ML20202A833

Rev 0 to Accident Analysis Methods for Application to Salem Nuclear Units

ML20202A833
Person / Time
Site:	Salem
Issue date:	03/14/1986
From:	Hsu D, Rosenfeld E, Rothrock D Public Service Enterprise Group
To:
Shared Package
ML18092B111	List: ML20202A833
References
NFU-0033, NFU-0033-R00, NFU-33, NFU-33-R, NUDOCS 8604110067
Download: ML20202A833 (127)

Text

f lE NFU-0033 E Revic3 ion 0 March 14 1986

I II PUBLIC SERVICE ELECTRIC AND GAS COMPANY I REPORT NUMBER

NFU-0033 REPORT TITLE: ACCIDENT ANALYSIS NETH00S FOR APPLICATION TO SALEM NUCLEAR UNITS APPROVAL REVISION O EFFECTIVE DATE 3/l4[8Io PREPARED BY 05/VME/d DATE /4!86 I REVIEUED BY A C A ~~ DATE /

REVIEUED BY d() 6 O DATE I AgeR0veo eY f // I d7 / / - oATE 3,4d, s, I

g C0ev NO. 8 I

I I

I nnnanaam P PDR

I I NFU-0033 Revision O tiarch 14, 1986 I

I ABSTRACT This report describes the methodology used by Public Service Electric and Gas Company (PSE&G) to perform transient and accident analysis for the application to the Salem pressurized Water reactors.

I .

I .

II I

I I

I I

I I

NFU-033 Revision 0 I March 14, 1986 TABLE OF CONTENTS E

me 10 INTRODlTCTION' 1-1 I 20 GENERAL PHYSICS INPUT 2.1 Moderator Temperature Coefficient 2.2 Baron Reactivity Uorth 2-1 2-1 .

2.3 I 2.4 25 Doppler Reactivity Coefficient Scram Reactivity Curve Hot Channel Factor 2-1 2-2 2-2 26 Effective Delayed Neutron Fraction 2-3 2.7 Prompt Neutron Lifetime 2-3 3.0 SAFETY EVALUATION 3-1 I

i 31 Uncontrolled Rod Cluster Control Assembly I 32 Withdrawal From a Subcritical Condition 3-2 Uncontrolled Rod Cluster Control Assembly Withdrawal at Power I 3.3 3.4 Uncontrolled Baron Oilution at Power Full Length Rod Cluster Control Assembly Orop 3-9 3-20 3-26 I 35 3.6 Excessive Heat Removal Oue to Feeduater Control Valve Malfunction Loss of External Load 3-36

) 3.7 Loss of Normal Feeduater 3-43 38 Loss of Reactor Coolant Flou - Pump Trip 3-58 i

3.9 3-64 Loss of Reactor Coolant Flou - Locked Rotor 3-75 3 10 Major Secondary System Pipe Rupture 3-84 3 11 Rod Cluster Control Assembly Ejection 3-91 40

SUMMARY

4-1 l 50 REFERENCES 5-1 APPENDIX A - COMPUER CODE DESCRIPTION A-1 I

I I

I I

NFU-033 '

R vision 0 March 14, 1986 I

LIST OF FIGURES Fiqure Page I

311 Rod Uithdrawal Transient at HZP. EOL:

Neutron Flux Versus Time 3-5 3.1 2 Rod Uithdrawal Transient at HZP. EOL:

Thermal Flux Versus Time 3.-6 3 1.3 Rod Withdrawal Transient at HZP. EOL: 3 Average Fuel, Clad and Coolant 3 Temperature Versus Time 3-7 3.2.1 Rod Uithdrawal Transient Uith Fast Uithdrawal Rate at HFP:

Nuclear Pouer Versus Time 3-12 3.2.2 Rod Uithdrawal Transient Uith Fast Uithdrawal Rate at HFP:

Pressurizer Pressure Versus Time 3-13 3.2.3 Rod Uithdrawal Transient Uith Fast I

Uithdrawal Rate at HFP:

Average Core Coolant Temperature l Versus Time 3-14 m 3.2.4 Rod Uithdrawal Transient Uith Fast g Uithdrawal Rate at HFP: g DNBR Versus Time 3-15 3.2.5 Rod Uithdrawal Transient Uith Slow Uithdrawal Rate at HFP:

Nuclear Power Versus Time 3-16 3.2.6 Rod Uithdrawal Transient Uith Slou Uithdrawal Rate at HFP:

Pressurizer Pressure Versus Time 3-17 327 Rod Uithdrawal Transient Uith Slou Uithdrawal Rate at HFP:

Average Core Coolant Temperature Versus Time 3-18 3.2 8 Rod Uithdrawal Transient Uith Slou uithdrawal Rate at HFP:

i DN8R Versus Time 3-19 3.3.1 Uncontrolled Baron Oilution Transient: $

Vessel Average Ccolant Temperature E Versus Time 3-22 3 3.2 Uncontrolled Baron Dilution Transient:

Pressure Versus Time 3-23 I

NFU-033 Revision 0 March 14, 1986 I

I LIST OF FIGURES (Continued)

Figure Pigg I 3 3.3 Uncontrolled Baron Oilution Transient:

Core Inlet Temperature Versus Time 3-24 Uncontrolled Baron Oilution Transient:

'I 3 3.4 ON8R Versus Time 3-25 Oropped RCCA Transient:

I 3.4.1 Core Heat Flux Versus Time 3-28 3 4.2 Oropped RCCA Transient Change in Average Temperature Versus Time 3-29 3.4.3 Oropped RCCA Transient:

Pressurizer Pressure Versus Time 3-30 3.4.4 Dropped RCCA Transient:

Change in DN8R Versus Time 3-31 3.4 5 Oropped RCCA Transient:

Core Heat Flux Versus Time 3-32

, 3.4.6 Oropped RCCA Transient:

l Core Average Temperature Change Versus Time 3-33 3.4.7 Dropped RCCA Transient Pressurizer Pressure Versus Time 3-34 3.4.8 Dropped RCCA Transient:

Cha,nge in DNBR Versus Time 3-35 351 Feeduater Contral Valve Halfunction Transtent:

Fraction of Nominal Neutron Flux Versus Time 3-38 3 5.2 Feedwater Control Valve Malfunction Transient:

Change in RCS Average Temperature Versus Time 3-39 3 5.3 Feedwater Control Valve Halfunction Transient:

Change in RCS Delta T Versus Time 3-40 3 5.4 Feeduater Control Valve Halfunction Transient:

Change in Pressurizer Pressure Versus Time 3-41 I .

l

NFU-033 R:;vicion 0 March 14, 1986-I LIST OF FIGURES (Continued)

Fiqure .P_g.gg 3 5.5 Feeduater Can'rol t Valve Malfunction Transient:

DN8R Versus Time 3-42 361 Loss of Electric Load Transient:

Neutron Flux Versus Time 3-48 3.6.2 Loss of Electric Load Transient:

Pressurizer Water Volume Versus Time 3-49 3.6.3 Loss of Electric Load Transient:

Pressurizer Pressure Versus Time 3-50 3 6.4 Loss of Electric Load Transient:

Average Core Temperature Versus Time 3-51 3 6.5 Loss of Electric Load Transient:

DNBR Versus Time 3-52 3.6 6 Loss of Electric Load Transient:

Neutron Flux Versus Time 3-53 3 6.7 Loss of Electric Load Transient:

Pressurizer Uater Volume Versus Time 3-54 3.6 8 Loss of Electric Load Transient:

Pressurizer Pressure Versus Time 3-55 3 6.9 Loss of Electric Load Transient:

Average Core Temperature Versus Time 3-56 3.6.10 Loss of Electric Load Transient:

DNBR Versus Time 3-57 3.7.1 Loss of Normal Feeduater Transient:

Core Average Temperature Versus Time 3-61 3 7.2 Loss of Normal Feeduater Transient:

Steam Generator Uater Level Versus Time 3-62 373 Loss of Normal Feeduater Transient:

Pressurizer Uater Volume Versus Time 3-63 3.8 1 Complete Loss of Flou - Pump Trip Transient:

Neutron Flux Versus Time 3-68 3.8 2 Complete Loss of Flou - Pump Trip Transient:

Core Flow Versus Time 3-69 I

iv

NFU-033 I Rsvision 0 March 14, 1986 E

I LIST OF FIGURES (Continued) j Fiaure ,P_;Lgg 3.8.3 Complete Loss of Flow - Pump Trip Transient:

Heat Flux Versus Time 3-70 3.8.4 Complete Loss of Flou - Pump Trip Transient:

DNBR Versus Time 3-71 3.8.5 Partial Loss of Forced Reactor Flow:

Core and Loop Flous Versus Time 3-72 3.8.6 Partial Loss of Forced Reactor Flow:

Neutron and Heat Flux Versus Time 3-73 I 3.8.7 Partial Loss of Forced Reactor Flow:

DNBR Versus Time 3-74 Locked Rotor Tr.ansient:

I 3.9.1 Nuclear Power Versus Time 3-79 3.9.2 Locked Rotor Transient:

Hot Channel Heat Flux Versus Time 3-80 3.9.3 Locked Rotor Transient:

Core Flou Versus Time 3-81 3.9.4 Locked Rotor Transient:

Reactor Coolant Pressure Versus Time 3-82 3.9.5 Locked Rotor Transient:

Clad Temperature Versus Time 3-83 3 10.1 Main Steamline Break Reactor Vessel Average Temperature Versus Time 3-86 3 10.2 Main Steamline Break:

i Reactor Coolant Pressure Versus Time 3-87 3 10.3 Main Steamline Esreak Core Heat Flux Versus Time 3-88 I 3 10 4 Main Steamline Break:

Steam Release Versus Time 3-89 I

I ~

I .

__ ____________m_____ _ _ . _ _ _ _ _ . _ _ _ _ _ _ . _ _ _ _ _ _ _ _

NFU-033 Revision 0 March 14, 1986 LIST OF FIGURES (Continued)

Fiqure Paug 3.10 5 Main Steamline Break:

Reactivity Versus Time 3-90 3 11 1 Rod Ejection Transient:

Core Power Versus Time at HFPBOL 3-96 3.11.2 Rod Ejection Transient:

Core Power Versus Time at HFPBOL 3-97 3 11 3 Rod Ejection Transient:

Core Power Versus Time at HZPEOL 3-98 3 11.4 Rod Ejection Transient:

Core Power Versus Time at HZPEOL 3-99 3 11 5 Rod Ejection Transient:

Temperature Versus Time HFPBOL 3-100 3.11.6 Rod Ejection Transient:

Temperature Versus Time HZPEOL 3-101 I

I I

I l

I I

I l

.1 I I

I NFU-033 Revision 0 March 14, 1986 I LIST OF TABLES Tahle q P_asg I 3.6 1 Time Sequence of Events for Loss of External Electrical Load with Pressurizer Spray and PORV's at BOL 3-46 3.6.2 Time Sequence of Events for Loss of 3-47 External Electrical load Without Pressurizer Spray and PORV's at BOL I 3.8.1 Time Sequence of Events for Loss of 3-67 l Reactor Coolant Flow I

E I

1 I

'I I

I I

I I

I vii I f

I NFU-0033 Revision 0 March 14 1986 10 INTRODUCTION I In order to gain a better understanding of the Salem Nuclear Generating Station operational transient phenomena. Public Service Electric and Gas (PSE&G) has performed a series of safety analyses to demonstrate the behavior of system components and core thermal I hydraulics under transient conditions. The results of these analyses have been compared with the vendor's I ~

calculations (contained in the FSAR) to demonstrate that PSE&G has the capability to check the vendor's transient pr~edictions independently with the intention l

of performing licensing safety and transient analyses.

I The methods and assumptions pertaining to use of DYN00E-P as the nuclear steam supply system simulator are considered adequate as presented in this report for use in safety analyses. The methods and assumptions I pertaining to both the fuel rod thermal and DNBR calculations are considered preliminary and are included to illustrate OYNODE-P's capabilities for lll l

supplying boundary condition forcing functions to fuel rod thermal and DNBR analyses. Updates to this report l will be made when the fuel rod thermal and DN8R models I and computer codes are verified to be adequate for safety related application.

A brief description cf the general physics parameters I used as input to the transient analysis is presented in Section 2. The upper and lower limits of the physics parameters are discussed. Typically. the specific I cycle design physics parameters will be within the bounding values.

lower limit in the safety analysis is dependent upon The choice of either an upper or the specific transient conditions. The general rule is I to choose the most limiting (conservative) values for the analysis.

A description of each transient analyzed is presented I in Section 3. Assumptions and methods used to analyze the accident are also described. Finally, a discussion of the results is presented in Section 4.

Appendix A gives the description of the computer programs that were used to analyze the accidents.

I I

I 1-1

NFU-033 Revision 0 March 14, 1986 2.0 GENERAL PHYSICS INPUT I The following is a brief review of certain physics parameters that are used as input for analyses presented in section three. These parameters are applicable to the Salem Units generic cycle Cthe I values may be found in the Final Safety Analysis Report (6)]. .The values chosen for the following I parameters vill be discussed in the pertinent analysis in section three.

2.1 MODERATOR TEMPERATURE COEFFICIENT an DYNODE-P uses two different methods to input a ,

the moderator reactivity coefficient. The fir 9t bases the defect on a change in coolant tempera-I ture. This Moderator Temperature Coefficient is defined as the change in reactivity per degree change in moderator temperature at constant fuel I temperature. (A negative a implies that an increase in coolant tempera 9ure results in a decrease in reactivity.) The second method is to I describe a as the Moderator Density Coefficient.

This is defined as the change in reactivity unit change in the moderator density at constant per fuel temperature.

The value and method of input for a is based on M

the transient and information supp1ied by Westing-I house in their analysis (6).

magnitude of a n Typically, the will increase over a core's lifetime due t0 the build-up of plutonium and other fission products.

I 22 BORON REACTIVITY UORTH The baron reactivity worth is defined as the I change in reactivity per change in baron concen-tration. This then is multiplied by the baron I concentration to give the baron defect. The values used for the baron reactivity uorth are

-16 0 and -8.0 pcm/ ppm.

23 DOPPLER REACTIVITY COEFFICIENT. a0 The Doppler reactivity coefficient, a is defined as the change in reactivity per degreh, change in 4 the effective fuel temperature. As the fuel  !

temperature increases. the resonance absorption cross sections of U-238 and Pu-240 therease. This I phenomenon. Doppler broadening, results in an increase in the number of fast neutrons which are I '

l I 2-1 i

i

NFU-033 Ravicion 0 Mnrch 14, 1986 parasitically absorbed in the fuel and therefore, a decrease in the reactivity. Consequently, the Doppler reactivity coefficient is negative.

That is, increasing fuel temperatures results in decreasing Doppler reactivity and vice versa.

