INIS-mf—14766 Nuclear Safety Research Association

&

Loss of Coolant Accident Analysis(Thermal Hydraulic Analysis)

-Japanese Industries Experience-

K. Okabe(Mitsubishi Heavy Industries ,Ltd)

presented atInternational Specialists' Meeting on Fuel Behavior

and Thermal Hydraulic Behavior in Accident Conditions7-9 November 1995 , Sofia

•vn. 71% 1 3

In Operation ' ft

Under Construction fi

Under Planning Q

TOTAL

Units

22

1

2

25

Capacity (MWe)

18,186

1.180

2,840

22,206

TOMARI P /S t l

OQTSURUGA P/S tk

TAKAHAMA P/St

Status of PWR plants in Japan

Overview of Loss of Coolant Accident Analysisin Japanese Industries

1970s1 o Introduction of computer programs from U.S

1980s' oThese programs are well verified and improvedthrough safety researches of Japan

oSome programs are originally developed

oApplication to the operational safety area

1990s1 oComputer programs are utilized in proposalsof new PWR safety design concepts

oFurther improvement of computational technique

Examples of Application to the operationalsafety area

o 1D drift flux model base computer program (CANAC)was developed by Mitsubishi

o Advanced trainning simulater based on CANAC and otherprograms was developedIt can simulate well PWR behavior in accident conditions andwidely used for the education and trainnings of PWR operators

o Emergency operating procedures were also developed based onthe thermal -hydraulic analysis capabilities

m PWR Full S c a l e S i m u l a t o r at T r a i n i n g C e n t e r

* . ^ £ z r --•- •:,-•• •••••• I '

CO

III. LOCA ANALYSIS WITH COMPUTER CODES

EVALUATION-MODEL (EM) CODES

- CONSERVATIVE, SIMPLIFIED MODELS.- ACCEPTED FOR DESIGN ASSESSMENT AGAINST

DESIGN-BASIS SCENARIOS.- CONSERVATISM NOT ENSURED FOR ALL POSSIBLE

SITUATIONS.- Do NOT PROVIDE BASIS FOR OPERATOR ACTIONS.

BEST-ESTIMATE (BE) CODES

- MORE REALISTIC, PHYSICALLY-BASED MODELS.- APPLIED TO BEYOND-DESIGN-BASIS SCENARIOS.- NEED CAREFUL ASSESSMENT & UNCERTAINTY

EVALUATION.- YET USER SENSITIVE - NEED QUALIFIED USER.

Example of verification and Improvementof U.S Computer programs

BASHdeveloped by westinghouse forreflooding analysis at largebreak LOCA of PWR

1 D/non-equilibrium two phaseflow and quench frontpropagation are modeled

improved byMitsubishi

JAERI• Large

fib

scale Reflooding Experiment

I. I . I . IM«I tot*

kI Data on droplet behavior / \V void fraction in large core et.al/

BASH-Mused for licensing calculation in Japannew plant, plant modification (SG replace et.al)

CCTF Experiment

BASH-M Calculation

1200

1100

700

600

500

8 ft(Elevation from

bottom of core)

"0 100 200 300Time after initiation of reflood (sec)

400

Comparison between CCTF experiment (C19)and BASH-M calculation

Display for Education at Training Center

Example of Proposals of New Concepts

°Various organizations propose new concepts for 21 st century

°Example of concepts proposed by MitsubishiD Steam Generators are used for the decay heat removal

at LOCA conditionsD Horizontal type SG is the best to maintain two phase

natural circulation under the Reactor Coolant System submerged

°Experiments and analysis are being done

Passive Safety System inside CV

Gravity injection pit

SecondaryContainment \ depressurization

vessel \ va|ves

Water level in CVk_

:eactor vessel

Reactorcore

Steam generator

Reactorcoolant pump

Cold leg

Condensate\ storage tank

Main steam line

Feed water line

Passive Core Cooling System usingHorizontal Steam Generators

(T) Reactor Vessel(2) Downcomer(3) Horizontal steam Generator

(Intact Loop)(4) Horizontal steam Generator (Broken Loop)(5) Hot Side Channel Head(e) Cold Side Channel Head(7) Pressurfzer(5) Containment Vessel(9) Accumulator(ft) Gravity Injection Pit(Q) Reactor Coolant Pump@ Primary Depressurlzation Valve (j

Flow directionunder normaloperation

Vent Line

SLlM-facility

fQuasi steady state natural circulation testI - comparison of loop flow - 11250

1000

Test

— — Break loop

M

« 750

o 500

I 250

0

Intact loop N2 gas injection starts

T

Water subtraction terminatesK(J0% inventory) V,

0 20 40 60 80 100 120

Time from the subtraction star t (min)

1250

5 1000

S 750

§ 500

8 250

TRAC-PF1 Calculation

0

Break loop N2 gas injection starts

— — Intact loop

Water subtraction terminates

0 20 40 60 80 100 120time from the subtraction start (min)

TMTrmM in ii—„>

SG I

II III Ih-CI I I I I I I I

Gravity injection

O

II I

I I I1 L

D C

Gravity injection

O I h — i n n n

R V

J_L

»M IN II

-» i i V-j

I S G

ITTTT

I

I LUXLLL

I I

; Breakpoint

'. CV condition is treated as boundaryconditions.(only large break case)

Nodal schematic .of TRAC-PFl/mod2 for SLIM test

TWO-PHASE NATURAL CIRCULATION(TYP. COOLANT INVENTORY a 90 - 50%)

REFLUX CONDENSATION(TYP. COOLANT INVENTORY = 50 - 30%)

— 20 —

TYPICAL SMALL-BREAK LOCA PHENOMENA

- NATURAL CIRCULATION & REFLUX CONDENSATION- STRATIFIED TWO-PHASE FLOW IN HORIZONTAL LEGS- ENTRAPMENT FROM STRATIFIED FLOW- COUNTER-CURRENT FLOW LIMITING & LIQUID HOLDUP- COEXISTENCE OF SUBCOOLED LIQUID AND STEAM

CD

00

100

>DCOh -!ZLU>

ooo

<a9 20toLUDC

00

BREAK LOCATION EFFECTS ONCOOLANT INVENTORY LOSS RATE(SIMULATED 2-IN. BREAK, FAILED HPI)

• CORE HEATUP

HOT LEGHORI2. BREAK

HOT LEGTOP BREAK

2000 4000 6000

TIME AFTER BREAK (s)8000

MASS & ENERGY BALANCE DURINGA SMALL-BREAK LOCA

oo\°

en

* o o

STEAMCONDENSATION-

DECAY HEATREMOVAL BYSGs

CORECOOLING

BREAK AHEA EFFECTS ON PRIMARY SYSTEMDEPRESSURIZAT1ON(SIMULATED COLD LEG BREAK, FAILED HPI)

0

• CORE HEATUPO CORE REWET

0.5%

ACCUMULATORINJECTIONSETPOINT

500 1000 1500 2000 2500

TIME AFTER BREAK (s)

Summary

°Previous safety researches enable us to• propose the reliable safety design for licensing• support the education/trainning of operators

oBest analytical model for two phase flow should beselected by the concerned problem characteristics

• 1D,2D,3DQ homogeneous , drift-flux , two velocity , three velocity• equilibrium , non-equilibrium

oFurther improvement of computational technique is requiredfor the efficient problem soluing in some area

3.50E-03

3.00E-03

2.50E-03

<5 2.00E-03GOCO

s 1.50E-03

Quasi steady state natural circulation testTRAC calculation accumulation ot N2 gas

N2 Flow direction

Time afterN2 injection

50sec

102sec

290sec