ML16057A812

Response to Request for Additional Information Regarding License Amendment Request to Adopt Dominion Core Design and Safety Analysis Methods and to Address the Issues Identified in Westinghouse Documents NSAL-09-5, Rev. 1

ML16057A812
Person / Time
Site:	Millstone
Issue date:	02/25/2016
From:	Mark D. Sartain Dominion, Dominion Nuclear Connecticut
To:	Document Control Desk, Office of Nuclear Material Safety and Safeguards
References
16-011A, CAC MF6251
Download: ML16057A812 (76)

Text

Dominion Nuclear Connecticut, Inc.

5000 Dominion Boulevard, Glen Alien, VA 23060 DominuIion Web Address: www.dom.com February 25, 2016 U.S. Nuclear Regulatory Commission Serial No.

16-011lA Attention: Document Control Desk NLOS/WDC R0 Washington, DC 20555 Docket No.

50-423 License No.

NPF-49 DOMINION NUCLEAR CONNECTICUT, INC.

MILLSTONE POWER STATION UNIT 3 RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION REGARDING LICENSE AMENDMENT REQUEST TO ADOPT DOMINION CORE DESIGN AND SAFETY ANALYSIS METHODS AND TO ADDRESS THE ISSUES IDENTIFIED IN WESTINGHOUSE DOCUMENTS NSAL-09-5, REV. 1, NSAL-15-1, AND 06-1C-03 (CAC NO. MF6251)

By letter dated May 8, 2015, Dominion Nuclear Connecticut, Inc. (DNC) submitted a license amendment request (LAR) for Millstone Power Station Unit 3 (MPS3).

The proposed amendment would revise the Technical Specifications (TS) to enable use of the Dominion nuclear safety analysis and reload core design methods for MPS3 and address the issues identified in three Westinghouse communication documents.

In a letter dated January 8, 2016, the Nuclear Regulatory Commission (NRC) transmitted a request for additional information (RAI) to DNC related to the LAR.

The RAI contained 18 questions. In a letter dated January 28, 2016, DNC responded to RAI Questions RAI-1 through RAI-8, RAI-13, and RAI-18. Attachment I is the DNC response to RAI Questions RAI-9 through RAI-12 and RAI-14 through RAI-16. DNC plans to submit the response to the remaining RAI question, RAI-17, by March 31, 2016.

During preparation of the RAI responses, Dominion identified several minor discrepancies within Attachment 5 of the May 8, 2015 LAR. In response to RAI-9, DNC has updated the analyses discussed in the RETRAN benchmark information originally provided in Attachment

5. The RETRAN benchmark information is not used in any analysis of record. Therefore, the discrepancies do not impact the no significant hazards consideration determination provided in the May 8, 2015 LAR. Attachment 2 provides an update to Attachment 5 of the May 8, 2015 LAR which corrects the discrepancies and includes the revised results in response to RAI-9.

If you have any questions regarding this submittal, please contact Wanda Craft at (804) 273-4687.

Sincerely,

Vicki, L.:di Hull

.~NOTARY PUBILIC Mark D. Sartain I

Commonwe~alth of virgin~ia Vice President -Nuclear Engineering My om....o

...........3.l201 COMMONWEALTH OF VIRGINIA)

COUNTY OF HENRICO)

The foregoing document was acknowledged before me, in and for the County and Commonwealth aforesaid, today by Mark 0. Sartain, who is Vice President -

Nuclear Engineering of Dominion Nuclear Connecticut, Inc.

He has affirmed before me that he is duly authorized to execute and file the foregoing document in behalf of that Company, and that the statements in the document are true to the best of his knowledge and belief.

Acknowledged before me this *Jda

• "*,*L_#,z 2016.

My Commission Expires:

S

[-*v/

• /

Notary Public

~Serial No. 16-011A Docket No. 50-423

~Page 2 of 2 Commitments made in this letter: None Attachments:

1. Response to Request for Additional Information Regarding License Amendment Request to Adopt Dominion Core Design and Safety Analysis Methods and to Address the Issues Identified in Westinghouse Documents NSAL-09-5, Rev. 1, NSAL-15-1, and 06-1C-03 (CAC No. MF6251) -

RAI Questions RAI-9 through RAI-12 and RAI-14 through RAI-16

2. RETRAN Benchmarking Information -Updated cc:

U.S. Nuclear Regulatory Commission Region I 2100 Renaissance Blvd, Suite 100 King of Prussia, PA 19406-2713 R. V. Guzman Senior Project Manager U.S. Nuclear Regulatory Commission One White Flint North, Mail Stop 08-C2 11555 Rockville Pike Rockville, MD 20852-2738 NRC Senior Resident Inspector Millstone Power Station Director, Radiation Division Department of Energy and Environmental Protection 79 Elm Street Hartford, CT 06106-5127

Serial No. 16-O11A Docket No. 50-423 ATTACHMENT I RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION REGARDING LICENSE AMENDMENT REQUEST TO ADOPT DOMINION CORE DESIGN AND SAFETY ANALYSIS METHODS AND TO ADDRESS THE ISSUES IDENTIFIED IN WESTINGHOUSE DOCUMENTS NSAL-09-5. REV. I. NSAL-15-I, AND 06-1C-03 (CAC NO. MF6251)

RAl QUESTIONS RAI-9 THROUGH RAl-12 AND HAl-14 THROUGH RAl-16 DOMINION NUCLEAR CONNECTICUT, INC.

MILLSTONE POWER STATION UNIT 3

t.

Serial No. 16-011IA Docket No. 50-423, Page 1 of 11 By letter dated May 8, 2015, Dominion Nuclear Connecticut, Inc. (DNC) submitted a license amendment request (LAR) for Millstone Power Station Unit 3 (MPS3). The proposed amendment would revise the Technical Specifications (TS) to enable use of the Dominion nuclear safety analysis and reload core design methods for MPS3 and address the issues identified in three Westinghouse communication documents.

In a letter dated January 8, 2015, the Nuclear Regulatory Commission (NRC) transmitted a request for additional information (RAI) to DNC related to the LAR. The RAI contained 18 questions. In a letter dated January 28, 2016, DNC responded to RAI Questions RAI-1 through RAI-8, RAI-13, and RAI-18. This attachment provides DNC's response to RAI Questions RAI-9 through RAI-12 and RAI-14 through RAI-16.

DNC plans to submit the response to the remaining RAI question, RAI-17, by March 31, 2016.

RAI - 9 (SRXB):

Reanalysis for the RETRAN Benchmarking Cases The licensee indicated in an e-mail dated June 30, 2015 (ADAMS Accession No. ML15349A808) that it identified a discrepancy between the MPS3 RETRAN Base Model Pressurizer Shell Heat Conductor and the Dominion RETRAN topical report (TR).

The MPS3 RETRAN base input deck models the pressurizer shell as a heat conductor, which differs from TR, VEP-FRD-41, which states that "Dominion continues to model the non-equilibrium wall as an adiabatic surface." Each of the five benchmarking cases supporting the LAR were reanalyzed with needed correction.

Discuss the results of the reanalysis for the five benchmarking cases, and provide the modified graphs showing the changes and text for the affected cases, including the affected loss of normal feedwater benchmark case.

DNC Response The results of the reanalysis of the five benchmarking cases originally submitted are included in an update to Attachment 5 of the May 8, 2015 LAR provided in to this letter. The updated Attachment 5 is provided with the changes noted by a change bar in the right hand margin of the affected pages. Most of the plots of key parameters are unchanged as there was no observable difference in the plotted values. Some selected plots for the Loss of Load event are revised because they exhibited differences in the period after the occurrence of peak reactor coolant system (RCS) pressure. The description of some items in input summary tables and results tables have been clarified to conform with the details of the analyses. The same conclusions originally stated in Attachment 5 of the May 8, 2015 LAR continue to be supported by the updated benchmarking analyses.

Serial No. 16-011IA Docket No. 50-423, Page 2 of 11 RAI -10 (SRXB):

Loss of Load (LOL)

Benchmark Analysis RCS Pressurization Rate On page 9 of Attachment 5 to the LAR, the licensee indicated that "the Dominion case trips slightly earlier than the FSAR [final safety analysis report] data because of the higher RCS [reactor coolant system] pressurization rate", and Table 4.1-2 showed that for the L OL [loss of load/turbine trip] analysis, the calculated peak RCS pressure for the Dominion case is 2, 717.19 psia [pounds-per-square-inch, absolute],

which is 12.22 psi lower than the peak pressure of 2, 729.41 psia for the FSAR case.

Discuss the differences of the models, input parameters or assumptions used in the LOL analyses for the Dominion case and the FSAR case that will result in:

(1) a higher RCS pressurization rate before the reactor trip for the Dominion case and; (2) a lower peak RCS pressure for the Dominion case against the ESAR case discussed above.

DNC Response The peak pressure value of 2,729.41 psia referenced in the RAI question above has changed to 2,705 psia in the updated benchmarking analysis provided in Attachment 2.

The loss of load/turbine trip (LOL) transient causes a sudden reduction in steam flow, resulting in an increase in the steam generator (SG) secondary pressure and a corresponding increase in RCS temperature and pressure. During the early phase of the LOL benchmark analysis, the Dominion rate of RCS pressurization is similar to the FSAR case. However, the Dominion case begins to pressurize slightly earlier as the primary side fluid expands into the pressurizer. The slightly earlier pressurization is attributed to differences in the SG primary-to-secondary heat transfer associated with the Dominion single-node steam generator (SNSG) secondary compared to the FSAR multi-node steam generator (MNSG) secondary.

For the SNSG, the secondary-side temperature corresponds to the saturation temperature for the secondary-side pressure, and will therefore increase with any increase in pressure.

The MNSG is subdivided into regions that may be either saturated or subcooled. The MNSG tends to better maintain heat transfer during transient conditions due to increased noding and modeling of dynamic effects for the liquid/vapor flow through the tube bundle.

These effects are expected to be small for the rapid changes associated with the LOL event, but do result in a slightly earlier heatup for the SNSG and associated increase in primary-side pressure.

The subsequent rate of pressurization is similar for both cases. However, for the Dominion case, because the pressure increase starts earlier, the reactor trip on high pressurizer pressure occurs slightly earlier.

Serial No. 16-011A Docket No. 50-423, Page 3 of 11 The peak RCS pressure, which occurs after the reactor trip, is closely related to the response of the pressurizer safety valves (PSV). It is noted here that the main steam safety valves (MSSV) actuate after the time of peak RCS pressure and, therefore, are not a factor for peak RCS pressure.

