Press Trip Setpoint = 1.720 psig t(sec)= 4.823 ALL Control Rods Inserted t(sec)= 6.016 MSIVs are closed t(sec)= 600.042 Operator takes control of HPCI inj. -- -'5.-30-2225 201 200 E = 175 --e---- CONTAIN 10 e CSABRE 0 SABRE 0 150 125 -10 ............, ....., 010000 20 3000 10 1000 2000 Time (see) Time (sec) 60 ..~ 150 501 ---CONTAIN 2 125 40 SABRE 30 E E = E S2100 e CONTAIN SABRE S75 010 1000 2000 Time (sec) Time (sec) Figure 5.5-1 Comparison of SABRE and CONTAIN results for small break LOCA. (SABRE Case 05)

ID CL E 0 CL r.. .2 a. LD 125 F 100 I---e--CONTAIN e CONTAIN SABRE"'a" '32111 I 1000 2000 000 U. I-3000 Time (sec) Time (sac)Figure 5.5-2 Comparison of SABRE and CONTAIN results for small break LOCA. (SABRE Case 05)Time (sec)I .2 *2 .5 a. I--Iaa-iw- /j/-&

5.6 Turbine

Trip ATWS from MEOD Rod Line -Comparison to GE TRAC Results In this section, SABRE is compared with TRACG results for a turbine trip ATWS initiated from the MEOD (maximum extended operating domain) rod line.1 TRACG includes a three dimensional neutronics model with multiple core hydraulic channels and a two-fluid flow model. TRACG is considered a "state-of-the-art" computer code for simulation of BWR behavior under unstable operating regimes. The TRACG results are therefore very useful for assessing the accuracy of SABRE in predicting BWR instabilities.

Both simulations are carried out with 100% turbine bypass capacity.

The GE calculation was carried out with a core model consisting of 8x8 fuel. Neutronic data bases available for SABRE include a U2C7 core (full core of 9x9 fuel) and mixed 9x9/t0x10 cores (mixed core of 9x9 and partial-length lOxlO fuel). The U2C7 core model is chosen for this benchmark as this is the most similar to the core model used in the GE study. Changes made to the base input deck are documented in Table 5.6-1. This SABRE calculation corresponds to SABRE Case 06 in the Computer Case Summary.

In this scenario, feedwater injection is available since the reactor is unisolated; however, feedwater heating is not maintained because of the loss of extraction steam from the turbine.

The loss of feedwater heating leads to high levels of core-inlet subcooling which drives the reactor into severely unstable operation.

Both TRACG and SABRE use a 60 second time constant to model the decay of feedwater temperature.

The sequence of events calculated by SABRE for this transient is presented in Table 5.6-2. Figure 5.6-1 shows the SABRE calculation results for the core-average power and core inlet subcooling.

Corresponding GE results obtained with TRAC-G are presented in Figure 5.6-2. The core-inlet subcooling results show good agreement.

The amplitude of the power pulses predicted by SABRE show fairly good agreement with the large-amplitude spikes calculated by TRACG. However, SABRE shows a lower frequency oscillation than TRACG and the SABRE calculation predicts a longer time to the onset of instability.

The lower frequency is probably due to the higher fluid inertia associated with the single-core-channel model used in SABRE. In a multiple-channel core model such as that used in TRACG, out-of-phase channel oscillations can occur among various core channels with the flow behavior essentially decoupled from the regions above and below the core. With a single core channel model core flow oscillations necessarily involve flow oscillations in the lower and upper plena and in the riser/separator region. The greater fluid inertia associated with the single core channel model used in SABRE probably contributes to the greater core stability shown in the SABRE calculation (i.e., the longer time to the onset of instability).

1 NEDO-32047, 'ATWS Rule Issues Relative to Core Thermal-Hydraulic Stability", General Elechic Company, January, 1992.

Table 5.6-1 Changes Made to Base SABRE 9x9 Input Deck in Appendix F.. Problem end time (F.2.1) 240 seconds Time step data (F.3) Max = 5 msec Min = 5 msec (0<t<10 sec) Max = 35 rnsec Min = 35 msec (10<t<120)

Max = 25 msec Min = 25 msec (120<t<135)

Max= 15 rrsec Min= 15 msec (135<t<180)

Max = 5 msec Min 5 msec (180<t) initial core power (F.6.1) 3323 MVW Rated core thermal power (F.6.2) 3323 M initial total core flow (F.7.1) 81.3 MLb/hre Initial guess for core channel flow (F.7.2) 75 MLb/hr Status of scram system (F.19.1) -1 (scram and ARI are failed) Trip recirc. pump 'A' on specified time (F.23.1) 0.0 seconds (get recirc pump on turbine trip) Trip recirc. pump 'A' on specified time (F.23.2) 0.0 seconds (get recirc pump on turbine trip) High water level set point for feedwater trip 54.D+09 inches (prevent loss of FW on high (F.24.2) level oscillations)

Delay for level set point set down (F.24.12) 11..D+09 seconds (neglect set point set down) Time at which turbine trip is initiated (F.26.1) 0.0 seconds Maximum combined steam flow (F.26.10)

=2(F.26.4)

= 2(14.631)

= 29.26 MIb/hr. (100% bypass capacity)

Turbine bypass capacity (F.26.17)

=(F.26.4)

= 14.631 MIb/hr (100% bypass _capacity)

Gap Conductance (F.36.4) 870 Btu/hr-t-°F (Table 2.1 of GENE-637 024-0893)

I Al Table 5.6-2 Sequence of Events Calculated by SABRE *** Kinetics fiLe is /dOO/appL/sabre3vO/data/u2c7.simtran.out

• SABRE data file is /home/earac/sabre_31/input/ec-atws-0505/c06.dat

• This is not a restart case 1 S A B R E -Version 3.1 (06) Turbine Trip ATWS from HEOD Rod Line (u2c7) -compare to GE t(sec)= .000. Scram is FaiLed t(sec)= .000 Low-Pres Condensate Injection Inop. t(sec)= .000 ARI is Failed t(sec)= .000 Main Turb Trip on specified time t(sec)= .000 Recirc pump-A trip on specified time Trip delay = .OOOE+00 sec t(sec)= .000 Recirc pump-B trip on specified time Trip deLay = .OOOE+00 sec I 4;10, I-. 0 (a 0 0 10.1 I.6 10" ' 2 40 60 80 10 120 4 6 8 0 2 4 Time (sec) 70 0 60 'M 50 0 40 ,1 0 U 320 C 2 -o 0 20 40 60 80 100 120 140 160 180 200 220 240 Time (sec) Figure 5.6-1 SABR E results for turbine trip ATWS with 100% turbine bypass capacity.

U2C7 core model (9x9 core). No feedwater flow reduction. (SABRE Case 06)

(9661 unrf '-f4l!!IS o3jrl1IPAH-Jgua9141 WOoo "dM oq O 9.rW Ifl sonssI o1n) S SAdp,, 'V-LtuOZE qm O IO'I IoJ u3M slinso-NDI

• K!ds sssd~q ouiqrtm %00[ 1p,.yA StALL dp.4 auiq-t4 g;/HMAE joj sjinsoj 210 Z"9"F, QungiA MM391 301 NDUOM ME e I Ma 891 51 on El

• a a 19 =SNI-MG ITN IH g ! fl3SI 3431 WOM NZ B DO m 91 egi m. got go a e o .... .....................

J ...i ..... .. ... ... ........... .......................

~..........................

.-...................

-~f3Qd 3~n a~z1wJ?~--

...........

I +/-.U S+/-t ::;.-::: I ..........

... ....... .. .. .. .. ..... ...... .. .........

...LtL 3i)Yd q I 5.7 ATWS Initiated by Pressure Regulator Failure Open--Comparison to GE Results The GE power-uprate ATWS analysis' for U2C7 shows that there are two limiting ATWS scenarios with regard to peak vessel pressure, peak clad temperature, and peak suppression pool temperature.

These two scenarios are the MSIV-closure and PREGO ATWS events. In this benchmark study, SABRE results for an ATWS scenario initiated by pressure regulator failure to maximum demand (PREGO) are compared against the GE power uprate ATWS analysis.

Section 5.8 compares SABRE against GE calculations for the MSIV-closure ATWS. The PREGO scenario involves failure of the pressure regulator to maximum steam demand as the initiating event. Scram and ARI are failed. The reactor depressurizes until the MSIVs close on low steam line pressure.

Recirculation pumps trip on high pressure following the MSIV closure, and the operator initiates boron injection to bring the reactor to Hot Shutdown.

Changes made to the base 9x9 SABRE input deck are identified in Table 5.7-1. This SABRE calculation corresponds to SABRE Case 07 in the Computer Case Summary.

Following the MSIV closure, the turbine-driven feedwater pumps continue injection to the Reactor Pressure Vessel (RPV) while there is sufficient steam pressure available within the main steam lines to power the feedwater turbines.

In the PREGO event, the turbine control valves and the turbine bypass valves open until steam flow reaches the maximum combined steam flow limit (see §F.26. 10). When the MSIVs close on low steam line pressure, steam continues to be extracted from the steam lines and pressure drops rapidly. Consequently, the turbine-driven feedwater pumps would not operate for any significant amount of time following closure of the MSIVs. In the GE ATWS analysis for the PREGO initiating event, it was assumed that feedwater would operate for 90 seconds following the start of MSIV closure.2 While this assumption is valid for the MSIV closure ATWS with proper operation of the pressure regulator, it is not a valid assumption for the PREGO ATWS. The assumption of feedwater operation for 90 seconds after the start of MSIV closure does not invalidate the GE Power Uprate analysis for the PREGO event because the assumption leads to a conservatively high suppression pool temperasture for the event. The availability of feedwater maintains level and fission power considerably higher than if feedwater flow was lost early in the event because of low steam line pressure.

