ML19209C615

Analysis Summary in Support of Early Reactor Coolant Pump Trip

ML19209C615
Person / Time
Site:	Rancho Seco
Issue date:	08/27/1979
From:	SACRAMENTO MUNICIPAL UTILITY DISTRICT
To:
Shared Package
ML19209C613	List: ML19209C612 ML19209C615
References
NUDOCS 7910160487
Download: ML19209C615 (49)

Text

so r

A'iALYSIS SQaqRy I:; SUPPORT OF AN EARLY RC PU'4P TRIP 1163 094 79101004-8 7

c CONTENTS Page I.

TRODUCTION.

1 II.

SMALL BREAK ANALYSIS.

2 A.

Introduction.

2 B.

Systen Response '.Jith nc Purps Running 2

C.

Analysis Applicability to Davis-Besse 1 11 D.

Effect of Prompt RC Pump Trip on Low Pressure ESEAS Signal.

13 E.

Conclusions 13 6

III.

IMPACT ASSESSME::T OF A RC PUMP TRIP ON NO:T-LOCA EVENTS.

15 A.

Introduction.

15 B.

General Assessment of Punp Trip in Non~LOCA Events.

15 C.

Analysis Concerns and Results.

16 D.

Conclu and Sun =ary 18 e

116.,

na5 o.-

e

c

?.NALYSIS SUw ARY I" SUPPORT OF AN EARLY RC PUMP TRIP I.

INTRODUCTION B&W has evaluated the effect of a delayed RC pump trip during the course of small loss-of-coolant accidents and has found that an early trip of the RC pumps is required to show conformance to 10CFR50.46.

A summary of the LOCA analyses performed to date is provided in Section II.

This discussion includes:

1.

A description of the models utilized.

2.

Break spectrum results with continuous RC Pump Operation.

3.

Break spectrum results with delayed RC pump trips including estimates of peak cladding temperatures.

4.

Justification that a prompt pump trip following ESEAS actuation on low RC pressure provides LOCA mitigation.

An impact assessment of the required pump trip on non-LOCA events has also been completed and is presented in Section III. This evaluation supports the use of a pump trip fol]owing ESFAS actuation for LOCA nitigation since no detrimental consequences on non-LOCA events were identified, i163 396'.

o e

II.

SMALL BREAK ANALYSES A.

Introduction Previous small break analyses have been nerformed assuaing a loss-of-offsite power (reacter coolant pump coastdown) coincident with re-actor trip.

These analyses support the conclusion that an early RC pump trip for a LOCA is a safe condition. However, a concern has been identified regarding the consequences of a small break transient in which the RC pumps remain operative for some time period and then are lost by some means (operator action, loss-of-offsite power, equipment failure, etc.).

This section contains the results of a study to further understand how the small break LOCA transient evolves with the RC pumps operative.

Specifically, section B.

describes the system response with the RC pumps running for B&W's 177-FA lowered-loop plants.

In-cluded in this section is the development of the model used for the analysis, a break spectrum sensitivity study, and peak cladding tem-perature assessments for cases where the RC pumps trip at the worst time.

Section C demonstrates the applicability of the conclusions drawn in section 3 to a 177-EA raised-loop plant (Davis-Besse 1).

The effect of a prompt tripping of the RC pumps upon roccipt of a low pressure ESEAS signal is discussed in section D

Finally, sec-tion E suc=arizes the conclusions of this analysis.

B.

System Response With RC Pumos Running 1.

It.troduction Recent evaluations have been performed to examine the primary system response during small breaks with the RC pumps operative.

During the transient with the RC pumps available, the forced circulation of reactor coolant will maintain the core at or near the saturated fluid temperature. However, for a range of break sizes, the reactor coolant system (RCS) will evolve to high void fractions due to the slow system depressurization and the high liquid (low quality fluid) discharge through the break as a re-sult of the forced circulation.

In fact, the RCS void frc: tion will increase to a value in excess of 90% in the short term.

In

)\\(0 r

the long term, the system void fraction will decrease as the RCS RCS depressurizes, IIPI flow increases, and decay heat diminishes.

With the RCS at a high void fraction, if all RC pumps are postu-lated to trip, the forced circulation will no longer be available and the residual liquid would not be sufficient to keep the core covered..A cladding temperature excursion would ensue until core cooling is reestablished by the ECC systems. The following para-graphs summarizes the results of the analyses which were performed for the 177-FA lowered-loop plants, to develop the consequences of this transient.

2.

Method of Analysis The analysis method used for this evaluation is basically that de-scribed in section 5 of BAU-10304, Rev. 3, "B&W's ECCS Evaluation Model"I and the letter J.H. Taylor (B&W) to S. A. Varga (NRC), dated 2

July 18, 1978, which is applicable to the 177-FA lowered-loop

p. ints for power levels up to 2772 FMt.

The analysis uses the CRA2T23 code to develop the history of the RCS hydrodynamics.

However, the CRAFT 2 model used for this study is a modification of the small break evaluation model described in the above ref-Figure 2-1 shows the CRAFT 2 noding diagram for small erences.

breaks from the above referenced letter. The modified CRAFT 2 model consists of 4 nodes to simulate the primary side, 1 node for the secondary side of the steam generator, and 1 node representing the reactor building.

Figure 2-2 shows a schematic diagram of this model. Node 1 contains the cold leg pump discharge piping, downcomer, and lower plenum.

Node 2 is the primary side of the SG and the pucp suction piping.

Node 3 contains the core, upper ple-num, and the hot legs.

Node 4 is the pressurizer and nodes 3 and 6 represent the reactor building and the SG secondary side, re-spectively. This 6 node model is highly sinplified compared to those utilized in past ECCS analyses. It does, houever, maintain RCS volume and elevation relationships uhich are important to properly evaluate the system response during a small break with the RC pumps running.

kk 0

0 The breaks analyzed in this section are assumed to be located in the cold leg piping between the reactor coolant pump discharge and the reactor vessel.

Section B.7 demonstrates that this is the worst break location. Key assumptions which differ from those de-scribed in the July 18, 1978, letter are those concerning the equip-ment availability and phase separation.

These are discussed below, a.

Equipment Availability The analyses which were performed assumed that the RC pumps re-main operative after the reactor trips.

For select cases, after the system has evolved to high void fractions (approxi-mately 90%) the RC pumps were assumed to trip. Also, the im-pact of 1 versus 2 RPI systems for pump injection were examined.

The aajority of the analyses performed assumed 2 HPI pumps.

However, as is demonstrated later, even with 2 HPI pumps avail-able, cladding temperatures will exceed the criteria of 10 CFR 50.46 using Appendix K cvaluation techniques.

Therefore, fur-ther analysis with only 1 HPI pump would only be academic.

b.