In the transient analysis, the data is taken from a generic figure of the Doppler power coefficient used in the FSAR (reference 6. figure 14.0-5) in which bounding power dependent values are given for the most and least negative Doppler coef-ficient. This Doppler power coefficient was then integrated and converted to a Doppler temperature l defect. Normally, a least negative Doppler a coefficient is assumed in the heat-up transients and a most negative Doppler coefficient is e assumed in the cool-down transients. g 24 SCRAM REACTIVITY CURVE. ascram(l)

The scram reactivity curve a is defined asthetimedependentreactivyggain(t), troduced into the core due to the insertion of control rods following a reactor trip signal. In these analyses, the scram curve used represents the reactivity insertion assuming the most reactive rod to be stuck in its fully withdrawn position.

2.5 HOT CHANNEL FACTOR. F q The nuclear defined heat as the flux ratio of hat channel factor.

the maximum F local h0a,t is flux in the core to the average fuel rod heat flux in the care. Incorporated into this value, besides uncertainty factors associated with core flux mapping and manufacturing tolerances are factors relating the axial and radial hot channel g factors g F,-KxFyy x Fy K = Factor representing mapping and manufacturing uncertainties.

3 5

Fyy = Ratio of radial peak power density to average peak power density in the horizontal plane of peak local power.

Fy = Ratio of the linear power density in E the horizontal plane of peak local B power to the average linear power density.

2-2 I

NFU-033 I Revision 0 March 14, 1986 I 26 EFFECTIVE DELAYED NEUTRON FRACTION.' B,pp I The effective delayed neutron fraction. B defined as the ratio of'all the delayed n$u(r. ons per fission to the total number of neutrons per is fission. This value is given as a beginning of cycle or end of cycle value. The difference is a I result of the inventory change of uranuim and plutonium over the cycle. As plutonium builds up, the number of delayed neutrons decreases, therefore, 0,pp decreases.

2.7 PROMPT NEUTRON LIFETIt1E. u The orompt neutron lifetime. Ou, is defined as the cverage time +'1r a fission emitted prompt

~

neutton to be absorbed, or to leak out from the I systen. This value is found to be slightly dependent upon core life in that there is a small change associated with fuel inventory changes.

I I

I l I

I '

I l I

I I

I 2-3

i l

l NFU-033 i

Revision 0 ,

30 SAFETY EVALUATION This section deals with the transient specific methods employed in performing a series of Salem generic safety analyses and presents comparisons of the results with l l

the vendor's results found in the Salem Final Safety  !

Analysis Report.(6) The following sections are I presented for each analysis:

I a. Description of the Accident - a brief synopsis of the accident including possible causes of the occurrence.

I b. Summary of Accident Analysis Nethodology -a brief discussion of the methods used to simulate the transient resulting from the accident.

c. Results - a presentation of the results, necessary comparison with FSAR results, and conclusions drawn.

I Specific physics parameters. as previously described, are chosen from the bounding values to give the most limiting condition'for that specific transient /acci-I dent. If a specific cycle design produces a physics parameter with a value exceeding the bounding value.

further evaluation vould be necessary. The evaluation I vill use methodology similar to that described in this section for the analyses.

I I

I .

I I

I I ,

I I 3-1 l

I - 1 J

NFU-033 E Revision 0 3 March 14, 1986 3.1 UNCONTROLLED ROD CLUSTER CONTROL ASSEMBLY UITHORAUAL FROM A SUBCRITICAL CONDITION I

3 1.1 Description of the Accident The accident is caused by the malfunction of the electrical circuits which supply current to the rod cluster control assembiy (RCCA). The maximgm reactivity l

insertion rate is 75. x 10 delta k/second; this value is greater than that occurring with the simultaneous l uithdrawal of the two control banks having the maximum combined worth at maximum speed. The neutron flux response to the reactivity insertion is charac-terized by a very fast rise terminated by the reactivity feedback effect of the negative Doppler coefficient. Conse-quently, the power burst is limited to a 3 3

tolerable level and the accident is terminated by a lou power trip.

I I

I I

I I

I I

3-2 I

I

I NFU-0033 Revision O 1

March 14, 1986 I

3.1.2 Summarv of Accident Analvs'is Methodoloav The uncontrolled RCCA vithdrawal from a subcritical condition was analyzed using I two computer codes. First DYNDDE-P,(3) a system simulation code which incorporates point neutron kinetics, including delayed neutrons and decay i I heat, was used to determine the pouer  !

history and system behavior. Following I

this, FRAP-T5,(4) a fuel rod analysis code, was used to calculate the hot channel fuel, clad and coolant temperatures using the DYNDDE results for I core power, inlet flow, inlet temperature and system pressure.

Conservative results were obtained by I using the following assumptions:

1 A Doppler coefficient of low absolute magnitude was used.

2 A positive moderator temperature coefficient (MTC) was used (i.e., 1 I PCM/*F).

3. The reactor was assumed to be initially at hot zero power (HZP).

4. The maximum positive reactivity I insertign rate assumed was 75.x10 delta K/second which is greater than that for the l simulataneous withdrawal of the I -

combination of the two control banks having the greatest combined uorth at maximum speed (45 inches /

minute).

I I l l

il I 3-3 I .

NFU-033 Revision 0 March 14, 1986

5. The most adverse combination of instrument errors, setpoint errors, and delays for trip signal actuation was assumed. A ten percent increase was assumed for the power range high neutron flux trip setpoint. raising the low g setting from the nominal value of 3 25 percent to 35 percent. The scram curve was based on the assumption that the most reactive rod was stuck in its fully withdrawn position.

I I

I I

I I

I I

l I

I I

I 3-4 I

- ,.,,- ,, - -..-, , ,.,- . , ,, ,,.-,---- - ,,--- --- ,, -..--,.--.r.---- - . - - . - - - - - - -

NFU-0033 Revision 0 March 14, 1986 3.1 3 Results The subcritical uncontrolled RCCA withdrawal transient was analyzed using the input assumptions described in the l I Salem FSAR.(6)

The DYNODE-P code was ussf to analyze the I case of a rapid (75.x 10 delta k/second) RCCA withdrawal at HZP.

The results of these calculations were compared with those in the FSAR.(6) The neutron flux is shown in Figure 3.1 1.

DYN0DE-P predicted a slightly smaller I peak neutron flux than the FSAR results.

This peak occurred slightly later than in the FSAR (less than 0.5 second).

The thermal flux and fuel temperature calculations performed by FRAP-T5 used I the DYN00E-P results as input.

thermal flux and the fuel temperatures predicted by FRAP-T5 are plotted against The FSAR predictions in Figures 3 1 2 and I 3.1.3. respectively. The thermal flux predicted by FRAP-T5 was considerably less than the FSAR prediction. The I underprediction was due to DYNODE-P's nuclear flux being slightly low and the FRAP-T5 modeling not accounting for end I of life conditions. The fuel temperature calculations were low for the same reasons.

I Since the maximum coolant temperature, thermal power and heat flux would not exceed the nominal full power values.

I the DNBR uould be higher than the design limit of 1.30.

il

,11 l

lI I

^5 I .

NFU-033 Revision 0 March 14, 1986 l I I

! }lr ,!

. i 3

g.. .,

I a .a x

Q N W  ! ..

4 I '

M 3:

l- >

IN I

I,

<d z .__

O y -

/

3 a

O ___---2:

x ----------- +

P d /' .. I B

I l ,.

l WN$WWW3-6 l

I

._._..-_.__,,.,--._,,,_--,,_._,,.v__,__. . . , . . . . _ , . _ . - . . _ , . , , _ , . . , _ , _ . _ , , . . , _ _ . . . . - . , _ _ _ . . . . _ _ . . . . . _ . .

man mum uma uma sus man man am som as aus aus sus em um em um um um FIG.3.1.2 ROD WITHDRAWAL TRANSIENT AT HZP,EOL:

THERMAL FLUX VERSUS TIME 1.2 .

REACTIVITY INSERTION RATE - 75 X 10-5 DELTA K/SECOND 1-i '

d z

s .

