ML20044E315
| ML20044E315 | |
| Person / Time | |
|---|---|
| Issue date: | 06/04/1992 |
| From: | Rosalyn Jones Office of Nuclear Reactor Regulation |
| To: | Newton R WESTINGHOUSE OPERATING PLANTS OWNERS GROUP, WISCONSIN ELECTRIC POWER CO. |
| References | |
| NUDOCS 9305240182 | |
| Download: ML20044E315 (24) | |
Text
-
.j
)
June 4, 1992 j
1 I
Roger Newton Westinghouse Owners Group l
c/o Wisconsin Electric Power Co.
P. O. Box 2046 231 W. Michigan St.
Milwaukee, Wisconsin 53201
Dear Mr. Newton:
i
SUBJECT:
POSSIBLE RECRITICALITY FOLLOWING A LARGE BREAK LOCA i
As a follow up to my staff's recent conversation with you on the subject of l
possible recriticality following a large break LOCA, we are enclosing a copy of my staff's evaluation on this subject. This evaluation includes the l
Brookhaven National Laboratory report on Post-LOCA Reflood Criticality l
Calculations.
We look forward to hearing from you further with regard to your plans to address this question.
If you need any additional information, please do not i
hesitate to contact us.
i Sincerely, j
i
/s/
i Robert C. Jones, Chief i
Reactor Systems Branch i
Division of Systems Technology j
Enclosure:
i As stated l'
DISTRIBUTION i
Central Files I
SRXB R/F i
AThadani RJones i
TCollins MChatterton MChatterton R/F y
1 SRXB: DST sk k SRXB: DST XB: DST 4
MCHATTERTON:jh TCOLjNS RJ0NE dU blYi? z-b 4ltt
&l4 fV
'f
\\
9305240182 920604
('
l TOPRP EMVW
{DR
~ _. _.....,, _.. _.. -, _
(
ENCLOSURE 1 3
I INTRODUCTION i
1 4
4 Licensing basis LOCA analyses are performed with the objective of conservatively estimating the fuel and cladding temperature.
As a result, the analysis assumptions deliberately minimize the water inventory in the reactor vessel throughout the transient.
Similarly, conservatively high fuel temperatures are used to maximize boil off and thus retard the reflood rate.
While the assumptions are conservative with respect to peak cladding temperature calculations, they are not conservative with l
respect to the potential for a return to criticality.
Minimizing f
water retained in the vessel following blowdown minimizes the amount of low boron concentration water involved in the core reflood.
Likewise, high boil off rates and slow reflood times assure that reflood water is highly concentrated in boron and that reactivity is added slowly.
t A study was therefore undertaken to assess the safety significance of this problem.
As is discussed below, we conclude that PWR characteristics are such that a significant safety concern is not expected, but the possibility of short term recriticality cannot be completely ruled out.
DISCUSSION The strategy for the problem was as follows:
First, static calculations were performed to establish a very conservative baseline estimate of reactivity (K-eff).
Second, key assumptions in the calculations were compared to those conditions expected during a LOCA (using large break LOCA experimental information j
whenever possible).
Third, an estimate was then made of the j
1 i
2 conservatism or non-conservatism in the assumptions (relative to criticality).
This served as the basis for evaluating the overall safety significance of the problem.
This approach was used because performance of transient criticality calculations for a large break LOCA were not considered practical from either a cost or schedule viewpoint, or necessary for an initial study.
Physics calculations were performed at Brookhaven National Laboratory.
Their report is attached as Appendix A.
The analyses were based on a 193-assembly PWR core model, and a "best estimate" TRAC-PF1/ MOD 1 calculation performed by INEL for the large-break in the cold-leg piping system between the emergency core coolant injection nozzle and the reactor vessel in a 4-loop Westinghouse RESAR-35 PWR (Reference 1).
Using the thermal-hydraulic parameters from the TRAC-PF1/ MOD 1 calculation, a
" snapshot" core multiplication factor (keff) for each of five points in the transient was calculated.
These calculations were done with the MCNP4 Monte Carlo code.
The five core statepoints considered in the analyses were:
Initial critical configuration (t=0.0 sec.)
- Core fully voided (t=32.85 sec.)