As shown in Figure 4.1-1 of updated (provided in Attachment 2 of this letter), the pressurizer pressure is modulated over a narrow band by the PSVs for the FSAR case, achieving pressures that are slightly higher than the Dominion case which has a relatively flat pressure profile when the PSVs open. Since the LOL event results in a very rapid pressure increase exceeding 60 psi/sec, even small differences in PSV response (e.g., delays, opening profiles, etc.) can noticeably affect the peak pressure. These differences are even more pronounced in the RCS cold leg and reactor vessel lower plenum where peak pressures exceed 2700 psia and are affected by differences in loop response (RCS loop, reactor vessel, and surge line loss coefficients, reactor coolant pump head dynamics, etc.).

Given the large and sudden increase in pressure from the initial value of 2200 psia for the pressurizer, the differences in peak pressure are relatively small.

RAI -11 (SRXB):

Locked Rotor (LR)

Benchmark Analysis Over-Pressurization and DNB On page 14, the licensee stated for the LR analysis that, based on the data comparison between the Dominion case and FSAR case, "both the initial under-prediction of the heat flux response, followed by an over-prediction during the rod insertion is indicative of the fuel rod heat transfer being modeled differently in the vendor methods than in the Dominion model."

It further stated that "the over-prediction of both nuclear power and heat flux will lead to conservative results at the limiting point in the transient for both RCS over-pressurization and DNB during rod insertion."

Discuss the differences of the fuel rod heat transfer models used in the LR analyses for the Dominion case and the FSAR case, and justify a higher peak RCS pressure of 2,680.75 psia and a higher peak cladding temperature (PCT) of 1,760.0 0F predicted for the Dominion case, compared to lower corresponding values of 2, 616.65 psia and 1, 718.3 °F (shown in Table 4.2-2 and Table 4.2-4, respectively) for the FSAR case.

DNC Response The Dominion calculated RCS peak pressure value is 2,680 psia and the peak cladding temperature is 1,760 °F in the updated benchmarking analysis provided in.

The Locked Rotor (LR) event is initiated by the instantaneous seizure of a reactor coolant pump rotor resulting in a rapid reduction in flow through the affected RCS loop, reduction in heat transfer to the SG secondary with a resulting temperature and

Serial No. 16-011IA Docket No. 50-423, Page 4 of 11 pressure increase of the RCS. As noted in the updated Attachment 5 Section 4.2 Summary for the LR benchmark transient, the response differences between the Dominion and FSAR cases are primarily attributable to loop friction losses and fuel rod modeling differences. Additional details relative to the higher peak ROS pressure and higher POT temperature observed for the Dominion cases are provided below.

For the benchmark analysis, when the LR event occurs, flow begins to decrease through the faulted RCS loop and reactor vessel core as shown on updated, Figures 4.2-3 and 4.2-5. About one second into the event, the reactor trips on low RCS flow, the unfaulted loops' reactor coolant pumps (RCPs) begin to coast down, and flow reversal occurs in the faulted loop.

It is during this interval between one and two seconds, that the rate of pressure increase is reduced and most of the pressure deviation develops between the FSAR and Dominion cases.

The RCS cold leg and reactor vessel lower plenum pressures trend toward the pressurizer pressure as shown on the figure below where the Dominion pressurizer pressure response (which is not available for the FSAR case) is added to Attachment 5, Figure 4.2-1, for reference. As shown on the figure, after the RCPs trip and flow decreases, the pressure difference between the cold legs and the pressurizer is reduced.

The differences between the FSAR and Dominion pressure responses suggests that there are differences in the loss coefficients (RCS loops, reactor vessel, and surge line) and related RCP dynamics contributing to much of the observed difference in peak pressure. It is noted that most of the difference develops before there is any appreciable reduction in reactor core power.

After this time interval, which ends at about two seconds, the rate of pressure increase returns to nearly the same value as before the RCPs tripped and the difference between the FSAR and Dominion pressures remains relatively unchanged through the point of peak pressure. The opening response for the PSVs will also have an effect on peak pressure and become noticeable during the LR event when pressure is changing very rapidly. Results of a sensitivity case in which the difference in peak pressure between the FSAR and Dominion cases was reduced, are discussed below.

Serial No. 16-011A Docket No. 50-423, Page 5 of 11 Figure 4.2-1 Supplement LR - Reactor Vessel Lower Plenum Pressure 2700.00 2600.00 2500.00

! 2400.00 2:3007.00 22007.003 2100.00 4 0.00 1.00 2.00 3.00 4.00 Time (sac) 5.00 Following the reactor trip, the control rods insert in approximately 1-4 seconds. The most significant reductions in core power occur after about 2 seconds, after which time core power remains higher for the Dominion case until both peak ROS pressure and peak POT have occurred.

This power response is affected by reactivity feedback from fuel Doppler effects. For the Attachment 5 benchmark analysis to be consistent with current FSAR methods, Doppler feedback effects for both cases were modeled using a Doppler power coefficient (DPC).

The Dominion method (VEP-FRD-41-P-A), however, uses a Doppler temperature coefficient (DTC) instead of a DPC. In order to better understand the difference in response between the FSAR and Dominion cases, a sensitivity case was performed using the Dominion DTC model. The resulting power response is shown below as a supplement to Figure 4.2-6 from Attachment 5. As shown, the agreement in core power, particularly after reactor trip, is much closer to the FSAR core response. In addition, the peak ROS pressure for the Dominion case was reduced from 2680 psia to 2666 psia and the subsequent hot spot model case results in a reduction in peak POT from 1760°F to 1739°F. It should be noted that the FSAR method also includes certain proprietary modeling approaches that modify heat removal from the fuel rods which likely contributed to the different response. Nevertheless, the sensitivity case indicates that the Dominion method produces results that compare favorably with the FSAR response given the very rapid primary-side heatup associated with the LR event.

Serial No. 16-011A Docket No. 50-423, Page 6 of 11 Figure 4.2-6 Supplement LR - Nuclear Power 0.80 Dom inion 0.6 z

0.40 0.20 000.00 1 0(0 2.(00 30(0 4.00 5.00 Tim~e (sec)

Serial No. 16-011IA Docket No. 50-423, Page 7of 11 RAI -12 (SRXB):

Loss of Normal Feedwater Benchmark Analysis Pressurizer Water Volume Response The pressurizer water volume response shown in Figure 4.3-6 indicated that Dominion analysis predicts the same trends as the FSAR data, but calculates lower values in the period from 63 to 900 seconds, followed with a strong in-surge during the second heat-up period in the transient. The calculated maximum water volume of 1588.96 ft3 is lower than the FSAR case of 1730.85 ft3.

The licensee indicated the deviations of the pressurizer water volume response are attributed to differences in the main steam safety valves (MSSV) modeling, as well as differences in the pressurizer spray models.

Discuss the differences in the MSSV model and pressurizer spray model used in the Dominion analysis and the ESAR case and justify the deviations discussed above for the pressurizer water volume response.

DNC Response The Dominion calculated maximum water volume is 1610 ft3 in the updated benchmarking analysis provided in Attachment 2.

The loss of normal feedwater (LONF) event results in a reduction of SG secondary-side fluid mass and challenges the ability to adequately remove decay heat and stored energy from the primary side. For the benchmark analysis, the SG secondary pressurizes until steam release occurs through the MSSVs following a reactor trip on low-low SG level. Auxiliary feedwater flow (AFW) flow is initially unable to adequately remove the primary-side energy resulting in a steady loss of SG fluid mass until decay heat is sufficiently reduced, about 2000-2500 seconds into the event. The primary-side temperature and corresponding fluid volume respond to the balance between the generation of decay heat and the removal of energy through the SG secondary via the MSSVs.

The Dominion MSSV model includes the effect of blowdown, which is not modeled for the FSAR case, resulting in a lower SG pressure and saturation temperature between approximately 70 and 1200 seconds, as shown in Figure 4.3-3. During this time, the difference between the SG pressure for the Dominion case and the FSAR case is about 50 psi. This corresponds to a saturation temperature difference of about 5°F, consistent with the primary-side temperature difference and associated pressurizer fluid volume early in this phase.

For the Dominion case, after 1200 seconds the MSSVs begin to cycle open and closed, raising the average pressure, but still remaining somewhat lower than the FSAR case. In addition, the lower SG pressures for the Dominion case result in slightly greater AFW flow, which varies with SG back-pressure, also contributing to differences in primary-side cooling. During the heatup phase, which lasts from about 200 to 2500 seconds, the increase in primary-side temperature and related fluid volume is greater for the Dominion case, but these values are lower at the start of this interval resulting in peak values that are lower.

Serial No. 16-011IA Docket No. 50-423, Page 8 of 11 Pressurizer sprays, heaters, and PORVs are assumed to function normally for the LONF event since this assumption yields more conservative results.

During the heatup phase, the pressurizer liquid volume increases significantly for both cases; however, the resulting pressure is more suppressed for the FSAR case.

The Dominion case results in opening of the pressurizer PORVs while for the FSAR case, pressurizer sprays alone are able to contain pressure. This could be attributable to differences in the condensation or spray models and/or higher assumed spray flow rates. It is acknowledged that although a conservative spray model response can contribute to higher pressurizer liquid volumes; this effect is considered to be small.

RAI - 14 (SRXB):

Control Rod Bank Withdrawal at Power (RWAP) Benchmark Analysis - Higher Core Power Rate of Increase The core power response in Figure 4.5-1 shows that for the RWAP 1 pcm/sec case, its rate of increase for the Dominion model is greater than the FSAR data. The faster power increase rate leads to the Dominion modeling tripping on high neutron flux at about 73 seconds, and the lower power increase rate for the FSAR case results in a reactor trip on an 0 TA T [Overtemperature delta T] signal at about 93 seconds.

Discuss the differences of the nodding, input parameters, models or assumptions used in analyses of the RWAP I pcm/sec case for the Dominion case and the ESAR case, and justify the greater increase of the core power rate observed in the analysis of the Dominion case.

DNC Response The Dominion case trips on high neutron flux at about 74 seconds in the updated benchmarking analysis provided in Attachment 2.

The RWAP event involves the inadvertent addition of core reactivity caused by the withdrawal of rod cluster control assembly (RCCA) banks. The RWAP benchmark case referenced in RAI-14, assumes a relatively slow reactivity insertion rate of 1 pcm/sec, resulting in a steady increase in core power and related primary system heatup. The power response, in turn, is affected by moderator and Doppler reactivity feedback effects.

For the benchmark analysis, moderator reactivity feedback is assumed to be zero for both cases. For Doppler reactivity, the Dominion model uses a DTC while the FSAR model uses a DPC with minimum reactivity feedback conservatively assumed for both cases. Although fuel temperature and core power both increase in a similar manner prior to the reactor trip, the relative effect of that increase and the associated modeling of the feedback effect may vary in the respective Dominion and ESAR Doppler models. It is also noted that the reactor core model used in the FSAR case incorporates proprietary mechanisms to modify the removal of heat from the core.

Based on the core power response shown in Figure 4.5-1, the differences result in greater reactivity feedback and a slower increase in power for the FSAR case.

