ML20138L573
| ML20138L573 | |
| Person / Time | |
|---|---|
| Site: | University of Michigan |
| Issue date: | 12/31/1982 |
| From: | Kerr W MICHIGAN, UNIV. OF, ANN ARBOR, MI |
| To: | |
| Shared Package | |
| ML20138L505 | List:
|
| References | |
| FOIA-85-587 NUDOCS 8512190303 | |
| Download: ML20138L573 (146) | |
Text
-
- , ' 9 j; % ,
.;...~,..,, - ,' 4 s . , , . .
. . . S.kM ; M .. . _ . .. .. Lai. uS6 L --
0;'
S$ }
.'i L JANUARY 1983 3
. .) -
y l. .
- d. Low Enrichment Fuel Evaluation and Analysis Program
[(
'J
.](.
j[ Summary Report for the Period j January,1981 - December,1982 i
9 li
- WILLIAM KERR, Project Director .
< L i
,s a
al$4 1
?9, i.{
Y.
q-LJ Q' Department of Nuclear Engineering O'
and the 4 Michigan-Memorial Phoenix Project
.j
$,)
t
!"3 B512190303 851203 PDR FOIA l AFTERGO85-587 PDR g
.. I a.
c.
__ - .. J 2 E2 -
. ~ . .
.. .. _, . 5. . , w ; .7,, _ .p,.,h a. 4_w.d.;@d.wkadEMa 4.
, . .- n. -.
~j . , ,( .
h
- 1. ~
1 n
-i i3 -
M 4, :
(1 d.3
.~.?
r] '
q ~
L't LOW ENRICHMENT 0 [:
Ji FUEL EVALUATION M-
- 7' 'AND N.
J! _. ANALYSIS PROGRAM N.-
'i 1.'1 l j l1; Summary Report for the Period January, 1981 to December, 1982
+0 li o
William Kerr, Project Director
<:j .
Department of Nuclear Engineering
,f.j j and the
.;j ! '- Michigan-Memorial Phoenix Project L
The University of Michigan
,,, Ann Arbor, Michigan
.n M.
l.1 Li l r '. -L.
January, 1983
[.. '.i !
L. i
- 7. ,'] , -
.v M,
{
~i
~. ; a$ .
- .k J
I i
e
\ .t105.$6: ' ' ' '
R r,~ " " ~ ' ' ' ' * * ~ ~ ~ ~ ~ ~ ~~ ~ '
v.3 .
.e pj k '
Nso t.
9; 9.3 v:
v:
e Fr fg PROJECT PARTICIPANTS -
E3..
s W
,y 4 William Kerr b
w N
re John S. King A
g John C. Lee Id
.. - William R. Martin 4
!Y
% Reed R. Burn g
g _
k Joo Hyun Baik
'J
);g Felippe Beaklini ~
p.
J. ;- Clifton Drumm
(
David Losey
} Gerald Munyan L}?
a ,a Joao Moreira .
r.
lj b.
James Rathkopf -
Gayle J. Stankiewicz h@
D _
f p
David Wehe 3 .
- ~
M
,-c Yh
')
s1 q
W :. .
> h
.k 11 4 ,_, ,(
)e e A* 'w . 5 , e-T w I
-~" Nw e~' *** e ' " ' - -'
- r. . :. ,
, , .. T .c .
- .,. -- _~ ,--
1
-s.r. -3~g-
'-f , ,& ~ & ; t,Q 7 ' j ? " " '~ ~
'1 . . . 2.
1 .
J-4
- .4 .
A k' TABLE OF CONTENTS d
M LIST OF FIGURES . . . . . . . . . ......... iv n-
\
8.;; LIST OF TABLES . . . . . . . . . . ......... V
- i
)_ I. INTRODUCTION . . . . . . . . . ......... 1 ;
a j
II. LEU DEMONSTATION EXPERIMENTS AT THE FNR .... 3 1
14 III. SIMULATION AND ANALYSIS OF THE TEST DATA . . . 7 3, A. HEU/ LEU Single Element Exchange ...... 8 B. LEU Critical Loading . . . ......... 8
,_ C. Thermal Flux Maps and Control Rod Worths . . 8 I!
(j- IV. GENERIC METHODS DEVELOPMENT AND VERIFICATION . . 9 (j " A. ENDF/B-IV LEOPARD Library ......... 10 3 B. Lumped Fission Product Correlation through h CINDER .. . . . . . ......... 10
$ C. Thermal Flux Maps j- D. Ex-Core Spectrum Calculations 19 36
'l E. Control Rod Worth Calculations . . . ... 39 2 F. Mixed LEU-HEU Control Rod Worths . . .... 45 G. Three Dimensional Capability for 2DB-UM . . 49 d; H. Effective Delayed Neutron Fraction . .... 49 y- V.
SUMMARY
AND RECOMMENDATIONS FOR FUTURE WORK . . . 51 REFERENCES . . . . . . . . . . . . ......... 54 APPENDIX A. FNR Demonstrati'on Experiments '
]j;' Part I: Beam Port Leakage Currents and Spectra d
J. APPENDIX B. FNR Demonstration Experiments *
(^! Part II: Subcadmium Neutron Flux Measure-gj ments O
- , APPENDIX C. Analysis of the Ford Nuclear Reactor LEU
,f Core f-
?;
'T-a w.
a
[ji s!
p -
S 4
h4 4
a
] iii 1,
Mh%d.t. -5:.cLC - ~
~
c- a : . . . -
} . .r : . , ,
3 a .s ,
$.} -
3 L!}
LIST OF FIGURES
.i.::)
1.j 1. Irradiated 233 U Resonance Integral . ..... 13 t:
9 2. Irradiated 233 U Thermal Absorption
- cd Cross Section . . . . . .......... 13 f.i n 3. Fission Product Group 1 Cross Section for the 3
FNR Fuel . .'. . . . . . ...... .... 14 y
G 4. Fission Product Group 2 Cross Section for the FNR Fuel . . ... . . . ...... . ... 15 i;"f ' '
'fj
- 5. Fission Product Group 3 Cross Section for the
,] ..
FNR Fuel . . ...... .......... 16 6:3 6. Fission Product Group 4 Cross Section for the
..,! FNR Fuel . . ...... .... .. . ... 17 _
[]a 7. D 2 O Fast Absorption Cross Section .. .... 25 24 8. D 2 0 Slowing Down Cross Section . .
} ...... - 26 ~}
-t
'i 9. Thermal Flux Distribution for the December
- ! 1981 LEU Core ..... ...... .... 27 -
]I 10. Fast Flux Distribution for the December 1981
- ? LEG Cora . . . . . . . . .......... 28 _
e_ '
a~ 11. SPND Thermal Flur Calculated by the ANISN Code 30 :
- 12. Linear Fit of log ($(E)/E) vs. E for HEU Fuel . 34 -
J:
4 13. Linear Fit of log ($(E)/E) vs. E for LEU Fuel .
e 35 f il 14. Neutron Temperature in the Aluminum Sample p: -
t :.<
- v. 15. Idealized FNR Geometry for ANDY Calculations . -
N 40
.e 16. Fast and Thermal Neutron Flux Distributions Calculated by the ANDY Code .... . ... 41 -
i.j 17. Neutron Flux Spectra Calculated by the ANDY -
0; Code . . . . ...... ... .. .. ... 42 f1 _
- 18. December 1981 LEU Core Map .
l:4 h- . ,
. ... ... ... 43 m' 19. Thermal Flux Distributions for the December -
bf 1981 LEU Core ..... ...... . . , , . 46 L.q ,.
RT 20. Fast Flux Distributions for the December 1981 LEU Core . . ...... .......... 47 1A ;
P]
- p. .,
p1 -
iv
, ~ ~ ~_~ : ^ ~ _
_ - _ _ - . - . - - . . - - _ _ _ . _a ~ ~
. , _ . , f;jf,.V e.-
s,, .,-.
. .: c;; f[7 '
?
, __[
+;s g ,_.. y w.
r - . ..- -*. , -
rj
, , 4,
$, l j LIST OF TABLES . i i
i 1. Thermal Flux Peaking in Special Element L-57 . 21
] 2. Thermal Flux Peaking in the Heavy Water Tank . 22 1
4 3. Rod Worths for the December 1981 LEU Core . . . 44 c.:
a.
L'l 4. Control Rod Worths for HEU/ LEU Mixed Cores . . 49 i.-) -
j 5. Effective Delayed Neutron Fraction for the FNR 52 3,
4 1s
~) .
1.l..
v,
- .f S--'
f e
i bN
~
i1-
- 1l
' s' L.
I4 5
!q li .'
9 M
,e n- -
..i
- h. ,
/
r.
's, W.
'! l
,${
r i. .
a
'f V
- . m y;, m'e , -
a:. .. n ' " e-i - a_. . d. a. a. . . r . . , s 7 ,.,...,, .; - ~ u. 2 i
_l .
.s s. .
a .-
~
ju i
] I. Introduction !
a ,
j The University of Michigan Department of Nuclear En- ,
{
],
gineering and the Michigan-Memorial Phoenix Project have i d been engaged in.a cooperative effort with Argonne National [
q 9;f Laboratory to test and analyze low enrichment fuel in the y Ford Nuclear Reactor (FNR). The effort was begun in 1979, a"
m as part of the Reduced Enrichment Research and 'l > st Reactor je (RERTR) Program, to demonstrate, on a whole-core asis, the j
.< l
.:s_ feasibility of enrichment reduction from 93% to ha.'.ow 20% in !
~
. o; ,
MTR-type fuel designs. !
7 ?- The key technical basis of the 1er en ic. Lent uranium, '
]_ creasing, at the same time, the uranium loading of each fuel l 1 element in order to compensate for the reactivity loss due, !
H to the larger 238 0 content. The required uranium loading l
., can be achieved by increasing the uranium density in the j fuel meat and by increasing the fuel volume fraction. At the same time it is necessary to insure that fuel elements !
operate within their thermal-hydraulic limits.
The first phase in our investigation ' performed in j
'I t preparation for the LEU fuel testing in the FNR. core.in- I s
cluded (a) initiation of development of experimental and i j, {
analytical techniques applicable for neutronic evaluation of l
" the MTR-type fuel elements, (b) selection of a LED design i for the FNR, (c) preparation of a preliminary FNR license .
amendment, and (d) a thermal-hydraulic testing program for
'q/ '
the MTR-type fuel elements.
o The 1979 Summary Report in- !
jj cludes a discussion of this initial phase of the FNR LEU
!II project.
'. Subsequent effort during 1980 was devoted to improving d-
<- and validating the experimental techniques and analytical
.y methods to be used in characterizing the high enrichment
- ). uranium (HEU) and LEU cores for the FNR. The experimental effort focused on the measurement of in-core and ex-core j-,. spatial flux distributions and the measurement of ex-core 4
1
^
;1 4 4w.. '
- +; + ,- a..d . :.