Phase Separation The present ECCS evaluation model created to evaluate small breaks without RC pumps operative,(quiescent RCS) uti-4 lizes the Wilson bubble rise correlation for all primary sys-tem control volumes in the CRAFT evaluction.

In this analysis, for the time period that the RC pumps are operative, the pri-mary system coolant is assumed to be homogeneous, i.e., no phase separation in the system.

In reality, the flow rates in the core and hot legs are low enough that slip will occur.

This will cause an increased liquid inventory in the reactor vessel compared to that calculated with the homogeneous model.

With the homogeneous assumption, core fluid is continuously circulated throughout the primary system and a portion of that fluid is lost via the break.

During the later stages o f the transient, a slip model will result in fluid being trapped in

'the reactor vessel and the hot legs.

The only method of losing liquid during this period will be by boiling caused by the core decay heat.

Thus, the assumption of homogenicty for the period with the RC pumps operative is conservative.

I163 099,

o Following tripping of the RC pumps and the subsequent loss-of-forced circulation, the system will collapse and separate.

The residual liquid will then collect in the reactor vessel and the loop seal in the cold leg suetion piping.

For this period of the transient, the Wilson bubbic rise model is utilized.

The homogeneous assumption for the period with the RC pumps operating applies to nodes 1, 2, and 3 in the CRAFT model.

Node 4, the pressurizer, and node 6, the secondary side of the steam generators, utilize the Ullson bubble rise model throughout the transient as these nodes are not in the direct path of the forced circulation.

3.

Benchmarking of the 6 Node CRAFT Model Studies were performed to compare the results of the 6 node model to the more extensive evaluation model for B&W's 177-FA lowered-

~

loop plants as described in the letter J.H. Taylor (B&W) to S.A.

Varga (NRC), dated July 18, 1978. The break size selected for 2

this comparison is a 0.025 ft break at pump discharge.

This break represents the largest single-ended rupture of a high energy line (2-1/2 inch sch 160 pipe) on the operating plants.

The break can be viewed as " realistic" or the worst that would be ex-nected on a real plant. Figures 2-3 und 2-4 are the results of this comparison, System pressure and percent void fraction shown in Figures 2-3 and 2-4, respectively, compare very well with those from the more extensive (23 nodes) CRAFT 2 small break model. As seen in these figures, the difference is not significant and is less than a few percent.

The computer time for this 6 node model is, however, significantiv decreased. The model utilized for this study is thus justified based on comparison of results to the more extensive small break model and desirable because of its economical run time.

4.

Analysis Result _s The break sizes examined for this analysis ranged from 0.025 ft2 to 0.2 ft2 in area and are located in the pump discharge piping.

Breaks of this size do not result in a rapid system depressuri-zation and rely predominantly upon the HPIs for mitigati f} }

n

a Table 2-1 summarizes the analyses performed for this evaluation.

The majority of the analyses performed utilized 2 HPI pumps through-out the transient.

The effect of utilizing 1 HPI pump is discussed in this section.

Figures 2-5 and 2-6 show the system pressure and average system void fraction transients for the break spectrum analyzed assuming continuous RC pump operation and 2 HPI's available.

In Figure 2-6, the average system void fraction is defined as V; - V2 Average system void, % =

x 100 y

1 V = total primary liquid volume excluding the pressuri-3 zer at time = 0, V

=t tal pri ary liquid volume excluding the pressuri-2 zer at time = t.

This parameter was utilized in place of the mixture height in that the coolant uill tend to be homogeneously mixed with the RC pumps operative. Under these assumptions, the core is cooled by forced circulation of tuo-phase fluid and not by pool boiling as in the case where the RC pumps are not running and separation of steam and water occurs. As shown in Figure 2-5, the system pressure re-sponse is basically independent of treak size during the first several hundred seconds into the transient.

This occurs because the forced circulation of reactor coolant maintains adequate heat transfer in the steam generators; the primary system thus depres-surizes to a pressure (about 1100 psia) corresponding to the sec-cndary control pressure (i.e., set pressure of SG safety relief valves). After some time (250 seconds for the 0.1 ft2 break), the system pressure will decrease as the break alone relieves the core energy.

Figure 2-6 shows the evolution of the system void fraction; values in excess of 90% are predicted very early (300 seconds) into the transient.

For the larger breaks the system high void fractions occur early in time.

For the smaller breaks it takes in the order of hours before the system evolves to high void fraction.

Core cooling is maintained during a small break with continuous RC pump h\\

_g_

f operation regardless of void fraction.

In the long term, the sys-tem will depressurize and the enhanced performance of the ECCS (HPI and LPI) will result in reduced system void fraction.

2 Figure 2-7 illustrates this long term system behavior for a 0.10 ft break.

For this case, the LPIS are operative at approximately 2300 seconds, and a substantial decrease in system void fraction results.

An arbitrary pump trip af ter approximately 2700 seconds would not result in core uncovery.

The potential for core uncovery due to an RC pump trip is thus limited to a discrete time period during which the na ural ev:1ution of the system produces high void frac-tions and prior to LPI actuation. For a 0.1 ft2 break, this time period is on the order of 2000 seconds.

For smaller breaks, this critical time could be a few hours even if the operator initiated a controlled cooldown and system depressurication as recommended in the small break guidelines.

Although the analyses described above used 2 HPI pumps, the effcet of only 1 HPI pump available on the system void fraction evolution while the RC pumps are operating is not significant.

Figures 2-8 and 2-9 show the impact of one versus two HPI pumps on system pres-sure and average void fraction transients for a 0.05 f t2 break with the RC pumps operative. As seen from these figures, the results with one HPI pump are not significantly different to the two HPI pump case and are bounded by the spectrum approach utilized. With one HPI pump, the system does depressurize more slowly (less steam condensation) and a higher short term equilibrium void fraction is achieved. Also, recovery of the core following a loss of the RC pumps uculd be significantly longer with only 1 HPI pump avail-able.

The majority of the analyses provided in this report uses two HPI pumps and demonstrates a core cooling problem with worst time pump trip given that assumption.

As analysis of one HPI available cases would only show a larger problem, such cases have not been exten-sively considered.

As demonstrated in section B.4, the resolution of this problem, forced early pump trip, provides assurance of core cooling for both one or two HPIs available cases.

Therefore, kk 0

there is no need for further pursuit of the single IIPI available case.

The effect of the RCP tripping during the transient was studied by assuming that the pumps are lost when the system reaches 90% void fraction. Loss of the RC pumps at this void fraction is expected to produce essentially the highest peak cladding temperature.

After the RC pumps are tripped, the fluid in the RCS separates and liquid falls to the lowest regions, i.e., the lower plenum of the RV and the pump suction piping.

At 90% void fraction, the core will be totally uncovered following the RC pump trip. Thus, the time required to recover the core is longer than that for RC pump trips initiated at lower system void fractions.

