_ _ _ _ _ _ _ _ _ _ ____ _ ___ _ _

,

*

TOPfCAL REPORT EVALUATION ON CADDS,

*,

!

Report Nos. and Titles: BAW-10098P, Rev.1, "CADDS - Computer Applicationto Direct Digital Simulation of Transients in PWRswith or without Scram." (Proprietory), December 1977

BAW-10098, Rev.1, "CADDS - Computer Application toDirect Digital Simulation to Transients in PWRs withor without Scram," (Non-Proprietory), January 1978.

Originating Orgainzation: Babcock & Willcox Company

.

I. Summary of Tooical Reports

Reports BAW-10098P and BAW-10098, Rev.1, describe the CADDS program used i

by Babcock & Wilcox to analyze reactor transients, with or without reactor scram,

in a pressurized water reactor. The CADDS code is identical to the CADD computer

code which was previously approved by the staff (Reference 1), with the exception

that steam generator modeling has been incorporated within CADDS. CADDS solves

the time-dependent neutron kinetics equations in conjunction with a thermal-

hydraulic solution for an average fuel pin during a transient. The code incor-

porates the major feedback mechanisms while accounting for some of the details

of single phase, nucleate boiling and film boiling heat transfer in the reactor

core. The pressure along the coolant channel is assumed constant for each time

step.

For a simultaneous solution of the entire primary system, the user couples

the core model with a one loop simulation of the reactor coolan't sys'te'm (hot leg,

steam generator, cold leg, reactor core, and pressurizer). A region-averaged.

model of each portion of the loop is utilized in determining the temperature

response in the loop for either variable or constant flow; the primary to the,

secondary system is determined by either of two methods: (1) solving a set

of thermal-hydraulic equations on both sides of the steam generator or (2) the.

8309120265 030031PDR TOPRP ENVBWC PDR 1

- _ - _ _ - _ _ _ - - _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ - _ _ _ ___ _ _ .

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

.

m

u'ser may input the itsam generator heat transfer to the secondary side as,

a function of time.

A reactor core can be evaluated independently (separate from the

primary system) by using as input reactor inlet enthalpy, reactor inlet.

flos, system pressure, and either rod and boron reactivity or heat' generation as functions of time. The use of the loop option requires

that only reactor system. flow.versus time and.ro'd an'd boron reactivity '(or'

'

heat generation) versus time be input.

The neutron kinetics model incorporates reactivity feedback effects ,

due to moderator density and temperature changes and Doppler broadening.

The fuel pin is represented by a detailed axial and radial sectionalization

to provide better thermal eva!uation for the neutronic feedback calculation.

The ability for the CADDS user to set reactor trip points allows

following accidents to completion with core shutdown. Using additional pro-

gram options, the user can (1) include decay heat and (2) ascertain the

cladding oxidation from penetration distance calculated from a metal-water

reaction based on the Baker ind Just parabolic rate law.

Thus, the CADDS program can be used to study typical anticipated

transients such as those caused by flow variation.

CADDS is not utilized to predict localized heat transfer p~henomena such

as hot channel DNBR. This is done by inputting core flows and . inlet

| thermodynamic conditions from CADDS int'o a detailed core heat transfer model., , _ _

The B&W RADAR Code (Ref. 2) has been approved for this purpose..

6

II. Summary of Reaulatory Evaluation*

A. Review Approach;

,

In various publications, Babcock & Wilcox has stated that CADDS:is used

for transient analyses of the following events: (The numbers preceding the

.

*

2

.- - .- .. - -

.

.

.

.

events refer to the Safety Analysis Report section in which the results

of the transient analyses are reported).

