ML20083H640

Nonproprietary Amend 1 to Topical Rept NF-1583.03-1, Methods of PRESTO-B Three-Dimensional,BWR Core Simulation Code

ML20083H640
Person / Time
Site:	Brunswick
Issue date:	12/23/1983
From:	Gitnick B CAROLINA POWER & LIGHT CO.
To:
Shared Package
ML19289B611	List: ML20083H548 ML20083H572 ML20083H594 ML20083H640
References
LAP-83-576, NF-1583.03-1-(N, NF-1583.03-1-(NP), NUDOCS 8401130283
Download: ML20083H640 (26)

Text

_

ENCLOSURE 5 TO LAP-83-576 NF-1583.03-1 Nonproprietary Version METHODS OF PREST 0-B A THREE-DIMENSIONAL, BWR CORE SIMULATION CODE AMENDMENT I (NONPROPRIETARY VERSION)

RESPONSE TO NRC QUESTIONS DECEMBER 1983 APPROVED BY: d #2~#d"8d y E. J. Gitnick Principal Engineer - In-Core Analysis CAROLINA POWER & LIGHT COMPANY 411 FAYETTEVILLE STREET MALL RALEIGH, NORTH CAROLINA 27602 .

i 8401130283 840103 PDR ADOCK 05000324 PDR (8688WRM/cfr)

INTRODUCTION This amendment to Topical Report NF-1583.03, " Methods of PRESTO-B:

A Three-Dimensional, BWR Core Simulation Code," is provided in response to requests for additicnal information as conveyed by Enclosure 2 of the letter to Mr. 8. E. Utley, CP&L, f rom Domenic B. Vassallo, NRC Division of Licensing, da';ed November 1,1983.

l 1 (8688WRM/ ccc)

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

s . . _ _

QUESTION 1 (NF-1583.03)

(Cover Page)

Typo: LWR should be BWR

RESPONSE

We regret any inconvenience caused.by this typographical error. The title of the Topical Report NF-1583.03 is: " METHODS OF PRESTO-B, A THREE-DIMENSIONAL, BWR CORE SIMULATION CODE."

i l

l i

l 1

(8688WRM/cfr)

L

QUESTION 2 (NF-1583.03)

(Page 4-2)

How is continuity in a cross-section at the boundary between burnup intervals -

in POLGEN assured? ,

RESPONSE

Continuity in a cross-section at the boundary between burnup ie.tervals is assured in POLGEN by overlapping the burnup ranges by one or more points in exposure. Five burnup points are used to determine the fourth order fit of cross sections as a function of burnup in each subdivision. This method assures continuous cross sections and piece wise continuous cross-section -

derivatives as a function of exposure.

2 (8688WRM/ccc) l

t QUESTION 3 (NF-1583.03)

(Section 4.2.2)

In the tratsient xenon model in PRESTO-B, the effective microscopic absorption cross-section for xenon, o x, is modified by a coef ficient g that is a function of void f raction and xenon density. How are the effects of burnup on og accounted for?

RESPONSE

Typical variation of o and g e in xthe range f rom 0 to 15,000 MWD /TU of burnup are:

'e 3 o:

x "x

This dependence is neglected in the parmaetric representations of g and c, by Eqs. 4.11 and 4.12 of Section 4.2.2.

Note that the microscopic Xenon cross-section is used only to account for deviations in local Xenon concentrations f rom the equilibrium value contained implicitly in the base cross-sections (Section 4.1).

The burnup dependence of the microscopic Xe cross-section is accounted for in the calculation of the base cross-section in RECORD.

s 3 (8688WRM/ccc)

. l QUESTION 4 (NF-1583.03)

(Section 4.2.2)

Provide a typical functional form of the spectral coefficient, g used in the xenon model.

RESPONSE

The spectral coefficient np used in Eq. 4.10 of the Xenon model is evaluated f rom RECORD calculations at 40% void and zero burnup for each nuclear fuel type. ng decreases slightly with burnup and also exhibits some void, dependen .1. The total variance is within within the range of 0-70% void, l 2, 3 0-15,000 MWD /T of burnup. The void an'd burnup dependence in g is neglected in PRESTO since the resulting correction f actor (1 X) 18 always close to unity (typically g

• ion, for equilibriua Xenon concentrat X). l 2, 3 The burnup dependence of the microscopic Xe cross-section is accounted for in the calculation of the base cross-sections in RECORD.