For benchmarking purposes, feedwater is maintained operable in the SABRE calculation for 90 seconds after the start of MSIV closure. This is accomplished by setting the low steam line pressure limit for feedwater operation to a negative value (see input F.24.3 in Table 5.7-1). In the GE and SABRE analyses, it is assumed that operators do not take manual control of feedwater injection prior to loss of feedwater from main-steam-line pressure decay. Shortly after the loss of feedwater injection, HPCI and RCIC systems automatically initiate when level drops to their set points. In the SABRE calculation, HPCI and RCIC flows were adjusted to maintain downcomer water level as close as possible to the level used in the GE calculation.

This was done in order to obtain a valid comparison for core power and suppression pool temperature.

HPCI suction is maintained on the CST for the entire transient because the 1 GENE-637-024-0893, "Evaluation of Susquehanna ATWS Performance for Power Uprate Conditions," September 1993. 2 GENE-637-024-0893, p. A-2.

-I-U 4 £ __EOPs instruct the operator to manually defeat the automatic suction transfer to the suppression pool on high suppression pool level of 23.83 feet. In the GE analysis, HPCI suction was maintained on the CST. In the SABRE calculation, the water level setpoint setdown for the feedwater system is neglected (the water level setpoint is "setdown" from +35" to +18" following a Level 3 scram signal 3). The setpoint setdown and any recirculation pump runbacks are neglected in the SABRE calculation in order to simplify the analysis and to add conservatism.

In the GE calculation, water level was initially controlled within the target band (-80" to -110') specified by the EOPs for U2C7 but then level was allowed to drift above the level control band as the transient progressed (level was about -55" at t=1000 sec).4 Since the GE REDY code controls water level through specification of the LPCI injection rate in the code input, it is difficult for GE to maintain constant level in their calculations.

Consequently, GE allowed level to drift somewhat during the simulation.

This was acceptable because with the GE methodology the somewhat higher level results in an overprediction of the suppression pool temperature.

A summary of the sequence of events calculated by SABRE is presented in Table 5.7-2. Figures 5.7-1 and 5.7-2 show a comparison of the GE and SABRE calculations for power, suppression pool temperature, steam dome pressure, core-average heat flux, indicated water level, and core average void fraction.

Key results for the PREGO ATWS are compared in Table 5.7-3. Comparison of results in Table 5.7-3 shows that the peak suppression pool temperature predicted by SABRE differs from the GE value by only 0.5 'F (176.5 °F for GE and 176.0 'F for SABRE). This corresponds to a difference of -0.6% in the pool temperature rise which is within the margin of error of the calculation.

SABRE predicts slightly higher values than GE for the steam dome pressure, and the core average heat flux. In the SABRE calculation, the low-pressure set point (861 psig) for MSIV isolation is reached at 11.4 seconds into the event while GE predicts that pressure drops to the set point at 12.2 seconds.

The most significant difference between the two calculations occurs after this point. SABRE results indicate that pressure begins to rise at about 15 seconds which is 3.6 seconds after the low-pressure isolation signal is reached. This SABRE result makes physical sense in that pressure should start to recover when the MSIVs are nearly closed (closure stroke is 4 seconds).

The GE results indicate that MSIV closure begins at 14 sec and pressure begins to rise at about 18 seconds into the event.5 This means that in the GE calculation there was a delay of nearly 2 seconds between the occurrence of the isolation signal (861 psig) and the start of MSIV closure.

This delay of -2 seconds appears too long based on the fact that the Susquehanna RETRAN model uses a delay of 0.05 seconds.6 Outside of this discrepancy in MSIV closure time, the SABRE and GE calculations show good overall agreement particularly with regard to power and suppression pool temperature.

3 Calc. EC-FUEL-0969, Section E.3. 4 GENE-637-024-0 8 9 3 , Supplement

1. 5 GENE-637-024-08 9 3 , Table A.3.1. 6 Calc. EC-FUEL-0520 Rev. 0 (NFE-B-01-002).

154 Table 5.7-1 Changes Made to Base 9x9 Input Deck in Appendix F Problem end time (F.2.1) 1050 seconds Time step data (F.3) Max = 5 msec Min 5 msec (0<t<30) Max = 30 msec Min = 30 msec (30<t) Initial total core flow (F.7.1) 87 Mlb/hr (Table 2.1 of Ref. 1) Initial guess for core channel flow (F.7.2) 80 MLb/hr Initial steam dome pressure (F.8) 1053 psia (Table 2.1 of Ref. 1) Initial downcomer subcooling (F.9) 29.49 Btu/Lbm 6 Status of scram system (F.19.1) -1 (scram and ARI are failed) SP level at which HPCI suction transfers to pool 23.83D+09 (prevent suction transfer to SP in (F.20.7) accordance with EOPs) Minimum HPCI flow (F.20.10) 250 gpm (reduced to make level control easier as power decreases)

Time at which operator takes control of HPCI injection 238 sec (use operator control of HPCI to try and match (F.20.14)

GE level vs. time. HPCI target level table (F.20.15)

Target specified to match GE water level. Time at which operator takes control of RCIC injection 238 sec (use operator control of HPCI to try and match (F.21.12)

GE level vs. time. RCIC target level table (F.21.13)

Target specified to match GE water level. High pressure trip for recirc. pump 'A' (F.23.7) 1214.7 psia (Table 2.2 of Ref. 1) High pressure trip for recirc. pump 'B' (F.23.9) 1214.7 psia (Table 2.2 of Ref. 1) Time delay for recirc. pump A' runback on low 1.D+09 seconds (Neglect runback) feedwater flow (F.23.12)

Time delay for reciro. pump 'A' runback on low 1.13+09 seconds (Neglect runback) feedwater flow (F.23.14)

Time delay for recirc. pump 'A' runback on low water 1.0+09 seconds (Neglect runback) level (F.23.16)

Time delay for recirc. pump 'A' runback on low water 1.03+09 seconds (Neglect runback) level (F.23.18)

Time at which manual trip of FW pumps is initiated 102 sec (90 sec after start of MSIV closure) (F.24.1) GENE-637-024-0893, p. A-2. Minimum steam line pressure for feedwater operation

-1.D+09 psia (prevent loss of FW on low steam line (F.24.3) pressure to agree with GE analysis)

Time at which pressure regulator failure-open is 0.0 seconds (to agree with GE caic.) initiated. (F.26.3) Maximum combined steam flow (F.26.10) 17.02 MIb/hr (LDCN 2750) ADS operability flag (F.27.1) 0 (EOPs instructs operator to manually defeat ADS in ATWS events) Time at which SLCS is initiated (F.29.1) 113 seconds (Table A.3.1 of Ref. 1) SRV actuation setpoints (F.31.4) Used UALs per power uprate ATWS analysis.7 Gap Conductance (F.36.4) 870 Btu/hr-ft 2-0 F (Table 2.1 of Ref. 1) Initial suppression pool level (F.48) 22 ft. (gives water volume. close to value in Table 2.1 of Ref. 1) Time at which Loop 1 of SPC becomes effective (F.50) 1000 seconds (Table A.2.1 of Ref. 1) Time at which Loop 2 of SPC becomes effective (F.51) 1000 seconds (Table A.2.1 of Ref. 1) 6 Calc. EC-ATWS-1001.

7 GENE-637-024-0893, "Evaluation of Susquehanna ATWS Performance for Power Uprate Conditions," Supplement 1, Table 1, September 1993.

Table 5.7-2 Sequence of Events Calculated by SABRE for Case 07

• Kinetics file is 1d00/appl/sabre3v0/data/u2c7.simtran.out

• • • SABRE data file is /Ihome/earnac/sabre-31/input/ec-atws-0505/cO7.dat

• • • This is not a restart case 1 S A B R E -Version 3.1 (07) PREGO ATWS -- U2C7 -- Compare to GE t(sec)= .000. Scram is Failed t(sec)= .000 Low-Pres Condensate Injection Inop. t(sec) .000 ADS actuation is defeated t(sec)= .000 ARI is FaiLed t(sec)= .000 Press Regulator FaiLure Open (Max Demand) t(sec)= 11.430 MSIV closure on Low reactor pressure Setpoint(psia)

= 875.700 t(sec)= 15.430 MSlVs are closed t(sec)= 17.675 Level Setpoint Setdown Setdown occurs when Level drops to 13.00 in. Delay for setpoint setdown = .10E+10 sec t(sec)= 17.675 Recirc pump-A runback on Low Rx lvt Setpoint for Runback = 13.00 in. Trip delay = .lOOE+10 sec t(sec)= 17.675 Recirc pump-B runback on Low Rx Lvt Setpoint forRunback

= 13.00 in. Trip delay = .100E+10 sec t(sec)= 18.120 Recirc pump-A trip on hi Rx press. Setpoint for trip = 1200.00 psig Trip delay = .230E+00 sec. t(sec)= 18.120 Recirc pump-B trip on hi Rx press. Setpoint for trip = 1200.00 psig Trip delay = .230E+00 sec. t(sec)= 102.023 Feedwater Trip on specified time t(sec)= 113.003 SCLS Initiation Number of OperabLe SLCS pumps = 2. Transport DeLay to vessel = 75.00 sec. t(sec)= 142.103 HPCI initiation on Low water Level Setpoint for initiation

= -38.00 in. t(sec)= 142.103 RCIC initiation on tow water Level Setpoint for initiation

= -38.00 in. t(sec)= 238.013 Operator takes control of HPCI inj. t(sec)= 238.013 Operator takes control of RCIC inj. t(sec)= 1000.013 Loop 1 of Supp Pool Cool Effective Service Water Temperature

= 88.00 F t(sec)= 1000.013 Loop 2 of Supp Pool CooL Effective Service Water Temperature

= 88.00 F

-r-o-r-Table 5.7-3 Comparison of GE and SABRE Results for U2C7 PREGO ATWS (Case 07)t Based on temperature/pressure heat flux rise predicted by GE, i.e. 100%x(SABRE-GE)/GE.