15,1.1 Decrease in feedwater Temperature

15.1.2 Increase in Feedwater Flow,

15.2.6 Loss of Nonemergency AC Power to the Station Auxiliaries

15.2.7 Loss of Normal Feedwater Flos

15.3.1 Loss of Forced Reactor Coolant Flow Including Trip of~

Pump Motor and Flow Controller Malfunctions

15.3.3 Reactor Coolan't Pump Rotor Seizure

15.4.1 Uncontrolled Control Rod Assembly Withdrawal From aSuberitical or Lcw Power Startup Condition

15.4.2 Uncontrolled Control Rod Assembly Withdrawal at'. Power

15.4.3 Control Rod Misoperation (System &lfunction or OperatorError)

15.4.4 Startup 'of' an Inactivs 'L60p or' Rec'irchliti~oriToop at an -~

Incorrect Temperature

15.4.6 Chemical and Volume Control System Fhifunction That Resultsin a Decrease in Boron Concentration in the Reactor Coolant(PWR)

15.4.8 Spectrum of Rod Ejection Accidents

15.8 Anticipated Transients Without Scram

To support our evaluation, we have applied the acceptance criteria pre-

sented in the respective sections of the NRC Standard Review Plan (NUREG-0800,

July 1981 Revisions). We have reviewed the applicant's suppo' ting' derivations~

r

and experimental data and have made a series of audit calculations using the

RELAP4/ MOD 6 computer code. Our conclusions regarding the use of the CADLS code

as described by BAW-10098P, Rev.1, and BAW-10098,'Rev'.l', are' stated in the -

Staff Position section of this report.

.

3

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

.

.

,5. F.eview of Analytical Fbeels

CADOS is a one-loop fixed nodalization code (with the exception cf ne

core and the steam generator wnere variable axial nodalization capability is

available) modeling a reactor core with a bypass, a steam generator, one hot-

leg and one cold leg node, and a nonequilibrium two-region pressurizer on the

hot leg. At B&W's request, this review was restricted to the multinode

Integral Economizer Once Through Steam Generator-(IE0TSG) option and the OTSG.

model was not evaluated. Although CADDS has a pump model, B&W also requested

that it not be reviewed. CADDS has no boron transport model, ECCS is not in- .

cluded, no provision was made for a hot channel DNB calculation. There are no

bubble rise models, and except for the pressurizer there are no metal heat

slabs so metal heat transfer cannot be accounted for. Inclusion of metal' heat|

would act to dampen the pressure response of the transients listed in Section II.

A. and its omission is therefore conservative.

The code has a partial initializer with the pressurizer initialized.at,

saturation given the initial pressure and mass of water while the steam generator

initialization depends upon the nodal option chosen. For the single node option,

a tube side heat transfer coefficient is computed using initial heat load and

average tube metal / coolant temperatures. In the case of the multi-model, the

primary side pressure, flow, inlet and outlet enthalpy, and the secondary side

pressure, feedwater flow rate and enthalpy are input by the user. .To obtain

the thermal steady state, the code adjusts the feedwater flow. The secondary_

side pressure distribution is calculated by solving the steady state momentum.

equation with input form loss coefficients. Input loss coefficients are also

used for the other parts of the system where a pressure drop is . involved.

A detail evaluation of the major models in CADDS follows. *

Core Neutronics Model

In addition to a power input table option, CADDS provides a kinetics

.

G

4

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

! -

! .

'

,opticn using the well-known point kinetics equations. The code permits uo

to twelve delayed neutron groups, and reactivity feedback due to moderatorI temperature changes taking into account the axial variation of the temperature

by the use of weighting coefficients. The reactivity feedback due to moderator

density changes also accounts for axial density variations and differences from

the average or hot fuel pin. The reactivity feedback due to fuel temperature

changes (Doppler) accounts for both the axial and. radial variation of the

temperature of a fuel pin by using weighting procedures. The reactivity feed-

back model also features trip reactivities and reactivity insertions which are ,,

input by the user in tabular form as' a function of time. The small reactivity

effect from subcooled boiling is not modeled which is conservative since the

inclusion'of this effect would act to depress the neutron flux at the core hot

i spots.

The method of solution of the CADDS point kinetics equations in conventional

in that it is based up the CHIC-XIN program-developed-by-the-Bettis-Atomic - -~;

Power Laboratory. A procedure is used to centrol the time step to ensure the

stability of the equations. However, the accuracy of the solution may be reduced

; if the input minimum time step is too large.'

The point kinetics model used by CADDS is satisfactory with respect to the

equations, method of solution, and reactivity feedback models.