I o Z 0.8 -

i 4 f l o l\ s 0.6 - I g

E I y X l I \

I g 0.4 -

1 5 l \

l l iE I s ggg

.2 - I s Legend

3. ("a

I s DYNODE -ow j

~~~ -__ _ FSAR_ _

.a

'o

' e 4 iO h 14 1s is 20 $

TIME, SECONDS i

l i

i FIG.3.1.3 ROD WITHDRAWAL TRANSIENT AT HZP, EOC:

i, AVERAGE FUEL, CLAD, AND COOLANT TEMPERATURE VERSUS TIME ,

.' 1000 l REACTMTY INSERTION RATE l = 75 X 10-5 DELTA K/SEC i

i 1 900-FUEL

\ l i

\

i k 800- l

\

! W I

(

g i

\

s I N

3 700- l

\

W i s etAo s l '

I /

Ms s ~

t

• 00- I ~ ~ , ,

i -

Legend zm=

Er yjy FSAR l c0OLANT rRAe e;a

~st 1'8 20 pD k k 8 10 h N 16 o

0 h ~

TIME, SECONDS l E o,

i

6 NFU-033 Revision 0 I March 14, 1986 I 3.2 UNCONTROLLED ROD CLUSTER CONTROL ASSEMBLY UITHORAUAL AT POUER 3.2.1. Description of the Accident The postulated accidental rod control I cluster assembly (RCCA) withdrawal is assumed to be caused by the malfunction of electrical circuits which supply the I current to the rod cluster control assembly. The. result would be an increase in the core heat flux. If it I

was not terminated by reactor trip, the primary to secondary power mismatch and the resultant coolant temperature rise could result in DNB. In order to avoid cladding damage in transients such as I this, the reactor protection system is designed to terminate such transients I before the DNBR decreases belou a value of 1.30.

3.2.2 Summary of Accident Analysis Methodology I '

The uncontrolled rod cluster control assembly withdrawal was analyzed using the system simulation code DYN00E-P,(3)

I ubich incorporates models of point kinetics. RCS, pressurizer, pressurizer relief and safety valves, steam generator I relief and safety valves. The core thermal hydraulics transient was analyzed using a modified version of COBRA IIIc-I MIT(2). In order to obtain conservative values of DNBR, the follouing assumptions were made:

1 A conservatively small (in absolute  ;

magnitude) value was assumed for '

the Doppler pouer coefficient.

2. A zero moderator temperature coefficient corresponding to I beginning of life is assumed.

This, combined with the minimum Doppler, allous the core ' power to increase faster at the beginning of the transient.

3 Initial conditions of maximum care I power, maximum reactor coolant average temperature and minimum reactor coolant pressure were assumed. These assumptions uere made to ensure minimum initial l margin to ONB.

.I .

f NFU-033 g Revision 0 E March 14, 1986

4. The maximum positive reactivity I

insertion rate which is greater ,

than that for the simulatneous withdrawal of the tuo control banks having the maximum combined worth at maximum speed. Tuo reactivity insertion rates were utilized in El E'

thig 10 analysis; specifically, 75 5*

delta k/second and 3. x 10 g delta k/second. g

5. The reactor trip on high neutron flux uas assumed to be actuated at 118 percent of nominal full power.

6 The coolant and pouer history were g then used as inputs to the COBRA g code to calculate DNBR.

3.2.3. Results (a) -5 Fast Withdrawal Rate--75. x 10 delta K/second.

The nuclear pouer, pressurizer pressure, average core coolant

~

temperature. and DNBR values during this transient are shoun in Figures 3.2.1, 3.2.2, 3.2.3, and 3 2.4, respectively. This transient was g terminated when the high neutron 3 flux set point was reached.

The pressurizer pressure predicted by DYNODE-P was slightly higher than that of the FSAR.(6) Houever, the pressure never reached the l pressurizer relief valve set point. 5 The nuclear power and average core coolant temperature predicted by g DYNODE-P were in good agreement g with those of the FSAR. The DNBR never fell below 1.30.

(b) Slow Withdraual Rate--3. x 10 -5 delta K/second The nuclear pouer and average core coolant temperature response are shoun in Figures 3.2 5 and 3.2.7 respectively. The comparisons between DYNODE-P and the FSAR are good. This transient uas termi-nated when the overtemperature 3-10 I

NFU-033 I Revision 0 March 14, 1986 I delta T trip set point was reached.

The pressure histories are plotted I in Figure 3 2 6. The set point for the pressurizer pouer operated relief valve was 2350 psia. The valve opened at about 36.53 seconds I into the transient and stayed open for about 2 0 seconds. The DNBR predicted by COBRA was always above 1.30, as shown in Figure 3.2.8.

In conclusion, the analyses shou that at the fast withdrawal rate, I protection would be provided by the high neutron flux trip. At the slau uithdrawal rate, protection would be provided by the over-I temperature delta T trip. For withdrawal-5 tes within the range I of 75 x 10 -5d elta K/second (fast) and 3. x 10 delta K/sec (slow),

it is expected that protection would be provided by one of the This was demonstrated I. above trips.

in the FSAR, where the results of analyses using rates covering this range, were presented. The DN8R I would remain above 1.30, and the integrity of the fuel would be maintained during an actual tran-I sient under similar conditions.

I I .

I .

I I

ll  ;

I 3-11 l

I .

e

! FIG. 3.2.1 ROD WITHDRAWAL TRANSIENT WITH i FAST WITHDRAWAL RATE AT HFP:

NUCLEAR POWER VERSUS TME 1.4 i

4 1.2 - ,'

4 _.s N i < s

< E \

2 1- \

\

o

! z S 0.8 -

U

< \

5 \

h

, 0.6- }

3: \

o w o_

l AE REACTMTY INSERTION N6 0 4- RATE = 75 X 10-5 DELTA K/SECOND \

d n s l

z \

! 0.2 - \ Legend x :o a DYNODE m to m FSAR y $. C 0 . . . . . . . . . . z=0 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 r $' "

TNE N SECONDS f" E

m l

M M M M M M M M W W W W W M M M m M

~ ~

sua num nas uma amm num uma em nas me aus num em nun uma amm num FIG.3.2.2 ROD WITHDRAWAL TRANSENT WITH FAST WITHDRAWAL RATE AT HFP:

i PRESSURIZER PRESSURE VER' SUS TME i 2310 l

REACTIVITY INSERTlON l 2300- RATE = 75 X 10-5 DELTA K/SECOND .

/m 2290- / \

/ \

/ \

@ 2280-

. / \

w K / \

I U$ f 2270- \

/ .

Ym we ,

g '

\

W 2260- 7 l U \

$ /

g '

$ 2250-

< y /

5

\

I k /

2240- / g l

a 2230- /

DYNODE gyy I

FSAR o 5. ?

! 2220 -T' , , , , , , , , , vHows.e 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 l

f" I TIME. SECONDS i ~ .

{ l

• 1 ,

m

i FIG. 3.2.3 ROD WITHDRAWAL TRANSIENT WITH i

FAST WITHDRAWAL RATE AT HFP:

AVERAGE CORE COOLANT TEMPERATURE VERSUS TME 590 589-h.

Ed 588-g REACTIVITY INSERTION g

W RATE = 75 X 10-5 DELTA K/SECOND Q.

]I-- 587-i WZ t-p

's .

i d j 586- /

/ g ao ' \

O / \

i o / .

g /

I m 585- '

o ,

o , -

, y _ _ _ _

N 584-5 R

j 583- Legend DYNODE ggy ,

Es^!L _ j.j-l o 0:5 i ts i 2:5 5 35 i 4:5 s 5.5 ^"

IIU l TNE N SECONDS i ~

m l

1

]

E E E E E E E E E E E E E M

um muu em uma uma num aus em um em man uma em man mum ame um um aus FIG.3.2.4 ROD WITHDRAWAL TRANSIENT WITH FAST WITHDRAWAL RATE AT HFP:

DNBR VERSUS TIME 2.6 1 /

REACTIVITY INSERTION /

2.4 ~ RATE = 75 X 10-5 DELTA K/SECOND

/

/

' 2.2 - -

I

/

/

i 2-i x oo /

uZ h /

vi 1. 8 -

/

I i

/

~_

1.6 - -

j

~

^

1.4 - Legend COBRA 3m=

- wam FSAR _

y $. ?

" 0. 8 1.2 * '

i 75 h 2.'5 $ 3.'5 4 4.5 5 o 0.5 1 t- o w TIME IN SECONDS i 'o l C

m

NFU-033 I Revision 0  !

l March 14, 1986 1 Il I

i l'

. flEl I S

/ g l

l N

-- g I

i A I WW \

I l

d'y W n I ,# t E

I i 4 is al, i i i

i i 5 g a

9 88 s

) ~

z q i

i I

in g 1 8 d

M 1

1

~

I n

i i i i I

WNrON .O NOLD4U mod WIDW 3-16 g

uma num muu num num amm uma num mum em nas amm man uma ame uma um um um FIGURE 3.2.6 ROD WITHDRAWAL TRANSENT WITH SLOW WITHDRAWAL RATE AT HFP:

PRESSURIZER PRESSURE VERSUS TME 2450 s

2400- I i

t

n. 2350-ui j\

x /

s' \ '

0 / \  ;

10 /

w g 1300- \

gx ,

' \

\

n! \

\

l h m 2250-

\

w \

\

i I (

2200- 8SSNo-o

. /

! ) 0 N0 i

i- \

rsAB, _ ' f 1 215 0 .

go is e {,h

, i U fo 25 'O TBAE N SECONDS Z8" l ~

l

i s

! FIG. 3.2.7 ROD WITHDRAWAL TRANSENT WffH SLOW WITHDRAWAL RATE AT HFP-i AVERAGE CORE COOLANT TEMPERATURE VERSUS TIME

! S ll 1

1 502- 5 '

f I Lt.

, tJ S00- '

f

. /

/ \

i 588-

\

s \

M~

l

, m, u _ _ .

\

H SOO*

co o

.! O o I I na 582-

\

b O -

I 580-

\

h $7s-

>4 \

l Legend

{. S79 i ggg 1 N 2 ;i?

e o' w O b

' BME N SECONDS e,

m i

i I

g g g g g M M M W E E

ums num aus sum amm uma e um num aus um num aus uma men um um FIG. 3.2.8 ROD WITHDRAWAL TRANSENT WITH SLOW WITHDRAWAL RATE AT HFP:

DNBR VERSUS TIME 3.2 l

3-f i

l REACTIVITY INSERTION 2.8 -

RATE - 3.0 X 10-5 DELTA K/SECOND l l

2. 6 - -

g I

2.4 - g l

• l l m@ 2.2 - I io e

2-l i

1.8 -

I i

~_ t 1.6 ~__ -

/

~- -

s Legend

~

i 1.4 - ' / COBRA

- 575 G^"- -

NNT l 12 . . . . . .

35

. :r u, o g-o 0 5 10 15 20 25 30 40 TIME IN SECONDS =aw i

e

'o 5

i .

i l

l

NFU-033 Revision 0 March 14, 1986 3.3 UNCONTROLLED BORON DILUTION AT POUER I

3.3.1 Description of the Accident Reactivity can be added to the care by inadvertently feeding primary grade Water E into the RCS via the reactor makeup 5 i portion of the chemical and volume contr'ol system (CVCS). The opening of

. the primary Water makeup control valve l

provides makeup to the reactor coolant system which can dilute the concentration of the baron in the reactor coolant, thereby increasing the reactivity. In order for makeup water to be added to the RCS at pressure, at least one cFarging g pump must be running in addition to a g primary makeup water pump. Inadvertent dilution from this source can be readily terminated by closing the CVCS control va ee.

3 3.2 Summary of Accident Analysis Methodology The system response of the baron dilution transient at power was simulated using a system simulation code. DYN00E-P,(3) uhich includes the point kinetics model with reactivity feedbacks, models of pressurizer, steam generators and CVCS l components. The ONBR calculation was g performed using a modified COBRA IIIc-

, MIT(2) code. The dilution rate of this a transient is limited by the maximum flou rate of the charging pumps. The equiva-g lent reactfvity insertion rate used was 1 17 x 10 delta K/second based on a g i

conservatively high baron concentration E l

of 1500 ppm at power. With the reactor in manual control and if no operator action is taken, the power and temperature rise will cause the reactor to reach the overtemperature delta T trip setpoint. The acceptance criteria for this accident are that pressures in the RCS and main steam system do not exceed 110*/. of t he design pressures , and that g the fuel clad integrity is maintained by ~g l limiting the ONBR to a value greater than 1.3. -

I 3-20 1

l m NFU-0033 Revision O March 14, 1986 3.3.3 Results No plots were presented in the FSAR for comparison. However, results for this analysis are presented. The average vessel temperature throughout the I transient is shoun in figure 3.3.1. The reactor trip on overtemperature delta T was predicted at 51.7 seconds which is in I, agreement with 52.0 seconds stated in the FSAR(6). Figure 3 3 2 shous the RCS pressure history. During the transient.

both the RCS and steam system pressures I never reached values higher than 106*4 of the normal operating pressures throughout the transient. Figure 3 3.3 shows the I neutron flux. In Figure 3.3.4, the minimum DNBR during the transient was shown to be 1.470 uhich was higher than the design value of 1.30. In conclusion, the fuel cladding, RCS and steam systems would remain intact throughout'the baron dilution transient.

I .

I I

I .

I I

I I 3-21

) FIG. 3.3.1 UNCONTROLLED BORON DLUTION TRANSIENT:

VESSEL AVERAGE COOLANT TEMPERATURE VERSUS TNE 590 1

588-u 358.-

g .

< 4 l @58.-

1 a m-h 582-w m

k 580-5

?c 578- I

! Legend DYNOOE ggy NNE 576 , '

N 0 W 20 4O 5'O 6'O 70

1. E. O WE, SECONDS Z8" O

c

" " " " " 8W sum . aus uma sus ums mas aus sus e e e

em num num amm mum num num um amm um mum um -aus---amm uma amm um uns amm FIG. 3.3.2 UNCONTROLLED BORON DLUTION TRANSIENT:

PRESSURE VERSUS TIME i

! 2400 I 2350- f l

f p2300-m ui l bh '

O

] g 2250-

)

4 2200-i xxz i [

! 3.y l- \ Legend DYNODE * $. S i eow 2150 , , , , ' ' ##

l -

O m 20 30 40 5O 6O 70

] ~

TNE, SECONDS m

4 i

1 l

FIG. 3.3.3 UNCONTROLLED BORON DLUTION TRANSIENT:

i -

NORMALIZED NEUTRON FLUX VERSUS TIME f

12 i )

j i-o.a -

8 w o.s -

i g o.4-m

o.1 -

k Legend ggg

! - DYNODE 2$C i

o . . . .

"E8 k

20 30 40 50 60 70 eow

, O l TNE, SECONDS ,

1 ~

. m t E E ,

~ ~

mum mum uma em men mm um mum man mm man mas amm um num FIG. 3.3.4 UNCONTROLLED BORON DILUTION TRANSIENT:

DNBR VERSUS TIME 5

4.5 -

4-3.5 -

x a

3-T y 2.5 -

l

) 2-4 t5-i~ Legend ggg

( COBRA Nbi

' " $. 8 t

g9 jo 3'o 4o 5'o 8 TIME. SECONDS

[0U

.-. o i m i

i I

~

1 i

l NFU-033 Revision 0 March 14, 1986 3.4 FULL LENGTH ROD CLUSTER CONTROL ASSEMBLY OROP  !

1 3.4.1 Description of the Accident A situation can occur in which a rod cluster control assembly (RCCA) drive mechanism becomes de-energized. No longer supported, the RCCA vill' drop into the' core. This analysis is concerned with the dropping of a full length RCCA into the core.

A single dropped full length rod assembly or assembly bank is detected by the following.

1 Sudden drop in core pouer level 2 Asymmetric power distribution

3. Rod bottom light (s)

4. Rod deviation alarm

5. Rod position indicator The importance of this accident lies in the possibility of a power overshoot resulting from the action of the automatic rod controller. Westinghouse design uses a dual controller which limits the pouer overshoot to a maximum of tuo percent. The essential feature of this rod controller is that it terminates rod withdraual well before the primary coolant average temperature is restored to an equilibrium condition. This not only minimizes the power overshoot, but also ensures extra margin to departure from nuclear boiling, ONB.

3.4.2 Summary of Accident Analysis A single RCCA drop was assumed in this analysis. The transient core response was simulated using the system simulation code, DYN00E-P(3). The core DN8 response was calculated using a modified COBRA IIIc-MIT(2) code. The DYNODE-P code simulated the neutron kinetics, reactor coolant system, pressurizer pressure, related relief and safety valves and steam generators.

Other assumptions made in this analysis

! include a zero moderator density reactivity coefficient corresponding to the BOL condition and the least negative Doppler feedback. This results in less 3-26

I NFU-0033 Revision O March 14 1986 reactivity feedback during the automatic controlled return to power strengthening the possibility of power overshoot. The rod drop was modeled as a ramp insertion of negative reactivity totalling the dropped RCCA reactivity worth of -0.25 percent delta k/k. Rod control was enabled in order to establish a power overshoot possibility.

3.4.3 Results A single RCCA drop with automatic rod controller was simulated for this transient. Figure 3.4 1 shows the DYNODE-P core heat flux prediction for a I ramp reactivity insertion due to the rod drop. DYNODE-P's rod controller estab-lished stable conditions following the rod drop faster than that predicted in the FSAR. This is also shown in Figures 3.4.2 and 3.4.3 for the average tempera-ture and pressurizer pressure response.

In Figure 3.4.4 the change in the departure from nucleate boiling ratio (DNBR) predicted by COBRA is compared to the FSAR.

I experienced.

A small pouer overshoot was Figures 3.4.5 through 3.4.8 shou the results of the core heat flux, change in the core average temperature, pressurizer pressure and change in DNBR for the same transient without the rod controller effects.

3-27

l i

! FIGURE 3.4.1 DROPPED RCCA TRANSENT:

i CORE HEAT FLUX VERSUS TIME i

i uo 1.0s- WfiH AtROMATic CONTROL L

l

, 1-a ,

z \ ,

w O o.es - I /

, z I o /

I i w E 1 /

!  ; o.so-f m '

I /

II- o.as- I/

l \ \l l h w o.ao-m

! 8

!' o.7s- Le9end INNODE SE 5' %

M<c EsA!L ,_ g;a r- "

0.70

%8" o

20 40 so ao ion 12 0 140 soo iso 200

THE N SECONDS 'o i

G l

l i

!M M M M M. m a g

m uma e am uma amm user uma seu um sums smur FIGURE 3.4.2 DROPPED RCCA TRANSIENT:

CHANGE IN AVERAGE TEMPERATURE VERSUS TIME 10 WITH AUTOMATIC CONTROL 5-0

t ,

U1 -

d ~

i La O

td ,

w w k i

@ V-i --2 f d Z

I O

f  :

Legend DYNODE gyy FSAR ok?

-30 , , , , , , , , ,

age O 20 40 60 80 10 0 12 0 140 160 180 200 eow i TIME IN SECONDS m

j i

FIGURE 3.4.3 DROPPED RCCA TRANSIENT:

PRESSURIZER PRESSURE VERSUS TIME 2280 2240-i s

/

2220- WITH AUTOMATIC CONTROL '

\

\

5 2200- -

, E \ ,-

! E \ ,' '

,M 2180- \ /

, /

I o" /

.l \ /

2180 - g l f

\ /

s-2MO- -

I

! 2i20- Legend ovwoot 2100 , , , , , , , , ,

l 0 20 40 80 80 100 12 0 14 0 180 180 200 75 O TIME IN SECONDS ggw

'o a

M M M M M M M M, M M M M M M M M M M M

mm e sum um amm uma amm em imm men sum um amm num FIG.3.4.4 DROPPED RCCA TRANSIENT:

CHANGE IN DNBR VERSUS TIME 0.7 0.6 -

'l 0.5 - WITH AUTOMATIC CONTROL i.

4 O' O.4 -

m z

Q z

w]

mo O.3 - -

\

g / \

j r I N O 0.2 --

l N

\

l 0.1 -

d \

] \

\ /

} }

g N 's

/ ' -

-TN Legend l 0.0 N j s

