ML20141D956

Forwards Revised Response to RAI 440.120 Re Rapid Boron Dilution Scenario

ML20141D956
Person / Time
Site:	05200003
Issue date:	05/14/1997
From:	Mcintyre B WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP.
To:	Quay T NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
CON-NRC-NSD-97-5126 NUDOCS 9705200249
Download: ML20141D956 (12)

Text

.

i w

t..

e

%d i

Westinghouse Ellergy Systems Ba 355 Pittsburgh Pennsytvania 15230 0355 Electric Corporation NRC-NSD-97-5126 DCP/NRC0864 Docket No.: STN-52-003 May 14,1997 Document Control Desk U.S. Nuclear Regulatory Commission Washington, DC 20555

/

ATFENTION: T.R. QUAY SUI! JECT:

REVISED RESPONSE TO RAI 440.120 FOR RAPID BORON DILUTION SCENARIOS

References:

1. Letter from NRC to Westinghouse (HufTman to Liparulo), "AP600 Boron Dilution Transient Analyses," dated 9/24/96.

l

2. NSD-NRC-97-5062 (DCP/NRC0809), "AP600 Shutdown Evaluation Report i

(WCAP-14837) and Response to RAI 440.53", dated 4/15/97.

Dear Mr. Quay:

Attached is the revised response to RAI 440.120 regarding rapid boron dilution scenarios. This revision reflects resolution of issues presented in Reference 1 and discussed durir,3 an October 25, 1996, telecon involving Messrs. IluiTman, Sun, and Attard of the NRC and Messrs. Corletti, Kemper, Ilill, Prokopovich, Carlson, Deutsch and Ms. Nydes of Westinghouse.

The resision to item b of the response was provided previously by Reference 2 and is repeated in this transmittal. In Revision 1 of the RAI 440.120 response, deletions from Revision 0 are noted with a line through the deleted text, while additions appear in bold italics.

With this revised response, the Westinghouse status for DSER open item tracking system item 3960 is changed to " Action N". NRC is requested to review this response and provide Westinghouse with feedback regarding the status of this item.

Please contact Robin K. Nydes at (412) 374-4125 if you have any questions regarding this transmittal.

Brian A. McIntyre, Manager Advanced Plant Safety and Licensing

/'

jml

/

M' Attachment cc:

William C. Iluffman, NRC (w/ Attachment)

~l Summer Sun, NRC (w/ Attachment)

Anthony C. Attard, NRC (w/ Attachment) lhlhlhlll hll Nicholas J. Liparuto, Westinghouse (w/o Attachment) 9705200249 970514 PDR ADOCK 05200003 E

PDR

AP600 Open items Tracking bAemi Database: Executivo Sousanary Date: 5/14/97 l'

Selecease:

brem no] between 3960 And 3960 Sorted by Item #

t hem DSER Seckon Tale /Desmption Resp (W)

NRC No.

Brancit Quessaan Type Detal Status Engmeer Status Status

.. - - - ~ -.

.._Leme.r No. /.- - Dane.-

> 3%0 NRRSRXB 15.

TEIAM Novenshtem/SAR-Chl5 Action N Action W NTD-NRC-97-512t>

[$sEr frora Bill Hufinia to NEk lM7E215 500 Baron DiluETransient Analyn requeds'M information to what was

~

{protided in response to RAI 440.120 (what w refer to as the rapid baron Waa= RAll EieNf addeuonal questions is in progress. ska 10/1996

~

{Saatus Updase: We had a video eh call with the NRC on 10/24L96 and base faished incorporsung all conuments frein that call except for ~

enhancmg the Funish Center Sceaano section to provide neore detads regardag the samall break LOCA analysis. Since the NOTRUMP test jsinnulations are behind schedule, the SSAR cues which will prooide the reference transsent for this RAI regision are not yet available. rha 1/l$/97

-i jRevised RAI response in review.1Etter in typing. rka 5/6/97.