..w ' .$.JW ' n . .y.:
p - . . . . .a :- .;..-
1
?e
- M ~
2 51 m
fj spectra. In the analytical area, work has continued to im-3 prove and verify the computer codes and calculational models
[d
). used to predict the neutronic behavior of the FNR. This.ef-fort included comparisons of predicted results and ex- ~
{j/
j perimental data for various FNR HEU core configurations as (4 well as predictions of the impact of the LEU fuel on the FNR -
':1 vi performance and operation. In addition, a series of .-
'is thermal / hydraulic tests were performed for the MTR-type fuel g elements and an amendment to th'e FNR Safety Analysis Report fj was submitted as part of the required License Amendment to
.s 3 the NRC to permit the use of the LEU fuel in the FNR. (Ap-
~
proval was granted in February 1981.) The 1980 Summary MJ Report 2 presents the details of this phase of the LEU -
)I project. . -]
i .
- g The continuation of the project into 1981 culminated, p with the loading of the LEU core into the FNR and the ]. J
?. achievement of initial criticality on December 8, 1981. The -
.q critical loading followed one-for-one replacements of HEU L 6
n fuel elements with LEU fuel elements in the center and -
] , , periphery of the FNR core. Following the critical loading, ly a approximately six weeks of low power testing of the LEU core le was performed including measurement of control rod worths, ~
i full core flux maps, and spectral measurements in-core and h ex-core. This was then followed by two months of high power -
k) testing (2MW),-during which similar measurements were taken.
[]
These measurements were performed as part of the demonstra- _,
tion experiments portion of the overall FNR LEU testing program. Analytical predictions of the neutronic behavior
[W have also been made and comparisons between measured data -
j and calculated results have been performed. This phase of
}'i}
the LEU project has' continued through 1982 and the present
- p report summarizes the experimental and analytical work per-
$q formed during the two year period.
+ -
l Section II presents the demonstration experiments and -
g testing portion of the current project, including a detailed s ,
O,.
oli _
I= $
- Il. i 1' ._
T ~ r L. 7 -. . _ _ _
= _ . _
3: ' , ' . ' .
w.,- .. -
m & n-l __
, . . . , .; 3 -_.--
l' _
~
. 3 3-l* r. discussion of the measured differences in various neutronic
'Llj characteristics between the HEU and LEU cores, includi'ng i spatial in-core and ex-core thermal flux distributions, ex-5
" ^
core (beam port) flux spectra.and intensity, control rod worths, temperature coefficients, and xenon worth. Special a
attention is given to the subject of the measurement of the A
9- spatial thermal flux distribution, which is still a topic of g current investigation. Section II makes extensive referen-3
., ces to two papers presented at the International Meeting
$ on Research and Test Reactor Core Conversion from HEU to LEU y
y Fuels, which was held at Argonne National Laboratory during g the period November 8-10, 1982. These two papers are in-E cluded as Appendices A and B to this Summary Report q Section III is devoted to the analysis of the FNR HEU .
and LEU core configurations and comparison with the data measured as part of the demonstration experiments portion of
- the LEU project. The comparisons between calculation and j L experiment include differential reactivity comparisons of single HEU and LEU elements, critical mass of the initial 7,, , LEU core, control rod worths, various reactivity coef fi-d ' ' cients, and spatial distributions of the thermal flux, both k! in core and ex-core. Similar to the discussion of the ex-perimental program in Section II,Section III makes exten-
) sive references to a paper 5 presented at the aforementioned International Conference, which discusses the analysis of
$ the LEU core and. comparison with experiment and which has k ._
been included as Appendix C to this Summary Report.
li}- The FNR LEU project has also been involved to a sig-zu nificant extent in the area of generic methods development d
for MTR-type research and test reactors.Section IV sum-f marizes the work performed in this area over the past two years, including significant results and the status of tasks
]l currently under investigation.
j Section V summarizes the current status of the overall project, including a discussion of the tasks currently under d,
1 l
L!
. . . . -. ;Q .
. j . .. .
- a. . a nzpl;samm :..A.:-xx:a.::..- .2 - . ,
a> :
- y'
'.) 4
.- _a.
21* 8 r-
'Q investigation. The principal unresolved issues are iden-di3 .
, tified and recommendations are made for future effort to
. .s.
l complete the current and planned tasks.
.< f.
$9
>a
.e4 '
'u
- e a
rf j d
.g .1 P49
'd Ja. .,
r t
.s ) .
i g.1 -
.'T
-i
- f. l
.M 1
- * ' se e ?-
F f 4,'
, , 1'b w
.I
. ' >3'$ ~
,M i 't 1.f+
/~,
S
.A.
ee
- +A r'.
9 I
,'il aY A
e _
IN'
'Ij ..
I * :j'.
- 7.1 o,
,.t. .,
, I
.I i L n+w-...-.- - au . ., . __ _
- .. - - - . . .- . . -. w .
.a +
~_
- 3. .
5 lt II. LEU DEMONSTRATION EXPERIMENTS AT THE FNR The Demonstration Experiments Program in this 1981-1982 j report period has by and large completed the measurements on 4
the first full LEU core required for a comparison with E
similar measurements made on the HEU core in the previous year. This has included mapping of subcadmium neutron fluxes in core and in H O 2 and D 2O reflectors, thermal neutron leakage current and spectrum at beam port exits,
} Rhodium cadmium fractions in core and in reflectors. The f .-
comparisons of fast spectra in the LEU fuel are still in h progress. Other comparison data are discussed in detail in two papers entitled FNR Demonstration Experiments, Parts I and II, presented at the Argonne National Laboratory Inter-I national Meeting of November 8-10, 1982. These are
{ {' reproduced as Appendices A and B in this report.
Particular effort was made to maintain the da*.a acquisi-kI -
tion instrumentation unchanged between the period of equi-librium HEU operation ud il the beginning of LEU operation in December, 1981 bnd throughout the LEU power operation ga during the first four months of 1982. Three LEU configura-j, tions were examined frcm January 1 through April, 1982. To
], verify the reproducibility of detection, a reloading back to j{ the equilibrium HEU core was examined in May, 1982.
fi k..
i During the course of the LEU observations it became in-L Li '
- reasingly evident that measured subcadmium flux profiles p were strongly affected by the "small core" geometry of the
$ fresh LEU load as opposed to the "large core" equilibrium j HEU core. A series of "high leakage" (HL-MEU) loadings were
] 2xamined beginning in July, 1982 and extending through Oc-
[ tober, 1982. These cores retained the small core (five ele-lj -
ment) width of the LEU core.in one dimension (North-South).
)i v
They demonstrate that this first FNR demonstration com-g parison.is complicated by the large change in buckling j necessitated by a clean LEU crticial loading. ,
While much of the data confirmed expected changes in ~
1
- 1 P . . . , . . . - . - - - , . - - . . _ , . , . . . - _ . - - _ - . - - - , , - - - - - _ - - - - --
.L. J. . J . - - _~ l:% , -u. ;. 4 . . .: , .. my q:wSa%::4%.=
. r - ~ ~ . ~. - :.:. . .-
9
.j -
g 6
'& l 9
I flux levels, the series of experiments of early 1982 un-covered several unexpected experimental results. Two of 9.! these are of particular importance. The foremost problem is If a lack of agreement between subcadmium profiles in the D20 ~~
d
- q reflector, relative to core center flux, when measured by re our Rhodium Self Powered Neutron Detector (SPND) and when m a. -
$j i measured by both iron and rhodium wire activation. As a
- 38. ' consequence, a number of simultaneous SPND and activation
-:t q observations were initiated in late 1982 but as yet the ex-5 g perimental discrepancy remains unresolved. Since the SPND
-has been the workhorse flux map instrument, and no wire ac-7)]
q tivations were thought necessary during the LEU power s
a cycles, this discrepancy leaves the measured D 0 reflector 2
] fluxes in doubt. A second, related problem is the apparent '
disagreement in the ratio of LEU to HEU beam port leakage
"..].
currents when measured (a) in the conversion of HEU (9-81) to LEU (1-82), as opposed to (b) the reverse conversion of LEU (4-82) to HEU (5-82). This disagreement is detailed in j Table I of Appendix A. It is presumed to be due to changes
]
() ~ ^^
in the LEU East-West loading geometries of (1-82) and '
(4-82). I*, is complicated by the fact that the leakage pat-9 torn was four.d to be sensitive to the source plane position j'. and beam departure angle in the D 2 0 reflector tank. It would be desirable to remeasure a given LEU-HEU conversion
- j, again if only to resolve these two questions.
3.j Despite these difficulties, the confirming results of d-Y the experimental progran are substantial. We may list high-y lights of these results here, and refer the reader to Appen-j->.
, dices A and B for detailed exploration:
t
$ 1. Replacement of a fresh HEU element by a fresh LEU 4
$ element at the center of the HEU equilibrium core '
reduced the midcore flux in that element by a factor 11 y of 1.192.036., Well within experimental uncertainty, i this is equal to the ratio of the LEU to HEU U-235 -!
y; . masses.
c:;
q 2. Whole core replacement of the 38 element equilibrium 2 HEU core by.a nearly unburned (<3%) 31 element LEU _
E I
. = . >
I a; --
. . . , . . . . . . . - - - - . :' . , . ~ . --.w~
~
- - - . - - . ~ -
j, . . .
,4 -
- l..
'i
,;I 7
.fi '
l core (LEU 4-82) reduced the subcadmium flux averaged fd "- over the five central core elements by a factor of 1.1720.02.
1 3 3. Simultaneous with the depression in core flux in 2),
j_ the reflector peak fluxes increased by factors of N 1.53 in H O and 1 y -
These fac$ ors are.17 in D 0 at midcore heightdominat3d by the chang
~
geometry, as demonstrated by a similar response from the high leakage HEU profiles (HL-HEU 7-82). Lower
',. factors are to be anticipated for larger equilibrium 5.j LEU cores.
t t 4. A mild hardening of the core spectrum was j demonstrated by measured cadmium fractions from
]pu. rhodium wire activations. No spectral hardening was evident for the leakage currents from beam ports 7.j - facing the outer edges of the D 0 reflector.
2 3 .
{ .
l l
i j..
r M
b
=
pg=
.L
~
.%4 fi g
- s qi i
~
14 _._: z -
._. _ - _ w - ._
..amm a . ; . u.. -
- 2. den
..L.,.,'
.w.- .. ._ .. *
. .s .
k .
f 8 G -
h.1 III. SIMULATION AND ANALYSIS OF THE TEST DATA
[j A. HEU/ LEU Single Element Exchange g,1 The single element exchange experiment was performed to -
}j ascertain some of the performance differences of the LEU and
j.b FNR. The exercise also provided an opportunity for verification of our analytical models. The element exchange f y((
J-was performed by substituting a fresh LEU fuel element for a l fresh HEU fuel element in an equilibrium HEU core. The swap ii was made twice at two different locations: at the center and I
periphery of the core. The exact element locations, the ex-j
(
perimental procedure, the experimental and analytical !
results, together with an explanation of the analysis are j presented in Appendix C. j
~$
.l In summary, the reactivity effects of the exchanges -
n predicted with the perturbation code, PERTV0 , agree very j
,1 l well with the experimental data. The LEU element was found !
7 to be less reactive than the HEU element at the center of i i the core but more reactive at the edge because of the !
" leaky" nature of the LEU fuel.
g B. LEU Critical Loadinq j
The first complete loading of an LEU core into the FNR i
-.i 3; took place on December 1981. The two-day loading procedure ;j
] was completed at about noon on December 8 with the placement S
a of the 23rd LEU element. At that point the reactor was >
lf slightly super-critical when all shim rods but not the
[
'4
- r.
