ML20083L749

From kanterella
Jump to navigation Jump to search
Analytical Justification for Treatment of Reactor Coolant Pumps During Accident Conditions
ML20083L749
Person / Time
Site: Arkansas Nuclear 
Issue date: 03/30/1984
From:
BABCOCK & WILCOX CO.
To:
Shared Package
ML20083L738 List:
References
TASK-2.K.3.05, TASK-TM 77-1149091, 77-1149091-00, NUDOCS 8404170352
Download: ML20083L749 (67)
v • d • e


Text

{{#Wiki_filter:, I ANALYTICAL JUSTIFICATION FOR THE TREATMENT OF REACTOR C0OLANT PUMPS DURING ACCIDENT CONDITIONS B&W Document No. 77-1149091-00 t Prepared for Arkansas Power & Light Company Consumers Power Company Duke Power Company Florida Power Corporation GPU Nuclear Sacramento Municipal Utility District Toledo Edison Company by Babcock & Wilcox Utility Power Generation Division P. O. Box 1260 Lynchburg Virginia 24505 h. hYDO O 9

) CONTENTS Page 1. INTRODUCTION........................... 1 2.

SUMMARY

AND CONCLUSION...................... 2 3. BACKGROUND............................ 4 4. SIGNAL SELECTION / TRIP SCHEME PHILOSOPHY 5 I 5. JUSTIFICATION FOR THE TREATMENT CF RC PUMPS 8 5.1. Conservative Analysis Performed to Evaluate the Effect of Delayed RC Pump Trip on SBLOCA........ 9 5.2. Best Estimate SBLOCA Analysis 10 5.2.1. SBLOCA Method of Analysis 10 5.2.2. Best Estimate SBLOCA Spectrum Results 13 5.2.3. Break Size -- Trip Time Sensitivity 16 5.2.4. Analysi s Resul ts.................. 17 5.2.5. Analysis Applicability to Raised Loop Design.... 18 5.2.6. Conclusions 19 5.3. Setpoint and Signal Selection 'for Non-LOCA Events 19 5.3.1. Best Estimate Analysis of a Single Double-Ended Steam Generator Tube Rupture 20 5.4. Other Considerations.................... 26 5.4.1. Potential Containment Isolation of RC Pump Se rv i c e s.................. 26 R EFE RE NC ES............................ 63 List of Tables Table 4-1. RC Pump Trip Results for 177-FA LL Plants 7 5-1. HPI and Makeup System Capacities................. 30 5-2._ MINITRAP2 Node Description................... 31 5-3. MINITRAP2 Path Description................... 32 5-4. SGTR Sequence of Events 33 - 111 - m

List of Figures Figure Page 34 5-1. CRAFT 2 Noding Diagram for Small Breaks 5-2. Critical Region for RC Pumps Trip, Break Size-35 Vs Time After Trip 5-3. " Realistic" Core Axial Peaking Distribution.......... 36 i 5-4. Average Primary System Void Fraction Excluding Pressurizer 37 38 5-5. RC System Pressure Vs Time.......b.......... 5-6. Liquid Volume in Reactor Vessel for 0.1-ft CLD 39 Break -- RCP Trip 91000 s 40 i 5-7. RC System Pressure Vs Time.......I.......... 5-8. Liquid Volume in Reactor Vessel for 0.2-ft CLD Break -- RCP Trip 9 400 s................... 41 42 l 5-9. RC System Pressure Vs Time 2 CLD 5-10. Liquid Volume in Reactor Vessel for 0.25-ft Break -- RCP Trip 9 300 s................... 43 44 5-11. RC System Pressure Vs Time.......$.......... i 5-12. Liquid Volume in Reactor Vessel for 0.3-ft CLD Break -- RCP Trip 9 300 s................... 45 5-13. S8LOCA Spectrum System Pressure Vs Time............- 46 5-14. -Liquid Volume in Reactor Vessel Following RC Pump 2 47 Trip for a 0.2-ft PD Break.... $........._..... Break at. 5-15. System Pressure Vs Time for a 0.2-ft 48 the Pump Discharge '...................... 3 Vs Time for a 0.2-ftgding Volume-Average TemperatureBreak at Pump Dis 49 e Maximum Hot Spot Cla 5-16. 5-17. Comparison of High Pressure Injection System Capacities 50 t for Raised-Loop and Lowered-Loop Design Plants 5-18. Mini-TRAP 2 Noding and Flow Path Schematic........... 51 5-19. Detailed Mini-TRAP 2 Noding/Flowpath Diagram i 52 for SGTR Model 5-20.- Steam Generator Tube Leaks -- Operator Action Outline..... 53 5 -21. Leak Fl ow V s Ti me....................... 54 5-22. Total Reactor Power Vs Time After Rupture of a Single 55 Double-Ended SG Tube..................... 56 5-23. Power Runback to 15%...................... 5-24. Pressurizer Inventory Vs Time Af ter Rupture.......... 57 58 5-24. Hot and Cold Leg Temperature Vs Time After Rupture 58 5-26. Core Outlet Pressure Vs Time After Rupture 5-27. Surge Line Flow Vs Time Af ter Rupture.............. 59 5-28. Total RCS Charging Flow Vs Time After Tube Rupture 60 61 5-29. SG Tube Flow Vs Time After Rupture 5-30.. Subcooling Margin Vs Time After Rupture............ 62 - iv - . -.-.-.LL ---.-

- - ~ . -. - -. -. -. - _. - - ~. -. - - 1 1 i ( 1. INTRODUCTION I I j The criteria for resolution of TMI Action Plan Item II.K.3.5, " Automatic Trip of Reactor Coolant Pump," were stated in lettersl.2 dated February 8, 1983 from Mr. Darrel G. Eisenhut (NRC) to all applicants and licensees with l B&W-designed nuclear steam systems (NSSs). Those letters requested each 4 l utili1;y to provide an individual submittal or reference to a generic submit-l tal providing the technical justification for the treatment of reactor cool-ant (RC) pumps during. transients and accidents. This report provides the technical justification for the treatment of RC f pumps during transient and accident conditions. This report presents the generic analyses of steam generator tube rupture for the 177-fuel assembly (FA) lowered-loop design and SBLOCA results of both raised-and lowered-loop 177-FA plant designs. These analyses support the tripping of all four i j RC pumps manually on indication of " loss of subcooling margin." Generic aspects of the B&W-designed NSSs as they relate to the treatment of l RC pumps are discussed in this report with the intention that this report - will be supplemented by each utility member of the B&W 0wners Group in pro-t I viding a response to NRC Generic Letter 83-10.1,2 Where an individual utility commitment is requested, it will be addressed on a plant-specific basis by each utility, and is not included in this report. t l i I i 9 1 i i i 1 j' "

2.

SUMMARY

AND CONCLUSION It is the position of the B&W Owners Group and B&W that the tripping of all four RC pumps is recommended following indication of a small break loss-of-coolant accident (SBLOCA) and that it can be achieved safely and reliably by the operator. This report demonstrates that the concept " loss of sub-cooling margin" is an appropriate signal to alert the operator of the need for pump trip and meets the intent of the criteria identified in Generic Letter 83-10. A loss of subcooling margin will occur for those small break LOCAs where a pump trip is required to show compliance with 10 CFR 50.46. As a result of best estimate SBLOCA analyses, it is concluded that times in excess of 10 minutes are available for manual operator action to trip the RC pumps following indication of a loss of subcooling margin. Fur-thermore, within the 10-minute time frame, the most limiting break size / trip time combination yields acceptable peak cladding temperatures far be-low the Ifmits of 10 CFR 50.46. This conclusion can be compared to tho l minimum 2-minute RC pump trip time predicted for the limiting break size l analysis using conservative methods with Appendix K assumptions. The in-l terrelationship between break size and RC pump trip time, which deter-I mines the critical region for conservatively predicted unacceptable conse-quences is shown in Figure 5-2. Adequate subcooling margin is maintained during steam generator tube rup-I ture events for ruptures up to and including the double-ended rupture of i a single tube, to ensure forced circulation throughout the event, if the operator follows procedures based on the Abnormal Transient Operating Guidelines (ATOG). Reducing the naed to trip the RC pumps for more likely non-LOCA events such as mild overcooling events is ensured by a judicious determination s -

of the subcooling margin setpoint. Procedures based on AT0G provide guidance for pump restart for those events where an unnecessary pump trip might occur. Consequently, reliance en the PORY for depressurization is unlikely. i The RC pump trip criteria based on loss of subcooling margin precludes operation of the RC pumps in a highly voided system. i i I l J 1 i 3-

