ML20214A788

RAMONA-3B Calculations for Browns Ferry ATWS Study

ML20214A788
Person / Time
Site:	Browns Ferry
Issue date:	02/28/1987
From:	Neymotin L, Saha P, Slovik G BROOKHAVEN NATIONAL LABORATORY
To:	NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
References
CON-FIN-A-3273 BNL-NUREG-52021, NUREG-CR-4739, NUDOCS 8705190595
Download: ML20214A788 (120)

Text

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

.- ./

NUREG/CR 4739 .

BNL NUREG-52021 RAMONA-3B CALCULATIONS FOR BROWNS FERRY ATWS STUDY i

j P. SAHA, G.C. SLOVIK, and L.Y. NEYMOTIN i i

4 i

1 4

i i DATE PUBLISHED FEBRUARY 1987 I

i i

1 I

4 l

I DEPARTMENT OF NUCLEAR ENERGY, BROOKHAVEN NATIONAL LABORATORY

! UPTON, LONG ISLAND, NEW YORK 11973

'l

) q Prepared for the U S Nuclear Hegulatory Commission i *)  %

p Office of Nuclear Hegulatory Hoscarch l p{J)l{(hLj g

Ij Contract No DE AC02 76CH00016

b9n"r ud[

l U70220 lt h 03Oooggry i in i PDR

. . _ . - . = . __. . ._. - .. -

NUREG/CR-4739 BNL NUREG-52021 AN.RX.1S

, l RAMONA-3B CALCULATIONS FOR BROWNS FERRY ATWS STUDY P. SAHA, G.C. SLOVIK, and L.Y. NEYMOTIN Manuscript Completed - September 1986 Date Published - February 1987

LWR CODE ASSESSMENT AND APPLICATION GROUP i DEPARTMENT OF NUCLEAR ENERGY BROOKHAVEN NATIONAL LABORATORY UPTON, NEW YORK 11973 l

t Prepared for the UNITED STATES NUCLEAR REGULATORY COMMISSION OFFICE OF NUCLEAR REGULATORY RESEARCH WASHINGTON, D.C. 20555 CONTRACT NO. DE AC02 76CH00016 NRC FIN A 3273 r__,.. . ._

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

4 t

t e

J i .

1 l

i i

e 4

1 l

1 4

.j i

I 1

i NOTICE This report was prepared as an artount of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof.or any of their employees, makes any warranty, expressed or implied or assumes any legal liability or respemsibility for any third party's une, or the results of such use,of any information. apparatus, product or pnaress disclosed in this report, or represents that its use by eurh third party would not infringe privately owned rights.

The views expressed in this report are not nesennarily those of the U.S. Nuclear

llegulatory Comminaion.

2 J

Available from Superintendent of Ilocuments U.S. Government Printing Office l',0. Ilox :17082 Wanhingion,1)C 2ml:17982 and

! National Technical Information Service Springfield. Virginia 22161 1

i

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

ABSTRACT Several aspects of the Anticipated Transient Without Scram (ATWS) ini-

. tiated by an inadvertent closure of all Main Steam Isolation Valves (MSIV) in

[ a typical BWR/4 are analyzed in the report. The analysis is performed using the Brookhaven National Laboratory code, RAMONA-3B, which employs a three-di-mensional neutron kinetics model coupled with a parallel-channel thermal hydraulics in representing a Boiling Water Reactor (BWR) Core.

Four different transient scenarios have been investigated: a) downcomer water level and reactor pressure control, b) manual control rod insertion transient, c) high pressure boil-off, and d) recirculation pump trip failure.

i Results of these calculations should provide better understanding of mitigative effects of operator actions during ATWS, thus helping in the development of adequate Emergency Procedure Guidelines (EPG) required for the BWR plant safety.

{ A few unresolved questions subject to future investigations are also dis-l cussed.

1 1

i 4

i i

I.

d l

l I - 111 -

i 1

- - , - . , , , . . , - - - , , . , , _ - - . - - , - , - - .,,,r ,_---,.,,-,-c.- . .

- - - +- -- - - - ,.,- . -- , , - - - - , - - - - - - . . - . . , - , , . - , , . . . , , , ,-

l TABLE OF CONTENTS Page l

ABSTRACT........................................................... iii l

EXECUTIVE

SUMMARY

.................................................. vii ACKN0WLEDGEMENTS................................................... xi LIST OF TABLES.....................................................

xiii  !

, LIST OF FIGURES.................................................... xiv l

r 1 INTRODUCTION....................................................... I j 1.1 B a ck g r o u n d . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 4 1.2 Obj ecti ve of Present Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.3 Brief Description of RAMONA-38............................. 3

{ 1.4 Report 0utline............................................. 4  :

i

! 2. TRANSIENT SCENARIOS DURING MSIV CLOSURE ATWS....................... 5 2.1 Ea rly Pl a nt Re s p on s e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S l' 2.2 Various Operator Actions................................... 6 i 2.3 Transients Simulated with RAMONA-38........................ 9 ,

i 4 3 RAMONA-3B INPUT M00EL.............................................. 11 i

i, 3.1 Neutronics Data and Core Modeling.......................... 11 l

l 3.2 Thermal Hydraul i cs and Systems Modeli ng. . . . . . . . . . . . . . . . . . . . 16 i ,

i 4 RAMONA-3B RESULTS AND DISCUSS 10N................................... 27 4.1 Level and Pressure Control (Transient 1) . .. ... . . . .. .... .... 27 [

} 4.2 Manual Rod Insertion (Transient 2)......................... 43 j

{,

4.3 High Pressure Boil-Off (Transient 3)....................... 60 i 4.4 Recirculation Pump Trip Failure (Transient 4).............. 67

+

l 5 CONCLUSIONS AND RECOMMENDATIONS.................................... 73 I 6 FUTURE W0RK........................................................ 75 i

i APPENDIX A: DETAILS OF RAMONA-38.................................. 79 i f APPENDIX B: DETAILS OF CROSS SECTIONS GENERATION.................. 97 4

APPENDIX C: 3-D TO 1-0 CORE MODEL COMPARISON...................... 103 APPENDIX 0: SENSITIVITY OF CONDENSATION M0 DEL..................... 105 i

i 4

-v-i i

i

(

EXECUTIVE

SUMMARY

The Anticipated Transient Without Scrai (ATWS) is known to be a dominant accident sequence for possible core melt in a Boiling Water Reactor (BWR). A recent Probabilistic Risk Assessment (PRA) analysis for the Browns Ferry nu-clear power plant indicates that ATWS is one of the dominant transients for core melt in a BWR/4 with the Mark I containment; the most dominant sequence being the failure of the long term decay heat removal function of the Residual Heat Removal (RHR) system.

Of all the various ATWS scenarios, the Main Steam Isolation Valve (MS!V) closure ATWS sequence was chosen for the present analysis because of its rela-tively high frequency of occurrence and its challenge to the residual heat re-moval system and containment integrity. Therefore, this transient has been, and continues to be, analyzed by various organizations using dif ferent com-puter codes. However, most of the prior efforts have been carried out using point kinetics codes.

Early deterministic analyses revealed large variations in predicted power levels during an ATWS with the downcomer water level lowered to the top of active fuel (TAF), as required by the Emergency Procedure Guidelines (EPG).

The RELAP5/M001.6 calculations predicted the power level of -81, which agreed with the General Electric prediction of -91 On the other hand, the Electric Power Research Institute using spatial neutron kinetics codes pro-dicted the power levels of 15-18%. Consequently, the Brookhaven National Laboratory (BNL) code, RAMONA-3B, with 3-0 neutronics was selected by the Severe Accident Sequence Analysis (SASA) program to perform best estimate ATWS calculations with plant specific neutronic macroscopic cross sections.

To accomplish this task, BNL has generated a set of cross sections for tho TVA Browns Ferry Unit 1 BWR reactor at the end of cycle 5 (E0C-5) fuel.

The RAM 3NA-38 plant hydraulic model was developed based on the plant specific data already extensively quality assured at the Idaho National Engineering Laboratory (INEL) during the RELAPS calculations.

The knowledge gained from applying the RAMONA-30 code to the ATWS problem is summarized below:

1. The assumption that the one-speed, time-dependent, continuous energy dtf fusion equation can be solved by substituting in a relationship that sepa-rates flux into a " shape function" and " amplitude function" does not apply to the HWR core. The feedback mechanisms of the BWR core continuously change the power shape, which completely violates the assumptions used in the point kine-tic model's derivation. Axial power shapes predicted at different times dur-ing the transient have indicated that the void feedback effect can instanta-neously alter the axial power shape. It has also been observed that the axial power shape can completely invert as a response to vessel depressurization or safety / relief valve (SRV) actions. Furthermore, the coupling between the thermal hydraulics and nodal power is so inter-related, it has been determined from this study that a 10 neutronics formulation is the minimum requirement for reliable prediction of ATWS, It should be noted that a correct prediction

- vit -

l

of the axial core power distribution is especially important during the natural circulation mode of reactor operation due to the tight coupling be-tween the core flow rate, core void distribution (which determines the natural circulation driving head and, therefore, the flow rate), and the core power distribution. The inability of the point kinetics model to simulate this be-havior is probably why the earlier calculations predicted powers of 8%,

while the spatial codes predicted -19% of rated power.

2. Level control (Contingency #7 of EPG) and pressure control reduce the reactor power from 30 to 19% during the course of the transient. High Pres-sure Coolant injection (HPCI) failure occurs in 23 minutes. The failure is assumed to occur by definition when the bulk PSP temperature reaches 190'F.

No Manual Rod Insertion (MRI) or Stand-by Liquid Control System (SLCS) was modeled c aing the transient.

3. Level and pressure control along with MRI will delay the time to HPCI failure. Although the reactor power can be reduced to 6% with 20 rods in-serted after the system depressurization, power spikes are expected because the void reactivity is comparable to the scram (or control rod) reactivity and both are smaller than the positive reactivity supplied from the Doppler and moderator feedbacks. By delaying the depressurization untti the time when

-32 rods have been inserted into the core, a reactor state would be reached where the negative scram reactivity would become the dominant feedback effect.

The scram reactivity should be large enough to lower the reactor power to the decay heat level. However, the pressure suppression pool (PSP) water tempera-ture is high and the HPCI failure will eventually occur unless the RHR cooling is activated. No SLCS was modeled during this transient.

4 The high pressure bollof f calculation (i.e., assumed HPCI failure) predicted that the power would drop to 4% at 150 seconds with the water level -4 feet below TAF. No Critical Heat Flux (CHF) condition was detected during the first 150 seconds. The combination of the Reactor Core Isolation Cooling (RCIC) and Control Rod Drive (CRD) flows is enough to sustain 2.7%

of rated power without loss of liquid inventory in the vessel. Thus, if the core thermal power remains below 2.7%, no fuel damage would be expected.

5. If the recirculation pumps do not trip during an MSIV closure ATWS, the predicted core power af ter 40 $ is around 80% of rated so that the steam flow rate is below the SRV maximum capacity which is 85% of full power steam flow rate. Of course, the power will drop when the jet pumps become un-covered. A peak pressure of 1140 psia was calculated. While the calculation showed that the reactor vessel was safe from an overpressurization failure, the water level was dropping rapidly. This calculation confirms the impor-tance of recirculation pump trip as the first operator action during an ATWS event.

, 6 The part of the EPG related to the ATWS issue must be re-evaluated l with the knowledge that the reactor power would be -19% of rated power with the water level at TAF, rather than 81 as first determined by the previous point kinetics calculations.

In conclusion, this report considers several aspects of the ATWS issue.

- vitt -

I I ,

The effects of lowering the water level to TAF and of the manual rod insertion have been studied along with the e f fect of the reactor pressure vessel depressu ri za tion. All of these mitigative actions utilize either the control rods or the void feedback mechanisms to introduce negative reactivity into the core in an attempt to bring the reactor power down, reducing the probability of a primary containment failure.

The results of these and similar calculations show the mitigative qual-i ity of dif ferent plant operator actions during ATWS and should help in the development of well defined emergency procedures for controlling the ATWS in such a manner that no plant damage or harm to the public would take place.

One important issue not yet adequately addressed in the SASA program is the effect of the SLCS activation on the progress of ATWS. It is recommended that best-estimate calculations with boron injection be performed in order to estimate the effectiveness of SLCS in bringing the reactor to hot shutdown condition. In our opinion, these calculations have to be perforned and assessed before the whole ATWS issue can be considered well understood.

1 i

l 1

1 4

- ix -

i

l l l l ACKNOWLEDGEMENT The authors would like to thank Dr. Robert Curtis and Dr. Bharat Agrawal of the Nuclear Regulatory Commission / Containment Systems Research Branch for their encouragement and guidance in performing this work. Discussions and i help supplied by Dr. Steve Hodge of Oak Ridge National Laboratory is grate-i fully ackilowledged. Furthermore, the lengthy discussions and advise offered by Dr. Hsiang-Shou Cheng of Brookhaven National Laboratory in developing the cross sections for RAMONA-38 deserves special thanks. The cooperation of the Tennessee Power Authority during this work was extremely useful and is deeply appreciated. The authors also express their gratitude to Dr. U.S. Rohatgi of Brookhaven National Laboratory for spending lengthy hours reading and comment-ing on the manuscript. Finally, the expert typing of Mrs. Dorothy Thompson and Mrs. Carmen Falkenbach is highly appreciated.

1 I

i i

l J

l i

-xi-

LIST OF TABLES Table Title Pa ge 2.1 Early Plant Response as Calculated by RAMONA-38.............

3.1 T VA S t a te Po i n t . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3.2 Sa fe ty a nd Re lie f Va l ve Se tpoi nts . . . . . . . . . . . . . . . . . . . . . . . . . . .

4.1 Stea dy-Sta te Nomi nal Ope ra ti n g Condi ti ons . . . . . . . . . . . . . . . . . . .

4.2 Se q ue nce o f Eve n t s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

LIST OF TABLES (APPENDICES)

A.1 R AMON A- 38 BWR Tra n s i e nt Ma t ri x . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

B.1 C ross Se cti ons Fo rma ts of RAMONA-38. . . . . . . . . . . . . . . . . . . . . . . . .

i l

- xiii -

i LIST OF FIGURES Fi gure Title Page 2.1 Heat Capacity Temperature Limit for the Browns Ferry Nu c l e a r P l a n t . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2.2 Transie nt Sequences Calcula ted Usi ng RAMONA-38. . . . . . . . . . . . . .

3.1 Burnup Evaluation Grid Used to Analyze Each fuel Type.......

3.2 BLEN02 Output of the Level 4 (24 levels total)..............

3.3 Comparison of BLEND 2 Generated Cross Sections Against TVA Plant Data............................................

3.4a Hydraulic and Neutronic Channels Used in Two Different Co m Mo de l s o f R AMON A- 38. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3.4b Hydraulic and Neutronic Channels Used in Two Different C o m Mo de l s o f R AMO N A- 38 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3. 5 Comparison of the RAMONA-3B Core Models Used in the ATWS Study Agai nst TVA Pl a nt Da ta . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3.6 Comparison of Reactor Power for RAMONA-3B/30 and R AM O N A - 3B / 10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3.7 Comparison of Core Inlet Flow Between RAMONA-3B/3U and R AMO N A - 3 0 / 10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3.8 Comparison of Predicted Total Reactivities from RAMON A-38/ 3D a nd RAMON A-3B/10. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3.9 RAMONA-3B Representation of a Boiling Water Reactor Vessel....................................................

3.10 RAMONA- 3d Ste a m Li ne Co n fi gu ra t i on . . . . . . . . . . . . . . . . . . . . . . . . . .

4.1.1 Transie nt 1 Wa te r Level Pre di cti on. . . . . . . . . . . . . . . . . . . . . . . . . .

4.1. 2 Transient 1. Water Injection into System with the Predicted Steam Leaving the Vessel and the Condensa-ti o n o f the Inje cte d Wa te r. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4.1. 3 Tra ns i e nt 1, Fl ow Ra te a t the SR V 's . . . . . . . . . . . . . . . . . . . . . . . . .

4.1. 4 Transient L. Condensation Rate on ECC Water Injected Into Vessel..............................................e 4.1. 5 Tra nsie nt 1, Core Inle t Subcool 1 n g. . . . . . . . . . . . . . . . . . . . . . . . . .

4.1. 6 T ra n s i e n t Sy s tem P re s s u re . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

- xiv -

i l

l l

l LIST OF FIGURES (Cont.)

l i Fi gure Title Page 1

4.1. 7 Transie nt 1 Core Ave ra ge Voi d Fracti on. . . . . . . . . . . . . . . . . . . . . .

! 4.1.8 Transient 1 Relative Power Prediction for the l F i r s t 10 s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4.1.9 Transie nt 1 Relati ve Powe r Pre di cti on. . . . . . . . . . . . . . . . . . . . . . .

4. 1. 10 Transient 1 Core Inlet Flow Rate to the Fuel Bundles........

4.1.11 Tra nsi e nt 1 Bypa ss Cha nne l Fl ow Ra te . . . . . . . . . . . . . . . . . . . . . . . .

4.1.12 Trans ie nt 1 Re acti vi ty Pre di cti ons . . . . . . . . . . . . . . . . . . . . . . . . . .

4.1.13 Tra nsie nt 1 Co re Ave ra ge Fue l Tempe ra tu re . . . . . . . . . . . . . . . . . . .

4. 1. 14 Tra ns ie nt 1 Ma ximum Co re Fue l Tempe ra tu re . . . . . . . . . . . . . . . . . . .

, 4.1.15 Transient 1 Relative Axial Powr Distribution at 800s

4. 1. 16 Tra nsie nt 1 Axi al Voi d Profi le at 800s . . . . . . . . . . . . . . . . . . . . . .

4. 1. 17 Transie nt 1 Relati ve Axi al Powe r at 1250s . . . . . . . . . . . . . . . . . . .

4. 1. 18 Tra nsie nt i Voi d Pro fi le at 1250s . . . . . . . . . . . . . . . . . . . . . . . . . . .

4. 1. 19 Comparison of PSP Water Temperature for Transient 1 and the Manual Rod Inse rtion Calculations.. ... ..... .. .....

I

{ 4.2.1 Co nt rol Rod Inse rti on Pa tte rn Pe rf o rme d at TVA. . . . . . . . . . . . . .

i Simulator MSIV Closure ATWS Session No. 7. . . . . . . . . . . . . . . . .

4.2.2 Control Rod Insertion Pattern used by RAMONA-38.............

4.2.3 Re l a ti ve Powe r fo r MR I Cal cul ati on . . . . . . . . . . . . . . . . . . . . . . . . . .

! 4.2.4 Core Avera9e Void Fraction for MRI Calculation..............

! 4.2.5 System Pre ssu re fo r the MR I Cal culati on. . . . . . . . . . . . . . . . . . . . .

j 4.2.6 Water Level Predi cti on for MRI Calculation. . . . . . . . . . . . . . . . . .

4.2.7 Core Inlet Subcooli ng for MRI Calculation. . . . . . . . . . . . . . . . . . .

4.2.8 Reactivity Predictions for the MRI Calculation..............

4.2.9 Core Flow into the Fuel Assemblies for the MRI Calculation...............................................

4.2. 10 Axial Power Distributions for 0, 269, and 389s for l

t he MR I C a l cu l a t i o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

t 4.2.11 Axial Power Distributions for 509, 749, and 1229s for MRI Calculation.......................................

I

~ XV -

i i

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

LIST OF FIGURES (Cont.)

Fi gure Title Pa ge, 4.2.12 Horizontal Power Distributions for MRI Calculation..........

4.2.13 Relati ve Power Prediction for MRI Calculation.. . ...... . .....

i 4. 2. 14 Core Average Void Fraction for MRI . Calculation..............

i I 4.2. 15 Reacti vi ty Pre di ctions for MRI Cal culation. . . . . . . . . . . . . . . . . . l 1

4.2. 16 Horizontal Power Distributions for the Calculation Wi thout Dep re ssu ri za ti on. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4.3.1 Water Level Prediction for the High Pressure Boil 4 O f f Ca l cu l a ti o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4.3.2 Water Injection and its Associated Amount of Conden-sation Along wi th the Exi sti ng Steam Flowrate. . . .... ..... .

4.3.3 System Pressure for the High Pressure Boil Off Calculation...............................................

4.3.4 Relative Power for the High Pressure Boil Off Ca l c u l a t i o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

j 4.3.5 Core Inlet Flow Rate to the Fuel Assemblies for the Hi gh Pre ssu re Boi l Of f Calculati ons. . . . . . . . . . . . . . . . . . . . . . .

i 4.3.6 Core Average Void Fraction for the High Pressure Boil Off Calculation...........................................

4.3.7 Components of the Total Reactivity for the High

, P re s su re Bul l Of f Ca l cu l a ti on . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

i

! 4.3.8 Core Inlet Subcooling for the High Pressure Boil Off Calculation...........................................

l 4.3.9 Condensation on Injected ECC Water for the High Pressu re Boi l Of f Calcul ati on. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4.4.1 Re lati ve Powe r for RPT Fai lu re Cal cu l ation. . . . . . . . . . . . . . . . . .

4 4.4.2 System Pressu re for PRT Failure Calculation. . . .. ... .... ... ..

j 4.4.3 Water Level Prediction for RPT Failure Calculation..........

4.4.4 Components of the Total Reactivity for the RPT Fa i l u re Ca l cu l a ti o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

! 4.4.5 Core Average Void Fraction for RPT Failure Calculation......

i 4

- xvi -

i

LIST OF FIGURES (APPENDICES)

Fi gure Title Page A.1 Simplified Flowchart of RAMONA-38 Calculational Logic.......

