ML20091A634

Errata to Evaluation of ATWS for Monticello Nuclear Generating Plant

ML20091A634
Person / Time
Site:	Monticello
Issue date:	10/31/1976
From:	GENERAL ELECTRIC CO.
To:
Shared Package
ML20091A631	List: ML20091A627 ML20091A634
References
NEDO-25016-ERR, NUDOCS 9105160421
Download: ML20091A634 (24)

Text

. - ._. .- .- . _ _ - . . _ _ - _.

4 NUCLEAR ENERQY OlYtSIONS O GENER AL ELECTGIC COMPANY s

• -- G ,-  :*  ;*

• t ' 5AN JOSE, CALIPORNIA HlN f \ ' G,.. ..

~3 ;,  ;,

E G E N E k AL $ .ELECTRIC .. .

[. , ,f..,

,c .

)) l 3 4 l u .< ,.c'

." J . ,_. ,

t ,

arrucasts rs ERRA.7TA ,'A.n.d _AD.D E N D_ A- . .. , .

PusLICATION No. NN#M

\ b AEj ;. ;l. W i

7hg Evaluation of Anticinatad' No,* 'S ' '

1 Tr'maien t s Wi tho"t E"Y** f * * "h * ' D ATg ' Oc tobe r" 1976 . :t .

Monticello Nuclear Generating Plant NOTE" Correct ed cor . tes of sie,'attKeeble ,.

- v 'l

s. *"y  %' :c,

• paHention'as speclMed below;qh ISSUE DATE **"*"'" ;L.

h m- ,y 4

-*. w

~ ""

R EF ERENCES , INSTRUCTIONS '

ITEM ta s cTion, p ae t.

(commscrious Awo A00stions) - 4. -

panaenapw. u w ei aw >r

" M "i y,; .'

Af fix the attached new ,pages over the corresponding.

( e

'u 4 s page.s, of your doctsnent I

s s-eu s 3-1' . , . ,

F5: '

39 y

44 3,33; ,2

. y W 3 14 a ~;i ,,g , ;. . ,

,,- N 3-18 v

3

• .' 3-21b 6  ; r ,

' JA , Q.,

3-22 eG. +

x' ,

'N '

t *C 3-26 .

,x w, , -

4 -

~ 'p; ; + p. - f.

u i 4 1

,. c.,3 ^ S -c

, 44 A.%.

s Aa d, s 4-24' '

r s , .n '

. , .:e7 4-3 , s

. W.

~~ .

4-5+ -

M 4-6. ,

• J s.i s . n..

..* 4 - 17 .. ,

C

~, _

p. . .,  % .-

4-18 5-1' @ '

' ""'; ' N' N -*

M"

, A-2 .J. . Q.' , 'g.; ,,gg A-5 '.M, , . .- ,;Q', ap, . y l

_ &yL, l

• - y- .

A

.. .a. ,* , ,

4' A-7^ ,m. w.

a4 .,4..;2, . ,r</v n,.' w

,s,$,. .

4 e4

- ag p. . ,

- .m" .

• no#

e,, go S i.', f.1 u

,w w;

, -i g+ .,' .

t- g_jg. ,

t.

.~. ,8 9 4.g, %m , . 4 >

. A-19' -

. cut

, y-

-t p "u% .',.

v.sm fi, ,4 i_' '

',:(W

' N 4, s

1, &

,,g. E e*L' 4 g gc

, ., p ; , ' ,',

sd 4 s i '

v 3- ,4,  :.tw Qjh h

• y + '; % ' .r .

. + :. . , , , ,

,vg.

, i, .. v. <s.. .g A ,

, %g(ma;4.g, . ,g.jgg,w

,,g,m ;g I

,9o -

,:j7e. < > g. , ,

,, w ,

e8, s b :7 y 'h

. i;; g . W *& ' I;O Ml } . .f

% , 1 '., ,n.  :) ;

b:3(, ,, n:

a .) g ay 'eL6

}-

,

• 4 e,<

.p.f,, y

.sy,

,.'g~,, p, .

{ ., ;.

9105160421 761116 s (

; PDR ADOCK 0S000263 3 0.ij , ,, . ,,.j . ~gl g.

f r

P PDR ~r-

, ee>-

o n f ' s~ A I .,2 r - . ,

. o\ y",t

1. 1

.y- . ' , , f ;/

AAGE

t

..

• SEDO-25016 4

3. EVALUATION OF EVENTS The following discussions summarize perf ormance of the plant during many transients events, including consideration in each case for the postulated

- failure of normal scram, and the effect of recirculation pump trip, ATWS rod

('-

injection (ARI), and operator actions. The previous reports have concentrated i

upon the most frequent (once in four years) cases with special attention to

+ the closure of all main steam isolation valves as a bounding case. That case

- still remains the most severe event from most viewpoints, and still receives the most parametric attention in this study. However, the scope of the events has been extended as requested by the NRC to cover all significant events expected at least once within forty years of the plant operation.