l i i l 3. BACKGROUND l, The treatment of RC pumps du *ing accidents and transients has received ex-j tensive attention over the past several years. The B&W Owners Group has 3 I performed analyses in response to IE Bulletin 79-05B evaluating the effect of a delayed RC pump trip using Appendix X assumptions during the course of a small break LOCA accident and has determined that an early trip of RC pumps is required to show conformance to 10 CFR 50.46 for a range of break sizes. Therefore, to be consistent with the conservative analyses per-formed, it is the position of the B&W Owners Group that all four RC pumps should be tripped if indications of a small break LOCA exist. The B&W Owners Group maintains that.it is highly - desirable to maintain RC pump operation during non-LOCA events, as an aid in the mitigation of the transient. Consistent with this philosophy, the concept of subcooling mar-gin was chosen as an indication for the need to trip all four. RC pumps. It is the intention of the report to demonstrate that this concept is consis-tent with tne B&W Owners Group philosophy for handling RC pumps during tran-sient conditions and complies with the intent of the criteria stated in i Generic Letter 83-10. The symptom approach of subcooling margin,. developed. l as part of the Abnormal Transient Operating Guidelines (AT0G) Program, is intended to replace the present guidelines of tripping solely on the pres-ence of a low RC - pressure engineered safety features actuation system (ESFAS) signal. 1 This report is based on the above positions and demonstrates that the con-f cept of subcooling margin is an appropriate indicator of the need to trip all four RC pumps -yet still allows continued operation for steam generator tube ruptures less than or equal to a single double-ended rupture. Justi-fication is ~also provided for manual initiation of RC pump trip on loss.of subcooling. 2 _4 v

i 4. SIGNAL SELECTION / TRIP SCHEME PHILOSOPHY The concept of subcooiing margin was chosen as an indication for the need to trip all four RC pumps during transient conditions. No partial or staggered RCP trip schemes are considered except for the extreme case where mechanical damage to the pump is likely as this adds to increased decision making on the part of the operator during transient conditions. A primary objective of the parameter and setpoint selection is the avoid-ance of RC pump trip for non-LOCA events particularly steam, generator tube rupture (SGTR). Realistic operator actions in accordance with the ATOG pro-cedures are shown in section 5.3 to avoid loss of subcooling 'and the need to trip the RC pumps for this event. Furthermore,f since subcooling margin would be quickly regained following makeup or HPI initiation, without loss of natural circulation even if the operator failed to' take actions to pre-vent RCP tripping and ESFAS. actuation, restart of the pumps would be al-lowed. Consequently, reliance on the PORY for depressurization is un-likely. A loss of subcooling will always occur for small breaks that have the po-tential to uncover the core and violate 10 CFR 50.46 criteria if the RCPs are tripped under certain two-phase conditions. This' is demonstrated in section 5.2. Hence, the loss of subcooling margin can be used as a key indicator for RC pump trip. 2 For most small break LOCAs (those larger than 0.05 ft ) rapid depressuriza-tion of the RCS with little or no decrease in -the RCS temperature causes the RCS pressure to quickly decrease to the saturation pressure. As shown in Table 4-1, for 177-FA plants, a loss of subcooling occurs within 17 sec-onds af ter initiation of the LOCA for small breaks 0.05 ft2 and larger. f ' {

1 i The RCP trip on loss of subcooling margin will exclude extended RC. pump operation in a voided system. The use of loss of subcooling is a suf-ficient indicator to assure that the RC pumps will be tripped for all losses of primary coolant in which RC pump trip is considered necessary. 'I Table 4-1. RC Pump Trip Results for 177-FA LL Plants (Conservative Appendix K Analysis) Time to reach indication, s Low RCS Conservative press. RPS low RCS Break RCS subcooled reactor trip press. ESFAS 70% void

size, margin before setpoint Saturation setpoint fraction sq ft accident (1900 psia) in RCS (1365 psia) in RCS 0.3 45.46 0.46

<0.1 7.75 130 0.2 45.46 0.46 <0.1 9.37 180 0.1 45.46 0.95 6.0 12.14 420 0.075 45.46 1.94 9.0 13.51 640 0.05 45.46 9.54 17.25 17.68 920 0.025 45.46 19.83 46.3 44.08 l 5. JUSTIFICATION FOR THE TREATMENT OF RC PUMPS Analyses of certain small break LOCAs, combined with the assumption of only one HPI train available, have demonstrated the potential for exceeding 10 CFR 50.46 limits if RC pumps are tripped while the RCS is in a highly voided (>70%) condition. Consistent with these results the B&W Owners Group has adopted the position of tripping all four RC pumps on indication of a loss-of-coolant accident. It is also concluded that there is a wide range of transients, especially SGTR, and LOCAs where it is desirable for an operator to maintain forced circulation cooling and mixing through opera-tion of the RC pumps. This section addresses the need for tripping the RC pumps upon indication of a loss of subcooling margin which is indicative of a small break LOCA and the appropriate timing of this action. The ability of the signal to discriminate between a small break LOCA of the appropriate size range of f interest and a more likely event, the steam generator tube rupture, is demonstrated by a realistic analyses of the design basis steam genenator tube rupture event. A summary of the SBLOCA analysis performed in response to IE Bulletin 79-05C, evaluating the effect of delayed RC pump trip -is provided in sec-tion 5.1. Section 5.1 includes model assumptions and conclusions obtained from that analysis. Section 5.2.1 describes the analytical methods and as-sumptions used in the best estimate SBLOCA analysis for the generic 177-FA lowered-loop plant type. The results of the best estimate SBLOCA analysis are provided in sections 5.2.2 and 5.2.4. Section 5.2.3 discusses the break size versus RCP trip times sensitivity relationships. A qualitative discussion of the applicability of the generic 177-FA lowered-loop results to the raised-inop plant is provided in section 5.2.5. Section 5.2.6 pro-vides the salient conclusions from the SBLOCA analyses. 1 ~. 5.1 Conservative Analysis Performed to Evaluate the Effect of Delayed RC Pump Trip on SBLOCA In response to Item 2 of IE Bulletin 79-05C, a conservative evaluation was performed to determine the effect of delayed RC pump trip during the course of small break LOCAs and concluded that early trip of the RC pumps was re-quired to show conformance to 10 CFR 50.46. A spectrum of small break LOCAs was analyzed for a 177-FA plant, both raised-and lowered-loop designs using Appendix K assumptions. The break 2 sizes ranged from 0.025-to 0.30-ft. A summary of timing results is shown in Table 4-1. The analysis method used for this evaluation is basically that described in section 5 of BAW-1010a, Rev. 3, "B&W's ECCS Evaluation Model," and the let-ter from J. H. Taylor (B&W) to S. A. Varga (NRC), dated July 18, 1978, which is. applicable to the 177-FA lowered-loop plants for power levels up ~ to 2772 MWt. A simplified 6-node CRAFT 25 model similar to that shown in Figure 5-1 was used for the analysis. Although this 6-node model is simp-lified compared to that described in the above referenced letter, it does maintain RCS volume and elevation relationships which are important to properly evaluate ' the system response during a small LOCA with RC pumps - i running. Studies were performed to benchmark the 6-node model against the 23-node small break LOCA evaluation model (referenced letter), and the results indicate that the simplified model is acceptable for the RC pump trip ' analyses. The following conclusions can be drawn from the previously described analysis:

  • If the RC pumps remain operative, core cooling is assured regardless of' system void fraction.

2 For breaks greater than.0.025 ft, the.RCS may ' evolve to system void frac-tions in' excess of 90%. At 40_ minutes, the 0.025-ft2 break has evolved to only a 47% void frac-tion. Thus, a _ delayed RC pump trip for-breaks 'less than 0.025-ft2 wilj ~ . not result in core uncovery. W --9'- ._,.m.

The potential for high cladding temperatures for a small break transient wi th delayed RC pump trip is restricted to a time period between that time where the system has evolved to a high void fraction and the time of LPI actuation. Even with two HpI pumps available, tripping of the RC pumps at the worst time (90% void fraction) results in a core uncovery period which cannot be shown to comply with 10 CFR 50.46, if Appendix K assumptions are utilized. There exists a combination of bred sizes and RC pump trip times which resulted in violation of 10 CFR 50.46 limits. A plot of break size ver-sus RC pump trip time which results in unacceptable consequences is shown in Figure 5-2. This curve indicates that a prompt RC pump trip upon re-ceipt of a low pressure ESFAS signal (which is approximately the same as loss of subcooling margin) will provide compliance to 10 CFR 50.46. The minimum time available for pump trip is approximately 2 minutes and is 2 detemined by extrapolation beyond the 0.2-ft case. 5.2 Best Estimate SBLOCA Analysis A re-analysis of the small break LOCA spectrum for 177-FA raised-and lowered-loop plant designs, discussed in section 5.1, was performed from a "best estimate" approach to evaluate the impact on required times for RC punp trip under realistic conditions. As expected, the time available for tripping the RC pumps following indication of loss of subcooling margin in-creased. The best estimate analyses described in section 5.2.1 is an exten-sion of the conservative analyses described in section 5.1. Realistic as-sumptions, described in section 5.2.1, were substituted along with a detailed methodology for evaluating the clad temperature response. 5.2.1. SBLOCA Method of Analyses (Best Estimate) The analytical model used for the best estimate evaluation is similar to that described in section 5.1. A simplified 6-node CRAFT 2 model, as shown in Figure 5-1, was used for evaluating the RCS themal-hydraulic responses to a - SBLOCA. Node 1 contains the cold leg pump discharge piping, down-comer, and RV lower plenum. _ Node 2 is the primary side.of the steam genera-tor and the pump suction piping. Node 3 contains the core, RV upper plenum _