~

- g N _ __ __ 908.BA - !E is l~

i

\_s FSAR 2$.?

xmo

! - 0.1 , , ,

30 40 50 60 70 g$8 20 10 f""

O TIME SECONDS O i

, 8

i i FIG.3.4.5 DROPPED RCCA TRANSIENT:

CORE HEAT FLUX VERSUS TIME t.10 l

f 1.05- WITHoUT AUTOMATIC CONTROL i Y z 1-

3 .

o '

Z u.

o 0.95- I Z lI

o i b I i yg o.90- g I b- g o.a5- I' '

i sI l

, p 0.ao- g

o j O \

\ Legend o.75-

' DYNODE @j'is s .

E ! ^!!. _ h.$ $

's __

o 2o lo s'o a'o ido rio tio 150 ido 200 g h'U TIME; SECONDS ,

O p s imus num muu samt aus ums e mui uma muu mas amm mim amm ums

um man sum sue mas semi um uma e man me uma em um uns - sua m FIG.3.4.6 DROPPED RCCA TRANSIENT

CORE AVERAGE TEMPERATURE CHANGE VERSUS TIME O -

N W'ITHOUT AUTOMATIC CONTROL

-s- N

% ~

w

, g N N

e-y ,

cr N '

y N s

j N w N o

w<

s 6 @ '

"R s

\

w h

U

\

i I s s

i w '

g 1 0

Legend 3

DYNODE $$$

n<c 3 FSAR g. [ 4 l "' "

-40 . .

[@ "

10 15 20 25 30 35 40 45 0 5 TNE. SECONDS 'o

C I

l

FIG.3.4.7 DROPPED RCCA TRANSIENT PRESSURIZER PRESSURE VERSUS TIME 2250 2200- WITHOUT AUTOMATIC CONTROL N

N N

2150 - N N

m n N N

2100 - g x N N

1

]

y 2050- g wm s

I Q N w g N

# N La 2000- N N x x N D s

. M s

]m1950- N j s 0-4 1900- N I N N

! N Legend 1850-

' DYNODE FSAR ggy N < c:

' 1800 , , , , , , , ,

0 5 to 15 20 25 30 35 40 45 @. $' $

! TNE, SECONDS ra s' $

,a i p E E E

FIG.3.4.8 DROPPED RCCA TRANSIENT:

CHANGE IN DNBR VERSUS TIME 1

0.8 -

m ,

Q)

~

0.6 - '

O ,-

sE

/

ww ao /

w /

@g /

4 /

r 0.4 -

U /

/

/

/

/

0.2 - WITHOUT AUTOMATIC CONTROL j

' Legend COBRA 3mz

/

/ [ GAR, _ f h, l O r , , ,

umo 5 10 15 20 25 g g-0 TIME SECONDS paw l

00 i .

I B

NFU-033

Revision 0 March 14, 1986 m l

g 3.5 EXCESSIVE HEAT REMOVAL DUE TO FEEOUATER CONTROL VALVE MALFUNCTION 3.5 1 Description of the Accident Excessive heat removal from the primary l system could be caused by a reduction in a feeduater temperature or excessive feeduater flou to the steam generators. g i This section will concentrate primarily on the excessive feeduater addition E

transient. The excessive feeduater flou could be caused by a full opening of a feeduater control valve due to feedwater control system malfunction or operator error. At power, this excess flow causes 3 a greater load demand on the RCS due to g increased subcooling in the steam genera-tor. Under automatic control, this increased load demand is balanced by the rod control!'r action. Reactivity is inserted to balance the core power to the increased load demand reducing the margin 3 to ONB. The overpower-overtemperature 3 trip protection is designed to Prevent any power increase uhich could lead to a DNBR less than 1 30.

3.5.2 Summary of Accident Analysis Methodology The excessive heat removal due to feed-water control valve malfunction was simulated using a system simulation code.

DYN00E-P.(3) The core ONBR was calcu-( lated using a modified COBRA IIIc-MIT.(2)

The OYN00E-P code simulates the core neutron kinetics, the pressurizer pres-sure, safety and relief valves, pre- E 5

, ssurizer spray and steam generator system. The feeduater control valve was e assumed to malfuction resulting in a step increase of 250 percent of nominal 5 feeduater flow to one of the steam generators. The reactor uas assumed to be operating at full power with automatic control and end of life conditions. This would give the largest reactivity feed- g back and result in the greatest power increase.

5 I

I 3-36 I

NFU-033 Revision 0 353 Results The power increase and the associated temperature changes in the primary system I are compared with FSAR(6) results in Figures 3.5.1 through 3 5 3 Figure 3.5.4 shows the pressurizer pressure I response which is more profound in OYN00E-P than in the FSAR.

generator level rises until the feedwater The steam I flow is terminated as a result of the high-high steam generator level trip causing a turbine trip and then reactor trip. The DNBR for the feeduater control valve malfunction transient is well above i the limiting value of 1.30 as shown in

, Figure 3.5.5.

lI

'I I

I

. I I

'I l

I I

I I

3-37 I .

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

. . , ,__,,__ ____ _ _ _ _ _ . _ . _ _ _ _ _ , _ _ _ . . _ . m___. . _ __.-

1 FIG.3.5.1 FEEDWATER CONTROL VALVE MALFUNCTION TRANSENT:

i FRACTIDN OF NOMINAL NEUTRON FLUX VERSUS TME i

i 1.2

"^

/ , -

\

E-f _-

, \

l 1-I N

l 5

O Z 0.8 -

! LA.

1 O Z

l O F

k O.8-l w 1 1M i *d I Z 0.4 -

E

! Z 0.2 -

Legend

! omooE xxz FSAR $$E o

0 2

4 8 8

era 10 12 14 18 18 20 g7" l, TME, SECONDS po i' ~

e

$ m l

i g

l um um nur en aus em sur aus um em uma mm. me aus um .uma sua sum ums FIG.3.5.2 FEEDWATER CONTROL VALVE MALFUNCTION TRANSIENT:

l CHANGE N RCS AVERAGE TEMPERATURE VERSUS TME 4

i g 2-

, w l

5 W '

s

'- ~ ~

' \

0 '

~- N

\

r

, U

\ g -

W

! YA a$e j g Z i <

O Legend ovm

""~

r i -8 . , , , , , , , ,

! 0 2 4 6 8 10 12 14 16 18 20 # [. w 1 TNE. SECONDS y8" C

i a

\

1

l

! FIG.3.5.3 FttuWATER CONTROL VALVE MALFUNCTION TRANSENT:

i CHANE N RCS DELTA T VERSUS TME

! ,o i - -%

i 5- -  %

j

- \

! __ \

4 0 i w t- .

_io .

! b e

m Z -ts -

IW

, 40 CZ

)

1 1 DYNODE i zwz

• * "3 i FS. A_R_ __ n <: c 1 -35 , , , , , , , , , or

,' O 2 4 4 8 10 12 14 18 18 20 *$

l TBAE, SECONDS %8"

! 'o

~

• m Ch i

( ,W M M M M M M M M W M W W W M M m m M

am sum uma sus aus an ums an as aus um ums um amm um man uma em ams FIG.3.5.4 FEEDWATER CONTROL VALVE MALFUNCTION TRANSENT:

i CHANGE N PRESSURIZER PRESSURE VERSUS TIME 200 1

15 0 -

i g E

W a .

M 10 0 -

g w

l [

4 5

h 50- \

/

. yB / \

.m w / \

k o

' \

il; w

~~

' ' ' ) - - _. - 1 h

i 5 ~50-i Legend

ovuooc ,,n i

73 ,

402

-e0 g ,

E U's i

4 a . " " 20 st TaE. secoNos ,n u

m i

) FlG.3.5.5 FELUWATER CONTROL VALVE MALFUNCTION TRANSIENT: .

i DNBR VERSUS TIME

) 5.5 1

4 5-i

) 4.5 -

l <-

i 3.5 -

I- 'f4. Ea w 3-i.

2.5 -

. /

/

2- /

/

' Legend i.s - -

COBRA xx=

mom n<c i FSAR 0*'

4

--- :r m o i %8d o i i i io 4 A is is 20 TIME SECONDS 'o e

I Ch i.

m M M e M M M M M W M M M M M M m m _ _ _ _ _ _ _ _ _ _ _ _ _

m

NFU-033 Revision 0 l March 14, 1986 3.6 LOSS OF EXTERNAL LOAD 3.6 1 Description of the Accident Loss of load can result from loss of I external electrical load or from a turbine trip. In either case, the offsite' power is available for the continued operation of plant components I such as the reactor coolant pumps.

During a turbine trip, the reactor would be automatically tripped if the power was I above 10 percent of rated power. The automatic steam dump system would accom-modate the excess steam being generated.

I If the main condenser was not available.

the excess steam generated would be dumped to the atmosphere and main feed-I water would be isolated. In this case feeduater flow would be maintained by the auxiliary faedwater system.

During a loss of external electrical load without turbine trip, no direct reactor trip dould be generated. The plant would be expected to trip from the reactor I protection system.

3.6 2 Summary of Accident Analysis Nethodoloqv The total loss of load transients were analyzed by employing the computer code I OYNODE-P(3) which included a point kinetics model coupled with the simula-tions of reactor coolant system. pres-I ,

surizer, pressurizer relief and safety valves, pressurizer spray, steam genera-tar and steam generator safety valves.

The ONBR analysis was done using the modified COBRA IIIc-MIT code.(2)

The initial reactor power and reactor coolant system temperatures were assumed I at their maximum values consistent with steady state full power operation inclu-I ding allowances for calibration and instrument errors. This resulted in the maximum power difference for the load loss. The initial reactor coolant system I pressure was assumed to be at a minimum value. This resulted in the minimum margin to core protection limits at the initiation of the accident.

I 3-43

NFU-033 Revision 0 March 14, 1986 The total loss of load is analyzed for the beginning of life conditions only.

The moderator temperature coefficient of l zero and a conservative Doppler power a coefficient were employed. Two cases were analyzed for this transient.

Case A:

~

Full credit was taken for the effect of pressurizer spray and power operated relief valves (PORV's) in reducing or limiting the coolant pressure.

Case B: No credit was taken for the effect of pressurizer spray and PORV's in reducing or limiting the coolant pressure. Pressurizer heater operation was assumed since heater operation maximizes pressure.

In both cases, no credit was taken for the operation of the steam dump system or the steam generator pouer operated relief B valves (PORV's) . The steam would be g

. released through the SG safety valves to limit the steam pressure on the secondary side to the setpoint value. Main feed-water flou to the steam generators was assumed to be maintained throughout the transient.

3.6.3 Results Case A: Figures 3 6 1 through 3.6.5 show the comparisons of DYNODE-P results with those of the FSAR for the transient which took full credit for pressurizer spray g and the operation of the pressurizer g PORV' s . Table 3.6.1 gives a comparison of the sequence of events between OYN00E-P and the FSAR.(6) l The neutron flux predicted by OYN00E-P matched the FSAR results as shown in E Figure 3.6.1 The trip actuation time 3 predicted by DYNODE-P was about one tenth second earlier than that of the FSAR.

The OYN00E-P predicted neutron flux after shutdown was lower than that of the FSAR.

The pressurizer uater volume inventory I

I 3-44 I

I NFU-033 Rnvicion 0 March 14, 1986 I predicted by DYNODE-P was higher than that of the FSAR, as shown in Figure I 3.6 2 The care average temperature predicted by DYN00E-P was in good agree-ment with the FSAR, as shown in Figure The temperature drop predicted by I 3.6.4.

OYN00E-P occurred earlier than the FSAR predicted; this was caused by the faster core shutdown predicted by OYNODE-P. The

, I ONBR predicted by COBRA was atuays above 1.3, as shown in Figure 3.6.5. Conse-quently, there would be no fuel damage.

Due to the operation of the pressurizer I spray and PORV's , the primary system pressure was always below 2550 psia, which is well below the RCS pressure design limit of 2750 psia. Therefore, I the reactor pressure vessel integrity would be maintained.

Case B: The results of the loss of load transient without pressurizer spray or PORV operation are shown in Figures 3.6.6 through 3.6 10. The comparison of the sequence of events Letween OYN00E-P and the FSAR are shown in Table 3.6.2 The pressure responses are shoun in Figure I 3.6.8. The OYNODE-P results match the FSAR results fairly well. The maximum pressure was louer than in the FSAR. As I shown in Figure 3.6.7. the pressurizer water volume predicted by DYN00E-P was higher than that predicted in the FSAR.

I The average core temperature is shown in Figure 3 6 9. The neutron flux was in good agreement between DYN0DE-P and the FSAR, as shoun in Figure 3.6 6 In conclusion, the DYNODE-P predicted RCS pressure results for both cases were 1 below the reactor vessel design limit so the vessel integrity would be maintained.

The COBRA IIIc-MIT predicted DNBR was atuays above 1.30 so the fuel integrity I would also be maintained.

ll I

I 3-45 I .

NEU-033 Revision 0 March 14, 1986 I

TABLE 3.6 1 TIME SEQUENCE OF EVENTS FOR LOSS OF EXTERNAL ELECTRICAL LOAD UITH PRESSURIZER SPRAY AND PORV's AT BOL Event Time (seconds)

OYN00E FSAR Loss of electrical load 0.0 00 Initiation of steam release from steam generator safety valves 9.1 90 Overtemperature delta T 9.0 9.1 Rods begin to drop 11 0 11.1 Minimum DNBR occurs See Fig. 3.6 5 Peak pressurizer pressure occurs 12.2 12.5 I

I I

1 I I

I I

l I

l I

I 3-46 I

I NFU-033 Revision 0 March 14, 1986 TABLE 3 6.2 TIME SEQUENCE OF EVENTS FOR LOSS OF EXTERNAL ELECTRICAL I LOAD UITHOUT PRESSURIZER SPRAY AND PORV'S AT BOL I Event Time (seconds)

OYNODE FSAR Loss of electrical load 0.0 0.0 Initiation of steam release from 8.3 90 steam generator safety valves High pressurizer pressure reactor trip point reached 6.0 60 Rods begin to drop 8.0 8.0 Minimum DNBR occurs See Fig. 3.6.10 E Peak pressurizer pressure occurs 8.0 9.0 I

. I

'I I .

I I

I

.I I 3-47

l I FlGURE 3.6.1 LOSS OF ELECTRIC LOAD TRANSIENT:

NEUTRON FLUX VERSUS TME 1.4 i

1.2 -

j i a

E -__

l' 3 z

i-

\

Ln-i o .

,' z o 0.s -

! P

o l

m

w ti-j i* g o.s-I LL.

z

- O

! m 0.4 -

b z

s O.2 - W11H PRESSURrztR spgAy ANO PORV AT BOL ~

- Legend i

- - ~ _ __ _

DYNODE rsAn {7g l

s ,b ,g 25 e74 O

TNE N SECONDS [$

~

i O W M g g *

• *

• M M ma y

FIGURE 3.6.2 LOSS OF ELECTRIC LOAD TRANSIENT:

PRESSURIZER WATER VOLUME VERSUS TIME 1350 1300-

/ _ N

\

\

h 1250-L.- \

i \

1200-

. N F ~

l 115 0 -

a '

h 1100-i b '

!0 -

1 1050- ' -

- WITH PRESSURIZER SPRAY AND PORY AT BOL

, coo. Legend

- DYNODE xxz

\ FSAR $ $

i 950 . . . . h5 0 5 to 15 20 25 g 7 w" i

TNE N SECONDS  ?"

~

CD m

i FIG.3.6.3 LOSS OF ELECTRIC LOAD TRANSENT:

i PRESSURIZER PRESSURE VERSUS TNE 2600 i

i 2500-

/

'g

! / \

1 /

, \

1 h 2400-4 g 1 '

y - \

$ \

M g 2300- f \

N ' \

i 1 '

- g

! wE

g 2200- \

m WITH PRESSURGER SPRAY AND PORV AT BOL

\

m \

210 0 - 'N

! s s

2x0- Legend DYNODE x :o z FSAR $ $

1900 , , , ,

o 5 30 ts 20 25 T" i

THE SECONDS f"

~

i

=

)

W W W W W W W W M M M W W W W W m W W

aus sus seu uun sum an amm aus sur aus sus as em sum uma sum um ens see FIGURE 3.6.4 LOSS OF ELECTRIC LOAD TRANStENT:

AVERAGE CORE TEMPERATURE VERSUS TIME 620 610 -

t.a_

!O o 600-a f s

- s a

g -

q 590- ,

, s s

~

' s s '

T " 5s0-W 8

W 570- wl1H PRESSUR12ER SPRAY AND PORv AT BOL O ,

l5 R

560- Legend DYNODE f.SAR_ _

((

$[$

550 0

• (o is 2o 25 ~Ed

,a a

THE N SECONDS C

2

l FIGURE 3.6.5 LOSS OF ELECTRIC LOAD TRANSIENT:

I DNBR VERSUS TIME i

8 7-l 6- -

5-E

"a$

w o -

j 4_

3-l 3 -

l '

2- WITH PRESSURIZER SPRAY AND PORV AT BOL ,-

[ggggd

]

j_ COBRA ggg FSAR E$?o

r o' S l o i i d 5 10 12 A is ts 20

! TIME IN SECONDS o i

e l

E O E E E E E E E E E E E E E W E W

NFU-033 Revision 0 March 14, 1986 5,

I ,

I f!E I

, =

l l l 4 1 5 i g 'g 2 h< 8 i

I I %h h I

lI l o"h S

a 1 i a, l Q>X e I

eE3 ____

.,1 I $,!

u j b y 1 -=

I H R

u_

I I U i i i i wMHON .30 NOuGY&fXAE NOWGN 3-s3 l

. FIG.3.6.7 LOSS OF ELECTRIC LOAD TRANSIENT

PRESSURIZER WATER VOLUME VERSUS TIME l 1300 l

1200-n /%

H /  %

) /* \

2 s 3 / N

' \

l 110 0 - -

, e '

- ~

s l s s j N< -

"D WITHOUT PRESSURIZER SPRAY AT BOL 'N s m

^

I w $ 1000-m l0 lE 900-l Legend i DYNODE @$$

FSAR h

800 . . . .

M' "

0 5 to 15 20 25 [@"

i 'e e

i M M M M W W W W W W W W W W W W W W W

an an um num amm as amm aus mer aun as aus saa en am uma em aus mas FIG.3.6.8 LOSS OF ELECTRIC LOAD TRANSIENT:

PRESSURIZER PRESSURE VERSUS TME 2600 p'\

\ .

2500-

/ \

/ \

\

g 2400-

a. ~

i W. \

o /

g

$ 2300- /

w / \

E ,' \

5 s YN 2200- g

$$ \

u s

s

( 210 0 - ~

~

~

~ ,

! 2000-Legend WITHOUT PRESSUR12ER SPRAY AT BOL DYNODE

~

EsA!!. _ g ;$. 7 1900 . . . . :rca o 0 5 10 15 20 25 gy[

TRAE. SECONDS Po I

o

~

e e

~

FIG.3.6.9 LOSS OF ELECTRIC LOAD TRANSIENT:

AVERAGE CORE TEMPERATURE VERSUS TIME ~

610 g, 600-4 D

'd b

/ \

O. #

2 g w /

\

F- Sgo- /

w / N YT / \

.8w /

s

' s N

- N O . _ _ _

N N

6 .

4 580- s N ,

Legend i v.ahouT ?RESSUR12ER SPRAY AT BOL DYWDE -

.,3 y y fSAR_ _ ,'

,U o

o i 3o S 2o 25 TNE. SECONDS a G

um e e e e sus a mus. m em m sur m m aus aus as aus use

ma sus em um aun as aun as sus em aus ums em aus aus som um ama mas FIG.3.6.10 LOSS OF ELECTRIC LOAD TRANSIENT:

DNBR VERSUS TIME 8

7-l J -

6-l

\

j 5-m

, W CD i 8Z

! $Q 4-i i

1 3-I -

j - __

2 2- -

- Legend l -N~

~ ~

COBRA -mz j ET m m WITHOLTT PRESSURIZER SPRAY AND PORV AT BOL

"<C

! FSAR '

0$o 14 16 18 20 eow k b d 10 12

! O 2 f"" o TIME. SECONDS I

i

)

_ _ _ _ _ _ _ _ _ _ ^ - - - - - _ _ - _

I NFU-033 Revision 0 March 14, 1986 3.7 LOSS OF NORMAL FEEOUATER 3.7 1 Description of the Accident A loss of normal feeduater can be caused l by a feeduater pump failure, valve 5 malfunction or loss oT offsite AC power.

This' accident would result in a reduction a of the heat removal capability of the secondary system.

g If the reactor were not tripped during the accident, core damage could occur from a sudden loss of heat sink. An alternative supply of feedwater must be supplied to the plant or residual heat following reactor trip 3 would heat the primary system uater to the point where water relief from the g

pressurizer would occur. Significant loss of water from the primary system could lead to core damage. Since the plant trips well before the steam genera-tar heat transfer capability is reduced, the primary system variables never g

g approach a DNB condition. The foliouing automatic plant res'ponses provide the necessary protection against a loss of normal feeduater:

1 Reactor trip on lou-low water level in any steam generator

2. Reactor trip on a steam flou-feeduater flow mismatch'in coinci-dence with lou water level

3. Two motor driven auxiliary feed- B water pumps uhich are started on: 5

a. Low-lou level in any steam generatcr.

b. Trip of all main feeduater pumps.

c. Any safety injection signal.

d. Loss of offsite power.

e. Manual actuation.

4. One turbine driven auxiliary feeduater pump which is started on:

I 3-58' t

NFU-033 Revision 0 March 14, 1986

a. Low-low level in any two steam generators.

b. Undervoltage on any tua reactor coolant pump busses.

c. Manual actuation.

.In the event of a loss of offsite power. the motor driven auxiliary I feedwater pumps that are supplied by the diesels and the turbine-driven pump that utilizes steam from the secondary system are I available. Both pump types are designed to start within one minute after the initiation signal, even I if a loss of all AC power occurs simultaneously uith a loss of normal feeduater. The auxiliary I pumps take suction from the auxil-iary feedwater storage tank for delivery to the steam generators.

I ,

3.7.2 Summary of Accident Analysis Methodolo9y The loss of normal feeduater transient is simulated using the DYN00E-P code. The I code. simulates the core kinetics, reactor coolant system including pressurizer, steam generators and feeduater systems.

I The program computes pertinent variables including the steam generator water level, reactor coolant temperature and I pressurizer water volume. Major assump-tions are:

1 The initial steam generator water I level (in all steam generators) at the time of the reactor trip is at a conservatively low level.

2. The plant is initially operating at 102 percent of rated power.

3. Reactor Coolant System (RCS) pumps are tripped at the time of the accident initiation to simulate a I loss of AC.

4. A conservative core residual heat generation rate based on long term i

I operations at the initial power level, is assumed.

II l 3-59 I .

NFU-033 Revision 0 March 14, 1986 Only one motor driven auxiliary I

5 feeduater pump is available one minute after accident initiation.

6. The steam relief from the steam generator is assumed through safety valves. No credit is taken for the power operated relief valves or condenser dump valves.

7. The initial reactor coolant average temperature is 4'F lower than the nominal value, since this results in a greater expansion of the RCS coolant during the transient and higher pressurizer water level in the pressurizer.

3.7.3 Results Initially, the water level in the steam generators would fall due to steam flow through the safety valves and the reduc- E tion of the steam generator void fraction g caused by pressurization after the turbine trip. One minute following the initiation of the steam generator lau lou uater level trip, the auxiliary feeduater pump was automatically started reducing the rate of steam generator uater level decrease. The FSAR predicted that at no 3

3 time was the tube sheet uncovered in the intact steam generators receiving auxili- g ary flou. In Fig. 3.7.2, it can be seen that the DYN00E prediction of this uater 5 level is above that of the FSAR: in other words, we predict this tube sheet to E remain covered also. The RCS coolant B water is not lost through the pressurizer relief or safety valves this is shown in a Fig. 3.7.3. The reactor coolant tempera-ture does not rise much higher than the g

initial value during the transient as shoun in Fig. 3.7.1. If the initial power is less than 102 percent rated power and the auxiliary feedwater CaPa-city is greater than that of one motor 3 driven pump, then the result vill be a g higher water level in the steam generator and increased margin to water relief from the RCS systemz Results of the analysis show that a loss of normal feedwater does not adversely affect the core, the RCS nor the steam system. The uater level in g l the steam generators receiving feeduater 3 l

1s maintained above the tube sheets.

3-60

ass an en nun uma um aus e um en am aus em se am um em aus um FIG.3.7.1 LOSS OF NORMAL I-EEUWATER TRANSENT:

CORE AVERAGE TEMPERATURE VERSUS TNE 680 ,

660-tg

! E '

i 3 640- i

!' k

[li a.

b 620-e4 /

ww /

Pc /

g 600-j l O m /

l o ,

/

580 Le@M f DW E ggg

~

FSAR y$?

560 7mo e$U i

9 )00 2dOO 3dOO 4dOO SdOO 6dOO 7000 TWE. SECONDS '

~

e s .

FIG.3.7.2 LOSS OF NORMAL HEDWATER TRANSIENT:

STEAM GENERATOR WATER LEVEL VERSUS TIME 40 I

STEAM GENERATOR p 33 WITH AUXILIARY FEE 0 WATER #

s

/

30-

. s d 25- g - e l g p l W /

I e

\

E 20- \ p

/

wy /

5 l 65 z is. g g STEAM GDIERATOR 2 WITHOUT AUXILIARY FEE 0 WATER b ~

$ \

\

5- g

. s y M- Fr N DYN00C E$.?

O ,

1000 2000 3000 4000 5000 8000 7000 3ae eow 0

^"

TNE, SECONDS ,

a e me e e um as e us, una es as e e e e as as aus me

a . .

I NFU-033 Revision 0 March 14, 1986 I

I i

I }l%

I o n

I w

-f*

I b's

<F-i E CA 3xb .-

i I l I

1 _

I f>o W l

$0 g

~$ 8 se a>

5 I 8 m

I 2% I og EJ l3 b I 9$ e,,

o l t

o ml5 gg -

I ms (A E

- N N

-8 I bhm s s

N rq W TD \ 'i o I E

I i . . ,

(

7- o I  !  !  !  !  ! i i G OI800 '3lTn0A M31VM H3ZIMOSS3Md 3-63 l __

.,_._ _,_.,. ,, , ,y, . . . _ - _ . _ , , , , , , _ _ . . . _ . . . . _ _ _ _ _ - . _ . _ , . ._

NFU-033 Revision 0 March 14, 1986 I

3.8 LOSS OF REACTOR COOLANT FLOU - PUNP TRIP 3.8.1 Description of the Accident A loss of reactor coolant flow may result from occurances such as an electrical failure in a reactor coolant pump or a fault.in the power supply busses. The immediate effect of a decrease in flou is g

, an increase in reactor coolant g i

temperature. Without a prompt reactor trip, this could result in departure from nucleate boiling (DNB) and eventual fuel damage. Three such reactor trip signals are provided for mitigating the loss of flou accident. These are:

1 Undervoltage or underfrequency on reactor coolant pump power supply busses.

2 Lou reactor coolant flou.

3. Pump circuit breaker opening. 3 Tuo loss of flow cases are considered in 3 this analysis. The first is a complete loss of flou. This results from all four 5

pumps being shut down at the same time uithout restarting. Due to hydraulic inertia of the fluid and the pump motor flywheel, the coolant flou experiences a coastdown effect. The reactor finally 3 trips on the undervoltage signal as stated in the FSAR.(6) 3 The other case considered is the partial loss of flow case. This is a situation

, in which two of the four pumps are shut doun at the same time allouing for coastdoun on only tuo loops causing an g

eventual low flow reactor trip.

3 I

I

I I

3-64

l 5

NFU-033 Revision 0 March 14, 1986 I 3.8 2 Summary of Accident Analysis Methodology I These transients were performed using the DYNODE-P code (3) to simulate the system response. The code calculated the core I power, core flou and heat flux during the accident. The COBRA IIIc/MIT code (2) was then used to calculate the departure from nucl~eate boiling ratio (DNBR).

In the preparation of the input for bqth DYN0DE-P analyses, the follouing I assumptions were made consistent uith the assumptions of the FSAR.

1 The moderator density reactivity coefficient is zero.

I 2 The most negative Doppler reactivity coefficient is used.

For the total loss of flow case, all four pumps are tripped allowing I the core flow to coast down. In order for the undervoltage signal to occur, a manual trip is input at I the time stated in the FSAR analy-sis. For the partial. lass of flow, only tuo pumps are tripped, thereby initiating a flou coastdown. A lou flow trip signal is created when I the loop flou drops to a fixed fraction of the initial value.

3.8.3 Results For the complete loss of flow. DYNODE-P I ,

predicted the neutron flux, core flou and heat flux as shown in Figures 3.8.1, 3.8.2 and 3.8.3 respectively. The core flow coastdown predicted by DYNODE-P matched the FSAR results to uithin three percent. The departure from nucleate I boiling ratio (DNBR) prediction by COBRA IIIc-MIT calculation is shown in Figure 3.8 4. The minimum DNBR is greater than ,

1 30, so the fuel cladding integrity a

g would be maintained and this accident would not violate any safety limits.

I t

lI 3-65 I

NFU-033 Revision 0 March 14, 1986 For the partial loss of flow, DYN00E-P predicted the core and faulted loop flou, neutron flux and heat flux as shown in g Figures 3.8.5 and 3.8 6. respectively. 3 The flow coastdown prediction by OYN00E-P differed only by about four percent for g the faulted loop. Figure 3.8.7 shows g that the DNBR prediction for a partial loss of flow by COBRA IIIc-MIT is greater than 1 30 at all times.

l A general chronological event table for the tuo transients is presented in Table 3.8 1 I

I I

I l

l B

e I

I I

I I

I 3-66 E

I NFU-033 Revision 0 March 14, 1986 TABLE 3.8 1 TIME SEGUENCE OF EVENTS FOR LOSS OF REACTOR COOLANT FLOU Accident Event Time (seconds)

DYN0DE FSAR l Partial Loss of Flow Coastdown begins 0.0 O.0 Low flow reactor trip 17 1.26 Rods begin to move 3.2 2 76 Minimum DN8R occurs 3.5 3.7 I Complete Loss of Flow Coastdoun begins 0.0 0.0 -

Rod Motion begins 1.2 12 Minimum DNBR occurs 2.1 27 I

I I .

I l I  ;

1 I l I

I l 3-67

..y.-. ~,,v_, r - - - . . _,. , , . , . .

4 4

FIG.3.8.1 COMPLETE, LOSS OF FLOW - PUMP TRP TRANSENT:

1 NEUTRON FLUX VERSUS TME 1.2 1- m g

d N

O s -

z o.8 - \

g

' z \

o p \

l y o.e- \

wk g ix \

l "$ \

z o.+ -'

o 0;

\

I 0.2 -

_________________ Legend DYN00E

. gy$

o 0

'  ;  ;  ; i i i 5 b "

5$.5 TnAE, SECONDS h0"

~

a E _ _. _ - _______ -

am aus aus aun amm um aus aus em ' sus em aus mas am sum uma uma amm use FIG.3.8.2 COMPLETE LOSS OF FLOW - PUMP TRP TRANSENT:

CORE FLOW VERSUS TME 1.1 t-N J \

<C \

f 0.9- \ -

O \

! Z N 4

b. s NN O

0.8- N z s 9

W s '

O  %

N ta h

G-N i

i 0.7 - N

• g  %

O N d s k h3 0.6 -

i N l O N i '

~

~

0.5 - s_ Legend DYNOOE @$$

- ts < c f.SAE. -

h$b 0.4 0

i

' j j 4

5

=

8 7 8 9 8 -E0 "P

TBAE. SECONDS 'o O

.a .

e

s FIG,3.8.3 COMPLETE LOSS OF FLOW - PUMP TRP TRANSENT:

HEAT FLUX VERSUS TME

)

'.2

,_ s i N N

i a N 4 \

' g -

N O 0.8 - \

z N

u. N o N Z N

9 '

s H

g 0.8 -  %

Y <

iO E 's

g 's

's s

! d 0.4 - '

o.2 -

Legend DYNODE $$$

n<c r_SA,R_ _ @ ['$

o  ;  ; i i i i i i i io TBAE, SECONDS F o5 C

mas amm aus sus sus em ama mas aus M WB m EEN W W W W M

E E E E E E E E E E E E E E E E E E E FIG.3.8.4 COMPLETE LOSS OF FLOW - PUMP TRIP TRANSIENT DNBR VERSUS TIME 2.6 2.4 -

2.2 - ,

1 i

2-

! Q' w$

bO

~ ,

1B-

/

/

/

1.6 - s p

's ~ /

/

~

's '

/

t4-

- Legend l COBRA 3,e l.

QAR_ _ f, l :r m o t2 O

0.5 1

1.5 2

2.5 3

3.5 4

4.5 5 ggo l TIME SECONDS e=wo i

l 1

y?oO

. T M

y$mI"

} N :r rP e

• M M_

dE M nDO R eN gy A S

L eof M 0

M 1

- ~

W M O

L -

• F ' s

" M RE W _

_ ~

OM ' i8 O '

L 0 T

C T F

g

_ E

' M AS O R - Mr EU C\

RS D N c. - M E V E C N 6 S

RS - D M OW -

N O

FO C E

FL - S M OF _ E SP N SO ~

4 T

M O

L O L -

L D AN _ M TA I

RE  % s

- AR ' N M

_ PO s 2

5. C s N 8 s 3  % M N

G I

's F

\ M

~ - - -

O 1 6 2 8 4 2 1

0 0 0 0 4g .Bd M

J hS o -

P Wb9 M

i

! ; 'l  ; l l

l um um man man ami aun num aus uma man ami um amm ums mas amm num sum ame FIG.3.8.6 PARTIAL LOSS OF FORCED REACTOR FLOW:

NEUTRON Ato HEAT FLUX VERSUS TME i

M l NERACE CHANNEL

] -_ -, HEAT FLUX j I' N N l N N s N N

i \ N j OA- \ \

i N

I N g N u

b \ s

~ s

~

h \  %

i \ '

b

\

0.4 - \

! C \ ,

\

NCUTRON FLUX > \,,,,,-

0.2 - -

Legend

I :o z l

$kc

! G^_a _ eE$

i 0 . . . .

-om i 0 2 4 8 '8 # ^3 i

TNE. SECONDS 'o O

m I

i

i1l l 1

y bW y -0 0 y g- C8 d A nR eB ^

L gO eC -

I 5

/ /

W '

O 5 L ' 4 F

R ' -

O - 4 T

C A

E _ 5 R

D 3

EMEI

~

CT R S O U

~ 3 S

D N

F S ' O C

FR OE 5.

2 E S,

E S V s M I

SR OB

' 2 T

LN '

L D A

I

.5 1

T '

R P

A

.1 7

8 3 5 0

G I

F O

6 5 4 3 2 8 7 1 1 1 1 1 1 t x@

wh I lll l lll l<

l NFU-033

= Revision 0 March 14, 1986 I 3.9 LOSS OF REACTOR COOLANT FLOU - LOCKED ROTOR 3.9.1 Description of the Accident This accident is the postulated instanta-I neous seizure of a reactor coolant pump rotor. Flow through the affected coolant loop is rapidly reduced causing a low react'or flow trip signal. Heat transfer to the shell side of the faulted steam generator becomes reduced. At first, the reduced flow results in a decreased tube I side film coefficient. Later after the trip, the reactor coolant in the tubes cools down uhile the shell side temperature increases (turbine steam flou I is reduced to zero upon plant trip) resulting in a decreased delta T.

I Follouing the reactor trip, heat stored in the fuel rod continues to be added to the coolant causing the coolant to expand.

I This effect combined with reduced heat transfer to the steam generator causes an insurge into the pressurizer and a pressure increase I throughout the RCS. The insurge of the coolant into the pressurizer compresses the' steam volume, actuates the pres-I surizer spray system and opens the pouer operated relief valves ( PORV' s ) and the pressurizer safety valves.

The danger resulting from a locked rotor transient lies in two areas. The first pertains to the reduction in heat removal

'g from the core. Without a prompt reactor

!g ' rip, the fuel cladding temperature vill l ,

increase such that substantial cladding i damage can evolve. The second concerns the rapid increase in system pressure.

This increase can jeopardize the integrity of the primary coolant system I without the effects of the pressurizer spray, relief and safety valves.

I I

3-75 k_. _ _ _ _ _ _ _ _ . _

NFU-0033 Revision 0 March 14. 1986 3.9.2 Summary of Accident Analysis Methodology Two computer codes were used to analyze this transient. DYN00E-P(3) was used to calculate neutron flux, the peak pres-

! sure, and the core flow follouing the g pump seizure. The thermal behavior of g the fuel rod located at the core hot spot was cal'culated using FRAP-T5.(4)

For conservatism, the pressure reducing ef fect of the PORV's and pressurizer spray was not included in the analysis. g At the beginning of the postulated 3 accident, the plant was assumed to be in operation under the most adverse steady state operating condition with respect to the margin to departure from nucleate boiling (DNB), i.e. maximum power level, minimum pressure and maximum coolant average temperature.

For the peak pressure evaluation, the a initial pressure was conservatively estimated as 30 psi above nominal g

pressure (2250 psia). To obtain the maximum pressure in the primary side, the l highest pressure occurring in the RCS uas 5 evaluated. This pressure was obtained by adding the loop pressure drop to the calculated pressurizer pressure.

In the fuel rod thermal analysis, DNB uas assumed to occur in the core. Results obtained from this analysis represented the upper limit with respect to clad temperature and zirconium uater reaction.

In the evaluation, the rod power at the hot spot was conservatively assumed to be three times the average rod power, i.e.,

Fq = 3.0 at the initial core power level. Furthermore, the axial power distribution was chosen to be a chopped g 1.55 cosine distribution. The core g coolant conditions were ramped from the nucleate boiling region to the film boiling region within .01 seconds after transient initiation. The film boiling heat transfer coefficient is representative of the louer range of the Bishop-Tang-Sandburg g correlation. The core pressure and W temperature conditions were set to 3-76 I

I

_ _ _ _ _ _ _ _ _ _____________j

I NFU-033 Revision 0 March 14, 1986 initial values (pressure = 2250 psia and temperature = 652'F) and held constant I throughout the transient since these are the most limiting. The fuel-clad gap was assumed to close, ramping the gap heat transfer coefficignt from nominal to I 10.000. BTU /hr-ft 'F (negligible resi-stance to heat transfer). In connection with these conservative assumptions, the I FRAP-T5 Licensing Audit Codes were used instead of best estimate codes for specific heat, thermal conductivity, I Poisson ratio. gap conductance, fuel deformation, and metal-water reaction calculations. The net effect of these assumptions was to deposit the maximum I amount of energy in the cladding.

The thermal acceptance criteria for the locked rotor accident are:

1) The maximum reactor coolant and main steam system pressures must I not exceed 110% of the design values.

I 2) The maximum clad temperature calculated to occur at the core hot spot must not exceed 2700*F.

3.9.3 Results The case of all loops operating with one locked pump rotor was analyzed. The nuclear power. hot channel heat flux, and core flow are shown in Figures 3.9.1, 3.9.2 and 3.9.3, respectively. The comparisons between DYNODE-P and the