Decay Heat Model, , ,

CADDS has as an option, the inclusion of a six group decay heat model to

the total power. The decay heat is based on infinite operation at the initial

power of the reactivity transient. The equations used in this calculation are:' similar in form to those used for computing the delayed neutron precursors. The- *

formalism used-for the decay heat model is satisfactory. The input constants

.

5

- . . . . - . .. - - - . - - - . , - - . --

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

*

i

lis'ed ir. tne topical report provide for decay heats the same as or in excessj t,

) to t*:se Of the 1973 AUS standard and are therefore acceptable.

| . . . . _ _ . _ _ _ . _ . .

Core-Heat Transfer Model

The principal use of the core heat transfer model in CADDS is to provide*

a heat source to the reactor coolant system. Detailed core heatup calculationsi

| for licensing purposes are performed using a sepa, rate computer code such as RADAR

which takes reactor coolant thermodynamic input from CADDS.,

The heat generation model assumes that the heat source is a separable,

function of space and time. The spatial variation, radial and axial, of the

h?at source can be input to CADDS and is invariant in time. The time depend-

ent variation of the amplitude of the heat source can either be input toi

I CADDS or computed from the point kinetics equations. Fixed fractions of the

generated heat are deposited directly into the coolant and into the cladding,

i Multinsde radial heat conduction through the fuel pin with an explicit gap

conductance is considered, but axial heat conduction is not. The material.

properties used in this calculation in general compare well with the data4presented in the MATPR0 and NSM handbooks . The U0 thermal conductivity

2

is predicted to within a few percent of the data base up to 1500k. Beyond

that temperature it is overpredicted by as much as twenty five percent but

better agreement is obtained for higher purosity fuel. The direction of

the error is conservative for reactor system thermal calculations. For the,

002 specific heat, the error is a few percent up to the melting point, while'

..

the Zircaloy-4 thermal conductivity is within two standard deviation limits

of'the least square fit given Reference 4 The Zircaloy-4- specific heat,

agrees very well with the data up to 1300k and beyond that in the B-phase. -

4

region is exactly that of the MATPRO model. For the Type 304 SS both thei

*

|

|i

!

| e--

+.. .

,

therm!1 conductivity and spe:ific heat in the temperature range O to 900*C cre '

predicted to witnin a few percent of the values given in the fiSM handbook.

The appropriate conditions associated with the heat transfer regions of con-

dottion or cor.ectior .(Cciburr ;. sclei:e ..' ling (ints) or film bciling(Quinn's modified Sieder-Tate or Groeneveld) are applied at the surface of

the cladding. Transition boiling is not considered. Superheat (Colburn)

is considered as an extension of the film boiling region. Switching

criteria for the various heat transfer regimes are based on the critical

wall temperature for necleate boiling, and the W-3 or B&W-2 critical heat

flux correlation in conjunction with the GE CHF correlation for film boiling -

which are standard options. However, heat transfer coefficients can also be

input by the user as a function of time. For each time step, the total

pressure is assumed to be constant along each flow channel and equal to the

reactor inlet pressure, unless the conservation of momentum option is used

(see next section).

The CADDS program does no't consider dimensional variations-in the gap------

between the fuel pellets and cladding in the heat transfer expressions for ;

.

the fuel pin. Rather, B&W reported that they input a constant but con-

servative value of gap conductance based on sensitivity calculations. The

heat transfer model is the same as that previously approved for CADD.

Core Thermal Hydraulics

Thermal-hydraulic conditions are calculated by simultaneous, solutio,n,,

of the mass and energy conservation equation with the assumption of uniform

pressure, as restricted to one-dimensional constant-area flov passage

applications. Assumption of a single pressure removes the need to solve.

the momentum equation..

S

.

7

.

The assumption of incompressible flow in tne coolant icops is appropriate

since the code is not used for two phase ' low analysis. The slip ootion model

is based on Thom data. The core bypass flow fraction is held constant durinc

the transient. Thermodynamic and transport fluid properties are those approved4

for CADDS.

Reactor Loop.

