ML20247B776

Forwards Addl Info Re Topical Rept YAEC-1659 on SIMULATE-3 Validation & Verification

ML20247B776
Person / Time
Site:	Yankee Rowe
Issue date:	09/01/1989
From:	Papanic G YANKEE ATOMIC ELECTRIC CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
BYR-89-138, NUDOCS 8909130143
Download: ML20247B776 (12)

Text

.,+ '

YANKEEATOMICELECTRICCOMPANY '%

7 "',@.g"jo'7?,"g'"

L"Y C 580 Main Street, Bolton, Massachusetts 01740-1398

.Ymx,ts

• N e. September 1, 1989 BYR 89- 138 United States Nuclear Regulatory Commission Document Control Desk Washington, DC 20555

Reference:

(a)- License No. DPR-3 (Docket No. 50-29)

(b) USNRC letter to YAEC, " Request for Additional Information Regarding the Topical Report YAEC-1659 on SIMULATE-3 Validation and Verification", NYR 89-242, dated July 31, 1989.

Subject:

Additional Information on YAEC-1659 SlMULATE-3 Verificat' ion and Validation

Dear Sirs:

Reference (b) requested additional information in support of the subject topical report.- Enclosed are the responses to your questions. We trust that you will find this information satisfactory. However, should you desire further information or clarification, please contact Mr. Richard Cacciapouti of our staff.

Very truly yours, YANKEE ATOMIC ELECTRIC COM ANY Geor- Papanic, Jr.

Senior Project Engineer Licensing GP/gbc Attachments cc: USNRC Region I USNRC Resident Inspector, YNPS Mr. M. W. Hodges, Chief Reactor Systems Branch Division of Engineering & Systems Technology Office of Nuclear Reactor Regulation 00 tI

{

\

t f

ADOCK 05000029

'.8909130143 PDR p 890901 PDC

p

!r j t t

RESPONSES ~TO REQUEST FOR ADDITIONAL INFORMATION CONCERNING YAEC-1659,. SIMULATE-3 VALIDATION AND VERIFICATION' QUESTIONS AND ANSWERS .;

'1) Q. The comparison of the measured and calculated axis 1 offsets' for the McGuire Unit 2 end-of-cycle 3 (EOC-3) xenon ' transient (Figure 3.5) shows that SINULATE-3 significantly overpredicts the observed axial flux difference. This EOC-3 discrepancy has been attributed to the lack of a true equilibrium state-point where the transient began. To further determine the cause of the apparent discrepancy, provide sensitivities of the amplitude of the xenon oscillation and stability index with respect to: (a) flux convergence, (b) source convergence, (c) thermal-hydraulic convergence, (d) doppler feedback, (e) time step size, (f) spatial nesh (12 vs. 24 axial nodes), and (g) Bank-D worth.

Also provide dotatis of the initial core conditions immediately prior to the McGuire xenon transieat, including the core power, core flow and inlet coolant temperature.

A. Figure 3.5 of YAEC-1659 shows measured and calculated data for two McGuire Unit 2 xenon transients in Cycle 3, one at middle-of-cycle (MOC) on the top of the page, and one at end-of-cycle (EOC) on the-bottom of the page. The data for these plots were also provided in YAEC-1659 as Tables 3.8 and 3.9, respectively. Upon review of the figure and tables, it is evident that the plotted data in the EOC figure does not correspond to Table 3.9 and is in error. A revised Figure 3.5 is attached which has the correct data from Tables 3.8 and 3.9 of YAEC-1659.

The revised figure does not show a significant overprediction of the observed axial flux difference. 'The minor differences observed at EOC are partially accountable to the lack of a true equilibrium condition before the transient. Notice in the MOC data that there is effectively no change in measured axial flux difference (-4.1 to

-4.4%) and no rod movement before the transient (from 0 through 19 hours2.199074e-4 days <br />0.00528 hours <br />3.141534e-5 weeks <br />7.2295e-6 months <br />). At EOC, there is a larger measured offset range (-3.9 to

-4.5%) just prior to the transient, accompanied by a small rod withdrawal at 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />. This is a non-equilibrium situation and results in larger differences between measurement and calculation, both before and during the transient. Ac L result, the overall axial offset prediction at EOC is slightly less accurate than that at MOC.