System void frac-tions in excess of 90% can possibly result in slightly higher ten-peratures due to the longer core refill times that may occur.

However, the peak cladding temperature results are not expected to be significantly different as the system pressure and core de-cay heat, at the time that a higher void fraction is reached, will be lower.

Table 2-2 shows the core uncovery time for the cases analyzed with the RC pumps tripping at 90% void fraction with 2 HPI pumps avail-able for core recovery. As shown, the core will be uncovered for approximately 600 seconds for the breaks analy:cd.

Figures 2-10 and 2-11 show the system pressure and void fraction response for 2 break with a RC pump trip at 90% void fraction. As the 0.075 ft seen in these figures, the system depressurizes faster after the RC pump trip, due to the change in leak quality, and the void fraction decreases indicating that the core is being refilled.

Figure 2-12 shows the core liquid level response following the RC pump trip. The core is refilled to the 9 foot level with collapsed liquid approximately 625 seconds after the assumed pump trip.

Once the core liquid level reaches the 9 foot elevation, the core is expected to be covered by a two-phase mixture and the cladding temperature excursion would be terminated.

1168 103 m

5.

Effect of 1.0 ANS versus 1.2 ANS Decav Curve An analysis was performed using the more realistic 1.0 ANS decay curve instead of 1.2 ANS decay curve.

The study was done for a 0.05 ft2 break with 2 HPI;s available and pumps tripped at 90%

system void fraction.

Figures 2-13 and 2-14 show a comparison of system pressure and average system void fraction for 1.0 and 1.2 ANS decay curves. As seen in Figure 2-13, the system pressure for 1.0 ANS case begins to drop from saturation pressure (s1100 psia) about 200 seconds earlier than the case with 1.2 ANS as a result of reduced decay heat. Also, the system will evolve to a lower average void fraction as shown in Figure 2-14.

After the pumps trip at 90% system void fraction, the case with 1.0 ANS decay curve has a shorter core uncovery time by approximately 200 sec-onds compared to 1.2 ANS case. This case demonstrates that the effect of a delayed RC punp trip may be acceptable when viewed realistically.

A peak cladding temperature assessment for this case will be provided in a supplementary response planned for September 15th, to the ISE Bulletin 7905-C.

6.

Effect of "o Auxiliary Feedvater Analyses have also been performed with the RC pumps available and no auxiliary feedwater. These analyses all assumed 2 HPI pumps were available.

The system void fraction evolutions for these calculations were not significantly different from those discussed with auxiliary feedwater. Thus the conclusions of tha cases with auxiliary feedwater apply.hh.. - -..

.2 Break Location Sensitivity Study A study was conducted to demonstrate that the break location utilized for the preceeding analyses is indeed the worst break location. As stated previously, the analyses were performed assuming that the break 2 was 1cohted in the bottom of the pump discharge piping. A 0.075 ft hot leg break was analyzed to provide a direct co=parison to a similar case in the cold leg. For this evaluation, the RC pumps were assumed to trip af ter the RCS void fractica reaches 90%. Figure 2.15. hows the average system void fraction transient and the core uncovery times for both the 0.075 ft hot end cold leg breaks. As shown, the cold leg break reaches 90% void fraction approximately 150 seconds earlier than the hot leg break. Also, the cold leg break yields a core uncovery time of 175 seconds longer than the hot leg break. The quicker core recovery time for the hot leg break is cc.used by the greater penetration of the IIPI fluid for this break. For a cold leg break in the pump dischargo piping, a portion of the llPI fluid is lost directly out the break and is not available for core refill. For a hot leg break, the full IIPY flow is available for core refill. Thus, as shown by direct compa - and for the reasons given above, hot leg breaks are less severe th,n breaks in the pump discharge piping. 8 Peak Cladding Temoerature Assessment As described previously, a RC pump trip, at the time the RCS Ed fraction is 90%, will result in core ancovery times of approximate 1; 600 seconds. The peak cladding temperatures for these cases were evaluated using the small break evaluation model core poscr shaps to demonstrate compliance with appendix K and 10CFR50.46. '_l s, adiabatic heatup assumption during the time of core uncoves . util: /.ed. This approach is extremely conservative in that the power s ..e and \\\\6b '\\05 7 local power rate (kw/ft) analyzed is not expected to occur durir.g normal plant operation. Furthermore, use of an adiabatic heatup assumption neglects any credit for the steam cooling that will occur during the core refill phase and also neglects the effect of any radiation heat transfer. Using a decay heat power level based on 1.2 ANS at 1500 seconds, the cladding will heatup at a rate will be 6.5 F/S under the adiabatic assumption. With a core uncovery period of 600 seconds and the adiabatic heacup assu=ption, cladding temperatures will exceed the criteria of 10CFR50.46. Use of a more realistic heat transfer approach with the entreme power shape utilized for this eval-untion is also expected to result in cladding temperature in excess of the criteria. In order to ensure compliance of the 177 FA lowered loop plants to the criteria of 10CFR50.46 a prompt tripping of the RC pumps is required. Section B. demonstrates that a prompt trip of the RC pumps upon receipt of a low pressure ESFAS signal will result in compliance to the criteria. An evaluation of the peak cladding temperature using a power shape encountered during normal operation for a reali6 tic transient response with delayed RC pump trip will be provided by September 15, 1979. \\\\63 \\D6.

C. Annivsis Apolicability to Davis-Besse I The significant parametric differences between the raised-loop Davis-Besse I plant and the preceeding generic lowered-loop analysis are in the high pressure injection (HPI) delivery rate and the amount of liquid volume which can effectively be used to cool the core. The liquid volume differential is due to the basic design difference; raised versus lowered loops. Because of the raised design, system water available after the RC pumps trip will drain into the reactor vessel. For the lowered loop designs, the available water is split between the reactor vessel and the pump suction piping. Thus, for the same average system void fraction, the collapsed core liquid level following an RC pump trip is higher for the raised loop design than for the lowered loop design. Figure 2-16 shows a comparison of the delivered HPI flow for the Davis-Besse I plant and the lowered loop plants. As shown, for a similar number of HPI pumps available, the Davis-Besse I pumps will deliver more flow. For the delayed pump trip cases presented in section B.4 of this report, the Davis-Besse I plant will take approximately 450 seconds to recover the core as opposed to :600 seconds for the lowered-loop plants. However, it is noted that the core recovery time is based on using two HPI's rather than one, as required by Appendix K. Use of only one HPI pump for Davis-Besse I will result in core uncovery times in excess of 600 seconds. The Davis-Besse I plant cannot be shown to be in compliance with 10CFR50.46 for a delayed RC pump trip. Prompt reactor coolant pump trip is, therefore, necessary to ensure compliance of the Davis-Besse I plant with 100FR50.46. 1.163 107.