First major peak in " Core Liquid Mass Fraction" plot (t=56 sec.)
Second major peak in " Core Liquid Mass Fraction" plot (t=100 sec.)
End of TRAC calculation for transient (t=150 sec.)
l 3
l i
The results for the core multiplication factor (keff) are as follows; t=0.0 sec keff = 1.0 l,
t=32.85 sec keff = 0.49 t=56 sec keff = 1.012 t=100 sec keff = 1.023 f
t=150 sec keff = 1.016-P These very conservative values were used as the baseline numbers l
for our safety evaluation.
l The key analysis assumptions and approximations which were l
l examined for their impact are:
L 1.
The control rods remained completely withdrawn from the Core.
2.
Boron concentration prior to LOCA was used throughout the calculations i.e. no ECCS water mixes with the reactor vessel water.
3.
The calculations were " snapshots" rather than transient reactivity calculations.
4.
No Xenon was assumed present in the core.
5.
The cross sections used were for fuel temperatures of 300 K.
6.
Water properties of the radial and azimuthal nodes were averaged.
l t
l 4
l The effects of each of these assumptions is addressed below.
1.
Control Roo Insertion l
The BNL study assumed that all the control rods remained completely withdrawn from the core.
If only a small portion of the control rods (about 40% depending on the core under I
consideration) did insert into the core, the excess reactivity would be e3iminated and a return to criticality precluded.
However, structural analyses (Reference 2) for the largest double ended hot leg break indicate that control rod insertion, for some of the rods, may be impeded by the hydrodynamic forces associated with the blowdown.
But since it is also unlikely that all of the l
rods will fail to insert, the BNL assumption is considered to be a significant conservatism.
2.
Boron concentration l
The boron concentration was not changed throughout the calculations.
All the water entering the core was assumed to be at the boron concentration which was present before the LOCA.
No account was taken of the fact that the ECCS water is at a much l
higher baron concentration.
The change in the boron concentration would be determined by the amount of water left in the lower plenum after blowdown and the mixing of it and the ECCS vater.
l If the boron concentration is increased due to the ECCS vater mixing with the initial concentration water, the excess reactivity can be eliminated very quickly.
For example, the ECCS water always has a boron concentration at least 800 ppm greater than that of the RCS and boron worth values for PWRs range from 8 to 10 pc=/ ppm.
Assuming a mixing factor of.25 and a boron worth of 8 pcm/ ppm, the change in keff would be.016 thus eliminating l
[
i
5 i
nearly all the excess reactivity in the worst case calculated.
A mixing factor of.25 would be comparable to the inwar plenum after blowdown being 75% full of lower Boron concentration RCS
- water, ECCS water injecting and mixing uniformly until the lower plenum is filled, and once the lower plenum is filled the concentration staying the same as the core fills.
In fact the average concentration would be constantly increasing as the higher concentration water continues to enter the lower plenum and raise the average concentration.
Hence even if recriticality is achieved, it is likely to be maintained for only a short period of time.
However, the amount of water remaining in the vessel after blowdown is not known.
Both test data and model data vary greatly, ranging from lower plenum empty to lower plenum approximately 90% full at the end of blowdown.
Various assessments of lower plenum liquid inventory following large cold leg breaks indicate the residual liquid volume fraction below the core region is between 15 and 75%. (Ref 2).
The INEL "best estimate" TRAC calculations used in the BNL analyses show the lower plenum empty at approximately 12 seconds and refilled presumably with ECCS water by about 35 seconds.
There are also questions about the "line water".
In some plants there is a significant volume of water in the HPSI, LPSI and SIS Header plus the SI tank lines (up to the equivalent of 90% of the lower plenum volume).
While the "line water" should be at the ECCS boron concentration, it may be at a lower concentration due to RCS water inleakage.
It is unclear how much "line water" is added to the residual water after blowdown and what its boron concentration is.
In addition, it is unclear if this line water has been adequately considered in the tests and calculations of blowdown.
6 i
Finally, it should also be noted that the BNL study was performed for BOC conditions.
Later in the cycle the difference between the P.CS bere concentration and that of the ECCS water will be greater.
This will require less mixing to produce the same reactivity effect.