A.2 RAMONA-3B Representation of a Boiling Water Reactor Vessel....................................................

A.3 Comparison Between the Experimental Data and the Code Pre-dictions of Axial Void Fraction for FRIGG Test 313009.....

A.4 Comparisons Between the Experimental Data and the Code Pre-dictions of Axial Void Fraction for FRIGG Test 313001.....

B.1 Overview of 3-0 Cross Sections Generation Procedure.........

C.1 Overview of 1-0 Cross Sections Generatio Method.............

0.1 Total Condensation Rate on the Injected Water (best-e s t i ma te ca s e ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

0.2 Total Condensation Rate on the Injected Water (Sensi ti vi ty Study - Hi gh Conde nsation Case ). . . . . .. . . . . . . .

0.3 Co re Inle t Subcooli ng (be st-e s tima te case ) . . . . . . . . . . . . . . . . . .

0.4 Core Inlet Subcooling (Sensiotivity Study - High Co n de n s a t i o n C a s e ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

0.5 Total Reactor Powe r (be st-e s timate ca se ) . . . . . . . . . . . . . . . . . . . .

0.6 Total Reactor Power (Sensitivity Study - High Co n de n s a t i o n C a se ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

0.7 Supp re s si on Pool Wa te r Tempe ra tu re . . . . . . . . . . . . . . . . . . . . . . . . . .

- xvit -

l

1. INTRODUCTION l

l

1.1 Background

Anticipated Transient Without Scram (ATWS) is known to be a doMnant accident sequence for possible core melt in a Boiling Water Reactor (BWR). A l

recent Probabilistic Risk Assessment (PRA) analysis (Mays et al.,1982) for l

the Browns Ferry, Unit 1, Nuclear Power Plant indicates that ATWS is one of the dominant transients for core melt in a BWR/4 with the Mark I containment; the most dominant sequence being the failure of long term decay heat removal function of the Residual Heat Removal (RHR) system. Therefore, extensive analysis has already been performed to understand the BWR behavior during an ATWS. Examples of some of the earlier work are given by Aronson et al.

(1973), General Electric (1981), Neymotin and Saha (1984) and Hsu and Diamond (1985).

Much work has also been done to prevent and mitigate any adverse conse-quences of an ATWS. Recirculation pump trip and boron injection are two prom-inent mitigative actions. In addition, the BWR Owners' Group has proposed a new set of Emergency Procedure Guidelines (EPG) (1982) which require the operator, in the event of an ATWS, to reduce the safety injection flow rate so that the downcomer water level would drop from the normal level to the top of active fuel (TAF) and remain there. This would consequently lower the core flow rate, resulting in an increased core void fraction and a decreased reac-tor power beca.use of the negative void-reactivity feedback. A lower reactor power would reduce the steam generation in the core leading to reduction of the heat-up rate of the pressure suppression pool (PSP). Even then, the reac-tor vessel may have to be depressurized by manual control of the safety / relief valves as the PSP water temperature increases beyond a certain limit. The de-pressurization should also increase the core void fraction and help lower the reactor power.

The U. S. Nuclear Regulatory Commission (USNRC) and the Electric Power Research Institute (EPRI) sponsored several projects to determine the reactor power during a BWR ATWS with the downcomer water level maintained at TAF. The results of these efforts have been documented by Harrington and Hodge (1984),

W. C. Jouse (1984), Dallman et al. (1985), Hsu and Diamond (1985), Chexal et al . (1984a,1984b), and Peterson et al. (1984). Various computer codes, name-ly, BWR-LACP (Harrington and Hodge,1984), RELAPS (Ransom et al.,1982), TRAC-BD1 (Taylor et al., 1983), NATBWR (Healzer and Abdollahian, 1983), N0DE-B (Kendrick and Fisher,1976), and RETRAN-02 (McFadden et al.,1981) were used for this purpose. However, the predicted reactor powers for essentially the same transient were significantly different depending on the codes and models used. This created a need for further studies.

In a BWR, the vapor void fraction varies significantly in space, particu-larly in the vertical (or axial) direction. Because of strong void-reactivity feedback, a space-time neutron kinetics coupled with a thermal hydraulic model is necessary for ATWS-type analysis. Even for the simpler case of a full ATWS with symmetric fuel and control rod pattern, a code with one-dimensional (axial) neutronics is required to calculate the changing axial power shapes.

During the transient, the axial power distribution influences the axial void distribution which determines the driving head for natural circulation flow rate, which finally affects the total reactor power. Thus, for a best-esti-mate prediction of reactor power during an ATWS, the analysis tool must be able to calculate the axial power distribution as a function of time. The re-fore, the point kinetics codes (e.g., RELAPS, TRAC-BD1, BWR-LACP, etc.) with the fixed power shape, am not suitable for best-estimate analysis of BWR ATWS.

In the earlier ATWS studies (Harrington and Hodge, 1984; Dallman et al.,

1985) sponsored by the USNRC under the Severe Accident Sequence Analysis (SASA) program, the point-ki atics codes, namely, BWR-LACP and RELAPS/ MODI.6, l pre dicted core powe r to be -9% (Harrington and Hodge , 1984) and -8%

(Jouse, May 1984 and October 1983), respectively. However, EPRI used spatial kinetics codes and predicted core power of 15-18% (Chexal et al., ( 1984a ,

1984b) and Peterson et al., (1984)). Therefore, with so many different pre-dictions of power levels, NRC requested Brookhaven National Laboratory (BNL) to perform similar studies using the space-time kineti cs code, RAMONA-3B (Wulff et al.,1984; Ccnnell et al.,1984) with plant specific cross sections.

This report presents the results of the BNL study and their impact on the various mitigative actions suggested for a BWR ATWS.

1.2 Objective of Present Study As mentioned above, a space-time neutron kinetics code is needed for the best-estimate analysis of BWR ATWS where the spatial power distribution varies with time. RAMONA-38, developed at BNL under USNRC sponsorship, is a BWR transient analysis code with three-dimensional neut-on kinetics and nonequi-librium, nonhomogeneous therinal hydraulics. As such, RAMONA-38 is the most appropriate tool currently available to analyze transients such as ATWS where a strong coupling between neutron kinetics and thermal hydraulics exists.

The purpose of the present study is, therefore, to provide benchmark-type BWR ATWS calculations for the USNRC SASA program using the RAMONA-38 code.

The Browns Ferry nuclear power plant has been selected as the model plant, since the other NRC-sponsored studies conducted at Oak Ridge National Labora-tory (0RNL) and Idaho National Engineering Laboratory (INEL) had already used the same plant as the model for the BWR/4 plants with Mark I containment.

This would facilitate direct comparisons of the results obtained at various organizations for the same transient (s).

All transients considered in this study were assumed to be initiated by an inadvertent closum of all Main Steam Isolation Valves (MSIV). It was also assumed that the automatic control rod insertion system would completely fail.

These particular initiating events wem chosen because of the relatively high frequency (approximately one per mactor year) of MSIV closure and the severe challenge that an MSIV closure ATWS poses to the reactor and containment systems of a BWR plant. Although started from the same initial events, the transients considered in the present study did proceed in di f fe rent ways depending on the operator actions and equipment malfunctions. Of particular importance were the effects of dropping the downcomer water level to TAF, depressurizing the reactor vessel when the PSP water temperature reached a certain value, and manually inserting the control rods one-by-one. In addi-tion, this study would quantify the effects of equipment malfunctions such as i

l l

I loss of high pressure coolant injection (HPCI) and failure of recirculation pump trip (RPT). Results of this study would not only impreve our understand-ing of the BWR behavior during an ATWS and help us in formulating correct mitigative actions, but it would also provide an opportunity for assessing the adequacy of point-kinetics codes for ATWS application. Note that some of the operator actions such as manual rod insertion cannot be realistically analyzed by a point-kinetics or even a one-dimensional neutron kinetics code. For such transients, a three-dimensional neutron kinetics code such as RAMONA-3B is essential.

1.3 Brief Description of RAMONA-3B RAMONA-3B (Wulff et al.,1984), developed at Brookhaven National Labora-tory (BNL), is a best-estimate BWR systems transient analysis code with three-dimensional neutron kinetics and nonhomogeneous, nonequilibrium parallel-channel thermal hydraulics in the reactor core. This code is the result of many modifications and improvements to the RAMONA-III code acquired by BNL from Scandpower A/S, Norway. RAMONA-3B is currently available through USNRC, on a royalty-free basis, to any U.S. organization for the analysis of U.S.

reactors.

The code is particularly suitable for the analysis of ATWS-type tran-sients which require close coupling between the neutron kinetics and thermal hydraulics. Neutron kinetics and thermal hydraulics are the two major parts of the code, linked by heat conduction in the fuel rods. In addition, the code includes models for boron transport and all necessary BWR components such as jet pump, recirculation pump, steam line with safety / relief and other valves, safety injection systems, and a simple plant control and protection system.

Brief description of the neutron kinetics and thermal hydraulic models used in RAMONA-3B is given below. (See Appendix A for further details.)

1.3.1 Neutron Kinetics The neutron kinetks model of RAMONA-3B starts from a two-group diffusion theory. Simplificatwes are made to the thermal neutron leakage term which reduces the formulation to the 1-1/2 group, coarse mesh diffusion model (Borresen, 1971) in a three-dimensional rectangular coordinate system. Six delayed neutron groups are accounted for, and the code predicts transient three-dimensional fission power distribution in the reactor core. The code can also be used in the axial one-dimensional mode as a user option.

Decay heat from fission products is computed from the 1979 ANS Standard 5.1. All feedback mechanisms between the neutron kinetics and thermal hydraulics, namely, the void, moderator, doppler, boron, and xenon reactivity feedbacks, are taken into account. (See Appendix A for more details.)

1.3.2 Thermal Hydraulics RAMONA-3B employs one-dimensional models for two-phase flows with unequal phasic velocities, subcooled, saturated, or superheated liquid phase and with transient boron concentration. Four equations of vapor mass, mixture mass, momentum and energy conservation describe the coolant dynamics in the vessel.

Two equations of vapor mass and momentum conservation describe the acoustic effects from valve closure in the (adiabatic) steam lines. One boron mass conservation equation is used to predict the transport of boron in the reactor vessel.

A single . pressure is used in the reactor vessel to comoute all phasic properties. This technique eliminates the acoustic effects in the vessel and contributes significantly to the computing economy of RAMONA-38. However, the important acoustic effects in the steam lines am retained.

Closed-contour integral momentum equations am used through each of the parallel core channels to calculate the axial fluid velocities. Also, the partial differential equation for the mixture mass conservation is transformed into a simple quadrature in space. These modeling techniques greatly increase the computing speed of RAMONA-3B without significant loss in accuracy.

l The code accounts for nonequilibrium vapor generation and condensation, unequal phase velocities, wall shear and wall heat transfer through appropri-ate correlations. These include a CHF correlation and models for the post-CHF regime. However, RAMONA-3B restricts the vapor phase to be at saturation

! all the time.

The code has individual component modeling to accommodate BWR systems of U.S. de si gn. This includes models for jet pumps, recirculation pumps and steam separators. However, all recirculation loops and all steam lines are lumped together to one recirculation loop and one steam line, respecti/ely.

The code tracks the mixture level in the reactor vessel downcomer and accounts for steam condensation on the safety injection water when the downcomer water level drops below the feedwater spargers.

The thermal hydraulics and the boron transport equations are integrated by the explicit first-or6er Euler method, whereas the steam line dynamics is calculated by the fourth-order Runge-Kutta-Simpson method.

1.4 Report Outline Early plant response and various operator actions during a BWR ATWS are first discussed in Chapter 2. Also described are - the transients simulated with RAMONA-38 as part of the present study. In Chapter 3, the RAMONA-3B in-put model is described. The results of RAMONA-3B calculations for -varicus transients are presented in Chapter 4 Conclusions and recommendations are presented in Chapter 5. Finally, suggestions for future work are given in Chapter 6.

The re are several appendices giving the details of RAMONA-3B models, method of generating the Browns Ferry specific nuclear cross sections, 3-D to 1-D core model comparison and sensitivity of condensation model. Reade rs interested in the details of the present study should find these appendices useful.

2. TRANSIENT SCENARIOS DURING MSIV CLOSURE ATWS There are many possible initiating events for an ATWS, which by defini-tion is a failure to scram the reactor after an " anticipated" but abnormal transient occurs. From the General Electric (February 1981) study, the

, severity of all ATWS transients can be bounded by accident sequences initiated l by:

a) MSIV closure, j b) turbine trip, and

[ c) an inadvertently-open relief valve (IORV) i during power operation. For the Browns Ferry plant, the MSIV closure ATWS is the most severe because essentially all the steam produced in the reactor ves-sel is discharged into the pressure suppression pool. This results in a faster PSP water heat-up rate than that during other ATWS events. Consequent-ly, the plant safety systems, in conjunction with correct operator actions,

. must be activated rather quickly to maintain core cooling and containment in-

! tegrity until the accident can be controlled. Therefore, only the MSIV clo-sure ATWS sequences will be discussed in this report.

2.1 Early Plant Response Closure of all main steam isolation valves causes a rapid pressurization of the reactor vessel. This increased pressure collapses the voids in the core, and because of void reactivity feedback, the reactor power increases.

It is assumed that the automatic scram system fails to insert any control rods at the high power signal. As the power continues to increase, the fuel tem-perature rises and the negative Doppler reactivity feedback turns around the power increase. Meanwhile, the rising pressure causes the recirculation pumps to trip and SRVs to lift. These actions reduce the core flow and reactor ves-sel pressure. The net effect is an increase in the core void fraction and a reduction in the core power. All of these occur very quickly (within about 10 seconds) and the operator does not have sufficient time to react.

As the MSIVs close, steam supply to the feedwater pump turbines stops.

Therefore, the feedwater flow into the reactor vessel ceases in approximately 10 seconds. However, steam continues to be discharged through the safety /re-lief valves into the pressure suppression pool. This causes the downcomer water level to drop and activate the safety injection systems, i.e., HPCI and RCIC. Because of these safety injections, the drop in the water level and vessel inventory slows down and the reactor power stabilizes at approximately 28% of the rated core thermal power. However, essentially all the steam generated in the core continues to be discharged into the PSP, and the pool water temperature continues to rise. The timing of these events, as calc 01at-ed by RAMONA-38, is shown in Table 2.1.

By approximately 2 minutes, the operator should realize that an MSIV clo-sure ATWS is in progress. He should, therefore, be expected to perform the appropriate mitigative actions. These will be discussed in detail in the following section. However, if the operator fails to take any action, the suppression pool water temperature would reach the potential HPCI failure tem-1

perature of 190'F in approximately 15 minutes. Therefore, correct operator actions are of vital importance in mitigating the MSIV closure ATWS in a Browns Ferry type plant.

TABLE 2.1 Early Plant Response as Calculated by RAMONA-3B Event RAMONA-3B (sec)

MSIV closure starts 0.0 S/RVs start to open 2.4 Recirculation pumps trip 2.6 MSIVs completely closed 5.0 Maximum fuel temperature reached 6.3 Maximum vessel pressure reached 7.8 HPCI and RCIC activate 18.7 r 2.2 Various Operator Actions As soon as the operator realizes that an MSIV closure ATWS is in pro-gress, he attempts to scram.the reactor. Depending on the cause of the orig-inal scram failure, he may very .well be successful and the reactor will be.

shut down. However, for the purpose of the present study, it is assumed that the reactor scram cannot be achieved. The operator would then be expected to attempt a manual rod insertion. This procedure' calls for inserting the con-trol rods one-by-one which can be done from the control . room (in Browns Ferry another person may be sent to the auxiliary instrumentation room'to attempt an alternative method of manual rod insertion). However, this was not modeled in the present study. In any case, the manual rod insertion is a slow process since it takes about one minute to insert one control rod. Thus, it may take on the order of a half hour to significantly reduce -the reactor power by manual rod insertion alone. Therefore, it is prudent to take additional miti-gative action (s).

The SLCS designed to inject sodium pentaborate solution into the BWR reactor vessel is one of the major ATWS mitigation systems. At Browns Ferry, .

it would inject -13% by weight of sodium pentaborate solution with -13,800 ppm of boron into the lower plenum through a single vertical sparger at the rate of 50 to 56 gpm. Also, at Browns Ferry the SLCS must be activated manu-ally. During an ATWS, as the PSP water temperature heats up to 110*F, the operator is expected to activate the SLCS. As boron starts to enter the reactor core, it would absorb thermal neutrons, thus lowering the reactor power. However, a reduction in power would decrease void fraction which, in

. - _ = .- _ . _ ._ _ _ __

1 1

turn, would attempt to increase the power again. This competition between the ,

boron and void feedbacks would continue until the boron concentration in the '

core becomes high enough to overcome the void effects. Only then will the reactor become subcritical.

l It is difficult to determine the time required for reactor power to drop at the decay heat level by boron injection. The sodium pentaborate solution that enters the lower plenum has a specific gravity of -1.1, and it may have a tendency of settling there depending on the natural circulation flow rate through the downcomer, lower plenum and the core. The lower the flow rate, the higher the chance of boron stratification (or settlement) in the lower plenum. Therefore, during the low flow natural circulation mode, boron con-centration buildup in the core could be slow, and consequently, the reactor shutdown could be delayed.

To reduce the reactor power level during an ATWS, the BWR Owners Group has suggested, in the new Emergency Procedure Guidelines (EPG), that the oper-ator throttles the safety injection flow rates, i.e., HPCI and RCIC, so that the downcomer water level drops to TAF and remains there. This would reduce the driving head for natural circulation and consequently reduce the core flow rate. A reduction in core flow rate would increase the core void fraction, which would reduce the core power. At the time of formulation of the new EPG, it was believed that the reactor power would drop to -9% of the rated power with downcomer water level at TAF. (One of the main objectives of the present study is to evaluate the effect of lowering the water level.) Detailed dis-cussions on the new EPG are given by General Electric Co. (1982), Harrington and Hodge (1984), Dallman et al. (1985), and Jouse (1984).

In addition to the water level control, the operator under the new EPG would be expected to minimize the SRV cycling by lowering the reactor vessel pressure below the minimum SRV opening setpoint. The operator is also expect-ed to monitor the PSP water temperature as the water heats up due to steam discharge into the pool and to lower the vessel pressure in accordance with the heat capacity temperature limit (HCTL) curve. This curve for the Browns Ferry plant is shown in Figure 2.1.

The PSP cooling should also be activated as the pool water temperature starts to increase. However, at Browns Ferry, certain interlocks and RHR system logic designed to enhance plant recovery from Loss-of-Coolant Accident (LOCA) would dramatically interfere. This is described in Section 4.1.4 of Harrington and Hodge (1984). Because of the complexity of maintaining the RHR system in the pool cooling mode with reactor downcomer water level at TAF, no PSP cooling is assumed in this study.

In summary, the operator would be expected to take the following actions to reduce the adverse effects of an MSIV closure ATWS:

a)- Initiate manual scram.

b) Activate the SLCS to inject baron into the reactor vessel.

4 i

i l

350 - -

300 -

F

~

250 -

b E

a w

200 -

i 1

i N K eo -

, 5 IS 10 0 -

i 50 -

l i

, i i i '

o 200 soo 600 800 2 00 REACTCR VESSEL PRESSURE (PSIA)

Figure 2.1 Heat capacity temperature limit for the Browns Ferry Nuclear Plant.

Reactor vessel pressure - suppression pool temperature combinations that lie within the shaded area are prohibited.

t

c) Throttle HPCI and RCIC to drop the downcomer water level to TAF and maintain there, as per the proposed EPG. This is called " level con-trol".

d) Control the reactor vessel pressure to minimize SRV cycling. Also, reduce pressure by opening SRV so that the heat capacity temperature

limit curve for the PSP is not violated. This is called " pressure control".

2.3 Transients Simulated with RAMONA-3B It is now clear that during an MSIV closure ATWS, a number of automatic and manual actions will take place. The automatic actions are: a) recircula-tion pump trip (RPT) at high pressure, and b) HPCI and RCIC activation at low downcomer water level. The manual or operator actions have been discussed in the previous section. An MSIV closure ATWS transient can, therefore, progress in many different directions. Because of resource limitations, only four transient sequences have been analyzed using the RAMONA-3B code. The tran-sients analyzed were selected after consultation with the BWR-SASA partici-pants, namely, NRC, ORNL, and INEL.

Four transient sequences analyzed are depicted in Figure 2.2 in an event tree format. The initial event for all transients is the closure of all MSIVs with total failure to scram. Moreover, every scenario is assumed to have the CRD hydraulic system operational to inject water into the vessel along with the assumption that the SLCS is not available. For Transient #1, it is assumed that the RPT, HPCI and RCIC systems will work as designed. The opera-tor will take the level and pressure controls as per the proposed EPG. How-ever, the operator will not resort to the manual rod insertion nor activate the SLCS. (This assumption is made primarily to isolate the effects of level and pressure controls from the boron injection eff;ct. It is recognized that

, in reality the operator would activate the SLCS, and a RAMONA-3B calculation including the effects of boron injection is highly desirable. This is pro-posed as future work in Chapter 6.)

Transient #2 is similar to Transient #1, except that the effect of manual rod insertion is now included. Thus, the operators are assumed to insert con-trol rods one-by-one while they are controlling the downcomer water level at TAF and vessel pressure in accordance with the new EPG.