3.1 CLOSURE OF ALL MAIN STEAM ISOLATION VALVES 3.1.1 Basic Event Description i

Automatic circuitry or operator action can initiate closure of the main steam isolation valves (MSIV). Normally, scram is initiated by position switches on the valves before they have traveled more than 10% from the open position, l Subsequent scram signals would be' initiated (if needed) from high neutron flux and high vessel pressure. The normal event displays very little, if any, neutron flux increase before shutdown is ef f ective. An abrupt vessel pressure rise occurs when the MSIV's close, lif ting the Saf ety/ Relief (S/R) valves for several seconds in their pilot actuated relief mode. Pressure is easily limited

{

below the design pressure of the primary system. Long term heat removal and inventory supply are provided by one of the S/R valves, the llPCI or the RCIC t

i system, and the RRR cooling capability (as long as necessary until the normal heat sink, the main condenser, can be utilized),

i 3.1.2 Response of Plant in its Present Configuration Figure 3-1 shows the reactor vessel pressure and neutron flux as a function of tire for the MSIV closure when credit for scram is not taken and when the postulated modifications are not assumed. As the MSIV's close the reactor 1-1

I ,. ',

NEDO-25016 t '

t 250 MONT.MSiv 74,4% Rt LIEF CAPACITY 1 NEUTRON FLUK 2 AVE SURF ACE HEAT FLUX 5 i50 -

2

! E 2 o

x 1 3 100 50 m

0 0 4 8 12 16 20 24 TIME hect Figure.3-3. MSIV Closure Transient With 2-RPT Modification t

effects neutronic shutdown, the reactor power is only due to decay heat.

Reactor core cooling is adequately provided by the HPCI and RCIC systems, the heat being removed by relief valve flow into the suppression pool. Cold snut-down is achieved by normal operator actions. The peak bulk suppression pool temperature reached is 153.2*F with 6 valves (153. '.*F with 8 valves) and is below the guideline value of Section 1.1. The containment pressure peaks at 7.6 pcig r

l

! (with 6 valves) and 7.4 psig (with 8 valves). Both of these peak values are l'

well below the guideline value.

! 3.2 LOSS OF NORMAL AC POWER l

I 3.2.1 Basic Event Description l

Loss of normal AC auxiliary power de-energizes all busses that supply power to the unit's auxiliary equipment such as the recirculation pumps, condensate l 3-5

E NEDO-25016 1 .

pumps and circulating water pumps. Coastdown of all pumps occurs and condenser vacuum is gradually lest. Turbine-generator trip occurs at the start if this is a general grid disturbance. Scram is normally activated by any of a number of signals (T-C trip, low vacuum, low water level) and MSIV closure occurs (low water level, low vacuum). Momentary opening of the S/R valves occurs to easily limit the pressure rise in the vessel. Long term heat removal and inventory supplies are provided by one of the S/R valves, the HPCI or RCIC system and RHR coaling capabilities.

3.2.2 Response of Plant in its Present Configuration When auxiliary power is lost, all pumps (circulating water pumps, f eedpumps and recirculation pumps) coast down immediately. The reduction in core flow begins to reduce the reactor power. The MSIV's close due to loss of condenser vacuum and contribute to the pressurization of the reactor. The short term response of the reactor is shown in Figure 3-7. Short term response is much less severe than the MSIV ATWS with RFI modification for the following reasons:

(1) The recirulation pumps trip at time zero rather than wait for the reactor pressure to reach the ATWS setpoint. (2) The f eedwater pucips are tripped at time zero which result in a lower core flow and lower core inlet subcooling, and thereby lower power without normal scram. The peak reactor pressure reached is 1194 psig (with 6 relief valves) which remains under the guideline number.

The peak fuel enthalpy reached is < 150 cal /ge, also below its guide.

In the long term, with the feed pumps lost, the reactor water level begins to drop. At the low low level, the KPCI and RCIC systems are initiated. However, l as the reactor core power is still higher than the combined flow capac1ty of i

IECI and RCIC, the reactor water level continues to drop. The combination of reactor power, core flow and water level eventually reached before the standby liquid control system can be assumed to be effective is such that the analytical model is not capable of continuing to predict the results of the event. Therefore, the results do not show reactor behavior during the time before manual neutronic shutdown becomes ef f ective and normal reactor water inventory is restored. If initiation of SLC takes place at 5 minutes after the reactor high pressure is reached, hot shut down would be achieved at about 20 minutes (with the normal configuration of only one SLC pump working).

3-9

.._.. _ _ .--. _ . _ _ _ - _ _ _ _ - _ ~ . __. . _ _ _ _ _ _ _ _ _ _ _ _ _ - . . _ _ _ _ .

t NED0-25016 In order to calculate the containment response, the initial relief valve flow predicted by the model is extrapolated through hot shut down. Assuming this steam dump into the containment, the containment pressure and suppression

. pool bulk temperature peaks were caltulated. The peaks reached are 21.8 psig i and 214'T respectively.

The results of this transient, if the unlikely ATWS event is assumed to occur, are uncertain because of analytical limitations and, since the suppression pool temperature is above the guideline, the results are considered to be unsatisfactory without the postulated codifications.