However the reactivity effect due to cooling l
the water from the operating temperature to the post-LOCA temperature will be greater because of the more negative moderator temperature coefficient.
Therefore it is not clear j
which point in the cycle is most important.
3.
No Transient Reactivity Feedback i
This study consisted of " snapshots" rather than transient reactivity calculations.
As the core approached criticality heat would be produced and the negative reactivity feedbacks would greatly alter the process.
These include the doppler fuel i
temperature, moderator temperature, and moderator density / void feedbacks.
Once criticality is approached the thermal-hydraulic parameters from the INEL calculation are not representative since that calculation included only decay heat.
This constitutes a very significant conservatism in the calculation.
4.
Cross Sections vs. Fuel Temperature The calculations were performed using cross sections for fuel i
temperature of 300 K.
Since the fuel temperatures are above 422 K in all cases, this introduces a conservatism due to doppler broadening.
BNL has stated that the effect of this will be approximately -0.008, -0.005 and -0.005 delta-k for the 56, 100, and 150 second cases respectively.
The calculated Keff with the fuel temperature corrections were 1.004 to 1.018 for the three points 56 seconds, 100 seconds and 150 seconds into the accident.
This is therefore a small conservatism in the calculations.
7 5.
Xenon considerations The calculations were d:r.e tith..o Xenon present.
BNL has stated that including Xenon would simply lower the boron concentration and would not affect the results.
The effect of this assumption was handled by considering the difference between the initial RCS boron concentration and the ECCS boron concentration instead of considering the actual concentrations.
Hence it not believed to introduce any significant conservatism or non-conservatism in the calculations.
6.
Averacina of Water Properties The water properties of the radial and azimuthal nodes of the INEL results were averaged.
The averaging over the radial and azimuthal nodes may have resulted in a more reactive core.
This assessment is substantiated by the results obtained.
The case at 150 seconds had a lower keff than the one at 100 seconds.
The total water volume was greater but the distribution of voids was more irregular for the case at 150 seconds.
Thus it appears that a case with a more uniform distribution of voids is more reactive.
The uncertainty in this regard however is too large to ascribe any conservatism to this assumption.
CONCLUSIONS The calculations performed by BNL provide a conservative baseline for assessing the potential for a return to critical following a large break LOCA.
Our assessment indicates that although, there is a*
,6snt neutron absorption capability in both the control rods and the borated safety injection system, the rate of i
negative reactivity addition from these sources is uncertain particularly during the early reflood stages of a large break LOCA.
This uncertainty in negative reactivity insertion rate I
8 stems primarily from uncertainty in the amount of initial primary system water which remains in the reactor vessel after blowdown, and uncertainty in the number of control rods which can be inserted.
It is cur belief that the likclihccd _I a large break LOCA in conjunction with failure of a large number of control rods to insert is small. We also recognize that inherent negative feedbacks would result if criticality did occur and that the safety injection boron nill limit any return to criticality to the short term.
For these reasons we do not believe that a significant safety concern exists.
However, because of the uncertainties in important parameters noted above, a short term return to criticality cannot be excluded.
Therefore, it is our recommendation that the Owners Groups be contacted.
They could play a significant role in resolving this generic concern by demonstrating that there would be no return to criticality or that the consequences of such a return to criticality are insignificant to plant safety.
This could be demonstrated by showing that the control rods will insert prior to reflood, that the thermal-hydraulics processes are such that boron concentration during reflood is sufficient to prevent recriticality, or that the effects of recriticality are not i
safety significant.
REFERENCES 1.
Quantifying Reactor Safety Margins, Application of Code Scaling, Applicability, and Uncertainty Evaluation Methodology to a Large Break, Loss-of-coolant Accident.
2.
An Assessment of the Return to Criticality Following a Large Break LOCA at the Palisades nuclear Plant, Combustion Engineering, February 1991.
I
J l
J APPCOIX A
)
l l
BROOKHAVEN NATIONAL LABORATORY j
i MEMORANDUM DATE:
April 15.1992 TO:
J.F. Carew q
j
.t..h-T N
FRO 51:
A.L. Aronson, M.J. Geiger, M. Todosow
SUBJECT:
Post-LBLOCA Reflood Criticality Calculations 1
1.0 INTRODUCTION
l The light water moderator / coolant inventory and density distribution in a typical PWR core and vessel vary dramatically during the course of a post LOCA reflood transient. The objective of the present study was to evaluate the core multiplication factor (k,ff) at several j
statepoints in such a transient to determine whether re-criticality might occur.
f I
t The analyses are based on a 193-assembly PWR core model similar to the Zion Unit 2.