Transient #3 examines the consequences of a HPCI failure. However, it is assumed that the RCIC system with -600 gpm of flow and the control rod drive (CRD) flow with -100 gpm are available. The downcomer water level during this transient would drop below TAF and the operator will not be able to per-form level control.

Finally, Trinsient #4 examines the consequences of a recirculation pump trip failure. The main objective of this study is to determine the maximum vessel pressure and the reactor power level in case the RPT logic fails, al-though the probability of such a failure is rather low.

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

1 I

Reci rculation Manual Rod Level Pressure i

Pump Trip HPCI RCIC Insertion Control Control i

' Transient 2 Transient 1 Success

' t 5 Transient.3 i

MSIV Closure ATWS

+

Transient 4.

p,339 ,

FIGURE 2.2 Transient Sequences Calculated Using RAMONA-3B 1

~ ~ - - - _ - - _ - .

3. RAMONA-3B INPUT MODEL 3.1 Neutroaic Data and Core Modeling Modeling of the reactor core is the most signi ficant aspect of a BWR power plant simulation because of the very strong nuclear feedback effects associattd with the core. This section will discuss how the core model was developed for the RAMONA-3B calculations presented in Chapter 4.

3.1.1 3-D Model Although the plant to be evaluated in this study was the TVA Browns Ferry (BF) Unit #1, Cycle 6, the core data available from TVA was for the BF Unit

1. 3, Cycle S. This was accepted for several reasons. First, the TVA staff assured us that the core loading and fuel strategy between the two BWR/4 reac-tors were almost identical. Both cores were in their equilibrium fuel cycle, which means that any set of cross sections generated would not only describe the BF- plants, but would also have general applicability to any BWR/4 in its equilibrium fuel loading phase of the P8X8R type fuel. Finally, the informa-tion needed for the BF Unit #3, Cycle 5, was readily available for analysis.

A complete description of the core was provided by TVA (Keys,1984). The neutronic data corresponded to the output from the utility process computer for several different operating points during the lifetime of Cycle 5. This information showed that five different fuel types had to be analyzed by a fuel assembly burnup (or depletion) code. The code used for this purpose was a 2-D neutron transport program named CASMO (Ahlin,1978). The processing strategy (or grid) used to evaluate each of the five fuel types can be seen in Figure 3.1. This grid allows one to evaluate the effect of burnup on the two-group macroscopic cross sections during the lifetime of a particular fuel type. For details on the cross sections format used in RAMONA-3B and on the CASMO-RAMONA-3B interface, see Appendix B.

The operational point chosen for analysis was at a core average burnup of E=8876 MWD /MT. This point was chosen since it was near E0C-5 when most of the control rods had been withdrawn, which poses the greatest challenge to core during an ATWS.

The CASMO output was processed by a code called BLEND 2. The core modeled by BLEND 2 was a full core using 24 axial levels in 764 fuel assemblies. This modeling is dictated by the TVA process computer, which uses the same nodali-zation to collect the data (i.e., the exposure and void histories). However, full core modeling with 18,336 nodes (24 x 764) would be very expensive, and should be avoided unless the problem necessitates it.

Furthermore, upon evaluation of the BLEND 2 output, no true 1/8 or 1/4 symmetry in the core arrangement was found (Figure 3.2). The RAMONA-3B core model generated for the 1/8 and 1/4 symmetry was made to be an average 1/8 and 1/4. This means that the cross section set which had the highest frequency of occurrence in any given location was used, and not the cross section set which belonged to any specific 1/8 or 1/4 quadrant.

i

. , , w -4

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

I 9

. VOID INSTANTANE0US Exposure in MWD /kgU HISTORY VOID-C0 EFFICIENT V,% a,% 0 1 2 3 4 5 6 7 8 9 10 12.5 15 17.5 20 22.5 25 27.5 30 0 xxx x x x x x x x x x x x x x x x x x x x x x x x x x 40/70- x x x. x x 0 0-Doppler . x x x x x 0-Control Rod x x x x x x

, 0-TM x x x x x 40/70-Control Rod x x x x x x 40 40- xxx x x x x x x x x x x x x x x x x x x x x x x x x -x 0/70- x x x x x 40-Doppler x x x x x 40-Control Rod x x x x x x 40-TM x x x x x 0/70-Control Rod x x x x x x 0 70- xxx x x x x x x x x x x x x x x x x x x x x x x x x

0/40- x x x x x 70 70-Doppler x x x x x 70-Control Rod x x x x x x 70-TM x x x x x 0/40-Control Rod x x x x x x TOTAL # POINT /PER FUEL TYPE - 198.

Figure 3.1 Burnup Evaluation Grid Used to Analyze Each Fuel Type. Each "X" Represents a Complete Set of Two-group Macroscopic CrossSection I

l PL APE  % CONOSIi10P4 DESCRIPflope 0 0 0 0 0 0 0 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 0 0 0 0 0 0 0 0 C 0 0 0 0 0 0 3 3 6 6 6 6 6 6 6 6 6 6 6 6 3 3 0 0 0 0 0 0 0 f 0 0 0 0 0 3 3 3 6  % 3 2 6  % 2 2 2 6 2 3  % 6 3 3 3 0 0 0 0 0 0 0 0 0 0 3 6 6 3 9 II 3 2 IQ 3 3 10 2 3 II 9 3 6 6 3 0 0 0 0 0 0 0 0 0 3 2 6 3 9 3 6 3 6 3 6 6 3 6 3 6 3 9 3 6 3 3 0 0 C 0 0 0 2 3 2 4 3 9 3 9 3 10 3 9 3 3 9 3 10 3 9 3 9  %  % 3 3 2 0 0 0 0 2 6 6 2 3 3 6 3 3 3 6 3 3 3 3 6 3 3 3 6 83 13  % 6 6 3 0 0 0 3 3 6 2 9 3 3 3 10 3 3 3 10 3 3 IS 3 3 3 10 3 3 13 9 2 6 3 3 0 2 2 6 2 9 3 6 3 9 3 6 3 9 3 7 7 3 9 3 6 3 10 3 6 3 9 2 6 3 2 3 6  % s 3 9 3 IO 3 9 3 10 3 10 3 3 10 3 10 3 10 3 10 3 9 3 9

• 6 3 3 6 3 13 6 3 II 3 6 3 3 3 6 3 3 3 2 6 3 3 3 6 3 13 3 6 II 3 6 3 3 6 3 's 3 10 3 3  % 9 3 3 3 9 3 3 IO 3 3 3 10 3 3 3 IO 3 3 3 6 3 3 6 6 2 6 3 6 3 9  % 7 3 9 3 7 7 3 30 3 6 3 9 2 6 3 6 2 6 6 3 3 6 4 9 3 9  % 9 3 10 2 9 3 6 3 3 6 3 9 2 10 3 9 3 9 3 9  % 6 3 3 6 2 3 6 3 3 3 6 3 3 3 6 3 3 3 3 6 3 3 3 7 3 3 3 6 3 2 6 3 3 6 2 3 6 3 3 3 6 3 3 3 6 3 3 3 3 6 3 3 3 6 3 3 3 6 3 2 6 3 3 6  % 9 3 9  % 10 3 9 3 9 3 7 3 3 6 3 9 2 10 3 9 3 9 3 9  % 6 3 3 6 6 2 6 3 6, 2 9 3 7 2 10 3 7 6 3 9 2 7 3 9 3 6 3 6 3 6 6 3

' 3 6 3 3 3 10 3 3 3 10 3 3 3 10 3 3 9 1 3 3 9  % 3 13 to  % 3 3 6 3 3 6 2 13 6 3 IS 3 6 3 3 3 6 2 3 3 2 7 3 3 3 6 3 3  % 6 3 3 6 3 .

3 6  % 9 3 10 3 l'0 3 9 2 10 3 IO 3 3 10 3 10 3 9 3 10 2 9 3 9  % 6 3  !

2 2 6 3 9 3 6 3 9 3 6 3 9 3 7 7 3 9 3 6 3 9 3 6 3 9 2 6 2 2 0 3 3 6 2 9 3 3 3 10 3 3 2 10 3 3 to 3 3 3 10 1 3 3 9 2 6 3 3 0 1 0 0 2 6 r  % 13 13 6 3 13 3 6 3 3 3 3 6 13 13 3 6 3 13 2 6 6 3 0 0 o 0 3 3 3  % 2 9 3 9 3 10 3 9 3 3 9 3 10 3 9 3 9  %  % 3 3 2 0 0 g

w 0 0 0 0 3 2 6 3 9 3 6 3 6 3 6 6 3 6 3 6 3 9 3 6 3 3 0 0 0 0 0 0 0 0 0 3 6 6 3 9 13 3 3 10 3 3 10 2 3 13 9 3 6 6 3 0 0 0 0 0 g

0 0 0 0 0 3 3 3 6  % 3 3 6 2 3 3 2 6 2 3  % 6 3 3 3 0 0 0 0 0 0 0 0 0 0 0 0 3 3 6 6 6 6 6 6 6 6 6 6 6 6 3 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 3 3 3 3 3 3 3 3 3 3

• 3 3 0 0 0 0 0 0 0 0 l

1 Figure 3.2 Blend 2 Output of the Level 4 (24 levels total). A number corresponds to a '

Cross Section Set Number (ranges between 1 to 13 in this example). Zero means no fuel at that location.

The BLEND 2 code was run twice; once to get a library of 13 cross sections sets which had the final selection criteria of 6E = S291 MWD /MT-and 6V = 0.39, and a second library of 20 cross sections sets which had a final selection criteria of 6V = 4233 MWD /MT and 6V = 0.31. The initial simulation was done

using a .1/8 core representation (101 channels with 24 axial levels = 2424 neutronic nodes), and 9 hydraulic channels (i.e., 216 hydraulic nodes). The boundary conditions can be found in Table 3.1, while the predicted steady state results of RAMONA-3B are shown in Figure 3.3. Two different libraries of cross sections sets were generated to determine the effect of the group's acceptance criteria (i.e..' 6E and 6V) on the overall result. Between these two, no effect of the criteria was evident.

TABLE 3.1 TVA Statepoint R-38 . R-3B i

TVA (13 x-sections) (20 x-sections)

! Power (MWt ) 3198 3198 3198 ,

System Pressure (MPa) 6.9154 (1003 psi) 6.9154 6.9164 Keff 1.00 1.0015 1.0054 Core Avg Void (%) 39 41 40 l Core Flow (kg/s) 1.29 E+4 (102.19 Mlb/hr) 1.29 E+4 1.29 E+4 i

! To finalize the core model, a nodalization study was performed to deter-1 mine the optimal representation. Based on that, the case with 20 cross sec-t ton sets, 24 axial levels,1/8 symmetry, and 9 hydraulic channels, became the

] reference calculation for all subsequent calculations (since it had the most amount of modeling detail). For the MSIV closure ATWS transient, the i RA;40NA-3B code with this model was run out to about 100s.- It was discovered that both calculations with 13 and 20 cross sections sets libraries produced the same results. Thus, the set of 13 was always used. This study also re-vealed that no less than 7 hydraulic channels could be used with the 1/8 core, and at least 8 hydraulic channels are needed for the 1/4 core symmetry case.

Finally, it was found that as low as ten axial levels could be used (by volume averaging when collapsing levels) for modeling of these transients. However, twelve axial levels were used in the final calculation.

Consequently, the nodalization study showed that the library of 13 cross

, sections sets can be used with 12 levels and 7 hydraulic channels for the 1/8 core symmetry (101 x 12 = 1212 nodes in core), and 8 hydraulic channels for the case with 1/4 core symmetry (i.e.,191 fuel assemblies x 12 levels = 2292 nodes in core). These models were used in performing the '-D calculations

}

l j

1

t 1

4 1

I I I I I I I I I i I i i I i 1

--- TVA PROCESS COMPUTER

- 13 CROSS SECTION SET x ---- 20 CROSS SECTION SET

.w 1 3: 1.5 -

o Q.

. - . _r ~ - -

l p I.O -

,/* -

1 <

I

\\ '

w m

0.5 _

! o,o i I I I I I I I I I I I I I i j O 20 40 60 80 100 120 14 0 160-1 HEIGHT (in.)

i 4

1-Figure 3.3 Comparison of Blend 2 Generated Cross Sections Against TVA Plant Data

discussed in this report. The hydraulic channel and neutronic channel nodings can be found in Figure 3.4 Comparisons with the plant data for all the finalized core models are shown in Figure 3.5.

3.1.2 10 Cross Sections While preparing for scoping studies, a set of 1-D (axial) cross sections was also generated from the core model with 1/8 core symmetry and 12 axial levels. Detailed information concerning the collapsing from the 3D to the ID core model is discussed in Appendix C. The comparison of axial power distri-butions for the 3-0, 1-D, and plant data is seen in Figure 3.5. To verify the 1-D set, the 1-D and 3-D test cases were run for the MSIV closure ATWS scenario. The results are presented in Figures 3.6 to 3.8.

In Figure 3.6, it can be seen that for the identical transient scenario the two different models produced almost overlapping power histories. In Fig-ure 3.7, the core inlet flow rate is also seen to be essentially identical.

These results verify the correctness of the equivalent 1-D set. However, the most convincing verification comes from Figure 3.8 where a plot of the total core reactivity produced in 1-D and 3-D calculations proves that, for _the same external stimuli, both core models produce nearly identical responses. This 1-D cross sections set was used for the High Pressure Boil-Off case discussed in Section 4.3.

The ability to run RAMONA-38 in either the 3-D or 1-D mode is a very use-ful option. The verification of the 3-D to 1-D collapsing method is straight-forward in RAMONA-3B, since all calculations can be performed with the same thermal hydraulics.

3.2 Thermal Hydraulics and Systems Modeling The typical RAMONA-3B representation of the Browns Ferry reactor (Figure 3.9) includes a downcomer, lower plenum, core, riser, steam separator and the steam dome with four steam lines lumped together (Figure 3.10). The combined steam line was modeled using six computational cells between the steam dome and the main steam isolation valve (this nodalization was based on results of a sensitivity study). Also, the steam line was equipped wi.th an MSIV, four banks of safety and relief valves and turbine valves. Each of the four SRV banks in the present calculation had specific opening and closing pressure set l

points, as well as rated steam flow rates, as shown in Table 3.2.

Hydraulic Channel-1 1 1 1 1 2 2 2 2 2 3 3 3 4 4 1 1 1 1 2 2 2 2 2 3 3 3 4 4 1 1 1 2 2 2 2 2 3 3 3 4 4 1 1 -5 5 5 52 3 3 3 4 4 1 5 5 5 5 3 3 3 3 4 4 5 5 5 6 6 6 6 6 4 4 5 5 6 6 6 6 6 4 4 5 6 6 6 6 4 4 6 6 4 4 4 6 4 4 4 4

Neutronic Channels a

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 -l29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 i

Figure 3.4a Hydraulic and' Neutronic Channels Used in two Different Core Models of RAMONA-3B

! I j l l

Neutronic Channels 1- 2 3 4 5 6 7 8 9 10' 11 12 13 14 15 16 17 18 19 20 21 22 23 24~ 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 Hydraulic Channels 1 1 1 2 2 2 2 3 3 3 3 3 3 4 4

. I 1 1 2 2 2 2 3 3 3 33 3 4 4

! 1 1 1 2 2 2 2 3 3 3 3 3 3 4 4 1 1 1 2 2 2 2 3 3-3 3 3 34 4 1 1 1 2 2 2 2 3 3 3 3 3 3 4 4 5 5 5 5 5 5 5 6 6 6 6 6 6 4 4 5 5 5 5 5 5 5 6 6 6 6 6 6 4 4 5 5 5 5 5 5 5 6 6 6 6 6 4 4 5 5 5 5 5 5 5 6 6 6 6 4 4-

, 7 7 7 7 7 7 7 6 6 6 6 4 4 7 7 7 7 7 7 7 6 6 4 4 7 7 7 7 7 7 7 6 4 4 3 7 7 7 7 7 7 7 6 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 Figure 3.4b Hydraulic and Neutronic Channels Used in two Different Core flodels of RAfiONA-38 1

I I I l.2 - -

/' + ~ --o-

~ A.

l g ,. A . ,g-o . :e g 1.0

) -

/ %Y

$0.8 -

/ DATA _

/ --- 1/4 CORE MODEL (3D)

$- --l/8 CORE MODEL (3D) gj 0.6 -

o ID SET FOR 1/8 CORE -

tr 0.4 - -

0.2 I I I O 100 200 300 400 HEIGHT (cm)

Figure 3.5 Comparison of the RAMONA-3B Core Models Used in the ATWS Study Against TVA Plant Data COMPARE 3D/1D MSIV CLOSURE ATWS RELATIVE POWER VS. TIME 3-RAMONASB_13D P 25 . ,,,,BAMQNAUB/1J2,.

g 3: .

o 8-D* l

$ 155 15 E

Z OEi i j mij H %%'.'#J' JOE",th' ' t'i'i O , . , , ,

0.0 200.0 400.0 000.0 800.0 1000.0 1200.0 TIME (S)

Figure 3.6 Comparison of Reactor Power for RAMONA-3B/3D and RAMONA-3B/10 l

COMPARE 3D/1D MSIV CLOSURE ATWS l

b FUEL BUNDLE INLET FLOW RATE VS. TIME p 10.0 g 3AYONA3B/3D N E0- 3AYONA3B/1D e

b &O-g N

M 4.o .' \

O 10-N ^

4 YrV i , ,w'wv V W W $^

0.0 ' , , , , ,

0.0 200.0 400.0 000.0 800.0 1000.0 1200.0 TIME (S)

Figure 3.7 Comparison of Core Inlet Flow Between RANONA-3B/3D and RAMONA-3D/10

l COMPARE 3D/1D MSIV CLOSURE ATWS PERTURB REACTIVITY VS. TIME 500.0 I '

+ a0-

, F f

l c

5 b* RAMONA3B/SD RAM,QH.A3B/1D ___

0.0 2d0.0 4$0.0 000.0 8d0.0 1000.0 1200.0 TIME (S)

Figure 3.8 Comparison of Predicted Total Reactivities from rat 10NA 3B/3D and RAMONA 3B/10

l l

STEAMLINE WITH MSIV, S/R VALVES l 1 "l STEAM DOME STEAM SEPARATOR D

g _ _ _ _ _ _

, _ _ _ _ _ _ _ _ ,n S ------

_ _ _ _ _ _ _ - N E

1. "C f

R _ _ _ _ _ _ _ _ o

______ M l _ _ _ _ _ . E i _ _ _ _ _ _ _ _ p C

JET O ___ pyyp R

E .--

LOWER PLENUM 2

_ ____ - - - _ - - = ..=

i t l ;1 i t i t

LOWER PLENUM i l

l Figure 3.9 RAMONA-3B Representation of a Boiling Water Reactor Vessel Byposs Volve E Sofety and Rellet g Volves (S/RV) a =

g Main Steam Tur bin e Stop Valve Isolation Valve ---

e i g (

i . . - - , , ,

i 6 Steam i ' ' Mo$i  ! ! 8 8 l 8 ! ! 8 l l l '

8 i ' I i i Dome l I 1 i e e e

[ ,

I 1 ' I I I I I i 1 f I I , e s I I  ! I I  ! i I I i 1 g ( ) g g g IMSIV node MIST node i

Figure 3.10 RAf10NA-3B Steam Line Configuration TABLE 3.2 Safety and Relief Valves Set Points

• Bank 1 2 3 4

1. (4 Relief) (4 Relief) (3 Relief) (2 Safety)

High 77.22 77.91 78.60 87.22 Pressure (o

10gening),

Pa Delay Time 0.0 0.0 0.0 0.0 (opening),

sec l

l Low 73.77 74.46 75.15 83.77 '

Pressure (closing),

105 Pa Delay 0.0 0.0 0.0 0.0 Time (closing),

sec Total 108.64 108.64 81.48 54.32 Rated Flow (kg/sec)

• It is known that this configuration does not presently exist at the plant; however, it was assumed to be comparable to previous calculations that needed to be benchmarked.

The feedwater spargers were located in the upper part of the downcomer, whereas the jet pumps and the recirculation loops were associated with the lower part of the downcomer region. Note that, in accordance with the RAMONA-3B approach, the two recirculation loops of the actual plant are lumped together (validity of this approach is based on the fact that the recircula-tion pumps are tripped simultaneously and no core inlet flow asymmetry is expected). Eleven hydraulic cells were used in modeling the one-dimensional downcomer.

The lower plenum (two connected components in RAMONA-38) was modeled with five hydraulic cells and was coupled to the seven parallel hydraulic channels representing the reactor core region for the 1/8 core model. Six of them were haated channels (in-bundle flow region) and the seventh was the integral by-pass channel. All channels were modeled using 12 axial hydraulic cells

-(corresponding to 12 neutronic. levels in the neutronic " channels"). The 1/8 core symmetry option was chosen for Transient il calculation based on the ,

assumption that a 1/8 slice of the core was representative of the whole core  !

(the slice is composed of the core region between two vertical planes which L have the core geometrical axis of symmetry in common). The one-dimensional l riser including the upper plenum was connected with the core exit and was  :

modeled using five hydraulic cells. l Each of the transients analyzed required a pressure suppression pool '

water temperature and inventory calculation. A stand-alone code was used that i treated the PSP with the lumped-parameter approach. Recently, a corresponding lumped-parameter model has been developed and implemented in RAMONA-38 (L.