3.2.3 Plant Response with the Postulated Modifications The short term response of this case is che same as the As-Built Case. The reactor pressure peak reached is 1194 psig (with 6 relief valves). The peak fuel enthalpy reached is < 150 cal /gm. When the dome pressure reaches 1150 psig i and the ATWS logic initiates the ATWS Rod Injection (ARI).

The reactor vill be in hot shutdown near 20 seconds by ARI, and combined flow of the RCIC and RPCI systems easily restore water level to the. normal range.

The cold shutdown condition can be reached by performing the normal manual actions. Contain=ent response with ARI function is shown in Figures 3-8 and 3-9. The containment pressure peak reached is 3.6 psig at 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> and the suppression pool bulk temperature peak reached is 153.4'T at 11 hours1.273148e-4 days <br />0.00306 hours <br />1.818783e-5 weeks <br />4.1855e-6 months <br />. Both of theFa are below their respective guideline values.

i g 3.3 LOSS OF NORMAL FEEDWATER FLOW k

3.3.1 Basic Event Description Inadvertant trip of all the feedwater pumps or water level controller failure (zero demand) have been postulated as potential causes of loss of all normal ,

l _

j feedwater flow to the vessel. Loss of auxiliary power also causes this event I as described above. Reactor core flow is reduced when the feedwater flow reduction occurs, dropping power gradually until it is totally shut down in ,

the normal case when scram is initiated from low water level. Continued 4

5 3-11

_ .~

SEDO-25016 gradual inventory loss occurs until isolation is initiated and the RCIC/RPCI systems are brouC h t on automatically to maintain proper water level to the conclusion of the event.

1 3.3.2 Response of Plant in its Present Configuration When the feedwater pumps are lost, the reactor water level drops to the Icw low level in a few seconds, causing the MSIV's to close, the recirculation pumps to trip and the HPCI, RCIC and RRR systems to initiate. From this point the short term response of the plant will be similar to that of the MSIV closure but milder. The peak vessel pressure and fuel enthalpy reached are l less than the re.pective guideline values of Section 1.1. The long term i

f r#~' response of the reactor is not analyced as it would be very similar to the case of loss of normal AC power transient.

3.3.3 Plant Response With the Postulated Modifications t

When the reactor water level reaches the low low level, the ATWS logic initiates ARI. Hot shutdown is achieved near 20 seconds. The minimum water level reached in this case would be about 1.5 to 2.0 ft below the low low level. The peak bulk temperature reached in the pool would be slightly less than the value for the MSIV closure (Viz. 153.4*F). This is because of the smaller power burst due to feedwater trip at the beginning and the recirculation pump trip being simul-taneous with the MSIV closure. Cold shutdown condition can be achieved in the same manner as MSIV closure transient; by normal manual actions.

I 1 ,

u 3.4 TURBINE-GENERATOR TRIP l

1 1

l 3.4.1 Basic Event Description I'

Less of generator electrical load initiates fast clesure of the turbine control 1

valves to provide overspeed protection for the unit. A variety of equipnent protection signals can lead to trip of the turbine stop valves directly. Both l types of events are very similar from the reactor point of view. Normally, l scram is initiated almost simultaneously with the start of fast valve closure.

Inherent control logic opens the steam bypass valves directing some steam l 3-14 1

. -.- - - - . . - - - _ _ . _ - - - . - . ~ . - - . - - - - -

NEDO-25016 I l

1 s

3.5.3. Plant Response With Postulated Modifications )

The closure of stop valves causes the vessel pressure to increase (Figure 3-11) resulting in an ATWS signal. When the dome pressure reaches 1150 psig, the recirculation pumps are tripped and the ARI is initiated. The tripping of the recirculation pumps causes an immediate reduction in power level. The peak vessel pressure reached is 1248 psig. The rod injection continues to reduce power to hot shut down. Further normal manual actions will bring the reactor i

to cold shutdown conditions. Heat is removed via the relief valves to the suppression pool. The RRR's are manually initiated at 10 minutes to remove excess heat from the pool. The pool reaches a peak bulk temperature of 153.3*F l and the containment reaches a peak pressure of 7.3 psig in about 11 hours1.273148e-4 days <br />0.00306 hours <br />1.818783e-5 weeks <br />4.1855e-6 months <br />, j which are within the guides of Section 1.1, 3.6 LOSS OF A FEEDWATER HEATER 3.6.1 Basic Event Description i

Historically, this event has been addressed in most SAR analyses even though on many plants no direct means of eliminating the action of a feedvater heater is provided, it is included here at NRC request, covering whatever chance of extraction steam loss or other heater interruption that could occur. Individual heaters (or linked groups) have traditionally been bounded by analyses with 100'F changes in feedwater temperature. Conventional designs for most plants including Monticello really limit that change to about 70'F or less. The expected plant behavior, should this event happen at full power conditions, is therefore a gradual power increase toward a new value consistent with the colder core inlet conditions. The change is gradual because of the thermal capacity of the heater and the mixing characteristics of the rc setor downcomer y and lower plenum. If the plant happened to be in automatic load control, the core flow would be reduced in such a way that steam flow to the turbine would be held essentially constant (although neutron flux would rise slightly above the initial value). In base-loaded, manual operation of the plant, power 1 increases somewhat more (without compensation by the flow control) and the bypass valves would be opened slightly, if needed to pass the excess steam.