Serial No. 16-011A Docket No. 50-423 Attachmentl1, Page 9ofl11 Selection of a 1 porn/second reactivity insertion rate is intended to be representative of a slow insertion rate case. The dynamic effects of the RWAP event at a given insertion rate are different for the Dominion and FSAR methodologies as noted above.

The critical characteristic of the event analysis is determination of the minimum departure from nucleate boiling ratio (DNBR) which occurs for a case where reactor trip times for the overtemperature AT (OTDT) and high neutron flux trip signals coincide. This point will occur at different reactivity insertion rates for the two methodologies.

RAI - 15 (SRXB):

RWAP Analysis - DNBR Calculations The results of MPS3 FSAR Chapter 15 non-LOCA analyses indicated the RWAP event is the most limiting event in terms of the margin to the safety limit DNBR in the categor'y of the anticipated operating occurrences (A QOs). Since the licensee also proposed to use the RETRAN and Dominion VIPRE-D method to perform DNBR calculations for assessing the fuel integrity during AOOs and accidents, the RETRAN benchmarking analysis for both RWAP I pcmlsec and 100 pcm/sec cases should be performed to include the results of the DNBR calculation by using the Dominion VIPRE-D method. The requested information includes a comparison of Dominion VIPRE-D analyses to the MPS3 FSAR analysis of record (A OR) showing that the calculated DNBRs for both cases are compatible with the AOR and the allowable range of the use of the NRC-approved DNBR correlation in VIPRE-D for the Dominion method is not exceeded.

DNC Response In a December 15, 2015 teleconference between DNC and NRC staff, the NRC agreed it would be sufficient to provide the requested comparison for one reactivity insertion rate case versus both cases as stated in the request.

The Dominion RETRAN and VIPRE-D results for the 1 pcm/sec insertion rate RWAP case are chosen to represent the Dominion DNBR evaluation of the RWAP transient.

The VIPRE-D model input file for this case was developed in accordance with Dominion VIPRE-D Topical DOM-NAF-2-P-A. The resulting transient DNBR plot was compared to the DNBR plot shown in MPS3 ESAR Figure 15.4-9. Both the Dominion and FSAR DNBR plots show a comparable trend.

The observed differences are primarily due to the dynamic effects associated with the RETRAN benchmark results, as noted in the response to RAI-14.

The core power rate of increase in the 1 pcm/sec case for the Dominion RETRAN model is greater than the FSAR data such that the reactor trip occurs approximately 20 seconds earlier. The inverse effect of power on DNB is clearly observed in the transient DNBR plot below, and the minimum DNBR values for the Dominion and FSAR cases are observed to be comparable.

In addition, the thermal-hydraulic conditions of the RWAP transient analyzed were confirmed to be within the validation range of the NRC-approved

~Serial No. 16-011IA Docket No. 50-423, Page 10 of 11 DNBR correlations utilized in the VIPRE-D model (WRB-2M and ABB-NV) consistent with the limitations on the use of DOM-NAF-2-P-A.

Figure 4.5-7 Supplement RWAP - 1 pcm/sec DNBR 3.9 i-l--1pcrnFSARDNBRl 1 pcm Diora 3.4 1.i

,, 2.9 2!

2;.4 I

0 20 40 60 80 100 120 TIME %SEC)

RAt - 16 (SRXB): Feedwater Line Break Analysis MPS3 ESAR (2012 Version), Section 15.2.8 discussed the feedwater line break (FLB) analysis for both cases with and without offsite power available. The FSAR AOR presented transient results including nuclear power, core heat flux, total reactivity, pressurizer pressure, total RCS flow, feedwater break flow, loop temperature, intact loop temperature, and SG pressure. FSAR Figures 15.2-13 and 15.2-19 indicated that a post-trip return-to-power will occur for the case with offsite power available, and core will remain subcritical throughout the transient time of several seconds for the case without offsite power available. Also, page 1 5-2-16 indicated that the FLB is the most limiting event in the decrease in secondary removal category. The analysis of the FLB will use a broad scope of the models in RETRAN, including feedwater break flow model, RC pumps coastdowm model, SG heat transfer model, and reactivity feedback model.

Perform the RETRAN benchmarking analysis for the FLB event for both cases with and without offsite power available. The information to be provided should show that: the values of the plant parameters and assumptions used in the Dominion FLB analysis are consistent with that used in the FSAR AOR; the results are compatible with the AOR; and there is no unexplainable thermal-hydraulic phenomena throughout the transients.

Serial No. 16-011A Docket No. 50-423, Page 11 of 11 DNC Response The additional benchmarking analysis has been performed for the FLB event for cases with and without offsite power available. The discussion of the event analysis, including inputs and assumptions and results, as compared with the FSAR analysis are included in Section 4.6 of the update to Attachment 5 enclosed in Attachment 2.

Serial No. 16-011A Docket No. 50-423 ATTACHMENT 2 RETRAN BENCHMARKING INFORMATION - UPDATED DOMINION NUCLEAR CONNECTICUT, INC.

MILLSTONE POWER STATION UNIT 3 RETRAN Benchmarking Information Page 2 of 61 TABLE OF CONTENTS

1.0 INTRODUCTION

AND

SUMMARY

3

1.1 INTRODUCTION

........................................................................................ 3 1.2

SUMMARY

............................................................................................. 3 2.0 MPS3 RETRAN MODEL......................................................................

4 3.0 METHOD OF ANALYSIS...................................................................... 8 4.0 BENCHMARKING ANALYSIS RESULTS................................................. 9 4.1 Loss OF LOAD/TURBINE TRIP............................

.......................................... 9 4.2 LOCKED ROTOR..................................................................................... 14 4.3 Loss OF NORMAL FEEDWAThR................................................................... 24 4.4 MAIN STEAM LINE BREAK..............

.......................................................... 31 4.5 CONTROL ROD BANK WITHDRAWAL AT POWER................................................ 42 4.6 MAIN FEED WATER LINE BREAK.................................................................. 49

5.0 CONCLUSION

S................................................................................ 61

6.0 REFERENCES

61 Pae3ol Page 3 of 61 1.0 Introduction and Summary 1.1 Introduction Topical report VEP-FRD-41-P-A, "VEPCO Reactor System Transient Analyses Using the RETRAN Computer Code," (Reference 1) details the Dominion methodology for Nuclear Steam Supply System (NSSS) non-LOCA transient analyses. This methodology encompasses the non-LOCA licensing analyses required for the Condition I, II, III, and IV transients and accidents addressed in the Final Safety Analysis Report (FSAR). The VEP-FRD-4 1-P-A methods are also used in support of reload core analysis. In addition, this capability is used to perform best-estimate analyses for plant operational support applications. The material herein supports the applicability assessment of the VEP-FRD-4 1-P-A methods to Millstone Power Station Unit 3 (MPS3) for the stated applications.

1.2 Summary This attachment provides a description of the RETRAIN base model for MPS3 and results Of benchmarking analyses using this model. The MPS3 model was developed in accordance with the methods in VEP-FRD-4 1-P-A, with certain noding changes noted below. This assessment confirms the conclusion that the Dominion RETRAIN methods, as documented in topical report VEP-FRD-4 1-P-A, are applicable to MPS3 and can be applied to MPS3 licensing analysis for reload core design and safety analysis. Dominion analyses of MPS3 will employ the modeling in VEP-FRD-4 1-P-A, as augmented with the noding changes listed below. Thus, VEP-FRD-41-P-A, as augmented, is the Dominion methodology for analyses of non-LOCA NSSS transients for MPS3.

The MPS3 RETRAN base model contains the following alterations in noding with respect to the modeling that is documented in VEP-FRD-4 1-P-A.

a) The MPS3 model explicitly models the safety injection (SI) accumulators.

b) The MPS3 model has separate volumes for the steam generator inlet and outlet plenums.

c) The MPS3 model includes cooling paths between dowucomer and upper head.

Pae4ol Page 4 of 61 2.0 MPS3 RETRAN Model The MPS3 RETRAN-3D Base Model and associated model overlays are developed using Dominion analysis methods described in the Dominion RETRAN topical report (Reference 1).

The Dominion analysis methods are applied consistent with the conditions and limitations described in the Dominion topical report and in the applicable NRC Safety Evaluation Reports (SERs).

The MPS3 Base Model noding diagram for a representative loop is shown on Figure 2-1.

Volume numbers are circled, junctions are represented by arrows, and the heat conductors are shaded. This model simulates all four reactor coolant system (RCS) loops and has a single-node steam generator (SG) secondary side, consistent with Dominion methodology. The SG primary nodalization includes 10 steam generator tube volumes and conductors. There is a multi-node SG secondary overlay that can be added to the Base Model for sensitivity studies although none of the analysis results presented herein utilize this overlay.

In addition to the base MPS3 model, an overlay deck is used to create a split reactor vessel model to use when analyzing Main Steam Line Break (MSLB) events, consistent with Dominion methodology.

This overlay adds volumes to create a second, parallel flow path through the active core from the lower plenum to the upper plenum such that RCS loop temperature asymmetries can be represented. This noding is consistent with the method described in VEP-FRD-4 1-P-A. A noding diagram of the split reactor vessel is shown on Figure 2-2.

The base MPS3 model noding is virtually identical to the Surry (SPS) and North Anna (NAPS) models with the exception of some minor noding differences listed as follows.

a) The MPS3 model explicitly models the SI accumulators.

b) The MPS3 model has separate volumes for the SG inlet and outlet plenums.

c) The MPS3 model includes cooling paths between downcomer and upper head.

The SI accumulators are part of the MPS3 model because injection from the accumulators occurs in the current FSAR analysis for MSLB. The use of separate volumes for the inlet and outlet should have little effect on transient response since the fluid temperature in these volumes is generally the same as the connecting RCS piping. The cooling paths are included to appropriately model upper head T-cold conditions.

The Dominion models, including the MPS3 model, have some differences compared to the vendor RETRAN model that was used to perform the current FSAR analyses. Table 2-1 and the subsequent text discussion provide an overview of these differences. Additional details Pae5ol Page 5 of 61 concerning differences between the Dominion MPS3 and FSAR RETRAN models are discussed in the benchmarking analyses in Section 4.

A description of the Dominion RETRAN methodology is provided in Reference 1, where specific model details are discussed in Sections 4 and 5 of that reference.

Table 2-1 RETRAN Model Comnarison of Key Characteristics Parameter Dominion FSAR Code Version:

RETRAN-3D in "02 mode" RETRAN-02 Noding:

Reacor VsselSingle flow path (special split core Reco esloverlay for MSLB only)

Multiple parallel flow paths Single node secondary. Five axial levels (10 nodes) for SG tubes Steam Generator primary side. Local Conditions Heat Multi-node secondary.

Transfer model available for loss of heat sink events.