Revised RAI response provided May 14 by NSD-NRC47-5126 (DCP!NRCOS64) W status changed to Action N. ska 5/14

}

I i

n i

e i

i Pcge: 1 Total Records: I

i NRC REQUEST FOR ADDITIONAL INFORMATION

=

=

1

^

1 Question 440.120 The staff is concerned with boron dilution events for PWR designs. A slow, inadvertent dilution due to a malfunction of the chemical and volume control system (CVCS) or faulty operator actions is a design basis event that must be shown to satisfy stringent acceptance criteria. Recently, the question of whether additional failures or scenarios other than the CVCS malfunction events might lead to inadvertent criticality and fuel damage has received considerable attention in Europe and the United States. For example, a preliminary study by the Finnish Center for Radiation and Nuclear Safety indicates that an inherent mechanism for boron dilution exists in the cold leg loop seats or transients and accidents, e.g., a small break LOCA, involving heat removal by reflux-or boiler-condensation natural circulation. Under certain conditions and scenarios, such as during the restart of RC pumps, substantial boron dilution could result in the core, leading to a reactivity induced accident.

a.

Although the AP600 design does not have a loop seal in the cold leg, has Westinghouse evaluated the possibility of accumulating deborated (a highly dilute slug) water in the reactor coolant loop, especially in the steam generator cold leg channel head, as a result of reflux / boiler condensation natural circulation in an accident? Address this concern.

b.

For those transients or accidents that may result in the accumulation of a deborated water slug in the RCS loop, provide an analysis to demonstrate that recriticality will not occur as a result of the deborated water slug entering the core, either through natural circulation or by restarting the pump (s). The analysis should include an evaluation of the degree of mixing between the deborated water slug and the existing borated concentration, the reactivity insertion, and the total reactivity. Describe the methodologies used in the analysis.

c.

If recriticality occurs, provide an analysis of the consequence, such as whether the calculated peak fuel enthalpy (due to insertion of reactivity) has exceeded the limiting value i

of 280 calories per gram.

d.

What emergency operating procedures are there to prevent the restart of RC pump that could result in criticality during transients and accident events? What are other protective j

measures?

)

Response

Several different scenarios for all PWR designs have been postulated that could cause the accumulation of unborated water in the reactor coo /ent system (RCS) loops. The postulated scenarios addressed herein, which include those that are unique to the AP600 design, are:

The "Finnish Center" scenario, which is addressed by item (a.) below, with supplementary j

analytical discussions given in item (b.).

440.120-1 Revisfort 1

o NRC REQUEST FOR ADDITIONAL INFORMATION The introduction of relatively unborated water is possible as a result of reverse break flow e

following a steam generator tube rupture (SGTR). This scenario is discussed in items (b.)

and (d.) below.

it can be postulated that the actuation of the AP600 core makeup tanks (CMTs) could

=

potentially yield pockets of coolant that may not receive the higher concentration borated water. Thus, subsequent loop recovery under cold conditions may be a concern if the critical boron concentration for the temperature of interest is higher than the boron concentrations present in the stagnant regions of the loop (s). This postulated situation is discussed in items (b.) and (d.).

During a dilution to criticality, if a loss of power occurs, the subsequent e 0 ;e / standby s

diesel generators startup and loading would allow the charging /make-up pumps to continue the dilution without the reactor coolant pumps (RCPs) in operation, thus providing the means to accumulate unborated water in the RCS loop (s). This situation, which has been referred to as the " French" scenario in other studies / reports, is discussed in item (d.).

Various RCS maintenance procedures have the potential for low, or zero boron concentration a

water to accumulate in the RCS. This situation is also discussed in item (d.).

Thesa five scenarios are addressed with respect to the AP600 design below. This response has been structured as items (a.) through (d.), which correspond to the four parts of this RAI question.

(a.)

"Finnish Center" Scenario The AP600 is not subject to the "Finnish Center" scenario during small break loss of coolant occ/ dents (LOCAs) evease. The AP600 design possesses no loop seals entering the reactor coolant pumps (RCPs). Only a small amount of low boron content water could collect in the 8

bottom of the four RCP casings (approximately 21 ft per RCP casing; this equates to approximately 1.8% of the AP600 reactor downcomer and vesselintet plenum volume) and potentially be present in small break LOCA scenarios for sudden transport into the reactor vessel upon RCP restart.