. FSAR(6) were good. The maximum RCS pressure predicted by DYNODE-P was I plotted against that predicted by the FSAR in Figure 3.9.4. The pressure predicted by DYNODE-P was slightly higher I than that of the FSAR but did not exceed the reactor vessel design pressure limit of 2750 psia.

I I 3-7.7 I 9

NFU-033 Revision 0 March 14, 1986 The maximum reactor coolant and steam g system pressure were lower than 110*4 of the design values. The clad temperature g as shown in Figure 3.9.5 was below the acceptance criteria value of 2700*F.

Therefore, there was no danger of the clad damage and the system pressure was well below the RCS design limit.

I I

I I

I I

I I

I I

I I

3-78

~-_ _.

[ mum mim um num man um amm am num num um um man uns man mum uma muu em FIGURE 3.9.1 LOCKED ROTOR TRANSIENT:

NUCLEAR POWER VERSUS TIME 12 1- N N \

Z \

2 O \ -

Z

u. 0.8 - \

O Z \

o \

U t

$ 0.6 -

wu \

\

f \

l 3 \

l o 1 0.4 - 1 e

l b da

\

2 s 0.2 - ~~ -----

_____ legend DYNODE 3,g UAE - .

O

;  ;  ;  ; i TIME N SECONOS

+ e 40

-p m

l

FIGURE 3.9.2 LOCKED ROTOR TRANSIENT:

HOT CHANNEL HEAT FLUX VERSUS TIME 1.2

{3 ,,_ --

s s

o s z

y O '

\

Z o.g -

9 N s

D

< N N

E ~

s

>< 0.6 -

y 3 s g

u. 's tt s j

b 's '

0.4 - .,

~

s 5

r 0.2 -

o -

Legend I

DYNODE EsAg _

0 . -

; j i a i 10 0 '

• i TBAE N SECONDS

^"

[$

, , , . - - == = == " " " " " " " "I t

W uma em uma um amm num mum e sum uma em sum um um num mum man umm FIGURE 3.9.3 LOCKED ROTOR TRANSIENT:

CORE FLOW VERSUS TNE 1.2 1-

\

A Q o.a-g _ _____ ____________

z 9

0 0.6 -

a g

o j O.4 -

w m

o o

0.2 -

Legend DYNODE ggy G A!L _  !.h'$

o i i i i i i i i i io ;lT THE N SECONDS c G

M A M_i

[

)

FIGURE 3.9.4 LOCKED ROTOR TRANSENT:

REACTOR COOLANT PRESSURE VERSUS TME 2700 1

,s ~ ~ ,

2600-

/ \

\

/

\

/ .

\

2500- / g g \

I \

E u / s 2400- j g

\

10 1

\

, E 2300- N N

N N

N N

I 2200- N N Legend DYNODE

~

P 2100 . . . .

1. f. w 2 3 4 5 6 7 8 9 10 O 1 i TNE N SECONDS y@"

l R

sus man ami uma mas e mas mum um e am mum muu num num uma amm man mum FIGURE 3.9.5 LOCKED ROTOR TRANSIENT:

CLAD TEMPERATURE VERSUS TIME 2200 2000- f'

/

/

E 1800-(n s

' N

/ s

@ 1800- /

Q j E /

1400-Eo_ l '

1200-i 1000-

. I 800-Legend FRAP FSAR ~  !$$

600 . . . . . . . . . o h. ? f 0 1 2 3 4 5 6 7 8 9 10 :r in o TIME IN SECONDS U

=

2 1

\

NFU-033 Revision 70 g s March 14, 1986 3.10 MAJOR SECONDARY SYSTEM PIPE RUPTURE <

3 10.1 Description of the Accident One of the most serious accidents con-sidered to be a limiting fault is the g main steamline break. The main steam g pipe is postulated to be completely severed at the outlet of a steam genera-

~

tor inside the containment at no load conditions with offsite power available.

The increase in steam flou through the break results in an increase in energy g removal from the primary system causing a g rapid drop of moderator temperature and.

reactor coolant system (RCS) pressure.

The cooldown of the moderator results in a positive reactivity insertion dus to the assumed large negative moderator temperature coefficient decreasing the g shutdoun margin. With the most reactive 3 rod cluster control assembly (RCCA) stuck in a withdrawn position, it is ccnceiv- g able'that the reactor could become g critical and return to power. Ulti-mately, the reactor is shut down by baron injection which results from actuation of E the low pressure safety injection system E and the accumulators.

3.10.2 Summary of the Accident Analysis The steamline break case described above was analyzed using OYNODE-P(3) which modeled the reactor core, pressurizer, steam generators, RCS and main steam supply system. Results were obtained by employing the following very conservative assumptions in the analysis:

1. The moderator initially contains i l lou concentration of boron. This' =

provides for a more negative moderator temperature coefficient and a lou concentration at time of boron injection.

2. Initially, the reactor is assumed to be in the subcritical zero power state. This assumption is made,so that the stored energy of the f g system is at a minimum and the m uater level in the steam generator is at a maximum. This results in a more severe transient.

3-84 I

= _ _ _ _ _

1 NFU-033 Revision 0 March 14, 1986

3. Conditions are similar to those at end of life (EOL). The main effect is a smaller effective delay neutron fraction, 8,77

4. The baron concentration injected into the system is conservatively

[ assumed to be at a concentration of 20,000 parts per million (ppm),

- which corresponds to the Technical r

Specifications lower limit for the Baron Injection Tank.

E t

5 Assuming maximum heat transfer in the broken loop steam generator and k g no reverse heat transfer for the

[

g intact steam generators.

3.10.3 Results The steamline break analysis was performed using the input assumptions L from the Final Safety Analysis Report CFSAR](6) in addition to the previous k . assumptions. The results are compared with the FSAR results in Figures 3.10.1 through 3.10.5. Figures 3.10 1, 3 10.2 L and 3.10.3 show the reactor vessel L average temperature, pressure and core heat flux, respectively. A complete pipe a severence is assumed. Therefore, the cross-sectional area of the steam pipe is used for the break area. The break flou

_ from the faulted steam generator is shoun

in Figure 3.10.4. The amount of steam released at the time of break differs by E about ten percent from the FSAR results for the faulted generator and one percent for the other generators (this flow is due to a common header design). The

reactivity change resulting from the coolant temperature change is shown in Figure 3.10 5.

F L These results show that the DYNODE-P model correctly simulates the steamline i g break transient. The reactor responds to E E the steamline break by becoming super critical. The subcritical condition is then restored at about 100. seconds 5

through baron injection, thereby, safely terminating the reactor's at tempt to return to power.

{

I 3-85 r-K 0

FIG. 3.10.1 MAIN STEAMLINE BREAK:

REACTOR VESSEL AVERAGE TEMPERATURE VERSUS TIME 560 ts.

540- \

5 \ l Q \ (

Z 520-g -

y z 500-

\

480- s

\

wb s

m N

• N g 460- N .

N N

h 440- N l

> N E N O s Legend  ;

b 4a0- s N xxz i I h 's ~ - n.

~

~_

g.;a rw 400 . . . . . .

80 90 10 0 110 0 10 20 30 40 50 60 TIME N SECONDS 70 y8" a

e e m e sun -

mas ung ,,, ,,,,

FIG. 3.10.2 MAIN STEAMUNE BREAK:

REACTOR COOLANT PRESSURE VERSUS TIME B 2500

'N

\

l g 2000- \

@ \

E \ .

y \

2 \

1500-

@ \

l

\

!T

> co

\

4 O

O 1000- \

O -

! E N i O s D ~~ '

l 6 E

'~

i 500-Legend ovuooc U A!L - kk!

oei 4

o . . . . . . .

so too iso

• p. o o

io 20 so 40 50 80 70 so i

TIME IN SECONDS %8" i: o

! E i

i

i I

! FIG. 3.10.3 MAIN STEAMLINE BREAK:

! CORE HEAT FLUX VERSUS TIME j 0.5 l

i f M 0.4-

z i

2 O .

Z in

, O

! Z 0.3 -

i O 5

1 w I $

i 0.2 -

b I

i w -

1 h -

0.1 - e l

o --

l

/ Legend

/ DWODE a :o z i W (D M 1

j 0.c

[/, , ,

30 40 50 60 70 80 90 10 0 110 GAE. - 25?

umo g $^ "

j 0 10 20 1

TWE N SECONDS . f5

=

i

• l

uma man uma um num amm man men ums num num amm uma amm uma amm ums FIG. 3.10.4 MAIN STEAMLINE BREAK:

STEAM RELEASE VERSUS TIME 3

10000 4

z 8000-O 1

g 6000-Z

~3 -

i O

, D.

.Y z 4000-tA G

E 2

i b g 2000- \

s

~ Legend J

~___~- -

-____ ovuooc ggg Nd.

FSAR

--- :r u, o?

H o' w 0 10 2O IO N $0 d0 YO $0 d0 15 0 11 0 TIME IN SECONDS  ?[

m 1

\ _ - --_-- -_

4 4

i FlG. 3.10.5 MAN STEAMLNE BREAK:

REACTNITY VERSUS TIME 0.5 -s

! l\

l I s I '

~ _

0 ' _ . _ _ _ . _ _ _ _ _ _ - -

I '

l l

h

'- d O

l

-0.5 -  !

O Y

e

o Z

_1 b

e l -1.5 -

[

N ewt _ era

-2 0 y 0 40 g

, i n

a 70 g'0 90 M H0 w$U 63 TME N SECONOS o O

, , , m e m m m m M " " " "

l E

NFU-033 Revision 0 March 14, 1986 3.11 ROD CLUSTER CONTROL ASSEMBLY (RCCA) EJECTION 3.11 1 Description of The Accident This accident is postulated as the g unlikely event resulting from a 5 mechanical failure of the control rod mechanism pressure housing. The result of this mechanical failure is the I ejection of a rod cluster control assembly (RCCA) and drive shaft, which results in a rapid reactivity insertion I and possibly adverse core Power peaking which could lead to localized fuel rod damage. The accident is mitigated by the reactor protection system high neutron I flux trip and the self limiting negative Doppler reactivity following reactor trip.

3.11.2 Summary of Accident Analysis Methodology The analysis of the RCCA ejection I accident is performed in two stages. In stage one a core transient calculation is performed using DYNODE-P(3) (a system I transient analysis code containing point reactor kinetics) to determine the system transient behavior and average power generation. Doppler and moderator reactivity feedback are included in the calculation. These reactivities are multiplied by a weighting factor to account for spatia! coolant and fuel I temperature distribution effects not explicitly represented in the computer code.

In stage tuo the average core energy addition predicted by DYNODE-P is I multiplied by the appropriate hot channel factors to perform the hot spot fuel and clad transient heat transfer calcula-I tions. The calculation is performed using the FRAP-TS(4) code.

The assumptions made in the rod ejection I analysis were taken from Chapter 14 of the Salem FSAR(6), particularly from Table 14.3-2d. Some of these values are I tabulated on the next page.

I 3-91 l I .

l

NFU-033 E Ravision 0 E March 14, 1986 Parameters Used in RCCA E.iection Accident Hot Full Pouer Hot Zero Power BOL EOL g Delayed neutron 0.44%

g fraction 0 55%

Moderator temperature coefficient -1. pcm/*F -26. pcm/*F Doppler Weighting factor 16 3.55 Ejected rod worth 0.2% delta K. 0.98% delta K E

The hot spot analysis was performed using the detailed fuel and clad transient heat g transfer computer code, FRAP-T5. The 3 pressure and core power histories were taken from DYNODE-P output and used as input to FRAP-T5 The hot spot uas modeled as a single node. Ten radial mesh intervals were used in the fuel, one in the gap and tuo in the clad. The B Dittus- Boelter correlation was used to 3 determine the surface heat transfer coefficient before DN8 and the Bishop-Tong-Sandburg correlation to determine the film boiling coefficient after DNB.

These values were input to FRAP-T5.

The hat channel factor during the tran-sient was assumed to increase from the steady state design value to the maximum a transient value in 0.1 seconds and remain g at the maximum value for the duration of the transient. Several other conser-vative assumptions were made. The heat transfer coefficient at the clad surface

( uas decreased from the nucleate boiling l region to the film boiling region in .01 g 3

seconds so that the maximum amount of energy was kept in the rod. This is consistent with the FSAR assumption that the core went into DNB at the start of the transient. The gao heat transfer coefficient was ramped from a nominal value to a higher value (negligible 3 resistance to heat transfer) repre- E sentative of the gap closing due to the 3-92 I

1 .

I

I NFU-033 Revision 0 March 14, 1986 expansion of the hot fuel. The bulk coolant temperature at steady state was I initialized to the saturation value, and the reactor coolant flou was reduced to 95.5% of nominal to account for 4 5% core bypass flow that is unavailable for heat transfer. The FRAP-T5 Licens;ng Audit I Codes were used instead of best est:, ate codes for specific heat, thermal I

• conductivity, Poisson ratio, gap conductance, fuel deformation and metal-uater reaction calculations.

The cumulative effect of these assumptions is to simulate the most limiting core conditions for the transient.

The acceptance criteria for the control rod ejection accident are:

1 The average hat spot fuel enthalpy must be less that 225 calories / gram I for non-irradiated fuel and 200 calories / gram for irradiated fuel.

I 2 Average clad temperature at the hot spot must remain less ' hat 2700*F to avoid clad embrittlement expected at temperatures above 2700*F.

3.11 3 Results Tuo cases of the RCCA ejection event were analyzed, namely the hot full power beginning of life (HFP80L) and the hot zero power end of life (HZPEOL) cases.

HFP80L The relative core pouer calculated by

. DYN00E-P in the HFP90L case is compared I

to the FSAR results in Figures 3.11 1 and 3.11 2 The fuel and clad temperatures predicted by FRAP-T5 are compared to the FSAR results in Figure 3 11.5.

The DYN00E-P results based on the FSAR data (Figure 3.11 1) showed a higher core I power after the reactor trip. Spatial kinetics were not included in the model.

However, sensitivity studies using different scram reactivity insertion curves indicate that the difference 3-93 I -

1

NFU-0033 I

Revision 0 March 14, 1986 between the FSAR and OYNODE-P results is due to different scram curves used in 3 both analyses. The results obtained from E

~ the DYNODE-P code when the FSAR scram curve is used and that obtaingj)from the a same code using the UCAP 8458 scram curve are presented in Figure 3.112 E against the FSAR result. The improvement in the DYN00E-P results illustrates the 3 sensitivity to scram insertion rate. g Figure 3.11.5 compares the fuel center- g line, fuel average and cladding tempera-tures predicted by FRAP-T5 to the FSAR 5 values. After performing several sensi-tivity studies the most conservative values of surface heat transfer coef-ficient, gap heat transfer coefficient and coolant bulk temperature were used in g the final FRAP-T5 analysis. The results are in good agreement and demonstrate 3 that the integrity of the cladding would be maint.ained. The maximum fuel enthalpy throughout the transient was 166. calo-ries / gram. These results were well '

uithin the acceptance criterion for this transient.

, HZPEOL i

I.

The relative core power predicted by DYNODE-P in the HZPEOL case is compared to the FSAR results in Figures 3.11.3 and 3.11.4. The fuel and clad temperatures E predicted by FRAP-T5 are compared to the 3 FSAR results in Figure 3.11.6.

The core power after scram (Figure 3 11.3) based on FSAR scram curve shoued I

'a similar trend to the HFPBOL results. E At approximately 1.4 seconds, the a DYN00E-P core power exceeded the FSAR

. prediction but displayed the same trend g as the FSAR throughout the rest of the transient. 5 As in the HFPBOL case the result.s of the sensitivity studies showed trend improvements when the DYNODE-P analysis was done with the UCAP 8458 scram curve. g This further verifies the scram curve g sensitivity.

I 3-94 l

nru vvas I Revision 0 f1 arch 14 1986 a Figure 3 11 6 compares the fuel c*at'r'ia*- '"*' var S* *ad c "ddiaS E temperatures predicted by FRAP-T5 to the FSAR values. As in the HFPBOL case the I most conservative' combination of surface and gap heat transfer coefficient and coolant bulk temperature was chosen for the final analysis. The results showed I good agreement with the FSAR results.

The maximum average fuel enthalpy throughout the transient was 154.

I .

calories / gram. These results are well uithin the acceptance criterion for this transient.

I I

I I

I I

I .

I I

I

)

3-95

I .

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

Revisi n 0 March 14, 1986 I

l Il 1'

,i l I

}li I I

I

j 2

j

/

I

/

j I BY / I r -

i U s' -Q 3

/

' I' q /

n /

I

/

j .s

/

I I i i i i i i i

• l cnw m nacanamoa woo 3-96

NFU-033 Revision 0 March 14, 1986 I

I .

i  ;

I ln I -

}lli I ,,

I I ..

I //

, . v) a O  !

zT i '/

l %= // -~

i

~W zs Oh

/

f' s j (f) s' ... ,/ .

E x #

/ ,,,.****

, 3 Q ,

O ** ..-

I %f m /

/

/

g t ,

n ww /

/

I *8 hO /

[E /

I

~Q

/

/

• l I s i 6 6 6 6 6 6 .

amsao.tn33 m 33,og3gog I 3-97

E

" anni,1986 I

I I'

! l I

}li

, , n I.

I l

l .-

l

/

/

t ' I

/ -S m b

, I d I w

] @> / .. l!! g nb .

I i  !

W l

/

I .g l 0

/

/

y. ,.

WMMON .30 NOLIDYW.d 'M3MOcl 3803

~'*

B

NFU-033 Revision 0 March 14, 1986 I

I I I }Iis I ,

o a I lli

'l /

ll So / -a I yg .

.i Z

m$

< /

I FM /

/

zu .w l os FF g /

i

~g m ri a .

l a}

Q .

,j i!i aL .Y  ;

iI @>3 tl .

/

t l

i lI c. 3:

nO I i

w.0 l -

I Oc W:

30o l

-2 co /

I _ - -

/

/

~~"*"*==- _, l

,I l

" g g # g g

"lVNINON 30 N0dOYMJ '83 Mod 3800 3-99

Revis on 0 March 14, 1986 I

I I

I I

1 -

g i

t f t 'ge , I; I y

/

, l I; -n g

/ g I I ~ tg e <l i

/ 1 g .,.

P I R \ll'll[ll'[j l

]

~1 I

s N i

s N .

s s

s s I

\ s g s s '

s s s

- amma,m g 3-100

~

sum num use a sua man muu me amm amm amm uma amm num uma mum seus ums nas FIGURE 3.11.6 ROD EJECTION TRANSIENT:

TEMPERATURE VERSUS TIME HZPEOL 4500 4000-

-- ~~____,~~~~___

/

3500- [,' _ _ _ _ _ , , , _

y. 4 . ' ' ~ ~ ~

N 3000- FUEL AVERACE TEMPERATURE tr //

8 i ,