Cycle 1 core developed by BNL. and a "best estimate" TRAC-PF1/ MODI calculation performed j
by INEL for the large-break. loss-of-coolant accident transient following the double-ended guillotine break in the cold-leg piping system between the emergency core coolant mjecuon j
1 j
nozzle and the reactor vessel in a 4-loop Westinghouse RESAR-3S PWR (NUREG/CR-5249.
1 EGG-2552).
l
)
1 2.0 CALCULATIONAL METHODOLOGY The multiplication factor (k r) for the various statepoints considered in this study was ef determined with the MCNP4 Monte Carlo code. Point-wise cross-sections based on ENDF/B-V were employed along with temperature dependent S(cr,$) thermal scattering kernel data for hydrogen bound in light water. The temperatures at which Doppler broadened cross-section or J
~
thermal scattering data were available are given in Table-l.
The geometry wn modeled in three-dimensions with fuel rods burnable poison rods and guide tubes explicitly represented (Figures 1-3). The top and bottom axial reflectors were approximated by homogenized zones (5-zones /- 15 cm. for the lower reflectors, 3-zones / ~ 20 cm. for the upper reflectors) based on the data in EPRI NP-1232. The radial reflector representation considered the stainless steel core baffle, core barrel and thermal shield along with associated water. In this way the CIAE1 three-dimensional geometry of the cylindrical fuel rods, rectangular fuel assembly lay down, and upper and lower reflectors of the core was preserved in the MCNP4 neutronics calculations. Vacuum boundary conditions were assumed to apply on the exterior surfaces.
The light water moderator / coolant propenies associated with cash unit cell (fuel rod, BPR, guide tube) throughout the core and reflectors were based on the TRAC modelling for the core / vessel. In TRAC, the inside of the vessel was represented by 3 radial,15 axial and 4 azimuthal zones. The core was modelled with 5 axial levels, and the equivalent core was contained within the outer radius of the second radial zone which extended to - the outer surface of the core barrel. The MCNP4 modelling preserved the essential characteristics of the l
coolant density distribution within the core and reflectors as far as their expected impact on the core multiplication factor were concerned.
Consequently, 5 axial coolant zones were accommodated in the MCNP model corresponding to the discretization in the TRAC-PFI/ MODI representation (Figure-4). Determination of the appropriate number densities for the light water for any given spatial region, however, required averaging of the TRAC results since incorporation of the full set of 5 axial x 4 azimuthal x 2 radial TRAC data was not feasible within the time frame available for the study. In addition, differences between the TRAC and MCNP4 models for the core (e.g. the TRAC cylindrical vs the MCNP4 rectangular geometry) would have required funher approximations, as well as necessitating a much more costly and l
complex full-core MCNP4 model, as opposed to the quarter-core representation employed. The l
full-core model would require 20 unique assembly-types vs the three assembly-types used in the present model, and 40 unique moderator densities vs the five densities used to represent the active core in the present model. These moderator densities must be assigned individually to unit
cells within each speci5e assembly-type. Therefore, while this level of detail is possible, the resulting modelis much more complex. and the corresponding running times for each calculation wnuld increase signincantly beyond the several hours /statepoint *eady required. It should also be noted, for example, that the one-sigma deviation of the TRAC region-dependent moderator density differed from the planar average by -5% for the t=100 second case. Therefore, the following approach was followed for each statepoint:
r the average water density for a given axial level within the active core was obtained by combining the void fraction, and liquid and vapor densities for each planar node from the TRAC results, and then appropriately averaging over the 8 nodes (4-azimuthal x 2-radial) within a plane.
the average water densities for the axial redectors were obtained in the same way by considering the TRAC results for the two axial levels immediately adjacent to the active core.
the average water density in the radial reDectors was taken as equal to that in the lower axial redector.