I 70 S50 ~50-2 40 " 10 C. 'L 125 q 30 .2. SABRE / // SABRE LL 20:7 10 M 100 Sw 0 0 200 400 600 800 10 Time (sec) Time (sac) 150 0D .., ........175 . 1400 -e-- GE 150 SABRE S1300 S1200 125 0. 125 E 1100 o 0 a 75 x 000 1 0. 2 3U05 r= 600 7 0 0 ....F .......25 0 1 500 0 10 20 30 40 50 0 10L 20 3 4 Time (sec) Time (sec) Figure 5.7-1 Comparison of SABRE and GE predictions for U2C7 PREGO ATWS event with boron injection. (SABRE Case 07)

S -G-- GE S-- SABRE "a- 0.8 0--- GE "SABRE 0.6 S-50 aL 0.4 a -1000 0.2 0 , .I .I .... .... -150 0 200 400 600 800 i000 2000 200 400 600 800 1C Time (sec) Time (sec) Figure 5.7-2 Comparison of SABRE and GE predictions for U2C7 PREGO ATWS event with boron injection. (SABRE Case 07)

IS-9 5.8 MSIV-Closure ATWS with Boron Injection

-Comparison to GE Calculations An evaluation of the Susquehanna plant performance under ATWS conditions was performed by GE as part of the Susquehanna power uprate project.!

In this section, SABRE results are compared against GE calculations, performed with the REDY code, for an MSIV-closure ATWS with boron injection.

Closure of the MSIVs is initiated at t=0. It is assumed that the operator does not take manual control of feedwater flow prior to its coast-down from depletion of main steam line pressure.

Once feedwater flow ceases, level drops rapidly, and HPCI and RCIC systems initiate.

HPCI and RCIC, with suction from the condensate storage tank (CST), are used to control level throughout the transient.

SLCS is initiated at 95 seconds into the event. In this scenario water level is allowed to gradually recover as the reactor approaches the Hot Shutdown condition.

Changes made to the base 9x9 SABRE input deck are shown in Table 5.8-1. The sequence of events generated by SABRE is presented in Table 5.8-2. This SABRE calculation corresponds to SABRE Case 08 in the Computer Case Summary.

GE and SABRE predictions for this event are compared in Figures 5.8-1 and 5.8-2. The SABRE calculation for the decay of fission power as a result of boron injection shows good overall agreement with the GE calculation.

This is confirmed by the close agreement in suppression pool temperature.

At 1000 seconds, the GE calculation predicts a suppression pool temperature of 178 OF. At the same time SABRE calculates a pool temperature of 177.3 'F. The short term steam dome pressure transient and core-average heat flux transients are compared in Figure 5.8-1 SABRE shows good agreement with the GE pressure calculation, but tends to under predict the peak heat flux. The SABRE heat flux curve is broader but less peaked. Differences between the key parameters, pool temperature, steam dome pressure, and core average heat flux are summarized in Table 5.8-3. This case was repeated with one SRV out of service (Case 8a in Computer Case Summary).

With one SRV out of service, SABRE predicts a peak steam dome pressure of 1351 psia at 10.5 seconds. GE calculates a peak dome pressure of 1340 psia at 9 seconds with one SRV out of service (Ref. 1, Table 3.1). SABRE overpredicts the pressure rise, as compared to GE, by 3.8%. The 1-D kinetics equations in SABRE are integrated using the LSODES ordinary differential equation solver. LSODES automatically selects the time step size and method order to optimize run time and satisfy the error criteria which are specified as part of the code SABRE input data (see §F.1.5 and §F.1.6). The fluid flow and heat transfer equations in SABRE are solved using first-order, semi-implicit integration with the time step size specified by the code user. Since the hydraulic step size is part of the SABRE code input (§F.3), sensitivity studies were performed to determine the effect of variations in step size. A sensitivity study is also performed on the kinetics error control parameters.

'GENE-637-024-0893, "Evaluation of Susquehanna ATWS Performance for Power Uprate Conditions," September 1993.

The results presented in Figures 5.8-1 and 5.8-2 were obtained using hydraulic time steps of 5 msec for t<30 sec and 25 msec for t>30 sec. A smaller hydraulic step size was used for the beginning part of the transient because core power and pressure change rapidly when the MSIV closure occurs. In the remaining part of the transient, a larger time step is used because reactor conditions change more slowly as boron is injected and the core approaches Hot Shutdown condition.

Sensitivity studies were performed on both the short-term time step and the long-term time step. Results of the hydraulic time step sensitivity studies are shown in Figures 5.8-3 and 5.8-4. Figure 5.8-3 shows the power spike generated by the MSIV closure with hydraulic time steps of 1.25, 2.5, 5, and 10 msec. The maximum neutron power and peak steam dome pressure corresponding to each of these time steps is listed in the table below. 08e 10 540 1310.2 08 5 516 1308.6 08f 2.5 504 1307.5 08i 1.25 497 1306.7 These results show that the incremental change in maximum power and peak pressure is decreasing as the step size decreases which indicates convergence to a solution.

Also, with a hydraulic step size of 5 msec the peak power appears to be overestimated by about 4%. Because of the short duration of the power spike, the differences in peak power computed with the step sizes listed in the above table will have negligible effect on the long term suppression pool temperature response.

As a result, any of the short-term step sizes ranging from 1.25 msec to 10 msec would be acceptable for the ATWS suppression pool heat-up analysis.

A step size of 5 msec is recommended however because it is the best compromise between computer run time and accuracy.

Aside from the peak power, there is little variation in the short-term solution with the step sizes considered.

The effect of variations in the long-term hydraulic time step size (t>30 sec) is also investigated.

Figure 5.8-4 shows the suppression pool temperature response with hydraulic time steps (for t>30 sec) ranging from 20 to 40 msec. The results do not show any noticeable dependence on step size for the range of steps considered.

Therefore, a long-term hydraulic step size anywhere in the range of 20 to 40 msec would be acceptable.

A sensitivity was also performed on the error control parameters for the kinetics solution (RTOL and ATOL in §F. 1.5 and §F. 1.6, respectively) carried out with the ODE solver LSODES. SABRE Case 08 was run with RTOL=ATOL=I.D

05. In SABRE Cases 08j and 08k the values of RTOL and ATOL were set to 5.D-05 and 5.D 06, respectively.

Figures 5.8-5 and 5.8-6 show that the core power and suppression pool temperature response are essentially insensitive to the changes in RTOL and ATOL.-ap 160 A sensitivity study was also performed on the number of boron mixing nodes used to track the transport of borated water through the reactor core and on the boron entrainment exponent b in Equation 2.4.6-1. The number of axial nodes for boron transport through the core can range from 1 to 10 (value is selected in code input). The results in Figure 5.8-7 show that the solution is independent of the number of mixing nodes when 8 to 10 nodes are selected.

All ATWS cases in this calculation package were carded out using b=1 (linear model) in Equation 2.4.6-1 except for Cases 081 and 08m. Case 081 was run with b=-1/2 and Case 08m was run with b=2. The entrainment fraction as a function of total core flow for b=1/2, b=l, and b=2 is'shown in Figure 2.4.6-1. Figure 5.8-8 shows the sensitivity of the suppression pool temperature response to the entrainment exponent b. The results show that there is little change in the pool temperature response as b is varied from /2 to 2.

Table 5.8-1 Changes Made to Base SABRE 9x9 Input Deck in Appendix F Problem end time (F.2.1) 1000 seconds Time step data (F.3) Max = 5 msec Min = 5. rnsec(t<30)

Max = 25 msec Min = 25 msec (30<t) Initial total core flow (F.7.1) 87 MIb/hr (Table 2.1 of Ref. 1) Initial guess for core channel flow (F.7.2) 80 MLb/hr Initial steam dome pressure (F.8) 1053 psia (Table 2.1 of Ref. 1) Initial downcomer subcooling (F.9) 29.49 Btu/Lbm 2 Status of scram system (F.19.1) -1 (scram and ARI are failed) SP level for HPCI suction transfer to SP (F.20.7) 23.83D+09 ft. (prevents transfer)

HPCI target level table (F.20.15)

Target water level for HPCI is specified to match GE reactor water level response.

RCIC target level table (F.21.13)

Target water level for RCIC is specified to match GE reactor water level response.

High pressure trip for recirc. pump 'A' (F.23.7) 1214.7 psia (Table 2.2 of Ref. 1) High pressure trip for recirc. pump 'B' (F.23.9) 1214.7 psia (Table 2.2 of Ref. 1) Time delay for recirc. pump 'A' runback on low 1 .D+09 seconds (Neglect runback) feedwater flow (F.23.12)

Time delay for recirc. pump 'A' runback on low 1 .D+09 seconds (Neglect runback) feedwater flow (F.23.14)

Time delay for recirc. pump 'A' runback on low 1 .D+09 seconds (Neglect runback) water level (F.23.16)

Time delay for recirc. pump 'A' runback on low 1 .D+09 seconds (Neglect runback) water level (F.23.18)

MSIV closure on specified time (F.25.1) 0.0 seconds (to agree with GE calc.) ADS operability flag (F.27.1) 0 (EOPs instruct operator to manually defeat ADS in ATWS events) "Time at which SLCS is initiated (F.29.1) 95 seconds (Table A.2.1 of Ref. 1) SRV actuation setpoints (F.31.4) Used UAL values from Table 1 of Supplement I to Ref. 1. Gap Conductance (F.36.4) 870 Btu/hr-ft 2-OF (Table 2.1 of Ref. 1) Initial suppression pool level (F.48) 22 ft. (gives water volume close to value in Table 2.1 of Ref. 1) Time at which Loop I of SPC becomes effective 1000 seconds (Table A.2.1 of Ref. 1) (F.50) Time at which Loop 2 of SPC becomes effective 1000 seconds (Table A.2.1 of Ref. 1) (F.51)2 Calc. EC-ATWS-1001.