- regulating rod were withdrawn. <
3-- Simulation of the critical loading sequence by the 2DB- '
7 ;
UM code using each of the two libraries (old library and -
ENDF/B-IV library) for the LEOPARD code 0 demonstrates the f
Q superiority of the new data set. The calculations made with .!
j '
the old library predict the final critical core configura- l
- <s
- ' tion to be sub-critical, who'reas the ENDF/B-IV library cor-rectly predicts the core to be slightly super-critical. A -
..a ;
j.
A L1 :
o '.i
__ J L L. : f. - . -
. ' L . _. . .. .
, , y_ g, ,, c ._ ,.. .-. - ... ... .- .. ... - . _ .
9
., more complete description of the results of the experiment,
. . ., the sequence of fuel insertion, and the analytical calcula-3 tion are given in Appendix C.
si
! C. Thermal Flux Maps and Control Rod Worths 1 Simulation of the measured thermal flux distribution f.. and control rod worths for various LEU and HEU configura-r tions was performed with the 2DB-UM code as described in
] detail in Appendix C. With a standard 6x6 mesh per assembly a,,
structure and LEOPARD generated two group cross sections for 1p the unrodded fuel assemblies, the calculations generally agree well with the measured flux in the core region. Cal-
((
culations reported in Appendix C, however, indicate that i' thermal flux peaking in the water hole in the special ele-
. ments is underpredicted in the calculations. In addition,-
j, the thermal flux distribution in the core region near the heavy water reflector appears to be slightly underpredicted in the 2DB-UM calculations. For the heavy water tank it-N; J
self, 2DB-UM results compare fairly well with the thermal i flux distributions determined from iron wire and rhodium "t --
wire measurements. As noted in Appendix C, however, there is a considerable discrepancy between the thermal flux in
'.~ the heavy water reflector determined from the SPND measure-1 ments and the corresponding wire activation measurements.
L Active investigation is underway to understand and resolve this discrepancy in thermal flux data in and around the heavy water reflector. Our recent activities in this area 7 are summarized in Section IV.C.
simulation of the shim rod worth data presented in Ap-pendix C indicates favorable agreement for rods A and C, jr while rod B worth is overpredicted to a significant extent 7 in the standard 2DB-UM calculations. This discrepancy ap-g pears to be also related to the accuracy in our everall flux
,! distribution calculation. Results of recent investigation
} in this area, including the parametric studies performed,
$ are discussed in Section IV.E.
nm.:a:. .. . , .
, . . _ .,: a.L~.upxL:q.~u l
10
- d:-
IV. GENERIC METHODS DEVELOPMENT AND VERIFICATION 2.*j A. ENDF/B-IV LEOPARD Library 9 . -
A new library for the LEOPARD code containing ENDF/B-IV
{l i
2,i data has been developed as reported in Ref. 9. The new li-h brary is necessary to remedy differences observed between N,
s data generated by the EPRI-HAMMER 10 and LEOPARD codes.
[g Modifications were made to the LEOPARD code to accommodate
.1
@ the new library and to update some of the physical constants
% " hardwired" into the code. The new library and the modified 4 version of the LEOPARD code have been verified extensively s "4
] by the simulation of critical experiments, comparison with 1 -
J1 a established benchmark codes, hnd modeling of the depletion ,
i of fissile fuel in pressurized water reactor (PWR) fuel. (,
2 } The verification process is described'in detail in Ref. 2. !
]' .
The effects of the ENDF/B-IV library have been sum-
, macized in Appendix C. On the average, the multiplication y constants obtained for the fifty-five cold, clean lattices G
-d with the new library were closer to 1.0 than were the 5, '
Q results with the old library. Better, but not per'ec::, -
y agreement between the EPRI-HAMMER and LEOPARD codes has also 4 been achieved. The comparisons between the measured con- -
$ centrations of actinides in PWR fuel .and those calculated
] with the two LEOPARD libraries indicate that their differen- ;
$ ces are slight and that both libraries model the burnup and j
depletion of the actinides to similar accuracy. The LEOPARD y code can now be used with more confidence because of the in-
~
f creased reliability of the ENDF/B-IV data over the old in-dustrial data set of the original library.
?
i
-y M l j B. Lumped Fission Product Correlation through CINDER
As a means to account for the poisoning effect of fis-j sion products easily and officiently, the LEOPARD code cal-H culates lumped fission product cross sections at each burnup -
.] step. The cross sections, expressed in terms of barns per .
f fission, represent the sum of the cross sections of all in- .-.l Os I
e ,
g l
- g. - . . .
, .7..-...., , .mgyg: _ _ :-5 A' ' ~ ' "
i .
.t
.. 11 3e j dividual fission product nuclides except 135 Xe and 149 Sm l l; , .
which are handled explicitly in the code. The aggregate 1 cross section for the thermal group is calculated by means !
$' of a third order polynomial in fuel burnup. In the fast
'$' 4 energy range, only the epithermal (group 3) cross section is l 3 considered, again by a polynomial but of the second order in l
- 3 burnup. The correlations that were previously in LEOPARD s were originally intended for only slightly enriched fuel
' l ..
(approximately 3%), and have been used to date in our
- neutronic analysis for the FNR core with a crude adjustment "I. f to account for greater enrichments.
l ff In an effort to verify and improve the calculation of I.
the lumped fission product cross sections, the EPRI-CINDER y code ll was modified to run on Amdahl 470V/8 computer at The
_' University of Michigan. The EPRI-CINDER cede is a point-depletion code that usas detailed fission p.oduct decay I chains. .
]"
j Simulation of PWR fuel with 2.6% 35 0 by both thw EPRI-CINDER and LEOPARD codes indicates significant disagreement 1- in the fission product cross sections obtained by the two codes.12 The value calculated by the LEOPARD code for the j_ epithermal absorption cross section, o3, is nearly 50%
. greater at beginning of life and 20% greater af ter fuel burn-
) up of 20,000 MWD / tonne of uranium than the EPRI-CINDER
, values. The thermal absorption cross section for neutron
- ] energy of 0.025 eV, on, shows even larger disagreement
[ nearly 60% over the entire depletion history.
Comparison of results from the original CINDER code 13 (the predecessor to EPRI-CINDER), EPRI-CINDER, and LEOPARD h simulations of 3.4% enriched PWR fuel indicates that the S. correlation incorporated into the LEOPARD code was probably
( based on calculations with the original CINDER code or a
.:., program similar to it. The. CINDER and LEOPARD results agree quite well with one another but not with those of EPRI-CINDER. The EPRI-CINDER results, with the ENDF/B-IV data
- /
/
\
. . ..-.: .. s.. . .w M .
S 12 -
j f base for fission products, are expected to be more accurate
] than the old CINDER calculations. This has been verified by
- an EPRI-CINDER simulation of a detailed experiment performed to measure the fission product poisoning in 233 U fuel ir-
~
[ radiated in the Material Testing Reactor.I4 The EPRI-CINDER d
q results agree fairly well with the experimental data.
j Figures 1 and 2 show the experimental and calculated values M;
for the fission product resonance integral and thermal ab- .-
sorption cross section, respectively, as a function of ir-radiation time.
/ ;
- i As in the case of the PWR-type fuel calculations, the j simulation of the four primary FNR fuel types--HEU and LEU, _,
, regular and special fuel elements--by the two codes, LEOPARD l 3 and EPRI-CINDER, yield significant disagreement in the cal-
! culated fission product cross sections.as seen in Figures 5
, and 6. Note that the LEOPARD correlation used to date produces only one curve for each enrichment, be it for a j special or regular fuel element. The LEOPARD cross sections
'{
.i are. larger tha'n the EPRI-CINDER results, as much as 90% in -
'^_1, .
the casa of. the HEU regular fuel element. Because of this drastic discrepancy, new LEOPARD lumped fission product cross section correlations have been formulated based on the
- k results of the EPRI-CINDER code.
H In the new fission product correlations for the LEOPARD
[]i code, cross sections are correlated as a function of fuel h burnup, where burnup is given in units of MWD / tonne of 235g j
o rather than in the traditional units of MWD / tonne of U.
g This yields more precise correlations than possible with -
L burnup variables chosen in other units. ' Figures 3 through 6 -
if show fission product absorption cross sections for all four .,
groups vs. fuel burnup in MWD / tonne of 235g, The actual correlations in the form of polynomials were ..
b.. derived from the burnup-dependent absorption cross sections L.
$ plotted in Figures 3 through 6. In the case of the fast-
,q, group cross sections, oj, a2, and a3, the data for all four -
m
,,..e s.p, .J. wp .os4" ' % ' b' M'M59'" 7
-74 =-WG r ~w =
i
. ;. . . - . - L.- . - . . - . . .
. ~.. ,,.,; . ., .
},n, _. .,;.
.,.s.. .+
^
i, .
t . . ..
1' - -
.i -
.i 1 13 a,
li gL
-j i
- 300 ,
EPRI-CINDER
-l t - O Experimental (Ref.14 -
d l
0
- i. O O
.~ G o O l
0 c 200
- eo O l
- 7m -
8 O o o w e e
- 0 i.*- N O m
-1 c _
y 5 w 100 - -
a 1
f,i
.a ily
- I i 0 '
, 0 5,000 10,000' Irradiation Time (hrs.)
Irradiated 233 U Resonance Integral
- Figure 1.
1 fL g
EPRI-CINDER O Experintental(Ref . le
'i 60 "
0 0
c 0
n4
- m
. m 40 o o
- ' C 0 b o k]I' ) - 8 o o
- !. 5 o 8o o8 e
a.. . .a 20 O
,li o 6,
F.
0 '
f l :
/
t 0 5,000 10,000 Irradiation Time (hrs.)
Figure 2. Irradiated 33 U Thermal Absorption Cross Section A
r i/ U ,.. . ..TA: 511S h . d u' d b.1,'la d N '/ "'S R dl. s b.l- - Ai EUS dt 2d[ bIA.I.!52d2?ibd ide d.E. ' Oui.! YO.U'. f.-
i I
I I
,I i
o e .
g.. . t so a IIEU Regular Element O O O LEU Regular Element A IIEU Spec Element '
e LEU Spec Element 4 -
n c: t O m go O. .
to O wt y _. - - - --
m _m w ~ " " .
__ m m _ , ,
~ -
_3-c
$4 e y
.Q o O. .
O e-4 4
O EPRI-CINDER Calculations N
O O' f t-I
~
r i
4 o
i o .
) o .
(
O 1 2 3 4 5 '6 7 Fuel Burnup (105 MtID/ tonne of 2350) i i
- I 4
- Figure 3.
i Fission Product Group 1 Cross Section for the FUR Fuel
- 4 g 9
e _ o.om.d m*-e -
- - . . S - : a , .L.l ' ..,g;;. cy . . g . . .:Q.: .
. 61 . . . .- . u. ; .4, :,n w b .a:u :i,~. . & . . , . . .r : . .. .
c- _ % ,
l .
.. .4 rl i
sn N
d' .-
_ _ , , _ ^ - - -
r- -
_ i_ e_ ^
O .
N.