The "Finnish Center" scenario as such is not re!:v:-! o/ concern to the AP600 because the steam generators are not cooling the RCS to petertS"y generate boron free condensate for any

!;"M:-! en extendediength of time during a LOCA event. During small break LOCAs in conventional PWRs, decay heat is removed through the steam generator safety valves forlong periods of time. In AP600, the pass /ve res/dualheat removalheat exchanger ( PRHR) rapidly becomes the dominant RCS heat sink following the generation of an "S" signal during postulated small break LOCA events. The steam generators become heat sources rather than heat sinks as the RCS depressurises. A potential means for generating a volume of unborated water during a small break LOCA is via operation of the PRHR. Steam condensed in the PRHR is delivered to the Loop 1 steam generator outlet plenum during small break LOCA events. However, with no 440.120 2 Revision 1 W westinghouse

j 1

o NRC REQUEST FOR ADDITIONAL INFORMATION sr i

~

s RCP loop seals the AP600 layout is such that the PRHR effluent will drain continuously from the steam generator channel head into the Loop 1 cold legs and flow into the reactor vessel. A deborated water slug cannot accumulate in the RCS loop cold legs. Within the reactor vessel the cold leg fluid entry point is above the direct vessel injection line elevation, which receives passive safety injection water from the cr r - 'ru; tr-t CMTs and/or accumulators with a high boron concentration which provides a significant reactivity margin to recriticality. E! cr *'r g _.. _ _ _ _ 7..

.g.; _. _ g _ _.. n. ___.g. : _ _..nj.q __.. g r., g:_ _ g. u:_ u _,.u.

.; u _ g.

i For the AP600 small break LOCA events, the dilute PRHR water must pass through and mix with anem4 hen approximately twelve cubic feet of borated water present in the downcomer prior to reaching the lower plenum and then the core. The relatively low flow rate of fluid from the downcomer into the core during the post RCP trip natura/ circuladon phase of AP600 small break LOCA events enables mixing to occur in the core and lower plenum. No unmixed " slugs" of highly dilute liquid from the PRHR are present in the downcomer to anter the core during LOCA design basis scenarios.

'- *'r 'M c-M - r c' r" b r 9 ^ "S^^ '.OC.^ r r-t e re ^.0S Mr cr * ? e er c,*'

"""" cerb r: :er; "*?e er '"r'r:r ec ^ ^ ?.' 20 '

r c'r r,tbc der cr r-berr-re cr-"rfr- "" et f

'-M Q :'h- '; ce tr re-*: urd """" r;r- "r-

"'r :rcr r cr A DS 'cu-t' M-

":r

ceuMrd, c ""u" er+::: ::r '- : re :-d ; edure r:r '---

rr--fr-Mr, rr r rM-*:-f f tr r-f'ut- ^^r -- "t' - r cr 1 in assessing the potential for boron dHution during LOCAs, an important considerethn is deWvery of boron from the CMTs (and the accumulators, when acdve) during the transient once they actuate on an "S" signal. The emNont boron concentration in the reactor vessel and RCS Increases signl6 candy from the In/dal value because both CMTs inject their total of 240000 lbm of water Inventory containing boron et the 3400 ppm (minimum) Technical Specl6 cation value t

into the RCS, wNch initleNy contains approximately 350000 Mwn of water. The minimalimpact l

of tNs PXS boron in}ection cn the RCS boron concentration occurs for the case in wNch the RCS is initia#y at its critical boron concentration for fudpower operation (approximately 1600 ppm, at beginning oflife). Under the conservative assumpdon that the break removes mass at the mixed everage baron concentration over the endre transient, the ambient boron concentration for tNs Mmiting case increases by 730 ppm when the CMTs have fumy injected their inventory.