Five core statepoints were considered in the analyses:
Initial critical conneuration (t=0.0 seeJ: The core is assumed to be at operating conditions with cross-sections and S(cr S) corresponding to the elevated temperatures given in Table-1 (i.e.. Tfavg)=988 K. Tyavg)=600 K.
T,,(avg)=560 K T (avg)=600 K), a soluble boron concentration of 1300 ppm.
fresh fuel, no xenon and all control rods fully withdrawn.
Core fullv voided (t=32.85 see): The soluble boron concentration is assumed to remain at 1300 ppm corresponding to the initial critical state. The average fuel and clad temperatures from the TRAC calculation are 582 K and 566 K.
respectively. The liquid vapor temperature is ~500 K.
1 I
First maior peak in " Core Liouid Mass Fraction" pjot (t=56 see 1: The soluble boron concentration is assumed to remain at 1300 ppm corresponding to the initial i
entical state. The average fuel and clad temperatures from the TRAC calculation are 496 K and 442 K, respecuvely. The liquid temperature is ~400 K.
Second maior Deak in " Core Licuid Mass Fraction" niot (t = 100 sec): The soluble boron concentration is assumed to remain at 1300 ppm corresponding to the initial critical state. The average fuel and clad temperatures from the TRAC calculation are 423K and 40SK, respectively. The liquid temperature is ~400K.
End of TRAC calculation for transient (t=150 sec.):
The soluble boron concentration is assumed to remain at 1300 ppm corresponding to the initial I
critical state. The average fuel and clad temperatures from the TRAC calculation l
are 422 K and 409 K. respectively. The liquid temperature is -400 K.
The MCNP4 calculations for all the transient statepoints (i.e., t > 0.0 see) were based on l
300 K cross-sections, thereby neglecting the negative reactivity due to Doppler broadening of the resonances (primarily those of U-235 and U-238 which have the largest magnitude and i
experience the greatest temperature change relative to the assumed 300 K). This approach was taken because generating temperature dependent nuclear data for resonance isotopes for use with MCNP is an involved and time consuming process. Nuclear data at 300 K is available in the base MCNP library provided by RSIC. In order to minimize the data generation effort for this study, additional data were produced only for the initial operating conditions. These two temperature dependent sets, therefore, bracketed the temperature range expected in the transient.
An ad-hoc correction was determined to account for this difference between the actual and MCNP4 library fuel temperatures. The S(cr,5) thermal scattering data were chosen at the closest available library temperature.
t j
3.0 RESULTS As noted in the previous section, the MCNP4 transient results must be adjusted to account for the difference between the actual statepoint and the 300 K MCNP4 library fuel temperature. The k,g correction is given by
I
~
i l
r (Ak/k) 6T 6 k,ti aT
- - (4. O x 10-5 a R/k/K) - ( T-3 0 0K),
where T is the statepoint fuel temperature and the Zion-2, Cycle 1 Doppler coefficient has been used. The k,g corrections for the transient statepoints are --0.0ll, - 0.008 - 0.005. and
- 0.005 Ak,g, for t=33,56,100 and 150 seconds, respectively. The difference between the j
actual fuel temperature and the 300 K library temperature is larger for the earlier statepoints and, consequently, the ak,g corrections are larger. The uncertainty in this correction is i
~
expected to be 20%', resulting in a maximum k,g uncensinty of 0.2% which is considered
~
l to be within the overall uncertainty of the calculations.
The results for the core multiplication factor (k,g) are given in Figures 5-7 for the l
l transient statepoints at t=56,100 and 150 seconds, respectively. The core k,g is presented as a function of the boron concentration which is assumed to increase as a result of ECCS boron injecuon. The reduction in core k,g resulting from ECCS boron injection is based on typied linear differential boron worths of -10 pcm/ ppm and -8 pcm/ ppm *. The estimated one-sigma confidence interval on k,g is 0.2%Ak/k. The core multiplication for the highly voided case at t=33 seconds was calculated to be k,g = 0.49 0.019 (one-sigma). Comparison of Figures 3 = 1300 ppm indicates that the k,g or the t = 100 second statepoint is slightly f
6 and 7 at C higher (by -0.7%Ak/k) than for the t=150 second statepoint. This is noteworthy since the moderator and fuel temperatures for the two statepoints are essentially identical. This difference is due to a comoination of (1) the MCNP4 calculational statistics ( 0.2 % Ak/k) and (2) the differences between the distribution of the moderator density for the two statepoints. The moderator density was relatively smooth in the 100 second case and varied axially from -0.95 to 0.75 gm/cc while in the 150 second case there was a substantial variation in the axial moderator density from ~0.95 to 0.20 gm/cc over the axial height of the core.
l t
~
- Typical Doppler coefficient uncertainties are 15 %.