- O '
C 'b o
.e4 m 'A. .
e o 3 u 5 HEU Regular Element 'l N O LEU Regular Element m A IIEU Spdcial Element f + LEU Special Element n H
..s
.O *
~
N O ..
4 O 6 EPRI-CINDER Calculations an o '
o' ' !'
.'t
-s
- e o
o . . . .
o O 1 2 3 4 '
5 6 7 f .'
.s Fuel Burnup (105 MUD / tonne of 235g)
?i.
- c Figure 4. Fission Product Group 2 Cross Section for the Fl!R Fuel t.
f
',Ti' e'
.m. .
~
_ fs.&.. L . . L.:2 ut. . .; - ,. s i:: ~- 3s :1s*7'
%Lu- "' ; " - ' ~ - * * ' - -'Gd " # ' '#'#*" " " * ' * " ' ' * ~ ' ' ' '
- t. .
t 1
i i
O O ,
C$.. '
i t
J 4
o o
cl. .
n c ' ' i m u = ,g .,, ,-N 22 o 8 -
x4 + -+ ~+
-g '4
-+.['
%- ~~-x -4 x-H N
en _ _ _- - -
c k
e o -- -
-_1 ^ ^ * .
.O o -
-v
. , g m
- e-p a
m e 5 HEU Regular Element .
o EPRI-CINDER .8 LEU Regular Element -
Calculations 4 HEU Spdclal Element 4 LEU Special Element o
- o. .
m Old LEOPARD + IIEU Correlation z LEU .f o
O - . .
f.
'l).O 1 2 3 i
4 5 6 7 Fuel Burpup (105 HWD/ tonne of 235g) F Figure 5 i-Pission Product Group 3 Cross Section for the FHR Fuel '
~ '
e g.
. I r
t
,r-4 W tJ -j W W 1 .J ._; _ ^8 I ' i ' ' '
.,,--a _ , , ,,, . . -; m ym-- , . a. . . %. < <.e...~,.....s.ma
~. -. . , -v.s. .xa.u.. w . +.
- . . . . w.- ...w. ....-....,..s..u__---,-,.a...t...a.
.. .3 - , .u . r. . . .
, 4 o
O.
o .,
O.
N }
L o .
O 5 IIEU Regular s'lement ..
d e . EPRI-CINDER O LEtl Regu.ar Flement Cagcugations- A IIEU Spec..at Element 4 LEU Spec..a1 Element ;
1 I _ o Old LEOPARD + IIEU ,
C O Correlation x II:U j o ojg 3
~1 g- .
m i m
~4
' = = gi %t ,,4,+ , 4 . .)
1 (m ' ' 'N ' ' ' 4
-+ ~
-x _ _.
k ' ' ' 'x . ,' _' ' '+ . p ,
,h
-M a
._a .o - -x _ _ ...
, = =
0 i O . h, .
o
_ ^_
un o
d,
- -- ?_ ^
4 . ..,i '
t Thermal absorption cross section for .1/40 ev. 1.
I o
o j
b.O l' 5 5 / 6 7
Fuel Purnup (10 5HWD/ tonne of 235 g3 Figure 6 Fission Product Group 4 Cross Section for the FNR Fuel .
4- o
.- e 4
.~. . . . . _ . - . . . . . . . _ _ . - - -. ,
i .
1
% 18 i
fuel types were used together to derive a single correlation j for each group. For the 0.025 eV cross section c , because o
- )
the differences between the fuel types are relatively large, H only the data for the HEU special fuel element were used to ]
] correlate the burnup dependence. Another correlation to
] represent the differences between fuel types was derived by ":
j taking advantage of the different amounts of fuel and water I
} present in the various fuel types. This correlation takes .,
j .th e form 9
d j
co (r) = co,HS Cj {1 + C2 exp(-C3 f I where r is the ratio of hydrogen atoms to 235 U atoms in the ' 3 unit cell at beginning of life, a ,HS o is the thermal cross section for the HEU.special element obtained as a polynomial -- j] in fuel burnup in units of MWD / tonne of 235 U, and C , C.2, 3 [ and C3 .are derived constants. The data used in deriving _, j these three constants were the values for ao for the four J, fuel types at a burnup of'100,000 MWD / tonne of 235 U. The 1 lumped fission product correlations derived in our study are y (.I
- 1
~ simple but sufficiently accurate. The maximum deviation be-1 tween the correlations and the EPRI-CINDER results is ob- '
served to be about 4% for low fuel burnup for the LEU regular element, with the deviations usually less that 1% 3 [.; for all four groups and four fuel types. e j With the new fission product correlations, LEOPARD cal- , i culations were performed to. generate few group constants for the 2DB-UM code. Comparison of the eigenvalues calculated ., by the 2DB-UM code with those of the identical test cases
- with a library utilizing the old burnup correlation indi-
- cates the effect is small but significant. "
The test case simulated the June 1977 HEU critical experiment, with the old correlations yielding an eigenvalue of 1.0139 compared -' with a value of 1.0182 onrained with the new correlations. . 1 W The HEU and LEU batch core depletion tests illustrate m % 'i 4- m t -
, ---M - , . - - ^ __ ,
_x #_ . g. g.4. y,,q,, ,....x. ; [_ i ... + . n f. l _ .J..n .. , [,- - Ic 19 iL "i i the differences more completely, for fuel depletion is the j basis of the correlation. The difference in the eigenvalue j, for the two correlations increases almost linearly with fuel , il depletion. The eigenvalues with the new correlations are
- .i t Jt
'.1 greater than those with the old as a consequence of the g overprediction in the absorption cross sections by the old . [' correlations. At the end of life at 200 days of depletion, g the differences amount to 0.68% for the HEU and 0.54% for
- 5. L the LEU core, a
'i wl The new LEOPARD fission product correlations duplicate ': l . the EPRI-CINDER results quite well for the four FNR fuel types analyzed, thus eliminating a known deficiency of the ]f LEOPARD code. The repercussions of the modification are not
- j. dramatic but significant enough to merit further study.
m C. Thermal Flux Maps
.7 A substantial amount of offort was made between Decem-s e ber 1981 and October 1982 to measure the thermal flux dis-tribution for various LEU and HEU core configurations, as described in Appendix B. As explained in Appendix B, three
[{ methods were used to measure the thermal flux, iron wire ac-tivation, rhodium wire activation, and SPND measurements. A substantial amount of effort was also made during the same
$ period to model the measured thermal flux distributions by 15 2DB-UM calculations, as described in Appendix C. The -4
}8 methods used to calculate the thermal flux distributions are n, j- described in Appendix C. 1 As noted in the appendices, there are several areas of q inconsistency in the measurements and the calculations, the main problem being the large difference between the wire ac-tivations and the SPND measurements in the heavy water tank. [k~ In order to resolve these inconsistencies, calculations were performed to determine the sensitivity of the thermal flux y- to various parameters. This study is not yet complete, but it appears at the present time that the thermal flux dis-k(
- g. tribution is most sensitive to the absorption and slowing
}
q r i
. ,m.
, y__ .
__.__.al, 20 l i l [. down cross sections in the non-lattice regions of the core 3 and in the H2O and D 2 0 reflectors. Moreover, these par-7 ticular cross sections have the most uncertainty associated
;5 with them because of the significant global spatial / spectral d coupling in these regions, especially in the D 20 tank. This d coupling makes the traditional unit cell approach for a -
4 generating cross sections somewhat uncertain. d j Parametric calculations have been performed to study 3 the thermal flux peaking in the special f uel elements and in y the heavy water tank, and the sensitivity of the SPND detec- .. e tor in the core and reflector regions. The results of these .l
$ calculations and comparisons with experiments are summarized below. -
Jil q t
- 1. Flux Peaking in the Special Element Water Hole The special fuel elements contain fewer fuel plates than do the regular fuel elements, and have in place of the y
center fuel plates a water hole, for the insertion of con-trol rods and sample holders. Table 1 compares the measured ] and calculated flux peaking in the special element at loca- ':j' ~ tion L-57 for the April 1982 LEU core. Comparison is made [l here in terms of the ratio of the thermal flux at the center .< g of the special element located at L-57 to the flux at the (j center of the core at L-37. s bi j The flux distributions. compared in Table 1 were ob- ., tained with the 2DB-UM code. The reference calculation, ' _i M m performed with a standard mesh structure of 6x6 meshes per S - assembly, and lattice and non-lattice cross sections s pj generated with the LEOPARD code, underpredicts the flux G peaking compared with the experimental data. In order to jj determine the adequacy of the 6x6 mesh structure to ac- *
- _, curately model the flux peaking, another 2DB-UM calculation -
} was performed with a 12x12 mesh per assembly structure. The - i .i
- 1. s 12x12 mesh calculation essentially had no effect on the ..
.;.! results. j' . , n-Li m- , I I l _ _ , . . - c. . F
;. ,j ,; .,, ,4 .,;s _g , ,,,, ,. . ,.- . . . . . . ._ e -
q _
;f -
i 3 21 dL 1 d Table 1 a q' Thermal Flux Peaking in Special Element L-57 3 it 5!
- d i Thermal Flux Ratio
(.j Measurement or Calculation L-57/L-37 lh 1" SPND Data . . . . . . . . . . . . .. . 2.05 e 6x6 2DB-UM calculation with LEOPARD i Non-Lattice Cross Sections . ... 1.65 gi 12x12 2DB-UM calculation with LEOPARD 1 Non-Lattice Cross Sections . .. . 1.66 0 6x6 2DB-UM calculation with EPRI-HAMMER
, E. Non-lattice Cross Sections . .. . 1.83 'i . a 4
], The cross sections for the water hole of the special-j; elements are computed with the LEOPARD code by including a
,4, large non-lattice region (half of the total cell) and.using " ' edited non-lattice cross sections for the waterhole. In or-der to determine the sensitiv-ity of the flux peak to the cross sections used in the water hole region, the cross sec-t/, tions for the water hole were generated with the EPRI-HAMMER j, code in the third calculation compared in Table 1 This jl resulted in a larger thermal flux peak in the special ele-ments, as is shown in Table 1, and better agreement with ex-
'y periment. However, the use of a one-dimensional transport Y code such as EPRI-HAMMER for this analysis may still be in-j{ adequate due to the complex geometry of the special element j' and surrounding fuel. Therefore, we will consider the pos-
,{ sibility of generating accurate cross sections for the spe-jL
.., cial elements using a two-dimensional transport theory code p which should be capable of treating this particular j' geometry.
- 2. D2 0 Tank Flux Peaking d- Thermal flux distributions in the D 02 tank have been 1
calculated with the 2DB-UM code to simulate the experimental data obtained with the SPND and. wire activations. Table 2 b m N O n,. - - , . , - . , , , , , - . , , .,--,-,,,-w ,,n.. .,,,,n,,,.wr,-,..-, , , ~ , - , , . . , , , , , _ - ,
l... ,
, . ~ .
-- - - - - ~
l , . 7.'l , a 22 4 O j compares the ratios of the thermal flux at position X in the {, D2 O tank to the flux at the center of the core at L-37 for {- the October 1982 HEU core and for the April 1982 LEU core. i The location of position X is shown in the FNR diagrams in t ( . Appendix B of Ref. 1. h
/ Table 2 E
Thermal Flux Peaking in the Heavy Water Tank 3 Measurements or Calculations Thermal Flux Ratio
;y D2 0 X/L-37 a
f, October 1982 HEU Core c Y.Q
~
Iron Wire Activation . . . . . .