l and the hot legs. Node 4 is the pressurizer and Nodes 5 ana 6 represent the containment and the steam generator secondary side, respectively. This model, although simplified compared to those utilized in small break LOCA evaluation analyses, maintains RCS volume and elevation relationships that are important to properly evaluate the system response during a SBLOCA with the RC pumps running. Key assumptiens which differ from those described in section 5.1 are discussed below. RC Pump Model The MIT two-phase pump model based on the steam-water test data was used in the CRAFT 2 model for two phase degradation calculations. This model pro-vides less severe flow degradation than that used in the previous analysis. Thus, it enhances heat removal by the steam generators during the forced circulation phase of transients, and results in a lower leak fluid quality. But the new pump model has a limited effect on the overall system response during the forced circulation phase of transients. HPI Flow The HPI system flows for generic analyses are shown in Table 5-1 and Figure i 5-17. For the best estimate analyses, two HPI pumps are assumed with 30% of one HPI flow going directly out the break. l Leak Discharge Model The selected discharge model is - the orifice equation for subcooled flow and the HEM for saturated flow both with a discharge coefficient of- 0.85. This l ' is a best estimate discharge model developed for the revised SBLOCA evalua-tion mode 16 in compliance with NUREG 0565. The conservative analyses 'de-scribed in section 5.1 used an orifice (subcooled) - Moody (saturated) dis-charge model with a - discharge coefficient of 1.0. . This latter model has significantly higher leak flowrates than that of the best estimate model. Equipment Availability ~ ' Both analyses. assumed that the RC pumps remain operative after the receipt of the RCS loss of subcooling margin indication. The RC pumps were assumed to trip when the RCS has evolved to. high void fractions. Also, the best l- - estimate analysis assumed that two HPI pumps and AFW are.available for core t

cooling as opposed to one HPI and AFW for the previous analysis. Approxi-mately 30% of one HPI flow is assumed to spill out the break. The AFW 1evel is raised to 50% of the operating level immediately af ter the pump trip. Power Level and Decay Heat An operating power level of 2772 MWt is assumed for all analyses, which is bounding for all 177-FA plants. The best estimate analyses were performed using the more realistic 1.0 times ANS decay heat curve instead of 1.2 times ANS decay curve, which was used in the conservative Appendix K anal-ysis. The decay heat has no impact on the peak cladding temperature (PCT) during the early phase of transients when the RC pumps are running, but will have a significant impact on PCT af ter the pumps are tripped and the core is uncovered. F0AM Calculations Following a RC pump trip at a relatively high RCS void fraction, steam-water separation occurs and the core is uncovered. In the conservative analyses described in section 5.1 the core is assumed to undergo adiabatic heatup until the liquid level reaches the 9-ft level (equivalent to approxi-mately 12 ft mixture level). The PCT calculations were based on an axial power peak above. the core midplane obtained from a power shape encountered during nomal operation (Figure 5-3) and on a heatup period from the RC L pump trip to the time of.9-ft core recovery. Use of an adiabatic heatup as-f sumption neglected any ~ credit ~ for the steam cooling that will occur during the core refill phase and for the radiation-heat : transfer. Therefore, the maximum clad temperature was overestimated and reflected in. the' available-I time for a manual RCP trip. This best estimate analysis, however, utilized a more detailed method to de-7 l temine the maximum clad temperature. ' This method included a FOAM 2 - analy-sis; to determine the. inner vessel mixture height. - The FOAM calculation in-cluded major sources of steam production within-the vessel, i.e., steam production due to decay heat ~and flashing. The steaming rates-due to flash - ing in the lower plenum are assumed to provide steam cooling. of. the core during refill of the downcomer. Once the mixture level rises into the core region, the core cooling 'is by pool boiling 'and steam cooling in the un-covered portion of the core. .- 12'- l

The axial power shape shown in Figure 5-3 was used in the F0 Aft calculation and was implemented with a radial peaking factor of.1.0. Thus, the resul-tant mixture height is representative of the average channel condi tions. This method is conservative since a higher peaking factor for the hot chan-1 nel would result in higher froth levels, thus, faster core recovery. Heatup Calculations 8 code in the manner de-The heatup calculation was perfonned using the THETA scribed in section 5 of BAW-10104.12 The following additional assumptions were utilized in the THETA evaluation: 1. The power shape shown in Figure 5-3 was used with a radial power factor of 1.65. This maximizes steam superheating and sets the peak local power at approximately 7 ft core elevation. 2. Coolant flow and mixture level were taken directly from the F0AM calcu-lations for the core region. Steam flow from flashing in the lower plenum was used during the refill of the lower plenum. 1 3. The fuel data generated by the TAC 029 code were used in this analysis. TAC 02 includes mechanistic fuel densification and fission gas release models, which predict realistic volumetric average fuel temperatures and pin pressures as a function of burnup. The fuel input data used in this analysis were calculated at the time of maximum fuel densification (N90 mwd /mtU burnup) at which the volumetric average fuel temperature is at a maximum. 5.2.2. Best Estimate SBLOCA Spectrum Results The break sizes examined for this analysis ranged from 0.1 to 0.3 ft2 in area and are located in the cold leg pump discharge piping. This break location has been shown to bound other locations in determining core un-covering times and associated cladding temperature excursions. Breaks of this size do not result in a rapid system depressurization to LPI actuation pressure before the system evolves to a high void fraction. In addition,- based on the ccnservative Appendix K analysis as discussed in section 5.1, - breaks.in this range yield the minimum amount of time available for manual RC ipump -trip following a SBLOCA. The following paragraphs - discuss the RCS behavior during'the transient for the break size analyzed..

0.1-ft2 Pump Discharge Break The average system void fractions for the break spectrum analyzed is shown in Figure 5-4. The average system void is defined as: Average system void (%) = Y1 - V2 x 100 where y1 = total primary system liquid volume excluding the pressurizer at time = 0, V2 = total primary system liquid volume excluding the pressurizer at time = t. This parameter was utilized to show the primary system liquid inventory since the primary coolant tends to be homogeneously mixed with the RC pumps running. Following the RC pump trip at 1000 seconds, the steam-water sepa-ration occurs. This results in steam discharge out the break and increases system depressurization as shown in Figure 5-5. As the primary system fur-ther depressurizes, the ECCS injection rate increases. Thus, the average system void begins to decrease approximately 200 seconds af ter the RC pump trip. As seen in Figure 5-6, the liquid volume in the reactor vessel increases rapidly following the RC pump trip as a result of water in the pump suction side flowing into the reactor vessel during the period of pump coastdown. As the primary loop flow ceases, the core refill rate decreases to a rate equal to ECCS injection less core boil-off. The core is refilled to the 9-ft level with collapsed liquid approximately 140 seconds after the as-sumed pump trip. Once the core liquid level reaches the 9-ft level, the core is expected to be covered by a two-phase mixture and the cladding tem-perature excursion would be teminated. Even with the conservative assump-tion of adiabatic heatup during the 140-second core uncovery period, the cladding temperature remains below the limit of 10 CFR 50.46. From the re-suits of this break size and the Appendix K analyses, it can be concluded that, for breaks 0.1 ft2 and smaller, more than 10 minutes time is avail-able for manual pump trip following a receipt of the loss of subcooling margin indication. l

0.2 ft2 Pump Discharge Break A comparison of the primary system pressure responses in Figures 5-5 and 5-7 indicates that the primary system depressurization is basically indepen-dent of break size during the first few minutes into the transient when the RC pumps are running. This is because the forced circulation of reactor coolant provides adequate heat transfer to the steam generators; the pri-mary system thus depressurizes to a pressure approximately equal to the set-point of the SG safety valves. The primary system depressurization rate in-creases following a pump trip at 400 seconds (80% void fraction) as the steam is vented out the break. As shown in Figure 5-8, as a result of leak flow and boiloff greater than the HPI fl ow, the liquid volume decreases from the pump coastdown at 420 seconds until the CFTs are actuated at ap-proximately 550 seconds. The collapsed liquid level rises rapidly and reaches 9-ft level at 630 seconds. A partial core uncovering time of ap-proximately 230 seconds necessitates a detailed heatup calculation to eval-uate the consequences of a delayed RC pump trip for this b.eak size. The heatup calculations are provided in section 5.2.3 and 5.2.4. 0.25-ft2 Pump Discharge Break The primary system pressure response -and the RV liquid volume are shown in Figures 5-9 and 5-10, respectively. The RC pumps are tripped at 300 sec-onds (80% void fraction). The core refill rate is slightly higher than 2 that of the 0.2 ft - break. Conservative calculations show acceptable re-sults of a pump trip at 300 seconds. However, similar to the 0.2-f t2 case, l the extended RC pump trip time will require detailed heatup calculations or justification which follows in section 5.2.3. l 0.3-ft3 Pump Discharge Break The 0.3-ft2 break was analyzed in order to assure an upper bound for the SBLOCA spectrum.. The primary system pressure response and liquid volume in the reactor vessel are shown in Figures 5-11 and 5-12, respectively. The RC pump trip is assumed at 300 seconds (90%' void fraction). The primary system-depressurizes rapidly following the pump trip, and the CFTs are actu-ated at 360 seconds causing _ a rapid recovery of the core at about 470 sec-I onds. Due to this fast recovery, this break size is not considered limit-ing for this analysis as discussed in.section 5.2.3. 15 - i t ~