-r"4p-Table 5.8-2 Sequence of Events Calculated by SABRE Kinetics file is IdOO/appl/sabre3vOldata/u.c7.

simtran. out

• SABRE data file is /home/eamac/sabre.31/input/ec-atws-O505/cO8.dat

• • • This is not a restart case 1 S A B R E -Version 3.1 (08) U2C7 MSIVC ATWS -- Compare to GE t(sec)= .000 Scram is FaiLed t(sec)= .000 Low-Pres Condensate Injection Ihop. t(sec)= .000 ADS actuation is defeated t(sec)= .000 ARI is FaiLed t(sec)= .000 MSlV closure on specified time t(sec)= 4.005 MSIVs are closed t(sec)= 4.810 Recirc pump-A trip on hi Rx press. Setpoint for trip = 1200.00 psig Trip delay = .230E+00 sec. t(sec)= 4.810 Recirc pump-B trip on hi Rx press. Setpoint for trip = 1200.00 psig Trip delay = .230E+00 sec. t(sec)= 95.020 SCLS Initiation Number of OperabLe SLCS pumps = 2. Transport Delay to vessel = 75.00 sec. t(sec)= 98.320 Feedwater Trip on low Stm Line Press FLow stops when press < 175.00 psia t(sec)= 110.445 Level Setpoint Setdown Setdown occurs when Level drops to 13.00 in. Delay for setpoint setdown = .10E+10 sec t(sec)= 139.045 HPCI initiation on Low water Level Setpoint for initiation

= -38.00 in. t(sec)= 139.045 RCIC initiation on Low water Level Setpoint for initiation

-38.00 in. t(sec)= 250.020 Operator takes control of MPCI in]. t(sec)= 250.020 Operator takes control of RCIC inj.

fjorj- /.'Hk Table 5.8-3 Comparison of GE and SABRE Results for U2C7 MSIV Closure ATWS (Case 08)t Based on temperturepressure/heat flux rise predicted by GE, i.e., 1oox(SABRE-)/GE.

100. ..200 90-GE 80- SABRE A 175 ILL M 70 US 60 o ,50E 400 ~40 '5 30 CL 125 GE 03 GE_ o 1SABRE iz20 1 a. 10- 100 0 -10 ....i .............. 0 200 400 600 Soo 1000 75 L ... 0 200 400 600 800 Time (9ec) Time (nec) 1500D ,.O .., ... 175 .j ., ., .. 1400 1-- -- --GE 1400 SABRE S1300 , IL a. o zS oE 1100 0 E 75 .B ooo w -e-- GE ,4 sR -E 90 so 50 8 0 0 [ ...5 .....210 ..15 ..2 0 2 5-.. 5 L 1 01 , .1 5 S Time (sec) Time (sec) Figure 5.8-1 Comparison of SABRE and GE predictions for U2C7 MSIVC ATWS event with boron injection. (SABRE Case 08)

S#~~~ -e* .SABRE

• 0.8 Z-- 0---- GE SABRE .0.66

• U. -50 3 10.4 S-100 6 U o 0.2 -150 -150....................

!.... I,,iil I,,,I,,, S0 200 400 600 800 1000 0 .. ... .... i .. 0 200 400 600 800 Time (sec) Time (sec) Figure 5.8-2 Comparison of SABRE and GE predictions for U2C7 MSIVC ATWS event with boron injection. (SABRE Case 08) 600 500 400 0-f 01 0 0~ L@300 200 100 I I I I I I I 0 hydraulic step size = 5 msec for t<30 sec ...............

hydraulic step size = 10 msec for t<30 sec hydraulic step size = 2.5 msec for t<30 sec .--hydraulic step size = 1.25 msec for t<30 msec 0 2 4 6 8 10 Time (sec) Figure 5.8-3 Effect of hydraulic time step size on power spike for MSIV closure ATWS. Time step size for kinetics solution is selected by ODE solver LSODES. (SABRE Cases 08, 08e, 08f, and 081)I I I T I I B I I16 f 200 190 180 LL 0 1ý70 0. 0 CL 0)170 160 150 140 130 120 110 100 90 80 0 250 500 750 1000 Time (sec) Figure 5.8-4 Effect of hydraulic time step size on suppression pool temperature response for MSIV closure ATWS. Time step size for kinetics solution is selected by ODE solver LSODES. (SABRE Cases 08, 08b, 08c, and 08d) lvi 200 F ' I' 190 .180 00 0U 170 160 a) E 150 o 140 0.

• 130 S120 110 10 Boron mixing nodes In core 100 9 Boron mixing nodes In core ..............

8 Boron mixing nodes in core 90 80 0 250 500 750 1000 Time (sec) Figure 5.8-5 Effect of Boron mixing nodalization on suppression pool temperature response for MSIV closure ATWS. Time step size for kinetics solution is selected by ODE solver LSODES. Hydraulic time steps are 5 msec for t<30 sec and 25 msec for t>30 sec. (SABRE Cases 08, 08g, and 08h) 2 4 6 8 Time (sec)Figure 5.8-6 Effect of 1-D kinetics error control parameters RTOL and ATOL on power spike for MSIV closure ATWS. Hydraulic time steps are 5 msec for t<30 seconds and 25 msec for t>30 seconds. (SABRE Cases 08, 08j, and 08k)I /C-'600 500 400 300 01 0 LL 200 100 0 0 10 0 250 500 750 Time (sec)Figure 5.8-7 Effect of 1-D kinetics error control parameters RTOL and ATOL on on suppression pool temperature response for MSIV closure ATWS. Hydraulic time steps are 5 msec for t< 30 sec and 25 msec for WO3O sec. (SABRE Cases 08, 08j, and 08k)200 190 180--f~a-z-171 L 0 0 CL 0 a. E 0 CL .5 0O 170 160 150 140 130 120 110 100 90 8U 1000 0-r-oL-rI / /0'-200 190 180 170 160 150 140 130 120 110 100 90 80 250 500 750 1000 Time (sec) Figure 5.8-8 Effect of boron entrainment parameter b on suppression pool temperature response for MSIV closure ATWS. Hydraulic time steps are 5 msec for t< 30 sec and 25 msec for t>30 sec. (SABRE Cases 08, 081, and 08m)0.-. 0 af.0 0) 0 0. C 02 2 as

5.9 Calculation

of Counter-Current Flow Limit-Comparison to Kutateladze CCFL Correlation The Ohkawa-Lahey void model is used in the SABRE two-phase flow Calculation.

In this drift-flux model, the functional form of the parameters C. (radial concentration parameter) and Vj (drift velocity) are such that the drift-flux model predicts the counter current flow limit in agreement with the Kutateladze correlation.'

In this section, the SABRE flow model is used to predict the counter-current flow limit (CCFL) for a heated channel with inlet flow blockage.

Calculation results are compared with the Kutateladze correlation in order to verify the two-phase flow model used in SABRE. The core channel model used in this benchmark study was constructed from subroutines taken directly from the SABRE code. The FORTRAN code used for the channel flow calculation is listed in Appendix E. In a channel with inlet flow blockage, liquid flows down into the top of the channel and replaces some or all of the water which is boiled off. If the flow is not counter-current flow limited, then a steady condition will arise in which all of the water converted to steam is replaced by the downward flowing liquid. If the channel power is high enough, however, the flow of vapor at the channel exit will restrict the downward flow of liquid (kinematic choking), and portions of the channel can become completely voided. In this situation, fuel rods can become damaged from loss of two-phase cooling.

The channel power associated with kinematic choking (CCFL) can be predicted from the Kutateladze correlation which is given by' .( 2 I LI V1/ 2 I -2 .(5.9-1) where *O= (5.9-2) Just below the power level associated with CCFL, the vapor and liquid flow rates at the channel exit are given by the following energy and mass balances, w, =jgpgA=Q/(hg -h,) (5.9-3) and w, =-wg, =jpeA (5.9-4) 'Ohkawa, K., and Lahey, RT., "The Analysis of CCFL Using Drift-Flux Models," Nuclear Engineering and Design, Vol. 61, pp. 245-255, 1980.

Substituting (5.9-3) and (5.9-4) into (5.9-1) leads to the following result for the channel power associated with CCFL: -2 Q1 Lv,2[2 "LIJ 112 (5.9-5) (ICCPZ ~ A 1/2 + 1/~Y 4 ( ,1/2h~A For a core channel with a flow area of 0.114 f 2 , and a system pressure of 1000 psia, Equation (5.9-5) predicts that the power associated with kinematic choking is 451 Btu/sec. That is, if the power of the blocked channel is less than 451 Btu/sec, the reverse flow of liquid at the channel exit will be unrestricted, and it will be equal to the inventory loss due to boiling. On the other hand, if the power exceeds 451 Btu/sec, then the flow of vapor at the channel exit will restrict the reverse flow of liquid into the channel. In this situation, there is a loss of coolant inventory accompanied by an increase in channel void fraction.

For benchmarking purposes, the core channel model developed from SABRE subroutines was run at a series of power levels some of which are below the critical power level of 451 Btu/sec and some above. Initially, the channel is filled with saturated liquid, so there is a sudden expulsion of liquid when the channel is instantaneously set to the specified power level. As the transient progresses, the liquid flow at the channel exit reverses its flow direction.

A no-flow boundary condition is specified at the bottom of the channel to simulate the blockage.

At the top of the channel, it is assumed that ah/az = 0, where h is the volume-weighted enthalpy of the two-phase mixture. The channel model consists of twenty-five axial nodes, and the heat addition is axially uniform.

Figure 5.9-1 shows the transient void fraction at the channel exit as a function of channel power. At 450 Btu/sec, which is just below the critical channel power predicted from Eqn. (5.9-5), the void fraction approaches a constant value as there is no restriction on liquid flow into the top of the channel. When the channel power is increased to 452 Btu/sec (1 Btu/sec above the critical power), an excursion in the void fraction occurs since liquid cannot enter the channel in sufficient amount to compensate for the inventory loss due to boiling. Figure 5.9-2 displays the liquid flow behavior at the channel exit for the same cases. The rapid decrease in reverse flow of liquid at the channel exit indicates the onset of kinematic choking. These results demonstrate that the fluid dynamics model in SABRE is consistent with the Kutateladze CCFL correlation.