~~~

Q -

W ' ~

w 2500- /

s -

1 W l

? l 4 2000- / CLAD TEMPERATURE h I 1500- g I

iOOO- Legend amz U r__saa. _ C 500 i , a i

• g;a i -

0 05 ' 1.5 2.5 3 35 4 rw t TIME N SECONDS ,

i O i 50 CD l

I

NFU-033 R vision 0 March 14, 1986 4.0

SUMMARY

For each of these analyses. it was necessary to verify that the mechanical and nuclear safety limits were not exceeded. Fuel and clad temperatures were determined I in certain analyses to verify that no fuel rod damage would be incurred. Another indication of possible fuel damage is the DNBR.. This value helps determine if I there is sufficient heat removal capability to avoid fuel damage. Another area of concern is the reactor coolant system pressure. This, like the DNBR. has a safety limit associated with it; that being 110*4 of the I design pressure. Other mechanical safety criteria exist that are inherent to specific analyses such as fuel enthalpy limits during reactivity insertion I accidents. These " failure consequence" limits may also be considered if the " failure threshold" limits such as DNBR are exceeded.

In Section 3.1. a rod withdrawal from a subcritical l

condition is presented. The DYNODE-P analysis predicted a safe reactor trip in response to the i uithdrawal. Important safety parameters for this analysis are fuel pellet and rod cladding temperatures.

Our analysis predicted that these temperatures would stay belou the safety limits.

I In Section 3 2, a rod withdrawal analysis performed at power is shoun.

1 For this transient. system pressure I and DNBR are the significant safety parameters that need examination. This transient was analyzed at a fast and a slau uithdrawal rate. DYN00E-P predicted a safe reactor trip and a transient maximum pressure that is below the relief valve setpoint. The DNBR prediction was calculated by COBRA and is greater than l the value determined to be a safe limit.

I In Section 3.3. an uncontrolled baron dilution transient analysis was presented. The concern for this I case is the ability of the reactor protection system to trip the reactor before any damage occurs. DYN00E-P predicted a successful trip with no limits exceeded.

I A single dropped rod accident is covered in Section 3 4. In this analysis, the possibility of a power overshoot exists due to the use of the rod controller model.

I The results show that no significant power overshoot would occur and that the plant transient behavior would be uithin the safety limits.

I I

4-1

NFU-0033 Revision O March 14 1986 Section 3 5 deals with excess heat removal from the primary system caused by a feedwater Control valve malfunction. The drop in temperature will cause a l reactivity insertion (in the presence of a negative 3 moderator temperature coefficient). Therefore. a ON8R analysis was performed to demonstrate the ability of B the system to prevent the occurence of a ONBR below g 1.3 The OYN00E-P analysis predicted a safe reactor trip without exceeding RCS pressure limits. The ONBR prediction by COBRA was found to be within safety -

limits.