5.2.3. Break Size -- Trip Time Sensitivity The combined results of this analysis and those from the analysis with the Appendix K assumptions, described 'in section 5.1, demonstrated that, for breaks 0.1-ft2 and smaller, more than 10 minutes time is available for man-ual RC pump trip following an indication of the loss of subcooling margin, j 2 For breaks larger than 0.3-ft, the rapid system depressurization to LPI setpoint assures quick core recovery. In the range between 0.1-f t2 and 0.3-ft2 extended core uncovering resulted in a limited time available for manual RC pump trip when evaluated using Appendix K assumptions and an adiabatic heatup estimation of clad temperature. The detailed method used in the best estimate analysis, taking credit for steam cooling, realistical-ly improves the previous results. The most limiting break size is selected by comparing the results of the analyses described in section 5.2.2. Fig-ure 5-13 shows the RCS pressure response for different break sizes. For break sizes of 0.3-and 0.25-ft2 the system pressure drops rapidly to the CFT setpoint. The pressure drop is even faster following the pump trip as the system collapses. Thus, a faster recovery is ensured. The 0.2-f t2 break results in CFT actuation at a later time, and the core recovery is slower since the CFT flow is a function of backpressure. For breaks be-2 tween 0.1-and 0.2-f t, the system inventory is lost at an even slower rate. An early trip will not result in complete core uncovering. There-l fore, steam cooling and a froth level provide significant core cooling. A worst case would be a trip at the later time, at very high void fractions, where the core becomes completely uncovered. Figure 5-14 shows the avail-2 break with the RC pump trip l able liquid volume in the vessel for a 0.2-ft at different times. The RC pump trip at 500 seconds represents the worst case for which the complete core uncovery period is the longest -(S95 sec-onds). As shown, the system collapses and the volume remains just below the core bottom elevation. An earlier trip (400 second trip) results in. l only partial core uncovering and thus higher heat transfer. Trips at later times (600-second-trip) results in shorter complete core uncovering time due to increased ECCS at lower pressure and thus higher recovery rates. 2 For breaks smaller than 0.2-ft, similar trends as those shown in Figure 5-14 will. result, however, shifted to the right on the time axis. The heat-2 break due up rates for these breaks will be lower than that for the 0.2-ft. ~_

to lower decay heat rates. Therefore, the available trip time for these breaks 'would be longer than that for a 0.2-ft2 break. The range of interest, therefore, is narrowed to break areas of approximate-2 ly 0.2-ft. The RCP trip for breaks larger than 0.2-ft2 will result in fas-ter core recovery relative to the 0.2-f t2 results shown in Figure 5-14. ( Furthermore, as shown in Figure 5-13, the LPI will actuate earlier, provid-ing even faster recovery. The limiting case, therefore, was selected as a 0.2-ft2 break at the pump discharge with the RC pumps tripped at 500 sec-i onds. This case, as shown in Figure 5-14, has the longest period of com-plete core uncovering and is considered representative of the most limiting condi tion. Therefore, this break size was chosen to evaluate the cladding temperature response for an extended core uncovery period to demonstrate ac-ceptaoility of the peak cladding temperature. 5.2.4. Analy' sis Results The limiting break configuration analyzed was a 0.2-ft2 break at the pump discharge with the RC pumps tripped at 500 seconds. As discussed in section' 5.2.3, this case resulted in prolonged complete core uncovering. l ~ Figure 5-15 shows the primary system pressure response. Following the ini-tiation of the break and reactor trip on low pressure, the system depressur-ized very rapidly to the secondary pressure. The depressurization con- - tinued at a slower rate as the two-phase mass and er:ergy was being released through the break. The RC pump trip at 500 seconds resulted in phase sepa-ration and higher quality discharge from the break, thus, a faster _ depres-surization of. the system. The available liquid volume drained to the bot-tom of the vessel and refilled the vessel to just below the bottom of the core, as shown in Figure 5-14. At this point the. core became completely uncovered. However, the steam flashing in the lower plenum provided par-l tial core cooling. The system continued to depressurize. until the. CFTs I were actuated at 595 seconds at which - time the core refill' started. The core was rapdily recovered and q'uenched. by a mixture level at about 680 l seconds. This break size /RCP trip time combination resulted in a core un-covery/ recovery transient that did not achieve LPI for the period of interest. The heatup calculations were performed using the THETA code in the manner described. in section 5.2.1. Figure 5-16 shows the hot spot cladding volume i - 17.- l

average temperature response. The cladding temperature dropped from its initial value following the reactor trip and remained close to the satura-tion temperature until the pumps were tripped at 500 seconds. The cladding temperature started to rise at about 505 seconds during the RC pump coast-down. The steam production due to flashing in the lower plenum provided steam cooling of the core. However, the hot spot cladding temperature con-tinued to rise and reached a maximum of 1532F at 662 seconds and began a rapid downturn as the hot spot quenched. The core was covered by a mixture level at 680 seconds and continuous core cooling was established there-after. 5.2.5. Analysis Applicability to Raised Loop Design The significant parametric differences between the raised-loop design plant and the preceding generic lowered-loop analysis are in the high pressure in-jection (HPI) delivery rate and the amount of liquid volume which can effec-tively be used to cool the core. The liquid volume dif ferential is due to the basic design di fference; raised versus lowered loops. Because of the raised-loop design, system water available af ter the RC pumps trip will drain into the reactor vessel. j For the lowered loop designs, the available water is split between the reac-tar vessel and the pump suction piping. Thus, for the same average system void fraction, the collapsed core liquid level following a RC pump trip is j higher _ for the raised loop design than for the lowered loop design. Figure 5-17 shows a comparison of HPI system capacities for the raised-loop design plant and the lowered-loop plants. As shown, the HPI pumps utilized in the raised loop design 'will deliver.more flow versus the RCS pressure below 1300 psia. The RCS pressure generally falls below this pressure when the RCS evolves to a high void fraction that may result in an extended pe-riod of core uncovering. When the RC pumps are tripped under such condi-tion, the HPI pumps will deliver more flow into the system and shorten the core recovery time. Therefore, the results and conclusions provided in sections 5.2.2, 5.2.3, 5.2.4, 5.2.5, and 5.2.6 are applicable to the raised-loop plant design. t I,

5.2.6. Conclusions The best estimate analysis of RC pump trip following a SBLOCA was perfonned with realistic assumptions described in section 5.2.1. The F0AM2 and THETA codes were used to calculate steaming rates, core mixture level, and clad-ding temperatures during the most limiting transient. The results of this analysis, described in previous sections, can be summarized as follows: 1. Following a SBLOCA, if the RC pumps remain operative, core cooling is assured regardless of system void evolution. However, continuous RC pump operation in a highly voided system is not desiraole for pump integrity reasons. 2. Prompt tripping of the RC pumps upon receipt of indication of loss of subcooling margin will maintain the PCT well below the ifmits of 10 CFR 50.46. 3. Based on the results of the analysis under realistic assumptions, an RC pump trip at any time following a SBLOCA for break sizes 0.05 and smaller will not result in PCT exceeding the 10 CFR 50.46 limit. 4. For breaks 0.2-f t2 and smaller, more than 10 minutes time is available for manual RC pump trip following indication of a loss of subcooling margin. 5. Small breaks larger than 0.2-ft2 cause a rapid depressurization of the RCS and early actuation of the CFTS and LPI system. Therefore, a de-layed RC. pump trip for break sizes larger than 0.2-ft2 will not result in PCT exceeding the 10 CFR 50.46 limit. 6. In summary, as a result of-a realistic analysis, at least 10 minutes l time is available for a manual RC pump trip - following a small break LOCA without exceeding the limits of 10 CFR 50.46. This conclusion is applicable to all.B&W 177-FA lowered-loop and raised-loop plants. 5.3. Setpoint and Signal Selection 'for Non-LOCA Events The RCP trip criteria, based on loss of subcooling margin, was developed with the intent of assuring that an indication for RC pump trip would occur