1 0.9 0.8 0.7 0.6 0 10 20 30 40 50 6C Time (sec) Figure 5.9-1 Prediction of channel power associated with kinematic choking by examination of void fraction response at channel exit.It uJ W Z C. C ~0 70)

X u.I (.1; Ui ,.Z ,r :3 S0 U I1 I1/-0.4 -0.45 -300 Btu/sec I -0.5"CD LOD La 350 Btu/sec -0.55 -0.6 400 Btu/sec -0.65 440 Btulsec 450 E _ 7 , i , , I , I , I , I , , , , I ' ' ' I -10 20 30 40 50 6C Time (sec) Figure 5.9-2 Prediction of channel power associated with kinematic choking by examination of liquid flow response at channel exit.70) 5.10 Suppression Pool Heatup from Decay Heat-Comparison to Heat Balance Calculation Benchmarking results are presented for suppression pool heat up due to decay heat generation.

In the SABRE calculation, an MSIV closure with reactor scram is initiated at t-O. Feedwater trips early in the event (45 seconds) on high RPV water level, and HPCI, with suction on the suppression pool, is used to supply coolant to the reactor. Initial reactor power is 3293 MWt, and the initial suppression pool level is 23 feet. RPV water level is controlled between 13" and 54". There is no removal of decay heat from the containment.

Changes made to the base SABRE input deck are identified in Table 5.10-1. This SABRE calculation corresponds to SABRE Case 10 in the Computer Case Summary.

In this transient, reactor water level is maintained near its initial value. In addition, the reactor is not depressurized; rated reactor pressure is maintained by automatic actuation of SRVs. Since there is little change in the reactor coolant inventory, and since HPCI suction is always on the suppression pool, suppression pool mass is essentially constant.

That is, water from the suppression pool is injected to the reactor, converted to steam by decay heat generation, and deposited back in the pool. Also, there is no sensible heat given off by the reactor vessel and internal structures because the reactor is maintained at high pressure.

For this situation, the rise in suppression pool temperature can easily be calculated from the following heat balance on the containment.

i.T(t) MwJT Cp,w ()d +M SwCPs[T(t)

-T2]+M 8 ,wwCps[T(t)-T 1] (5.10-1) where Qdy(t) = integrated decay heat at t seconds after shutdown (Btu), T(t) suppression pool temperature at time t (CF), Mw = initial mass of water in suppression pool = 7.96x10 6 lbm, Cpw specific heat of water Btu/lbm-*F, MN1 mass of steel in drywell = 1.6x10 6 lbm (§ D.18.14), Ms,ww = mass of steel in wetwell = 0.695x10 6 Ibm (§ D.18.16), CP'S specific heat of steel = 0.14 Btu/lbm-°F

(§ D.2), T12 initial temperature of suppression pool and air space = 90 'F, and T2 = initial temperature of drywell = 120 'F. The temperature dependence for the specific heat of water is approximated by the following linear relation'1 Data from ASME Steam Tables, 5th Edition, p. 278.

CP,w(T) = 0.998 + 1.813x10-[T

-901. In Eq. 5.10-1, the integrated decay power Q, is taken from p. 351 of PP&L Technical Report NPE-89-001, "The PP&L Approach To Risk Management And Risk Assessment (Second Revision)," January 1989. The decay power in (5.10-1) is taken from an independent source in order to provide an independent check on the decay power model used in SABRE. Table 5.10-2 gives the suppression pool temperature as a-function of time after shutdown as calculated from Eq. 5.10-1. The sequence of events calculated by SABRE is presented in Table 5.10-3. Figure 5.10-1 compares the results in Table 5.10-2 against the SABRE calculation for suppression pool temperature.

The good agreement between the results provides validation for the SABRE suppression pool temperature calculation, the containment heat structure calculation, and the SABRE decay heat model.

7a4tA /77 Table 5.10-1 Changes Made to Base 1Ox1O SABRE Input Deck in Appendix G Maximum number of time steps between detailed 2000 edits (F. 1.-1) Minimum number of time steps between detailed 100 edits (F. 1.2) Problem end time (F.2.1) 100000 seconds Number of print interval sets supplied (F.2.2) 1 Time step data (F.3) Max = 5 msec Min = 5 msec (0<t<30 sec) Max = 50 msec Min = 50 msec (30<t) Pdnt intervals for general edits (F.5) 5000 seconds Initial core thermal power (F.6.1) 3293 MWt Time at which manual scram is initiated (F.19.6) 0.0 HPCI low water level initiation set point (F.20.2) +13 inches Time at which MSIV closure is initiated (F.25.1) 0.0 Normal CRD flow rate (F.33.1) 0.0 gpm (want injection flow from HPCl only) Initial heat load from RPV to drywell (F.55) 0.0 Btu/sec (set to zero so all energy goes directly to SP) Set point for loss of drywell cooling on high drywell 0.0 psig (trip drywell cooling at t=0) pressure (F.58)

Table 5.10-2 Suppression Pool Temperature Response Due to Decay Heat Addition (from Eq. 5.10-1)0. 101,186,095.

126,051,035.

170,970,387.

211,407,189.

283,315,635.

347,305,404.

406,553,360.

542,006,444.

664,652,182.

887,624,525.

1,090,604,471.

1,456,580,992.

1,785,291,761.

2,089,293,397.

-J -.7v. v 102.8 105.8 ill.i 115.9 124.4 132.0 139.0 155.0 169.4 195.6 219.3 261.8 299.7 334.5 1500 2000 3000 4000 6000 8000 10000 15000 20000 30000 40000 60000 80000 100000 Table 5.10-3 Sequence of Events Calculated by SABRE

• Kinetics file is /dOO/appt/sabre3vO/data/u2c10.simtran.out

• • • SABRE data fiLe is Ihomeleamac/sabre.31/input/ec-atws-0505/clO.dat

• This is not a restart case 1 S A B R E Version 3.1 (10) Suppression pool heatup by decay heat (U2C10) t(sec)= .000 Low-Pres Condensate Injection mnop. t(sec)= .000 MSIV closure on specified time t(sec)= .000 Scram initiated on specified time Scram time (sec) = 2.80 t(sec)= .000 DW CooLer Trip on Hi DW Press Trip Setpoint = .000 psig t(sec)= .000 HPCI Suction Trans to SP on high SP level SP water level = 23.00 ft t(sec)= 2.805 ALL Control Rods Inserted t(sec)= 3.090 HPCI initiation on Low water level Setpoint for initiation 2 13.00 in. t(sec)= 3.890 Level Setpoint Setdown Setdown occurs when Level drops to 13.00 in. Delay for setpoint setdown = .11E+02 sec t(sec)= 3.890 Recirc pump-A runback on Low Rx tvL Setpoint for Runback = 13.00 in. Trip delay = .300E+01 sec t(sec)= 3.890 Recirc pump-8 runback on tow Rx tvt Setpoint for Runback = 13.00 in. Trip delay = .300E+01 sec t(sec)= 4.005 MSIVs are closed t(sec)= 8.320 RCIC initiation on tow water level Setpoint for initiation

-38.00 in. t(sec)= 8.320 Recirc punp-A Trip on low Rx level Setpoint for Trip = -38.00 in. Trip delay = .900E+01 sec. t(sec)= 8.320 Recirc pump-B Trip on Low Rx level Setpoint for Trip = -38.00 in. Trip delay = .900E+01 sec. t(sec)= 39.735 Main Turb Trip on high water Level Setpoint(inches) 54.000 t(sec)= 39.735 RCIC Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 39.735 HPCI Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= ".935 Feedwater Trip on high Level Trip Setpoint = .54E+02 in. t(sec)= 801.535 HPCI initiation on Low water Level Setpoint for initiation

= 13.00 in. t(sec)= 931.485 HPCI Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 2405.935 HPCI initiation on Low water Level Setpoint for initiation

= 13.00 in. t(sec)= 2541.285 HPCI Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 4492.535 HPCI initiation on Low water Level Setpoint for initiation

= 13.00 in. t(sec)= 4628.185 HPCI Trip on hi water Level Trip Setpoint .54E+02 in. t(sec)= 6925.585 HPCI initiation on Low water Level Setpoint for initiation

= 13.00 in. t(sec)= 7054.185 HPCI Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 9333.285 HPCI initiation on Low water Level Setpoint for initiation

= 13.00 in. t(sec)= 9442.785 HPCI Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 11690.435 HPCI initiation on Low water Level Setpoint for initiation

= 13.00 in. t(sec)= 11816.285 HPCI Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 14327.485 HPCI initiation on Low water Level Setpoint for initiation

= 13.00 in. t(sec)= 14438.435 HPCI Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 16912.235 HPCI initiation on Low water Level Setpoint for initiation

= 13.00 in. t(sec)= 17036.135 HPCI Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 19741.485 HPCI initiation on Low water Level Setpoint for initiation

= 13.00 in. t(sec)= 19854.385 HPCI Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 22561.035 HPC! initiation on Low water level Setpoint for initiation

= 13.00 in. t(sec)= 22683.485 HPCI Trip on hi water Level Trip Setpoint .54E+02 in. t(sec)= 25578.335 HPCI initiation on Low water Level Setpoint for initiation

13.00 in. t(sec): 25692.935 HPCI Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 28619.435 HPC[ initiation on Low water Level Setpoint for initiation

13.00 in. t(sec)= 28741.185 HPC[ Trip on hi water Level Trip Setpoint = .54E+02 in. 1 4D -

t(sec)= 31835.335 HPCI initiation on Low water Level Setpoint for initiation

= 13.00 in. t(sec)= 31951.635 HPCI Trip on hi water Level Trip Setpoint = .54E÷02 in. t(sec)= 35081.685 HPCI initiation on Low water Level Setpoint for initiation 13.00 in. t(sec)= 35203.035 HPCI Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 38477.385 HPCI initiation on Low water level Setpoint for initiation

= 13.00 in. t(sec)= 38595.185 HPCI Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 41622.135 HPCI initiation on Low water Level Setpoint for initiation 13.00 in. t(sec)= 41724.585 HPCI Trip on hi water level Trip Setpoint = .54E+02 in. t(sec)= 44693.485 HPCI initiation on Low water Level Setpoint for initiation

= 13.00 in. t(sec)= 44806.885 HPCI Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 47918.635 HPCI initiation on Low water Level Setpoint for initiation

13.00 in. t(sec)= 48025.635 HPCI Trip on hi water Level Trip Setpoint .54E+02 in. t(sec)= 51142.335 HPCI initiation on Low water level Setpoint for initiation

13.00 in. t(sec)= 51254.135 HPCI Trip on hi water Level Trip Setpoint = .54E÷02 in. t(sec)= 54456.435 HPCI initiation on Low water Level Setpoint for initiation

13.00 in. t(sec)= 54565.885 HPCI Trip on hi water level Trip Setpoint = .54E+02 in. t(sec)= 57794.985 HPC; initiation on Low water level Setpoint for initiation.