Neymati n , 1985) .

i i

I l

l

4 RAMONA-3B RESULTS AND DISCUSSION This chapter presents the RAMONA-3B results of the four transient se-quences discussed in Section 2.3. They are referred to as a) Level and Pressure Control (Transient 1) b) Manual Rod Insertion (Transient 2) c) High Pressure Boil-off (Transient 3) d) Recirculation Pump Trip Failure (Transient 4) 4,1 Level and Pressure Control (Transient 1)

As described in Section 2.3, the transient was assumed to be initiated by an inadvertent closure of all MSIVs starting at the full power corresponding to the end of Cycle 5 (Table 4.1). The SRVs were supposed to operate as de-signed and the recirculation pumps to be tripped upon reaching the high pres-sure set point [7.83 MPa (1135 psia) with 0.53 sec delay]. The feedwater was lost by 8 seconds and the plant was left in an automatic' mode of operation un-til 150 seconds. At this time, operator recognized that an MSIV closure ATWS was in progress and started to act according to the Emergency Procedure Guide-lines, i.e., the downcomer water level was gradually lowered to the TAF level.

At the same time the reactor vessel was slightly depressur'ized [to approxi-mately 6.5 MPa (943 psia)] in order to avoid excessive SRVs cycling.

Lowering the water level was accomplished by reducing the ECCS (HPCI and RCIC systems) injection flow rates. The safety injection systems had been activated by a low water level signal earlier in the transient. This level is approximately 2.7 meters (8.9 ft) above the TAF elevation. Thereafter, the downcomer water level was kept at TAF by controlling the HPCI and RCIC injec-tion flow rates until the end of the transient calculation. Along with the water level control, another safety measure exercised by the plant operator was modeled in the calculation: the reactor pressure vessel was depressurized according to the PSP heat capacity temperature limit curve (Figure 2.1). It should be noted that the present transient scenario did not include such miti-gating activities as boron injection or manual rod insertion (the latter is covered in the following section). The sequence of events, including some im-portant system parameters, is shown in Table 4.2. The transient calculation was run up to 1500 seconds.

Selected results of the calculation are shown in Figures 4.1.1 through 4.1.19. The discussion will start with the results directly related to the plant response to the lowering of the water level, which is the main EPG-im-pcsed operator action. From the very beginning of the transient, the vessel water inventory was depleting because the ECCS water injection could not make op for the steam losses through the S/R valves. This resulted in the water level dropping by approximately 4 meters (-13 ft) below its steady-state location by the time the operator assumed control of the plant (150 sec). At tnt p9 int, the operator reduced the HPCI and RCIC flow following the EPGs in

TABLE 4.1 Steady-State Nominal Operating Conditions Thermal Power, MWt 3198 Pressure, Pa (psi) 6.92 x 106 (1003.7)

Mass Flow Rate of Coolant at: ,

Core Entrance, kg/sec (Ibm /s) 12870 (28373)

Steam Line/Feedwater, kg/sec (Ibm /s) 1592 (3510)

Feedwater Temperature, C ('F) 191.2 (376.2)

Core Inlet Subcooling, C ('F) 11.5 (20.7) .

Core Bypass Flow / Total Core Inlet Flow, % 10.6 Recirculation Drive Flow, kg/sec (lbm/s) 4304

=

TABLE 4.2 Sequence of Events Event (sec)

MSIV closure starts 0.0 S/RVs start to open 2.4 Recirculation pumps trip 2.6 Maximum core-averaged fuel temperature is reached, 724 C (1335"F) 4.1 MSIVs are completely closed 5.0 Maximum system pressure is reached, 8.9 MPa (1276 psi) 7.8 HPCI and RCIC injection activated 18.7 Operator assumes control 150.0 a) Water level dropped and maintained at TAF b) Operator begins manual rod insertion 3

1 i

order to further lower the level to TAF. As a consequence of this action, a ,

drop in the core flow rate occurred (because of decrease in driving head) I which resulted in an increase in the core void fraction. The negative void i reactivity feedback reduces the total reactor power, which is the ultimate l 9 goal of the operator's actions.

It can be seen in Figure 4.1.1 that the rate of the level drop increases considerably when the operator reduces the safety water injection rate.

Finally, the downcomer water level reaches the TAF elevation at about 240 seconds and is kept there until the end of the transient. We have assumed a

" perfect" operator for this simulation. Two lines on Figure 4.1.1 mark axial locations of the TAF and the feedwater spargers in the downcomer region in the RAMCNA-3B elevation referencing scheme.

The operator actions are depicted more clearly in Figure 4.1.2 showing the S/RV flow rate (mass leaving the system) and the combined HPCI and RCIC flow rate (mass entering the system), thus reflecting the vessel inventory as well as the water level histories. The anticipated reduction in SRV cycling resulting from the initial vessel depressurization at 150 seconds by the operator can be seen in Figure 4.1.3 It should be mentioned that in the course of this project a need for a model specifically representing steam condensing on the cold safety injection water.was recognized and a corresponding model was developed and implemented into the RAMONA-3B code (see Appendix A). Since the cold water is injected through the feedwater spargers, the condensation starts as soon as the down-comer water level drops below the sparger elevation. The condensation model takes into account condensation on cold water jets (96 jets each of 44 mm diameter in the present calculation) and on the downflowing water film which is composed of the water exiting from the steam separators and injected ECC water. The film may be either subcooled or saturated depending on the conden-sation rate on the cold water jets. Figure 4.1.4 shows the predicted conden-sation rate as a function of time. The oscillations seen in the figure re-flect variations in the ECC injection rate that resulted from the water level control activities because the condensation rate is proportional to the ECC injection rate.

An important system parameter af fected by the condensation on ECC water in the upper downcomer region is the water subcooling at the core. entrance.

The reactor power is sensitive to the core inlet water subcooling through the moderator temperature and void reactivity feedbacks. (A sensitivity study in-vestigating the effect of core water inlet subcooling on reactor power was conducted with RAMONA-38. The results indicated that a reduction in core thermal power was achieved through an increase in the core void fraction. See Appendix D for further details.) As seen in Figures 4.1.1 and 4.1.4, af ter approximately 40 seconds from the initiation of the transient, the highly sub-cooled water was being injected into the steam environment in the upper down-comer region because the feedwater spargers were uncovered. Therefore, the core inlet subcooling, as shown in Figure 4.1.5, becomes a function of the steam separator water return flow, ECC water temperature, ECC injection rate, '

and the condensation rate. By 25 minutes of the transient, the subcooling at the core inlet was about 10'C (18 F).

3-D BROWNS FERRY WATER LEVEL IN DOWNCOMER VS. TIME 4

COLL.WAT. LEV.

j af WAT. LEV a  : 2 y 0-[l TOP OF DOWNCOMER . 0 3

W d 4 ,

-- .- FEED WATER SPARGEP $

- -s.ss g d

$ - qg ~19%

-14%

a2i b[hl) 3 m ,.... ..

0.0' .. .. .. ., ... , .,

0.0 200.0 400.0 600.0 800.0 1000.0 1200.0 1400.0 1000.0 TIME (S)

Figure 4.1.9 Transient 1 Relative Power Prediction i

l

3-D BROWNS FERRY b FUEL BUNDLE INLET FLOW RATE VS. TIME y 12.0 10.0 5 -22.05 0

$ E0i g CC 1a Q 6.0i -13.23

= 3 4.0i g -16% -lag b

cr., E05 y.ovv......r......-....

- 4 41 0.0 ' .. ,. , .. , .. ..

0.0 200.0 400.0 000.0 800.0 1000.0 1200.0 1400.0 1600.0 TIME (S)

Figure 4.1.10 Transient 1 Core Inlet Flow Rate to the Fuel Bundles 3-D BROWNS FERRY BYPASS CHANNEL FLOW RATE VS. TIME 16000 m INLE? FLOW W OUTL6T FLOW h 1000.0-

~~~ ~~

-2204.6 E

500.0- -1102.3 m -

~

) )

-600.0 ,. .. , .. ., , ..

0.0 200.0 400.0 000.0 800.0 1000.0 1200.0 1400.0 1800.0 TIME (S)

Figure 4.1.11 Transient 1 Bypass Channel Flow Rate

3-D BROWNS FERRY REACTIVITIES VS. TIME 2000.0 m

y 1000.0 i - ---=1--- - - f <JGirAMD0?PL 'E w

~

$ 0.0 r

.. _ . . .._.._--n *~'$w W **-

D -1000.02

> ~

AL.LLLLhkLhk ukkhk

{_ , PITU ' ' " D Pl l 0

ps -3000.0-DM

-4000.0' , ,. ., ., .. ., .,

0.0 200.0 400.0 000.0 800LO 1000.01200.01400.01000LO TIME (S)

Figure 4.1.12 Transient 1 Reactivity Predictions

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

The combined effect of the total power governing parameters is shown in Figure 4.1.12, where different components of the core reactivity are plotted as a function of time. It is seen that the positive reactivity insertion due to the Doppler and moderator temperature effects (fuel and moderator tempera-ture are functions of system pressure) is balanced by the negative void reac-tivity feedback due to void increase (see the corresponding plot of the core average void history in Figure 4.1.7). Note the effect of the vessel depres-surization at 900 seconds on the three major reactivity components in Figure 4.1.12: the negative void reactivity increases (core voids grow due to spe-cific volume increase at lower pressure) and the positive Doppler and modera-tor temperature reactivities also increase (fuel and moderator temperatures drop). Core average and maximum fuel temperature histories are shown in Figures 4.1.13 and 4.1.14.

As previously discussed by Neymotin and Saha (October 1984), the axial power distribution varies significantly during the course of a BWR ATWS. To illustrate the extent of the variation, two sets of power and corresponding void distributions (void distribution being the major factor defining the power distribution) at time of 800 and 1250 seconds are shown in Figures 4.1.15 through 4.1.18. The results clearly demonstrate that the point kinet-ics model can not be used to perform ATWS calculations because of the strong spacial feedback effects. These strong feedback effects, particularly the void, dramatically alter the neutron flux and axial power shape which, in turn, feeds back to the thermal hydraulics. Thus, the use of at least 1-D neutronics is required for ATWS calculations. Figures 4.1.15 through 4.1.18 prove that the basic assumption of point kinetics, i.e., the separability of the flux into a shape and time function, is not valid. The inability of the point kinetics to account for the tightly coupled spacial relationship between the thermal hydraulic parameters and power distribution was one of the major i reasons for the earlier point kinetics prediction of -87 of rated power.

l The most significant parameter to monitor during the transient is the PSP (i.e., wet well) water temperature, which is directly related to the total reactor power and core steaming rate. The ability of PSP to condense the steam produced in the core determines whether or not the primary containment will rupture due to overpressurization of the dry well. The wet and dry wells are connected by 12 two-ft (0.61 m) diameter vacuum breakers which open when the wet well pressure exceeds the dry well pressure by more than 0.5 psi (.003 MPa). In Figure 4.1.19 the wet well, or PSP, water temperature is shown to have reached the assumed HPCI failure temperature at approximately 23 minutes.

(The dashed lines in Figure 4.1.19 show the results of the Manual Rod Inser-tion (MRI) calculation described in Section 4.2). The events occurring after the HPCI failure are discussed in an ORNL report (Harrington and Hodge, July 1984). In this report the overpressurization event has been identified to occur about 20 minutes after the HPCI failure. Therefore, the overpressuriza-l tion of the primary containment is predicted to be approximately 43 minutes af ter the beginning of the transient.

Finally, the effect of dropping the water level to TAF has been estab-lished to be about 19f. (averaged over time, Figure 4.1.9) with the resulting HPCI failure time of approximately 23 minutes. This means that other mitiga-l tive actions must be taken as required in the EPG. The effect of the SLCS and manual rod insertion must also be included in the future ATWS calculations.

l 3-D BROWNS FERRY CORE AVERAGE FUEL TEMPERATURE VS. TIME 800 m

U 700i -1292.0 e 8004 M

500f -932.0{

400i

- = ::::::::::: . :

300 ' ,. ..

0.0 200.0 400.0 800.0 800.0 1000.0 1900.0 1400.0 1800.0 TIME (S)

Figure 4.1.13 Transient I Core Average Fuel Temperature 4

, 3-D BROWNS FERRY MAXIMUM CORE FUEL TEMPERATURE VS. TIME 1400 t

u 1200i -2192.0 8

8 1000i N

l4 8005 -1472.0{

400 i - 752.0 soo ' , ,. .. .. ., , .,

0.0 200.0 400.0 000.0 800.0 1000.0 1200.0 1400.0 1800.0 TIME (S)

Figure 4.1.14 Transient 1 Maximum Core Fuel Temperature

3-D BROWNS FERRY POWER VS. CORE HEIGHT AT T = 800.0 S a

10 ~

2 y 1-5

, N M-m 0 $2 d.4 d.8 d8 1 FRACTIONAL CORE HEIGHT Figure 4.1.15 Transient 1 Relative Axial Power Distribution at 800s 3-D BROWNS FERRY VOID VS. CORE HEIGHT AT T = 800.0 S OS e

g 0.4 -

e 0 02-0.0 , , , ,

0 02 0.4 0.8 OS 1 FRACTIONAL CORE HEIGHT Figure 4.1.16 Transient 1 Axial Void Profile at 800s

3-D BROWNS FERRY POWER VS. CORE HEIGHT AT T = 1250.0 S L4 12 -

1-l =

02-g 02-i I

0.4 ' , , , ,

O 02 0.4 0.8 OS 1

{

FRACTIONAL CORE HEIGHT Figure 4.1.17 Transient 1 Relative Axial Power at 1250s 3-D BROWNS FERRY VOID VS. CORE HEIGHT AT T = 1250.0 S as h a6-E g 0.4 -

e 9 "2-0.0 , , , ,

0 02 0.4 0.8 G8 1 FRACTIONAL CORE HEIGHT i

j Figure 4.1.18 Transient i Void Profile at 1250s I

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

i.

A15tPEB MCI FAILUR1 l

ire 0.y . . ..

g,;d .o~...~, ,, .-

l iso ,s*,

c 5,no y

e '

y ,

sto

- u... . s

...u.... .:< o 10 . .

e 700 400 600 800 1000 1700 1400 1800 1808 fileg (3) t Figure 4.1.19 Comparison of PSP Water Temperature for Transient 1 and the Manual Rod Insertion Calculations 4

42 -

The purpose of the calculation presented in the next section wass to study the effect of manual rod insertion separately and to isolate its mitigative ability.

4.2 Manual Rod Insertion (Transient 2)

The scenario for this transient is identical to the level and pressure control transient discussed in Section 4.1 The only difference is that the operator is assumed to begin inserting control rods one-by-one at 150 seconds.

The control rod insertion strategy used during the calculation was based on information from Figure 4.2.1. This data was taken from video tapes given to BNL by Dr. Steve Hodge of ORNL. These video' tapes were made by ORNL at the Browns Ferry Simulator on August 20, 1983 under the auspices of the USNRC/SASA program. The actual insertion strategy used during the simulator session can be found in Figure 4.2.2. It should be noted that the control rod insertion strategy used from the simulator was chosen because it was a response repre-sentative of a trained plant operator. An alternative way would be to program RAMONA-38 to insert rods at maximum worth locations during this transient; however, the information gained would be of minimal value since no operator would have internal core power and neutron flux information when choosing which control rod to manually insert. Additionally, an optimum control rod insertion strategy during an ATWS does not exist for the general core, since the core condition is constantly changing, so that the optimum strategy would be substantially different for B0C, MOC and E0C. However, there are several general procedures that the operator would follow when facing the ATWS.

An operator would typically select the high worth control rods around the as possible. A control rod center of core and has its maximum worthinsert them between as symmetrically notches 30 and 8 (i.e., 4.5 f t and 10 ft from the bottom of the core, respectively). Thus, the operator would first choose a control rod between these notches to drive into the core and stop insertion af ter the 10 f t (or notch 8) mark. These attributes were also observed in the simulation session from which the control insertion strategy was taken.

The video also revealed that besides the MRI, the operators must perform many tasks during an ATWS event. These other duties (such as level control) sometimes take the operators away from the control panel where the switches controlling the rod insertion are located (the insertion process discontinues when the switches are not physically attended). Consequently, the process of insertion is constantly being interrupted, which results in slowing down the control rod insertion process. While the maximum speed of insertion is 3 in/s, the video revealed that the average speed of insertion was only 2 in/s.

Furthermore, it should be noted in Figure 4.2.2 that the control rods selected af ter 1229 seconds were chosen by the authors following the same strategy used by the operators at the simulator session, ideally, the MRI calculation should be performed with a full core model since it is well known that light water reactor cores are neutronically loose-ly coupled. Then the effect of each particular rod as it is being inserted would be a large undertaking. However, the operators are trained to perform symmetric insertions (with at least 1/4 symmetry) whenever manual rod inser-tion is required. This behavior was evident during the simulation.

r l

ROD INSERTION PATTERN AT SIMULATOR

! (TRANSIENT #7, AUGUST 20, 1983) l l

l

" I 7

=

s1 g l

I 7 a

f a 11 13 i 3 3 '15 6 17 1

rt 12 l-l-1"' . T--'

l 27 4 18  ! 16 5

' 3 '

10 9 14

" l h

! ts L 1: J '

i

! 11 l

j at i

! s I a ' L J i

i i

]

at as to 14 ts a 2s a m a e as a se as J

j

• NUMBER INDICATES ORDER OF INSERTION i

}

Figure 4.2.1 Control Rod Insertion Pattern Performed at TVA Simulator MSIV Closure ATWS Session No. 7 3

RAMONA-3B ROD INSERTION PATTERN I 1 m

I 3'

I I l I l a

a E D D E 3

j A F B F A l l

c ,_!_,..__._____..____.__;.__._

# I A F 6 F A

• D D E E  !

18 I

,  ! G

}; u l- 1: -_1 i

"  ! H i e _ l

.I '

! = ' Ui J i

! i

\

m os to 14 1s  ::: 2s a s4 a e as a s4 m R00 SEQUENCE TIME MOVEMENT A 150 to 269 s 5.3 to 10.4 f t j B 269 to 389 s 0 to 10.4 ft l C 389 to 509 s 4

0 509 to 749 s 1

E 749 to 989 s l

F 989 to 1229 s G 1229 to 1469 s. "

H 1469 to 1709 s a

! 1709 to 1949 s a 1

J Figure 4.2.2 Control Rod Insertion Pattern Used by RAMONA-38 4

1

i i

Furthermore, the TVA data did point to a reasonable quarter core symmetry (see Section 3.1). Thus, the quarter core symmetry option of -RAMONA-38 was chosen for this transient along with the corresponding cross section sets discussed in - Section 3.1. Moreover, it should be kept in mind when analyzing the re-sults, that in the present calculation four control rods were being driven in simultaneously at 1/4-th of normal speed, except for the boundary cells rods (see Figure 4.2.1). However, at the completion of each 4-group rod insertion, the total control rod worth inserted was corresponding to the actual case.

The RAMONA-3B results are presented in Figures 4.2.3 through 4.2.9. Much of the basic physical behavior discussed in Section 4.1 also applies here; 4

hence the purpose in this section will be to discuss features which are dis-tinctly related to the manual rod insertion case. As mentioned earlier, this transient has the same scenario as discussed in Section 4.1 with the manual rod insertion effect superimposed.

The relative power as calculated by RAMONA-38 can be seen in Figure i 4.2.3. Immediately after recirculation pump trip, the reactor power drops

, to approximately 30% of the rated power. The power is then slowly reduced to

! an average power of 14% at 240 seconds as a result of lowering the water level l to TAF (see Figure 4.2.6) and the inserted rods. The effect of the six con-trol rods which were inserted by 389 seconds did result in a lower power level as compared with the unrodded case (Transient 1) where the power level was about 20% at this time. As seen in Figure 4.2.8, the amount of negative reactivity introduced into the core by the scram (or control rods) was not significant by 389 seconds, but it was slowly increasing.

Between 600 and 900 seconds, the average relative power was approximately 12%. By 749 seconds a total of 12 rods had been completely driven into the core. The large amplitude oscillations of the void reactivity seen in Figure 4.2.8 are caused mainly by S/RV valves opening and closing. This cyclic be-havior can also be seen in Figure 4.2.5, where the system pressure is plotted.

Furthermore, this total core behavior is superimposed upon the local void power competition. Local void collapsing leads to local increase in power be-cause of the increase in neutron moderation. Eventually this results in a local average void increase, causing a drop in the local power because of the i lack of neutron moderation. This internal competition continues in all neigh-boring nodes, which accounts for all smaller " wiggles" seen in the core average void (Figure 4.2.4), void reactivity (Figure 4.2.8), etc. Implicit to void generation is its associated effect on the hydrostatic head across the core and its effect on core flow. In Figure 4.2.9 the higher oscillations in core flow correspond to S/RV action, while the smaller oscillations are a re-sult of the internal void / power struggle. The resulting core flow behavior is dictated by this internal phenomenon.

The reactor power is about 10% of rated between 900 and 1229 seconds and by 1229 seconds 20 control rods have been inserted. The PSP heatup rate, as seen in Figure 4.1.19, is slightly different from the case without rod inser-tion. The overall effect of the control rods on the core's feedback mechanism can be seen in Figure 4.2.8, where the rods have changed the worth of the void reactivity, as compared to the case without the rods (Figure 4.1.12). The scram (or control rod) reactivity is slowly becoming the main shutdown mechanism with the switchover point near 20 rods. Once this switchover takes  !

MANUAL ROD INSERTION WITH DEPRESSURIZATION RELATIVE POWER VS. TIME 1.0 OA-k na-m g OL4- -lii%

~12:

p / ~ 5%

0; 0^~ l

,b 'u"w

---u. am s. ,. u ,, , :7;;

0.0 , , , , , , ,

0.0 300.0 400.0 000.0 800.0 1000.0 1200.0 140Q0 1800LO TIME (S)

Figure 4.2.3 Relative Power for MRI Cal'ulation c MANUAL ROD INSERTION WITH DEPRESSURIZATION VOID FRACTION VS. TIME 0.80 Z OM-o 1 0.50 -

k 0.46 -

O

> 0.40 -

036 , , . -r , , ,

0.0 300.0 400.0 000.0 80u.0 1000.0 1200.0 1400.0 1800.0 TIME (S)

Figure 4.2.4 Core Average Void Fraction for MRI Calculation

i-MANUAL ROD INSERTION WITH DEPRESSURIZATION SYSTEM PRESSURE VS. TIME 8

m a-(

3 Ti N

-1160.3 si -870.2 E Si I

A 4- - 500.2 04 850.0 4dQ4 edno Odo010bE013bao 14 boa 18004

, TIME (S)

Figure 4.2.5 System Pressure for the MRI Calculation i

f I )fANUAL ROD INSERTION S

WITH DEPR m URIZATION WATER LEVEL IN DOWNCOMER VS. TIME 10 COLL,WAT. LEV.