4 (8688WRM/ccc)

QUESTION 5 (NF-1583.03)

(Section 4.4) thder what conditions are the coefficients b2 and b3 determined. For example, are they deterte.ined for the full power equilibrium value of samarium concentration?

RESPONSE

The basic cross-section data input in the form of POLGEN polynomials contain the ef fects of equilibritsa Sa-149. The coefficients b2 aad b3 are used to account f or deviations f rom the equilibrita concentrations and are determined as: -

b2" 2. 3 b =

3 Separate sets of coefficients are determined for use in zero power and at-power cases; the first is generated from CZP - 0% void data and the latter f rom HFP - 40% void data. .

5 (8688WRM/ccc)

4 QUESTION 6 (NF-1583.03)

(Eq. 4.2.1)

Provide more detail on the manner in which the correction term, h 4 , is

' ob tained.

RESPONSE-The correction term h4 of Eq. 4.21 is defined as:

h4= 2,3 i

where AZa II) is the control rod absorption tg obtained f rom single bundle RECORD cakculations, using Eq. 4.20, and AEa is the corresponding quantity derived from 4-bundle 5-group RECORD-11D2 calbulations with one rodded and three unrodded bundles.

AEa (4) is obtained by performing the 5 group to 2 group collapsing based on the hundle wise average group fluxes of the 4-bundle model. Typical values based on data for BWR 8x8 fuel are:

Hot, operating condition h4= 2, 3 cold condition h4=

6 (8688WRM/ccc)

QUESTION 7 (NF-1583.03)

(Section 4.5)

How will the introduction of the GE hybrid control rod affect the calculation of control rod reactivity and burnup?

RESPONSE

The presect design of the GE hybrid control rod includes a few full-length hafnium rods in the high flux region. This design still, however, uses a large number of boron bearing rods for overall reactivity control. At the present time, the RECORL cross-section library is being modified to analyze

, partial or complete hafnium control rods. Once these modifications are complete, a study will be conducted which evaluates the neutronic performance of the hybrid rods compared to present designs. The results of this study will determine whether or not the present PRESTO-B control rod model should be modified to account for reactivity and burnup performance of the GE hybrid control rod.

7 (8688WRM/cf r)

QUESTION 8 (NF-1583.03)

(Eq. 4.2.1)

Provide the functional form of f(E) and typical numbers showing the dependence of f(E) on the 2-group constants, (E}.

RESPONSE

The function f(E) of Eq. 4.21 is defined as:

f(E) = 2,3 Typical numerical values are:

Void fraction a f(E) 0 0.4 2,3 0.70 8

(8688'AM/cf r)

QUESTION 9 (NF-1583.03)

(Section 4.5.3)

Provide additional details of the model that describes control rod history effects on the two-group constants, including the explicit dependence of the group constants on the variables of interest.

RESPONSE

Control rod history effects are modeled by correction terms ALa and avIf2to the corresponding nodal thermal group cross-sections. These co,rection r teris are updated in each burnup step as follows:

Initial values f or f resh fuel ara 2 = 0 AVEf2=0 Control rod adjacent to node during step i -> 1 + 1:

AEi +1 "

where AEg = noda] exposure increment 2, 3 a and b are fuel type dependent coefficients (separate constants for AEa 2

and avEf ) evaluated as illustrated in Fig. 9.1.

2 No control rod adjacent to node during step i -> 1 + 1 :

0 ~

i+1 AI(1) and AI(3) are fuel type dependent constants evaluated f rom RECORD cale tlations with rodded depletion f rom zero to 10,000 MWD /T (Point 1) and

. unrodded depletion from 10,000 to 20,000 MWD /T (Point (1) to Point (3),

Fig. 9.1) .

An evaluation of this model against explicit RECORD calculations for four different control rod histories is shown in Figs. 9.2, 9.3 and 9.4.