.D. Effect of Promat RC Pump Trip on Low Pressure ESFAS Sig.lal As demonstrated by the previous sections, the ECC system can not be demonstrated to comply with ICCFR50.46 using present evaluation techniques and Appendix K assumptions under the assumption of a delayed RC pump trip. Thus, prompt tripping of the RC pumps is necessary to ensure conformance. Operating guidelines for both LOCA and non-LOCA e7ents have been developed which require prompt tripping of the RC pumps upon receipt of a low pressure ESFAS signal. Because no diagnosis of the event is required by the operator and ESEAS initiation is alarmed in the control room, prompt tripping of the RC pumps can be assumed. The effect of a prompt reactor coolant pump trip on an ESFAS signal has b e e-examined to ensure that the consequences of a small LOCA are 2 bounded by previous snall break analyses which assume RC pump trip on reactor trip. As shown by Table 2-3 at the time of low pressure ESFAS initiation, keeping the RC pumps running results in a lower average system void fraction. This occurs because the availability of the RC pumps results in lower hot leg temperatures and thus less flashing in the RCS at a given pressure. Thus, a prompt trip upon receipt of an ESFAS signal will result in a less severe system void fraction evolution than cases previously analyzed assuming RC pump on reactor trip. E. Conclusions The results of the analyzes described in this section can be summarized as follows: 1) 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 1 1 6 8 1 0 8J fractions in excess of 90%.

3) At 40 minutes, the 0.025 ft break han evolved to only a 47% void fraction. Thus, a delayed RC pump trip for breaks less than 0.025 f t will not result in core un overy.

4) The potential for high cladding temperatures for a small break transient with delayed RC pump trip in restricted to a time period between that time where the system han evolved to a high void fraction and the time of LPI actuation.

5) Even with 2 IIPI pumps available, tripling of the RC pumps at the worst time (90% void fraction) resulta in a core uncovery period which cannot be shown to comply with 10CFR50.46, if Appendix "

assumptions are utilized.

6) A prompt RC pump trip upon receipt of a low pressure ESFAS signal will provide compliance to 10CFR50.46,

7) The above conclusions are applicable t o both the B&W 177 FA lowered and raised loop NSS designs.

~.

III. TMPACT ASSESSMENT OF A RC PU?T TRIP ON NO"-LOCA EVENTS A. Introduction Some Chapter 15 ' events are characterized by a primary system response similar to the one following a LOCA. The Section 15.1 cvents that result in an increase in heat removal by the secondary system cause a primary system cooldown and depressurization, much like a small break LOCA. Therefore, an assessment of the conse-quences of an imposed RC pump trip, upon initiation of the low RC pressure ESFAS, was made for these events. B. Cencral Assessment of Pumn Trip in Non-LOCA Events Several concerne, have been raised with regard to the ef f ect that an early pump trip would have on non-LOCA events that exhibit LOCA characceristics. Plant recovery would be more dif ficult, dependence-on natural circulation mode while achieving cold shutdown would be highlighted, manual fill of the steam generators would be required, and so on. However, all of these draubacks can be accommodatco since none of them will on its own lead to unacceptable consequences.

Also, restart of the pumps is not precluded for plant control and cooldown once controlled operator action is assumed.

Out of this scarch, three major concerns have surfaced which have appeared to be sub-stantial enough as to require analysis: 1. A pump trip could reduce the time to system fill /repressurization or safety valve opening following an overcooling transient. the time available to the operator for controlling ILPI flow and the margin of subecoling were substantially reduced by the pump trip to where timely and effective operator action could be questionable, the pump trip would becone unacceptable. 2. In the event of a large steam line break (maximum overcooling), the blowdown may induce a steam bubble in thp RCS which could impair natural circulation, with severe ccusequences on the core, es-pecially if any degree of return to power is experienced. 3. A more general concern exists with a large steam line break at EOL conditions and whether or not a return to power is experienced following the RC pump trip. If a return to critical is experienced, natural circulation flow may not be sufficient to remove heat and to avoid core damage. .} "s 10._--- -

Overheating events ucre not considered in the impact of the RC pump trip since they do not initiate the low RC pressure ESFAS, and therefore, thcee would be no coincident pump trip. In addi-tion, these events typically do not result in an empty pressurizer or the formation of a stcan bubble in the primary system. Reactivity transients were also not considered for the same reasons. In addi-tion, for overpressurization, previous analyses have shown that for the worst case conditions, an RC pump trip will mitigate the pressure rise. This results from the greater than 100 psi reduction in pressure at the RC pump exit which occurs after trip. C. Anal.,is cf Cencerns and 5.crul:s 1. System Reprensurization In order to resolve this concern, an analysis was perf ormed for a 177 FA plant using a MINITRAP model based on the case set up for TMI-2, Figure 3.1 shows the noding/ flow path scheme used and Tabl,e 3.1 provides s descriptio-of the nodes" and flow paths. This case assumed that, as the result of a small steam line break (0.6 f t. split) or of some combination of secondary side valve failure, secondary side heat demand was increased from 100% to 138% at time zero. This increase in secondary side heat demand is the smallest which results in a (high flux) reactor trip cnd is very similar to the worst moderate frequency overcooling event, a failure of the steam pressure regulator. In the analysis, it was assumed that follouing EPI actuation on low RC pressure ESFAS, main feedwater is rcmped down, MSIV's shut, and the auxiliary feedwater initiated with a 40-second delay. This action was taken to stop the cooldown and the depressurization of the system as soon as possible after HPI actuation, in order,to minimico the time of refill and repressurization of the system. Both HPI pumps were assumed to function. The calculation was performed twice, once assuming two of the four RC pumps running (one loop), and once assuming RC pump trip right after HPI initiation. The analysis shows that the system behaves very similarly with and without pumps. In both cases, the pressurizer refills in about 14 to 16 minutes from initiation of the transients, with the natural circula- \\y63 '\\ -

t

g..