~~~~ # ~~*

6-id j 0- Top 0F DOWNCOMEn - 0,0 I 3

~

I

_gn =_ - - - : -- _ :_ : : : _- __ _ __+_ e _ _d 15,12

-19,68

-10 , , , , , ,

04 300.0 400.0 000.0 aca0 100a0 1800.0 140E0 i

TIME (S)

Figure 4.2.6 Water Level Prediction for MRI Calculation

MANUAL ROD INSERTION WITH DEPRESSURIZATION CORE INLET SUBCOOLING VS. TIME 30 -54.0 6

20- 36.0 5 10 - llbfftllllfWf g -18.0 g N J m h.

0 - 0.0 OD 950.0 4 $ 0.0 8d O 4 000.0 10 BOLO 13b0414kMLO 1800LO TIME (S)

Figure 4.2.7 Core Inlet Subcooling for MRI Calculation MANUAL ROD INSERTION WITH DEPRESSURIZATION REACTIVITIES VS. TIME 2000.0 0 1000.0 - e. P ^^ ' " " --- ' - ' = - -

ER 0.0 Q*

f , .,_ _

E -'"" s glg $ 4# ,P Wy D -2000.0i y i l\

N -3000.0 i

-4000.0- , , , , , , ,

OD 200.0 400.0 600.0 800.0 1000.0 1200.0 1400.0 1600.0 TIME (S)

Figure 4.2.8 Reactivity Predictions for the HRI Calculation

MANUAL ROD INSERTION WITH DEPRESSURIZATION b FUEL BUNDLE INLET FLOW RATE VS. TIME y 1&O -33.07 8

N O

$ ag0- -22.05 r4 7 N

N k

4 50- ) -11.02 o

J:lJr'J:JJ2AJJ J d d } J,Y 0.0 , , , , , . - 0. 0 0.0 200.0 40060 800.0 800.0 1000.0 1900.0 14004 TIME (S) i Figure 4.2.9 Core Flow into the fuel Assemblies for the MRI Calculation 1

place, the void reactivity seeks a level which compensates for the left-over positive reactivities from the Doppler (i.e., fuel temperature effect) and moderator temperature feedbacks (Figure 4.2.8).

At 1228 seconds, the PSP temperature has reached a point (-160*F) where the operator had to depressurize the system to comply with the HCTL of the PSP system. From Figure 4.2.5, it can be seen that the blowdown continued to about 1390 seconds, when the pressure reached approximately 3.5 MPa (507.6 psia). This depressurization resulted in an increased (negative) void reac-tivity so that the reactor power dropped to about 5% of the rated power with 1.5% coming from the decay heat in the core. The depressurization rapidly increased the void (Figure 4.2.4), and so the loop hydrostatic head, causing higher flow through the core (see last 100 seconds in Figure 4.2.9).

In Figures 4.2.10 and 4.2.11 the changes in axial power distribution are plotted for several times. It is seen that the axial power profiles (i.e.,

the shape function) change not only significantly during the transient but also invert under certain events such as depressurization. It should be in-ferred from this plot that power, core flow, void distribution, and reactivity effects are interrelated so tightly that only spatial kinetics codes can be used to predict these types of transients.

The horizontal or planar power distribution at different times for the core is shown in Figures 4.2.12a through 4.2.129 . The effect of the control rods can be followed since each figure corresponds to the moment when a rod insertion was completed. An obvious result was that the center of the core has been essentially turned off (Figure 4.2.12f), leaving the outer ring to become the major source of power.

As it is seen in Figure 4.2.8, the depressurization was started before the scram reactivity became the dominant controlling mechanism. Hence, after the depressurization when the reactor becomes harder to control because of the increased steepness of the derivative da/dP, the reactor should have enough scram reactivity (i.e., 20 rods inserted at 1229 s) to prevent possibility of void / power oscillations. To separate the effects of MRI and depressurization, another calculation was performed without depressurization to isolate the control rods effect. Some of the results of this calculation are presented in Figures 4.2.13 through 4.2.15. The horizontal or planar power distributions beyond 1469 seconds are shown in Figures 4.2.16a through 4.2.16c.

j in Figure 4.2.13 it can be seen that with 28 rods completely inserted (by 1709 seconds) the power dropped down to 6% of the rated power because the scram reactivity (see Figure 4.2.15) became the dominant shutdown mechanism.

Hence, if the operator can postpone depressurization until at least 32 rods l are inserted, the core would probably have enough scram reactivity to reduce the chance of developing large power oscillations.

The RAMONA-3B calculations have shown that manual rod insertion during an ATWS increases the amount of time available before the PSP water temperature reaches the assumed HPCI failure temperature of 190*F (see Figure 4.1.19).

Also, af ter about 32 rods are inserted the feedback mechanism contributed from the scram reactivity should be the dominant negative reactivity feedback mechanism. However, the PSP water temperature is high (Figure 4.1.19) by the

i

! MANUAL ROD INSERTION POWER VS. CORE HEIGHT AT SEVERAL TIMES a

389 S

, ~~0 .

' ~

2

rol M p ---- -

1- '

N,N 02 0

/ ,

02 0.4 08 0A 1

FRACTIONAL CORE HEIGHT 1

i i Figure 4.2.10 Axial Power Distributions for 0, 269, and 389s for the MRI Calculation i

MANUAL RCD INSERTION POWER VS.' CORE HEIGitT A SEVERAL TIMES a

A 12 - f 2

! y 1-E M-l 1 0 . . , ,

0 02 0.4 OA OA- 1 FRACTIONAL CORE HEIGHT Figure 4.2.11 Axial Power Distribution for 509, 749, and 1229s of MRI Calculation i

i ,

e-, n, -e ~~-,e-- ,-s w w- em---w-,- ,ne---,,, -----,,,,,,---o-- -wm -_w,--w - - - - - - - - - - --aw,,-- - - - - - - + ,--

MINil28MTAL P0bER - 9difet4LIND TO 100 183 IIS 132 101 SR 103 131 105 98180 IM 79 72 SS %4 IN 138 112 110 m 110 102 123 11% 133 lit 120 102 90 44 130 117 125 98 IOS 100 117 100 132 128 130 15% 118 IIPs SI 105 IM 107102 98 125 18% 103 '.03 133 119 10% W 101 St W liO IN 801 100 185 132 SS 98 117 133 97 93 M S2 IIPe lit 102 123 112 lie 103 128 113 131 114 120 98 90 49 125 W ll2 110 127 103 187 105 IM liS 127 100 SS 73 38 103 IM IOS 502 98 IM lil 93 m lia 104 109 93 8%

SI 103 IN M M lis IM 93 SS 98 104 90 85 W llPo 93 I M 11% 17e 119 114 h 83 35 67 9%

li2 St 104 104 127 18% IM 90 m OS S4 W W SS 91 St IM 109 102 90 59 et 87 m 85 M 97 98 82 SS 5%

SS M 87 SS 95 G2 73 46 94 49 47 44 %3 %0 33 T=0 (A) ea]N t 20NT AL POb(R Ol5f AllhJf lON los 103109 74 66 SI 114 102 102 119 138 SS 78 97 "I i

107 21 90 6* S3 80 86 119 117 1%% 130 132 113 109 5%

126 103 102 SS 62 75 98 107, 139 133 ISS 129 13% 117 58 97 IIS 93 82 76 106 105 10% 110 147 13% 119 110 115 SS 99 tot 180 SS SS 103 IM 103 108 131 152 112 107 110 59 9e 98 SS lit 102 107 98 127 125 1%9 132 139 ll% 103 57 18% 07 100 102 120 97 ll% 112 141 133 147 I M 115 m 45 95 Ill 99 98 9p 127 IIS lol 106 13S 128 127 108 75 95 96 184 95 98 120 134 102 96 ill 120 112 73 88 95 87 lia 115 139 126 125 103 93 98 77 63 103 83 96 102 129 189 137 107 104 72 61 l

80 99 82 99 93 125 IIS til 97 6%  !

76 78 87 03 88 100 ' % 90 11 60 5 l

Bl el 82 IPe OS 94 77 %9 51 45 %S ** 43 41 3S Ju}gg (B)

Figure 4.2.12 Horizontal Power Distributions for t1RI Calculation

( A) T=0.0s, (D) T= 269s (C) T=389s, (D) T409s, (E) T=749s, (F) T=1229s, (G) T=1410s

HOR 120NTAs PobER Ol57Rl8UTl0N SI 68 h SS SS 74 IIS 503 105 124 145 OS SI 105 S4

%C 69 St %S %S 7% h 120 120 ISI 130 l%I 128 Ils SS

%d 63 h %S SS 71 90 109 145 140 ISS 13e l%S IN M 61 43 7% 80 70 104 IOS 107115 ISS 843 IM IM IN S3 as 32 5 79 h a02 129 IOS 112 130 104 120 187 128 95 7e E3 79 104 98 108 800 131 838 ISO 142 ISO IM 183 SI 102 79 93 90 IM 90 117 IIS 148 148 ISO 135 125 91 49 l Se 105 W 97 100 133 122 IOS 812 1%S 130 135 187 to et 93 113 SS 102 I M 146 800 102 lit 130 122 St SS 92 85 128 128 149 135 lb 110 100 104 03 87 102 SI 95 105 137 IM 150 118 lit 70 SS 79 37 SI 93 90 135 I M IM 308 GO 75 79 30 87 m 109 115 SS 77 0%

N 82 m 57 91 90 93 52 St %S 46 %S 45 49 37 T=389 (c)

MORIZONTAL P0hER DISTRIGUTION M 26 32 38 48 72 113 104 107 127 149 90 m 107 SS 27 44 41 37 48 72 m 128 122 ISS 148 145 th 128 80 40 SO SS 30 S2 00 98 110 148 143 170 143 ISO 138 8%

S4 73 es 63 87 102 107 100 liG ISO 147 138 IM IM 95 64 77 98 75 82 102 138 80s IIS 8%# les IM IM 12% GS 75 el 78 102 SS 107 101 133 t h 183 1%S 195 127 lit 63 100 79 92 90 120 90 lit 119 192 145 182 140 I M m SO 88 105 95 98 108 135 IM 100 185 149 t h 142 IM 03 82 9% Bl4 97 103 131 149 110 105 IN lh IM 93 87 % 07 IN 122 152 139 137 113 103 107 85 90 103 83 97 107 139 131 152 110 115 30 87 to 99 83 m 100137 IM IN 10e 70 77 to 90 00 95 110 lit SS 78 GS

.3 .3 .s W. af Si a 53 T=509 52 47 47 46 %S %S 30 (D)

Figure 4.2.12 NOR120 petal Poh4R OllIRlEUfl0N 87 18 M 31 43 70 11% 107 113 1 M ISO N Si lit 80 IS M M M 37 70 m IM IM 1m ISI ISS th 130 et 1 30 M 33 M 4S es 98 114 198 153 let 193 ist 148 es M as 77 % SS M 108 lit th 170 IS7141 133130 70  ;

3 El %I 98 71 M 13nt lil 128 ISI 100133 lM lb 72 Se Ss 98 m es 103 104 15 141 173 ISS 147 133 IM es 30 83 19 87 l13 90 125 lb 141 IM th ISI 139 101 5%

77 M #7 M 90135 IF7113 Ifl ISO le 193 lM M 75 8710e W 103133 th IIS Ill 130 143 l b to 83 90 55 128 t h ISS 14% 143 I M 109 115 M 74 lol el M 108147135 ISS lh 171 m 71 30 ee > 95 103 142 133 IM 18 3 h 7e el 98 90 SB ll4 128 103 M M m e5 37 91 95 W 87 SS T- M s 53 40 40 %7 40 44 M (E)

HORIZONT AL Pot 44 015fRl8Ut t0N 8 0 la il 33 SG 108 M 109 137 I M 101 W IN 8%

0 to il 18 M S3 70 11% IM 147 ISe IM l%S 141 M (

12 Il 14 le 29 43 72 M 140 IS% 190 10% l?S ISS h 3 19 18 le 29 h 45 W to lio 178 10% ISl 195 ISI 79 30 M 33 %l 47 %S M 90 183 193 I M 142 148 144 77

%9 S0 49 78 78 77 90 123 137 177 IM 179 190 I M 73 79 St 73 SB 103 87 lit IIS IM lel IN 10% ISf 109 SS 77 93 88 90 98 t h 177 113 IM les 195 let $ 30 M

,- iii M iOS iM i.3 iM ii. iM iM it. M 95 92 87 127 131 IM 154 194 lM lie 123 98 79 107 95 102 IIS IS% 147 173 I M 133 91 16 M M iO4 ii,iS7 i%. i4, in .i m 87 90 W 109 127 135 ll3 90 W 9 .i 9. iOO iO4 iO4 95 .0 T=1229s So w w St w So %

(F)

Figure 4.2.12 j l

l i

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

seAllopeN 80h4R Ol%taimetIcBe

1. 1 M l# M %8 13 tot 800 400 174 140 les 401 103 99 77 $4 30 34 47 79 90 llS let l*w 157 639 lie 107 M 95 51 M I? W 70 9e te ISO If? twe IM IM t06 96 49 99 90 W % 43 e4 es 100 43# IM 809 se to SS GJ S? > 40 Td 47 19 87 98 all IM 90 97 90 W M 94 90 603 9? ttPe 95 sto la? toe 95 118 93 el we IM 108 IN til IM 109 177 109 IN 100 107 800 Se 70 39 I M t*d 178 114 117 I M lit IOS 107 lit 100 108 se St f M 641 697 17% lit III ik 110 100 lb tel 99 &#

190 199 979 IS? 157 ISS I M 157 Ill 94 95 67 96

, is is,i.eisi.iiM i.in n, >- ei i,0 i.i . iii i n i.s i M 1 ,s iO .9 i ii i,i i 9 in n ,ii. ii. iO. 29 u I IT l IS I 31 IN 119 10's 90 60

., ,. ,, .. s. ,, T=1ti10s (G) i Figure 4.2.12 MANUAL. ROD INSERTION NO DEPRESSURIZATION RELATIVE POWER VS. TIME 0.4 k "-

n O J Q.

M 0.2 -

~

l 0.1 -

O.0 , , ,

0.0 600.0 1000.0 1500,0 2000.0 TIME (S)

Figure 4.2.13 Relative Power Prediction for MRI Calculation 56 -

i MANUAL ROD INSERTION '

NO DEPRESSURIZATION VOID FRACTION VS. TIME 0A0 y 026k o

1 g 0.46-e a40i 9 Om64 020 ' . . .

O.0 600.0 1000.0 1500.0 200a0 TIME (S)

Figure 4.2.14 Core Average Void Fraction for MRI Calculation MANUAL ROD INSERTION NO DEPRESSURIZATION REACTIVITIES V3. TIME 2000.0 O 1000.0i *.f^' ="" " '= '-= gg p g g a v oD:rs u n -

a gof m.~..~.~.~. T.~ -

-1000.0i D -2000.0i

- . n a Tff l _000ao.

-400a0' . . .

0.0 500.0 1000.0 1500.0 2000.0 TIME (S)

Figure 4.2.15 Reactivity Predictions for MRI Calculation

HORilottT Al. POWER olSTRISuf10N 9 9 13 20 38 55 95 97 106 135 166 99 94 120 61 9 11 18 le 25 50 65 110 119 162 153 162 143 136 65 tw 12 15 le M %0 65 92 137 148 176 15% 16e 1%3 64 23 2 20 30 34 38 %6 79 97 I%I 136 129 129 13% 66 36 31 37 %% %e %2 56 72 83 103 130 107 18% til 64 59 60 57 79 76 76 70 99 91 75 78 119 11% 105 54 96 7% 37 92 112 90 103 100 112 70 70 105 Ill 80 %3 96 11% 10% 106 800 144 I M 107 101 117 102 187 97 64 97 112 137 187 123 157 177 12e lli IIS 123 108 70 109 117 109 157 159 197 lie ISI 141 116 106 76 59 139 109 129 1%5 192 ISI 210 170 159 95 70 112 123 IIS 133 143 199 195 105 l52 9%

113 115 130 130 8%2 165 173 1%S 110 De t l

126 127 130 136 148 139 123 76 78 71 70 69 69 66 55 (A)

= * : .*0N' A POWER olSTRieutl0N

.2 :2 IS M 45 88 I48 148 155 199 6 7 151 1%5 lb 9'

3 15 25 36 72 f. 6 ISO 173 237 N7 M5 219 218 10*

t  !*

9 22 37 55 92 IM 195 20% 2*e N9 255 2M 107

.N 23 N 33 30 *1 60 103 133 198 196 190 193 210 103 l' 12 34 ** 48 46 65 87 106 139 tel 153 l6t 187 98 6' 62 57 % 70 74 72 104 107 97 90 170 169 159 87 95 '5 83 30 93 76 90 9e 115 79 98 141 156 11% 63 W l0* 92 IPe 79 90 SS 83 99 I.7 11% 1%% 16 92 W l03 II6 86 77 > 93 90 33 tot 125 120 82 0' .:= 95 lli SS 6? 60 93 90 99 96 79 65

38 :09 II* 102 105 68 66 81 90 62 55

N

0* 97 95 99 87 93 82 5*

0 13 1;9 90 91 9 92 77 60 =9

9

5 i:0 101 93 93 69 *2 63 59 5, .- ,, 3, T=1709s (B)

Figure 4.2.16 Horizontal Power Distributions for the Calculation Without Depressurization (A) T=1469s. (B) T=1709s. (C) T=1816s

l l

l l

l l

l l

l HOR 120NT AL PohER DISTRIOUTION 83 12 17 25 30 64 10% 100 104 126 160 117 119 148 75 12 l* 15 23 32 63 77 IIS 186 135 135 478 164 163 79 19 16 20 23 36 51 79 95 130 lit 150 159 198 169 82 30 29 26 37 40 %6 S2 el 96 137 137 140 147 159 79 49 41 97 S3 M %9 62 76 07 109 140 til 135 4%S 77 90 83 7S 93 m 06 77 80% 96 m 82 139 137 12e 70 129 99 110 103 114 90 102 97 182 76 99 122 132 96 S3 125 1%S 122 109 90 117 808 98 93 116 109 131 118 08 125 141 156 113 9B 103 109 92 Si tra 123184 77 146 IM 129 1%e llS > 73 Ill 10% ". 99 79 63 190 l*e IM 137 139 79 85 100 109 72 60 IS* a7G 1%% 132 185 130 112 819 102 66 IM l% 165 135 12% 125 120 100 76 62 in i6. i60 i%6 i 32 ii6 *

• T=1816s 50% 9e e5 w 66 57 **

(C)

Figure 4.2.16 l

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

E

~ ~

end of the. ATWS with the manual rod insertion, and hot shutdown is -not yet accomplished.

4.3 High Pressure Boil-Off (Transient 3)

The high pressure boi,1-off. calculation was started from the same ' steady state condition as -the base calculation (Transient 1) discussed in ~Section - -

4.1. . An MSIV closure . ATWS scenario similar to that of Transient I he: been chosen: MSIVp close in 5 seconds, control rods fail to insert on SCRAM'sig-nal, feedwater is lost in 8 seconds, and the recirculation , pumps .are tripped on the high vessel pressure signal. The specificity of the high pressure boil-off transient stems from the assumption that the HPCI system (the primary source of safety injection with 320 kg/s (5000 gpm)) is assumed to be.unavail-

) able. So only the RCIC system which provides approximately 39 kg/s_ (600 gpm)-

of water is available. . Another source: of the vessel inventory make-up water, i namely, the Control Rod Drive (CRD) cooling water. with the flow rate of about '

+

6 kg/s (13.2 lbm/s) is also available. The reactor was left in automatic mode j with the vessel pressure oscillating between the safety-relief ' valve pressure set points. No operator actions were assumed to be undertaken.

1 This transient was modeled with 1-D ' neutron kinetics option 'of RAMONA .

38. It .has been .shown in Section 3.1.2 and Appendix D' that 1-D axial core

4. model is adequate for transients where a planar symmetry is maintained. In i the present transient, both recirculation pumps trip, while the SCRAM fails' 1

and the SLCS and MRI are not activated. Thus, it can be assumed that the

! initial planar symmetry in the core will be- maintained throughout -the transient.