The single-loop model includes volumes to model the hot leg, steam-

generator primary side, cold leg, reactor core, and core bypass. Inlet, outlet,

and average conditions are maintained for each volume so that transport delays

and specific volumes may be determined. Changes in the reactor system fluid

volune are used in determining surge flow to and from the pressurizer. Except

for the pressurizer and local core channels, all volumes are essumed to be in

subccoled (single-phase) condition. The coolant flowrate at the core inlet

is input to the code as a func. tion of time. The flowrate is modified in the

core and steam generators to account for density changes .jn the channels resulting

from heat transfer. This is done to preserve continunity. Since CADDS does not

solve the momentum equation, local pressures cannot be detennined directly from

CADDS cutput. Local pressures must be determined outside the code by making a

separate accounting of the contribution of reactor coolant pumps, flow losses and

evaluation on reactor system pressures. The surge line flow is calculated

simultaneously with the system pressure, taking into account the surge line~ ~~

quasistatic pressure drop and is based on the system expansion or contraction

assuming that the pressure time derivative is spatially independent. The

pressurizer pressure in computed in tandem.

No critical flow checks are made for the surge line in CADDS. This is ,

acceptable since the surge line and reactor loop are liquid filled for the

transients analyzed. The reactor system is " stiff" and pressure gradients large

enough to produce critical flow are not produced, This was verified by the staff

audits..

k

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

\. .

.

n

Steam Generator

There are two steam generator models, single-node and multi-node. In

tne single-node model,, two differential equations describing the average-

l primary-side coolant enthsipy and the average tube temperature are solved.; i

Heat demand on the secondary esidemust be input as a function of time; this ;1

i .

heat demand is applied to the tube temperature equati^on.'

In the multi-node IE0TSG model, the one-dimensional conservation of

mass, energy, and (secondary only) momentum, and fluid state equations are;

j used on both primary and secondary sides. The feedwater flow and steam -

t

outlet pressure on the secondary side must be input as a function of time.

The steam generator flow on the primary side is calculated in a man-1

ner similar to that for the core, that is, the input flow is modified for

density changes within the steam generator to preserve continunity. On

the two-phase shell side, thermal equalibrium is assumed and slip flow is

3 calculated ~using the Thom corr 51etion. Two phase -friction-mul-tipliers-based- - - -

en the Martinelli-Nelson correlation are used. Heat transfer across the.

tube is computed using a three node radial conduction approximation. On

the primary side the regime is subcooled forced convection (Colburn) while

en the secondary side the regimes are subcooled forced convection (Colburn),4

nucleate boiling (Thom) transition boiling (McDonough-Milich-King), stable

film boiling (Hao-Morgan-Parker-Howard), and superheated forted convection

(Hao, et a t.). Switching ceriteria between the various flow regimes are, ,

i critical wall temperature and the following CHF correlations: BAWL for low

and the Griffith countercurrent flow CHF for extremely low flow conditions.

The switch between the stable film and the transition boiling regimes is.

accomolished by using the maximum heat flux.

;

I4

.

'

9

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

_ - - _ - . _ - - . - _. _ . ___. - . - . _ _. - . _ . . - . . ..

.

e

I*

. Pressurizer |

i In tne nonequilibrium model, tne energy and nass conservation laws are

solved for an all liquid and all vapor region. Included in the energy and

mass balances are the energy transferred to the pressurizer walls, the

energy tranferred across the steam-liquid interface, condensation and

evaporation. The forcing function is the surge rate. Safety valves,

! relief valves, heaters and spray models are 'available to the user. The

I program permits the pressurizer to completely fill with liquid and the

! liquid to escape out the safety valves. The heater model assigns a time -

i

constant to heater surface thermal response, while the spray flow is,

obtained from a table input activated by pressure. The safety valve

; model also includes a relief line pressure drop with an input loss;

! coefficient. For steam relief, however, the flow is obtained from ai

table input. Constant film coefficients between pressurizer vapor and

liquid and the vessel wall'are input. Standard input values for internal

pressurizer heat transfer were verified to be conservative by sensitivity,

studies..

; --.

1

C. Limitations on use of the CADDS code for Transient Analysis

:

15.1.1,15.1.2,15.2.6, and 15.2.7 Decrease in Feedwater! Temperature, Increase in Feedwater Flow, Loss of Non-

emeroency AC Power, and Loss of Normal Feeawater Flow- , ,,,

!' For the events, there are two ways in which CADDS can be used:i

i.i

(1) When the single-node steam generator model is employed, the heat demand*

.

is input as a function of time. (The input is obtained by modeling with|

| a system code such as POWER TRAIN, or using a conservative approximation,

..