The one-halfInch break case presentedin SSAR Chapter 15 represents a smeW break LOCA case sub}ect to posable boron dlhadon due to condensadon. In general, the smeNer the break sise of the postufeted LOCA, the longer the RCS remains elevatedin pressure and the steem generators and MiHR remain effeedve heet sinks. Condensadon occurs in both the PRHR andin the steem generators kving tNs break transient. During the inhtfal 11500 seconds of the transient, the natural cheuladon Mow through the steem generators and PRHRis essentleNy singk phase huld. No mechanism exists for any pockets of dHute Neuid to occur during tWs natural circulation period before the steem generators drain.

440.12o s Revision 1 9

NRC REQUEST FOR ADDITIONAL INFORMATION Tit y

The non-PRHR-loop steam generator continues as a heat sink while it drains between 11500 and 13700 secords, Condensate formedin the tubes In this time period flows back into the vessel upper plem;m as part of the liquid backflow associated with the steem generator draining. Of the total 24000 lbm of non-PRHR-loop steam generator drain Hquid flow into the reactor vessel upper plen:tm dudng this pedod, approximately 4000 lbm are condensate. The condensate is weH mixed i. the hot leg will' the Hquid accompanying it, and no pockets of dilute liquid occur.

The upper plerwm is fuH to the hot leg level and contains approximately 30000 lbm in this time interval. Since the ambient boron concentration has increased due to CMTInjection, the liquid in the upper plenum exhibits a higher than initialboron concentration event with the ongoing dHution of the totalmass in the upperplenum. Furthermore, during this steam generator drain pedod, a natural circulation tiow of almost 200 lbm/sec directs mass from the upper plenum into the PRHRloop hot leg, and flow is from the core into the upper plenum. Therefore, boron dilution due to the 4000 lbm condensate backflow is not of concern.

Dunng the initial 11000 seconds of the transient, the PRHR receives essentlaHy single phase Hquid from the hot leg. The average quality of fluid entering the PRHRis less than 1% during this time Interval. From 11000 - 17000 seconds a significant amount of condensation occurs in the PRHR, and a smsHer amount occurs in the PRHR-loop steam generator. Dudng this interval, 20000 lbm of steam present in the two-phase mixture circulating into the PRHR is condensed there. A large liquid flow rate is predictedin this time pedod from the upper plenum through the PRHR, so the condensate is well-mixed with liquid at the ambient boron concentration in the PRHR retum line and cold legs.

Between 11500 and 13700 seconds, approximately 5000 lbm of steam condenses in the PRHR4oop steam generator tubes and drains into the hot leg together with steam generator liquid. The fluid from steam generator draining enters the PRHR along with the fluid from the vesselupper plenum. After 13700 seconds the steam generators become heat sources, and no further condensation occurs there. Overall, the total mass flow through the PRHR in this interval is greater than 430000 lbm. Since the total condensate combined due to PRHR and steam generator heat transfer in the Intervalis about 35000 lbm, less than 10% of the PRHR outlet flow is condensate. That portion which is condensate is weH-mixed within the circulating flow retuming to the coldleg. Moreover, the amblent boron concentration in the reactor vessel and RCS is higher than the initial value because both of the core makeup tanks have injected their entire inventory of boron at the 3400 (minimum) ppm Technical Speci6 cation value. The increase in baron content from the injection of CMT boron to en RCS which initially contains Hquid at a maximum boron concentration of 1600 ppm means that the 40000 lbm of condensate produced does not cause a decrease in the RCS boron content below the inittel value. After ADS actuation at 16600 seconds, the PRHR is no longer effective in condensing steam, and no mechanism for further boron dilution occurs. In fact, RCS baron concentration is increased by actuation of the accumulators shortly after ADS actuation.

The smsM break LOCA cases presented in SSAR Chapter 15.6 do not model the PRHR once automatic depressudiation system (ADS) stage 4 is active. This approach, which is consistent 440.120-4 Revision 1 W westinghouse

NRC REQUEST FOR ADDITIONAL INFORMATION

=.