tion case refilling about one minute before the case with two of four pumps running (See Figures 3.2,3. 3). In both cases, the system is highly subcooled, from a minO:um of 30*F to 120'r and increasing at the end of 14 minutes (ref er to Figure 3.4). f[ It is concluded that an RC pump trip following ILPI actuation b will not increase the probability of causing a LOCA through f the pressurizer cede safetics, and that the operator will have the same lead time, as well as a large margin of subcooling, to f control HPI prior to safety valve tapping. Although no case ( vith all RC pumps was made, it can be inferred from tbc one f loop case (with p=ps r=ning) that the subcooled margin will The 4 be slightly larger for the all pumps running case. 1 pressuriner will take longer to fill but should do so by 16 minutes into the transient. Figure 3.t shows the coolant temperatures (hot leg, cold leg, and core) as a function of 3 l~ time for the no_ RC pumps case. 2. Effect of Steam Bubble on Natural Circulation Cooling For this concern, an analysis was performed for the same 99 generic 177 FA plant as outlined in Part 1, but assuming that } 2 as a result of an unmitigated large SL3 (12.2 f t. DER), the j i excessive cooldown would produce void formation in tbc primary i, P system. The intent af the analysis was to also show the b extent of the void formation and where it occurred. As in E the case analyzed in Part 1, the break was symmetric to both generators such that both would blow down equally, maximizing la 9 the cooldown (in this case there was a 6.2 f t. break on cach e loop). '1here was no 11SIV closure during the transient on either steam generator to maximize cooldown. Alse,, the tur- [ bine bypass system was assumed to operate, upon rupture, until isolation on ESFAS. ESFAS was initiated on low RC pressure and also actuated IIPI (both pumps), tripped RC pumps (when applicabic) and isolated the MrWIV's. The AFU i was initiated to both generators on the low SG pressure signal, with minimum deley time (both pumps operating). l This analysis was performed tuice, once assuming all RC pumps running, once with all pumps being tripped on the IIPI actuatien (after ESFAS), with a short (45 second) delay. In both cases, voids were formed in tbc hot legs, but the dura- } } b'h \\,

tion and size were smaller for the case with no RC pump trip (ref er to Figure 3. 7). Although the RC pmap operating case had a higher cooldown rate, there was less void forma-tion, resulting f rca the additional system mixing. The coolant tempetatures in the pressurizer loop hot and cold legs, and the core, are shown for both cases in Figures 3.5, 3.6. The core outlet pressure and SG and pressurizer levels versus time are given for both cases in Figures 3.8, 3.9. This analysis shows that the system behaves very similarly with and without pumps, although maintaining RC pump flow does seem to help mitigate void formation. The. pump flow case shows a shorter time to the start of pres-suriser refill than the natural circulation case (Figure 3.9), although the time difference does not seem to be very large. 3. Effect of Return to ?ower There was no return to power exhibited by any of the EOL cases analyzed above'. Previous analysis experience (ref. Midland FSAR, Section 15D) has shown that a RC pump trip will nitigate the consequences of an EOL return to power condition by reducing the ecoldown of the primary system. The reduced cooldown substantially increases the subcritical cargin which, in turn, geduces or eliminates return to power. D. Conclusions and Su rary A general assessment of Chapter 15 non-LOCA events identified three areas that warrauted further investigation for tapact of a RC pump trip on ESFAS low RC pressure signal. 1. It was found that a pump trip does not significantly shorten the time to filling of the pressurizer and approximately the same time interval for operator action exists. 2. For the maximum overcooling case analyzed, the RC pump trip increased the amount of two-phase in the primary loop; however,thepercentvoidformatiodisstilltoosmallto affect the ability to cool on natural circulation. 3. The suberitical return-to-power condition is alleviated by the RC pump trip case due to the reduced overcooling effect. Based upon the above assessment and analysis, it is con-cluded that the consequences of Chapter 15 non-LOCA events are not increased due to the addition of a RC pump trip on ESFAS low RC pressure signal, for all 177 FA lowered loop plants. Although there were no specific analyses performed for TECO, the conclusions drawn from the analyses fer the lowered loop plants are applicable. 6 4 O t

\\\\63\\\\Y

e Table 2-1. Analysis Scope With AFU Available Continuous RC Break location pump operation RC pump trip @ 90% void Break size, (ft2) Cold leg Hot leg 2 HPI 1 HPI 2 HPI 0.025 X X 0.05 X X* X X* 0.075 X X X X 0.10 X X X G.20 X X em use es en een kAnalyzed with both 1.0 and 1.2 ANS dccay curves. 9 V e Table 2-2. Impact Assessment of Break Spectrum With RC Pumn Trip at 90% Void Break size (ft ) Core uncovery time (sec) 2 0.10 550 0.075 625 0.05 575 o tes : 1. Two HPIs ara _labic during the transient. 2. Core uncovery time is the time period following pump trip re-quired to fill the inner RV with water to an elevation of

9. ft in the core which is ap-proximately 12.ft uhen swelled.

) { }} \\\\ ' 21 -

Tabic 2-3. Compa'rison of Systen Void Fractions at ESFAS Signal System void fraction ' ^ Break size, 2 (ft ) Pumps on Pumos trioned 0.02463 0.0 0.04 4.47 0.05 0.04 0.055 6.74 0.07 8.06 0.075 0.90 0.085 8.45 0.10 2.17 7.97 0.15, 10.70 0.20 6.78 1163 iI7 MINITRAP2 NODE DESCRIPTIO:{ N_0DE NUMBER DESCRIPTION 1,33 Reactor Vessel, Lower Plenum 2,34 Reactor Vessel, Core 3,35 Reactor Vessel, Upper Plenum 4,10 Hot Leg Piping 5-7,11-13 Primary, Steam Generator 8,14 Cold Leg Piping 9.32 Reactor Vessel Downcomer 15 Pressurizer 16,24 Steam Generator Downcomer 17,25 Steam Generator Lower Plenum 18-20,26-28 Secondary, Steam Generator 21,29 Steam Risers 22,30 Main Steam Piping 23 Turbine 31 Contcinment MINITRAP2 PATH DESCRIPTION PATH NIDiBER DESCRIPTIO;i 1,2 Core 45,46 Core Bypass 3,5,5,11,12,44 Hot Leg Piping 6,7,13,14 Primary, steam Generator 8,15 RC Pumps 9,16 Cold Leg Piping 10,43 Downcomer, Reactor Vessel 17 Pressuriner Surge Line 18,19,26,27 Str .erator Douncomer 20,21,28,29 Set .~try, Steam Generator 22,30 Aspirator 23,31 Steam Riser 24,32 Steam Piping 25,33 Turbine Piping 34,35 Break (or Leak) Path 36,37 HPI 38,39,43,44 AFW 40,41 Main Feed Pumps 42 LPI Table 3.1 1163 ii8. N* E bW Mhb 6 ggg

Figure 2-1. CPJJT2 lioding Diagram for S:: tall Break h h3 T iE U L 3 4 h (A / ,6 s <s < s ( 3 h ' ~ b Q p. e-s r T 7 w = g; m,,~1.......,,.. et.. .r.: g n,,...,_...-. b b Q .s tu a r.sm tre. r.t> g .,..c....,_, 7 (3 0,. s. ..t. . t..i t ca 122 r.es ... e _.... I.th-at ,e pt.s .t s Co... tu.e.t seier sr.ca. } ode No. Identificaticm Peth No. 7dentification 1 Ikwnecmer 1,2 Core 2 Zever Plenus 3,4,18,19 Eot Les Piping 3 Core, Core Bypass, Uoper 5,20 Bot lec Upper Plenus, Upper Head 6,21 SG Tubes 4,14 Uot les Piping 7.22 SCAver Eezd 5,15 - Stez:= cenerator Upper S Core Byps::s Eead, SC Tubes (Upper Ealf) 9,13,24 Cold Leg Piping 6,16 SO Tubes (Lover Half) 10,14,25 Pt=ps 8,20 SG Lover Head 11.12,15,16,26,27 Cold Leg Piping 3.11.19 Cold Leg Pipi=g (Pu=p Suetion) 17,31 Dovncomer 10.12.20 Cold Leg Piping (Pump Discharge) 23 III 13 Upper Downec=cr 28,29 Upper Down:orer - (Above the ( of Nozzle Selt) 30 Pressurizer 21 Pressurizer 32 Vent Valve 22 Containment 33,34 Leak & Return Path 35,36 EPl 37 Containmen: Sprays g e 1163 119