9 (8688WRM/cfr)

e FIGURE 9 1 2 33 10

,w -

FIGURE 9.2 C

1. .

.5 -

I I I 1 I I f .

1 I I 2000 10000 12000 EXPOSURE 20000 RH-05 ( MWD /TU )

O C

1.

.5 -

I I l 1 I .

1 I f f I 2000 4000 6000 8000 10000 12000 1400016000 18000 2000d RH-04

~

EXPOSURE

( MWD /TU )

d C

1.

.5 -

f t i I I I .

I *I I I 2000 10000 EXPOSURE 2OOCKT RH-03 ( MWD /TU )

A C

1.

.5 -

I I I I 1 I .

1 I I

, 2000 10000 EXPOSURE- . 20000' RH-02 ( MWD /TU )

Control rod histories used for comparison of PRESTO-B's control history correction with explicit RECORD results.

11

FICURE 9.3 -

n 02 RECORD - runs (J. Haugen )

-1 (10-3 cm ) ---

PRESTO - model 4 -

x RH -02

+ RH -O3 A RH -04 o RH -05 3 -

~

PJ 2 -

7 P ~ ~ ~-

l

/ /%

/ ,/ ~%g /

,/ ,/% ~ s'

/ l

! 1 -

e 7 e-

/

/

/

,//~~~~~~'~

b /

/ ~~____3 c-_ _ _._ _ _

i i i i i i i i i E 2000 4000 6000 8000 10000 12000 14000 16000 18000 2OOOO(MWDITU)'

A fo2 for the rod histories shown in fig.1.

FIGURE 9.4 i

OVE f2 RECORD - runs (J. Haugen )

(30-3 cm-1) j PRESTO - model 8 -

x RH -02

+ RH -03 7 -

A RH-04

o RH-05 t

6 -

~

u 5 -

l 4 -

'/

- /

- ^N~~

3 -

s' ,

' s~~~

/ K / ,

l ,/ 7 ~ ~ ~ ~~/

2 - W /'~~a/

/

/

j

/ / ~~~,_~~~~

1

, ~~~3-- _ _,

l /

i i i i i i E 1 i i 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000 (MWD /TU)"

ape f2 for the rod histories shown in fig.1.

QUESTION 10 (NF-1583.03)

(Section 6.1)

The thermal energy content of the reactor vessel walls is neglected in the reactor vessel energy balance equation. Since the pressure vessel has a thermal time constant of between one and ten minutes, the neglect of the reactor vessel thermal energy leads to inaccuracies of quasistatic changes in reactor conditions over comparable durations of time. Specifically PRESTO-B may not be used for heatup/cooldown calculations.

RESPONSE

The PRFSTO-B heat balance equations have been developed by assuming that the reactor vessel control volume is in equilibrium. Variations from the equilibrium assumption can be introduced by modifying the radiation power term input. Qrad, provided that the time constant is sufficiently long to permit the feedwater system to come to equilibrium. In practice, however, neglecting the vessel heat conduction is slightly conservative for cooldown transient.

The heatup transients need to be addressed on a case-by-case basis.

I i

l l

I l

l I 14 (8688WRM/ccc)

QUESTION 11 (NF-1583.03) l I

(Section 6.2.9) f Provide the bases f or the correlation between the effective Doppler j temperature and the power density used in PRESTO-B.

RESPONSE

The PRESTO-B doppler fuel temperature model applies a linear variation in fuel temperature as a function of loca? power. In addition, PRESTO-B permits a quadratic variation in average fuel temperature as a function of exposure.

This relation is shown in NF-1583.03 Section 6.2.9.

The exposure dependent average fuel temperatxtes are generated in the fuel performance code, by assuming constant operations at the lattice average linear heat generation rate, f rom zero to the lattice physics cut-off exposure. This permits the modeling of average fission gas release, and pellet and clad dynamics.