E s

with the NOTRUMP validation against AP600 Integral test results, conservatively eliminates a depressudiation source. To identify possible boron dilution effects, a two-inch coldleg break case in which the PRHRis modeled throughout the transient was examined. This case exhibits more condensation in the PRHR than occurs in the one-halfInch break SSAR case described above (53500 lbm). Due to the RCS depressudzation, the steam generators become heat sources rather than heat sinks within 500 seconds of transient time. Of the totalPRHR condensation, 35000 lbon forms before accumulator injection, and 14500 lbm more condensate forms before the actuation of ADS stage 4 at 2200 seconds. Until this time, the PRHR flow is two-phase natural circulation. The liquid flow entering the PRHR exceeds the condensation amount through 1800 seconds, so no pockets of dilute boron liquid wiH form in the piping.

As previously noted, infection of borated water from the CMTs increases the ambient boron level above its initial value. A time of minimum baron in the RCS isjust before the start of accumulatorinjection, when the dHute PRHR return stream can be accommodated by partial CMTInjection alone. The CMTinjection at the time of accumulatorinjection is enough to increase the emblent baron concentration by at least 540 ppm, which is adequate to maintain the RCS boron level above the initial concentration. The accumulators rapidly inject borated water into the RCS, increasing the downcomer, core, and RCS boron concentrations. By the time that the condensate iness dominates the PRHR return flow stream, both the accumuistors have injected completely, raising downcomer baron concentration high enough to accommodate the 7000lbm condensate thatis delivered from the PRHR after 1800 seconds and before ADS Stage 4 actuation.

By 2200 seconds into the transient, the accumulators and CMTs have delivered theirInventades of borated water to the RCS, and the baron content in the downcomer is approximately 900 ppm above the maximum initial concentration. The 3000 lbm of PRHR condensate which enters i

the downcomer prior to the start ofin containment refueling water storage tank (IRWST)

Injection at a rate aversging 3 lbm/sec compdses less than 10% of the totallower plenum /downcomer mass inventory. Furthermore, dudng the Intervalbetween the CMT empty time at 2460 seconds and the start ofIRWSTInjection at 3400 seconds, only 20% of the downcomer mass inventory passes from the downcomer node into the lower plenum.

Therefore, no pockets of dilute liqu/d form. When the IRWST becomes active, it provides highly borated water to the vessel downcomer. Therefore, the dilution associated with the PRHR conde setlon during ADS stage 4 operation is insignificant. Based on the examination of the above cases, no boron dHutton occurs within the core for any AP600 postulated I.OCA scenado.

(b.)

Transients or Accidents Addressed by Analysis The safety-related method for decay heat removal for the AP600 consists of heat transfer to the IRWST by the PRHR, and borated makeup water addition to the RCS from the CMTs. Operation of the CMTs require that the RCPs are tripped. As the residual heat from the core is removed by the PRHR and CMTs, boric acid is added to the RCS by CMT injection flow. The RCS flow associated with the operation of the PRHR and CMT systems is caused by the thermal driving j

440.120-5 Revision 1 I

\\

NRC REQUEST FOR ADDITIONAL INFORMATION l

heat established by the convective heat transfer. Analyses have been performed (Reference 440.120 2) to investigate the flow behavior throughout the RCS while the PRHR and CMT systems are removing core decay heat, M c~'cr to quantify the resulting boron distributions that could form as convective flow rates approach stagnation. For this study a loss of normal feedwater (LONF) transient was chosen.

The Reference 440.120-2 analysis effort utilized the TRAC-PF1/ MOD 2 code to perform transients that are very similar to the design basis LONF transient thatis presentedin the SSAR Section 15.2.67 (Reference 440.120-3) mc'c ence 2.). A description of the AP600 TRAC-PF-1 thermal /hydrauMc and neutronic models is presented in Reference 440.120-2 Sections 3.1 and 3.2, respectively. Conditions corresponding to beginning of life, equilibrium cycle, no xenon were assumed, as this would be the most limiting plant conditions in the event core recriticality were predicted. Benchmarking between the TRAC-PF1 code with the SSAR data, which is based upon output from the Westinghouse LOFTRAN-AP code, indicated good agreement. A detailed discussion of the thermal /hydrauHc compedson between the TRAC-PF1 calculations and the SSAR data (f.e., reference LONF)is presentedin Reference 440.120-2 Section 4.3.1 i