~ _. - = - k for those SBLOCAs where pump trip was required to meet the criteria of 10 1 CFR 50.46. The spectrum of SBLOCA analyses discussed in the previous sec-tions demonstrate that a loss of subcooling will always occur for small breaks that have the potential to uncover the core if the RCPs are tripped f under certain two-phase conditions. The actual value of the setpoint is de-termined on a plant-specific Dasis to ensure that this indication will a1 4 low continued forced RCS flow during realistic SGTRs up to and including the design basis SGTR -- a single double-ended rupture. The setpoint also includes consideration for minimizing the indication for need to trip RC pumps for more likely non-LOCA events such as a mild overcooling transient due to excessive steam or feedwater flow. While optimum selection of the loss of subcooling setpoint is expected to discriminate against the more likel'y overcooling, severe overcooling events such as a large steam line break can result in a loss of subcooling and an indication for manual tripping - of the - RC pumps. The FSAR safety analyses 1 aee performed assuming the most limiting case of. pumps on or pumps off, i therefore tripping of the pumps will not result in consequences more severe l than previously analyzed. A best estimate analysis of the steam generator tube rupture event has-been performed to demonstrate that steam generator tube leaks up to a single i double-ended rupture will not result in sufficient loss of subcooling if ATOG procedures.are followed, to produce an indication for the need to trip l the RC pumps. 5.3.1. Best Estimate Analysis of a Single Double-Ended Steam Generator Tube Rupture ). l. A single double-ended steam generator tube rupture (SGTR).was simulated for a generic.177-FA lowered-loop plant design. The analysis is intended to l demonstrate that sufficient.1oss of subcooling margin will.not occur result-ing in a need to trip the RC pumps if an operator follows the ATOG for-the event. The results of: this-analysis.will support the. bases for utilizing manual. tripping of the _ RC pumps on the criteria of loss of.subcooling margin. - 20 3

l 5.3.1.1 SGTR General Characteristics t A SGTR is a loss-of-coolant accident (LOCA) that allows reactor coolant to leak into the secondary side of the steam generator where it is released into the steam plant and can lead to significant offsite doses if this i steam is released to the environment. For a complete severance of one SG l tube, a leak rate of approximately 400 gpm at normal system pressure and temperature would be expected. The leak from a failed tube cannot be isolated and reactor coolant will con-tinue to be lost until the plant is completely cooled and depressurized and the primary loops have been drained. l Since a tube rupture can exhibit the same general characteristics as a small break LOCA, the general procedures for LOCA mitigation must be fol-2 lowed. A continuous cooldown and depressurization of the RCS is essential to avoid opening of the SG safety valves minimizing the risk of releasing radiation. Forced circulation by the RC pumps will provide a continuous uninterrupted cooldown and depressurization of the RCS. The following best estimate analysis of a single double-ended SGTR for a generic 177-FA lowered-loop plant was performed to demonstrate that opera-tor actions as per AT0G to control RCS inventory, perform plant runback, and initiate -low power reactor trip, preclude loss of subcooling margin and RCP trip during a single double-ended SGTR event. 4 5.3.1.2 Method of Analysis (Best Estimate) The analysis was performed with the TRAP 210 computer code. A description of the TRAP 2 model simulating the RCS at:d the steam generator (SG) secon-dary _ system is shown in Figures 5-18 and 5-19 and Tables 5-2 and 5-3. The model has been developed to predict the system behavior during a SGTR event and simulate important operator actions as described. in AT0G. In addition, i the model utilizes a non-equilibrium _ pressurizer model capable of p; edict-ing two-phase pressurizer inventory and mixture level. 4 . Operator actions leading to a cold shutdown, systems initial conditions, and operation and other input assumptions are described in the following i subsection. P a.

5.3.1.2.1. General Operator Actions Although the ATOG guidelines may differ slightly among 177-FA lowered-loop plants, the most important operator actions and stages in treatment of the SGTR are similar. The operator actions for a SGTR event are summarized in Figure 5-16. ) Event Identification The first stage for mitigation of SGTR is prompt recognition of the event and determination of the affected steam generator. The occuirence of secon-dary radiation alarms (steamline monitor or condenser air ejector) almost simultaneous with SGTR and RCS pressure and pressurizer level drops are unique indicators that a SGTR has occurred. This analysis assumed the op-erator diagnoses the SGTR following the radiation alarm and pressurizer low level alarm and begins to take prescribed action. Plant Control at Power In the second stage of the event, the RCS pressure and pressurizer level must be stabilized so that the plant may be run back without tripping. A trip at high core power may result in venting radioactive steam to the en-vironment through the secondary safety valves. Normally, the makeup system (MJ) will automatically increase MU flow to stabilize pressurizer level. For a double-ended rupture (DER) of a single tube the leak flow ( 400 gpm) 13 greater than the MU flow. The operator is instructed to take action to increase makeup flow or HPI and terminate letdown in order to stabilize the RCS. Once RCS pressure and pressurizer level are stable, plant runback to hot zero power (HZP) should begin. This analysis assumes that the operator starts a second makeup pump and iso-lates letdown in order to stabilize RCS pressure. If the output of two makeup pumps does not match the leak flow, the operator is assumed to shift suction to the BWST and initiate full HPI flow. Analysis 8 has shown that for a DER of a single tube, the operator has approximately 11 minutes to stabiliae RCS pressure before the reactor trips automatically. Once the system is stabilized, the operator is assumed to runback the power. _

=- Plant Runback to Hot Zero Power Operator action should be initiated to stabilize RCS pressure and pressuri-zer level while conducting a plant runback to low power without tripping. RCS inventory st'uld be monitored during the plant runback. This analysis assumed operator action to reduce the MFW demand to match a 10% per minute runback in core power level. Upon reaching a low power level where the available turbine bypass (TB) ca-pacity is sufficient to avoid lif ting the steam safeties, the plant is tripped. This action was assumed at 15% power. Although the turbine by-pass capacity may vary from plant to plant, reactor trip near these power levels will exhibit a negligible difference in plant response. Cooldown and Depressurization The initial objective of the cooldown is to bring the RCS hot leg tempera-ture to a value ( 540F) that corresponds to a saturation pressure which is below the steam safety valve setpoint. This action will limit atmosphere radiation releases. Below 540F the SG with the tube rupture can be iso-lated. The cooldown should be continued to cold shutdown conditions at a i rate of 100F/ hour using the good generator and DHRS while maintaining the RCS subcooled. 5.3.1.2.2. Assumptions and Initial Conditions The following assumptions and initial conditions were used in this analy-sis. Core power is 2568 MWt. (The transient is not expected to be sensitive to initial power.) Offsite power is available throughout the transient. Reactor coolant pumps operate throughout the analysis. Initial pressurizer level is 180 inches -(indicated).. This is conserva-tive with respect to B&W 177-FA plants because this corresponds to an ini-3 .tial liquid volume of 820 ft. No B&W 177-FA plants operate with a lower nominal pressurizer level. l ! l l .~.

  • HPI and makeup flow characteristics are as shown in Table 5-1.

The re-sults are not sensitive to flow capacity as long as the flow from two HPI pumps can match the leak flow rate at rated power. TF primary to secondary leak flow is conservatively modeled as subcooled c' ; charge with a discharge coefficient of 1.0. .itial plant condition -- Power level 2568 MWt Hot leg temp 605F Hot leg presrure 2160.2 psia 582.7F Tavg RC system flow 36,597 lbm/s Pressurizer level (as measured from the lower tap) 180 inches Total steam flow 3258 lbm/s Subcooling margin 42F 5.3.1.3. Results of Analysis The primary-to-secondary leak flow rate which resulted from the double-ended rupture was approximately 50 lbm/s (see Figure 5-21). Operator ac-tion was modeled to manually start a second makeup pump and isolate letdown 1 minute after indication of rupture. The charging flow of two makeup pumps was not sufficient to match the leak flow. Subsequently, pressurizer level continued to decrease and HPI injection was initiated manually at t=4 min. to stabilize RCS inventory. This is shown in Figure 5-28. A runback in reactor power of 10% per minute was initiated when the plant was considered stable (t=7 min., see Figure 5-22). Main feedwater was man-ually ramped to match core power (Figure 5-23). The ICS would normally ramp power and feedwater flow to maintain a constant Tavg, however, model-T ot decreased. T old to remain constant while ing difficulties caused h c This forced T vg to decrease causing the system to contract more during the a runback than would actually have occurred. However, this had little effect on the analysis since the total system contraction is from full power to zero power, and thus the pressurizer outsurge remains approximately the same.

9 When reactor power and steam load were within the turbine bypass capacity, 15%, the reactor and turbine were tripped. The turbine bypass system con-trolled secondary pressure while the loss of heat source on the primary side caused the RCS to contract. The RCS contraction caused a pressurizer outsurge and the pressurizer level (Figure 5-24) decreased. The pressur-izer level continued to decrease until the energy removed by the turbine bypass system was equal to the energy generated in the core at which time pressurizer level began to increase due to high pressure injection. The minimum indicated post trip pressurizer level was 13 inches. Once pres-surizer level was increasing, the analysis was terminated because it was as-sumed the operator would cooldown and depressurize the RCS while maintain-ing minimum subcooled margin. The subcooled margin at termination of the analysis was 91F. Figures 5-21 through 5-30 show the transient responses of pertinent parameters. Plant Condition at Termination of Analysis --Table 5-2 lists the sequence of events for this analysis. Power level 68 MWt l Hot leg temp 545F l Hot leg pressure 2000 psia 544F l Tavg System flow 37,546 lbm/s Pressurizer level (indicated) 21 inches Bypass steam flow 53 lbm/s Subcooling margin 91F 5.3.1.4. Conclusions Since the unique indication of a SGTR lends itself to quick diagnosis, the operator has time to increase makeup or initiate HPI to stabilize the RCS inventory. If no action is taken to -increase makeup, the reactor will eventually trip on the variable low pressure trip. The pressurizer may sub-sequently. drain because of insufficient pressurizer. inventory to accommo-date the RCS contraction. -l ~