+

.

10 .

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

*.

,

cf the heat removal rate as a functicn of tice.) This model is to be'

used for 177 FA Plants.

(2) If an IE0TSG (205 FA Plants) is being analyzed, the CADDS multi-node

steam generator model can be used, provided the steam outlet pressure

on the secondary side is input as a function of time. (This pressure

vs. time data i-s obtained from nodelina with a code such as POWER TRAIN

or by using a conservative approximation.) The multi-node model may

not be utilized for analysis' oE Plants ,which do not have IE0TSEs (177

FA Plants).

(3) The CADDS multi-node steam generator does not contain an auxiliary

feedwater model and therefore carnot ce used for analysis of plant

conditions for which auxiliary feedwater would be actuated without

additional justificationc

When POWER TRAlti'or the like_ is used to provide CADDS input, it is

probable that the CADDS calculations will produce different steam generator

primary inlet conditions than the POWER TRAIN calculatior(s. .If the dis-

crepancies are conservative, this.is acceptable;-otherhise iteration is' ~~

necessary to remove the discrepancies. -

_

s

&

15.3.1 and 15.3.3 Loss of Forced Reactor Cool' ant Flow, andReactor Coolant Pump Rotor Seizure

'*. z

Because CADDS has no flow model, another code must be used ~to provide '

CADDS with reactor flowrate following a loss of pumping power. The B&W PUMP.'

. .

#

|

.

coce ".3s been aprreved for crcvicinc the flow coastcoun. OU'P does not

cor.:ain a naturai circulation mocei however anc cdditional cualified

input would have to be obtainec for natural circulation calculations.

15.4.1,15.4.2,15.4.3, and 15.4.8 Uncontrolled Control RodAssembly Withdrawal (at Startup) Uncontrollec Control RodAssembly Withdrawal at Power, Control Rod Misooeration, andSpectrum of Rod Ejection Accident

,

.

The CADDS neutron point-kinetics approach for handling reactivity

excursions must be demonstrated to meet the Standard Review Plan requirements-

for conservatism regarding power ' distribution and reactivity-ccefficient

weighting factors, etc. on a case by case basis.

15.4.4 Startup of an Inactive Loop

With the CADDS single-loop capability it is possible to analyze the

situation in which, with one ' pump running in each of two loops initially,

the second pump in each of the two loops are started si5ultaneously. How-

ever, CADDS cannot treat asymmetric pump startup events, in which pumping

power is different in each of two loops. B&W has analyzed two pump startups

in plant FSARs. The effect of two pump startup on the reactor system should

bound that of single pump startup.

15.4.6 Chemical and Volume Control System Malfunctions, ,,

CADDS does not contain models to calculate the reactivity effects of

boron transport and dilution. When the code is utilized for analysis of

boron dilution seperate justification must be provided for the reactivity.

vs. time table which is input to account for changes in boron consentration.

15.8. Anticipated Transients Without Scram ( for above listed svents)

Whenever CADDS is used for analyzing any of the foregoing events, but

.

12

- __ _ _ __

.

.

witnout allowance for scram, the foregoing comments for each event remain

applicable.

General Limitations

With respect to all of the above-listed events, the following general

conditions apply to use of CADDS:,

1. Because CADDS requires the primary loop (except for the reactor core

and pressurizer) to be in the subcooled condition, CADDS is not con- ~

sidered acceptable for use whenever the hot leg temperature exceeds

the saturation temperature. Forthermore, due to the assumption of a

constant bypass fraction, CADDS is not acceptable for use for cases

having two phase flow in the core without further specific justification.

2. CADDS'does not consider axial core pressure gradients for core heat

transfer calculations; therefore, for all events in which minimum.

DNBR is an issue, a code such as RADAR must be combined with CADDS

for calculating DNBR.