(pages 4-26 through 4-47). An acceptable comparison of the neutronic model was obtained with Westinghouse reterence core data. Speel6cally, the TRAC-PF1 calculatedpower distdbutions and rod worth values agreed within 6 to 7% of reference Westinghouse calculations. This degree of agreement for the neutronic modelis acceptable given that this study focused on the mixing aspects of baron in the AP600 design and a detaHed neutronic response as e result of a boron dilution was not necessary. A retum to cdt/cnHty was not chsNenged; thus a high level of agreement with the reference Westinghouse data is not necessary. Furthermore, the TRAC-PF1 calculated reactMty was normalized to the reference Westinghouse data, as discussed on Reference 440.120-2 pages 5-48 and 5-49.

The results (see Reference 440.120-2 Section 5.1) of the loss of normal feedwater transients indicate that all regions of the RCS become sufficiently borated following RCP trip and CMT actuation as a result of RCS flow remaining high enough in all regions of the AP600 primary side system for a sufficient duration. The affects of reduced decay heat were also included in the analysis (Reference 440.120 2 Section 5.2/. The low decay heat analysis arbitrarily assumed 1% of the ANS 1979 decay heat curve. Reduced heat generation in the core results in the passive cooling systems to lose their thermal driving head earlier in the transient, thereby providing a shorter duration for the CMTs to inject the higher concentration boron into the RCS.

The results demonstrate that boron concentrations throughout the RCS were greater than the critical boron concentration required for cold (2OO'F) N-1 rods inserted (most reactive RCCA assumed to be stuck out of the core), no Xenon conditions. Therefore, it can be concluded that subsequent RCS loop recovery, following CMT actuation and RCS cooldown to equilibrium temperatures, will not pose a recriticality potential.

Additional analysis were performed as part of the Reference 440.120-2 study to quantify the volume of unborated water that could collect in the RCP casings and steam generator channel head without resulting in localized core inlet boron concentrations to decreast to the critied 440.120-6 Revision 1 W Westinghouse

9 NRC REQUEST FOR ADDITIONAL INFORMATION Mh boron concentration following the restart of the RCPs. These additionalanalyses are discussed in Reference 440.120-2 Section 5.3. The affects of nominal and reduced decay heat situations were also considered. The initial conditions for these investigations were obtained from the pseudo-equilibrium conditions (i.e., transient times > 4000 seconds) for the loss of normal feedwater transients discussed previously.

The findings of this unborated water investigation can be directly applied to the SGTR reverse break flow scenario and also supplement the previously discussed "Finnish Center" scenario, (which is the subject of part (e.) of th/s responsel, as discussed below.

A high order solute tracker, which is described extensively in Reference 440.120-4 (and is also included as Appendix D of Reference 440.120-2), and discussed to a lesser degree in Section 2.2 of Reference 440.120-2,120 2". was employed to significantly reduce numerical diffusion. This high order solute tracking method employed for the unborated slug investigation has been benchmarked against experimental mixing data from a 115 scale model of a three-loop Westinghouse PWR, c4-dbcmcd 'ur*"c " The benchmark against the experimentaldata is describedin Reference 440.120-2 Section 4.2.3. The results of the comparison between the TRAC-PF1 high order solute tracker with the experimental data clearly demonstrate that the high order method is conservatively under predicting the mixing that would occur, as indicated by the experimental mixing data. This is primarily due to the fact that the high order solute tracker calculations do not account for the mixing that results 4m from the impinging jet of coolant onto the downcomer walls of the reactor vessel. As such, the application of the high order solute tracker to the mixing transient calculations discussed below have has significant conservatism inherent in the results. Furthermore, the mixing that would occur from the highly turbulent flow caused by the RCP impellers has not been credited. Thus, larger volumes of unborated coolant could be shown to be acceptable if the mixing that would occur from these ignored effects (i.e., inlet coolant jet impingement on the downcomer and RCP impellers), were j

explicitly modeled.