The best estimate doppler temperature is based on a form presented in Reference 1, this being; TD=aTA + (1 - a) T3 where TD , TA, and Tg are the doppler temperature, fuel pellet average temperature and surface temperature respectively. The interpolation constant is set to a = 0.85. ,

The PRESTO-B doppler fuel temperature input is balanced to account for the following effects;

1) Raported bias between the fuel performance code and measurements. When COMETHE-IIIJ is used, the bias reported in Reference 2 is appropriate,

2) the spatial distribution of resonance absorption within the fuel pellet,

3) the fuel temperature dependency on linear heat generation rate, and,

4) a need to conservatively estimate power feedback for licensing applications.

When COMETHE-IIIJ is employed aa the fuel performance code, the effective i doppler temperature at the lattice average linear heat generation rate is represented as; TDOPPLER = 0.85 TCOMETRE + 43.8 K where TCOMETHE is the COMETHE radially averaged fuel temperature.

15 (8688WRM/cfr)

Figure 11.1 compares the PRESTO-B and best estimate power to doppler fuel temperature relationship for a GE P8x8R lattice at 8 GWD/MTU. The power range of 1 to 15 IGi/f t. spans conditions of normal operation and allows for 20%

overpower on the peak core node, during a power transient. As can be seen, the linear power to toeperature relation is sufficicntly accurate to be used for steady state applications.

Re f erences:

1) S. L. Forkner, et.al. , "Three-Dimensional IRR Core Simulation Methods,"

Tennessee Valley Authority, TVA-TR78-03, June 1,1978.

2) EPRI, " Evaluation and Modification of COMETHE III-J," EPRI-NP-2911, March 1983.

I l

16 (8688WRM/ccc)

1

~

FIGUltE 11.1 4

DOPPLER TEMPERATURE VS.

LINEAR HEAT GENERATION RATE CLHGR)

*=T BEST EST. =T PRESTO-B i

1 i 1500--

i t400- 4 i T 1300 -

! E  :

! H 1200-

• l P  : ,

E 1100- ,

R -

) A 10005

• T  :

• 1 U 900- *

R -

! E 800i *  !

i  :

! 700-

K  :

).

600-i 5005....,..........................,........,....,....,....,....,....,

! 1 2 3 4 5 6 7 8 9 10 tt 12 13 14 15

, LHGR CKW/FT) i ,

I l - _.

QUESTION 12 (NF-1583.03)

(Table 6.1)

Is the set of recommended values for thermal-hydraulic parameters valid for all steady state applications? Were all calculated void distributions i-presented in figures 11.9 through 11.37 obtained with the given set of recommended parameters?

RESPONSE

The set of thermal-hydraulic parameters of Table 6.1 are valid for steady state applications. (Ref. Table 11.3).

All calculated void distributions presented in Fig. 11.9 through 11.37 were obtained with the same set of Thermal Hydraulic Model Parameters as follows:

Ap = 2400.

G1 = 0.22 G2 = 0.2 B =

3 B

• 2

=

vg

=

v2 v0

%=

R O

" 2*3 Rg =

<=

6;==

e The data for the velocity correction of the slip correlation (B;, B2, viand v2) differ slightly from the recommended set. Reanalysis of the FRIGG data, using the parameters of Table 6.1 gave the following result:

Data Set Average Deviation (%) Std. Deviation (%)

Table 6.1 -1.1 t2.2 Original analysis, Figs. 11.9-11.37 0.6 t2.1 Experimental Uncertainty 12.0 The parameters of Table 6.1 give a slight underprediction of void at low flow for the FRIGG data but have been optimized against BWR low power, low flow conditions and are thus recommended for actual BWR analysis.

18 (8688WRM/cfr)

QUESTION 13 (NF-1583.03)

(Eq. 7.6)

How is the formula used to account for 3-D effects near the tip of a control -

rod arrived at? What are the typical errors in eigenvalue and peaking factor introduced through its use?

RESPONSE

Peaking f actors obtained f rom 2-D RECORD calculations for the rodded conditions are generally higher than for the unrodded condition. In PRESTO, such peaking f actors are used to calculate the maximum linear heat generation rate for assemblies adjacent to inserted control blades. Rodded condition

• peaking f actors are used up to a point located one node belcw the end (tip) of the blade. A gradual, linear, transition from rodded to unrodded condition is assumed for the last (one acde long) part of the blade. This is modeled through Eqs. 7.5 and 7.6.