3-18 3 - - - - - - . - - - - - . - - - - - _ - - . , . ,y. ,,,,---,,.-w,,,,- , - - , - . - - - . - ~ - , , .

, t

?

NEDO-25016 10 psi and et c.ize; valve steadily blows 164 lbs/see of steam into the pool.

The guides for suppression pool sempetaLure and containment pressure shown in Section 1.1 would eventually be exceeded if scram could not be initiated  ;

and there were no modifications or manual corrective actions taken.

3.7.3 Plant Response With Postulated Modifications Should the normal scram not take place, the manual initiation of recirculation

! pump trip and AR1 in 5 minutes af ter receiving the high torus temperature alarms would reduce the power and shut down the reactor. The high temperature alarms are initiated when the pool temperature reaches 110*F. The bulk pool l tempcrature reached in this case is 120*F at hot shutdown, well below the l guideline value. The long term pool temperature response is discussed in the appendix in answer to Questions B.4, 5 and 6.

3.8 LOSS OF CONDENSER VACUUM 3.8.1 Basic Event Description The reduction or loss of vacuum in the main turbine condenser can be caused by loss of cooling water pumps or inef f ectual operation of the vacuum support

- equipment, it sequentially trips the turbine stop valves closed (which normally scrams the reactor) and, if the event is severe enough and the reduction of flow f rom the turbine still is not enough to help condenser performance, the steam bypass valves are closed. These actions would occur normally over a period l of several minutes or at worst, 20-30 seconds. The initial part of the event is ,

the same as a turbine-generator trip since all systems funtcion in the same way

' as they do in that event. The long term behavior.of the event is similar to any isolation unless enough vacuum can be maintained to preserve bypass flow l

thereby permitting decay heat removal through the condenser instead of relying upon the pool and the shutdown cooling systems.

3.8.2 Response of Plant in its Present Configuration I This case begins identically to the T-G trip and then the bypass valves would ,

i close. The response without the postulated modifications is less severe than MSIV closure. The peak vessel pressure would exceed the guide of Section 1.1.

3-21

._ _ ._ . . _ . _ . . _ - . __.--_m . _ . _ _ _ _ - , . _ _ _ . _ _ _ _ _ - _ _ _ _ _

NEDO-25016 i

3.8.3 Plant Response With the Postulated Modifications The event results in short term peak values less severe than the MSIV case.

The All the AWS logic is activated quickly by the high pressure transient.

in j longer term nature of this case (assuming vacuum continues to deteriorate) converted to a nearly nomal isolation by action of the AWS rod injection which completes the nuclear shutdown less than 20 seconds into the event.

3.9 FEEDWATER FAILURE - MAXIML'M DEMAND 3.9.1 -

Basic Event Description A postulated failure of the feedwater/ water level controls in the direction of maximum demand results in a moderator temperature and void fraction decrease causing a reactor power' increase at the same time water level increases toward high level protection. The feedwater pumps are tripped as well as the main turbine when level reaches the high trip setpoint. Scram nomally occurs with the turbine stop valve closure, . limiting any further power increase in such a The resulting pressure way that satisfactory thermal margins are maintained.

rise is controlled by the turbine bypass (throughout) and S/R valves (momentarily).

Final aspects of the event are similar to the loss of nomal feedwater since RCIC/HPCI system initiation eventually are expected to occur.

3.9.2 Response of Plant in its Present Configuration When the failure of the feedvater in the direction of maximum demand occurs, the high level turbine trip (with bypass) and feedwater pump trip will occur near 143 seconds. The feedwater pump trip results in the water level dropping So, to the low low level causing the MSIV closure and recirculation pump trip.

without the postulated modifications the short term response is similar to the g

! MSIV clo1ure but milder. The long term response is similar to the loss of AC power event.

3-22

. _ _ _ __ _ _ - -. _ _ _ _ . _ _ _ . . . - . - . _ _ _ _ _ , ~ . _ _

EDO-25016 3.10.2 Plant Response k'ith the Postulated Modifications - Not Applicable 3.11 St*MMARY OF TRANSIENT ANA1YSES ATk'S impact of all the key transients expected in the life time of the plant were considered in this section. The results show the impact of modification, in terms of reactor pressure, fuel transient and containment conditions by

.j

! comparison to the guides of Section 1.1. The specific postualted modifications considered are recirculation pump trip and ARI. The results of the analyses can be summarized as follows:

1. For most pressuri:ation transients the reactor pressure exceeds the comparison guide for the plant in its present configuration.

2. For all transients, the recirculation pump trip modification effec-tively citigates the short te rm AT'a'S response . The reactor peak pressure satisfies the 1500 psig guide value and the peak enthalpy of the hottest fuel satisfies the guide value of 280 cals/gm.