The results of the RAMONA-38 calculation for this transient are shown in-

! Figures 4.3.1 through 4.3.9. Figure 4.3.1 shows that the downcomer water i level was dropping very rapidly reaching the safety injection systems set ,

point level, i.e., Level 2. at approximately 15 seconds.- The TAF elevation

was crossed about 58 seconds later. The water level' continued to drop since i

the combined RCIC and CRD water- injection ' was not sufficient to make up the vessel water inventory being depleted by. steam losses through the steam line-safety and relief valves. The corresponding water injection and steam line.

flow rate histories are shown in Figure 4.3.2. The rate of downcomer water i

level drop is proportional to the difference between these curves.

4 After an initial spike in the vessel pressure and reactor power due to i the MSIV closure (Figures 4.3.3 and 4.3.4), .the core flow ' rate' was - continu-ously decreasing (Figure 4.3.5) following the water level drop in the down-comer. (After the recirculation pumps trip, the core flow rate becomes a function of the downcomer static head.) As a' consequence, the core, voids started to _ increase, causing a strong negative reactivity insertion.' .(The average core in-bundle void fraction histories are shown in Figure 4.3.6.).

Again, the histories of. the in-bundle void fraction, core void reactivity (Figure 4.3.7), and power (Figure 4.3.4) demonstrate the tight . coupling between the thermal hydraulics and reactor neutronics typical for " the BWR core.

As previously discussed in Section 4.1, the. core inlet subcooling is an r important parameter contributing both to the axial core power distribution and 1

- , , . - - . _ - . . . . , ~ , . .,c, ,,.w, , _ , _ . ___,._..--._..w..., . . , _ _ _ , , - , . . , . . ._-e,.

HIGH PRESSURE BOIL-OFF WATER LEVEL IN DOWNCOMER VS. TIME a

COLL.WAT. LEV.

Q v 0-WAT: LEV ,,,,,,,

-0 3

M -

M

. t

-a- -

-6.6}

s 4- g -

-13.1 $

.-. . - - - . g

-e ,, ., , ..

, .,. , -19.7 OD 20.0 40.0 00.0 80.0 100.0 120.0 140.0 160.0 TIME (S)

Figure 4.3.1 Water Level Prediction for the High Pressure Boil Off Calculation HIGH PRESSURE BOIL OFF WATER INJECTION, STEAM OUT & ECCS CONDENSATIO:

a00a0 -4409.2

+ COY)ENSATION

(

e 1500.0-

i ~ WATER IN M T m i M__ O U T -3307.0 g . ,

1000.Ok -2204.6 q -

j

% 500.0- -1102.3 d M

4 k

0.0 - - -- -

0.0

'El

-600D , , , , ., ., ,.

OD 20.0 40.0 00.0 80.0 100D 120.0 140.0 10 0.0 TIME (S)

Figure 4.3.2 Water Injection and its Associated Amount of Condensation Along with the Existing Steam Flowrate

HIGH PRESSURE BOIL OFF SYSTEM PRESSURE VS. TIME 9 -1305.3 Q

0 85- -1232.8 M

a- -1160.3 N 5 vs- -1087.8 k 7f -1015.3 04 SIno MLo dLo eo' D 1dGO 1$0A 1I00 1860 TIME (S) r Figure 4.3.3 System Pressure for the High Pressure Boil Off Calculation HIGH PRESSURE BOIL OFF RELATIVE POWER VS. TIME a

2 a-m h 1-e m

~

W no alto MLo e'o .o e'n o Ida0 1$0a 1400 tono TIME (S)

Figure 4.3.4 Relative Power for the High Pressure Boil Off Calculation i

HIGH PRESSURE BOIL OFF b FUEL BUNDLE INLET FLOW RATE VS. TIME p 10.0 , -22.0

)

g an-U .

g a0- -13.2 g k= 5 w.

k 105 - 4.41 a0 . ., ., , .,

n/L, . , ,

- 0.0 0.0 30.0 40.0 00.0 80.0 100.0 130.0 140.0 180.0 TIME (S)

Figure 4.3.5 Core Inlet Flow Rate to the Fuei Assemblies for the High Pressure Boil Off Calculations HIGH PRESSURE BOIL OFF CORE VOID FRACTION VS. TIME 0.6 -

- Jh ~~

o 0.4 -

{

k B 02-B 0.0 a'0.0 4'0.0 o'O.'O s'O.'O ido.o i k 0 1U.0 100.0 TIME (S)

Figure 4.3.6 Core Average Void Fraction for the High Pressure Boil Off Calculation

HIGH PRESSURE BOIL OFF REACTIVITIES VS. TIME 2000.0-

, SCRAM 1000.0 e ,- DOFFLw--ron m r

+

0.0 h/- ___ _ .-_~_

- VOID--

s.: -1000.0-~

b

-9000.0-

-300a0 . , . . . ..- . l 0.0 20.0 40.0 80.0 80.0 100.0 180.0 . 140.0 100.0 TIME (S)

Figure 4.3.7 Components of the Total Reactivity for the High Pressure Boil Off Calculation HIGH PRESSURE BOIL OFF CORE INLET SUBCOOLING VS. TIME SO '

-54 O

90- -36 v .

M M 10 -

i

-18 _

b Q.

O_ -

W N 0,0

-10 , , , , , , ,.

0.0 30.0 40.0 80.0- SEO 1040 130.0 1400 18 0.0 TIME (S)

Figure 4.3.8 Core Inlet Subcooling for the High Pressure Boil Off Calculation

HIGH PRESSURE BOIL OFF CONDENSATION ON ECCS C

U SD- -13.2 Z 4.0 - - 8.82 $

l, O -

M P.0 - - 4.41 b

Q 2;

g 04 0.0 20.0 40.0 80.0 80.0 100.0 130.0 1440 18 & O 0,0 TIME (S)

I i

Figure 4.3.9. Condensation on Injected ECC Water for the High Pressure ~

Boil Off Calculation l

l

the total _ reactor power. Unlike the previously discussed transients when the HPCI injection flow was available, the safety injection flow rate for this calculation is . very small,. which results in a low core inlet subcooling (Figure 4.3.8). The low injection flow rate is also accompanied by low con-densation rate after the downcomer level drops below the feedwater spargers, as seen in Figure 4.3.9. Incidentally, this figure also shows that the feed-water sparger level was crossed at approximately 39 seconds.

The calculation shows that during the first 150 seconds, the Doppler (fuel temperature), void (core flow rate, vapor generation rate), and moder-ator temperature (core inlet temperature and reactor vessel pressure) reactiv-ity effects (Figure 4.3.7) brought the total reactor power down to approx-imately 4.1% of the rated power (Figure 4.3.4). It is worthwhile to mention that throughout the transient the Doppler and void reactivity effects were the principal contributors to the total ccre reactivity.

An important result is that by 150 seconds only about 1.2% in the pre-dicted 4.1% total power was produced by fission in the core, while the rest was the decay power. The results indicate that the reactor power was almost brought down to the decay heat level due to the large amount of negative reactivity caused by the core voiding after the water level in the downcomer dropped 1.2 meters (4 ft) below TAF. The core, however, could regain power in response to any positive reactivity insertion, such as a drop in void frac-tion.

i Gradually, the reactor should arrive at a quasi-steady state condition; the power will oscillate around the value corresponding to the power needed to evaporate the 45 kg/s (99.2 lbm/s) of combined RCIC and uRD injected water which represents 2.7% of rated power. The quasi-steady state would be estab-lished in the following manner: any decrease in power leads to a lower vapor generation and, consequently, to a decrease in the core voids. At the same time, build-up of the downcomer water level and natural circulation driving head begins, and the core flow rate increases. As a result, power starts to increase and, eventually, the core vapor generation exceeds the RCIC and CR0 water flow rates so that the downcomer water level, the driving head, and the core flow rate start to decrease. Consequently, the void fraction in the core increases, dropping the total reactor power back to the water injection con-trolled power level (-2.7%). It is clear, however, that regardless of the water injection flow rate, the power cannot drop below the decay heat level.

Another issue associated with the present transient scenario,_ besides the '

total reactor power, was the thermal hydraulic behavior of the _ core at low flow conditions. The question was whether the fuel and cladding temperatures would reach a point where the fuel rod damage could occur. In the RAMONA-3B calculation, no CHF condition in the core was predicted up to 150 seconds when the calculation was terminated because of intermittent full-core flow re-versal. At this point a stand-alone sensitivity study based on the results obtained in the RAMONA-3B calculation was initiated in order to address the issue of the fuel rod temperatures.

The following conservative assumptions have been made for the simplified core thermal hydraulics model:

i

e there is a level separation in the core; e the core mixture level resides at the elevation where the equilibrium flow quality equals to one; e below the water level there exists a film boiling flow heat transfer regime; e only pure superheated steam cooling takes place in the core above the water level; e radiation heat transfer is negligible.

The stand-alone steady state calculation was performed based on the above assumotions and using the decay heat power distribution predicted by the RAMONA-3B code with a core flow rate of 45 kg/s (i.e., 99.2 lbm/s from the RCIC and CRD). The post-CHF heat transfer coefficient for the superheated steam cooling above the water level was calculated according to the correla-tion suggested by Anklam (1981).

The results obtained for the core power of 2.7% of rated reactor power and flow rate of 45 kg/s (99.2 lbm/s) showed that the maximum temperature at the fuel rod center was 996 C (1825'F), which is approximately 10% above the corresponding temperature at the reactor nominal power condition. However, the cladding surface temperature was predicted to be around 980*C (1796'F).

Note that this is the lower end of the temperature range where the exothermic steam-zirconium reaction should be taken into account while predicting the cladding temperature and the hydrogen generation (Camp et al.,1983). There-fore, the temperature of 980 C (1796'F) is high enough to cause concern, although it may not immediately lead to fuel rod failure.

In sumary, the following conclusions are drawn from the results pre-sented in this section:

1. Total reactor power drops to 4.1% of rated power by 150 seconds (1.2%

fission and 2.9% decay power) with the downcomer water level at about 1.2 meters (4 ft) below TAF.

2. No CHF condition is expected up to 150 seconds into the transient time. Stand-alone long term estimate does not show fuel and cladding temper-atures high enough for immediate fuel rod failure. However, the cladding tem-perature is high enough to cause concern about the fuel rod integrity because the steam-zirconium reaction could begin.

4.4 Recirculation Pump Trip Failure (Transient 4)

The final calculation to be discussed is a recirculation pump trip fail-ure transient. Although it is recognized that the probability of a recircula-tion pump trip (RPT) failure is very remote, it was decided to analyze this transient in order to determine the peak system pressure in the event of an ATWS in conjunction with the RPT failure.

The assumed sequence of events for this transient is the same as dis-cussed in Section 4.1, except that the recirculation pumps fail to trip on all possible trip signals (i.e., high pressure, low water level, operator action 1

)

l

during ATWS, etc.). For completeness, the sequence is briefly repeated here:

MSIV closure in 5 seconds, failure to SCRAM, loss of feedwater flow in 8 sec-onds, recirculation pumps fail to trip, and HPCI/RCIC injection starts when water level reaches Lo-Lo level (or Level 2).

The core model used for this calculation had seven hydraulic channels and ten axial levels with 101 fuel assemblies represented (i.e., 1/8 core model).

The results of the calculation are presented in Figures 4.4.1 through 4.4.5. In Figure 4.4.1, the core thermal power rises up to about 260% of the initial power during the first few seconds, and reaches a power level of 80%

at 40 s. The vapor generation rate corresponding to this power is below the maximum rated SRV capacity of 85% of full power steam flow rate. This elimi-nates the possibility of an over-pressurization accident leading to vessel failure. Figure 4.4.2 indicates that the reacter vessel pressure reaches a peak of about 1340 psia (i.e., 9.24 MPa) and plateaus out to about 1313 psia (9.05 MPa), which is well below any brittle fracture stress point for the reactor pressure vessel. However, the steaming rate is so high that the ves-sel water inventory is rapidly depleting (see Figure 4.4.3). Of course, the power will drop as the downcomer water level reaches the top of the jet pumps.

It should be noted that the maximum system pressure for this transient did not achieve a pressure high enough to challenge the vessel's integrity.

The response of the core is slightly different from the other calcula-tions presented in this report. In Figure 4.4.4 the void reactivity caused the initial increase in power as a result of the core void collapse. This void collapse occurred due to the high pressure that resulted from the MSIV closure. However, after 20 seconds, the power is determined by the internal competition between the positive reactivity from the Doppler effect (i .e. ,

changes in fuel temperature) and the negative reactivity effect from the moderator temperature feedback.

Consequently, the effect of the recirculation pump trip failure on the events and timing in an ATWS is very severe. The predicted high power results in a fast depletion of the reactor pressure vessel water inventory. This can eventually lead to fuel damage which would occur at the decay heat level. The calculation also showed that the concern about a possibility of a reactor ves-sel over-pressurization accident was not justified.

l l

l RELATIVE POWER .VS. TIME 3

(it: 25-o . .

A 2-r4 h

m 1 p

0.5 ' . . . . .

0.0 10.0 20.0 30.0 40.0 60.0 60.0 TIME (S) l Figure 4.4.1 Relative Power for RPT Failure Calculation SYSTEM PRESSURE VS. TIME

'- 1377 g 95 -

1305 L

bU Q T4 .

E M 8- -

b 1160E~!

v3 .

m m

~

A 7i -

1015 8.5 ' , , , , , 9142 0.0 10.0 20.0 30.0 40.0 50.0 60.0 TIME (S)

Figure 4.4.2 System Pressure for RPT Failure Calculation

WATER LEVEL IN DOWNCOMER VS. TIME 10 COLL.WAT. LEV.

g - W.A..L,L,E,Y,_ . , , _ , , ,

a ^

N

> !S 0-

] TOP 0F DOWNC0f1ER

~

0 d

.k  : TAF

~

gi

-16.4y

-10 . . . . .

0.0 10.0 20.0 30.0 40.0 50.0 60.0 TIME (S)

Figure 4.4.3 Water Level Prediction for RPT Failure Calculation REACTIVITIES VS. TIME 100a0 m SCRA:i g 50no-DO 'E w g P, E ,

y _,,-- _-

k --- . _,' ,

23 /

g --

4

-100a0 , , , , ,

04 10.0 30.0 30.0 40.0 60.0 00.0 TIME (S)

Figure 4.4.4 Components of the Total Reactivity for the RPT Failure Calculation 4

l VOID FRACTION VS. TIME

z; a44-O E

0.43-4 N 0.40- f p

5

> 033-038' M M no 10.0 aICL0 MLO SED TIME (S) i Figure 4.4.5 Core Average Void Fraction for RPT Failure Calculation

5. CONCLUSIONS AND RECOMMENDATIONS The knowledge gained from using RAMONA-3B with its spatial neutron kinetics and two-phase thermal hydraulics is summarized below.

1. The assumption that the one-speed, time-dependent, continuous energy diffusion equation can be solved by substituting in a relationship that sepa-rates flux into a " shape function" and " amplitude function" does not apply to the BWR core. The feedback mechanisms of the BWR core continuously change the power shape, which completely violates the assumptions used in the point kine-tic model's derivation. Axial pcwer shapes predicted at different times dur-ing the transient have indicated that the void feedback effect can instanta-neously alter the axial power shape. It has also been observed that the axial power shape can completely invert as a response to vessel depressurization or safety / relief valve (SRV) actions. Furthermore, the coupling between the thermal hydraulics and nodal power is so inter-related, it has been determined from this study that a 1-D neutronics formulation is the minimum requirement for reliable prediction of ATWS. It should be noted that a correct prediction of the axial core power distribution is especially important during the natural circulation mode of reactor operation due to the tight coupling be-tween the core flow rate, core void distribution (which determines the natural circulation driving head and, therefore, the flow rate), and the core power di st ribution. The inability of the point kinetics model to simulate this be-havior is probably why the earlier calculations predicted powers of -8%,

while the spatial codes predicted -19% of rated power.

2. Level control (Cor,tingency #7 of EPG) and pressure control reduce the reactor power from 30 to 14% during the course of the transient. High Pres-sure Coolant Injection (HPCI) failure occurs in -23 minutes. The failure is assumed to occur by definition when the bulk PSP temperature reaches 190 F.

No Manual Rod Insertion (MRI) or Stand-by Liquid Control ' System (SLCS) was modeled during the transient.

3. Level and pressure control along with MRI will delay the time to HPCI failure. Although the reactor power can be reduced to 6% with -20 rods in-serted after the system depressurization, power spikes are expected because the void reactivity is comparable to the scram (or control rod) reactivity and both are smaller than the positive reactivity supplied from the Doppler and moderator feedbacks. By delaying the depressurization until the time when

-32 rods have been inserted into the core, a reactor state would be reached where the negative scram reactivity would become the dominant feedback effect.

The scram reactivity should be large enough to lower the reactor power to the i decay heat level. However, the pressure suppression pool (PSP) water tempera-l ture is high and the HPCI failure will eventually occur unless the RHR cooling i

is activated. No SLCS was modeled during this transient.

4 The high pressure boiloff calculation (i.e., assumed HPCI failure) predicted that the power would drop to ~4% at 150 seconds with the water level -4 feet below TAF. No Critical Heat Flux (CHF) condition was detected during the first 150 seconds. The combination of the Reactor Core Isolation Cooling (RCIC) and Control Rod Drive (CRD) flows is enough to sustain ~2.7%

of rated power without loss of liquid inventory in the vessel. Thus, if the core thermal power remains below 2.7%, no fuel damage would be expected.

5. If the recirculation pumps do not trip during an MSIV closure ATWS, the predicted core power after 40 s is around 80% of rated power so that the steam flow rate is below the SRV maximum capacity which is 85% of full power steam flow rate. Of course, the power will drop when the jet pumps become un-covered. A peak pressure of 1340 psia was calculated. While the calculation showed that the reactor vessel was safe from an overpressurization failure, the water level was dropping rapidly. This calculation confirms the impor-tance of verifying the recirculation pump trip as the first operator action during an ATWS event.

6. The part of the EPG related to the ATWS issue must be re-evaluated with the knowledge that the reactor power would be -19% of rated power with the water level at TAF, rather than 8% as first determined by the previous point kinetics calculations, performed elsewhere.

_o

6. FUTURE WORK This report considers several aspects of the ATWS issue. The effect of lowering the water level to TAF and the effect of manual rod insertion have been studied, along with the effect of the system pressure reduction. All of these methods utilize either the control rods or the void feedback mechanisms to introduce negative reactivity into the core in an attempt to bring the power down to a level which would reduce the probability of a primary contain-ment rupture followed by a fuel mishap. The results of these calculations should provide better understanding of mitigative effects of different operator actions during ATWS, so that eventually adequate procedures will be developed to ensure the BWR plant safety. However, one very important issue remains to be resolved, and that is the effect of the SLCS system on the ATWS. It is recommended that best estimate calculations be performed for verification of the effectiveness of the SLCS. Best estimate calculations verifying the assumed mitigative response of the SLCS system must be performed before the ATWS issue can be considered technically understood.

Finally, it is recommended that the NRC sponsor work to independently develop procedures to use during an ATWS using their state-of-the-art codes with spatial neutron kinetics.

l l

REFERENCES

'A. Ahlin, _ M. Edenius and H. Haggblom, "CASMO: A Fuel __ Assembly' Burnup_

Program," Studsvik Report- AE-RF-4158, June 1978.

T. M. Anklam, "0RNL Small-Break LOCA Heat Transfer Tests Series I: Rod Bundle Heat Transfer Analysis," 0RNL/NUREG/TM-445, August 1981.

A. Aronson et al., " Status Report on BNL Calculations of ATWS in BWRs," BNL-17608, RP1022, 1973.

A. , L. Camp et al . , "Li ght Water Reactor Hydrogen Manual," NUREG/CR-2726, l SAND 82-1137, June 1983.

I B.' Chexal and W. Layman, " Reducing BWR Power by Water Level Control During an ATWS: A Quasi-Static Analysis," NSAC-69, May 1984a.

B. Chexal et al., " Reducing BWR Power by Water Level Control During an ATWS:

( A Transient Analysis," NSAC-70, August 1984b.

R. J. Dallman et al., " Severe Accident Sequence Analysis Program - Anticipated Transient Without Scram Simulations for Browns Ferry Nuclear Plant Unit 1,"

NUREG/CR-4165, February 1985.

General Electric Co., " Assessment of BWR Mitigation of ATWS, Volume II, <

(NUREG-0460 Alternate No. 3)," NED0-24222, February 1981.

I i General Electric Co., Prepublication Draft, Emergency Procedure Guidelines, j BWR 1 through 6, Revision 3, December 1982.

1 .

i R. M. Harrington and S. A. Hodge, "ATWS at Browns Ferry Unit One," NUREG/CR-3470, ORNL/TM-8902, July 1984 j T. M. Harrington, " Evaluation of Operator Action Strategies for Mitigation of MSIV Closure Initiated ATWS," ORNL letter report, November 1985.

l J. M. Healzer and D. Abdollahian, "NATBWR - A Steady State Model for Natural Circulation in Boiling Water Reactors," EPRI NP-2856-CCM, February 1983.

C. J. Hsu and D. J. Diamond, "The Effect of Downcomer Water Level and Vessel-Pressure on Boiling Water Reactors During an ATWS," Brookhaven National Labor-atory,1985, to be published.

J W. C. Jouse, "INEL BWR Severe Accident ATWS Study," Eleventh Water Reactor Safety Research Information Meeting, Gaithersburg, Md., October 1983

! W. C. Jouse, " Sequence Matrix for the Analysis of an ATWS in a BWR/4; Phenom-j ena, Systems and Operations of Browns Ferry Nuclear Plant Unit 1," Draft, May

1984.