A loss of electric load accident was analyzed in Section 3 6. This transient causes an increase in the system pressure. The analysis was performed with and without pressurizer spray and pressurizer power operated relief valves. The DYN00E-P analysis predicted a successful reactor trip and showed that system pressure would be maintained below safety limits for both situations. The COBRA predicted DN8R was also 3 found to be uithin the safety limits. 5 In Section 3.7. the loss of normal feeduater transient is presented. In this transient. a major concern is that ample heat removal capability should be available to the primary system. 0YN00E-P predicted that the intact steam generator tube sheet would not be E uncovered, therefore. satisfactory heat removal vould 3 be available.

In Section 3.8. the analysis concerns a reactor coolant pump -t r i p whiCh causes a loss of reactor Coolant flou.

Tuo situations were considered: A partial loss of flou (uhich is a situation where tuo out of the four reactor E E

coolant pumps are tripped) and a complete loss of flow (in which all four of the reactor pumps tripped).

Because of the loss of flou. ON8R becomes most siginificant in this analysis. The DYN00E-P analysis predicted a reasonable system transient behavior for both cases. *The ON8R predicted by COBRA did not fall below the 1 3 safety limit value.

A locked rotor transient is presented in Section 3 9.

The core flou resulting from a locked rotor and the g reactor trip predicted by the OYN00E-P code compared g well with corresponding values from the FSAR. The clad temperature was found to be within the limits for prevention of clad damage.

I I

4-2

I NFU-0033 Revision 0 March 14, 1986 Section 3 10 documents the analysis performed for a steamline break accident. During a steamline break I accident, the possibility exists that the reactor previously in a shutdown state could go critical and return to power. Our analysis predicted that for the assumed initial conditions, the reactor would go critical. However, due to baron injection, it would eventually return safely to a subcritical condition.

The steam release that results from the break is reasonably calculated in our analysis usirs OYN00E-P.

A rod ejection accident is presented in Section 3.11.

There were tuo cases performed for this analysis. One case considered the reactor to be at ful1 power and beginning of life. In the other case the reactor was assumed to be at hot zero power and end of life. In I both situations, the OYN00E-P calculation resulted in a reasonable system transient and a successful reactor trip. In both cases, it was assumed that DNB existed at the beginning of the transient. FRAP-T5 uas run in I order to obtain results for clad and fuel temperatures.

The results predicted that there would be no fuel l

melting or clad failure.

In these analyses, some parameters do not directly relate to the cause or forcing function of that I accident, but are very dependent upon the reactor core design. There are other parameters which* relate directly to the for.cing function or the cause of each transient and therefore, are identified as transient I specific parameters.

parameters must be considered to determine uhether a particular transient should be analyzed.

For each reload, all of these These analyses have been performed to be compared with the Westinghouse analyses presented in the FSAR. PSE&G has obtained the same results and reached the same conclusions in these analyses as the vendor presents in the FSAR. Since Westinghouse is an NRC accepted institution for accident and transient analysis, ue (3 feel that this demonstrates our capability and

'3 DYNODE-P' S adequacy for performing such analyses to NRC standards.

I I

I I

4-3

NFU-033 Revision 0 March 14, 1986

5.0 REFERENCES

1 U. 8. Henderson; "The Nuclear Design of the Salem Unit I Nuclear Power Plant Cycle 1 " UCAP-8458:

December 1974 Westinghouse Electric Corporation:

Pittsburgh, Pennsylvania.

2. J. U. Jackson.and N. E. Todreas: " COBRA I IIIC-MIT-2: A Digital Computer Program for Steady State and Transient Thermal-Hydraulic Analysis of Rod Bundle Nuclear Fuel Elements:" MIT-EL 81-018:

June, 1981: Massachusetts Institute of Technology:

Cambridge, Massachusetts.

3. R. C. Kern. et at: "DYN00E-P, A Nuclear Steam I Supply System Transient Simulator for Pressurized Water Reactors;" Versions 4 1 (NAI 81-33), 5.1 (NAI 82-23), and 5.2 (NAI 82-41): Utility Associates International (formerly Nuclear I 4.

Associates International); Rockville, Maryland.

L. J. Siefken, et al; "FRAP-T5, A Computer Code I

• for the Transient Analysis of Oxide Fuel Rods,"

NUREG/CR-0840: June 1978 Idaho National Engineering Laboratories: Idaho Falls. Idaho.

5. J. C. Lai: " Modification of COBRA IIIC-MIT:"

Internal Memorandum: NFG 82-069: November 1982:

Public Service Electric and Gas Company Hancock's Bridge, New Jersey.

6. "Public Servic's Electric and Gas Company - Salem Nuclear Generating Station, Units I and II: Final Safety Analysis Report" United States Atomic Energy Commission Docket Numbers 50-272 and 50-311; January 1981: Westinghouse Electric Corporation: Pittsburgh, Pennsylvania.

d I

I 5-1 l

NFU-033 I Revision 0 March 14, 1986 I

APPENDIX A In preparation of this report, three computer codes were used in these transient and accident analyses. 0YN00E-P was applied to perform the Nuclear Steam Supply System simu-I lation and COBRA IIIc-MIT and FRAP-T5 were employed to obtain the thermal-hydraulic response of the coolant channel and hot spot analysis. OYN00E-P simulates the NSSS tran-sient and obtains an average core and reactor coolant system (RCS) response. In the fuel rod transient analysis, a hot rod was modeled using FRAP-T5 code. Hot rod power history, core inlet coolant temperature, enthalpy and RCS pressure I history from DYN00E-P are used as input to FRAP-T5 FRAP-T5 code calculates the fuel temperatures, fuel The enthalpy. cladding responses and other parameters which give indications of whether or riot the rod integrity is main-tained. Data from DYN00E-P output is similarly used for input for COBRA IIIc-MIT. Through the use of COBRA IIIc-MIT. thermal hydraulic response of the hot channel is I obtained. The main output of this code is the departure from nucleate boiling ratio (DNBR).

The following is a brief synopsis of these codes. More detailed descriptions of these codes are listed in the references at the end of this report in Section 5.0.

1. Nuclear Steam Supply System Simulation I DYN00E-P(3) is a Fortran IV Computer program which simulates the dynamic response of a Nuclear Steam Supply System (NSSS) of a pressurized water reactor (PUR) under accident or transient conditions.

0YN00E-P includes a simulation of the major components of a PUR NsSS uhich significantly influence the re-I sponse of the system to transient conditions. Geometry options are provided to permit representation of any of the current PUR designs.

The major features of OYN00E-P are:

Point kinetics model as well as one dimen-I sional kinetics model for core power tran-sients with major feedback mechanisms and decay heat represented. An initially sub-critical core can be modeled.

- Power forced mode option for hot channel analyses.

I A-1 I I

NFU-033 Revision 0 March 14, 1986 Multinode radial fuel rod and multinode axial coolant channel representat~ ions in the core.

- Conservation of mass, energy, volume and boron concentration for the Reactor Coolant System. Conservation of momentum is optio-nal.

- Detailed pressurizer model including spray and heater systems and safety and relief valves.

- Explicit representation of the shell side of l the steam generators including conservation W of mass, energy, and volume.

(

- Explicit representation of the main steam system with isolation, check, dump, bypass, and turbine valves including conservation of mass, energy, momentum, and volume.

- Representation of the Reactor Protective and ,

High Pressure Safety Injection Systems.

Representation of the major control systems.

Provisions for simulating a variety of transients and accidents including a break in the main steam system, steam generator tube ruptures, and ATUS events.

Self initialization.

Full range of uater properties including supercritical pressures.

The basic input parameters involving initialization g are: 3

" Core geometry and initial thermal-hydraulic characteristics.

Primary system data including initial Reactor Coolant System (RCS) pressure and pressurizer E level, core inlet enthalpy. RCS flou distri- E bution, RCS baron concentration, and core bypass flou.

- Initial core power level and distribution.

- Hydraulic characteristics and RCS steam generator and main steam system volume distributions.

Initial steam generator pressures and levels and heat transfer data.

A-2

I NFU-033 Revision 0 March 14, 1986 I The basic input parameters involving the transient response are:

Core kinetics characteristics including control rod motion.

Reactor Coolant System inertias, pressure loss coe.fficients and pump hydraulic, and torque characteristics.

Control system characteristics.

Main and auxiliary feeduater characteristics.

Valve characteristics.

Safety systems characteristics.

Transient power demand.

Transient load demand.

I The output consists of two edits the first is the major edit consisting of data printed at select time points during the transient, the second is a transient i

l summary table. The major output consists of the follouing list of parameters:

Core variables Avorage Power l Fuel rod temperature and heat flux l Coolaht enthalpies, temperature, and mass l Kinetics variables including k,pp I RCS variables l Mass, energy, and baron distribution of the coolant loop flou rates Pressurizer pressure and level Safety system variables, setpoints, and valve I status Pressure control system variables Reactor coolant pump speeds, torques. and developed heads Steam generator variables Pressure and levels Masses Heat loads Feedwater and steam flous I -

Main steam system variables Pressure and mass distributions Steam flous I

A-3

NFU-033 E Revision 0 E March 14, 1986 The transient summary table is located at the I

end of the output. This includes:

Time

- Relative neutron power Pressurizer pressure K

- C8bb average heat flux Average and maximum fuel temperature g Total steam generator heat load 3 Core inlet flou and enthalpy Relief plus safety valve flou Pressurizer uater level Maximum transient values for parameters listed above and time of occurance Maximum steam generator secondary side pressure and time of occurance Trips generated during transient and time of generation Table containing times at which restart files were written I

I I

e I

I I

I A-4

I NFU-033 Revision 0 March 14, 1986 A2 Fuel Rod Transient Analysis The Fuel Rod Analysis Program - Transient (FRAP-T5)(4) is a Fortran IV Computer code that calculates the transient performance of light uater reactor fuel rods during hypothesized reactor transients and reactivity initiated accidents. The code performs a steady state calculation to initiate the transient calculation of temperature, pressure, failure, and deformation his-I tories of fuel rods. The models implemented by FRAP-T5 include:

Heat conduction Heat transfer from cladding to coolant Elastic-plastic fuel and cladding deformation Cladding oxidation Fission gas release Fuel / cladding mechanical interaction I -

Transient fuel rod gap pressure Cladding annealing Heat transfer between fuel and c! adding The code has an option that automatically provides a detailed uncertainty analysis of the calculated frel rod variables due to uncertainties in fuel rod fabri cation, material properties, power, and cooling.

Th+ basic required input parameters for FRAP-T5 con-sists of the following:

Data describing fuel rod designs specifically, I that pertaining to fuel pellet fesign, cladding design. and information on the fill gas and plenum spring.

Gas / gap information, fuel thermal distribution, fuel and cladding thermal-mechanical pro-pertles, and orLginal fuel burnup at specified burnop level.

The fuel ead power data. includir.g povar I distribution and Iinear1y averaged ead pcuer history.

i A-5 l

i

NFU-033 l Revision 0 E March 14, 1986 The output is formatted in prerequested time edits.

Much of the data is given in terms of distribution throughout the rod. Included in each edit is:

- Fuel and clad temperature distribution.

Fuel and cladding thermal-mechanical responses to transient, including deformation and metal-water interaction and information on heat transfer.

Fuel gap thermal-mechanical response to transient.

- Coolant thermal-hydraulic response to transient.

I I

I I

B I

I A-6

NFU-033 Rsvicion 0 March 14, 1986 A3 Fuel Channel Thermal-Hydraulic Analysis The COBRA IIIc-MIT-2(2) computer program computes the flou and enthalpy in rod bundle nuclear fuel element subchannels during both steady state and transient I conditions. It uses a mathematical model which con-siders bcth turbulent and diversion crossflow mixing between adjacent subchannels. Each subchannel is assumed to contain one-dimensional, two-phase, sepa-  !

rated, slip-flow. The two-phase flou structure is {

assume'd to be fine enough to define the void fraction as a function of enthalpy, flow rate, heat flux, I pressure, position, and time. At the present time, steady state two-phase flow correlations are assumed to apply to transients. The mathematical model neglects sonic velocity propagation therefore, it is limited to transients where the transient times are greater than the time for a sonic wave to pass through the channel.

The equations of the mathematical model are solved by using a semi-explicit finite difference scheme. This i scheme also gives a boundary-value flow solution for j both steady state and transients.

The features of COBRA II'Ic/MIT-2 can be summarized as follows:

It can consider transients of fast to inter-mediate. speed. No sonic velocity propagation effects are considered.

t The numerical scheme performs a boundary value solution where the boundary conditions are the inlet flow, inlet' crossflow, inlet I

enthalpy, ano exit pressure.

The numerical solution has no stability limitation on space or time steps.

The transverse momentum equation includes I temporal and spatial acceleration of the diversion crossflou.

Fuel pin model options allou calculation of I fuel and cladding temperatures during tran-sients by specifying power density.

I -

Forced flou mixi.ng due to diverter vanes or utre uraps is included.

The numerical procedures allow more complete analysis of bundles with partial flav blockages.

A-7

NFU-033 g Revision 0 m March 14, 1986 The inclusion of the temporal and spatial acceleration of the diversion crossflou provides a more complete physical model with only a small increase in the complexity of the numerical solution. The use of fuel rod heat transfer models coupled with subchannel analysis methods provides a more complete way of performing transient thermal-hydraulic analyses of rod bundle nuclear fuel elements. By selecting appropriate heat transfer corre'lations the fuel temperature re-sponse to selected transients can nou be analyzed in

- much greater detail.

A modification was made to the original COBRA IIIc- 3 MIT-2 code. The spacer-grid factor used in the Salem E FSAR (6) was employed in the modified COBRA IIIc-MIT-2 code (5). The modified COBRA IIIc-MIT-2 has been utilized to analyze all the transient cases in this report.

The basic input parameters for COBRA IIIc-MIT are:

Parameters referring to the fuel rod and coolant channel geometry.

Operating conditions and transient driving func- .

tions of pressure, enthalpy, flou, and power.

Friction factor correlations.

Void fraction correlations.

Loss coefficients..

Fluid flow mixing parameters.

- Fuel nod heat transport and heat transfer corre-lations.

- Critical pouer ratio (CPR) and critical heat flux ratio' correlations.

The output of COBRA Illc-MIT is broken up into time edits. The user determines the details to be included in these edits. The information available for the output edits are Channel results Cross flow tables

- Fuel temperature tables ONBR or CPR This output can be specified for all channels, rods or fuel nodes analyzed or for any channel (s), rod (s), or fuel node (s) of interest.

A-8

I .

I NFU-033 Revision 0

'PSEG The Energy People

" " " " " ~

I E I

I I

,3 ACCIDENT ANALYSIS METHODS FOR APPLICATION TO SALEM NUCLEAR UNITS I

I I

I I

.I_._ __ . _ __ _ __ _

~

=

1