- ( .82 g Rhodium Wire Activation . . . . .88 ~
Rhodium SPND . .... . . . . 1.23
- d. Calculation . -
. ........ .81 b April 1982 LEU Core i , f' 3 ,
n . Rhodium SPND g
.... . .. 1.61 *i 6x6 Mesh Calculation Control Rods Out i LEOPARD D 0 Cross Sections .901
., 12x12 Mesh Caldulation (
.; Control Rods Out
'l LEOPARD D 0 Cross Sections .904 u 6x6 Mesh calculation -1 l',
3 Control Rods Out
- j LEOPARD D.30 Cross Sections SI j.
Zero Axial Buckling ( l in the Heavy Water Tank 1.28 *
,i 6x6 Mesh Calculation ;; control Rods In p,
4 LEOPARD D 0 Cross Sections .895 6x6MeshCalculation
?; Control Rods Out '4 ,'3 ANISN D 2 0 Cross Sections . 1.26 b +
s ; A ' 4 The reference 2DB-UM calculations for both core con- l a figurations were performed with 6x6 meshes per assembly and , f On _ _ _ , ,. ,;.,,- -- a uw =- -,- - r -
+ ... . g ;1 <
l__ 7 -r , y .g., .__ ..m-, - -_ . 4 .,dt ., s, c1. . ,,g4 , ;7-r . ... j y a ji 23 -
]b n
I-] lattice and non-lattice cross sections generated with the j LEOPARD code. From Table 2 it can be seen that the flux peak in the D20 tank measured by wire activations is sig-G e L. nificantly less than that measured by the SPND. The cal-culations agree well with the wire activations, but not with { the SPND measurements.
;] ' A 2DB-UM calculation with 12x12 meshes per assembly
}l was done to assess the adequacy of the 6x6 mesh structure Q" for the LEU core, which indicates a very slight difference compared with the 6x6 mesh result.
(jj L In order to determine the sensitivity of the D2 0 flux peak to the axial buckling in the D20 tank region, the D 20 tank buckling was artifi-
;, cially adjusted. The results of the 2DB-UM calculation for ) ,
zero D2 0 tank buckling is also given in Table 2. Although the increase in D2 0 flux is substantial, this parametric ; j variation is quite unrealistic, and gives only an upper , j bound to the buckling effect. Flux m.asurements are normally made with regulating rod partially inserted, but the partial insertion of control
?"- rods is' difficult to model in a two-dimensional calculation.
l 4 Normally, the fluxes are calculated with the 2DB-UM code ji with the rods assumed fully withdrawn. In order to inves-l;1, tigate the effect of the control rod insertion on the spa-J' tial flux distribution, the flux distribution was recalcu-
, lated for the LEU configuration with the rods assumed fully L inserted. It is expected that the actual flux would lie somewhere between the calculation with rods withdrawn and that with rods inserted. The insertion of the control rods
,j ' shifted the in-core flux away from the D2 0 tank, but had al- ;
j; most no effect on the flux within the D 02 tank, as is shown lL in Table 2. l I h The D2 0 cross sections that have been used in the cal-
?,
a culations were computed with the LEOPARD code, by using the j method discussed in the previous section of this report. 3 These cross sections have also been calculated with the y-s l p
Eli.n., - m c _ .._ _ _ ._ . . . . . . _ . - . . . . 1 . y' f 24 ANISN code,15 a one-dimensional transport theory code. In this ANISN calculation, the core and reflector regions were modelled in slab geometry with a 30 group cross section set which was collapsed from a 123 group set with the XSDRN code.16 The fast. group absorption cross section from the $ ANISN code was significantly less than the LEOPARD value, - and the ANISN slowing-down cross section was significantly larger than the LEOPARD value as is shown in Figures 7 and
- 8. The fast absorption cross section calculated by the LEOPARD code is 4.7x10 ~4 cm ~l, an order of magnitude greater ,
than the value calculated by the ANISN code. The ANISN cal-culation also showed that the D2 0 fast absorption and slow-9 ing down cross sections vary with penetration into the D20 tank, illustrating the spatial / spectral coupling in the D 20 t tank. - Flux distributions calculated with the 2DB-UM code - 7 using the heavy water cross sections from the ANISN run , h showed a substantially larger thermal flux peaking in the y) - heavy water tank than the calculations with LEOPARD D 0 2 , ll . cross sections. The use of ANISN heavy watec cross sections . 3 also caused a shift in the in-core flux toward the heavy [ 3 water tank, bringing the flux calculation into better agree-ment with the SPND measurements, but into larger disagree-
]
ment with the wire activations. Comparisons of the flux 7
- ! distributions calculated with the 2DB-UM code using 'i LEOPARD- and ANISN generated 2 D 0 cross sections for the 7
g December 1981 LEU core are shown in Figures 9 and 10. The j
- j comparisons are for the thermal and fast flux distributions, _
$ respectively, in a north-south scan through the core. The 7
, ratio of the thermal flux in D 20 tank position I to that at the core center obtained in this calculation with the ANISN cross sections is compared with the other calculations for , the April 1982 LEU core in. Table 2. -
4 Future work in this area will involve the determina-tion of more accurate D2 0 cross sections as a function of { 9 a -- 'i v 9 - [. -
--w
.. . . uwa==.aw a .= . . . .x . :.... ua.a= wa. . .a.x a arzwa..:., .. acc2r:sra ua . + . ..
m -. - .. - -
.~,~- r- e-- m-y , , , c ,
; ,. 4 C i ,
.l l
l l
. z.
L 10 i i i i i i i i i
'r .
f { .'t
. I.'
8
.a e
'o 8 -
w r 8
- t e ainsu
= 6 -
.i:
, 4, O ,
b , g '.
- }-
"2 O
4 - u s M U /
&2-
.o O
e ,g,
.i' o oca -
q ,. x O C
\
O C r c O 0 ' ' ' ' ' ' ' ' t I O 3 6 9 12 15 18 21 24 27 30 i. i. Distance into Heavy Water Tank (cm) f l Figure 7. D2 0 Fast Absorption Cross Section L.. , l 5
'r
l ll l l
- e j ri E,I.h,. ' 4 - ' ;
- - 0 d
o. u i
~.
- i. t
'e R
I
'Lt u ;
c 1 O: 1 _ - a 0 . U - _ Co c 3 d d N i i 7
- 2 i
l c I i 4 - t' 2 n o , e._ _ I i 1 ) m c i t c 2 ( S e k s n s a o . I
' 8 T r 1 C r
D e n P t w _ T a o i _ D l f D O I T ' 5 y g
. ' T 1 v n S a i .,
T Ic w o
~ l 1
N A I o l S
' 2 t ,
1 n O - i 2 J e D f c n . I a 8 9 t s e i r D u 3 g I i F _ 6 f
- i. ' ..
3 s e
- - Ca J. - -
4 0 oo 0
. 6 2 8 4 1 2 1 1 0 -
2
~.. n4 eI 8te1O4~v 8 t./ maoo g8 l4 4
4 0 o4huao4 j
'. ~ ~
i w L l: i. f. 1 h e 1 R_-.
~
i i'
g
.- . . p.a . . . - - -
. g.*%~~ _3 . '- .
.3-
.f f f;*
- j j* ,
' " ^ . .
.u =. . . = .. : = aaw.= = u- , . a . " --- " - s;.
~2~ pn ,n - y pm
. r. -
t t f, . l DEC 1981 CORE L-50 TO L-45 ..
- I I I I I I I I I I I I
.i 4
3.5 -
"2" " "' "- "2 -
r ( X - 33.0 - (i _s 4 - - u.
$ 2.5 -
2 ., m M 2.0 - p'N"""n? b ;- Cross {
$ 15 r *
' s\"**** "
x. 3 W t t A*
- 4 / \ *n
-8 g l
Ld to n
/ tropapp x _ R 5
E - g 2DB-UM Calculations D)O Cross Section 0.5 - . l 1 0.0 ' ' ' ' ' ' ' ' ' ' O 10 20 30 40 50 60 70 80.90 100 110 120 130 140 ' POSITION (CM) , Figure 9.
%ernal Flux Distribution for the December 1981 LIII core .
e
_I '
. ,.C g (,
- i! i l !a J 1.
i
- ; Ilt ,l I ! -
m.
= =
W a. a. w= xu a, 0
- u. - ~ - - - 4 1
a I 0
- n. 5 ,
31 e r o a1 4 O 2 0 C U
- l' '
2 1 E. 1 w. L O 0 1 1 1 8 9 1 r 0 e T , 0 b m a. 1 e ; 0 \ c e 0 , D 3
' 0 9
D _ 5 s t h t e
't 0 s'g, i l '
8 ) a o r L g L t ( c f c o, s n ' 0 7 n o n o ) E i
= i t
o i t i t u R t 0 c a 0 i b l ' s i O 2< u o r 4 D s c 6 P t
= l- s C
- , t M j i a $ -
R W S = I C ' 0 D O C p 1 t a M 5 x D a 1 R A s A c
/
U
- l u = ,
P s B 0 F 8 O0 o
*' 2r t c
/' /
D 2 4 t 9 LD C s
- a. 1 e
S 0 3 F 0 a C , 0 1 w E 0 2 e r u D 3 g
=.
, 1 1 ' 0 1
i F e 0 _
= O. 0 0
- 0 0 0 0
= s 5 4 3 2 1
0
=. XD'h p N 4 ' w ;> F < J.L
. D
.JC u )
- u. -
]
c - a -_ w _
]
[ (
,. ,I li! L
,-a
.___.%._ m. s
,. - ;.~ 7 _. _. -
- m. - - -
1 "* I F
;i 29 ,
j 'L !-j position within the D 0 tank. Thi' calculation is compli-2 { ]_ " cated by the c'omplex geometry of the .ank, which neces-sitates at least a two-dimensional transport code to be ade-
)
quately modeled. For example, a two-dimensional transport ; 7.i l~
}
code could be used to model the core, D 0 tank, H O reflec-2 2 f
] L tor, and beam tubes in cylindrical geometry, which is still an approximate model of the actual configuration. Ad-j
[ j ditional future work will include three-dimensional diffu- : _ . sion theory calculations in the core and reflector regions,
; using few group D 2 0 cross sections that vary axially and ]
radially within the heavy water tank. diffusion theory results with other transport theory cal-Comparison of the (
- I
-!' culations (including Monte Carlo) for the heavy water tank 1 l
j will also be made. j
- 3. SPND Detector Simulation
- I, 4
Modelling the SPND detector and paddle in the core, i, l ~ water reflector and D2 0 tank has also turned out to be a j ,, difficult problem. In preliminary calculations to date, the { ANISN code has been used to model the SPND and paddle in the i 3' core and reflector regions in cylindrical geometry with ?O , z groups. The detector is surrounded by a region of moderator C J, .j l material and a region of core material. A white boundary !
- 3
^ condition is used at the external boundary of the core ! region. The thermal flux in the SPND detector and Inconel
. i paddle is shown in Figure 11 The flux has been normalized (
,] at the surface of the Inconel paddle. !