! Kendrick and Fisher, " Advanced Recycle Methodology Program," EPRI RP-118-1, CCM-3, Electric Power Research Inst., January 1976 1

l

i

T. A. Keys (TVA Staff Nuclear Engineer), letter dated March 6,1984;

Subject:

Browns Ferry 3,- Cycle 5 (BF3, CYS) Nodal Data, sent to John Carew, Brookhaven National- Laboratory.

S. E. Mays et al., " Interim Reliability Evaluation Program: Analysis of Browns Ferry Unit 1 Nuclear Plant," NUREG/CR-2802, EGG-2199, July 1982.

J. H. McFadden et al., RETRAN-02, A Program for Transient Thermal-Hydraulics Analysis of Complex Fluid Flow Systems," NP-1850-CCM, Electric Power Research Inst., May 1981.

L. Neymotin and P. Saha, "A Typical BWR/4 MSIV Closure ATWS Using RAMONA-3B with Space-Time Neutron Kinetics," Paper No. A3, Proc. International Nuclear Power Plant Thermal Hydraulics and Operations Topical Meeting, Taipei, Taiwan, October 22-24, 1984. J. Chao and C. Chiu, eds. (ANS).

C. Peterson et al., "A Transient Analysis of Power Reduction During BWR ATWS Using Transient Codes," EPRI report in publication.

V. H. Ransom et al., "RELAP5/M001 Code Manual Volume 1: System Models and Numerical Methods, "NUREG/CR-1826, Draft, EG&G, September 1981.

0.D. Taylor, et al ., " TRAC-BD1: An Advanced Best Estimate Computer Program for Boiling Water' Reactor loss-of-Cbolant Accident Analysis," NUREG/CR-2178, November 1983.

W. Wulff, H. S. Cheng et al., "A Description and Assessment of RAMONA-3B: A Computer Code with Three-Dimensional Neutron Kinetics for BWR Systems Tran-sients," NUREG/CR-3664, BNL-NUREG-51746, January 1984 l

l l

APPENDIX A RAMONA-3B Description RAMONA-3B (Wulff et al.,1984) has been developed at Brookhaven National Laboratory under the sponsorship of the U.S. Nuclear Regulatory Commission for analyzing BWR systems transients. RAMONA-3B is the only currently available best-estimate BWR system transient code designed to predict three-dimensional l power in the core coupled with the fuel and cladding temperature and vessel thermal hydraulics phenomena. The code employs the 1-1/2 group method with either three-dimensional or one-dimensional diffusion neutron kinetics, coupled with a one-dimensional, non-equilibrium, non-homogeneous two-phase model for thermal hydraulics.

A.1 RAMONA-3B General Features The RAMONA-3B capabilities cover a variety of BWR normal and abnormal system and accident transients including those with full, partial or no scram, control rod drop accidents (CRDA), main steam isolation valve closure, turbine or recirculation pump trip, transients induced by change in feedwater condi-tions, failure of pressure regulator transients, etc. (See Table A.1). The code can also simulate a BWR reactor at different steady-state conditions:

zero power, hot standby, and full power. RAMONA-3B has been recently inter-faced and tested with a balance of plant (B0P) code MINET and validation tests have been successfully performed (Van Tuyle et al.,1985). Addition of the B0P calculation capability to the RAMONA-3B code will allow simulation of entire BWR plant transients.

RAMONA-3B has a wide variety of the models for simulating the plant con-trol and protection systems. Major among them are:

a. feed-water controller,

b. reactor vessel pressure controller,

c. High Pressure Coolant Injection (HPCI) and Reactor Core Isolation Cooling (RCIC) systems, ci , steam line by-pass and turbine stop valves, and safety and relief valves (S/RV),

e. reactor scram, main steam isolation valve (MSIV) closure, and recirculation pump trips on different signals specific for the BWR being modeled,

f. Standby Liquid Control System (SLCS) for boron injection.

Table A.1 RAMONA-3B BWR Transients Matrix r Control Rod Drop Accident Full or Partial Scran Transients with f' Manual Rod Insertion Reactivity Changes Boron Injection Moderator Temperature  !

and/or Density Change  !

f Turbine Stop Valve Closure I

(Generator Load Rejection, r

Transients with Turbine Trip, etc.)

Pressure Changes i MSIV Closure j

Failure of Pressure Regulator

( Faulty Operation of Safety / Relief Valves

' Changes in Feed-Water Mass Flow

[InadvertentActuationofSafety Transients with Coolant Inventory and/or Mass l Injection System Flow Rate Changes l

( Recirculation Pumps Trip Transients with Coolant Temperature Changes Loss of Feed-Water Heater

A.2 RAMONA-3B Models and Solution Methods Neutron kinetics and thermal hydraulics are the two major parts of the code which are linked together by the heat conduction in the fuel rods. The coupling and interaction among the thermal hydraulics, neutronics, and heat conduction as employed in RAMONA-3B is illustrated in Figure A.1 As seen, the feedwater, HPCI, RCIC and SLCS are passive boundary conditions for the vessel thermal hydraulics, whereas the steam line communicates with the vessel in a two-way manner. The vessel thermal hydraulics calculation provides infor-I mation required by the fuel cross sections model (distribution of the void fraction, moderator temperature, and boron concentration). Remaining parame-ter for the cross sections calculation - the fuel temperature distribution -

is supplied by the heat conduction model. The neutron kinetics calculation ends with the prediction of the core fission rate distribution and, conse-quently, the thermal energy which is transferred into the core coolant in two ways: through heat conduction in the fuel rods (fuel pellet, gas gap and cladding) and as heat deposited directly into the coolant.

A.2.1 Neutron Kinetics The RAMONA-3B code builds the core region out of a number of vertical stacks of neutronic cells (i.e., neutronic channels). Each neutronic channel represents one or several actual fuel bundles. Each neutronic cell is assign-ed a cross sections set depending on the fuel cell history prior to the begin-ning of the transient (fuel composition, burn-up, and void history).

The neution kinetics model of RAMONA-3B starts from i;he following two-group, three-dimensional, time-dependent diffusion equations:

Fast Neutrons:

I 1

3*1 (A-1)

= v . Di v&1-E21*1 - Eal*1*(1~8)["1f1*1+"2f2*2)+i=1Ac$

E E vy at 1 i

Thermal Neutrons:

1 - 3*2 (A-2) v2 at = v . D 27*2 + E21*1 - Ea2*2 l

j

Control Rods Cross Sections .

Neutron Kinetics Direct Power a o a n Heating Deposition Fuel ,,

Temp.

Heat Conduction in Fuel Rod g Heat Flux Moderator -~~-

Temp. Feedwater Flow

, and Entnalpy Void Fraction Vessel

_ HPCI, RCIC Thermohydraulics Boron Concentration SLCS Pressure " " Steam Flow Steam Line Figure A.1 Simplified Flowchart of RAMONA-33 Calculational Logic

Delayed Precursors:

ac 9

at *8["1f11+"2f22)-Acij 1 E 4 E 4 (A-3) where i = 1 to I (I < 6).

Postulating that the thermal neutron leakage term, i.e., v.02742, can be either neglected. or assumea to be constant, ^.he model is further reduced to the well-known 1-1/2 group,' coarse mesh diffusion model. The boundary condi-tions at the core periphery are specified usin; parameters related to the ex-trapolation length for the fast flux and aM for the thermal flux.

The three-dimensional power distribution is calculated as a sum of the prompt and delayed energy generation rates. The prompt component is propor-tional to the instantaneous fission rate, whereas the delayed energy genera-tion rate is calculated using the 1979 ANS Standard 5.1 for decay heat from fission products. fhe cross section dependencies on fuel and moderator tem-peratures, core void fraction, boron concentration and control rods pattern are taken into account in the neutron kinetics calculation. RAMONA-3B calcu-lates the two-group cross sections in the following form:

{=(1-f) d aya +f d aa n + ag(T g -Tgg) n=1 n=1 (A-4)

+a f (/TJ - /q ) + or(c g.a)

The first two terms in equation (A-4) represent the void feedback linear-ly interpolated between the unrodded and rodded cases, respectively.. The third and the fourth terms account for the moderator and fuel temperature re-activity feedbacks, respectively. The last term in the correlation calculates the effect of presence of boron in the moderator should the SLCS be activated.

The RAMONA-3B capabilities in the neutronics area have been recently en-hanced by developing a procedure for collapsing the three-dimensional core kinetics model into a one-dimensional model. The benefits of using a 1-D core model come from reduction of the core modeling complexity, as well as the com-putation time; this approach can be used for those transient applications where the fuel and control rod arrangement and the core flow pattern remain approximately axis-symmetric throughout the transient.

The 3-D to 1-D collapsing procedure is based on averaging the1 macroscopic cross sections -for the nodes located on a horizontal plane. Several static calculations are performed first'by perturbing the core void fraction and mod-erator and fuel temperatures. The same calculatons are repeated for a com-pletely rodded core. The .3-D -nodal values of the neutron flux and diffusion theory parameters (i.e., D 2 Ea....); are - further used to produce Lan equivalent . set of 1-D cross.3,, D .

sections. The diffusion coefficients are'calcu-lated by averaging either the nodes inversed diffusion coefficients or. the -

neutron leakage rate D* grad (+). A complete description of the neutron' kinet-ics 3-D to 1-D collapsing methodology can be found in Moberg et al.,- (1983).

RAMONA-3B editing capability of the neutron kinetics results has' been recently expanded by implementing continuous calculation of different compo-nents of total core reactivity. This capability allows to individually evalu-ate effects .of the fuel temperature, moderator temperature, core void frac-tion, and control rods on the total. core reactivity. Besides providing excel-lent information for analysis of the core response to different perturbations, the edits can supply reactivity functions for point kinetics codes used in scoping studies.

The total core reactivity is calculated from the " inverse'in-hour" equ'a-tion using the total reactor power as the amplitude function. .The different reactivity components are calculated using the perturbation theory which can extract the contributions from the fuel and moderator temperatures, core aver-age void, and the control rods worth during the transient.

A.2.2 Heat Conduction in Fuel Rods Thermal energy storage and heat conduction in the fuel rods are computed using a discrete-parameter model written separately for the fuel pellet, gas gap, and cladding:

pc h = v . kvT + q (A-5)

The three equations are linked through appropriate heat flux and temperature boundary conditions on the structure interfaces. No axial . heat conduction in the fuel or cladding is allowed. Heat capacities (pc) of the fuel and. clad-ding, along with the thermal conductivity of the cladding, are considered to be constants. The thermal conductivity of the fuel as well as the gap thermal conductance are prescribed functions of fuel temperature.

A.2.3 Thermal Hydraulics A typical BWR vessel schematic as represented in the RAMONA-38 code is -

shown in Figure A.2. The model includes a one-dimensional downcomer and a lower plenum. The exit of the lower plenum is connected with the reactor core inlet. To be compatible with the 3-D neutron kinetics, the core is composed i -M-i

STEAMLIAE WITH NSIV,S/R VALVES Iji STEAM DCME STE!JI SEPARATOR l

D R l g I i W S - - - - - - -

______J W

' fd s

E - +.

R - - - - - - -

0 1

N

- - - _ -_ E p

i C ,

u E,s O PUMP R

E --

LOWER PLENUM 2 - - - - - - -

I I 8

I I I

' I LCWER PLENUM I Figure A.2 RAM 0t1A-3B Representation of a Boiling Water Reactor Vessel.

of a number of parallel heated channels modeling the in-bundle flow region.

There is also one core flow channel representing the integrated by-pass flow region. The outlet of the core is connected with the riser (and upper plenum region) which ends with a steam separator residing in the steam dome. The feed-water spargers are typically located in the upper part of the downcomer, whereas the jet pumps are lo'c ated in the lower part of the downcomer region.

The code has a lumped-parameter model for the Pressure Suppression Pool (PSP) that allows prediction of the PSP water inventory and temperature during transients when 'the S/R valves are open and steam is released into the PSP.

The reactor vessel thermal hydraulics of the RAMONA-3B code is based on the following one-dimensional, four-equation model:

Vapor Mass:

h(ap)+V-(jo)=r g gg y (A-6)

Mixture Mass (writte.n in the form of the equation for volumetric flux):

p p Dp Dp 7+dm" o g

ry -[ M+ Dt ) (^- }

Mixture Momentum:

aG G Gl at

["P gg * ( -"}P1t)"- - 90m - EN 2 D (A-8)

Mixture Energy:

h[apu gg +(1a)pu]+h[anvh gg ggg +(1a)pvh]= gg + q(1 a) (A-9)

Two further simplifications are made before the set of equations (A-6) through (A-9) is solved in RAMONA-3B. First, the mixture mass equation (A-7) is combined with the energy equation (A-9) and integrated over the entire vessel; this, along with the equations of state, gives a temporal equation for the average vessel pressure in terms of the vessel boundary conditions (feed-water and ECCS water injection rates, and steam line flow), total vapor gener-ation rate, and vapor and liquid compressibility. This pressure, which is a function of time, but not of space, is used to compute the steam and water properties inside the reactor vessel. However, the code has an option forcal-culating the vessel pressure distribution by back-solving the mixture momentum equation (A-8) for pressure. In the latter case the steam-water properties can be calculated as a function of local pressure.

The second simplification is made by integration of the mixture momentum equation along each of the parallel channels in the core (including the by-pass channel). This results in a set of closed-contour integral momentum equations which, together with the mixture mass equations, i.e., Equation A-7, are solved to calculate the two-phase flow field in the entire vessel. Thus, the above two simplifications considerably reduce the computation burden of the thermal hydraulics calculation without significant loss in nodeling accuracy.

The code hydraulics is based on a slip model approach allowing for a non-homogenous two-phase flow in the vessel. The slip model of the form:

v = Sy g g

+v g (A-10) s is used to calculate the relative veloci ty between the vapor and liquid phases. Non-equilibrium vapor generation and condensation are accounted for through appropriate correlations. The vapor phase is assumed to be at satura-tion, whereas the liquid phase can be either subcooled, saturated or super-heated. The code's constitutive relations package includes correlations for the single- and two-phase wall friction, form losses and wall heat transfer, including the post-CHF regime.

The RAMONA-3B code has a condensation model specifically designed to pre-dict the condensation rate when the subcooled water (either feedwater or ECCS water), injected through the feedwater sparger, enters the steam environment in the upper part of the downcomer region of a BWR-4 type reactor. The in-jected water crosses the downcomer in the form of horizontal jets produced by the feedwater sparger nozzles, and then mixes with the saturated water exiting from the steam separators. The mixture flows downward along the standpipes and along the upper plenum and core shroud, in the form of a film. The con-densation rate is calculated both on the liquid jets and on the liquid film.

The condensation heat transfer on a subcooled liquid jet of diameter D is calculated using the correlation for the Stanton number as follows (Theophanous,1979):

Stj = 0.02 (L/D)-0.5, (A-11) where L is the length of the liquid jet exposed to steam. This corresponds to an average over the length of the jet heat transfer coefficient

5 = pg ,j t,j p,)St)

Y C

, -(A-12)'

3 where gV ,j is the liquid velocity at the nozzle exit.

This leads to the steam condensation rate on N liquid jets of temperature ~

T:g T v, jet " ~A T0T

• P t j p,j g,j_. M (TL C V 3 ,3 - T sat IIN fg , (A-13) where AT0T is the totai-jets surface area.

The second part of the present model accounts for condensation on the downflowing mixture composed of the original jet water, the newly formed con-densate carried by the jets, and the saturated liquid exiting from the steam r separators. The film velocity is assumed to be equal to the free-fall veloci-ty, and the average film thickness is calculated through the flow rate and film velocity. A constant value of 0.0073 is und for the Stanton number fol-lowing Linehan (1968) for the condensation on the -liquid films.

Thus, the total steam condensation rate becomes:

Tv , TOT =V T , jet + ry,fj),=

( A-14 ) --

[0.02.No.jp,34,3(Tg t

C V

,3 -Tsat) 0.0073D bH P ,f Yt,f C f(T,,f-Tsat))

t

-w H

fg where H is the film length from the spargers to the downcomer water level, and Db is the core barrel outer diameter..

The transient boron concentration is computed from the boron mass balance equation:

3[p(1-a)c) g B or *Y*IPc3)=B i gg (A-15) solved in parallel with the liquid and tter mass conservation equations.

Boron is assumed to propagate with the lia .lo locity, and no boron stratifi-cation in the lower plenum is allowed.

l For the steam line the code uses the mass and momentum equations with the assumption of an adiabatic process. Note that the steam lint pressure is a function of both time and space. Therefore, the acoustic effects in the steam line produced by its valves openings or closures are taken into account.

h

i

^

A.2.4 Solution Method- i

\

After all partial differential equations ~ are transformed into ordinary I

, differential equations, the initial or steady-state conditions are obtained by i setting all derivatives to -zero,~ and then iterating to obtain the eigenvalues of the system-of equations. For the transient calculation, different integra-tien methods are used _ for the different. pads of the code. Specifically, the-Gauss-Seldel . iteration is used to integrate the . fast neutron equations, ex-

~

plicit integration for delayed neutron equations, an iterative predictor-cor-rector method for heat conduction, the explicit first order Euler method for the _ vessel- thermal hydraulics, and finally, the fourth-order Runge-Kutta-Simpson method for the steam line dynamics. The neutron kinetics and fuel rod hiat conduction equations are " integrated with a " master" time step, whereas tha vessel thermal-hydraulic and steam line equations use substeps. There are

• several time step control _ procedures assuring -stability and accuracy of the

~

calculation.

A.3. Code Structure and Computer Requirements The RAMONA-3B code is -written in FORTRAN-4 and consists of one main pro-gram with about 120 subprograms. It is operational on a CDC 7600 computer,_

i Scope 2.1 operating system and requires 153K (octal) words of small core mem-j ory (SCM) and 462K (octal) words of large core memory (LCM).

As mentioned above, the code calculation logic is separated into a i steady-state calculation and a transient calculation, each of which, in turn,-

is subdivided into thermal-hydraulic and neutron kinetic sections.- These divisions lead to a user's choice of 'four modes of calculation and an effi-

! cient' use of computer memory through segmentation loading. Operational fea-I tures include restart capability (either from steady-state or transient) and plotting.

The current version of the code allows to use 75 hydraulic. channels cou- i

. pled with 384 neutronic channels and 24' axial cells - to model . the reactor

l. core. The computer run time for a typical 3-D BWR ATWS calculation can be evaluated using the following information. Approximately 8000 CPU seconds l were spent for 1000 seconds of a MSIV closure ATWS transient time; there were j

112 hydraulic cells in the vessel -and steam line models and 1212 neutronic -

cells in the 3-D core.

A.4. RAMONA-3B Assessment and Aoplication The RAMONA-3B assessment matrix includes comparisons of the code results

with data from separate effects and system tests- (FRIGG experiments .(Nyland et l al.,1968); assessment using the FIST tests, Hwang et al., -1983, is underway) l as well as plant transients (Peach Bottom-2 Turbine Trip Test Series,.

Carmichael et al.,1978 and Slovik et al.,1984). Wulff et al., 1984 (p.318) provides a list of the operational transients at different US Boiling Water Reactor plants selected for the future code assessment purposes. Assessraent of the steam line hydrodynamics has been done using both comparisons with ,

analytical solutions for sample problems and with Peach Bottom-2 tests results (Wulff et al., 1984).

V 1-As an example of.the FRIGG-test assessment, Figures A.3 and A.4 show the axial void distribution in the electrically heated vertical. bundle 'as predict -

ed by' RAMONA-38. The calculations were . performed using the Bankoff-Malnes -

correlation '(Jones,1961). As seen from the : comparisons . in Figures A.3 and A.4, the code . predictions agree very well with the test- results. (Since the assessment' work was done in . parallel ~with assessment of. another BWR ~ code, TRAC-BD1/M001 (Taylor et al . , 1984), L the figures show two calculation re-

- sults). The results obtained from the ' FRIGG-test assessment indicated' that.

. the slip model ~ performance is adequate. In order to improve treatment of the l subcooled boi, ling phenomena, a mechanistic model for subcooled boiling (Saha, j 1981)-will be implemented into the code in the near future, t- In the code application area RAMONA-3B has been used -to analyze center

- and 'off-center control rod drop . accidents (Cheng et al.,1982 and Cokinos et al., 1984) .and MSIV Closure . Full- and Partial.-ATWS . (Wulff et' al.,- 1984).

RAMONA-3B has- been, and is currently being, used for the analysis of-a number - '

of' MSIV Closure ATWS - scenarios. (including' the manual rod insertion case, Slovik et. al.,1985) for-the Browns Ferry plant under the 'present NRC Severe Accident Sequence Analysis (SASA) Program (Neymotin et- al.,1985).

In sumary, RAMONA-3B code with its 3-D space-time neutron kinetics (and

. a 1-D neutronics capability) is considered to be a powerful tool for best-es-i timate predictions of the Boiling Water Reactor behavior during transients i where - the core power distribution is expected - to vary significantly with L time. Results-of the RAMONA-38 code application calculations performed so far

have proven to be very useful .in different areas of the nuclear safety research.

RAMONA-3B is available~ to any U.S. organization, on a royalty-free basis, for the analysis of U.S. reactors.

1-1 l

1 i

i I

.FR' IGG RUN NO. 313009 l SUQ-4.4,.G-1107 KG/M2-S, 0-2.98 MW, P-50.0 H Lf.GDdO ews-?e te-m

.....jR 0 C9L - -

8_. d -

. t; -

c .

[ d-c -

( O N

>O-i 8 / . . . . . . i .