Figure 2-2. CRAFT 2 N0 DING DIAGRAM FOR SMALL BREAKS (6.N0DE MODEU CFT 4 ~ 2 6 5 3 1 N-1 s ~ / J// s es M@ LEAK PATHS 8 & 9 M@ ( x un 2:=:i 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 Fot Legs 5 Containment S Pumps 6 Secondary SG 6 Vent Valve 7 Pressurizer 8,9 Leak & Return Path 1168 120.

• w e,

CORE PRESSURE VS TIME,177-LL, 2772 Et, PUMPS ON 2 0.025 FT BREAK 23 1400E MODEL 2 20 ~ 0.025 FT BREAK ~~~ ~ 6 tt00E MODEL 18 n 'E K 10 E a 14 2 S 8 12 10 8 0 500' 1000 1500 2000 2500 Time, sec Figure 2-3 ))b'b - 2s -

PERCENT SYSTEM V010S VS TIME, PUMPS ON 100 80 e qE S 60 9 2 40 1 (1} M b >UD.N I ~ E 5 0.016 ugg['.\\0010 Yp e 29 g. P ' G 'g,015 f* i/ 0 i 0 400 800 1200 1600 2000 2400 Time, sec Figure 2-4 e BREAK SPECTRUM-RC PRESSURE VIITH THE RC PUMPS OPERATIVE AND 2 HPI PUMPS 2500 2000 m i 1500 J (r ti e'- . ' %. ~ 0. 0 2 5 F T p 1000 2 N \\, \\ ~ s s~N \\ N 500 ~\\ + 2 1 Lph,*"N x-- -.u -0.05 FT .N 2 0.075 FT 2

• \\.~ 0.2 FT2 0.10 FT 0

f 0 500 1000 1500 2000 2500 3000 3500 Time, sec Figure 2 5 0 e s 8 BREAK SPECTRUM-AVERAGE SYSTE!4 V010 FRACTION WITH THE RC PUMPS OPERATIVE AND 2 HPI PU!APS 100 \\ y.o.3<o -;,a,ga,_, / / l' lo /+'. / , s..(,, 80 -1l / y N.l 4% ~ / a e ! ~ll & / % s,' /

s

%t E g 60, / / D/ o + h = a/ v/ Es /. /.-

/

/ / 40 -:/ f

g

/ /. / Mp '. c g.@ E / / / o +/ / 20 .f./. 0 0 400 800 1200 1600 2000 2400 2800 Time, sec Figure 2-6 1168 124 2 BREAK WITH CONTINUQUS RC PUMP 0.1 FT OPERATION AND 2 HPl PUMPS 2500 100 7 ~ ~ - _,_ _ _ j LPI l N 60 - [ l\\\\ 2000 / s a / N \\ a / \\ .E / N 60 / g - 1500 [-a e / s / \\ s_ i 5 g 40 -\\ -/,' s / 1000 g 8 / 's N's / / LPI 20 N 500 N l w ~. ~ ~ ' - I' 0 O 0 500 1000 1500 2000 2500 3000 Time, sec Figure 2-7 . 1168 125

RC PRESSURE FOR 0.05 FT2 BREAK AVAILABLE 1 HPl VS 2 iiPI'S 30

---

2 HPI'S, PUMP DN, HOMOGENE0VS

~ ~ 25 ^ ~ 20 o O .2 E. 15 i n' u l O \\ 10 - s---- w o%o%o = ~s o_ s o N s 'N o s s N O g 5 ~~D 0 i 0 500 1000 1500 2000 2500 3000 Time, sec Figure 2-8 1163 126 2 AVERAGE SYSTEM V010 FRACTION FOR 0.05 FT AVAILABLE 1 HPI VS 2 HPI'S 100 ,o--- -o -o -_ o o' 7 p,,- 0 = B0 s' l// ,f

---

2 HPI'S, PUMP DN,HOMOGENE0US o

f ~ 60 / -o-1 HPI, PUMP Oil, HOMOGENEOUS , S / ./ -e m s' 40 a 8 / l / / 20 0 i e 0 400 800 1200 1600 2000 2400 2800 3200 Time, see Figure 2.? '\\\\63 121 2 RC PRESSURE FOR 0.075 FT, PUMPS OFF 3 90i'. SYSTEM V010 30 _ _ _ _ 2 HPl'S, PUMP DN, 25 HOMOGENEOUS . - 2 HFI'S, PUMP GFF G 90 f'r V010, 2 PHASE ^ 20 ~ O .2 E 15 S l i -\\ _.,N = t )0 N h9.s N.' 5 Ns

• s ' ' ~ '**** %

N. N.% 0 O 500 1000 1500 2000 2500 3000 Time, sec Figure 2-10 I AVERAGE SYSTEM V010 FRACTION FOR 2 0.075 FT, PUMPS OFF e 905 SYSTEM V010 100 ~ / /

• ~.%.

\\*N,% p 80 f e / / q l 2 l- / 60 S / / / / 2 HPI'S, PUMP DN, HOMOGENEOUS w / 5 40 / 2 HPI'S, PUMP OFF G 905 VOID,. u p ,/ 2 PHASE / 20 I 0 O 400 800 1200 1000 2000 2400 3200 Time, sec Figure 2-11 1163 129 AVAILABLE L10U10 VOLUME VS TIME 2 FOR 0.075 FT BREAK WITH 1.2 ANS DECAY HEAT CURVE 3000 ^ m { 2500 Ea

E 2000 3

9' LEVEL OF ACTIVE CORE o 1500 E E: 0 1000 i i RC PUMPS OFF 500 0 400 800 1200 1600 2000 Time, sec ~ Figure 2-12 e \\\\63 \\b

2 RC PRESSURE VS TIME FOR 0.05 FT BREAK WITH 1.0 ANO 1.2 AUS BEFORE AND AFTER PUMP TRIP 2 0.05 FT, 2 HPI'S 1.2 ANS, PUMP DN 3000 - 2 0.05 FT, 2 HPI'S ~~ 1.0 ANS, PUMP DN 2 2500 - 0.05 FT, 2 HPl*S 1.2 ANS, PUMP 0FF 2 0.05 FT, 2 HPI S