The McGuire data consisted of rod position, axial flux difference and power level on an hourly basis. Power was effectively full power, with a lowest power level of 99.2%. No data on core flow or coolant inlet temperature variations was available and these parameters were assumed to remain constant during the transient. No data prior to the transient was available. Considering the fact that the data is only provided hourly and is incomplete, the comparisons are considered good and represent the relative accuracy achievable when trying to predict actual plant axial offset, with continuous changes in rod motion, power and temperature conditions. This is quite distinct from the use of xenon transients in licensing 1

w ~

z; -

=

1,s:'

{, J

, v

• ~

l . applications,.'which,is discussed-later.-

% ' While the correction.of Figure,3.5: obviates the. remainder of fx the2 ' question, 'an evaluation . of the sensitivities - L of the ' modelt E' '

demonstrates its validity. To increase the impact'of the sensitivity-f to L the : parameters, a more severe ' xenon transient is chosen with 2

greater, changes in.: axial offset than the McGuire transients. This

' transient is 'an MOC case from equilibrium,' full-power ' conditions. The transient consists.of instantaneous rod insertion to.50%.and-power reduction to 150%: power. for 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />, followed by instantaneous rod' 3 . withdrawal and a' return to' 100% power. A ' reference case and

'

sensitivity'. cases were run with the latestiversion of' SIMULATE-3.--

L The.following comments are relevant' concerning the axial offset

. changes resulting';from the changes, in each of the following parameters:

l a) . Flux Convergence.. -

The' acceptance criterion for flux convergence-was tightened by an order of magnitude (0.00005 to 0.000005)'. No change in' axial' offset'was found, b) . Source ' Convergence The acceptance criterion for' source convergence was tightened by 'an order .of magnitude (0.00005- to 0.000005). No change in axial offset was found.

c) Thermal-Hydraulic Convergence - The number of hydraulic iterations was increased by 50% (20 to '30) . No change in axial offset was found.

d) Doppler Feedback - Two cases were run to study . doppler '

effects. In the first,.the core average fuel temperature was increased by 100*K -and, in the second, it was decreased by 100*K from the nominal value of 968*K. Doppler feedback is maintained'.

in all cases and the linear relationship between local. power and local . fuel temp 3rature is unchanged. The effe't of fuel temperature' change is shown in Figure 1. Notice that there'is a change in axial offset and the time dependence of the oscillation.

e) Time Step Size - The time step size was reduced by 50% (1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> to 1/2 hour) . Minor changes in axial offsets were observed, as seen Figure 2. The relatively'small'effect of time step size results from the fact that the average of the beginning and end of step powers is assumed in the SIMULATE-3 calculation.

f) Spatial Mesh - The number of axial nodes was doubled (12 to

24) . Again, only slight changes in axial' offsets were observed, as seen Figure 3. SIMULATE-3 is less sensitive to spatial mesh than the previous generation of nodal codes since there'is a true intranodal flux solution.

g) Bank D Worth - Measured rod worths are well predicted by SIMULATE-3 in YAEC-1659 for McGuire and Farley. Considering the fact that there is little rod insertion in the McGuire 2

s lg

,, lt ! ,

1 l' transients, bank worth is not considered.a significant factor.

,4,

' SIMULATE-3' utilizes,a fourth-order polynomial representation of.the intranodal flux solution in both.the f'ast and thermal groups. As c such~,*its. sensitivities'to-convergence parameters and. geometry are

-less than.that of simple nodal-solutions using. constant nodal flux assumptions.

l f '2) Q. Provide the following iodine, and xenon constants used ' in : the SINtiLATE-3 analysis: fission yields, decay. constants and microscopic absorption cross sections.

A. These parameters are provided in Table' l.

I'

' 3) Q.-What numerical approximations are used in the determination of the xenon . buildup and depletion? Indicate the effect of these 1 approximations on the calculated xenon concentration and power axial offset. ,

A'.' The SIMULATE-3 solution to the xenon buildup and depletion equations makes the following assumptions:

o The iodine production rate and xenon destruction equations are solved analytically using the average of the beginning and end of step core power levels during . the time step interval selected.

o The spatial distribution of the neutron flux used in the

~ buildup and destruction equations is' assumed to be unchanged during the - time interval selected. The end of step flux distribution is used, not an averaged distribution. The impact of this assumption can be seen by varying the time step interval, as shown in Figures 2 and 4, which contain the axial offset and xenon concentrations, respectively, for 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> and 1/2 hour time step cases.

o- 'All iodine decays into xenon (i . e . , iodine has a neutron absorption cross section of zero).

o' Xenon absorption is proportional to the thermal flux (i'. e , an equivalent thermal absorption cross section is defined such that the total xenon absorption rate is preserved.)

4) Q. Rave radial power oscillations been observed or calculated for the two (Figure 3.5) McGuire transients? If so, provide core maps showing typical radial powers.