Reactivity Model Doppler-only power coefficient Dopper eedack Doppler temperature coefficient that and a Doppler temperature is a function of TFUEL.

coefficient effect driven by moderator temperature.

Moderator Feedback Moderator temperature coefficient Moderator density coefficient ANS 1979 Standard U-235 with 1500 day burn.AN 199Sadr Decay Heat Q = 190 MeV/fission.

Boundss additional 2o uncertainty Attahmet 2pg.

6 of 61 Figure 2-1 MPS3 Base Model Nodalization Diagram MO Slt*Vo la mc pg. 7 of 61 Figure 2-2 MPS3 Split Vessel Nodalization pg8o6 pg. 8 of 61 3.0 Method of Analysis Validation of the Dominion MPS3 RETRAN method involves comparison of RETRAN analyses to the MPS3 FSAR analysis of record (AOR) for select events. The Dominion analyses presented herein are not replacements for the existing AORs. These events represent a

broad variation in behavior (e.g.

RCS

heatup, RCS cooldown/depressurization, reactivity excursion, loss of heat sink, etc.), and demonstrate the ability to appropriately model key phenomena for a range of transient responses. The transients selected for comparison with their corresponding MPS3 FSAR section are provided in Table 3-1. For each transient, an analysis is performed using the Dominion MPS3 RETRAN model and compared with the current FSAR analysis. Initial conditions and inputs are established for each benchmark to provide an adequate comparison of specific transient behavior.

Table 3-1 Transients Analyzed for FSAR Comparison Transient MPS3 FSAR Section Main Steam Line Break 15.1.5 Loss of Load/Turbine Trip 15.2.3 Loss of Normal Feedwater 15.2.7 Locked Rotor 15.3.3 Control Rod Withdrawal at Power 15.4.2 Main Feedwater Line Break 15.2.8 pg9o6 pg. 9 of 61 4.0 Benchmarking Analysis Results A summary for each transient comparison is presented in the following sections. Included in each section is an input summary identifying key inputs and assumptions along with differences from FSAR assumptions. A comparison of the results for key parameters is provided with an explanation of key differences between the Dominion and FSAR cases.

4.1 Loss of Load/Turbine Trip The Loss of Load/Turbine Trip (LOL) event is defined as a complete loss-of-steam load and turbine trip from full power without a direct reactor trip, resulting in a primary fluid temperature rise and a corresponding pressure increase in the primary system. This transient results in degraded steam generator heat transfer, reactor coolant heatup and pressure increase following a manual turbine trip.

The LOL transient scenario presented here was developed to analyze primary RCS overpressurization. It is initiated by decreasing both the steam flow and feedwater flow to zero immediately after a manual turbine trip. The input sumnmary is provided in Table 4.1-1.

Table 4.1-1 ILOI Tnniit Sulmmary Parameter Value Notes Initial Conditions_______

Core Power (MW) 3723 Includes 2% uncertainty RCS Flow (gpm) 363,200 Thermal Design Vessel TArG (F) 576.5 Low Tavg plus uncertainty Pressurizer Pressure (psia) 2200 Includes -50 psia uncertainty Pressurizer Level (%)

52.5 Low Tavg Target plus uncertainty SG Level (%)

50.0 Nominal SG tube plugging (%)

10 Maximum Pump Power (MW/Pump) 5.0 Maximum Assumptions/Configuration_______

Reactor trip only Hi Pzr Pressure is active Automatic rod control Not credited Pressurizer sprays, PORVs Not credited Main steam dumps, SG PORV Not credited AFW flow Not credited Reactivity Parameters Doppler Reactivity Feedback Least Negative Moderator Feedback Most Positive p.lo~

pg. 10 of 61 Results - LOL Pressure in the RCS increases during a LOL due to degraded heat transfer in the steam generator and is alleviated only when the pressurizer safety valves (PSV) open as well as the main steam safety valves (MSSV). The pressurizer pressure response is shown on Figure 4.1-1, RCP outlet pressure in Figure 4.1-2, and the peak RCS pressure values are listed in Table 4.1-2. The Dominion case predicts a pressunizer pressure and RCP outlet pressure response that agrees very well with the FSAR results past the point of peak RCS pressure.

Following the initial decrease in primary system pressure, the FSAR pressure levels out where the Dominion case results continue to decrease. The difference is due to differing secondary safety valve modeling in the vendor model, specifically in that the Dominion model includes the modeling of blowdown in the main steam safety valves and the vendor model does not. Hence, more energy is removed through the secondary system in the Dominion case once the main steam safety valves actuate than is removed from the secondary system in the vendor model.

Figure 4.1-3 shows the power response is nearly identical both before and after the reactor trip on high pressurizer pressure and control rod insertion. The Dominion case trips slightly earlier than the FSAR data because of the higher RCS pressurization rate.

The Dominion model vessel inlet temperature, Figure 4.1-4, and coolant average temperature, Figure 4.1-5, agrees in trend and rate of increase although the response lags the FSAR response before the inlet temperature peaks at a slightly lower value. This indicates that the FSAR steam generator heat transfer degrades sooner than what is predicted by Dominion model and is attributed to the difference expected between the use of a multi-node steam generator (MNSG) in the FSAR model and the single-node steam generator (SNSG) model employed in the Dominion model. Overall, both the Dominion model and FSAR models exhibit similar trends in the temperature responses and the differences have no effect on peak RCS pressure.

Table 4.1-2 LOL RCS Overpressure Results Parameter Dominion FSAR Sequence of Events:

High Pressurizer Pressure Setpoint 5.6 6.2 Reached (see)

Peak RCS Pressure (sec) 9.2 9.9 Peak RCS Pressure (psia) 2705 2725 I

p.lo6 pg. 11 of 61 Figure 4.1-1 LOL - Pressurizer Pressure 2700 2600

• .2500

" 2400 S2300 2200 2100 2000 010 20 30 40 Time (sec) 50 Figure 4.1-2 LOL - RCP Outlet Pressure 3000 2900 2800 2700

• 2600 22500 22400 2300 2200 2100 2000 0

10 20 30 40 50 Time (sec) p.1o6 pg. 12 of 61 Figure 4.1-3 LOL - Nuclear Power 1.2 1.0 S0.8 0.

0.2 0.0 010 20 30 40 Time (sec) 50 Figure 4.1-4 LOL-Vessel Inlet Temperature 590 580 570

'07560

~3550 E

)I-540 530 520 510 4 0.00 10.00 20.00 30.00 40.00 Time (sec) 50.00 p.1o6 pg. 13 of 61 Figure 4.1-5 LOL - Vessel Average Temperature 600 595 590 o-585 580 E

5-575 570 565 560 0

10 20 30 40 50 Time (sec)

Summary - LOL The Dominion MPS3 analysis provides results that are similar to the FSAR analysis for the LOL event. The RCS peak pressures are essentially the same although the pressure diverge somewhat later in the event after pressure relief begins due to differences in MSSV modeling. There are small differences in the RCS temperature response due to differences in the SG models, however, this has no effect on the RCS peak pressure. The Dominion MPS3 analysis is presented for benchmark comparison, and does not replace the existing AOR.

p.1o6 pg. 14 of 61 4.2 Locked Rotor The Locked Rotor / Shaft Break (LR) event is defined as an instantaneous seizure of a Reactor Coolant Pump (RCP) rotor, rapidly reducing flow in the affected reactor coolant loop leading to a reactor trip on a low-flow signal from the Reactor Protection System. The event creates a rapid expansion of the reactor coolant and reduced heat transfer in the steam generators, causing an insurge to the pressurizer and pressure increase throughout the reactor coolant system (RCS).

The LR transient scenario presented here was developed to analyze primary RCS overpressurization.

It is initiated by setting one RCP speed to zero as the system is operating at full power. The reactor coolant low ioop flow reactor trip is credited, with a setpoint of 85% of the initial flow. The input summary is provided in Table 4.2-1. Most of the input parameters are the same as those used in the FSAR Chapter 15 analyses.

Table 4.2-1 LR Input Summary Parameter Value Notes Initial Conditions Core Power (MW) 3723 Includes 2% uncertainty RCS Flow (gpm) 363,200 Thermal Design Flow Vessel TAVG (F) 594.5 Nominal + 5*F Pressurizer Pressure (psia) 2300 Includes +50 psia uncertainty Pressurizer Level (%)

64 Nominal SG Level (%)

50 Nominal Assumptions/Configuration Reactor trip

_______Only Low RCS Loop Flow is credited Automatic rod control

_______Not credited Pressurizer sprays, PORVs

-Not credited Main steam dumps, SG PORV

-Not credited AFW flow Not credited SG tube plugging (%)

10' Max value Reactivity Parameters Doppler Reactivity Feedback Most Negative Dominion model adjusted to use FSAR

__________Doppler Power Coeffcient Moderator Feedback Most Positive

' Original benchmark case inadvertently assumed 0% SG tube plugging Results - LR RCS Overpressure Case Pressure in the RCS increases during a LR event due to degraded heat transfer in the steam generator and is alleviated only when the pressurizer safety valves (PSV) open. The magnitude of the Dominion model pressure response both in the reactor vessel lower plenum, Figure 4.2-1, and at the RCP exit, Figure 4.2-2, is greater than the FSAR model response, while following the same trends as the FSAR data. At the limiting point in the p.1o6 pg. 15 of 61 transient response, the Dominion model conservatively predicts a pressure approximately 63 psi greater than the FSAR model in the reactor vessel lower plenum. The difference between the Dominion model and FSAR model's peak responses is the same at the RCP exit as in the lower plenum.

The Dominion faulted loop flow response (Figure 4.2-3) and unfaulted loop flow response (Figure 4.2-4) are in good agreement with the FSAR model response up to or just beyond the point of rod insertion. Following reactor trip there is some divergence in the unfaulted loop flow trends, which are consistent with the core heat flux predictions and assumed minor differences in the loop friction losses between the Dominion and FSAR models.

With respect to the faulted loop flow response, the maximum reverse flow seen in the FSAR model is slightly greater than seen in the Dominion model, which is also attributed to small differences in the loop friction losses between the Dominion and FSAR models.

For the total core inlet flow response (Figure 4.2-5), the Dominion model predicts a lower flow than the FSAR model for approximately the first 4 seconds of the transient. After 4 seconds the FSAR and Dominion model core flow responses cross and the Dominion model predicts a slightly higher core flow rate. The limiting point in the transient occurs prior to 4 seconds such that RETRAN-3D produces a more limiting response than the FSAR model for the Locked Rotor/Shaft Break event.