This model has been evaluated by analysis of pin-wise gamma scan data (Hatch I) as illustrated in Figs.13.1 and 13.2. Pins located in the N-N corner were chosen to represent typical peak locations for the rodded condition. The calculated curves were obtained by multiplying the nodal La-140 distributions with N-N corner pin peaking f actors evaluated by Eq. 7.5.

Calculated points obtained without the correction formula (Eq. 7.6) are also shown for comparison. Thest are seen to produce pronounced, unrealistic spikes on the calculated curves.

Thus, the formula used to account for 3-D effects near the tip of a control rod improves the peaking factor calculation. It has no effect on calculated eigenvalues since the nodal 2-group cross-section data are unaffected.

19 (8688WRM/cfr)

FIGURE 13.1 l l La-140 Rel. Oirstr.

CONTROL BLADE (NOTCH 14) 1.5 9

4 k

.  % x .

/ J

? /

10 < r y h'

Measurement l

/- \

N Calculation 1

/

' \

0.5 p

'/

/

/

/

./

BOTTOM 14 16 18 20 22 TOP AXIAL NODE INDEX (PRESTO)

Comparison of measured and calculated axial distributions of single pin La-140, Hatch I, HX373.

$ Denotes calculation without PRESTO control rod tip correction formula.

20

FIGURE 13.2 La-140 Rel. Distr.

I 1.5 CONTROL BLADE (NOTCH 24)

Measurement

/l \

'^#

/ \ -

1.0 '

g e /y / \

Calculation \,

\

/ \

0.5 ,

l ,

/

l/ k

\

BOTTOM 4 6 8 10 TOP AXIAL NODE INDEX (PRESTO) ,

Comparison of measured and calculated axial distribution of single pin La-140. Hatch I, HX393.

9 Denotes calculation without PRESTO control rod tip correction formula. 21

e QUESTION 14 (NF-1583.03)

(Section 8.1)

The f our f uel bundles surrounding a TIP string can be af fected directly by the insertion of one, two, three or four control blades. Do the a-factors defined in (8.1) account f or all these possible control rod configurations?

RESPONSE

The a-factors are generated in RECORD f 'r completely rodded and completely unrodded configurations. Insertion of one, two, three or four control rods (completely rodded) is modeled by applying the rodded a-factors to the rodded bundles and the unrodded a-factors to the unrodded bundles. In practice, the configurations which are of most interest are unrodded, one rod or two diagonally adjacent rods. Comparisons to many 3-D TIP distributions have demonstrated that no additional configuration dependent adjustments are required.

22 (8688WRM/ ace)

QUESTION 15 (NF-1583.03)

It is stated that a detailed fuel perfornance code, such as COMETHE, will be used to provide average fuel tamparature to PRESTO. The COMETHE code has not been approved for use in plant safety analysis. Will this code be submitted .

for NRC review?

RESPONSE

The use of COMETHE by CP&L is historical. At the time of initial development of CP&L BWR steady-state analysis ' capability, COMETHE was an EPRI sponsored code which provided a substantial improvement in predictive capability compar ed with other available codes. EPRI has since transferred their development efforts in this area to FCODE. Ongoing investigations will determine whether COMETHE, POSHO-THERMAL, FCODE or an alternate methodology is suitable for our safety analysis applications. .

The PRESTO-B fuel temperature is used to ?rovide power feedback to the cross-sections, and as such has no other use. The PRESTO-B fuel temperatures are not intended for the evaluation of stored energy or clad performance ano, '

therefore, do not require review for these applications. The need to submit COMETHE for staff review should be viewed in the context of its limited usage by CP&L for PRESTO-B input.

CP&L would prefer not to submit COMETHE foroNRC review. The subject of fuel performance will be addressed in future submittals concerning BWR system transient modeling.

i l

l 4

23 (8688WRM/ccc)

L' l

E!! CLOSURE 6 TO LAP-83-576

?

i

). .

I lt i

F

- __ , en, ,