13.00 in. t(sec)= 57906.785 HPCI Trip on hi water Level Trip Setpoint .54E+02 in. t(sec)= 61195.585 HPCI initiation on Low water level Setpoint for initiation 13.00 in. t(sec)= 61306.585 HPCI Trip on hi water level Trip Setpoint = .54E+02 in. t(sec)= 64627.385 HPCI initiation on Low water Level Setpoint for initiation

= 13.00 in. t(sec)= 64739.835 HPC! Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 68105.135 HPCI initiation on Low water Level Setpoint for initiation

= 13.00 in. t(sec)= 68217.485 HPCI Trip on hi water LtveL Trip Setpoint = .54E+02 in.

t(sec)= 71285.835 HPCI initiation on Low water Level Setpoint for initiation

13.00 in. t(sec)= 7138i.385 HPCI Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 74337.085 HPCI initiation on Low water Level Setpoint for initiation.

13.00 in. t(sec)= 74442.585 HPCI Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 77489.285 HPCI initiation on LoW water Level Setpoint for initiation

= 13.00 in. t(sec)= 77589.985 HPCI Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 80625.285 HPCI initiation on Low water Level Setpoint for initiation

= 13.00 in. t(sec)= 80729.535 HPC! Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 83804.035 HPCI initiation on Low water Level Setpoint for initiation

= 13.00 in. t(sec)= 83907.135 HPCI Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 86992.585 HPCI initiation on Low water level Setpoint for initiation

= 13.00 in. t(sec)= 87097.185 HPCI Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 90205.685 HPCI initiation on Low water Level Setpoint for initiation

= 13.00 in. t(sec)= 90310.185 HPCI Trip on hi water Level Trip Setpoint = .54E+02 in. t(sec)= 93431.535 HPCI initiation on Low water Level Setpoint for initiation 13.00 in. t(sec)= 93536.935 HPC1 Trip on hi water level Trip Setpoint = .54E+02 in. t(sec)= 96676.785 HPCI initiation on Low water Level Setpoint for initiation

= 13.00 in. t(sec)= 96782.535 HPCI Trip on hi water level Trip Setpoint = .54E+02 in. t(sec)= 99936.935 HPCI initiation on Low water Level Setpoint for initiation

= 13.00 in.

-la-aIA-ý400 350 300 250 200 150 100 50 10 20 30 Time (hr) Figure 5.10-1 Comparison of SABRE-calculated suppression pool temperature against heat balance calculation (Eqn. 5.10-1) for MSIV-closure event with no decay heat removal.0 E 0 C. C) 0. C PAULi tsb 5.11 Small Break LOCA -Comparison to GE SAFER/GESTR-LOCA Results In this section, SABRE is compared with results from the GE SAFER/GESTR-LOCA Analysis for Susquehanna.'

This benchmarking case compares break flow and reactor level with GE results for a 0.1 f-2 liquid break in the recirculation discharge piping. In this scenario, offsite power becomes unavailable coincident with the LOCA. The GE analysis assumes that MSIVs remain open until they close on low reactor water level. This maximizes the loss of reactor coolant from the RPV. HPCI is assumed to be inoperable, and no credit is taken for RCIC injection.

The water level setpoint for ADS initiation and MSIV closure used in the GE analysis is 366.5" above the bottom of the vessel which corresponds to TAF (-161"). The SABRE calculation is run until level reaches TAF. The kinetics file used in the SABRE calculation corresponds to U2C7 which is the initail power-uprate cycle. Changes made to the base input deck are documented in Table 5.11-1. This SABRE calculation corresponds to SABRE Case 11 in the Computer Case Summary.

In this scenario, a 0.1 ft 2 liquid break occurs at t-0. Feedwater injection coasts down to zero in 5 seconds as a result of the loss of offsite power. Recirculation pumps also trip at t-= because of the loss of offsite power. Although closure of the MSIVs would follow from the LOOP, the isolation valves are maintained open in accordance with the GE calculation.

GE assumes the MSIVs stay open until they close on low RPV level so as to maximize the loss of coolant inventory.

The critical flow out the break is computed using the Homogeneous Equilibrium Model (HEM). In order to be consistent with the GE methodology, a flow multiplier of 1.25 is used for subcooled liquid in the SABRE calculation.

The liquid at the break is subcooled in the time frame considered in this benchmark study. This critical flow model and multiplier are used by GE (GE-NE-187-22-0992, Table 3-1) for nominal LOCA calculations (not Appendix K calculations).

Note that GE uses a multiplier of 1.0 in nominal calculations if the coolant is saturated.

The sequence of events calculated by SABRE for this SBLOCA is presented in Table 5.11-2. Figure 5.11-1 shows a comparison of calculation results for break flow, reactor water level, reactor pressure, and steam line flow. SABRE tends to somewhat overpredict the break flow and rate of level decrease as compared to the GE results, but overall the results show fairly good agreement.

'Pappone, D.C., "SAFER/GESTR-LOCA Analysis Basis Documentation for Susquehanna Steam Electric Station Units 1 and 2," GE-NE-187-22-0992, September, 1993.

rrP%'3"- S~Table 5.11-1 Changes Made to Base Input Deck in Appendix F Problem end time (F.2.1) 175 seconds (Run until level drops to TAF) Hydraulic step size (F.3) Max = Min = 5 msec for t<10 sec Max = Min = 25 msec for t>1Osec Initial core power (F.6.1) 3457 MWt (GE-NE-1 87-22-0992, Table 4-2) reak Area (F. 18) 0.1 ft 2 (GE-NE-1 87-22-0992, Section 6.1.1.3) Break flow coefficient (F.18) 1.25 for subcooled water 1.00 for saturated water (GE-NE-1 87-22-0992, Table 3-1) Time at which manual scram is initiated 0.0 (scram on LOOP) (F.19.6) HPCI operability flag (F.20.1) 0 (HPCI is inoperable)

RCIC operability flag (F.21.1) 0 (RCIC is inoperable)

Trip recirc. pump 'A' on specified time (F.23.1) 0.0 seconds (get recirc pump trip on LOOP) Trip recirc. pump 'B' on specified time (F.23.2) 0.0 seconds (get recirc pump trip on LOOP) Feedwater flow versus time flag (F.24.15) 1 (Feedwater flow is specified in accordance with GE-NE-187-22-0992, Appendix C; FW starts to coast down immediately due to LOOP) Number of points in FW flow versus time table 7 (F.24.16)_

Level setpoint for MSIV closure (F.25.2) -999. (Setpoint is specified to prevent closure of MSIVs on Level 1. In GE LOCA analysis, low level setpoint is TAF, but since problem is run until level reaches TAF there is no need to close MSIVs in SABRE calculation)

ADS operability flag (F.27.1) 0 (Setpoint is specified to prevent ADS actuation on Level 1. In GE LOCA analysis, low level setpoint is TAF, but since problem is rum until level reaches TAF there is no need to actuate ADS in SABRE calculation)

SRV area and actuation set points (F.31.4) set points are specified in accordance with Table 4-5 of GE-NE-187-22-0992.

Gap Conductance (F.36.4) 870 Btu/hr-ft 2-0 F. This value is appropriate for a full core of Siemens Nuclear Power (SNP) 9x9-2 fuel (Table 2.1 of GENE-637-024-0893).

The GE SAFERIGESTR-LOCA Analysis for Susquehanna was carded out for SNP 9x9-2 fuel.

S~/88 Table 5.11-2 Sequence of Events Calculated by SABRE

• Kinetics file is /dOO/appl/sabre3vO/data/u2c7.si mtran.out

• • • SABRE data file is /hore/eamac/sabre_31/input/ec-atws-OSOS/cll.dat

• • • This is not a restart case 1 S A B R E -Version 3.1 Case 11 -0.1 ft2 tiq break LOCA (u2c7) -Benchmark to GE t(sec)= .000 Liquid Break Break Area = .100 ft2 MuLtiplier on break flow 1.250 t(sec)= .000 HPCI is inoperable t(sec)= .000 RCIC is Inoperable t(sec)= .000 Low-Pres Condensate Injection Inop. t(sec)= .000 ADS actuation is defeated t(sec)= .000 Scram initiated on specified time Scram time (sec) = 2.80 t(sec)= .000 Feedwater Flow determined from Table Data in Table ends at t = .10E+10 sec t(sec)= .000 Recirc pump-A trip on specified time Trip delay = .OOE+00 sec t(sec)= .000 Recirc punp-B trip on specified time Trip delay= .OODE+00 sec t(sec)= .870 DW Cooler Trip on Hi DW Press Trip Setpoint = 1.720 psig t(sec)= 2.805 ALL Control Rods Inserted t(sec)= 4.390 Level Setpoint Setdown Setdown occurs when Level drops to 13.00 in. Delay for setpoint setdown = .11E+02 sec 3 0 U. @2 a)Time (98c)900 2 8500 25 50 75 100 25 150 175 Time (sec)@2 @2 0 S -I U S E B 0 Time (see)E a,-e-- GE SABRE Figure 5.11-1 Comparison of SABRE results against GE SAFERIGESTR-LOCA analysis for 0.1 ft 2 liquid break. (SABRE Case 11)CL A. IL cc 175 Tine (Sec)2'0 ,.. .....I. ..I ' ' ' I ..I .....