I i
At reduced critical boron concentrations, the moderator temperature coefficient is more negative and the insened positive cooldown reactivity is larger. Reduced boron concentrations occur when the enre is les reactive as a remh of venon buildup, fuel burnup, reduced enrichment and/or the addition of bumable poisons. In order to estimate the effect of reduced boron on the transient k,ff, MCNP4 calculations were also performed at an initial boron concentration of Cg = 900 ppm for the t=0 and t = 150 second transient statepoints. The transient increase in core multiplication increased from ~0.010 for the C3 = 1300 ppm case to ~0.012 for the Cg = 900 ppm case. The transient.ik,ff s plotted versus initial critical i
boron in Figure - 8. For this 150 second case, therefore, the insened reactivity did not increase substantially as the boron concentration - C was reduced by 400 pprn.
3 4.0 DISCUSSION The initial calculations for evaluating the potential for recriticality during the reflood i
stage of a large break LOCA have been completed. The calculations were performed with the MCNP4 Monte Carlo program using a detailed three-dimensional model based on a 193-assembly PWR core similar to Zion-2 Cycle-l. The thermal-hydraulic input data is based on an INEL best-estimate TRAC-PF1/ MODI large break LOCA calculation. This TRAC i
calculation assumes the core does not return to critical and, consequently, the MChT4 calculations reponed here do not include the effects of changes in the thermal-hydraulic state resulting from neutronic feedback. The core multiplication was determined at several statepoints during the refill and reflood sis:ges of the LOCA. These steady-state calculations indicate that.
if the negative reactivity resulting from the ECCS boron injection is neglected, the reactor will return to a critical (k,ff= 1) state as a result of the core cooldown and reflood.
l After a critical state is reached, the positive reactivity insenion will be panially balanced l
by several negative reactivity components as the core power increases. These include the doppler fuel temperature, moderator temperature, and moderator density / void feedbacks. In addition, the ECCS boron injection will provide a strong shutdown reactivity. The calculations carried out in the present study provide a conservative estimate of the insened reactivity by neglecting this time-dependent neutronic feedback, which was not included in the INEL LOCA analysis. This negative feedback will result in a substantial reduction in the conservative
t reactivity estimate provided here, and depends on the specine details of the redood and ECCS boron injection.
I The total reactivity insenion sincluding both positive c.nd negative comnnnents) and resulting increase in core power following the LOCA may be obtained by performing a time-dependent coupled neutronic/ thermal-hydraulic calculation. This calculation should include the t
time-dependent changes in moderator and fuel temperature, moderator density and ECCS boron injection. The MCNP4 analysis developed in the present study may be used to determine the input reactivity coefficients required to conven these changes into core reactivity, as well as the cooldown and renood reactivity.
JFC:jg cc:
R. A. Bari, BNL M. S. Chattenon, NRC T. E. Collins, hTC H. E. Polk, hTC C. L. Snead, BNL G. J.
Van Tuyle, BNL i
i
i 1
i l
Table-1 AVAILABLE TEMPERATURE DEPENDENT CROSS SECTION AND THERMAL SCATTERING DATA l
l Isotope Cross Section S(a,S) l l
Temperature (*K)
Temperature (*K) l U-235 988,300 l
U-238 988,300 l
O-16 300 H-1 300 600,500,400.300 l
zr 600.300 t
Cr 560,300 i
Fe 560, 300 Ni 560,300 l
l B-10 300 l
l B-11 300 l
l t
l
f 000000000000000 0 0 0 0 0 0,0,0 0 0 0 0 0 0 0 0000000000000 00 00000 000000 00 0o 000000000000000 o000000000 00 000 i
o0000 0 0000000 00 000000000000000 0000000000 0 0 000 000000000 000 000 o000000000 00000 00000000000000o o00000000000000 0000000000000 00 00000000 0000000
_._._.r._.