.. =. - Operator action, as per ATOG, to stabilize RCS pressure and pressurizer level, runback the plant, and trip the reactor does preclude sufficient i loss-of-subcooling margin to result in the indication for a need to trip the RC pumps. 5.4. Other Considerations i 5.4.1. Potential Containment Isolation of RC Pump Services A primary objective of the parameter and setpoint selection is the avoid-ance of reactor coolant pump trip for the more probable of the non-LOCA events. Loss of subcooling margin has been shown to occur for all small break LOCAs of concern and may also occur for design basis overcooling type transients which may exhibit a similar primary system response as a LOCA. Item I.1.e of the enclosure to Generic Letter No. 83-10 expresses the con-cern that " Transients and accidents which produce the same initial symptoms as a LOCA (i.e., depressurization of the reactor and actuation of engi-neered safety features) and result in the termination of systems essential for continued operation of the reactor coolant pumps (i.e., component cool-ing water and/or seal injection water)." ...In particular, if a facility design terminates water services essential for RCP operation, then it should be assured that these water services can be rastored in a timely man-ner once a non-LOCA situation is confimed, and prevent seal damage or fail-ure." The following contains some of the generic guidance provided -in the Abnor-mal Transient Operating Guidelines (ATOG) to assure water services when RCP operation is desired. The potential for containment isolation signals resulting from various transient types is discussed as well as guidance for the restart of RC pumps under degraded service conditions. ~5.4.1.1. ATOG Guidance ATOG emphasizes protection of the. RCP seals and motor to prevent damage. It is desirable to trip the pumps to prevent mechanical damage in case they must be -restarted at a later time. Preserving the pumps for long-term,

~ _ _ _ _ _. _ _ _ i cooling or cooldown is desirable, and it is recommended that they be shut down if high vibration or loss of auxiliary cooling water services occurs. (See a discussion of cooling water services and the effect on pump opera-tion under Containment Isolation Signal Guidance below.) Limits on con-tinued pump operation are given in " Plant Limits and Precautions." The rules for RCP trip to prevent mechanical damage are applicable in all cases except (1) when the pumps were not tripped immediately (i.e., within two minutes after the subcooling margin was lost) or (2) when severe inadequate core cooling (ICC) exists. In these cases the operator should try to re-store the RCP service which is lost. Operators are instructed to perfom the remedial action of verifying or re-establishing RCP seal injection and service water flow if potential i sol a-tion signals have been verified (i.e., ESF signals). The sections of AT0G I which permit continued RCP operation or anticipate RCP restart give instruc-l tions for the verification or reestablishment of seal injection and cooling water. AT0G provides specific instructions for RC pump restart. The Equipment Op-eration chapter of ATOG shows the conditions under which the RC pumps can be _ restarted. For situations where inadequate core cooling is not a con-cern, confirmation that no RC pump-damage will occur is among the require-i ments for pump restart. (If inadequate core cooling is a concern, RC pump restart is required even if damage can occur). Some sper.ific actions which must be taken to verify pump integrity before restart include: I ' ~If component cooling water is isolated and pumps are allowed to run, bear-ings and-stator will rise -in temperature. If pumps are then tripped, l they should not be restarted until component cooling water is restored and bearings have cooled down to normal operating temperature. i If pumps are tripped before bearings: get hot (exceed alam setpoint), pumps can be restarted and run once component cooling water is. restored. i ' Unless bypassed, interlocks will prevent pump restart until component l cooling water and seal injection are available to the pump. ll Controlled bleed-off flow should.be reestablished prior to pump restart. ~ l - t- . _. - ~ -.. _.,, 4 m,.....---.

Other considerations for RCP restart are verified on a piant specific basis. The Plant Limits and Precautions Documents, Plant Setpoints Docu-ments, and RCP vendor instruction manuals provide further instructions. 5.4.1.2. Generic Containment Isciation Signal Philosophy In considering all B&W Owner's Group plants generically, it must be recog-nized that some differences exist. For example, (1) the plants utilize three different RCP vendors (Byron-Jackson, Bingham, and Westinghouse), (2) the plants utilize various motor suppliers (Siemans-Alli s, GE, Westinghouse, and AEG), (3) the 205-FA plants operate with a lesser subcool-ing margin than the 177-FA plants, (4) system configuration and design dif-fer slightly and, (5) as a result of recommendations for increased RCP re-liability and response to NUREG-0737, Item II.E.4.2, each utility has opted for various isolation signals which affect RCP cooling water services avail-abili ty. The containment isolation signals presently being used, depending on plant and function, are: Low RC pressure (1500-1700 psig). Low RC pressure or high containment pressure (3-4 psig). Low-low RC pressure (400 psig). ESF interlocked with RCPs not running. High-high containment pressure (25-38 psig). ESF with seal bleedoff redirected to quench tank. High radiation. Low CCW surge tank level. Non-LOCA transients such as a severe overcooling resul ting from a steam line break inside containment have the potential to produce the same ini-tial symptoms as a LOCA. If it were desired to continue with RCP operation for these transients, low RC pressure (1500-1700 psig) and high containment pressure (3-4 psig) actuation have the greater potential to cause unwanted isolation of RCP services and require timely reestablishment of these ser-vices. Using high-high containment pressure as an isolation signal would prevent. isolation of the service functions for the majority of such tran-sients and accidents and yet maintain isolation capability for design basis 1 _ _ - - _ - _ _

accident (large LOCA and SLB). The other isolation signals mentioned above are used to preclude RCP or equipment damage. The cooling water services supporting the RCP with the potential of being isolated are: l Seal injection. Seal bleedoff (resulting from seal injection provided to pump). Component cooling water to seal area coolers.

  • Component cooling water to RCP motors and oil coolers.

For a running RCP, seal integrity can be maintained with either seal injec-tion or component cooling, within operating parameters, to the seal area coolers. More than ten years ago, B&W removed the seal injection isolation signal from the makeup and purification system design and recommended that the signal be removed from the operating plants. Presently there is only one plant with seal injection isolation which is signaled to close on 400 psig RC pressure. Four hundred psig RC pressure is an indication of a large break LOCA where RCP operation is not anticipated. The seal bleedoff is important to seal integrity in that it provides pres-sure staging across the seals and removal of the heat generated by the ro-tating seals. Closure of the seal bleedoff ifne with the RC pumps running can cause significant seal damage. In this situation with a low seal face leakage rate, the heat generation rate at the upper seal will triple due to tripling of the seal P and there is not enough flow to remove this heat. AT0G instructions call for quickly reestablishing seal bleedoff flow or tripping the pumps. Some plants have initiated a change to redirect the seal bleedoff flow to the quench tank upon receipt of a containment isola-tion signal. Another alternative to minimize the occurrence of ESF closure is to actuate close only on hi-hi containment pressure (the same signal that actuates containment spray fl ow). For plants utilizing ESF signals (1500-1700 psig RC pressure, 3-4 psig containment pressure) for isolation, timely reestablishnent of seal bleedoff flow per the AT0G -instructions is required for continued RC pump operation. With proper seal injection and seal return, integrity of the seals can be maintained indefinitely on loss of component cooling water to a running RC pump. However, the RCP motors cannot be run indefinitely on a loss of cool-ing water. If it is desired to continue RCP operation, cooling water must be reestablished within a given time frame as indicated in AT0G to preclude damage. The time frame varies depending on the type of motor. Table 5-1. HPI and Makeup System Capacities 1. HPI Flow Vs Pressure Pressure, psig

gjgg, gpm 0.0 300 600 900 1200 1500 1800 2100 2300 1 HPI 550 525 500 470 440 405 365 320 280 2 HPI 1070 1025 970 915 850 780 705 615 535 2.

Makeup (MU) Flow Vs Pressure (One Pump) Pressure, psig 0 1000 1600 2100 2205 2265 2300 3000 Fl ow, gpm 317 250 200 150 53 53 32 32 i j

Table 5-2. MINITRAP2 Node Description Node number 1 Reactor vessel, lower plenum 2 Reactor vessel, core 3 Reactor vessel, upper plenum 4,10 Hot leg piping (including " candy cane") 32,33 " Candy cane" and upper SG shroud 5-7,11-13 Primary, steam generator tube region 8,14 Cold leg piping 9 Reactor vessel downcomer 80 Pressurizer 16,24 Steam generator downcomer 17,25 Steam generator lower plenum 18-20,26-28 Secondary, steam generator tube region 21,29 Steam risers 22,30 Main steam piping 23 Turbine 31 Containment 34 OTSG Tube 1 l f 1 l 1 l u.-': .i..;D.'-.e i. - : :, ?. - : : ~ & D ' -i. '. T> c./ ur: - '. +.: A~c : -.;. pgin.;; ; , i:l;i . -s . s. a A ': - n. %.. s ~. ' - y .7... >, NlS'$.

e
  • Table 5-3.

MINITRAP2 Path Description l'f;- 1 ". . ~, - ., ~. A, .Y Path number Description Y*f 74 v;. '. ? 1 Core j.t. i.

l' 2

Core bypass 1;.(, J b7--N 3 Upper plenum, reactor vessel .p.. 4,11 Hot leg piping ij.)

u. - A

[ 5,12 Upper steam generator shroud U:j. 45,46,47,48 Top of hot leg " candy cane" c.. 6,7,13,14 Primary heat transfer region, SG ( k*,. 8,15 RC pumps }.' '. '.. q. h.4 9,16 Cold leg piping 7.. E '

r. 9,, 8. ~ -

0 10 Downcomer, reactor vessel p.t.lp.? Q 17 Pressurizer surge line

kg/.