3. For all transients , ARI successfully mitigates the long term response.

The containment pressure and suppression pool temperature satisfy their respective comparison guide values.

4 In the loss of feedwater heater and rod withdrawal error transients all the key variables are within the comparison guides even without the benefit of the postulated codifications.

i.

l 3-26

--. - ..-.. - .~ .. .-.- -...--.. - - - - - - . - . - ~ - - - - - - . . - . - - - . - . .

V NED0-25016 i

4. SYSTEM DESCRIPTION WITH POSTULATED MODIFICATIONS All of _ the anticipated transients, which would require mitigation with the plant in its present configuration in the unlikely event of an ATWS, quickly reach at l least one of two conditions which are readily sensed and from which the actions of the postulated modifications may be initiated. These conditions are:

1. High vessel pressure, and 2. Low water level. The vessel pressure was chosen to be slightly above the relief valve setpoint. The value used in the 4

4 analysis was 1150 psig. The low low level point chosen is that level at which the recirculation pumps already trip and HPCI and RCIC are initiated. A simpli-fied block diagram of the postulated modification is shown in Figure 4-1. The overall requirements for these modifications are as follows:

A.- The system should be diverse from current RPS.

B. No cingle component failure in the instrument channels or logic shall

- cause inadvertent injection of all control rods.

C. The system should be testable in service. ,

D. The system should be designed so that as much as possible no single component f ailure can prevent 2-RPT and ARI.

E. All hardware should be high quality and be environmentally qualified.

Ce t '.n r.anual actions are required of the operator. Paragraphs 4.1.3 and 4.2.3 l show that capability to manually initiate recirculation pump trip and ATWS Rod  ;

Injection is available as a backup to automatic initiation. Suppression pool cooling must be initiated manually within 10 minutes of the ATWS event (see paragraph 4.5.2).

Certain alarms and indications are given to the cperator to allow him to per-i form the required manual actions within the time limits. The response to request number 2 of reference 5 discusses the information available to the operator. Additionally, annunciator windows have been added that alarm when I

4-1

- . . . - - . - - - . . . - - . . - . . . . . - - .-- - .----. - -. ~.

i . .

SED 0-25016 t

the reactor water level or reactor pressure reach the ATVS setpoints.

Therefore, at the beginning of the ATWS event, when the recirculation pumps are signalled to trip and the ATWS Rod Injecticn is automatically initiated, the operator is alarmed that an ATWS has occurred. He then has sufficient tim t to perform the required manual actions.

4.1 TRIP OF FIELD CIRCUIT BREAKERS OF BOTH RECIRCULATION PUMPS Since normal scram is assumed to be unavailable f or reducing the reactor power and since the transient event is one in which power reduction is necessary, another method of reducing the power is needed for the first 15 seconds of the event. The trip of both recirc pumps causes a quick reduction in core flow which increases the core void generati_'n, thus introducing a negative reactivity thus decreasing the power. In short term cons;derations, the quick power reduction brings the reactor pressure, neutrcn flux and fuel surfact heat flux down in time to acceptably limit the peak pressure, clad oxidation and peak fuel enthalpy. The analysis is done tripping generator field breakers and the results are applicable to alternate breakers installed between the -g set and pump motor, since the field breakers result in a slower trip of the pumps and yield more limiting results.

4.1.1 Performance Characteristics Logic Delay f or Trip (Sec)* 10 53 Pump Inertial Constant (Sec) 10 3 4.1.2 Autcmatic ATWS Actuation Higher Reector Dome Pressure Setpoint (psig) <1150 Reactor Low Water Level Setpoint Low-Low i

• Including dynamic response of the sensors, logic, action of the breakers and collapse of generator field. ,

4-2

l NED0- 2 5016 4

4.1.3 Manual Actuation High Torus Water Average Temperature Alare 1110 Setpoint ('F)

High Reactor Dome Pressure Alarm Setpoint 11150 (psig)

Reactor Low Water Level Alarm Setpoint Low-Low 4.2 ATWS ROD INJECTION ATVS cod injection (ARI) is a means of predccinantly diverse blade injection which is motivated mechanically by the normal hydraulic control units and control rod drives, but which utilizes totally separate and diverse logic.

The advantage of this method is that the in.tial signals of high vessel pres-sure or low water level are used to dump separate valve (s) which cause the pilot air header to bleed down. This bleed down takes approximately 15 seconds after which the reactor is shut down by rod injection. Although this type of rod injection does not eliminate the short term consequences of the assumed failure of normal scram action, it does reduce the long term consequences to nearly those of normal scram situations. The short term consequences are controlled by the early trip of the recirc pumps.

4.2.1 ARI Perf ormance Characteristics Delay Af ter Air Trip (Sec) 151 Logic Delay for Rod Injection (sec)* 1053 Rod injection Rate After Delay Same as normal scram 4.2.2 Automatic Actuation High Reactor Doce Pressure Setpoint (psig) 11150 Reactor Low Water Level Setpoint Lcw-Low

• Including dynamic response of the sensors and logic.