This high order solute tracking scheme was not employed for the previously discussed loss of normal feedwater transients, as the natural convection flow tends to distribute the boron being Infected by the CMTs quite rapidly. This eliminates sharp fronts in the baron concentretion and results in a steadily rising system baron concentration in a uniform wey. Thus, numerical diffusion plays a very smallrole,cc '": bcrc~ *renc;c-* :cc dete"~cd tc be ~c! "; cc~:cct":c, cnd nurcr!ce' d!!!u& p!r/c c ' cr/ cm!! ic!c, if any, in driving the solute distribution within the system. As such, the runs not modeling unborated slugs of coolant were not repeated with the high order solute transport methods, since the expected results would be basi: ally the same.

The results of this unborated slug analysis, where the RCPs were started in the loop containing the unborated water (see Reference 440.120-2 Section J.3.f/, yielded unborated volumes i

greater than 115 ft for the situation where nominal decay heat had been assumed, and 8

8 J

unborated volumes greater than 66 ft for the situation where the decay heat had been assumed 8

to be 1% e' tM ANS 1979 curve. In contrast, one RCP casing can collect less than 21 ft before be m ped to the cold leg connection to the RCP casing. In the absence of cold leg lov seal op 'g, volurnes of unborated water larger than 21 ft per RCP casing, would begin to 8

440.120 7 Revision 1

.4 NRC REQUEST FOR ADDITIONAL INFORMATION di spillinto the cold leg piping to be mixed with the borated coolant in the RCS before reaching the reactor vessel. Thus, the maximum volume of unborated water that could collect in a steam generator channel head region cannot be greater than 42 ft'(i.e., two RCPs per steam generator j

outlet channel head; this equates to approximately 3.5% of the AP600 reactor vessel inlet i

plenum volume). The analysis results presented above indicate that approximately one and one-half times this credible value can be accommodated (i.e., this volume can theo'retically accumulate and not result in the core inlet boron concentration dropping below the critical concentration following RCP restart in the affected adjacent loops) under low decay heat conditions, and more than two and one-half times as much under nominal decay heat conditions.

Unborated slug analyses were also performed assuming that the unborated slug of coolant existed in one loop, and the RCPs were restarted in the opposite loop, as describedin Reference 440.120-2 Sect /on 5.3.2. The findings from this set of analyses is directly applicable to SGTR recovery, as the recovery procedures regarding RCP restart will identify that the RCPs in the intact RCS loop must be restarted first. This analysis demonstrated that the resulting mixing due to the reverse flow through the faulted steam generator and associated RCS loop can accommodate e&~r'/ arge volumes of unborated water in the faulted steam l

generator U tubes and channel head and localized core inlet boron concentrations remain we4 above the critical boron concentration.

(c,)

Recriticality has not been predicted for the rapid boron dilution mechanisms / scenarios addressed herein.

(d.)

Protective Measures; EOPs, Others The Emergency Operating Procedures (EOPs) will be written by the combined license applicant.

Westinghouse provides input to the EOPs through the Emergency Response Guidelines (ERGS).

The ERGS stipulate that prior to restarting RCPs, there must be indication of subcooling based upon core exit thermocouple readings, and indication of pressurizer level. These two conditions allow for single phase natural circulation. The results of the analyses discussed in part (b.) of j

this response demonstrate that adequate mixing of the boron injected from the CMTs occurs under natural circulation conditions, even assuming a low level of decay heat.

RCP re start is specifically addressed for the steam generator tube rupture accident by ERG AE-3 (Reference 440.120-5). Steps include a note of caution regarding the potential of inadvertent criticality following any natural circulation or PRHR cooldown if the first RCP started is in the 09 reduced when the first RCPs restarted are those in ruptured loop. This potentialis 0M:

/

the intact loop, which is supported analytically as noted previously in section (b.) of this response.