18,19,26,27 Steam generator downcomer and plenum 3 Q5. - 5 +. ' l !'[ ', } 20,21,28,29 Secondary heat transfer region, SC '. f;. ~.-e 22,30 Aspirator

  1. .7.l<

23,31 Steam riser, steam generator

.x*

? - .v e y 24,32 Main steam piping ?2{j.;. - Q: 25,33 Turbine piping Q e;- a.. c. - 34,35 Break (or leak) path p(x.,/ M I.. N' -s 36,37 HPI u. q,[$(s.: f.i 38,39,43,44 AFW 1-

p 40,41 Main feed pumps
f..rY ?

lE 42 LPI f T. fe 49 MSL crossover

4. 9 H,

50,51 Turbine bypass valves $, w--{. 'l 52 Pressurizer spray .i.(j " ~ .. g 53 Letdown W

A -

.9 54 Makeup IJpper OTSG tube (steady state only) ..[') 55 2! 56 Lower OTSG tube S M* 1V,

y..':

57 Lower end of ruptured 0TSG tube D...:. - F. i 58 Upper end of ruptured 0TSG tube '.M. ;. I. Q1 Pressurizer heaters y. ..;(; Q2 RC pump heat

, \\,,

.e T -' .:l "4 32 - Sg m x 2 -.,. A' q.;.

. w * :

'n '4 .I '. ' ' ' } gi;. i '; ; 4; _ _ _ s, 437.' (. * ^ __. ;. 5.,.,- ' ' l,' :- 3 ; _;,, _.i L ;..;.; ;,~. :. ; ;. - .;_.,._._. y ^;;. i4 w ;,..a.sf-.,.:.,. .u

i Table 5-4. SGTR Sequence of Events Time after rupture, Event min:sec l SGTR at full power 0:00 1 MSL radiation monitors and/or condenser air ejector 0:00 radiation monitors trip high secondary activity alarms Low pressurizer level alarm 0.20 Operator starts second ma'keup pump; letdown isolated 1:00 Operator switches pumps to HPI mode and begins HPI 4:00 Reactor runback begins 7:00 , Reactor and turbine trip 15:30 Minimum' indicated pressurizer level of 13 inches 16:35 Analysis terminated ' 17:40 l

Figure 5-1. CRAFT 2 Noding Diagram for Small Breaks (Six-Node Model) CFT l O h 4 2 6 5 3 1 g 1 -O o nO vc ^ 4!F v n i ( v n n WU v Node No. Identification Path No. Identification 1 PD piping, DC, LP 1 Core 2 Primary SG 2 LPI 3 Core, UP, hot legs 3,10,11 HPI 4 Pressurizer 4 Hot legs 5 Containment 5 Pumps 6 Secondary SG 6 Vent valve 7 Pzr surge line 8,9 Leak & return path 3 -

Figure 5-2. Critical Region for RC Pumps Trip, Break Size Vs Time After Trip i l D l \\' 5 rs ~(_ \\m 0.2 ,g G Critical Note: Time T=0 is reactor j Region trip time. ~~~ -~~ A u os 0 100 1000 10000 l Time After Trip, s i i l i r 5 1

Figure 5-3. " Realistic" Core Axial Peaking Distribution 1.4 1.2 e' g 1.0 T< 0.8 3 0.6 Eo 0.4 z 0.2 i i i e i i I 0 20 40 60 80 100 120 140 Core Elevation, inches b 'S -f---

i Figure 5-4. Average Primary System Void Fraction Excluding Pressurizer l 80 / 1 l

  • t /

g / E 60 / 1 k l,' s // a 40 - /,/ / 2 0.1-ft 7 2 0.2-ft 7f 2 20 - --- 0. 25-f t 2 -X 0.3-ft I I I I I I 200 400 600 800 1000 1200 Time After Break, s I i l.

Figure 5-5. RC System Pressure Vs Time (0.1-ft2 CLD Break -- RCP Trip at 1000 s) 2000 - 2000 - 'G" 1600 - J ii as E 1200 - S U RCP N Trip g 800 - 400 - 4 I I I I I I l i I I I I 200 400 600 800 1000 1200 Time After Break, s

Figure 5-6. Liquid Volume in Reactor Vessel for 0.1-ft2 CLD Break -- RCP Trip @ 1000 s t l 1800 l Volume 0 9 ft core ele. "a 1600 J 5 o u 1400 '53 s QJ ".5 1200 V 1000 Pump Trip 800,- l t,, i i i ~ 'r,00 900 1000 1100 1200 1300 Time After Break, s i 1 l ~ Figure 5-7. RC System Pressure Vs Time (0.2-ft2 CLD Break -- RCP Trip 9 400 s) 2000 Ea J 1500 $i E 1000 RCP Trip v7 a 500 0 0 100 200 300 400 500 600 700 3 Time After Break, s

Figure 5-8. Liquid Volume in Reactor Vessel for 0.2-ft2 CLD Break -- RCP Trip 9 400 s l 2000 l 1800 mg Volume 0 9 ft core elev. J 1600 5 'E u 's 1400 3 a 3 1200 s 1000 Pump Trip i 800 l I s-l T r a 200 400 600 800 1000 Time After Break, s i t I. - -

C Figure 5-9. RC System Pressure Vs Time (0.25-ft2 CLD Break -- RCP Trip @ 300 s) 2000 Ea. J ~ Ii 1500 E a. e3 1000 g RCP Trip E 500 0 0 100 200 300 400 500 600 i Time After Break, s % s- .1

Figure 5-10. Liquid Volume in Reactor Vessel for 0.25-ft2 CLD Break -- RCP Trip @ 300 s 2200 l 2000 3 1800 Volume 0 9 ft core E elevation ,= o 1600 5 M 1400 1200 l 1000 Pump Trip l l T i i i i i 100 200 300 400 500 Time After Break, s -

Figure 5-11. RC System Pressure Vs Time (0.3-ft2 CLD Break -- RCP Trip 9 300 s) 1 E 1500 u, [ 1000 gRCP Trip Pu a E 500-- U 0 I I I I I 0 100 200 300 400 500 600 Time After Break, s . (

Figure 5-12. Liquid Volume in Reactor Vessel for 0.3-ft2 CLD Break -- RCP Trip @ 300 s 1800 Volume 0 9 ft core elev. f 1600 i 1400 J 5 's u 1200 ~53 a QJ 5 1000 A 'm E 200 600 \\ ump Trip P i i 400 - l 7 i e i i 100 200 300 400 500 Time After Break, s . l

Figure 5-13. SBLOCA Spectrum System Pressure Vs Time 2000

0. 2-f t2 0

-- -. -0. 25-f t2 m 0.3-ft2 I 1500 r e \\ y Pump Trip i 1000 s g s~_'s,~~_ .g Pump Trip CFT Setpoint m s g 500 (%'N l %~ LPI Setpoint N l 0 i i i i e i i e i i i i i i e i q,100 200 300 400 500 600 700 800 -l Time After Break, s 4 4 I

Figure 5-14. Liquid Volume in Reactor Vessel Following RC Pump Trip for a 0.2-ft2 PD Break 2400 i ~r~~,,' // I Top of core l 2000 12 d volume ./ i t l-l f - 10 ~ / / d d J 1600 / / - 8; I r- ~ ~.-. .J.) e .c ,= ~ j l ~ 6.? e E CFT y g 1200 l / 4g e i u I l 2 2 / Bottom of c, ore 0 m 800 volume ,f CFT Trip at: I --- - - 300 s 400 - - - 400 s j CFT 600 s -- ---50 0 s ) g-0 300 400 500 600 700 800 900 Time After Break, s

2 Figure 5-15. System Pressure Vs Time for a 0.2-ft Break at the Pump Discharge 3000 2500 2 E. 2000 J h, v 0 1500 I u RCP Trip 5 U 1000 l5! CFT Setpoint 500 LPI Setpoint i I I I I I i 100 200 300 400 500 600 700 800 Time After Break, s e 1 48 -

Figure 5-16. Maximum Hot Spot Cladding Volume-Average Temperature Vs Time for a 0.2-ft2 Break at Pump Discharge 1775 u J 5 by 1250 3 i t E l E ,2 725 o Y c. C j h 2 RC Pump Trip u i 200 I I I I a 100 200 300 400 500 600 700 800 t ~ l l l l _.