., , This type of one-dimensional calculation is not a very ;
-l accurate representation of the actual problem. Measurements
~j , in the D 20 tank are made in aluminum tubes filled with water
? that only penetrate a short distance into the heavy water
.- tank. The measurements are made near the core-heavy water i
,1 tank interface and near the axial heavy water-light water
[!, interface. The flux in the region near the detector is [ 3a highly asymmetric, and the flux depression is dependent upon , y the three-dimensional geometry of the problem. The measure-l.' r
! l f
p _ _ _ _
.__.m e' I V O113.s . fi;U. . . . L + , .L .nD > u -2, 12 ;~;.lf.,*: ' SAL!,hD2.Ifaft$51rd51.'.ei'.e.iit . L ..
! i I4
'l, 9 . ,L i 'j.
y , I'
~
t-1.2 i i , i i i i C
, PO y It' rot 1ET. l.
L PP DDT.*: I' E- 'i 1.1 E9-X 3 $. m 1 SPND It! af. t.. [3
._3 "/ $ h i ICP.
id N N n
- ~ ~
1.0 _ _ .l d ' SPID IN
# SPND IN 2 y 0.9 -
"2 -
m r / o r.. t-Ld G 0.8 -1 - 4 .t Ld s ) 30 group ANISN E Calculations in 0.7 - cylindrical Cec = try -
.T-0.6 O.00 0.05 0,10 0.15 0.20 0.25 0.30 0.35 0.40 t, Radius (cm) I Figure 11. SP!ID Thermal Flux Calculated by the ANISH Code l
. t r.
.C! Q ' _J w ) i ____ ._
n_ ..s AdC- s L_ ~ . ~ w. . . - . . . - ...~..?
- i. .
%y.< L ;)
31 d.., t L h . ments within the core are made in the water gaps between the
, fuel plates.
n It is difficult in a single calculation to model both j the SPND detector and the region surrounding the detector. g Because of the large resonance absorption in rhodium at 1.3 (j ev, a transport theory calculation is required to model the 1 detector. The geometry of the detector is well suited for a j cylindrical geometry calculation, although the modelling of l - the environment surrounding the detector requires difforent c ,
-types of calculations. In the core region, a slab geometry transport theory treatment is necessary to model the fuel
, plates, clad, and water channels. In the heavy water tank, l
o
- two or three-dimensional diffusion theory calculations are necessary to model the interface of the core,, heavy water,.
j and light water regions. Transport theory calculations are
~
f required to model the beam tubes in the heavy water tank. Figure 11 compares three calculations done with the i
~
ANI5N code in cylindrical geometry to investigate the effect g of the surrounding medium on the thermal flux within the H
-d SPND detector. The rhodium emitter wire, aluminum oxide in-
) sulator, Inconel sheath, and Inconel paddle are explicitely
. represented. In the first calculation, the detector is sur-j rounded by a region of core material. In the second cal-
;. N 4 .. .
culation, the detector is surrounded by 16 cm of D 20 and a 4 region of core material. In the third calculation, the 1-4 detector is surrounded by 4 cm of H2O and a region of core 9"t material. The core material in the second and third cal-
- g. culations provides a source of neutrons for the calcula-tions. The three calculations are normalized at the surface k.j < of the Inconel paddle.
I
;L These calculations reveal some information about the
, self-shielding of the thermal flux within the SPND detector. l %g , The self-shielding factors for the SPND surrounded by a medium of core material, D20, or H 2 O are .75, .72, .72, respectively, which are not significantly different from 1 j
_ _ . _ _ _ ._ _ , . - . L u. .a ; . < . . . _ _ . 32 each other. The bulk of the self-shielding is due to the
/]
c.j rhodium, with only,a small contribution coming from the In-2 conel sheath. This is reasonable, considering the large D resonance in Rh-103 at 1.3 eV. Additional calculations g 7 g which more accurately account for the complicated geometry ] of the region surrounding the detector are needed to deter- , [] mine the flux depression factors for the SPND detector sur- !
- l. rounded by a medium of core material, D2 0, or H20. The flux ij depression factor is a measure of the decrease in the flux j
.i.
at the surface of the detector due to the presence of the detector in the medium. The sensitivity of the SPND is determined by the combined effect of the self-shielding fac-to'r and the flux depression factor. h To model the environment of the detector in the lat-i tice region, the EPRI-HAMMER code was used in. slab geometry - j to compute a thermal . spectrum in the f uel, clad, water chan-
! nel, and non-lattice region of a special element. An aluminum sample holder was inserted into the water hole of
~
the special element to simulate the holder used to guide the detector. The spectra were analyzed by fitting a strai'ght j'
~~
line to a plot of log (4(E)/E) versus energy to determine a j neutron temperature characterizing the Maxwellian distribu-tion for neutron flux 6(E). Fits of log ($(E)/E) versus d, energy for the HEU and LEU configurations are shown in J. Figures 12 and 13, respectively, indicating that the flux is
] nearly Maxwellian from .01 eV to .95 eV. .l Figure 14 shows a plot of the neutron temperature as a j function of position within the special element. The figure ',q shows.a decrease in the temperature of about 7'K in the ., 9
- water gaps between the plates, where the spectrum is ex-pected to be softer, and a further drop in the water hole,
- and an increase in the aluminum sample holder. The calcu-lated temperatures within the HEU and LEU fuel meat are 340'K and 347*K, respectively, indicating that the thermal i{ spetrum in the LEU fuel is somewhat harder than in the HEU -
1
~).
a...
(a,3(._ ~ . - ..... . -- . - - - , . .- ~ . - p< - J . - 1 Kj 33 h
]' fuel, as expected. The calculated temperatures within the aluminum sample holder in HEU and LEU fuel are 330*K and 333*K, respectively.
_ D. Ex-Core Spectrum Calculations ' J1 As part of our effort in developing models to represent (- the complex geometry in the l'NR heavy water tank, both one-dimensional discrete-ordinates and three-dimensional Monte h 6- Carlo calculations were performed. Both scalar and angular fluxes were obtained in the ANISN calculations to compare with the flux spectra measured at the beam ports. Only
.l preliminary Monte Carlo calculations of the scalar flux have j been performed 'with the ANDY code IO to date.
f- 1. ANISN Calculations 4_ The crystal diffractometer sensurements yield the ther-j mal neutron spectra at selected beam ports in the D2 0 tank. j . As reported in Appendix B, an insignificant change in the ] L- neutron temperature'(i.e., the temperature corresponding to 9 the log ($(E)/E) fit) at a specific port were observed be-tween the HEU and LEU fuel; however, a large difference in i neutron temperature was observed when the diffractometer was j "I moved to a dif f erent beam port. Although this was only one set of measurements and the core configuration was the so-j called "high-leakage" HEU core, the difference is large f enough (40*K) to warrant additional investigation. Since .;q this temperature difference was observed for I and J ports, A which " view" the core at angles of 63* and 105* from the h 21 north-south line, respectively (hence J port is actually
- 11. looking away from the core), the thermal neutron spectrum may depend on direction, as well as distance into the D 0
( tank. As a preliminary attempt to simulate this geometry, a one-dimensional ANISN calculation was performed utilizing a 2 20 group (6 fast, 14 thermal) library collapsed from a 218-f .. group ENDF/B-IV SCALE library I7 An S8-P3 calculation was performed for a full-core north-south traverse, including the H2 O reflector, core, and D 02 tank. The resulting ther-4
,< M >
. ' en &w .w s de l. - ;L -4 .. eu- s.'n n'n.. . .. . . . O, m%L1s,i?.WS, Dt t
, 1
, J,
']
-4 o
. .\
34 t
'i
,z .
,,.T
%'h*
4 me
,4
. ',)
..g
.s
=4
' a.j y,
< g 6
6 6 6 e 1 2 J" g
- 2 n- % u o
e *
.s -
_ g
- g . m k
.3.
a k
=
j
}
R I
- 5
. w t -
u a
> w 9
i W 3 I .e a 4
.+
g we b
$ be I , - O M l m N et 9 4
- 9 " g 3
M 4 g g b - t'. I
, b b ,
~
. .j
* =
U >. N
-f k -
;4 MM s
e
=
e la. I - a
. 21 c-l M tin e
iJ
, = , I g
.I'
.e h e 4 0 (utam et 9 m l
.s 3
\
"1 .
1
*I .
l.- i: ,, ,
~
9
\.
I e e end W : T M .a M m .MAnM
^
,_y__- &=-
.z_,
, 7 . . . _ . . . . .
a, 4 x 1 35 3 ja 3 il. J' s l3 o 5
. I s
6 ai lj 6 g , j ., j i- e e 1 . J - u i .g O k
- O
%,e
. M I
I
. J, v> >
I .
.u I a s
m m i
> M
. = -
u D .E g g
- w b o g t. .u
- w 0 t w aa 88 :
a J r 38 ya
- 5 e .
I
. ~
24 - O
, dG g L. m a: ,
I f , ' s (3/ (2) $ ) sot
'I e
P9 8
, . .* d; -; '; h . , , .1.
t
&_GA. . -
. - . ' . ' . _ ? ~c . 3. g .~ ' . a d _ c ;.
.L. , .
_ . _ _ . . ~
', . ... ;1 ,. .: [ 1. ~.:1L: - i:. .'. . w. .
t.
.f s ..
I ki i t: k . 3 3 3 g i ! AIUMIPUM gggggy, EAMPIP. - IlOID'm (CIAD-fur'Ir-CI.AD-P O. . . . ) f o A o 7 Je* cF c'i M i - , l^** 42 e h LEU See - g u 8 k
$ see j
r - 5 _] PPU
~
~
3ae - % EPRI-IIAleER Calculations in ~ 2 Slab Geometry i se I 8 I I ese tan am 3 ., 3, ).
. g Distance from the center of the Sample IIolder (cm) i r.
' h, Figure 14. Neutron Temperature in the Aluminum Sample ',
Ilolder in HEU and LEU Fuel j, v.
. . .i .
s' . t I
- s. -
. r- .U - I
- I ! 1 6 1 y _
i , . , , i
g-- , m, . . . -
._;.4
- s. .g. . . _ ~ - , _ . ._ i 1
37 4 ] mal spectra were fit to Maxwellians and a spatially-dependent neutron temperature was then obtained.