A. It~ is not known for the McGuire units whether radial power oscillations have been observed or calculated. Westinghouse provides 3

' j.-

,.g 1 ,

the following general statement'in RESAR-3concerning radial power oscillations. These statements have also found to be true .for operation of our Maine Yankee plant:

"The core is ' designed so that diametral and azimuthal oscillations due to spatial xenon effects are self-damping and

- no operator action or control action is required to suppress.

them. The stability to diametral oscillations is so great that this excitation is highly improbable. Convergent azimuthal oscillations can be excited by prohibited motion of individual control rods. Such oscillations are readily observable and alarmed, using the excore long ion chambers. . Indications are also continuously available from the'incore thermocouple and loop temperature measurements. Moveable incore detectors can be activated to provide more detailed . information. In all presently proposed cores these horizontal plane' oscillations are self-damping by virtue of reactivity feedback etfects designed into the core. "

5) Q.Does the EOC-3 (Figure 3. 5) comparison suggest a systematic underestimate of the core axial offset by the measurement system? If so, what is the effect of this underestimate on the core operating margins?

A. Figure 3.5 of YAEC-1659 was found to be in error and has been~

replaced. The new figure shows no systematic error.

6) Q. What is the effect of the SINULATE-3 overprediction of the observed xenon oscillation amplitude (Figure 3.5) on the intended core reload safety analyses?

A. As indicated previously, Figure 3.5 of YAEC-1659 has been replaced.

[

The new figure shows no systematic error.

Free-running xenon oscillations are typically used in the core licensing process. The oscillations are a mechanism to generate a spectrum of power distributions to span the large range of axial offsets required for trip limit and setpoint evaluation. The I oscillations are excited by depletion with heavy rod insertion to a high xenen/ iodine imbalance and correspondingly bottom-peaked power shape. With doppler feedbacks purposely omitted, instantaneous rod withdrawal in this situation leads to a divergent, free-running

~ oscillation. The xenon distributions saved from these oscillations are used in calculations at selected power level and rod insertion

' combinations with appropriate feedback effects. These calculations allow us to infer a conservative total peaking for the particular power level and rod insertion combination as a function of the axial offset indication of the excore detectors.

This process has been demonstrated to produce total peaking factors versus axial offset which are conservative relative to those seen under normal operation in which axial offset is controlled. The licensing requirement of the process is to provide conservative total l

4 I 1

1

l. I H

Y. - _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ b

.,; , 8f g -

peaking'as'a function-of axial offset. Following the axial.~ offset i: history of1 a particular xenon' oscillationlin time, with changing power.' level,-temperature and rodded conditions, is a more. difficult test of the core modelling and provides an added level of confidence in the model accuracy.

j ~ 7) Q. The xenon transients presented in Figure 3.5 are relatively mild when

1. compared to design transients with respect to magnitude and rate of bank reactivity insertion. How would SINULATE-3 he ' expected' to compare with measurement for the stronger design transients?

L A. As ' indicated previously, Figure 3.5 of YAEC-1659 has been ~ replaced.

The new. figure shows no systematic error. Since this is 'the case, it is expected that there is no significant systematic error for stronger transients.

L .

-Licensing. applications, however, rely upon a series of H

benchmarks, comparisons and conservative assumptions'to assure that-applications to conditions under which measurements are not available will not be affected by minor trends in predictive capability. For instance, for bank withdrawal transients, the following factors

,. provide assurance-that the licensing analysis is conservative:

l o Bank . worths are measured at the beginning of ~each cycle 'to assure that both the rate of reactivity insertion with rod motion - and the total reactivity insertion in the licensing analysis is conservative.

o The maximum rate of bank withdrawal assumed in the licensing l analysis is conservative relative to that achievable by the hardware at the plant.

o Reactivity coefficients for moderator temperature, doppler and power are measured at the beginning of each cycle and during the cycle to assure that the values assumed in the licensing analysis'are conservative.

o .The. total peaking associated with bank withdrawal is the most adverse of the possible axial offset conditions, accounting for axial offset'and excore system measurement uncertainties. The total peaking is conservative based on its generation from a free-running xenon oscillation technique, as described above.

o Core parameters are conservatively represented with appropriate uncertainties in initial core conditions and kinetics parameters.

o Benchmarking to tests and higher order analysis is performed to demonstrate that the core power response modelling used in the licensing calculations is conservative.

5

a

. r. :

.4 4 .

Table 1 3: Yields,' Decay Constants and Microscopic Absorption Cross Sections

.for Iodine, Xenon, Promethium and Samarium

---

Yields --------------------

Nuclide Iodine ' Xenon ' Promethium U-235 0.06288 0.00248 0.01080 U-236 0.06133: 0.00172 0.01347.