I

3. Because CADDS has only single-loop capability, it cannot calculate

reactor temperature maldistributions that occur whenever ~one of the

actual two loops is asymmetric with the other. Examples o.f when

asymetry occurs are: decrease in feedwater temperature to only

one steam gene.rator (15.1.1), increase in feedwater flow to only one .

steam generator (15.1.2), loss of feedwater flow to only one steam~

.

generator (12.2.7), coolant flowrate variations different in one

loop than in ,the other (15.3.1,15.3.3, and 15.4.4)._ A different

method must be used to show that such maldistributions are accep-

able (e.g., bounded by acceptable symmetric events).

13_ _ _ _ _ _ _ _ _ _ _ _ _

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - -_

.

4 Reverse flow situations cannot be analyzed with CADDS, due to specific

ma thema tical features in the code.

5 NYPFLO Option 3 stablized flow equations with numerical weighting|

L factor) has not been demonstrated to be reliable and is therefore notacceptable for use. B&W has agreed not to use this option.

|.

6. Reactor system pressures calculated by CADDS should be modified account

for elevation, flow losses and pump effects before being reported in '

FSARS.

7 CADDS does not model emergency core cooling systems and therefore cannot

be used to analyze plant conditions for which these systems would be

actuated.

Code Oualification

In support of their code qualification work B&W has performed sensi-

tivity studies with CADCS and submitted some comparisons with (limited)

measured data. These are summarized in the next section. As part of 'this

evaluation, ANL made audit calculations using Bellefonte plant model with

RELAP4/ MOD 6 for Loss of Feedwater, Control Rod Withdrawal, and Loss of 7brced

Reactor Coolant Flow transients..

B&W Studies.

For the purpose of determining the effect on the results of uncertainties

in various aspects of -the model, B&W performed sensitivity studies of'

fuel-cladding gap conductance on peak. system pressures; time step size on~

the accuracy of. kinetics solutions, and heat transfer coefficient between.

. . . _ ... . -

_ _ - _ _ - - - - - - - - - - - - - - - - - - - - - - - - - - - -

.. _ - , . _ _ _ = - _ - - -. . . - ..

.

*

;

j . liquid and vapor in the pressurizer on reactor system pressure. The results

of these studies were factored into guidance for the user on input selection.

CADDS results were compared with measured data obtained after ai

turbine trip at 72% power at Oconee Unit I and after a generator trip at

96% power at Three Mile Island Unit I. Both transients resulted in a high

e

pressure trip. In the Oconee case, CADDS calculated a peak reactor power

1 level of 75% while the measured value was higher, 82%. However, CADDS

calculated higher peak factor coolant pressure and temperature. In the

Three Mile Island case. CADDS calculated all three parameters to be very close

to the measured data. -

ANL Audit Calculations

1. Loss of Feedwater

As may be seen from Figure 1, ANL obtained similar results to those

l of CADDS inasmuch as the ANL RELAP4/M006 calculations obtained peak-

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

pressures within 10-20 psi of those obtained by B&W. Reactor trip times

within 1 second of those by B&W were ob.tained.

The oscillatory behavior in the CADDS pressure response once the

primary safety valves open, results from the input assumption that the valves

blowdown to a pressure below the opening pressure before closing. ANL assumed

that the valves opened and closed.at the same pressure.i

; 2. Control Rod Withdrawal. . .-

ANL analyzed the transient generated by the withdrawal of the most,

reactive single control rod group. ANL's calculations yielded a faster rate'

: of flux increase than that reported by B&W and therefore reached the flux trip' *

setpoint much earlier (the RELAP4/ MOD 6 calculations typically tripped at,

|

roughly 9.5 s while the B&W tripped at approximately 11 s). Therefore, the

{ peak pressures obtained by ANL were lower than that of B&W. These results

15

, - . . _

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

.

-.

'

may be seen in Figures 2 and 3.,

3. Loss of Forced Reactor Coolant Flow

AHL analyzed a four pump coastcown transient such as due to loss of

offsite power (reactor scram at transient initiation). ANL obtained similar

flow vs time and power vs time behavior. See Figure 4 It should be noted

however that loop flow is not calculated by CADDS but is input from the PUMP

- code which already approved.

Ill. Quality Assurance Inspection .