E E,
a
~l i
i id 5
m
.w n
h3
-:a si N!
,l l
000000000000000
(
0 0 0 0 0 0.0 0 0 0.0 0 000
(
oo@;oo@ooo@oo
@oo <!
i o0000000 0 00 0000
(
oooo@oooo o @oo o o ('
00@o 0 0 0,0 00 00 0 0 0 (!
000 0 00 000 0 0000 0 (l t
o000 000G0,00000 0 (!
0000000 00 0 0 000 0 (l l
00@,000000 0 0 0@o 0 (i 0000'0000 0 0 @o 00 0 (!
l 00000000 0 000 0 00 (!
l oo@oo@ooo @oo@oo n l
0 0 0 0 0 0 0 0 0 0;o0000 (j
000 0 00000000000 n
E h
~
b' E
.hi L
il lis
~h l
3 i
333
P Assembly Width N
flodel llounda ry Vacuum Doundary Condition L.-
me 4
q7J rl s
m Figure - 3 1
6 MCNP4 PHR Quarter-Core i I i
I I Model for LULOCA i
Criticality Analysis 1
+
1 l
i
-4 t
',l t
f j
f l
I I
L l
l j
s.
h t
- h 1
-c.
.u 1
m 1
I j
il :
l
.}, Mr
. )
ll.a i!
i i,
,]' J:
l l
!i!.h j;
- );
~
p.
Ik) del Boundary lE Ej{
Ill E L. J L 113 mea a
a
=a-
-4 l
e 0
0 l
0 2
0 N
0 9
O C 1
R E B
M OS P
B 0)
P 0 M
/
6 8 P M
5 P
C 1
S=
P
(
V T N
0 O
1 F
0 I
F@
0 T e
7 A 1 R EN 5
T O
N t
KI E
E R
l T
0 C B
G 0 N l
FA 6 O M
f 1
P OR C
P T
N
/
N N O
M 0 R C
OE 0 O P
5 B 8
IT C 1
e 2
AN 0
IRO 0
0 0
AC 0
=
V 4
R 1
O F
R R
F N
0 A
E E
K 0
M 4
1 6
2 3 G
0 9
91 I
S 1
0 0
E N
O l
1
m v
0 r
0 m
0 2
e
-r 0
w N
0 OC 9
1 RE B
OS M
e P
B 0 0 M
/
0 P
)
0 8 P M
1 S1 P
C P
(
V=
T N
0 O
1 F@
0 T e
0 I
F 7 A 1 R E N T
+
O N
6 K
E I
0 C T
B r
w E
l F A 0 N R
6 O M
w iG 6
O R C
P 1
I F
P 4
N T N
/
O M
N 0 R C
O 5 B 8
E 0 O P
ITC 1
A N e
2 0
I RO 0
m 0
0 AC 0
=
V 4
R 1
O F
R F
R E
E 0
A K
0 M
4 1
6 2 3 G
0 9
91 I
S 1
0 0
E N
O 4
FIGilRE - 7 VARIATION OF K-EFF VS. BORON CONCENTRATION @ T=150 SEC.
K-EFF 1.04 i
i I
1-O.96 i
1 0.92 1300 1400 1500 1600 1 1700 1800 1900 2000 BORON CONCENTRATION (PPM) e 8 PCM/PPMB e 10 PCM/PPMB
)
ONE SIGMA ERROR - 0.002 4
FIGilRE - 8 TRANSIENT CHANGE IN K-EFF VS.
SOLUBLE BORON CONCENTRATION I
K-EFF(T-150 sec) - K-EFF(T-0 sec)
K-EFF 1.07 0.015 1.06
- 1.05 1.04
- 1.03
-- 1.02 i
1.01
~
1 l
0.005 800 900 1000 1100 1200 1300 1400 BORON CONCENTRATION (PPM)
-- TRANSIENT K-EFF
-+-
- T = 0.0 s e c.
E e T 150 sec.
ONE SIGMA ERROR = 0.002
-. ~,...
. -. ~., _. -.
-