Figure 5-17. Comparison of High Pressure Injection System Capacities for Raised Loop and Lowered Loop Design Plants 1200 k Raised Loop \\ 2 HPI \\ 1000 g s \\ s s o' 800 \\ % \\ 3 \\ Lowered Loop 2 HPI I 600 Raised Loop \\ \\ b ? 1 HPI \\ k \\j 400 \\ \\\\ Lowered Loop \\ 200 1 HPI I V li i i i i i i 200 600 1000 1400 1800 2200 2600 RCS Pressure, psia l l t 1 1 __

Figure 5-18. Mini-TRAP 2 Noding and Flow Path Schematic [s] y g -- c 23 O MSIO TBS TBS e 5-p .___@_______jsi l_. _ _ _ _ _ 9_ _ _ _ _. 22 l 3o L ~ ~ _I O Mssy 33 32 g lL EMov-a g "h 01 % g 80 o Ms!V 29 r-28 11 O g-g 09 g-- @[ jg @( 3 n 7 F

1. -

27 12 9 O ts 24 Je e( 8C p 3 CFT 2 = S 2s is d 3, g @l D 12 ='~ 2, OT e e I@~ ~ 2 14 8 ig tL@ g" O i l l e E-g2 l h.

Figure 5-19. Detailed Mini-TRAP 2 Noding/Flowpath Diagram for SGTR Model t 5 20 34 6 19 7 18 \\ J J = Node Description 5,6,7 Primary OTSG Tube Region 18,19,20 Secondary OTSG Tube Region 34 OTSG Tube Flowpath Description .55 OTSG Tube (steady-state only) '56 OTSG Tube 57 Lower End of Ruptured 0TSG Tube 58 Upper End of Ruptured 0TSG Tube

Figure 5-20. Steam Generator Tube Leaks -- Operator Action Outline EvtNT '0ENT!FICAflCN SGTR $rePrce! AFFICTED OfSG I - A00lfl0NAL FAILURE! r $PECIAL CA$ts sifN FAILURES CCRRECT IF PO$$1BLE if PLANT CONTROL Af PostR - INCREA$( uu OR NPI $fAllLIZE RC PRES $URE & PRES $URilfR LEVEL if PLANT RUNSACA TO NZP

  • PREVENT LIFilNG as!v'$

OR if REACTOR TRIP REACfoll TRIP if if l FOR A L ARGE TUIE LIAE. OR FOR A $31ALL TUS( LIAE WifM ' $NALL LEAR Of fMOUT C0leEN. RC PuuPS RUNNING ANO TN( $Eg ' R APIO CC.60084 ANS 9 CONCEN!ER OPERAflNG:

  • RAPIO CJ0L00m ANO O(PR($$URilAfl0N DEPRES$UtilAfl04 TO PROCEID TO COLO $NUf008N

- $00F ANO 1000 P$la fifM A NOReAL RATE OF 4AINTAIN ACS $USC00 LING C00LOCe4 . gfPA$$ ($A$ ACTUAfl0N ITPA$$ $LIIC if i!0 tait $TEin ANO FIE0uNES y 0F AFFECT!D GENERATOR. svPa$$ $LalC (RE!f0RAfiON !!01 Aff FEEDWATER. FEED g OF FEEDeATER ANO PERl00lc GEN (Raf04 A$ NE[DtB 70: FLIO9Af(R A40 $ffAE $fEAulNG OF AFFECito $G . salNTAIN Of!G LEVEL l!DLAfl0N 13 $PICIFICALLY TO THE CON 0EN$tR uAt et REQUIRED REQUIRiol. NAfuRAL CIRCULAfl04 RC$ COOLING Raft REPAIR FilLURE As PO!!IBLE If if RC$ C00L0084 70 COLD RC$ CDCt00tPI 70 COLO $NUf00tNs $NUf00m af 100F *R

  1. COE I 2 04 3 900t ! QR 2 x________

l l Figure 5-21. Leak Flow Vs Time After Rupture 60 s h 50 L B 2 3c 40 .3 I I I I I i i i 100 200 300 400 500 600 700 800 900 1000 Time After Rupture, s ] i.

Figure 5-22. Total Reactor Power Vs Time After Rupture of a Single Double-Ended SG Tube i \\ 1.0 V 5 0.8 E c. H 0.6 lii 0.4 z 0.2 f I 0.0 S 0 200 400 600 800 1000 Time After Rupture, s o - E5 - i A

i i

  • cs_ 5 U i G $ 5, e 3.

E 0 0 0 0 0 0 0 0 0 0 1 0 8 7 6 2 2 1 1 1 ~ 0 I 0 1 1 wo / l e F )s ru r 0 s e 0 t 0 e a 1 5 r w r

  • \\

P d e 1 e w o \\. g e o N ,\\ t

  • {

e F P V L k n e 0 s N I 0 c t i r a o a o 9 b H M C s e n r u .\\ u R tp r y u e l 0 R .N w I l 0 o a 8 r P ~ u e n t \\ a f s \\ m A 3 2 I 0 m r e e 5 t 0 i a 7 T \\ e w r d u 7 g e g e f. i d F d N d e 0 nl 0 I al 6 o N s rr et \\ wn /. oo \\ P c 0 N b 1 0 5 s' e to N 0 0 0 0 0 0 0 8 6 4 2 1 l-OC 5%4[ {* E* bI g, e [ i lll'

Figure 5-24. Pressurizer Inventory Vs Time After Rupture A 938 ) [ 750 o 3 8 563 - 3 5 375 - E n. 188 - I f f i t 200 400 600 800 1000 Time After Rupture, s 57 -

I i Figure 5-25. Hot and Cold Leg Temperature Vs Time After Rupture k 600 m Hot Leg ? B 550 > ~ N Co1T1 Leg "u 500 4b0 6b0 8b0 0 1000 200 Time After Rupture, s Figure 5-26. Core Outlet Pressure Vs Time After Rupture E 2200 t y 2 0 E 1900 "vs8 8 1600 0 I I I l l 0 200 400 600 800 1000 s Time After Rupture, s

s..

Figure 5-27. Surge Line Flow Vs Time After Rupture 200 - 100 - C Ji ~ ~ NW I L S -100 -200 I i -300 200 400 600 800 1000 Time After Rupture, s l 1. O

Figure 5-28. Total RCS Charging Flow Vs Time After Tube Rupture 80 i e 3 60 J b 40 C V 20 A I i i i 0 200 a00 600 800 1000 Time After Rupture, s ) \\ \\...... _. _ _..

Figure 5-29. SG Tube Flow Vs Time After Rupture Upper Tube End 37.0 ( l 36.0 t S g 35.0 s I E 8 C 33.6 c. I $s O { 3 26.3 EE \\ 21.8 m Lower Tube End 17.5 f 13.1 d - t

8. 7, I

i i i / 200 400 600 800 1000 i Time After Rupture, s

\\ Figure 5-30. Subcooling Margin Vs Time After Rupture 90 80 + u-70 E i Tr. 5-60 E' 50 8 ( v"3 40 6 30 ( 20 1 10 I I I I I I I I I 200 400 600 800 1000 / Time After Rupture, s f

REFERENCES i 1. D.G. Eisenhut to all applicants with Babcock & Wilcox (B&W) Designed i Nuclear Steam Supply Systems (NSSS's), Letter, " Resolution of TMI Ac-tion Item II.K.3.5 ' Automatic Trip of Reactor Coolant Pumps' (Generic b Letter No. 83-10e)," February 1983. 2. D.G. Eisenhut to all Licensees with Babcock & Wilcox (B&W) Designed j Nuclear Steam Supply Systems (NSSS's), Letter, " Resolution of TMI Ac-tion Item II.K.3.5, ' Automatic Trip of Reactor Coolant Pumps' (Generic Letter No. 83-10f)," February 1983. 3. R. B. Davis to ' B&W 177 Owners Group Technical Subcommittee on TMI-2 ) Incident Related Tasks, Letter, " Responses to IE Bulletin 79-061 Ac-tion Items" (Analysis Summary in Support of an Early RC Pump Trip), September 12, 1979. 4. G. E. Anderson, Jr., " Guidance for Post-LOCA Tripping of Reactor Cool- { ant Pumps," B&W Document 51-1132119-04, January 7,1983. 5. J. J. Cudlin, et al., CRAFT 2-FORTRAN Program for Digital Simulation of { a Multinode Reactor Plant During Loss of Coolant, BAW-10092, Rev. 3, Babcock & Wilcox, Lynchburg, Virginia, October 1982. 6. N. K. Savant et al., B&W's Small Break LOCA ECCS Evaluation Model, Ap- \\ pendix B, BAW-10154, Babcock & Wilcox, Lynchburg, Virginia, November 1982. 7. FOAM 2 --- Computer Program to Calculate Core Swell Level and Mass Flow Rate During Small Break LOCA, BAW-10064, Rev. 1, Babcock & Wilcox, Lynchburg, Virginia, June 1975. f 8. THETA 1-B -- Computer Code for Nuclear Reactor Core Thermal Analysis, BAW-10094, Rev.1 Babcock & Wilcox, Lynchburg, Virginia, April 1975. 1 9. TACO 2 -- Fuel Pin Performance Analysis, BAW-10041P, Babcock & Wilcox, Lynchburg, Virginia, August 1979. ( 10. TRAP 2 -- FORTRAN Program for Digital Simulation of the Transient Be-havior of the Once-Through Steam Generator and Associated Reactor Coolant System,

BAW10128, Babcock & Wilcox, Lynchburg,
Virginia, August 1976. 0 i

i .. _}}

Retrieved from "https://kanterella.com/w/index.php?title=ML20083L749&oldid=4364606"