, Neutron j" temperatures were calculated for the angular fluxes as well j as the scalar flux, to ascertain whether the neutron gj temperature was a function of the neutron direction, as
) would appear to be the case for the measured spectra. The ANISN results, which must be considered preliminary, predict j a temperature decrease of approximately 20*K for the angular j_ fluz at an angle of 130* from the north-south axis (hence i; similar to J port) versus the angular flux at an angle of j 50' from the north-south axis, which is more like I port. i" The obvious drawbacks of a one-dimensional calculation for 4 this complicated geometry force us to conclude that although j there may be a directional temperature effect which appears 1 to be consistent with experiment (and physical intuition),~ we would prefer to defer further quantitative conclusions f or more detailed calculations (e.g. , Monte Carlo or 2-D "1
^ discrete ordinates). As an aside, it may be worthy of men-tion that the ANISN predicted neutron temperatures within
( , the core agreed (within 10*K or so) with completely in-9~ dependent THERMOS calculations of the fuel temperature, f which were performed to determine the effect of the in core sample holder on the fuel thermal spectrum. I 2. o ANDY Calculations i e As part of our effort to model the complex geometry of g the FNR heavy water tank, Monte Carlo calculations were per-j' formed with the ANDY code. Our effort to date has been limited to idealized simulation of a FNR geometry, with the L; primary purpose of estimating the degree of neutronic cou-pling between the core and the heavy water tank. Prelimi-li [.. t nary calculations have been performed with and without representing, in an idealized fashion, the beam tubes, as an ?l initial offort to evaluate the impact of beam tubes on the /; ' flux distribution in the surrounding medium. ' t il The ANDY code is a general purpose, multi group Monte }. I 4
_. :a --.:w - -=- w ~~ --
,lj
); -
.i j
j 38 s a q Carlo code, with a simple three-level topology for geometry I fij specification. As part of the initial investigation, several test calculations were performed with the ANDY code,
'I which compared favorably with either analytical solutions or - .l , the corresponding calculations with the ANISN code. For ] simulation of the FNR geometry, 27 group, Po cross sections ,.
[i in the ANISN format were obtained from the SCAI.E package for j use with the ANDY code. j The idealized FNR geometry used for our ANDY calcula- , j tions is shown in Figure 15, with a beam tube extending , j halfway into the heavy water tank. The core is represented 2 by a homogeneous mixture of fuel, water, and aluminum, with the 6 mm thick aluminum wall of the heavy water tank ex-
] } plicitly represented. The beam tube, when represented, is '
f assumed t'o be vacuum. Only fixed-source calculations, with , either a uniform or a cosine-shaped fission source distribu- .
; tion for the core region, were performed. } We present in Figure 16 fast and thermal flux distribu-tions for the idealized FNR with a beam port, calculated for a uniform source distribution, with and without importance.
sampling. The flux distributions, averaged over the cross sectional area of the core and heavy water tank, and over the circular cross sectional area of the beam tubes, are
" ~
plotted along a north-south scan of the core. For the case
~ ] with importance sampling, the particle weight that was as-1 signed is indicated for each region. With a limited number -- ] of particle histories simulated, the information presented J
1 in Figure 16 is of limited statistical significance. The ,
- general trend, however, appears to be reasonable. Even with j a uniform source distribution for the core region, which
~ ? tends to overestimate the flus in the reflector regions, thermal flux peaking in the heavy water region is not
( pronounced. This result needs to be, however, compared with - ] 1 more accurate calculations. In Figure 17, neutron flux spectra obtained with the ANDY code are plotted for a point - 1 - p{ .. . u' -
3 .,._.. .. ..,.
.. .~. u.i.+. .x i 39 h
1'
!- at the middle of the core, at the core-heavy water inter-j_ face, and at the middle of the heavy water tank. The results corsespond to the case without importance sampling plotted in Figure 16, and softening of the spectrum in and
])~ . around the heavy water tank is clearly visible. J ] Unfortunately, because of poor statistics in our ANDY i~ results, with limited numbers of particle histories simu-
! lated, neither the degree of coupling between the core and j- heavy-water reflector nor the ef fect of the beam port can be quantified. Further study is underway to perform similar
$ calculations with a deterministic code in an idealized two-I dimensional geometry to compare with the ANDY results ob-tained so far. Once the degree of neutronic coupling be-
,' tween the core and reflector regions is quantified, further
] Monte Carlo calculations may be performed for a more realis- '
tic geometry confined primarily to the heavy water tank. ) E. Control Rod Worth Calculations
- Full length rod worth measurements were made on the 27
~
element f resh LEU core in December,1981. The fuel loading for this core is shown in Figure 18. The control rod worths for this core were calculated with the 2DB-UM code as described in Appendix C and the results are compared with the measured rod worths in Table 3. As noted in Appendix C, the 2Ds. calculation agrees well with the measurement for
- l. rods A and C, but the calculation overpredicts the measure-ment for the B rod by 14.2%.
The rod worth was calculated in each case by computing the reactivity difference between a rod-in case and a rod-
, out case. The calculation was done with 6x6 meshes per as-sembly with control rod cell cross sections obtained with
- the EPRI-HANNER and TWOTRAN II codes, as discussed in earlier
- reports.1,2 Several parametric calculations were performed to determine the sensitivity of the rod worth to various L
a pa rame te r s . It appears that the most sensitive parameter is 4 the slowing down cross section in the heavy water, as noted
- h _ _ _ _ _ ___._ ...__._______________;
M--.- ts.9&nLG u,a : y3i au r - -. .. ru n % ;.. a L. C. G - :1 . _. i. ' ; ;.dL ' ' ' - ;G'.w Girik C:~ e caloWX -:..; M .O a.k K .....;.
.c --
1 l I t. l f t,
- p. ,
O. y,....-------------..------.-------------- e I o i ;
/s t s t 0.6sm,' ,
g -}
#8t*C -
e
.' l h
/ < r
, s
/ , I 0 ,/ '
I
' . g I --- , -- .. ... - - - - _ . _ . _ - ,
I , g 8 , 5 ; 4 0.3m s 3 8 g i 8 g e l
- -e 6 0 e 8 8 ,.
l ' 8 T { e
' Peam !
8 og,i s une. , owe Ileavy 8 'rtabe 1,., e I Pbrth * ' t S
- Mater i i
e o I O.6m ,l m e 8 l 8 I t , ! t / ) a a t ' s e/ a a I e i 4 I.
- 4 l / '
.i 8 I
. 3
- 8 a k
I
, a -
, e t ,e {
- 0. 3m e' 8 -
1
"'I4 ' '
e l i ,e g e
* ' . I e
. . - . . . . - - - - - . - . . . . .. . E' #
.!f b 0.381 m h 0.405 m ~~ N ^~
- 0. 318 ~i-- 0.2 d
.i j.
6 men . risjure 15 Idealized FMR Gecanetry for At3DY Calculations '
.L.
. , ), . '
{,.,,,_ , , , , , - i e i ! C._.) .
- F I ,
? .t
[
..,,,y.. . .
=
4 i-t 4 41 e.-
?
j 27-troup A::DY Calculations .4 3- ,1 M) RFWITl#CE SRFt DeG (104 histories) i
)
,, RF(NtTRNCE SRPLING (5 x 103 histories)
,o's*
- i. p
'. & s 4
- j #
Fast \ s riux S g g
\t t
j Particle 4,
- t'e ight t 1 -
t 1 1 1 2 2 3 2 4 1 ' t 'k t { u i 5 *-- t ater Core !t
' "**" == water t: ster
>' l*
O 8 8 i \ .I t
= it j t-
,l. x . t , s
- ( -a 6 w
2 b c
\t 1
. o.
6 %
! a m
5 5- - l1 u t t i ( {i Thermal Flux s \ ll i,
's% *% '
5 t ..
- t g
4' %s,s s,-
. l . . . -
!' 2MMI 50,M ~10,M M,00 15UXI 130,2 Position (cm)
, . Figure 16. Fast and Ther=al :*eutron Flux Distributions Cciculated J by the A::DY Code
..' I 5 0
.rna.a.%; nLL'uGGmn%26- -
MM-U - - R W F.% '& 1 2 ;> K Q 5 M .;; Ocz.u:h.v: ;. h . . r q
.i i
8 I n flIODL.E OF CORE i CORE REFLECTOR INTERFACE
-- flIDDLE OF HERVY WRTER TANK i J e
3 27-group AIIDY Calculations ~
!i R.
a
- u. .t y -
A ' y a V u
>.gB be K -
7 '
.g N
o L ' ' _ r---. - 2 e* _ _._r . l 1 ((~
- t o - - - -
-7_ ---
j
- J r g
g ---
------u-.....r-------------
. t.
=
. -- , _r-- -
Ix10-2 1x10-1 l lx101 lx102 lx103 lx104 lx105 lx106 lx107 ticutron Energy (eV) . Figure 17 t 13eutron Flux Spectra Calculated by the ANDY node - ,I t t i - f ' I t 1 _j ! 5 1 !
- d .
s t ., i i : ' e + ' ;
.,....__s7_...
. , .~. .. m ,a :
-- ..,.~ , ,1.
. . q.p.
..~ . . . _, ... . _ . - . _
4- *
. ,. ?
a 43 i 4 j-
,e a
l , 6, .
'; Heavy water Tank ,r -
v.-;
- -}
i
]
1 - A C
) .j 1
B ,
,i .t t i 6 e i.
J '
.g I
.d. .
b 8
?
^
Regular Special Empty ',1 Elernent Element Location
- l
~1.
2 i 1 i ., l' ' Figure 18. . December 1981 LEU Core Pap 1 f7-sp [4
- , - . - - n.. ~ .- .. , , . - - -.- .-w, , - . . , -4_, , ,-.,-.-,--,-,-v--- - . - , - - - - , - - - - - . , , - -
.,--7,---- -, ,-, ,- ,--r,,,,n,--,
I
- ~. a a u..
~~
.- - , .. ,- - . . ~ .
c . L ..,
}
44 l
. i Table 3 a -
g Rod Worths for the December 1981 LEU Core J Reactivity Worth (%4k/k)
,' Measurement or Calculation m A Rod B Rod C Rod ki , .
! Experimental data . . . . . . . 2.22 2.32 2.28 j .
6x6 Calculation - LEOPARD D 0 Cross Sections
~
Equilibri b Zenon . . . . 2.28 2.65 2.25 u 12x12 Calculation . 1 LEOPARD D 0 Cross Sections Equilibri b Zenon . . . . - 2.81 - d:I 6x6 Calculation ) LEOPARD3D 0 Cross Sections ' Zero Zenoft . . . . . . . . 2.32 2.66 2.29 I 6k6 Calculation . ANISN D 0 Cross Sections - Equilib$iumZenon . . . . 2.43 2.34 2.41
; in the discussion of the parametric investigation which fol-lows. . -
~
.. { ' In order to assess the adequacy of the 6x6 mesh struc-ture, a 2DB-UM calculation with 12x12 meshes per assembly - Li was done to model shim rod B. This calculation yielded an - 1
- } inferior result compared with the measurement than the 6x6
.l calculation, as is shown in Table 3. This is not too l' surprising since the rod cross sections were computed from _ EPRI-HAMMER calculations specifically designed for 6x6 2DB-UM geometry. Another 2DB-UM calculation with zero xenon concentration was also done to determine the offeet of xenon ~ concentration on the rod worth. As can be seen in Table 3, -: p, the potential effect of time varying xenon concentration is - f quite small. I l The effect of the D O 2 slowing down cross section on -< l
- the rod worth calculations was investigated by calculating l
.. the D2O cross sections with the ANISN code, as described in . l Section IV.C, above. Replacing the LEOPARD generated D 0 2 1 .1 .
. 6
j . ..,_ a.w a ._ _ , w s c.c __ ,,. ... . .:. ... f :-. g . .p.. y.,. -.