The nuclear power response, Figure 4.2-6, predicted by the Dominion model agrees well with the FSAR data, with the Dominion model response slightly over predicting power during rod insertion following the reactor trip on low RCS flow. Similarly, the Dominion model core heat flux response, Figure 4.2-7, also slightly over predicts the FSAR model's response in the same time frame during control rod insertion. Additionally, the Dominion model heat flux response shows a slightly larger decrease at the initiation of the event over the decrease seen in the FSAR data. Both the initial unader prediction of the heat flux response, followed by an over prediction during the rod insertion is indicative of the fuel rod heat transfer being modeled differently in the FSAR methods than in the Dominion model.

However, the over prediction of both nuclear power and heat flux will lead to conservative results at the limiting point in the transient for both RCS overpressurization and DNB during rod insertion. Overall the nuclear power and heat flux predictions are very similar.

A summary of the LR transient analysis comparison is provided in Table 4.2-2.

p.1o6 pg. 16 of 61 Table 4.2-2 LR RCS Overpressure Results Parameter Dominion FSAR Sequence of Events:

Low RCS Flow Setpoint Reached (sec) 0.1 0.1 Rods Begin to Drop (sec) 1.1 1.1 Peak RCS Pressure (sec) 3.8 4.1 Peak RCS Pressure (psia) 2680 2617 Summary - LR RCS Overpressure Case The Dominion Millstone analysis provides responses that are similar to the FSAR analysis for the LR event, with the Dominion model predicting higher peak RCS pressures.

Differences are attributed to ioop friction losses and fuel rod modeling differences. The Dominion MPS3 analysis is presented for benchmark comparison, and does not replace the existing AOR.

p.1o6 pg. 17 of 61 Figure 4.2-1 LR - Reactor Vessel Lower Plenum Pressure 2700 2650 2600 2550 2500

• , 2450 2400 2350 2300 2250 2200 0

5 10 15 20 Time (sec)

Figure 4.2-2 LR - RCP Outlet Plenum Pressure 2700 2650 2600 2550 2500 v

2450 S2400 2350 2300 2250 2200 05 10 15 20 Time (sec) p.1o6 pg. 18 of 61 Figure 4.2-3 LR - Faulted Loop Normalized Flow 0

It-

")

N.

0Z 1.20 1.00 0.80 0.60 0.40 0.20 0.00

-0.20

-0.40

-0.60 Time (sec)

Figure 4.2-4 LR - Unfaulted Loop Normalized Flow 0

U-

~0ci)

N 0z 1.16 1.06 0.96 0.86 0.76 0.66 0.56 0.46 0

5 10 15 20 Time (sec) p.1o6 pg. 19 of 61 Figure 4.2-5 LR - Core Inlet Normalized Flow 1.10 1.00 0.90 0.80

\\.*

.2 0.70 IL

___ 0.60 zO0.50

-Dominion 0.40 FA 0.30 05 10 15 20 Time (sec)

Figure 4.2-6 LR - Nuclear Power 1.11 1.01 0.91 0.81 0.71 0..

0.61 N S0.51 Oz 0.41 0.31 kI"--Dmn~

0.21-*._I--SA I

0.11 05 10 15 20 Time (sec) p.2o6 pg. 20 of 61 Figure 4.2-7 LR - Core Heat Flux x

II "1-N tu 0Z 1.12 1.02 0.92 0.82 0.72 0.62 0.52 0.42 0.32 0.22 0.12 05 10 15 20 Time (sec) p.2o6 pg. 21 of 61 LR Peak Cladding Temperature The Locked Rotor event is also analyzed to demonstrate that a coolable core geometry is maintained. A hot spot evaluation is performed to calculate the peak cladding temperature and oxidation level. The Dominion Hot Spot model is described in Topical Report VEP-NFE-2-A, "VEPCO Evaluation of the Control Rod Ejection Transient."

(Reference 2). The Dominion Hot Spot model was used to evaluate the MPS3 PCT and oxidation level for the LR event.

The Dominion hot spot model is used to predict the thermal-hydraulic response of the fuel for a hypothetical core hot spot during a transient. The hot spot model describes a one-foot segment of a single fuel rod assumed to be at the location of the peak core power location during a transient. The hot spot model uses boundary conditions from the LR system transient analysis to define inlet flow and core average power conditions. The hot spot model uses MPS3-specific values for fuel dimensions, fuel material properties, fluid volume, and junction flow areas.

The hot spot model is run to 0.1 seconds and a restart file is saved. Upon restart, the fuel/cladding gap conductance (thermal conductivity) is modified to simulate gap closure by setting the gap heat transfer coefficient to 10,000 Btu/ft2 -hr-°F for a gap conductance of 2.708 Btulft-hr-°F. The hot spot model input summary is provided in Table 4.2-3. Most of the input parameters are the same as those used in the FSAR Chapter 15 analyses. Where differences from the FSAR inputs exist, they are indicated in the Notes column.

Table 4.2-3 Hot Spot Model Input Summary Parameter Value Notes Computer Code Used RETRAN-3D FSAR uses VIPRE Initial Conditions Ratio of Initial to Nominal Power 1.02 RCS Flow (gpm) 363,200 Hot Spot Peaking Factor 2.60 Assumptions/Configuration Pre-DNB Film Heat Transfer Coefficient Thorn Time of DNB (sec) 0.1 Post DNB Film Boiling Heat Transfer Bishop-Sandberg-Coefficient Tong___________

Fuel Pin Model Post DNB Gap Heat Transfer Coefficient 10,000 (Btu/hr-ft2-°F)_________

Gap Thermal Expansion Model activated?

Yes Zircaloy-Water Reaction activated?

Yes Attahmet 2pg.

22 of 61 LR Peak Cladding Temperature Results The peak cladding temperature obtained from Dominion's MPS3 hot spot model for the locked rotor event is 1760 0F. The maximum zircaloy-water reaction depth is 3.60875E-06 feet, which corresponds to approximately 0.19% by weight based on the nominal cladding thickness of 1.875E-03 feet. A summary of the LR Peak Cladding Temperature Hot Spot analysis comparison is provided in Table 4.2-4. The cladding inner surface temperature is shown in Figure 4.2-8.

Table 4.2-4 LR Hot Spot Results Parameter Dominion FSAR Peak Cladding Temperature 1760 0F 1718 0F Maximum Zr-water reaction (w/o) 0.19 0.22 The Dominion peak cladding temperature and maximum oxidation values are comparable to the FSAR values. The Dominion MPS3 analysis is presented for benchmark comparison, and does not replace the existing AOR.

p.2o6 pg. 23 of 61 Figure 4.2-8 LR Hot Spot - Cladding Inner Surface Temperature 1750 1550

'-1350

,- 1150 E

950 750 0.1 5.1 10.1 15.1 Time (sec) p.2o6 pg. 24 of 61 4.3 Loss of Normal Feedwater The Loss of Normal Feedwater (LONE) event causes a reduction in heat removal from the primary side to the secondary system. Following a reactor trip, heat transfer to the steam generators continues to degrade resulting in an increase in RCS fluid temperature and a corresponding insurge of fluid into the pressurizer. There is the possibility of RCS pressure exceeding allowable values or the pressurizer becoming filled and discharging water through the relief valves. The event is mitigated when Auxiliary Feedwater (AFW) flow is initiated and adequate primary to secondary side heat removal is restored. This analysis shows that the AFW system is able to remove core decay heat, pump heat and stored energy such that there is no loss of water from the RCS and pressure limits are not exceeded. The LONE input summary is provided in Table 4.3-1.

Table 4.3-1 LON Iput Summary Parameter Value Notes Initial Conditions Core Power (MW) 3723 Includes 2% uncertainty RCS Flow (gpm) 363,200 Thermal Design Flow Vessel TAVG (F) 583 FSAR value RCS Pressure (psia) 2300 Nominal + 50 psi Pressurizer Level (%)

71.6 Nominal + 7.6%

SG Mass

- 89000 Dominion model adjusted to be consistent with FSAR analysis Assumptions/Configuration Low-Low Level Reactor Trip Setpoint 0%

Percent of narrow range span Pressurizer: sprays, heaters, PORVs

-Assumed operable AFW Temperature (F) 120 Max value AFW Pump configuration

-2 motor-driven pumps feed 4 SGs Auxiliary feedwater flow rate (gpm)

-Variable as function of SG press.

Local Conditions Heat Transfer model active SG secondary side

__________________________FSAR= multi-node SG Decay Heat

-FSAR decay heat constants are

_____________________________applied for this case Reactivity Parameters Doppler Reactivity Feedback Most negative Dominion model adjusted to use

_________FSAR Doppler Power Coeffcient Moderator Feedback Most Positive I

p.2o6 pg. 25 of 61 Results - LONF The results for the LONF comparison analysis are presented in Table 4.3-2 and Figures 4.3-1 through 4.3-7.

The loss of feedwater flow to the steam generators (SG) results in a reduction in SG level until a reactor trip occurs on Low-Low SG level. Normalized power is shown on Figure 4.3-1 and normalized core heat flux in Figure 4.3-2. The nuclear power response and heat flux response predicted by the Dominion model are in excellent agreement with the FSAR data, indicating that the scram on low-low steam generator level occurred at essentially the same time shown for the FSAR data. The results continue to demonstrate good agreement through the end of the event.

Figure 4.3-3 shows the steam generator pressure response. The Dominion steam generator pressure is initialized at a slightly different pressure than the FSAR model because the Dominion model initial condition is adjusted to minimize the steam generator area adjustment. Between 10 and 34 seconds the FSAR pressure increases more rapidly to a pressure --43 psi greater than the Dominion model prediction when the steam line is isolated.

This difference is attributed to differing heat transfer degradation in the MNSG model used in the FSAR analysis versus the SNSG model used in the RETRAN-3D model. Steam line isolation occurs at nearly the same time, causing pressure to increase rapidly. The peak pressure is limited by the main steam safety valves (MS SVs), resulting in an almost identical peak pressure in both the Dominion and FSAR responses. However, the Dominion model pressure decreases following the peak value, where the FSAR model response remains at a constant value near the peak value, due to differences in MSSV modeling.

Figure 3.1-4 shows the steam generator liquid mass. The steam generator liquid mass depletes faster in the Dominion cases than in the FSAR cases. This is consistent with the increased relief flow as shown in the steam generator pressure response.

The response in the pressurizer is shown in Figures 4.3-5 and 4.3-6. Between the FSAR and Dominion model, the pressure responses are in good agreement until around 45 -

50 seconds where the Dominion pressure is lower than the FSAR., reflecting less heat transfer degradation during this period. This is followed by a second pressure peak that is higher for Dominion than the FSAR. Based on the sharpness of the Dominion peak compared with the FSAR data, this difference is most likely driven by differences in the pressurizer spray models and primary to secondary heat transfer.

For the pressurizer water volume, shown in Figure 4.3-6, the Dominion model results follow the same trends as the FSAR data, but drops lower in the period from 63 to 900 seconds, then demonstrates a strong insurge during the second heat-up period in the transient while peaking at a somewhat lower value than the FSAR. The difference seen in the pressurizer pg. 26 of 61 volume results is primarily due to the previously discussed MSSV modeling differences and the resultant increased steam release from the Dominion model compared to the FSAR model as well as possible differences in the pressurizer spray models.