4-3 i _ _ . _ _ _ _ _ _ _ _ _ - _ _ - -

... - . . . . . . - - . . - . ~ . _ . . - - . ~- .- - . - -- ,---~ .----_~~ - ---.~,-.-.-

?

.. . NEJ0-2$016 4 [

i

-- - - - es=== eng I ,

5 f l

• 4 1

l I

HIG H HIGH '

LOW R E ACT OR Rt ACT OR TORU$

WATER L(VEL ppgggggg ggypgggyypg ,

I _

1 I r- J 1

1 r i f l

i MANUAL ,

i f .

i "5 TART HPCI 1 RCic 3 r 3 r 1r RHR -

(LPCI MODE )

't, INITI AT E ATWS ROD INJE CTION

'2. TRIP GENER ATOR FIELD BRE A k,E R$ OF RECll4CU.

LATION PUMP MG SET $

w 'NEW POSTULATE D

{. MODIFICAflONS

If f M$ IN PLACE FOR f WHICH CRE DIT 18 NEEDED Figure 4-1. Monticello Postulated ATWS Modification block Diagram

' 4-5

- - _ . . . - _ _ . . - . . - . - . . _ . - . - . . - . . . ~ . - . - . . - - . . - .

1 l

l .

SEDO-25016 4.4.2 Manual Actuation tiigh Reactor Dome Pressure Alarm Setpoint (psig) 11150 Reactor Lov Water Level Alarm Setpoint Low-Lov I tiigh Torus Water Average Temperature Alarm 1110 Setpoint ('F)

Tiroe Required for Manual Initiation After 110 Alarm (Min) e 4.5 Dl'.*ERSITY, TESTABILITY , SEPARAT10S , REDUSDASCY In order to properly explain the diversity tetseen the normal RPS and A*NS Icgic schemes, it is necessary to first describe the two individual logic schemes. Within each discussica the process sensor variables which initiate RPS and ATWS are discussed as well as the logic ccnfiguration and testing methods. A summary discussion en the diversity of sensor inputs to RPS and the diversity betvcen the LPS logic and the ATWS logic is provided.

Each transient that is expected to be more frequent than one in 40 years is 6

discussed in the attached appendix (answer to question B.13). The discussion of each transient consists of:

1. A description of the transient and how it can occur.

2. The order cf normal scram parev.ter trips that will be generated within the first three minutes after the transient given a normal scram does not occur.

3. The order that the isnticipated Transient Without Scram (ATWS) trips of high reactor pressure or low reactor water that . 11 occur if the j transient is severe enough, given that normal scram does not occur.

4.5.1 Oescription f the Reactor Protection System (FIS) Logic l

The Pf 5 logic is described in this section. The description is brief but covers the major design concepts. More detailed information is found in the Monticello FSAR.

4-6

1 a _

l . . - NEDO-25016 l i The main steam line radiation detectors are located along the main steam lines so they can monitor the radiation levels inside the main steam lines. The detector output is processed through an electronic system to produce an output i that is proportional to the radiation activity inside the main steam line. ,

1 i When the radiation trip point is reached, the bistable in the system provides  ;

i an output to the nortnal scram system. Calibration of the system is persormed  ;

by making artplification adjustnents to generate a trip output while inputting 1 a known standard signal.

a  !

t 4.5.3.2 RPS Sensor Diversity 5-Diversity for the RPS sensor inputs is achieved because of the diverse input I device trips that operate in diverse environnents and have diverse calibration l procedures and calibration standards.  ;

I 4.5.3.3 Diversity Between RPS Logic and ATJS Logic The diversity between the RPS logic and ATWS logic is achieved by funccional ,

application of the logic elements, and location of the logic elements. Some .

logic elements used in synthesizing the design may be the sace, such as trip units, pressure tratismitter, pressure switches, relays, power supplies, vires, However, if common elements are used, the application i

terminal boards, etc.

in relation to the trip status will be diverse. Diversity between the RPS ,

and ATVS logic is shown in the table below. l I

1 i

Location Logic Centacts End Diverse logic  !

Svstem of louic Power Source During Oteration Status Equation

"*~ ~

RPS RPS cabinets 115 voit AC closed energized g ATWS ECCS 125 voit DC open de-energized Two-out-of i cabinets two or two-i out-of-two l

l 4,6 RELIABILITY IMPROVDiENT DUE TO ARI 3 L

i i

l The proposed addition of a scram air header trip initiated by the ATWS logic provides a diverse neans of tripping the reactor protection logic system.

4-17 )

m - , .~ , . - ,-,,--,n--,- n,,n- ..n,,,-,--,-..e,,.._-.-v. - . - - . - - - . . ~ . - - - - ~ - - . - - - - - - - - - - - - - - -

iED0- 2 50l t, i .

This modification reduces the top majot contributor f.o scram unreliability, i.e., censon mode f ailure of the eight scram centacters. It is estimated that this modification veuld result in approximately a one (1) crder of magnitude reduction in scram system unraliability.