! 0.00.5 1.0 1.5 2.0 2.5 3.0 3.5 1.0 4.5 i RXIRL LENGTH (M) l l Figure A.3 Comparison Between the Experimental Data and the Code Predictions of Axial Void Fraction for FRIGG Test 313009.

i

! FRIGG RUN NO. 313001 i 5U4-5.0, G-1492 KG/M2-S, 0-1.5 MW, P-49.6 l

d -

t.t Gi.NO

1. 99A-?9 (9-m w- ...I.fhC-801.(.M09),,,

,d-

• EXPERIMENT o . .

[m o d-c

• tr Ls. N l a. d -

j E-- "

i O~

O

! d ' ~

0.00'.5 l'. 0 l.5 2'. 0 2.5 3.0 3'. 5 4.0 4.5  :

, RXIRL LENGTH (M)

Figure A.4 Comparisons Between the Experimental Data and the Code Predictions of Axial Void Fraction for FRIGG Test 313001.

I t

4 s

NOMENCLATURE ac,bc,cc coefficients used in rod controlled void feedback a0,bO,c0 coefficients used in rod uncontrolled void feedback ci delayed neutron precursor concentration, m-3 CB boron concentration (ppm) .

B boron source strength D diffusion coefficient, m D

d* +V 9h f friction factor fd control rod fraction Gm mixture mass flux, kg/m2s g acceleration due to gravity, m/s2 j volumetric flux density, m/s p pressure (Pa) q linear wall heating rate, W/m q gamma heating rate, W/m3 S slip ratio St Stanton number Tf fuel temperature, 'R Tg moderator temperature, 'R u specific internal energy, J/kg v velocity, m/s vo bubble rise velocity, m/s

Greek a void frei. tion 6 total delayed neutron fraction ry- evaporation rate, kg/m 3s-A decay constant for delayed neutrons - s-1 E macroscopic neutron cross section,- cm-1

+- neutron flux, m-2 s-1 2

4,_ two-phase mixture friction multiplier o density, kg/m 3 v mean number of neutrons in fast or thermal group

,' Subscripts 1 fast-group neutrons 4

2 thermal-group neutrons l

a absorption f fission g vapor (saturated) i index for delayed precursor i liquid i m mi xtu re I

i a

4 l  :

REFERENCES L.A. Carmichael and R.O. Neimi, " Transient and Stability Tests at Peach Bottom Atomic Power Station Unit 2 at End of Fuel Cycle 2," EPRI NP-564,1978.

H.S. Cheng and D.J. Diamond., " Analyzing the Rod Drop Accident in a Boiling Water Reactor," Nuclear Technology, 56, p.40,1982.

D.M. Cokinos and J.M. Carew, " Comparisons of a Center- and Off-Center BWR Con-trol Rod Drop Accident," Transactions of American Nuclear Society, 47, p.405, 1984 W.S. Hwang, et al., "BWR Full Integral Simulation Test (FIST) Phase 1 Test Re-sults," NUREG/CR-3711, EPRI NP-3602, GEAP-30126, November,1983.

A.B. Jones, " Hydrodynamic Stability of a Boiling Channel," KAPL-2170, October 2, 1961.

J.H. Linehan, " Interaction of Two-Dimensional Stratified, Turbulent Air-Water and Steam-Water Flows," ANL-7444, May, 1968.

L. Moberg and L. Nordin, " User Manual for FRAM: A Data Processing Link Between 3-D and 1-0 RAMONA-3B Neutronic Models," Scandpower, Norway, October, 1983.

L. Neymotin, et al., " Evaluation of BWR Emergency Procedure Guidelines for BWR ATWS Using RAMONA-3B Code," Third International Topical Meeting on Reactor Thermal Hydraulics, Newport, Rhode Island, October 15-18, 1985.

O. Nyland, et al., " Hydrodynamic and Heat Transfer Measurements on a Full-Scale Simulated 36-Rod Marviken Element with Uniform Heat Flux Distribution,"

FRIGG-2, R4-447/RTL-1007, 1968.

P. Saha (1981), "A Simple Subcooled Boiling Model," ANS Transactions, Vol . 39, pp. 1058-1060. Also, BNL Memorandum " Improved Subcooled Boiling Model for TRAC," dated June 2, 1981.

G.S. Slovik and P. Saha, "A RAMONA-3B Code Description and Nodalization Study for the Peach Bottom-2 Turbine Trip Test 3," 12th Water Reactor Safety Re-search Information Meeting, October, 1984 G.C. Slovik , et al ., " Application of RAMONA-3B to BWR ATWS," 13th Water Reac-tor Safety Research Information Meeting, October,1985 D.D. Taylor, et al ., " TRAC-BD1/M001: An Advanced Best Estimate Computer Pro-gram for Boiling Water Reactor Transient Analysis," NUREG/CR-3633, EGG-2294, April , 1984.

T.G. Theofanous, "Modeling of Basic Condensation Processes," presented at the WRSR Workshop on Condensation, Silver Spring, Maryland, May 24-25, 1979.

G.J. Van Tuyle, et al., "RAMONA-3B/MINET Composite Representation of BWR Thermal-Hydraulic Systems," Third International Topical Meeting on Reactor Thermal Hydraulics, Newport, Rhode Island, October 15-18, 1985 W. Wulff, et al., "A Description and Assessment of RAMONA-3B MODO Cycle 4: A Computer Code with Three-Dimensional Neutron Kinetics for BWR System Tran-sients," NUREG/CR-3664, BNL-NUREG-51746, January,1984 I

I l

I l

i e

f 4

i

i.

_y y . . - -_ w - y , .,4 _.- ,.- - e

APPENDIX B Details of Cross Section Generation The neutron kinetics methodology in RAMONA-3B uses a library of sets of two group macroscopic cross sections evaluated at a specific exposure and void

- history. This is done for two reasons. First, the integrated values of ex-posure and void history do not change enough during a transient to make the cross sections sets a strong function of these variables. Secondly, the com-putational expenses required to functionalize a series of tables to fit all the variables and evaluate each node during a transient is not worth the gains of increasing the accuracy. Hence, RAMONA-3B has adopted a well known and tested feedback model (Diamond,1976), (Cheng et al .,1980) and ("BEAGL-01",

1984) which uses a format usually called the BNL TWIGL format:

I(a,Tp,Tm,F E,V) = F(a + ba + ca 2

) + (1-F)(a' + b'a + c'a2 )

+ P(Tm - mT *) + R (/TF - /Tp*) (B-1) where a = instantaneous void, Tp = fuel temperature, Tm = moderator temperature, F = control fraction, E = exposure, V = void history, P = 6E/6Tm.

R = 6E/6/T F, a,b,c = coefficients used in rod controlled void feedback, a',b',c' = coefficients used in rod uncontrolled void feedback,

• = reference case The functional form of the feedback model has been presented in Eq. B-1 and the processing of the neutronic data using BLEND 2 to fit the necessary coefficients of Eq. B-1 can now be described.

BLEND 2 takes the grid information for each fuel type and its location in the core along with the corresponding (E,V) data from the TVA process computer and evaluates the coefficients found in Eq. B-1. This procedure is schematic-ally shown in Figure B.l.

4 a

3-D CROSS SECTION GENERATION (MACROSCOPIC CROSS SECTIONS MICBURN AS A FUNCTION OF FUEL TYPE, ANALYSIS 60 m CASMO VOID FRACTION, FUEL TEMP.,

EFFECT MODERATOR TEMP., CONTROL FRACTION, EXPOSURE, VOID HISTORY.)

i (COLLAPSES CASMO CROSS SEC-

' NODAL EXPOSURES TIONS AROUND SPECIFIC EX-AND + BLEND POSURE AND VOID HISTORY.")

, VOID HISTORIES e

i

, r RAMONA-38 t

i

Figure B.1 Overview of 3-D Cross Sections Generation Procedure i

i d

As a simple example, we can follow the BLEND 2 procedure of determination of the macroscopic cross section coefficients for the moderator temperature coefficient. Each "X" in Figure 3.1 represents a full set of 2-group param-cters like those shown in Table B.1. To find the moderator temperature coef-ficient corresponding to the fast group diffusion coefficient D , the z evalua-tion is performed by intergolating between the points found in Figure 3.1 cor-responding to reference TM and perturbed moderator temperature Tgg.

I g ) - D i(a,T M,Tp Et ,V1 )

Di (a,Tgi,Tp Et ,V P=

TM1 - Tg where Tgt = perturbed moderator temperature, TM = reference moderator temperature, Tp = reference fuel temperature, a = reference void fraction (i.e., a = V ),1 D i

= fast group diffusion coefficient, Et ,Vt = exposure and void history associated with the particular node.

Each coefficient found in Eq. B-1 (for every parameter in Table B.1) was de-termined in a similar manner for each node.

Once each node is assigned its set of macroscopic cross sections corre-sponding to the set found in Table B.1, BLEND 2 does a constrained minimization process using the exposure and void history of each node to group members of a fuel type (a group is called a chain). The following criteria are used:

Ej - Ej < 6E Vj - Vj < 6V Vj where a chainEj, head, whileare the exposure Ej,Vj void historyof are the properties of a nodenodes other arbitrarily of thechosen sameas fuel ty pe. These evaluations are performed to determine whether each node falls within the criteria of acceptance. If it is selected, then the node is added to the chain with Ej and Vj as the chain's head. Several sweeps are conducted with increasing bounds of acceptance for 6E and 6V until the number of chains drops below a predetermined number of cross sections sets. The cross sections sets are then averaged and assigned an identification number.

The identification numbers and their locations are printed out by BLEND 2 (see Figure 3.2). The cross sections are automatically printed in the RAMONA-3B format.

l l

1 I

l 1

Table B.1 Cross Sections Formats of RAMONA-3B

0) = (1-f)(a)) +a 21 * + 831 a ) + f (a16 + a26 2 +a 36" )

+a 47 (Tm -Tmo) a+a 66

• I D2 * (I ~ )(uS1 + #61 a+a 71 a ) + f (a46 + a56

+a l2(T, - Tmo)

C el = (1-f) (a22' #32 2 + a 42 2 ) + f (a76 + aj7a+a27" )

+a 52 (Tm -Tgg) + a62 I ~\ o)

St1

• E31

^I r1 = (1-f) (a72 + ay3 a+a 23 a ) + f (a37 + a47 .i + a57 * )

+a 33 (Tm -Tmo) + a43 I -

o)

E *8 a2 = (1 - f) (a53 + a63 73 a ) + f (a67 + a77 " + "18* )

+ a)4 (Tm -Tmo) 2 a+a

= (1-f) (a24 + a34 a+a 43 2)

VE 2 fg 44 a ) + f (a?.8 + a38

+a 54 (Tm -

mo) + a64 (

o) vrf2 = (1-f)(a74 + a15 2 +a 25 " )

• I I 58 + a68 a+a 73 a)2

+a 35 (Tm -Tmo) 1

- 100 -

._ ____.__J

l I

References "BEAGL-01: A Computer Code for Calculating Rapid LWR Core Transients," EPRI NP-3243-CCM, October 1984 (prepared by Brookhaven National Laboratory).

H. S. Cheng et al., "The Use of MEKIN for Light Water Reactor Transient Calcu-

lations," BNL-NUREG-28785, Brookhaven National Laboratory, November 1980.

D. J. Diamond, Ed., "BNL-TWIGL, A Program for Calculating Rapid LWR Core Tran-sients," BNL-NUREG-21925, Brookhaven National Laboratory (1976).

i l

{

- 101 -

APPENDIX C 3D to 10 Collapsing The 1-D set of cross sections was generated from the 1/8 core symmetry model with the initial steady state point corresponding to the TVA operational condition evaluated in Section 3.1. An overview of the procedure is presented in Figure C.1. Since there are a maximum of 8 coefficients to fit for any group diffusion parameter (see Table B.1), there were 8 static runs required.

These calculations are performed to generate input for the FRAM (Moberg, 1983) code. The caseq were:

1. Initial steady ' state (or base case);

2. Uniform void perturbation of a = a0 + .12;

3. Uniform void perturbation of a = a0 .12; i 4 Initial steady state with all control rods inserted;

5. All control rods inserted with uniform void perturbation of a=a0 + .12;

6. All control rods inserted with uniform void perturbation of a=a 0- .12;

,1 l 7. Uniform perturbation to fuel temperature Tp=Ty-150'C;

8. Uniform perturbation to moderator temperature T g =T g - 25*C.

The values of the uniform perturbations for the void, fuel and moderator i temperature came from experience gained by running the 3-D case prior to the I generation of the 1-D equivalent set. _

The option used in FRAM was an exact fit for the group coefficients with the 1/D averaging algorithm for the group diffusion parameter. The weighting

, function was chosen from the adjoint flux approximation.

i A comment about the power buckling correction used in FRAM would be appropriate. Even with rigorous collapsing methods, proper treatment of side reflectors, etc., the calculated 1-D cross sections set may still not repro-1 duce exactly the average axial power shape and Keff calculated in the orig-inal 3-D case. To correct for this, the so-called power / buckling method is  :

I employed in FRAM. In this method, the fast radial buckling terms are defined

so that the fast neutron balance on each horizontal plane and the Keff of the original 3-D case is preserved. Details of this procedure can be found in the FRAM manual.

j References 4

L. Moberg and L. Nordin, "FRAM: A Data Processing Link Between 3D and 10

RAMONA-3B Neutronic Models," prepared for BNL, October 1983.

1 l i

l

- 103 -

]

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

1-D CROSS $[Cilots ontvitu

. 0m-n 5 % "!!it'E'."J M A .

v m ,. 'M"litYMt.'

o FMR yEAWMgO h0 bl UR$

il Mnou-se Eiko Figure C.1 Overview of 1-0 Cro'ss Sections Generation tiethod

- 104 -

_ _ - - _ . _ l

APPENDIX D Sensitivity of the Condensation Model The condensation model for the case where the water level in the down-comer drops below the feedwater spargers has been described in Appendix A. In order to quantify the effects of condensation on the subcooled liquid injected through the spargers as predicted by the RAMONA-3B code, an additional calcu-lation has been run following the Transient I scenario." In this calculation the condensation rate was assumed to be high enough to bring the injected water to the saturation state before this water merges with the water residing in the lower part of the downcomer region.

The following figures show some selected results of both calculations up to 500 sec of the transient time. It is seen from Figures D.1 and D.2 that the total condensation rates in the downcomer differ significantly in the two calculations. The oscillations in the condensation rate predicted in the calculations after 200 seconds are caused by the changes in the ECCS water injection rate. (As was mentioned before, the water level had been kept at TAF by adjusting the water injection rate.) The corresponding plots of the cater subcooling vs. time at the core inlet are shown in Figures D.3 and D.4 The sensitivity of the total reactor power to the condensation rate on the ECCS water can be appreciated when observing Figures 0.5 and 0.6, where the total power is plotted as a function of time. Note that the core inlet subcooling is the parameter relating the condensation rate to the neutron kinetics and therefore to the power. It is seen, for example, that between 500 and 600 seconds, the total reactor power was approximately 20% in the RAMONA-3B best estimate calculation, whereas the high condensation (sensitiv-ity) calculation predicted about 12%. The difference in the power predicted in both calculations led to a difference in the Pressure Suppression Pool water heatup rate, as seen in Figure 0.7.

1

- 105 -

3-D BROWNS FERRY m CONDENSATION ON ECCS Q 100.0, 220.5 B

80.02 Cr3 gg 60.0- -132.3 4a0- b 20.02 - 44 1 04 ' . . . . . 0.0 i O 0.0 100.0 200.0 300.0 400.0 600.0 000.0 TIME (S)

Figure D.1 Total Condensation Rate on the Injected Water (Best-Estimate Case)

HIGH CONDENSATION g CONDENSATION ON ECCS g 400.0 881.8 0

b 300.0- -661.4

- =

g 200.0- ,

-440.9}

u 100.0 - -220.5 e

o 0.0 , , , . , 0.0 0 0.0 100.0 200.0 300.0 400.0 500.0 000D TIME (S)

Figure D.2 Total Condensation Rate on the Injected Water (Sensitivity Study - liigh Condensation Case)

- 106 -

3-D BROWNS FERRY CORE INLET SUBC00 LING VS. TIME 30  %.0 6 :1 30- -36.0 v ,

ll

-18.0[

l30- O.

no 3dno ad&O 3500 4500 Odno

- 0.0 8000 TIME (S) l Figure D.3 Core Inlet Subcooling (Best-Estimate Case)

HIGH CONDENSATION CORE INLET SUBCOOLING VS. TIME So %0 6 }

20- -36.0 a

M 10-

-18.0{

l 0-00 ido.O e50.0 300.0- 4dn0 Sdno

- 0.0 Sono TIME (S)

Figure 0.4 Core Inlet Subcooling (Sensitivity Study -

High Condensation Case)

- 107 -

3-D BROWNS FERRY RELATIVE POWER VS, TIME 8,

tai g si y tai 1-

%s' L; 4AAA -

a o e & .60. A0 .50. 0.

TIME (S)

Figurs D.5 Total Reactor Power (Best-Estimate Case)

HIGH CONDENSATION RELATIVE POWER VS. TIME 3

b in:

o a-C=3 1-d 0

1H .-? _

I

--s 4 0.0 1d0.0 S$0.0 3d0.0 4do.0 650.0 .0(LO TIME (S)

Figure D.6 Total Reactor Power (Sensitivity Study - High Condensation Case) l

- 108 -

I

I I I I i i 200 - -

180 - -

160 - -

w -

• BASE CALCULATION D - - -

D l40 3.53 F / min

$ N

~

g120 -

p -

HIGH CONDENSATION -

g 100 -

2.40 0 F / min ,

0- -

80 -

I I I f f f l 0.0 10 0 200 300 400 500 600 TIME (s)

Figure D.7 Suppression Pool Water Temperature

- 109 -

NRC Fonu 335 " " " " " ' '

, U.S. NUCLE AR REGUL ATORY COMMlWON

"hfUREG/CR-47b '"'

Bl%IOGRAPHIC OATA SHEET BNL-NUREG-52021 /

< , , m A N o >U ,, , ~ ,-n - N . . . . . .. .... . ,

RAMONA-3B Calcula ions for Browns Ferry ATWS Study

2. a.

j 1 AUTHOH(S) S. D ATE REPOHT COMPLE TED/

P. Saha, G.C. Slovi and L Y. Neymotin * "'" eptember S 1986 il Pt Ht OHMING ORGAN 1/ATION N A? AND M AILING ADDRESS (laciuw /,o codes DATE HE POHT ISSU$d Department of Nuclear nergy * "Novemp d I "'" 1986 Brockhaven National La ratory ,

Upton, NY 11973 y ,,,, ,,,, ,/, j a aa-a y o 11 NPONbOHINti OltGANI/AflON NAMt ANh MAILING ADOHE55 Itactuo, tp Co*f Division of Accident Evald tion 10 P7JkCTITASK/WOHK UNif NO U.S. Nuclear Regulatory Co .ission gnu so.

Washington, D.C. 20555 /

A3273

/

1J $ TVt O> Ht v0H f we n.oo CQ& E ne o Itoerousser d.arso Technical Report (Formal) g j/

14 buPPLEMEN TAHY NOTE S 14 (Le.ve cravel I6 AISTH ACf (100.=< sten ne seus

[

Several aspects of the Anticipated Tranglent Without Scram (ATWS) initiated by an inadvertent closure cf all Main Steam Isolation Valves (MSIV) in a typical BWR/4 are analyzed in the report. The analyi.is isjperformed using the Brookhaven National Laboratory code, RAMONA-3B, which erploys a three-dimensional neutron kinetics model coupledwithaparallel-channeltherpal-hydraulicsinrepresentingaBoilingWaterRe-actor (BWR) Core.

4 Four different transient scenario'k have been investigated: a) downcomer water level and reactor pressure control, b))

c) high pressure boil-off, and 4) recih, culation manual control pump trip rodfailure.

insertion transient, h

Results of these calculatio/ns should effects of operator actions d0 ring ATWS,\ provide thus helping in thebetterdevelopmentunderstanding of adequate of mitigative Emergency Procedure Guidelipes (EPG) required for the BWR plant safety.

f A few unresolved / quest [ons subject to \uture investigation IF KE Y WOHOS ANO DOCUME NT AN Af 515 17. DESCHiPtOHs Anticipated Transient ithout Scram (ATWS)

BWR/4 Emsrgency Procedure idelines (EPG) k Manual Rod Insertto Three-Dimensional eutron Kinetics \g MSIV Closure Nuclear Reactor aafety i /.e allt N fil it H5 Ol'F t ND6 II if HMS 3 8 AV AI TY $T ATEMENf 59 $E R6 f Y CL Ah5 t,ans rceored J l No Of P ALE 5

,. ,E ,,. A .,a...... .. .H.cE w ac e oaw an .. ...

- 111 -

~ . _ - . -

m..._. _ _ - -

-.-___.-_.e_.

_e . _ _ _ - _ _ . . _ _ _ _ -

t g

ISL, "

'J ~ 74177

,IL ' ivc-OAC 1 ur M-ACM 1 % "> 1 '

• 1 s1 ~

jc :a

,rtIcy t, . r c , f c ,-

u,-

961 < s.pga gguc a

-1 1 s

p it64 '-A a y s ,~

i %p' s }

b

~

f}Q t'. g--

1 a

a 4

5 1

1t l.

l 1

i i

i 5

i I

I E

l' 4

2 r

6 6

e 5 I 1

s 1

1 J

f I

f P

1 1 -