ML19312E431
| ML19312E431 | |
| Person / Time | |
|---|---|
| Issue date: | 04/04/1980 |
| From: | Denise R Office of Nuclear Reactor Regulation |
| To: | Gammill W Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML19312E432 | List: |
| References | |
| NUDOCS 8006040398 | |
| Download: ML19312E431 (2) | |
Text
-- w:'".
+
. W1 ;d:i
, f.h,W{wbS~-
- y
'4 p
f Central Files '
x a
[
h MgM
~ N CPB'R/F M
s N
y, NRR R/F _
}
C
^ <
~]..
7 RPDenise N
Apg 479gg.>
g
?. -,
c i
. 5.Sf b
a :- 1 f
..:c i n '
v y y :9,;
MEMORANDUM FOR: William Gammill, Acting Assistant Director for i~ % e
_*,./-
0perating Projects DOR i
?
_ Nj w
c FROM:
Richar[P.'Denise, Acting Assistant Director for ci y
Reactor Safety. DSS j?'
SUBJECT:
REVIEW OF TOPICAL REPORT XN-75-27 SUPPLEMENT 2 " EXXON
' '+9 Y NUCLEAR NEUTRONICS DESIGN METHODS FOR PRESSURIZED WATER -
~;
. REACTORS * (TAR-1722) p-x,
,4~.3
.c-c.. M.- -
.f.
.(,.
f.
The Reactor Physics Section o'f' the Core'Perfomance Branch has completed
. D.
i?
l its review of XN-75-27 Supplement 2.
The initial report and Supplement 1 were approved by the staff in January 1977. Subsequent to this, Exxon i t, analysis of the D. C. Ccok Unit 1 Cycle 2 startup physics program in-W.
dicated that additional model refinements were necessary in order to
^
more accurately predict critical boron concentration and radial power
. w :i distributions. These model improvements are described in Supplement 2
..M "1
~
and compared to more recant. operating reactor measurements, resulting in
- Q~ ';..),
an improvement in the prediction'of critical boron concentration and
.M6M radial power distribution. 1 ' ".,
,7 yr c.,cN
. ;( _-
r
^
Based on information presented in the report, discussions with Exxon, J(
and responses to questions by the staff, we conclude that' the modified design methods used for PWR neutronic analysis for cores similar to -
those analyzed in the report are acceptable and may be referenced in ue future licensing actions. We also concur that the improved modeling of -
A
't three-dimensional effects in~ the two-dimensional calculational models.
ge described therein eliminates the need for the empirical baffle model,
i previously required.
f,'
3 r: 2-r,, +..+ r ;m.Q~.
.m -
The staff does, however, recommend continued review of comparisons of^, gff%'
calculated physics data with measured data from future physics startup
~ MUf tests and operating reactor' data to provide continuing assurance of the w% JP T..
model applicability. -
",?~c 4
(
4'J r' O
~ t,
.T "ortgkal Signed By.
7
^
( S 4 (gchard P. Den.
,y g rP 7
qv v
gy x
d,
~^4 1 Richard P. Denise, Acting Assistant.
MT/; -
q' h<
.. Director for Reactor Safety..
7 ~ W S:
u
- [i, Division of Systems Safety.
h:."
~
g
'N]7{'"DI i
'80 b6d4'03 77 i M7Y u
y
.y.
n.
- 4..
- ~:-
ca$ tact
- *g 4
i_ m an.,naas.i.
. omcc.
- DSSICPB D1S/AD/RSb h
su ma$c - LKo.pp; pay
. DFieno'Of'
.f
.KKn.iel _
RfDenia}
.w.
ac"d.
M'
- ~ -- '
t
~
J" MIMM oarc >
04/ / : /80
'04/ M /80 04/c2/80 '
- 04/ yp!80 '
u V
_4 B.
Burnable Poison f ransition Effects Additional improvement in the prediction of radial power distribution is obtained by more properly defining the neutron spectrum used to calculate microscopic cross sections in fuel assemblies which have previously contained burnable pcison.
For example, the Cycle 1 core in D. C. Cook contained a heavy loading of burnable poisons.
Prior to Cycle 2 the burnable poison rods in each of these assemblies were removed and the assemblies reinserted into the reactor for Cycle 2.
The presence of burnable poison in an assembly causes a harder spectrum than in an identical assembly not containing burnable poison. This results in a higher plutonium content per unit of burnup.
In addition, the presence of the additional plutonium in the assemblies that have had the burnable poison removed produces a harder spectrum and, therefore, a different set of microscopic cross sect. tons than an assembly of equivalent burnup which has never contained burnable poison.
In the initial D. C. Cook Cycle 2 calculations the isotopic concentrations in the fuel were correctly transferred from Cycle 1 to Cycle 2.
However, the microscopic cross section sets used were from spectra typical of fuel which had not contained burnable poisons in Cycle 1.
This results in microscopic cross sections wnich produce reactivity worth which is slightly too high for these assemblies. Since all of the assembl.ies that had the burnable poison removed were placed in the center of the core, the calculated reactivity, and therefore power, was slightly overpredicted at the core center. ENC has modified their procedure so that the microscipic cross sections from Cycle 1 to Cycle 2 are properly spectrum weighted. This modified procedure is acceptable to the staff since it is more realistic and results in improved power distribution predictions. The new procedure shifts the calculated power toward the core periphery by 2 to 3%.
C.
Inclusion of Actinides in PDQ-7 Model Initially, the two-dimensional diffusion theory calculations of reactivity and power distribution did not account for the higher
1sotopes such as the actinidas. Specifically, the isotopes Np-237, Pu-238, Am-241, Am-243 and Cm-244 were not represented in the PD07/HARNONY calculation.
In order to improve the analytical prediction of the intial power distribution in Cycle 2 of D. C.
Cook Unit 1, the effect of these higher isotopes was included and found to decrease the critical boron concentration by 18 ppm and to I
l imi, cove the power distribution by shifting the calculated power to the core edge by about 2%.
D.
Power-to-Activation Rates for Flux Map Processing Codes An improvement is made to the calculated power-to-activation rate
(
correlation which is used in the flux map processing codes to infer powers from measured activation rates.
In the original representation, the instrument thimble and its associated water are homogenized, l
1.e., there is no discrete representation of the detector and its environment. The improved correlation incorporates a transport theory correction to the cross sections of the homogenized region l
to account for heterogeneous effects in the instrument thimble.
The effect of this on a core basis is the' lowering of the predicted power in the fresh fuel with respect to the burnt fuel. The overall measured power distribution is thereby raised in the center and lowered on the periphery of the core resulting in about a 3 to j
4% improvement in the agreement between measured and calculated assembly powers. This revised calculational technique is acceptable since it incorporates a more realistic model and results in improvement between measured and calculated power distribution.
E.
Non-Uniform Dancoff Effect l
A modification is made to the calculation of the Dancoff effect in XPOSE to account for nonuniformities in the fuel rod lattice such as control rod guide tubes and instrumentation thimbles. The Dancoff factor accounts for fuel rod shadowing effects in the i
calculation of resonance escape probability. The improved Dancoff calculation yields a 4% larger value than the original model.
{..