The proposed change does not significantly re-duce the potential f or miscali-bration of all scram sensors as given in WASH-1400. Hevever, it is felt that l the probability of miscalibration of several sets of diverse sensors was overly conservative in the WASH-1400 analysis. Current preliminary analyses indicate that this failure mode may drep frcm tt.e list of major contributors.

1 I

e i

l 4-18 l

KEDO-25016

5. SUMHARY The ATWS events were analyzed, showing the need for plant modifications for events if the failure to scram is postulated. The events analyzed covered those I transients expected to occur within 40 years, coupled with a failure to scram.

l'ne ATWS Frevention system identified includes automatic recirculation pump trip, ATWS Rod Injection, manual suppression pool cooling and reactor vessel

' water level maintenance by the core cooling systers. The ATVS Prevention system provides satisf actory recovery of the plant for all transients analyzed. The ATWS Rod injection feature decreases scram unreliability by approximately one order of magnitude. Responses to all identified unanswered NRC questions are provided. In many cases, the responses are generic and reference is given.

i

(

i i  !

3 5-1

i. .

1 NEDO-? 5016

\

i l Resp 0NSE TO QUESTION 1 QUESTION l

Provide the peak torus water temperature reached during the MSIV closure ATUS. Pro-3 vide and justify a torus water temperature limit. If the calculated temperature l exceeds the limit, discuss the plant modifications needed to keep torus water tem- 1 d perature below the proposed limit. If the peak torus water temperature exceeds 1700T i l f discuss plant modifications needed to keep this temperature below 1700T.

i

RESPONSE

1 i

The peak bulk pool temperature reached following the postualted MSIV closure  ;

.' ATWS event with the postulated modifications is 153.4't.

The torus water temperature limit during S/RV discharge is dependent on the con-figuration of the discharge pipe where the steam enters the suppression pool. A ,

. local limit of 170'T for an ATVS event has been established for pipes with a t

t

single discharge point, such as a uniform cross section pipe. Data presented in NEDO-21078 Test Results Employed by General Electric for BVR Containment and ,

Vertical Vent Loads. October 1975, indicatas that additional discharge points  !

(greater energy dispersion) generally improves the thermal performances of the discharge device. Therefore the rams head discharge thermal performance would be expected to be better than a single pipe discharge. Even so, the same thermal i limit of 170 T (ATWS) is currently recommended for the rams head discharge.

11 the calculated pool temperature exceeds the thermal limit established for the ATVS event, then modifications could be made to the NSS to reduce the energy ,

2 .

i 4 released to the pool during the event (for the existing discharge device), or the i

j 31scharge configuration could be modified to raise the thermal limit to a higher temperature.

t

= ,

1 A-2

NEDO-2!016 RESPONSE TO QUESTIOS B.1 QUESTION Section 7.1 of NEDO-20626 identifies the systems relied upon to mitigate the con-sequences of An'S. Der nstrate the diversity of these systems and the

• initiat-ing signals f rom the Reactor Scram System. Further discuss the reli. ty of

', these systems to perf orm their f unct ' ens during an ALT, event.

RESPONSE

The systems identified in SEDO-20626 for B category plants are not appropriate for Monticello. The appropriate systers fer Monticello are specified in Sec-tion 4 of this report. The response to the diversity question is contained in Sections 4.5 and 4.6.

i A-5

I NEDO-25016 i 1

l 4 l

e. RHR flow and temperatur1 vhen RHR is used for suppression pool cooling and f decay heat removal ,

1

. f. Storage capacity of each source of water used to maintain level and recove  :

f 1 energy from vessel ,

i 4 i

g. Operacor actions including the time action taken j

l QUESTION B.5 i i

In an October 7, 1974 letter from I. T. Stuart to V. Stello, CE stated (responses to Question 4) that the condensate storage tank would provide water for KPCI and  !

RCIC for 24 minutes and that the suppression pool is not needed as a scurce of water for an ATVS event. If this is the case, explain how the plant can be brought to a cold shutdown condition. l i

RISPONSE These questions will be addressed belov first for all the transients analyzed in

< Section 3 except the inadvertent opening of a relief eAlve. Discussion of 1 the latter transient follows. The case of a S/R valve f ailing to reclose during I an AT%'S event constitutes a Single failure in addition to the common mode failure and is not analyzed for Monticelas, an ATWS "C" plant.

Discussion for Transients Other Than Inadvertent Opening of a Relief Valve In section 3, the time evolution of plant variables in ATWS transients until after achieving hot shutdown was discussed. In this state, the reactor power is due only to decay heat which is removed by steam flow through relief valves into the suppression pool (or, in the case of transients like turbine trip with bypass available, by steam flow into the main condenser). The reactor inventory in the hot shutdown state is made up by EPCI and/or feedwater system flows. The bource of the HFCI system flow is the condensate storage tanks which is (75,000 gal-lons each) sufficient for at least fifty minutes worth of full HPCI flow (3000 gpm). Similarly, the Monticello main condenser wet well is large enough to ,

i supply the feedwater system for an approximation of 3 minutes at its full flow r