5.12 Check of SABRE Mass Energy Balances with Core Spray Injection In this test problem, the Core Spray injection table in the SABRE input deck is modified to allow initiation of Core Spray flow at 8000 gpm while the reactor is operating at normal conditions.

The case is run for 250 seconds at which time conditions are essentially at steady state (there are only some minor fluctuations in core power and core flow at this time). The SABRE mass and energy balances are checked against hand calculations to ensure that the mass/energy addition associated with Core Spray injection is accounted for properly in the code. Changes made to the base input deck are documented in Table 5.12-1. This SABRE calculation corresponds to SABRE Case 12 in the Computer Case Summary.

The sequence of events calculated by SABRE for this test problem is presented in Table 5.12-2. In Table 5.12-3, the final general edit (at 250 seconds) for the run is shown. From the information in this general edit, the mass and energy balances for the first control volume in the upper plenum (the volume which receives the Core Spray injection flow) can be checked. Given the gas and liquid flow rates into the first upper plenum computaional cell, and the enthalpies associated with the inlet flows, the out flow of gas and liquid is computed by a hand calculation.

The results of the hand calculation are compared to the values generated by SABRE to ensure that mass and energy are conserved.

The steady-state mass and energy balances for the first upper plenum cell are Wet +We =Wcs +Wgc + WIC +W WB +W, J (5.12-1) hgWgu +hff WB (5.12-2) where Wsu = gas flow exiting Upper Plenum Cell 1 (Lbmlsec), Wt = liquid flow exiting Upper Plenum Cell 1 (Lbm/sec), We C = gas flow exiting Core Cell 27 (Upper Reflector) (Lbm/sec), WC = liquid flow exiting Core Cell 27 (Upper Reflector) (Lbnbrsec), W B = gas flow exiting Bypass Cell 5 (Lbm/sec), WIB = liquid flow exiting Bypass Cell 5 (Lbm/sec), h9 = saturation enthalpy of steam at reactor pressure (Btu/Lbm), hf = saturation enthalpy of water at reactor pressure (Btu/Lbm), W,, = core spray injection flow (Lbmn/sec), and hcs, = enthalpy of core spray injection flow (Btu/Lbm).

Solving (5.12-1) and (5.12-2) for Wgu and Wju gives aO= d, -al 2 d 2 (5.12-3) all a 2 2 -a 2 1 a 1 2 and a, 1 d 2 -a 2 1 d 1 (5.12-4) a 1 1 a. -a 2 l a, 2 where alI =-- 1, a, 2 = 1, a 2 , = h9, a 2 2 = hf d 1 = Wcs +Wgc +WTc +WgB +WB and d 2 = hcs Wcs + hg Wgc +hfWc + h WSB +hf WtB" The temperature of the water in the suppression pool for this case is 90 'F. The density of water at this temperature is 62.11 Lbm/ifO. A Core Spray injection rate of 8000 gpm corresponds to (8000gal ft min 62.11 Lbm 1 L WcN l07.1 Lbm/sec.

7i-ý.4805ga1" 60-sec ft From Table 5.12-3, reactor pressure is 1021.575 psia. The gas phase and liquid phase enthalpies at this pressure are 1192.122 Btu/Lbm and 545.865 Btu/Lbm, respectively.

Also, the core spray injection enthalpy is approximately the saturation enthalpy of 90TF water which is 58.018 Btu/Lbm. Using these specific enthalpies and the mass flow rates from Table 5.12-3 in Eqs. (5.12-3) and (5.12-4) results in Wgu (hand calculation)

= 3040.512 Lbm/sec, and W~u (hand calculation)

= 25853.32 Lbm/sec.

The SABRE calculated values from Table 5.12-3 are Wgu (SABRE) = 3041.943 Lbm/sec, and Ww (SABRE) = 25851.323 Lbm/sec.

The differences between the hand calculation and SABRE are Agu = (100%) (3041.943

-3040.512)/(3040.512)

= 0.05% Al = (100%) (25851.323

-25853.32)/(25853.32)

= -0.01% The small differences indicate that the mass and energy addition associated with the Core Spray injection is properly accounted for in SABRE.

For Core Spray injection, the injection enthalpy used in SABRE is the saturation enthalpy at the suppression pool temperature which is 58.018 Btu/Lbm in this test problem. In this problem, Core Spray injection is artificially initiated at rated reactor pressure to cheek the nodal mass and energy balances.

If Core Spray injection could actually occur at rated reactor pressure, the injection enthalpy should be corrected for the added pressure imparted by the pump. However, in reality Core Spray only injects at low reactor pressures, and under these conditions, the enthalpy correction associated with the pump head is small and can be neglected.

Table 5.12-1 Changes Made to Base SABRE 9x9 Input Deck in Appendix F Problem end time (F.2.1) 250 seconds Core Spray injection table (F.22a.2)

Core Spray Flow (gpm) AP (psi) 8000. 0. 8000. 1000. 8000. 10000.

S/907/ Table 5.12-2 Sequence of Events Calculated by SABRE *** Kinetics fire is /dOO/appL/sabre3vO/data/u2cC9.simtran°.out

• • • SABRE date fiLe is /home/eamac/sabre_31/input/ec-atws-O505/cl 2.dat

• This is not a restart case 1 S A B R E -Version 3.1 (12) Check Mass/Energy BaLance with Core Spray (U2C9) t(sec)= .000 Low-Pres Condensate Injection Inop. t(sec)= .000 Initiation of Core Spray FLow Reactor Press = 1035.30 psig Supp Chamber Press = .50 psig Table 5.12-3 General Output Edit at 250 seconds I SABRE -Version3.1 0 (12) Check Mass/Energy Balance with Core Spray (U2C9) ** problem time and solver parameters time (sec) no. of steps 250.01 16353 step size (sec) .0300 0"** 1-D Kinetics Results Volume No. Axial Power profile (fiss) 25 .1665 24 .4933 23 .8539 22 1.0362 21 1.1744 20 1.2737 19 1.3401 18 1.3830 17 1.4008 16 1.4204 15 1.3918 14 1.3471 13 1.2883 12 1.2177 11 1.1383 10 1.0568 9 .9762 8 .9020 7 .8400 6 .7933 5 .8022 4 .8253 3 .8086 2 .7177 1 .3524 Axial Power profile (decay) .1514 .4487 .7770 .9439 1.0715 1.1646 1.2291 1.2736 1.2968 1.3242 1.3091 1.2816 1.2435 1.1971 1.1447 1.0923 1.0427 .9996 .9683 .9536 .9530 .9551 .9363 .8330 .4093 Fast FLux .149008E+01