• ~ 1.0 ANS, PUMP OFF

,- 2000 o. d a, v 1500 "'g a-I \\ \\ 1000 - ~ -- na' ,C,7 - ,s..,, ~ ' ~~s.% 500 ' h-k g q, 0 0 500 1000 1500 2000 2500 3000 Titte, sec Figure 2-13 1163 13I _ m._

2 PERCENT SYSTEM V010 FRACTION FOR 0.05 FT BREAK VitTH 1.0 At101.2 ANS BEFORE AND AFTER PUMP TRIP 100 -;.Ws'@~._ -}& " e.'/ n0 ~ ~ o

i. /

2 sf 60 /[ 2 ,,0.05 FT, 2 HPI'S, PUMP ON, S. 7f-1.2 ANS 4 E 2 0.05 FT, 2 HPI'S, PUMP DN, f 5 40 / 1.0 ANS / o 2 5 o ,_0.05 FT, 2 HPI'S, PUMP OFF, f 1.2 ANS ,/ 2 0 g 0.05 FT, 2 HPI'S, PUMP 0FF, 1.0 ANS 0 i 0 400 800 1200 1600 2000 2400 2800 Time, sec Figure 2-14 S S 0 63 t3 2 AVERAGE SYSTEM V010 FRACTION VS TIME FOR A 0.075 FT BREAK, BREAK LOCATION COMPARISON PUMPS OFF e 905 V010 100 UNC0VERY TIME = G25 I le 7 SEC $g BD g h/h /e UNC0VERY TIME = j f 60 /p 450 SEC a /g@ N = v / E / r ~ <n /

• 40

/ g* ~c 0 k'N i 20 - // 0 0 400 800 1200 1600 2000 2400 Time, see Figure 2-15 O e \\\\63 \\b COMPARISON OF DEllVERED HIGH PRESSURE INJECTION FLUl0 TO RV FOR PUMP OISCHARGE BREAK \\ 1200 - N 1100 N (# \\ Cp 1000 \\#h/ 900 N R= 600 \\g i 700 \\ 600 \\

1. t#p#

\\ 0 [= 500 \\

1. pf

\\ (E0 gggp I 5 400 5 k NP/ 300 N j s LOW Loop 200 's HPI 100 j 0 I 0 200 600 1000 1400 1800 2200 2600 3000 Pressure, (psia) Figure 2-10 e, :.. v

vsv i su M .21

• D4

..r.-~~:. l 3! .L1 3o g .....--.------------.--f '( - - - - O - e........, O tuov f lc'OV-t+ l 1 C-@--

---

- u r.w2, yg 15

.2 I ,t. 1 ie y- _of,_, a2 It S Jo g* --h> ~ g, 1 Q 10 f e] ] @{ }@ O e 4 m 97 11. b L If M 9~ g u, cn - e e 3[ ] <o 1 ( ]O h lb s 26 ^ + y n et @ e n-as-c e _7 c 1 6 I 2 R g .O 'i@ e 1 g Figure e 3.1

• 1

\\\\63 \\'9 IIINITRAP2 Noding and Flow Path Scha.m.c _ ~,.., - ,q

PRESSl":12Ef; l. ] EiEA" CE:.EI;Alci: L !C'J10 I E7:L :l I CU3 Ti-l::SIENT 11."E (102.' f P. Ei:3 0F L1TE,D.6 FT STE A:'.LI::E E:El.Ti (DOU::0!;;G l'CCEkATE 2 FREU. ). (RC PIL'? TRIP) S.G. 10BE REC 10N ooeogo 0 FULL 0 5 0 0 G PRZR. FULL .=, g g 40 D u '5 e g g D Q O O 4 g O g 30 g g O 0 S e 0 00@e e 5 0 e e U O $ 20 S e a w E D G 5 G U 6 KEY e a 0 0: PRESSUR17.ER 10 6 o: STEAM gel'ERATOR ' A' O 6 6: STEAM GENERATOR 'B' i t i i i -i 0 2 4 6 8 10 12 14 16 Transient Time (Minutes) b .] } b 'd l Figure 3.2 . 41 -

D bs D b Dh b c. n.- I _ _a eu , u) w ~ ~ \\ m m o a b .,t ..c = a g gr.. g c, ry t. i, t._, Lu Z J \\ N.: Z 8 g v c2 f9 d EdW 'E

s a

.3

.s; :- 3.- rc c3 CC & Us - Qt Q, l co -4

r.

N V \\ NI Or k Iu tu tu M N o- \\ cc >- >-- \\ o x l \\ \\ \\ c2 0 g w M g j \\ J w g g y O O d > m cu y 4 -- s g \\ w

2 s

s x C w z s > - ^ s \\ \\ tu J D. g g g =r .J

2 -

y s v g \\ <x. ric \\ g o tu t-- \\ s M o' os - >~. .E D M n. \\, 's '\\ N 5', m !!i = o o_ \\ g g u w o \\ g C Cd O co t2.1 \\ s \\ o to 0; n: a: s J -t cm n CO u$ ~co D ~" ~ \\ ~D LD M \\ 14J 23 RJ. 6 >Z

2

W -

t-s p w - v M.8 - p\\ g - - CD CD J C\\_ \\ o n. u \\N = c.o m u O \\ E VJ CD gg e < CD o \\ g Lu J V3 O g N s s%s o = m %g g 4 2 < - a g ) C.D tu g \\ g ce tu er s b D b w cc u_ N -w ce o. => u. m a: GG O m a w w ~ o m o o C. E e G O m O O O G O CEI l t p _ _. go i o o o c2 c3 O W M N e (11) I D A0'I rit nh!1 Jo:cicuc3 mcai s "J ari nissa y Figure 3.3 g i, j ~[ ~ -