Table 4.3-2 LONF Results Parameter Dominion FSAR Peak PZR Liquid Volume (ft3) 1610 1730 Figure 4.3-1 LONF - Nuclear Power pg. 27 of 6l o1

~0 Z) 1.20 1.00 0.80 0.60 0.40 0.20 0.00 1

10 100 1000 Time (sec)

Figure 4.3-2 LONF - Normalized Core Heat Flux 10000 1.20-1.00-0.80 S0. 60 LI) 0.20 0.00-1.00 10.00 100.00 1000.00 10000.00 Time (sec)

Attahmet 2pg.

28 of 61 Figure 4.3-3 LONF - Steam Generator Pressure Ct)

Co 09 1330 1280 1230 1180 1130 1080 1030 980 930 880 110 100 1000 Time (sec) 10000 Figure 4.3-4 LONF - Steam Generator Liquid Mass 100000 90000 80000 70000 60000 co50000

-~40000

• J30000 20000 10000 0

110 100 1000 Time (sec) 10000 Attahmet 2pg.

29 of 6l Figure 4.3-5 LONF - Pressurizer Pressure 0-.

E

--0 2500 2450 2400 2350 2300 2250 2200 2150 1800 1700 1600 1500 1400 1300 1200 1100 1000 1

10 100 1000 Time (sec)

Figure 4.3-6 LONF - Pressurizer Water Volume 10000 110 100 1000100 Time (sec) 10000 p.3o6 pg. 30 of 61 Figure 4.3-7 LONF - Loop Average Temperature 595 590 E

I-585 580 575 570 110 10O0 1000 Time (sec) 10000 Summary - LONF The Dominion analysis provides results that are similar to the FSAR analysis for the LONF event. The major differences result from the main steam safety relief valve modeling, which results in higher steam releases and a subsequent increase in heat transfer following the reactor trip. In addition, the steam generator nodalization and related heat transfer along with other modeling differences such as pressurizer spray also affect the transient response.

These effects are cumulative resulting in a somewhat smaller long-term pressurizer insurge and higher pressurizer pressure peak compared to the FSAR results. The Dominion MPS3 analysis is presented for benchmark comparison, and does not replace the existing AOR.

pg31fl pg. 31 of 61 4.4 Main Steam Line Break The Main Steam Line Break (MSLB) event is a rupture in the main steam piping resulting in a rapid depressurization of the SG secondary and corresponding cooldown of the primary.

The temperature reduction results in an insertion of positive reactivity with the potential for core power increase and DNBR violation.

The MSLB transient scenario presented here is modeled as an instantaneous, double-ended break at the nozzle of one steam generator from hot shutdown conditions with offsite power available. The input summary is provided in Table 4.4-1.

Table 4.4-1 MSLB Input Summary Parameter Value Notes Initial Conditions Core power (MW)

~-1%

H-ZP Pump power (MW) 0.0 RCS Flow (gpm) 363,200 Thermal Design Flow Vessel TAVG (F) 557 H-ZP nominal RCS Pressure (psia) 2250 Nominal Pressurizer Level (%)

28 HZP nominal SG Level (%)

50 Nominal Assumptions/Configuration Heat transfer option Forced HT Map FSAR uses a proprietary heat (note 1) transfer formulation Main feedwater flow (% HFP value) 100 initiated at time 0 sec Auxiliary feedwater flow rate (gpm)

Max initiated at time 0 sec SG tube plugging (%)

0 Minimum value Reactivity Parameters RWST Boron Credited FSAR does not credit boron from

____________the SI system Accumulator Boron Not Credited Doppler Reactivity Feedback Doppler Only FSAR - Doppler power defect Power defect, plus DTC included in moderator DTC model density feedback disabled Moderator Feedback Moderator Moderator density feedback

____________________________density feedback_______________

1 - Dominion method maximizes heat transfer coefficients for the faulted SG secondary side.

p.3o6 pg. 32 of 61 Results - MSLB with Offsite Power Available The faulted loop steam flow and steam generator pressure responses shown in Figure 4.4-1 and Figure 4.4-3 match the FSAR data reasonably well with the steam flow and pressure in the Dominion model remaining somewhat higher than the FSAR data. This is partly caused by the slightly larger break junction area and the higher initial steam pressure for the Dominion model. In addition, the Dominion model uses conservatively high heat transfer coefficients in the faulted steam generator, which allow the faulted steam generator to pull heat faster from the primary side.

The Intact loop steam flow (Figure 4.4-2) shows a different response due to differences in the MSIV closure. In the Dominion model, the MSIVs close linearly over 10 seconds, while the FSAR model uses a delay of 10 seconds to conservatively increase RCS overcooling. The initial steam flow is higher for the Dominion case, decreasing below the FSAR value as the MSIVs close. The steam generator mass and pressure responses, shown in Figure 4.4-8 and Figure 4.4-4, reveals the differences in MSIV modeling with the Dominion model releasing somewhat less liquid inventory prior to valve closure.

For both the faulted and intact loops the main feedwater and auxiliary feedwater responses (Figure 4.4-5) give an excellent match to the FSAR data. The steam generator inventory (Figure 4.4-7) for the faulted loop depletes faster in the Dominion model than in the FSAR case due to the higher steaming rate from the faulted steam generator and the quicker and more conservative return to power.

The nuclear power and core heat flux responses (Figure 4.4-9 and Figure 4.4-10) calculated by the Dominion model peak higher and more quickly than the FSAR data.

This response is contributed to by the greater cooling effects of the faulted steam generator on the RCS due to its higher steam production. The quicker return to power is also a result of differences in the nodalization and mixing at the core inlet and outlet between the Dominion model and the FSAR model. The return to power also drops off approximately 50 seconds sooner in the Dominion model. This is also caused by the higher steam rate in the Dominion model which causes the faulted steam generator to dry out sooner. The power response for both models is not affected by the delivery of boron to the RCS. This is because the FSAR model does not credit boron and in the Dominion model boron does not reach the RCS from the SI system until after the termination of the transient. Overall, the Dominion model results in a more conservative response for core heat flux and power.

Attahmet 2pg.

33 of61 The pressurizer pressure response (Figure 4.4-12) agrees very well with the pressure predicted by the FSAR model for the first 50 seconds of the transient, after -which the FSAR data falls approximately 100 psi lower than the pressure calculated by the Dominion model. This difference is a result of using only a single upper head leakage path in the Dominion model. The upper head leakage is taken from the three intact loops and does not credit any flow from the lower temperature, faulted loop. This causes the upper head temperature to remain slightly higher than would actually be the case, which allows a vapor bubble in the upper head to form sooner and become larger. This in turn prevents the RCS pressure from falling lower.

The pressurizer drains at approximately the same rate for the Dominion model and FSAR models (Figure 4.4-13). However, for the Dominion model the pressurizer begins to refill approximately 100 seconds sooner. The quicker refilling is a result of the higher and quicker return to power which causes the RCS temperature to rise sooner in the Dominion model. This causes the RCS fluid inventory to expand which results in the pressurizer refilling sooner in the Dominion model than is seen from the FSAR model.

Table 4.4-2 MSLB with Offsite Power Results Time (sec) From Start of Transient Event Dominion FSAR Steam Line Ruptures 0

0 Manual Reactor Trip 0

0 Increase MFW to 100% of Nominal HFP00 Value 0____0 Initiate Maximum AFW to Faulted Steam Generator 0

0 Main Feedwater Isolation 7.5 8.2 MSIVs Closed 12.5 13.5 Pressurizer Empty 15.5 20.5 Criticality Attained 33.5 28.0 Safety Injection Flow Initiation 47.9 72.8 Faulted Steam Generator Dries Out 298

-350 p.3o6 pg. 34 of 61 Figure 4.4-1 MSLB - Faulted Loop Steam Flow 3000 2500 i-' 2000 E-1500 500 0100 200 300 400 500 600 Time (sec)

Figure 4.4-2 MSLB - Intact Loop Steam Flow 2400 1900 cic)1400 E

S900 400

-100 0

10 20 30 40 50 Time (sec) 60 70 80 90 100 Figure 4.4-3 MSLB - Faulted Loop Steam Generator Pressure pg. 35 of 61 1200 1000 800 400 200 0

100 200 300 400 500 Time (sec)

MSLB -Intact Loop Steam Generator Pressure 600 Figure 4.4-4 1100 1050 1000 950 a900 S850

~-800 750 700 650 I ------- Dominion I!

4 qmmm Im*

m Im

• v*m m N4 4Lira *mml

• m m

1*mD,Blf$ w m

W W

Immm m 0

100 200 300 Time (sec) 400 500 600 p.3o6 pg. 36 of 61 Figure 4.4-5 MSLB - Faulted Loop Total Feedwater Flow 1400 1200-1000-C,) 800-o2 600-C,)

C,)

S400-200O I

I I

I Dominion I-"

FSAR I

0 10 20 30 40 50 Time (sec)

Figure 4.4-6 MSLB - Intact Loop Total Feedwater Flow 1200 1000-U)

E Ut) 800-600-400-200-Dominion

-- -- FSAR 0

0 10 20 30 40 50 Time (sec) p.3o6 pg. 37 of 61 Figure 4.4-7 MSLB - Faulted Loop SG Liquid Mass 180000 160000 140000

-120000 100000 v80000 S60000 40000 20000 0

163000 162000 161000 160000

~159000 158000 155000 154000 0

100 200 300 400 500 Time (sec) 600 Figure 4.4-8 MSLB - Intact Loop SG Liquid Mass 0100 200 300 400 500 Time (sec) 600 p.3o6 pg. 38 of 61 Figure 4.4-9 MSLB - Normalized Core Power 0

n~

Z 0.20 0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00 0

100 200 300 400 500 Time (sec) 600 Figure 4.4-10 MSLB - Normalized Core Heat Flux U.)

.N

('

0Z 0.20 0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00 010O0 200 300 400 500 Time (sec) 600 Attahmet 2pg.

39 of61 Figure 4.4-11 MSLB - Reactivity Feedback 100 200 300 400 0

500 600 100

-100

-300 o-500

• _: -700 n, -900

-1100

-1300

-1500 Time (sec)

Figure 4.4-12 MSLB - Pressurizer Pressure 2500 2300 2100 1900

-,1700

.- 1500 S1300 o.1100 900 700 500 0

100 200 300 400 500 Time (sec) 600 p.4o6 pg. 40 of 61 Figure 4.4-13 MSLB - Pressurizer Liquid Volume 800 700 600 i;; 500 E400,,

200 i

Dominion 100 I

~--,,-FSAR 0100 200 300 400 500 600 Time (sec)

Figure 4.4-14 MSLB - Faulted Loop Vessel Inlet Temperature 550 530

£L 490 470 a) I-I 430 410

!--.-Dmno

/-

-- FSAR 390...