U-238. 0.06934 0.00027 0.01618 Pu-239 t 0.06400 0.01180 0.01223 Pu-240 0.06514 0.00772 0.01330 Pu-241 0.07001 0.00232 0.01473 Pu-242- 0.07036 0.00236 0.01586

---

Decay Constant (seconds'2) -------

Iodine Xenon Promethium 0.00002924 0.00002100 0.000003626 Microscopic Thermal Absorption Cross Section (10' barns) at 0 MWD /MT Enrichment (w/o U-235) Xenon Samarium 1.60- 15'4.347 4.61859 2.40 .43.024 4.32364 3.10 136.137 4.14064 6

FIGURE 3.5 McGUIRE UNIT 2 CYCLE 3 XENON TRANSIENT MOC Transient 20 220 18 + ^ o A AA AA A ^^^

16 -

~

A^^4 A 000 0 A

- 200

_ AAAA - 180 10 - - 160 g 8-g 6- - 140 @

g 4' o SIMULATE-3 Data '

a.2- O Measured Data - 120 [

g ,0 - _

A Bank D Position -

- 100 0 E -4 .  :

^

Y

- 80 g

- 60

' ~

- 40

- 20

-20 , , , , , , , , , , , , , , , , , , , , , O O 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 Titne (hrs)

EOC Transient 20 220 18 " - A ^ AA ^^^^^^ A AAA A - 200 16 -

2[ AA - 180 10 - - 160 v 8-g f 6-4~ o SIMUIATE-3 Data

- 140 $

a 2-O-

O Measured Data - 120 [

g A Bank D Position - 100 0 C. 4 .

- n -

'it .g _

- 80 f

- 60

• -~~"

_ 40

. - 20

-20 , , , , , , , , , , , , , , , , , , , , , 0 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 Time (hrs)

,llJg

.e 0 -

2 -

d, " 8 1

-7 ~

~

.E 6 1

t r 4 1

nt i

e ne si n s r nac a :s 2 )

s Tr Tf n oE ot ef

) 1 h

t r

+

nnk .

m 1 eic /

eXe a b f d

w rCimd 0 G

i u

gO T e FMsFv Rt ele e

r

) 1 0

0 5

7

(

Wsp e m

Pf fp Oo g H 8 T i

lal D ica px i yA 86 8 h .

T 60 68 91

=

UU= =8 U

l W 6 FF F TT ep p sm a

T m

m .

CeT T e eh sgw 4 aio BHL oOA b 2

- }  : . ,. . ' :  : : _  :  : - ~ 0 3 2 1 0 1 2 3 4 5 6 7

1. 4 0 0 0 0 0- 0- 0- 0- 0- _

.i9g< .

ll!

0 2

~

2 8

d 1 F 6 1

4 1

is t

n e n t

ie

/ .

n s a n r at s 2 T r Tec 1

)'

n f o oft n nE .

2e ie a eX ez rC u imS i

, 0 igO Tp )j 1 FM se t R t vS .

ee W

P fsmf i g la OT 8

c ipia x -

yA s T s pe pt t

eS 6 S e em i imT .

Tr u r

u o oH 4 H2/

1 1 oa .

k 2

; - , i - ,: ' : : : : _

_ _ ~

-  : ~ 0 3 2 1 0 1 2 3 4 s. 6 7 0 0 0 0 0- 0- 4 0- 0-4 4 gE9Q<

r  ! l

0 2

i 8

4 1 C -

6 1

f 4 1

t n nt ies ie n s f a

r a n f' 2 r s T Tt n oec

) 1 o f n tn f .

3e iE eX eh uC Tms r 0 e

igO M ) 1 FM sl va .

R eaitt W sp P f f

S g  ;

8 a O l

l c

ipiax _ .

y A T -

lyy 6 l

l l iiaa

(

xx AA .

_ ee dd oo NN 4 24 1 2 oa .

( . 2

, -[

- : -[  : :_:~0 3 1 ..,:'.':2 0 1 3 4 5 6 7

0

• 0 O 0 -

0- 0 0-0- 0-

_ %a9s.N

l Y c.

l E

i S

y .

)

l) -

\)

E w E.E

.9 m -

EF

@Om J N -

H .5 p j 88 :

-c E *uJ g

+

v*Fe p X m .u a >W S O C n.

E .o 3 8 Q w.

W EE *F

_- 8 F :r hO r >

a.o n .

Noc E.g o

i >

h A m F$

-F Y '

4 M

,- e OO

, , , , , , .- . . . , , o i M

• $

• S 6 a (0t x) uopellueouoo uoueX j

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