• \\,L.,

' ^ rneuw w..- Il OO c wt n i ^ C 3O M -C l ~ L O ul X f OC ( C'.3 C") O I ~. P-* 'i' I OO 4 m f-I cc t -J ~ H o OO G m en ) N g ce ct z a. o-v Z w I OO G B -o o 0 O C o w.-. 3 x I w e . e. a 00 G w cr: ^ w o o I n CC J >~! O V3 .- w e Q Q o o a 1- -, stl cc m "s tn to 5 ~C p a M I AJ Cf I OO cc z w 90 ( w g O C ( 8 I R! m I OO G. "M co L: w w a m *-- o G o m .a I OO G Om t c: m + w O O G w w m g -= cu o i Og 4 E a co. re w l OO G o ca >_ cu i O O q w = m a I OO G m" g o 00 G 8S i _ e OO G v i e3 G I o G I CD G g - m l G I a:) G I ED G / CO G /' - d. g s _ t e o e o o e o M M o o o a m m v (3,) ajnlejartmal lueloo3 figure 3.4 }}()O l30 f.'";L t. ':1 : E" ?i : ".i., '.' ~. ' L' S 2.- _ i U. i i. 2 S ' it i !- l '. i. r llH. i2 2 fi g,,:.., g (l ?i2 FP. r f f'n P ""il. E. S T E ' "'. I ':E C"E,'?. ' '!"" l T I C '. I'_0 ) pg r.n ni. 9 v n !,n ) Jgg I I PRESSURIZER l E."PTIES L P [10T LEG BEGit'S TWO-PHASE i I\\ j li \\s i CSL FL900 TaiJ 003 U-s j K FL0iS CEGil'S \\ l \\ T I-{ HOT LEG SUBC00 LED g ^ 500 9 2\\ .k as DSg hN b\\ s C s "'s a -'~N E 400 / \\ N g ss/ \\ \\ \\ \\ KEY t Q \\ 0: il0T LEG (PRZP. LOOP) 5 \\' 300 o: CORE g a: COLD LEG (PRZR LOOP) O SATURATION LINE e il i 200 0 2 4 6 8. 10 12 Transient Time (lainutes) Figure 3.5 ibb _ f, f, _

CGntl.!!T TE;iTEiil.TU.,' S VE?1US 1: '.:::' :1 i!": (107 FP, BE'il!. Hi:G OF L IFE,12.2 F i iN.,i.E E;;3 i";Pi!;.:E,!!:7 ' T I C.'.TED STE."'L !!:E 0.0E! ',P.C FU;.:P li:!P j 7 0 0 g... = _ _... _ PRESSURIZER E!!.PT IES fH0TLEG l BEGH S TiiG-PHl.SE j\\ r.) \\ CLf,E FLOCD TA!!K j 000 'N' s FLO7! CECll'S \\ I HOT LEG SUSC00 LED C a s { ( I o O as U \\ 5 500 o o A s c 9 s n a o o 6s bs 9Yh R a o N o o 6 U 5 o b s f a o

b. % s' o

400 o o o a e U o o 3 U a e o KEY 3 o: il0T LEG (PRZR LOOP) 300 3 0: CORE 6: COLD LEG (PRZR LOOP) o SATURATI0il Lit!E 200 i f _11 0 2 4 6 8 10 Transient Ti:::e (fliniites) Figure 3.0 ilot3 140 o 43 -

S il..' " L U L _' t i V 0 L U'.c 'li it' US Ti', o.:. Ii '.i 11; r 162 F P... tuli...ii.L Di liiE.12. 2 i I LLuJLE L 2 I,UPl:. ~.L, U,: 'lil'

• 1! 3 Siil."L l:

+' 12 l I q h i i i il il 10 l 'i' i t I l i; 1 I I I \\ l 1 0 .I ^ \\ m t n 6 I t, ** g i m 4 f \\ g / \\ i i' \\ l I f I \\ l [.} \\ l I \\ I I 3 \\ l 2 l lg ( I / \\ l y'gf,E I \\ i % \\ \\ s gf. k I/ \\ I I jj/ \\ tI s h.___ Lb _ h__3 0 0 2 4 6 8 Transient Time (tainutes) Figure 3.7 g 0: HOT LEG (PRZR) - RC PUI:1P TRIP 0: HOT LEG 'B' LOOP-RC PU!dP TRIP 1,41 A: Il0T LEC (PRZR LO0l') - NO TRIP G: Il0T LLG 'B' LOOP-N0 TRIP - I:. 6 -

Crif. Oi ! t E 7 Pi, U :' : H : <Os it:- ' "I 11 r 2 (102~ f I'. 0:C f:. :!::: f;T L I,' E.,17. ? T T E:""LE Et,0 hu, I(. i, b;,.l116!,1L0 5i Ets: L #l,2 C.;L A ?;) 2 5 t: 0,._ _ t l f f. 2000 9 =- t o 5 1500 - 0 O O m O E o KEY 2 o RC PUJPS TRIP -a g 6: i;0 PU:1PS TRIP S 1000 a ao o o o o o 500 o o o op c O o o o6 O O oOgQo o ob O i i .t -- 0 2 4 6 8 10 12 Transient Time (Minutes) figure 3.8 ) ) ()'b ..,f.- .....- ~ ~. d () s <3 q. J w O G) b.= O z-( La J Q .-- C : ^ a m c.a - >. ~ W hJ J

• CL. DT
• K CD C'

.so Q 4 .M p W E*~J

n. u.

(. n s g LLI ~ O Dg ~ (.,. i. C')

5 g a m

,: a to -

^ C1 tb C - -- (6 l CC LAJ - Ce rc i. - cc m C# W W /* % '. f

.3 Cb 6.-

to 3 c._ n.

c..

r. 1 m Ct. CE A E

• E L

i M D l H T D D Cr Cc D (b Cb d ^ Lu Cb A M

- En

• 25 O OQ e

Z O = ca a> =a -J w CL CC v v W LAJ

"3 C

> LtJ O O ) J (D -~ LLJ J Cr Z LAJ LAJ s U J CO v v > > D LAJ V LJ, (, j C3 CD a., a - O ) E = C.:* rg 8 C.1 W m m O o o.J a H & <l >== ==& LA. .t Lf3 CC C Cr rr C Cr N 14.s LAJ tAJ uJ GJ LAJ c N N.i.* Z ) N N y w ' q M Cr rc C ') C;, D -a g") C C: Q LaJ w en v> u g n w u. 6-M "** dV to to < < j 9 LAJ L4.J LaJ LA.s LAJ J CL: s In u. e-W A LA. [b Ib Co M O C1 g O a Ce z M LAJ O 2 at Q M v LAJ to I art C3 Cr Cc LA s to LAJ Cu Z LAJ .Z.= uJ C3 O. LL J E L= d N 4 bO LLJ N LAJ & O H N v 1 h l o 3 9.i o N l C5 o c, e O %T tg y \\\\63 i 0 \\ naan n,.a, me 3 s i Fin.ure 3.9 - z, - o 8

REFERENCES ~ I B. !!. ann et al., "B&W's ECCS Evaluatlon Model," 3AU-10104, Rev. 3, August 1977. 2 Letter, J.H. Taylor (B5W to S. A. Varga (NRC), July 18, 1978. R. A. liedrick, J.J. Cudlin, and R.C. Foltz, " CRAFT 2 - Fortran Frogram for 3 Digital Simulation of a Multinode Reactor Plant During Loss-of-Coolant," 3A'.' _.M92. Eev. 2, April 1975. J.F. Wilson, R.J. Grenda, and J.F. Patterson, "The Velocity of Rising Steam 4 in a Bubbline Two-Phase Mixture," ANS Transactions, 5, (1962). \\'> > 9 u 63 144 ,,9 -}}