0 100 200 300 400 500 600 Time (sec) pg41ol pg. 41 of 61 Figure 4.4-15 MSLB - Intact Loop Vessel Inlet Temperature 560 550 540 530

• m _520

.~510 E 500 I-- 490 480 470 460 0

10O0 200 300 400 500 600 Time (sec)

Summary - MSLB This section presents a comparison of a RETRAN-3D Main Steam Line Break transient calculation with the Millstone model using the Dominion RETRAN transient analysis methods (Reference 1) compared to the FSAR results. The Dominion MPS3 analysis is presented for benchmark comparison, and does not replace the existing AOR. The key observations from these comparisons are that:

1) The peak power and heat flux reached with the Dominion methods is higher than the FSAR result.

2) Core and steam generator nodalization effects asymmetric transients such as a MSLB.

p.4o6 pg. 42 of 61 4.5 Control Rod Bank Withdrawal at Power The Control Rod Bank Withdrawal at Power (RWAP) event is defined as the inadvertent addition of core reactivity caused by the withdrawal of rod control cluster assembly (RCCA) banks when the core is above no load conditions. The RCCA bank withdrawal results in positive reactivity insertion, a subsequent increase in core nuclear power, and a corresponding rise in the core heat flux. The RWAP event described here is terminated by the Reactor Protection System on a high neutron flux trip or the overtemperature AT trip (OTAT), consistent with the FSAR analyses.

The RWAP event is simulated by modeling a constant rate of reactivity insertion starting at time zero and continuing until a reactor trip occurs. The Dominion analysis involves two different reactivity insertion rates, 1 pcm/sec and 100 pcm/sec that match the reactivity insertion rates presented plots in the FSAR. Most of the input parameters are the same as those used in the FSAR Chapter 15 analyses. Where differences from the FSAR inputs exist, they are indicated in the Notes column.

Table 4.5-1 RWAP In ut Summary Parameter Value Notes Initial Conditions________________

Core Power (MW) 3650 Nominal RCS Flow (gpm) 379,200 Minimum Measured Flow Vessel TAVG (F) 589.5 Nominal RCS Pressure (psia) 2250 Nominal Pressurizer Level (%)

64 Nominal SG Level (%)

50 Nominal Initial Fuel Temperature Minimum Uses current FSAR analysis conductivity adjustments Assumptions/Configuration Reactor trip

-High neutron flux or OTAT Automatic rod control

-Not credited Pressurizer level control Not credited Pressurizer heaters

-Not credited Pressurizer sprays, PORVs Active SG tube plugging (%)

10 Max value Reactivity Parameters Doppler Reactivity Feedback Least Negative _________________

Moderator Feedback Most Positive Zero MTC for cases from full power Results - RWAP 1 pcm/sec Case Figure 4.5-1 shows the core power response. The core power rate of increase for the Dominion model is greater than the FSAR data. This leads to the Dominion modeling p.4o6 pg. 43 of 61 tripping on high neutron flux at about 74 seconds. The FSAR case rises in power at a slower rate, which trips on an OTAT signal at about 93 seconds. The difference in reactor trip mechanisms between the Dominion and FSAR cases is reasonable considering the breakpoint for switching between OTAT and high flux as shown in FSAR Figure 15.4-10.

The pressure response also affects the OTAT setpoint such that the lower FSAR pressure (see below) will act to reduce the setpoint.

The pressurizer pressure response is shown in Figure 4.5-2. For the Dominion model, the pressure rises faster than the FSAR result. At about 42 seconds, the Dominion model reaches the pressurizer relief valve setpoint and begins to cycle. The FSAR more slowly increases in pressure and reaches the relief valve set point around 10 seconds prior to the reactor trip. The difference in pressure response can be attributed to the difference in core power response as each cases pressure response initially mimics the energy generated by the core as seen in Figure 4.5-1 and the higher spray flow assumed in the FSAR analysis, which acts to suppress pressure. The same can be seen in the vessel average temperature response where the FSAR case lags the Dominion response, yet reaches a temperature approximately 5 degrees higher than the Dominion case due to the FSAR case tripping later in the transient.

Table 4.5-2 RWAP 1 pcm/sec Time Sequence of Events Event Tm scns Reactivity Insertion at 1 pcm/sec 0.00.I RatrTrip Signal Initiated 7."9."

• Trip on high neutron flux

• • Trip on OTAT Results - RWAP 100 pcm/sec Case Figure 4.5-4 shows the core power response for the current FSAR analysis and the Dominion model. The Dominion model trips on a high neutron flux at about 1.17 seconds, compared to about 1.29 seconds for the current FSAR analysis. The 100 pcm/sec transient is a fast transient and the time period before the reactor trip is so brief that any differences in fuel pin heat transfer modeling assumptions have little impact on Doppler reactivity feedback. Overall, the Dominion model peaks at a higher, thus more conservative power level.

p.4o6 pg. 44 of 61 The pressurizer pressure response is shown in Figure 4.2-5. The Dominion model matches very well with the FSAR analysis. The main difference being that the Dominion model peaks at a higher pressure than the FSAR analysis. This correlates with the power response shown in Figure 4.2-4 where the Dominion model peaks at a higher overall nuclear power.

Figure 4.2-6 shows the vessel average temperature. For the 100 pcm/sec case the Dominion model matchs very closely with the FSAR analysis Table 4.5-3 RWAP 100 peru/see Time Sequence of Events Event Time (seconds)

Dominion FSAR Reactivity Insertion at 100 pcrn/sec 0.0 0.0 Reactor Trip Signal Initiated 1.17*

1.29*

• Trip on high neutron flux p.4o6 pg. 45 of 61 Figure 4.5-1 RWAP - 1 pcrn/sec Nuclear Power 1.40 1.20 1.00

~50.80

.' 0.60 oz 0.40 0.20 0.00 2400 2350 2300 2250

.S* 2200 2150 2100 2050 2000 1950

~

I I

I I

I I

I I

I I

-" -- FSAR Dominion tI iIi ii 0

20 40 60 80 100 120 140 Time (sec)

Figure 4.5-2 RWAP - 1 pcrn/sec Pressurizer Pressure V

I I

I '

I Dominion 0

20 40 60 Time (sec) 80 100 120 p.4o6 pg. 46 of 61 Figure 4.5-3 RWAP - 1 pcm/sec Vessel Average Temperature U-0 Q)

I-U) 0.E 0)I-610 605 600 595 590 585 580 575 570 565 0

20 40 60 80 100 Time (sec)

Figure 4.5-4 RWAP - 100 pcm/sec Nuclear Power 120 1.40 1.20 1.00 S0.80 o

4)

N~ 0.60 E

0z 0.40 0.20 0.00 0

2 4

6 8

10 Time (sec)

Attahmet 2pg.

47 of 61 Figure 4.5-5 RWAP - 100 pem/sec Pressurizer Pressure 2400 2350 2300

• 2250

&2200

~2150 2100 2050 2000 1950 0

2 4

6 8

10 Time (sec)

Figure 4.5-6 RWAP - 100 pcm/sec Vessel Average Temperature 600 595 590 585

.~580 S575 570 565 560 0

2 4

6 8

10 Time (sec) pg. 48 of 61 Summary - RWAP The Dominion Millstone model provides results that are similar to the FSAR analysis for the RWAP event. At higher insertion rates, the results match very well. At lower insertion rates, the power increases at a greater rate in the Dominion model than the FSAR model.

However, the temperature increases to a higher peak in the FSAR analysis. The Dominion MPS3 analysis is presented for benchmark comparison, and does not replace the existing AOR.

p.4o6 pg. 49 of 61 4.6 Main Feedwater Line Break The Main Feedwater Line Break (MFLB) event is defined as a break in a feedwater line large enough to prevent the addition of sufficient feedwater to the steam generators to maintain shell side fluid inventory in the steam generators. If the break is postulated in a feedline between the check valve and the steam generator, fluid from the steam generator may also be discharged through the break. Depending upon the size of the break and the plant operating conditions at the time of the break, the break could cause either a RCS cooldown (by excessive energy discharge through the break) or a RCS heatup. The FSAR analysis presents the RCS heatup scenario.

A major feedwater line rupture is classified as an ANS Condition IV event as discussed in FSAR Section 15.0.1. A main feedwater line rupture is the most limiting event in the decrease in secondary heat removal category. Based on a number of prior analyses, it is concluded in FSAR Section 15.2.8 that the most limiting feedwater line rupture is a double ended rupture of the largest feedwater line, occurring at full power with and without offsite power available. Cases both with and without offsite power available are simulated for the benchmark analysis herein.

The MFLB transient is initiated in the Dominion model by opening the break on steam generator 1 and stopping, main feedwater to all four steam generators (SG) as the reactor is operating at full power. Upon transient initation, the break path opens and allows blowdown from the faulted SG secondary side inventory to the atmosphere. The input parameters are the same as those used in the FSAR Chapter 15 analyses as shown in Table 4.6-1 below.

The results for the MFLB transient need to demonstrate that the reactor core remains covered, the RCS does not overpressurize, and the AFW system is able to adequately remove decay heat.

p.5o6 pg. 50 of 61 Table 4.6-1 M4FLB Input Summary Parameter Value Notes Initial Conditions Core Power (MW)

3723, Includes 2% uncertainty RCS Flow (gpm) 363,200 Thermal Design Flow Vessel TAvG (F) 594.5 Nominal + 5 0F RCS Pressure (psia) 2300 Nominal + 50 psi Pressurizer Level (%)

71.6 Nominal + 7.6%

SG evl %)62 Nominal + 12% (Faulted Loop)-

SG evl %)38 Nominal - 12% (Intact Loops)

SG tube plugging (%)

10 Maximum Pump Power (MW/pump) 5.0 Maximum Assumptions/Configuration Low-Low Level Reactor Trip Setpoint 0%

% narrow range span in faulted SG Pressurizer: sprays, heaters, PORVs Not credited AFW Temperature (F) 120 Max value Auxiliary feedwater flow rate (gpm)

.Variable as function of SQ press.

All MFW assumed lost at time of Main Feedwater 0

bra Reactivity Parameters Doppler Reactivity Feedback Most Cnevtvasmto Moderator Feedback Negative Cosraiesumtn Results - MFLB Case With Offsite Power Available The results for the MFLB case with offsite power available are presented on Figure 4.6-1 through Figure 4.6-8.

The nuclear power response (Figure 4.6-1) predicted by the Dominion model is in good agreement with the FSAR data, with the reactor trip occurring on low-low steam generator level.

There is a return to power between approximately 100-200 seconds due primarily to moderator reactivity feedback effects during the primary side cooldown prior to steam line isolation (SLI). After that time, the core remains subcritical for the duration of the transient.