A-7 I

______7

! . i i

1 NED0-25016 i

e

' even when there is no steam flow into the condenser. Therefore, the HpCI system  ;

alone can maintain e.akeup water to ccepensate for reactor inventory loss due to decay heat for more than 40 hours4.62963e-4 days <br />0.0111 hours <br />6.613757e-5 weeks <br />1.522e-5 months <br /> after hot shutdown. l f k' hen hot standby is achieved by ATk'S Rod Injection (ARI), the situation of reactor  ;

depressurization can te achieved by mcaipulating the ADS system and relieving the I

stored energy of the reactor into the suppression pool. This would result in a suppression pool temperature rise. k'henever the main condenser is available for -

duty steam flow into the condenser can be established through the turbine bypass line, thus avoiding further steam flow into the suppression pool. The reactor depressurization rate can be controlled by manipulating t'ne ADS valves or the 1

. pressure regulator setpoint.

Energy release during reactor depressurization down to 150 psig is expected to result in a suppression pool temperature rise of about 40'T (with only one RHR heat exchanger in the pool cooling mode). The action of one pJ1R heat exchanger ,

can accomplish a suppression pool cooling rate of s6'F/hr when a bulk pool tem-  !.

perature of 160'T is assumed). Therefore if the reactor depressurization (to  :

150 psig) is perfomed over a period longer than 16.8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />. the energy release into the suppression pool vould not exceed 160*F when both RHR heat exchangers are in the pool cooling mode.

l In the Monticello plant the RHR system heat exchangers can assume the function of ,

directly cooling the reactor water only af ter the reactor is depressurized. Thus.

the ITCI system will assume the long term cooling function through reactor depres-l surization. The RPCI system is automatically initiated by the reactor low low i water level and does not require specific operator action.

Figures A.1 and A.2 show the long term traces of reactor power .and reactor pres-sure prior to the start of depressurization for the ?$1V closure transient.

Reactor water icvel and containment conditions are shown in Figures 3-4, 3-5 and 3-6. These plots are typical of pressurization transients which are not accompa-nied by feedwater ass.

(

A-8

, . i 1  !

(

NED0-25016 [

i i l

Figures A.3, A.4 and A.$ show the plots of retctor power, pressure and water level

{

]4 for the case of loss of AC pcwer ATWS. Containnent temperature and pressure i s plots are shovn in Tigures 3-8 and 3-9 respectively. These plots are typical of 3

pressuritation transients accompanied by loss of feedwater flov early in the i 1

, transient.

1 I

t i Discussion for the Case of Inadvertent Opening of a Relief Valve I J

In this case hot shutdown in achieved at approximately 350 seconds with ARI.

The reactor will centinue to blow down through the inadvertently opened reitet valve.

The feedwater system can continue to make up the reactor inventory loss for some i

time after hot shutdown is achieved. Thereaf ter the HPCI (initiated by Low-Low water level) will assume this function. Since the HPCI pumping capacity is more

than twice the rated relief valve flew, reactor inventory can be adequately maintained. Moreover, the condensate storage tank can supply the KPCI system in l this reactor water makeup function for more than one hour. Beyond this time, h the suppression pool forms a second source of water for the HPCI system which can [

be utilized by opening appropriate valves. Since in this situation a closed flow 3

path is established, reactor water inventory can be maintained indefinitely. I When ARI function is assumed, the suppression pool tee;.erature at hot shutdova is 120'F. As mentioned before, there vould be anothe*, estimated rise of 40 F as the reactor blows down to a pressure of 150 psig. (This rise would be somewhat less than 40'T vhen effect of both RHR heat exchangers in the pool cooling mode is '

considered.) Thus, the suppression pool temperature vould reach 160 T before the

) reactor pressure is below 150 psig. At reactor pressures equal to or less than this pressure, the steam mass flux through the relief valve vould be low enough that the 170'T pool temperature limit vould not be critical. Higher teeperature limits are justifiable for these conditions. Tigures A.6 and A.7 show the long term plots of reactor power and vator level.

l t

I l

2 A-11 eer- re.wi+v&--d-yrw-ee---.-.-1rN t-- --? guy v pr$.'N--w'-d-s+ -a-- *4-=m- * *-""-m-u

• NED0-25016 PIsp0SSE TO QUESTION B.16 QUESTION Tne sensitivity of peak pressure to relief valve capacity is presented in Table 6-3 of UEDO-20626. For each product line, what is the minimum relief capacity?

For the minimum relief capacity plants, what is the relief capacity of each valve?

What is the probability that a relief valve will not open upon reaching the pressure setpoint? Identify B class plants with lower relief capacity than that used in NED0-20626. Previde AT.!S analyses using the plant with the least relief capacity as basis.

Resp 0NSE Monticello has eight relief valves of combined capacity equal to 74.4!.' of the rated reactor steam flow. The analysis teported in Section 3 takes credit only for six of the eight valves in the reactor peak pressure calculations. The con-tainment conditions are not very sinsitive to whether six or eight valves are used in the analysis. The failure of a relief valve to open in addition to an ATWS is not considered applicable to the Monticello plant.

A-19

_ __________________ _ __ _ _