Regarding the postulated loss of AC power during a dilution to criticality (also referred to as the

" French" scenario), it is assumed that the emergency diesel generators startup and provide 440.120 8 RevislOn 1 W Westinghouse i

l

o NRC REQUEST FOR ADDITIONAL INFORMATION

@E iil power for the CVCS pumps. The addition of unborated makeup water to the RCS would continue without the RCPs in operation, thereby providing the means to accumulate unborated water in the RCS loop (s). "c r /::

/t should be noted that the AP600 design includes a Battery Charger input Voltage Low signal which causes the Memineralized water supply isolation valves to close and aligns the SAI boron acid tank to the makeup pumps. Therefore, this postulated scenario is not a concern with respect to the AP600 design, as logic exists to prevent such an occurrence.

Concerns of RCP restart following maintenance that has a potential for the formation of low, or zero boton concentration water to accumulate in the RCS, are recommended to be addressed procedurally, for any PWR design, as discussed in Reference 440.120 6. The means to prevent such a maintenance initiated scenario is that steps be included as part of the maintenance procedures to remove / mix this low, or unborated water volume. Measures include, but are not limited to, verifying that sufficient mixing will be present outside of the vessel, using feed and bleed, or drain and fill of the affected area (s).

Regarding other protective measures, there will be interlocks in the logic controlling the RCP power supply. These interlocks, together with the AP600 ERGS /EOPs, will preclude the inadvertent restart of the RCPs following the actuation of the passive core cooling systems.

Conclusions it has been demonstrated that after a loss of heat sink event, following CMT actuation and RCS cooldown to equilibrium temperatures, RCS loop recovery will not pose a recriticality potential, as the boron addition from the CMTs becomes sufficiently mixed with the entire RCS. It has also been shown that unborated water accumulated upstream of idle RCPs, as is postulated during the "Finnish Center" scenario, to volumes approximately one and one-half times that physica!!y possible to stagnate in the AP600 RCP casings of one of the steam generators, will not result in recriticality following RCP restart. Furthermore, conservatively calculated unborated volumes greater than the entire primary side of a steam generator can be accommodated without recriticality concerns under reverse RCS loop flow circumstances (i.e., the RCPs are restarted in the loop opposite from that containing the unborated coolant). The exceptional mixing that occurs in the RCS loops under the reverse flow configuration is utilized procedurally for those instances where it will be apparent to the operator where the low, or unborated water volume may exist (e.g., the Mr

ener"c' ide ruptu'e SGTR ERGS). Therefore, the AP600 design has shown that substantial boron dilution can occur, however unlikely, without leading to recriticality. Even though analyses indicate recriticality would not occur, additional steps have been prescribed to minimize boron dilution potential, thereby maintaining a " defense in depth."

440.120 9 Revision 1

s NRC REOUEST FOR ADDITIONAL INFORMATiGM J

References 440.120--1 Andreychek, T. S., et al., "AP600 Low-Pressure Integral Systems Test at Oregon State University Test Analysis Report," WCAP-14292, Revision 1, September 1995.

440.120-2 Macian, R., K. Ivanov, and G. E. Robinson, " Analysis of Boron Dilution Transients in the AP600", The Pennsylvania State University, Nuclear Engineering Department, June 1996.

440.120-3 Simplified Passive Advanced Light Water Reactor Plant Program, AP600 Standard Analysis Report, Section 15.2.7, " Loss of Normal Feedwater Flow," Revision 5, February 29,1996.

440.120-4 Macian, R., and John H. Mahaffy, "High Order Numerical Modeling of Solute Transport in System Codes", The Pennsylvania State University, Nuclear Engineering Department, September 1995.

440.120-5 " Revision 1 of the AP600 Emergency Response Guidelines," NTD-NRC 95-4525 (DCP/NRCO376), 009e* Me. ?" S2 ^^?. August 9,1995.

440.120 6 Burnett, Toby, et al., " Risk of PWR Inadvertent Criticality During Shutdown and Refueling," NSAC-183, Westinghouse Electric Corporation, Electric Power Research Institute, December 1992.

SSAR Revision: None l

1 l

440.12410 Revision 1 l

W westinghouse