.371805E+01

.580333E+01

.723971E+01

.820228E+01

.882661E+01

.918388E+01

.933023E+01

.931393E+01

.918568E+01

.884869E+01

.836858E+01

.780092E+01

.718252E+01

.654420E÷01

.591683E+01

.533153E+01

.481615E+01

.439973E+01

.410874E+01

.392580E+01

.379520E+01

.358695E+01

.299384E+01

.172827E+01 Therm Flux AbsXSThrm Liq_Boron (cm**-1) .4776567E+00 .OOOOOE+00

.7767447E+00

.0000E+00

.8904190E+00 .OOOOOE+00

.1106688E+01 .OOOOOE+00

.1287336E+01

.00000E+00

.1423535E+01 .OOOOOE+00

.1524954E+01

.00000E+00

.1597801E+01 .OOOOOE+00

.1641539E+01

.0000E+00

.1606372E+01 .OOOOOE+00

.1603438E+01 .OOOOOE+00

.1583638E+01

.00000E+00

.1546144E+01 .OOOOOE+00

.1493503E+01 .OOOOOE+00

.1428661E+01 .OOOOOE+00

.1357483E+01

.0000E+00

.1281727E+01

.0000OE+00

.1208923E+01

.0000E+00

.1146666E+01 .OOOOOE+00

1097549E+01 .OOOOOE+00

.1122467E+01 .OOOOOE+00

.1158305E+01

.0000E+00

.1108408E+01

.00000E+00

.9299905E+00

.00000E+00

.6823881E+00

.00000E+00 4

1*** fuel/clad conditions volume no. avg fuel temp (degf) 25 610.968 24 739.025 23 893.477 22 977.697 21 1044.502 20 1094.160 19 1128.255 18 1150.747 17 1160.327 16 1171.052 15 1156.441 14 1133.665 13 1104.058 12 1069.122 11 1030.614 10 992.013 9 954.792 8 921.305 7 893.976 6 874.038 5 877.678 4 886.301 3 875.719 2 832.060 1 678.794 clad temp clad surface heat flux (degf) temp (btu/ft2-sec) 554.433 551.679 5.496 563.054 554.926 16.282 571.388 557.371 28.183 575.389 558.408 34.204 578.356 559.134 38.769 580.459 559.630 42.051 581.858 559.951 44.252 582.761 560.156 45.683 583.142 560.242 46.287 583.565 560.336 46.959 582.988 560.207 46.042 582.077 560.001 44.598 580.869 559.724 42.694 579.409 559.384 40.407 577.753 558.989 37.834 576.039 558.570 35.197 574.332 558.141 32.596 572.746 557.731 30.206 571.413 557.378 28.219 570.418 557.109 26.748 570.601 557.159 27.018 570.132 556.330 27.734 567.064 553.521 27.175 562.274 550.227 24.123 55 57 547.878 11.893 film coefficient (btu/hr-ft2-f) 4395.0 7564.6 9952.3 10964.0 11672.8 12156.8 12471.0 12670.9 12754.4 12846.8 12720.7 12519.6 12249.5 11916.9 11531.2 11122.1 10703.3 10303.5 9958.8 9695.7 9744.5 9797.9 9501.9 8793.0 4240.1 W w tcrit (deg f) 570.52 570.56 570.67 570.88 571.13 571.42 571.75 572.09 572.46 572.84 573.24 573.64 574.03 574.42 574.79 575.15 575.49 575.81 576.11 576.40 576.67 576.85 576.85 576.85 576.85 t (rewet) (deg f) 642.90 642.90 642.90 642.90 642.90 642.90 642.90 642.90 642.90 642.90 642.90 642.90 642.90 642.90 642.90 642.90 642.90 642.90 642.90 642.90 642.90 642.90 642.90 642.90 642.90 ivolume no. fuel templ 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 (degf) 624.86 785.11 983.66 1094.06 1182.60 1248.93 1294.71 1325.00 1337.94 1352.43 1332.69 1301.99 1262.20 1215.43 1164.13 1112.96 1063.90 1019.98 984.31 958.38 963.11 974.58 961.57 905.77 710.67 fuel temp2 (degf) 597.07 692.91 803.23 861.25 906.30 939.28 961.68 976.36 982.59 989.55 980.07 965.22 945.80 922.71 897.00 870.97 845.61 822.56 803.58 789.64 792.19 797.96 789.81 758.30 646.89 heat gen (btu/ft3-sec) 879.31 2604.86 4508.80 5472.04 6202.38 6727.41 7079.60 7308.40 7405.02 7512.61 7365.85 7134.81 6830.24 6464.43 6052.81 5630.93 5214.94 4832.63 4514.79 4279.50 4322.78 4437.43 4348.03 3859.79 1895.22 heat genl (btu/ft3-sec) 791.38 2344.37 4057.92 4924.84 5582.14 6054.67 6371.64 6577.56 6664.52 6761.35 6629.26 6421.33 6147.22 5817.99 5447.53 5067.84 4693.44 4349.37 4063.31 3851.55 3890.50 3993.68 3913.23 3473.81 1705.70 heat gen2 (btu/ft3-sec) 967.24 2865.34 4959.68 6019.24 6822.62 7400.15 7787.56 8039.23 8145.52 8263.87 8102.43 7848.29 7513.27 7110.87 6658.09 6194.03 5736.43 5315.90 4966.26 4707.45 4755.06 4881.17 4782.83 4245.77 2084.74 1"** steam dome fluid conditions voL no. density (Lb/ft3) 1 2.280 void fract temp (deg f) 1.000 547.431 0"** separator fluid conditions vol no. density (tb/ft3) 1 19.045 void fract temp (deg f) .619 547.177 0"** riser fluid conditions vol no. density void fract (Lb/ft3) 1 18.973 .621 temp (deg f) 547.177 ww enthatpy (btu/lb) 1192.135 w(bubl rise) (lb/sec) .000 w(stm line) (lb/sec) 3041.938 w(cond) (lb/sec) .000 cond eff (M) .000 w-break (tbm/s) .00 enthalpy (btu/Lb) 593.774 enthaLpy (btu/Lb) 594.080 boron conc (ppm) .000 boron conc (ppm) .000 w-gas (Lb/sec) 3041.941 3041.937 gamma heat (btu/sec)

.000 gamma heat (btu/sec)

.000 w-liquid (lb/sec) 25851.293 25851.380 w-liquid (lb/sec) 25851.380 25851.515 4 NI w-gas (lb/sec) 3041.937 3041.932 0*** upper pLenum fLuid conditions voL no. density void fract temp enthaLpy boron conc gamma heat w-liquid w-gas (tb/ft3) (deg f) (btu/Lb) (ppm) (btu/sec) (Ib/sec) (Lb/sec) 3 19.106 .618 547.177 593.511 .000 .000 25851.515 3041.932 2 21.125 .572 547.177 585.759 .000 .000 25851.680 3041.924 1 21.125 .572 547.177 585.759 .000 .000 25851.323 3041.943 23910.489 3876.241 1"** by-pass fluid conditions voL no. density void fract temp enthaLpy boron conc gamma heat w-Liquid w-gas (Lb/ft3) (deg f) (btu/Lb) (ppm) (btu/sec) (tb/sec) (tb/sec) 5 42.227 .093 547.177 549.090 .000 5046.392 2499.201 24.909 4 43.685 .060 547.177 547.865 .000 14636.932 2507.095 17.097 3 46.392 .000 546.076 544.430 .000 14435.574 2524.296 .000 2 46.716 .000 541.614 538.708 .000 10566.370 2524.306 .000 1 46.948 .000 538.318 534.520 .000 5747.670 2524.314 .000 2515.479 .000 0*** core fluid conditions voL no. density void fract temp enthaLpy boron conc gamma heat w-Liquid w-gas (Lb/ft3) (deg f) (btu/Lb) (ppM) (btu/sec) (ib/sec) (Lb/sec) 27 16.548 .676 547.177 606.055 .000 .000 21411.288 3851.332 26 15.861 .692 547.177 610.109 .000 333.722 21410.798 3851.356 25 15.940 .690 547.177 609.624 .000 988.617 21439.565 3822.364 24 16.179 .684 547.177 608.187 .000 1711.219 21525.362 3736.448 23 16.610 .675 547.177 605.706 .000 2076.795 21673.900 3587.731 22 17.161 .662 547.177 602.705 .000 2353.979 21854.244 3407.240 21 17.830 .647 547.177 599.318 .000 2553.242 22058.730 3202.654 20 18.613 .629 547.177 595.659 .000 2686.908 22280.554 2980.749 19 19.512 .609 547.177 591.824 .000 2773.740 22514.007 2747.224 18 20.532 .585 547.177 587.878 .000 2810.409 22755.050 2506.149 17 21.676 .559 547.177 583.891 .000 2851.241 22999.301 2261.885 16 22.973 .530 547.177 579.854 .000 2795.538 23247.123 2014.071 15 24.405 .498 547.177 575.898 .000 2707.851 23490.138 1771.096 14 25.951 .462 547.177 572.111 .000 2592.254 23725.559 1535.740 13 27.618 .425 547.177 568.508 .000 2453.414 23950.957 1310.430 12 29.432 .383 547.177 565.047 .000 2297.186 24164.311 1097.187 11 31.393 .339 547.177 561.758 .000 2137.068 24364.110 897.521 10 33.461 .292 547.177 558.707 .000 1979.180 24550.029 711.771 9 35.679 .241 547.177 555.827 .000 1834.078 24722.274 539.742 8 38.035 .188 547.177 553.136 .000 1713.440 24881.955 380.323 7 40.533 .131 547.177 550.624 .000 1624.135 25031.177 231.390 6 43.299 .068 547.177 548.181 .000 1640.565 25172.669 90.218 5 46.388 .000 546.140 544.513 .000 1684.081 25263.243

.000 4 46.600 .000 543.225 540.768 .000 1650.152 25263.268

.000 3 46.805 .000 540.350 537.099 .000 1464.856 25263.279

.000 2 46.985 .000 537.780 533.840 .000 719.268 25263.339

.000 1 47.073 .000 536.513 532.240 .000 .000 25263.177

.000 27778.658

.000 1l** Lower plenum fluid conditions vot no. density void fract (Lb/ft3) 2 47.073 .000 1 47.069 .000 0*** jet pump fluid conditions vot no. density (lb/ft3) 0 47.065 temp (deg f) 536.513 536.573 void fract temp, (deg f) .000 536.634 0"** downcomer fluid conditions voL no. density void fract (Lb/ft3) 1 47.065 .000 0*** pressure drop results jet pump inlet (psi) 11.11 outlet (psi) 1.18 elevation (psi) -5.39 wall frict (psi) 1.68 spac frict (psi) spaciaL acc (psi) .00 1*** misc. reactor parameters downcomer subcooting (btu/Lbm) downcomer collapsed Level (inch) steam dome pressure (psia) steam dome fluid volume (ft3) downcomer fluid volume (it3) pressure regulator setpoint (psia) downcomer bubble rise vet (ft/sec) steam flow exiting vessel (lbm/sec) number of srvs open steam condensation rate (lbm/sec) condensation efficiency core power (mw) core power (% of rated) core inlet flow (mlb/hr) by-pass intet flow (mlb/hr) total vessel fluid mass (Lbm) crd outlet enthaLpy (btu/tbm)temp (deg f) 536.634 tower plenum .00 5.40 2.27 .02 .00= 13.461 = 50.796 = 1021.575 = 8420.787 = 5913.213 -966.434 = 1.157 = 3041.938 -0 = .000 = .000 -3038.40 = 88.30 -90.92 = 9.06 -529466.18

= 532.43'V enthalpy (btu/tb) 532.240 532.316 enthalpy (btu/tb) 532.392 enthaLpy (btu/Lb) 532.392 boron conc (ppm) .000 .000 boron conc (ppm) .000 w(inject) (tb/sec) 1928.169 gamma heat (btu/sec)

.000 .000 gamma hea (btu/sec)

.000 h(inject) (btu/tb) 351.879 w-gas (tb/sec) .000 .000 .000 w-gas (lb/sec) .000 .000 mixture tvt (inches) 50.796 w-liquid (Lb/sec) 27778.658 27778.689 27778.718 t w-liquid (Lb/sec) 27778.723 27778.718 subcool (btu/Lb) 13.461 upper plenum .00 .00 .71 .00 .03 w-break (tbm/s) .00 core .81 1.09 3.01 4.02 2.72 .79 boron (ppm) .00 by-pass 7.71 .00 4.67 .00 .00 riser 1.54 .00 1.34 1.74 .04 separator 3.79 .00 .82 .36 -.01 4"