The response for pressurizer pressure and pressurizer water volume are shown on Figure 4.6-2 and Figure 4.6-3.

The Dominion results trend well with the FSAR results for pressurizer pressure and water volume. One difference is a brief increase, in pressurizer pressure and associated insurge into the pressurizer around the point of reactor trip for the Dominion case.,This increase occurs due to differences in the primary-to-secondary heat transfer following the reactor and turbine trips between the MNSG FSAR model and the Dominion SNSG. The SNSG responds more quickly to the decrease in secondary Side level following the loss of main feedwater compared to the MNSG, which initially p.5o6 pg. 51 of 61 experiences less reduction in SG level and associated heat transfer. This effect only occurs for a relatively brief duration. Eventually, steam line isolation (SLI) occurs on low steam line pressure resulting in a primary side heatup as the intact SGs repressurize.

Pressurizer pressure increases until the pressurizer safety valve (PSV) setpoint is reached and remains essentially constant at the PSV relief pressure until a downturn in pressure occurs near the end of the transient.

This indicates the termination of the event as sufficient cooling is being provided by auxiliary feedwater (AFW) for the removal of primary side energy.

The hot leg and cold leg temperature response is shown on Figure 4.6-4 for the faulted loop and on Figure 4.6-5 for the intact loops. There is good agreement between the Dominion and FSAR cases with temperatures exhibiting the same trends throughout the.

event and deviating only slightly prior to SLI, which has a negligible effect on the overall results for this comparison due to the long term nature of this event. As noted for the pressure response discussion above, the temperatures are decreasing at the end of the transient indicating adequate long term heat removal.

The Dominion RCS flow fraction results are shown on Figure 4.6-6. Since power to the reactor pumps is not lost for this case, flow is maintained throughout the transient and varies only with coolant conditions. The Dominion case is in good agreement with the FSAR data throughout the transient.

The secondary system pressure response is presented on Figure 4.6-7 where SG pressure increases briefly following the reactor trip then decreases due to the loss of fluid mass through the feed line break. After SLI occurs, the intact SG pressure increases to the MSSV setpoint while the faulted SG pressure continues to decrease to atmospheric pressure as the remaining fluid mass is depleted. The Dominion and FSAR cases show good agreement as both the magnitude and trends of faulted and intact loops are consistent following the point of reactor trip and subsequent SLI.

Figure 4.6-8 shows excellent agreement between the main feedwater break flow rate response in both the Dominion and FSAR case. One difference is seen around the point of reactor trip over a period of approximately 12 seconds that is related to the steam generator modeling differences. As discussed relative to the pressurizer pressure response, the Dominion SNSG model results in a faster reduction in liquid level and more rapid increase in break flow quality such that flow falls off more quickly as the break is uncovering. After this brief transition period the break flow rates continue to agree well and this difference has a negligible effect on the overall transient response.

pg. 52 of 61 Figure 4.6-1 MFLB - Nuclear Power (case with power) 1.2

• 0.8 w

o0.6 n- 0.4 0.2 110 100 1000 Time (sec) 10000 Figure 4.6-2 MFLB - Pressurizer Pressure (case with power) 2600 2400 2200 2000 1800 1600 1400 1200 1000 10000 1

~~~10 10 00 100 Ta me (sec) 1000 p.5o6 pg. 53 of 61 Figure 4.6-3 MFLB - Pressurizer Liquid Volume (case with power) 2000 1600 1400

.o 1200 100 1000 600 400 200 180 i

~

~

iiI i i i

" i 00 N

W I

omr°' li L

10 10 100 1000 lime (sec) 10oo 10000 Figure 4.6-4 MFLB - RCS Temperatures - Faulted Loop (case with power) 650

• '600 E 550 500 450 10 10 100 1000

-time (sec) 1000 10000 pg. 54 of 61 Figure 4.6-5 MFLB - RCS Temperatures - Intact Loops (case with power) 700 650

• '600 a,

E550 500 450 10 10O0 1000 10000 Tlime (sec)

Figure 4.6-6 MFLB - Normalized RCS Flow (case with power) 1.15

____l ii i

i I i 1.1

• i

• i i C

o 1

C

• 0.95 S0.9 0.8 0.8 0.75 0.7 10 100 "time (sec) 10 000 10000 pg. 55 of61 Figure 4.6-7 MFLB - Steam Generator Pressure (case with power) 1400 1200 1000 800 600 400 200 10 100 1000 lime (sec) 10000 Figure 4.6-8 MFLB - Feed Line Break Flow (case with power) 8000

)

7000

"= T 6000

/

.G5000 2000 1000 lime (sec) 10000 1000

q I

~~~Attachment 2

p.5o6 pg. 56 of 61 Results - MFLB Case Without Offsite Power Available The results for the MFLB case without offsite power are similar to the case with power available but are generally less limiting for long-term primary side heat removal since the RCPs are not running and adding heat to the primary side fluid.

The nuclear power response (Figure 4.6-9) predicted by the Dominion case is in good agreement with the FSAR data. As shown for this case, there is no return to power during the early portion of the cooldown due to less reactivity feedback and the reactor core remains subcritical for the duration of the transient.

The responses for pressurizer pressure and primary side temperatures are shown on Figures 4.6-10 through 4.6-12. As discussed above for the case with offsite power, the Dominion case exhibits a brief increase in pressure around the time of reactor trip but otherwise the response is similar to the FSAR case with long-term pressure maintained at the PSV setpoint. The hot leg and cold leg temperature response shown on Figure 4.6-11 and Figure 4.6-12 also demonstrate similar trends. One difference is that the cooldown that occurs prior to SLI is more pronounced for the Dominion case, which is primarily attributed to higher primary to secondary heat transfer. This is the result of a somewhat slower rate of flow decrease following the RCP trip for the Dominion case, resulting in maintaining better primary side heat removal during that phase. In addition, SLI occurs slightly later in the Dominion case, which also enhances heat removal prior to the time of isolation. Similarly, the delay in break isolation delays the point of steam generator dry-out, such that additional heat is extracted through the break. As shown, these differences have little effect on the long-term temperature response as the Dominion and FSAR temperatures agree very well through the end of the transient. This case results in lower long-term temperatures, as the RCPs trip due to the loss of offsite power and do not contribute any pump heat to the system.

The secondary system pressure response, presented in Figure 4.6-13, is similar to the response for the case with power. Since there is less primary side heat generation and heat removal for this case, the SG depressurizes more quickly and SLI occurs earlier in the transient, compared to the case with offsite power available. Long term trends are similar with heat removal via the MSSVs on the intact SGs. There is good agreement between the Dominion and FSAR cases with the FSAR case depressurizing slightly faster prior to SLI.

The Dominion RCS flow fraction results are in good agreement with the FSAR result as shown on Figure 4.6-14, where the loss of flow associated with the loss of power and associated RCP trip are seen.

As noted above, the flow decreases somewhat more p.5o6 pg. 57 of 61 quickly for the FSAR case, which appears to affect the intermediate temperatures but does not impact the long term temperature results.

Figure 4.6-15 shows good agreement between the main feedwater break flow rate response in both the Dominion and FSAR data. The small differences seen around the point of reactor trip are due to differences in the Dominion SNSG and the FSAR MNSG as discussed above for the case with power available. That is, the Dominion SNSG model results in a faster reduction in liquid level and more rapid increase in break flow quality such that flow falls off more quickly as the break is uncovering.

After this brief transition period the break flow rates continue to agree well and this difference has a negligible effect on the overall transient response Figure 4.6-9 MFLB - Nuclear Power (case without power)

I i

• I i

1

! I I

t l i E

o nihI gO.

ii i

i ii U04 ii*

~

0.2 i

110 100 1000 10000 "Time (sec) pg. 58 of 61 Figure 4.6-10 MFLB - Pressurizer Pressure (case without power) 2500 2400 2300 2200

• 21 00 2000 1900 1800 1700 1600 1500 110 100 1000 10000 Trme (sec)

Figure 4.6-11 MFLB - RCS Temperatures - Faulted Loop (case without power) 700 650 600 5O0 45O 1

10 100 1000

-time (sec) 10000

SI pg. 59 of 61 Figure 4.6-12 MFLB - RCS Temperatures - Intact Loops (case without power) 700 650 u_600

S50, 45O 110 100 1000100 Tlime (see) 10000 Figure 4.6-13 MFLB - Steam Generator Pressure (case without power) 1200

____i*

t

}-,

.*, 00 i

I*

200

~

Dii ! M ip j ll

-- FSA*Rlaa edoPi I

11i i 1 0

1 Do Ill F* tei ap 10

-*~~~

10101000 10000 "lime (sec) pg. 60 of 61 Figure 4.6-14 MFLB -Normalized RCS Flow (case without power) 1.2 i

i " -'

1

!i FSAR 0u.8 006 I

°0.:

i 1 }

!a

--I 10 10O0 1000 10000 Figure 4.6-15 MFLB - Feed Line Break Flow (case without power) 8000 7000 li i

i'i° i

• o T* i *,

0o n

o o

5000O

,-.-t-ii

-fi i

I i

ii -

Time (sec) pg. 61 of 61 Summary - MFLB The Dominion Millstone model provides results that are similar to the FSAR analysis for the MFLB event. Two cases are analyzed, one with offsite power available and another without offsite power. Some small differences are observed early in the transient for RCS temperatures, which are attributable to differences in the Dominion SNSG model and the FSAR MINSG model; however, these differences have a negligible effect on the long-term primary side heat removal and associated temperature response. All acceptance criteria are satisfied for both cases.

5.0 Conclusions This attachment presents benchmarking transient analyses performed with the MPS3 RETRAN model developed in accordance with VEP-FRD-4 1-P-A. These analysis results are compared with current Millstone FSAR results. The following conclusions are drawn based on these analyses.

1) It is demonstrated that the Dominion RETRAIN-3D model and analysis methods can predict the response of transient events with results that compare well to FSAR results.

2) Where there are differences between the Dominion results and the FSAR results, they are understood based on differences in noding, inputs, or other modeling assumptions.

3) The Dominion Millstone RETRAN-3D model is consistent with current Dominion methods (Reference 1). These methods have been applied extensively for Surry and North Anna licensing, engineering and plant support analyses.

4) The RETRAN comparison analyses satisfy the applicability assessment criteria and provide further validation of the conclusion that Dominion's RETRAN analysis methods are applicable to Millstone and can be applied to Millstone licensing analysis for reload core design and safety analysis.

6.0 References

1)

Topical Report, VEP-FRD-41-P-A, Rev. 0.2, "VEPCO Reactor System Transient Analyses Using the RETRAN Computer Code," March 2015.

2)

Topical Report, VEP-NFE-2-A, "VEPCO Evaluation of the Control Rod Ejection Transient," December 1984.