I

During the course of the revi.ew NRC Region IV conducted programmatic

I inspections of B&W's transient computer codes including CADDS. Based on the

results from these inspections it was concluded that an effective system has

been developed and is being implemented for the orderly development, verification,

and use of computer codes. While some further inspection in specific areas is

necessary no problems which would alter the above conclusion are anticipated.3

(Referece 6)-

IV. Staff Position

j We have reviewed the methods and assumptions described in BAW-10098P <

,

Rev.1, and have concluded, subject to the limitations stated in Section II. C

of this report that the CADDS code contains acceptable models and methods for

transient analysis of the specific events listed in Section II.: A. The adequacy

of the results'obtained with CADDS will be highly dependent on the adequacy of'the.

plant parameters and input into the code. For this reason the staff requires-

that results obtained with the CADDS code be accompanied with appropriate

descriptions and justifications for the code input..

The sta ff concludes that topical report BAW-10098, Rev.1 is an acceptable:

I non-preprietary version of BAW-10098P, Rev.1.t

!

.

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

.

.

.

y Figure 1E'

Pressurizer Pressure2, Loss of Feedwater

-

n y' '\ y' , '

. , ' s ,ie,

P. '/%~

m~; X CADDSoc. v;

5 E Q ANL AuditE -

= *IC Etc) N r*

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

c.

G, ',

, , , , ,- -0.0 20.0 10.0 (,0. 0 B0. 0

TitC Sl.C

- . .

O

O

17

.

.

.

.

|c

rv:- Figure 2

Xe

# Neutron Power

n - .c ^ .._2-

Ewade c-

~

'.0 I. D lb.0 1$.0 2b. 0 2h.0 3b.00

TIME SEC

X CADDS.

UANLAudit.

.

8-N

Figure 3X

MN3 9 t-:-:-:-:C \ Pressurizerf:X X

c.L X Pressuree

-'

c)b

?n o

$ 8-E s

- .-

1|

. . i i . . .|' O.0 5.0 10.0 15.0 20.0 25.0 30.0

TIME SEC

.

Control Rod Assembly|

Withdrawal

.

-

18

.

.

.

.

i

1

I

li

<

'

Figure 4

.

tes .s

N\

"

sc O PC ILCB.

~

\

s: \-

b Aill Audit\u

a~

s- 4e \ - CA DDS - ~ - - _ _ _ _ _ . . . _ _ _ . . _ . _ _ .'

-

NEt1CICL FDI[t ,

ta gN

20 - %~

5 0- _a- _ _

,

i , , ,c ,

C !.C 2.0 3.0 4.C 5.8Time, see

.

Loss of Forced Flow

- . ..

9

9

*

.19_. ._

~

|.-

|.

t

.

References

1. R. H. Stoudt and S. E. Busby, "CADD-Computer Application ofTransients in Water Reactors" BAW-10076P- A, Babcock and Wilcox,December 1984

2. C. D. Morgan, H. C Cheatwood and J. R. Gloudemans, " RADAR-Reactor -

Thermal and Hydraulic Analysis during Reactor Flav Coastdown,"BAW- 10069A, B,ab, cock and Wilcox,0c.tober, .1974.

_

3. "RELAP4/ MOD 6 A Computer Program for Transient Thermal Hydraulic ;

Analysis of Nuclear Reactor and Related Systems - Users Fhnual"CDAP TR OD3, EG&G Idaho, Inc., January 1978).

4. D. L. Hagrman., G. A Geymann and R. E. Mason, "MATPRO-Version 11(Revision 1) A Handbook of Materials Properties for Use in theAnalysis of Light Water Reactor Fuel Rod Behavior," NUREG/CR-0497TREE-1280), Rev.1, R3 and R4, EG&G Idaho, Inc. , February,1980.

,

5. J. C. Spanner, et al . , " Nuclear System lhterials Handbook," Vol .1,Design Data, TID 2666, Hanford Engineering Development -Laboratory,June 1978.

6. Memorandum from U. Potapovs, NRC Region IV, to B. Sheron, NRC, NRR," Programmatic Inspection of POWER TRAIN, CADDS, and TRAP-2 (VITS82-162)," October 21, 1982.

.

. . .-

!

i't

;l

*i

.

,

i

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