; .w
-~
45 i f- (
, cross sections with the ANISN generated cross sections had ; the effect of increasing the flux within the D 0 tank, thus f 2
4 l causing a shift in the flux away from the south facit of the 1 core, as discussed earlide in connection with Figures 9 and l
- 10. The flux is shifted away from the B rod and toward the '
}
- . A and C rods, causing an increase in the worth of rods A and C, and a decrease in the worth of rod B.
]- The results of 2DB-UM control rod calculations with the ANISN generated 2D 0 cross sections are compared with the g experimental data in Table 3. Agreement between measurement [- and calculation for the worth of rod B is much better with { 7 the ANISN D2 O cross sections than with the LEOPARD D2 0 cross sections, but the agreement for A and C rods is worse with the ANISN cross sections. A comparison of the rod-in and rod-out flux profiles through rods A and B is shown for the thermal and fast flux, respectively, in Figures 19 and 20. ] The ANISN 2D O cross sections were used in the 2DB-UM cal-
}- culations to generate these plots. .? ; The rod worth calculations at this time are not con-4- clusive. There is, perhaps, a need for an additional full-f length rod calibration on a fresh LEU core to further test the calculational, capability. There is also a need for an
.; accurate spectral calculation to determine the spatial de-pendence of the slowing down cross section in the D2 0 tank. n F. Mixed LEU-HEU Control Rod Worths A limiting factor in determining the maximum allowable fuel burnup in the FNR core is the shutdown margin. In or- _ er to determine the possibility of using partially burned
; LEU fuel in an equilibrium HEU core, the control rod worths ,
} for various mixed core configurations have been calcu- ' {~ lated. These calculations will be useful for studying an op-j timal fuel management strategy for the transition from HEU to LEU fuel. Application of mathematical programming to the ] e optimization of the fuel shuffling scheme of the FNR will require correlations to calculate the rod worth for various 1 - / .
----- -, m -, . .-._r--.
~,m.., .,,,..-.w._ _,-__.,,%,_m,,..,._,, w w e ., , ,,m,,,,_gy,.. ., - - --. .ww,,_,y,v_,. -
.._.m.e .* .- u. t.. .. : e. a..cm_. . n . mu m . :. o , c.tamam..m..m . ..1 .. .
k l f
/
- DEC 1981 CORE L-50 TO L-45 4.0 , , , , , , , , , , , ,
I? O CORP D "0 3.5 - 2 2 - s't ' L
- x g s l-33.0 i J -
La. * ( L' ~ i < L.D P-ROD
~ -
g NT
-s E 2.0 -
N - F .
\
J \ Ld ' 2 1.5 -
,i s
-i \ -
F- / s 4 ' g I I s , d E 1.0 - f -7 I N, - f s' I I
, P-Ron
,g III 2DB-UM
! 0.5 - i, , c.icui.tions - 1 0.0 ' ' ' i ' O 10 20 30 40 50 60 70 80 90 100 110 120 130 140 i t Pigure 19. POSITION (CM) .
. (
Thermal Plux Distributions for the December 1981 TEU Core i i.
. .V.
km
.r j I ; -1 1 i i ; 't it- ! -1 1 *I.-
8 __j ' I a ' r E
y- 7 y _ 9 . ;_ 7 ~ .m. _m;es.g.,7_ < .
. . . .m . . .
.j c
i DEC 1981 CORE L-50 TO L-45 ' 6.0 i , , , , , i i i i i i 11 O COPP. DO ff O
?
2 i 5.0 - -
.i
.i X /
D ' (N '
.nb0 -
I s - Ls. s P-ROD i g . ._ F OUT I s 7 i (n ) ' f 4 f ^tp-ano I N"
' 3.0 -
ni i w / ', a : .i 1 c
\ 'o -
, H i l s 4, 2.0 i < i y; w , x s = i
\
s 9' 1*0 - 2DB-UM Calculations s - g .; 1 I e I I _.. I I I i i i i 1-- I 0.0 1 i 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 . .L , ' 1 1 Position (ca) i
- Figure 20 Fast Flux Distributions for the December 1981 IEU Core ' l ..
l' E t
.+
s l ?'
... ~.-... - . . -. - - . - . . ~,
3 1
'! 48 ? mixed HEU-LEU core configurations. Calculations of the type d
4 reported here of fer an initial step toward achieving this l end. The calculations wjll also be useful in determining if _l
] partially burned LEU fuel from a batch core can be burned up I in an equilibrium HEU cycle without violating the shutdown margin requirements. It is hoped that the partially burned } LEU fuel discharged from the FNR batch LEU core can be effi-O ciently utilized in future mixed HEU-LEU cycles. Rod worth -
correlations are necessary to dete'rmine the extent to which . the partially burned fuel can be used, without violating the shutdown margin. l i 9 The calculations were performed by replacing the .. j eenter five fuel elements in an equilibrium HEU core by' LEU L fuel of various burnups. The burnups of the remaining HEU . fuel elements were adjusted by multiplying by a constant factor to keep the eigenvalue constant. Table 4 shows the control rod worths for FNR cycle 2115, and for the mixed ~- )) - cores with LEU fuel of burnups of 4%, 6%, and 8% in the ' -
} eenter five lattice positions. Replacement of HEU fuel by -
i LEn* fuel in the center five lattice positions decreased the-h rod worth by several tenths of a percent. The utilization j of LEU fuel with higher burnup resulted in a larger decrease Il in the rod worth than the lower burnup LEU fuel did. Many more calculations of this 'ype t will be required to develop ~ 'r.! sufficient correlations for an optimization study. Also,
, alternate methods for maintaining a constant eigenvalue -
i.( might be investigated, such as adding or removing fuel from i the outer region of the core. _, 1 G. Three Dimensional Capability for 2DB-UM - V, h The lack of an officient three-dimensional global ~ j analysis capability has been a serious limitation at the } University of Michigan for several years. Without this -- ] capability, accurate predictions of axial flux / power shapes . q and control rod worths.are problematical. This is especial- _ s ly true for the FNR with the D 02 tank on one face which 3 - a
~
gh r.r _ m _ - , ., _. -2%
%s y
. ?. . .
= w,- k .
. - .7
,.,, , .-- ~ .- . -
z--~: - ..- --~= ^ ~~ y,3,y-r, p: 31 4g m.n * -
~
g - 1, l J , 49 r -
! Table 4 N- ~
Control Rod Worths for HEU/ LEU Mixed Cores i
)?
j_. Total Rod Worth lj Core Configuration (%ok/k) A Ll ' Ll Reference HEU core . . . ...... . 6.629
. 4% Burnup LEU Fuel in Center 5 Elements 6.409
- 6% Burnup LEU Fuel in Center 5 Elements 6.366 j- 8% Burnup LEU Fuel in Center 5 Elements 6.323 v:
l 4' . . 3 makes the problem difficult to analyze in two dimensions. Ij Therefore, effort was initiated in August, 1982 to develop a j three-dimensional capability for the 2DB-UM code. Since a i primary goal of this effort was to retain the efficient
'j- macroscopic depletion capability of the 2DB-UM code (among .1 -
other features installed in 2DB over the years), the deci- [_. sion was made to make the 2DB-UM code three dimensional s, rather than beginning with the three-dimensional production code 3DB 20 and modifying it to be consistent with the 2DB-UM code. The major portion of this work has recently been com-q pleted and the 3DB-UM code appears to be working correctly, at least for several simple test problems, when compared with the 3DB code. In the process of incorporating the 3DB code into the 2DB-UM code, the disk input / output routines r -- were completely rewritten with most.of the storage main- !;[ tained in the memory with the result that the 3DB-UM code is ik,
- a factor of three faster than the 3DB code. Whether this
}4 advantage persists for the large full-core FNR problems S .- remains to be seen. The principal work remaining on this j task is to install the modifications necessary to allow the , i s 3DB-UM code to use the identical cross section data base o{" l (cross sections parameterized as a function of fuel type, $;~ burnup, fuel temperatures, etc.) that is currently used by
- 1. the 2DB-UM code.
1
)
F,F ,O C r. r -..-- , .- - - . - - . , - - . . . , , , . . . . - - . . , , - . . . . . . - , - . _ . . . - . . - - - . . , - . - ~ . . . - _ , - , , , . - .
,a...._-.- . _ _ _ . ._ . _ . _ . . _ _ .- . . ~ . ._
g ,. j .'. - Q ;.
- 50 Q H. Effective Delaved Neutron Fraction d
fj Calculation of the effective delayed neutron fraction 3 s 8ef, for the FNR core was undertaken in 1981 by J. Moreira 9 as a M.S. project.21 Three different methods were used in this study: a) first-order perturbation theory, b) eigen-value method, and c) non-leakage probability method. The eigenvalue method is based on the work of Kaplan and Henry 22 and the actual application of the method to the FNR con-figuration is described in Ref. 2. ] ' The third method was originated in Ref. 23 and involves Q writing the non-leakage probability dg of fission neutrons ~ as - (l~ h 9 Yfd> (1) 0 ( where S is the physical fraction of delayed neutrcns, and
'l! (p and [d are the probabilities that the prompt and -
delayed neutrons, respectively, will not leak out of a given
; volume during slowing down. Substituting Eq. (1) into time- "
fj ^ dependent diffusion equation for flux 4 coupled with the ~' f precursor balance equations, and taking the inner product of - the terms in the resulting equations with the adjoint flux 4 4*, we obtain, ) YL h " j R (2) F6-f(Qt*,QPo> i 4 where P is the production operator. Since the term P4 is 3l proportional to power, the ratio of the inner products in
- ) Eq. (2) can be interpreted as the ratio of probability of non-leakage of delayed neutrons from the core region to the _
corresponding non-leakage probability for any fission d:3 neutrons. Thus, S,gg can be obtained as 4 _ ( l) (3) h.. a'._.-
< cV,' ?
~
lt ' H ~ m ,n . _ _ -
. - - . - . . - - . . ~ , ..
}. - .u .
. . . . i .. _ le.nju;.u :.
. v ..
1 51 n y - j Two-dimensional calculations for FNR configurations in-1 dicate that method (a) with four energy groups, under-i}- predicts 8,gg by 2-3% compared with methods (b) and (c). O 3 The results from the latter two methods are in agreement j with one another to within 1%. To resolve the difference
]j between method (a), and methods (b) and (c), further effort was undertaken in 1982 to perform the first-order perturba-1 tion calculations wi'ht a larger number of energy groups. By j modifying the LEOPARD code, six group cross sections were
;[ obtained, with the .second and third groups in the standard
] LEOPARD structure replaced by four groups. A first-order
!" perturbation calculation with the six groups cross sections indicates a reduction, by a factor of two, of the difference 1
noted earlier in the 8,gg values calculated.
]- The values of 8,gg/S calculated for several FNR con-i- figurations are summarized in Table 5. In addition to show-ing the differences due to different calculational methods, o_ Table 5 also indicates that the ratio 6,gg/8 decrease as the core size increases. Direct eigenvalue or non-leakage
] ~
probability calculations with perturbed fission spectrum,
~
accounting explicitly for the delayed neutron spectrum, have not been performed for the 39-element equilibrium core con-I- 1 figurations. For both the HEU and LEU configurations, in p . this case, first-order perturbation theory calculations were
.. only performed. Based on the comparisons between the per-turbation theory and eigenvalue results obtained f or the i batch core configurations, values of the correction factor,
$ S,gg/S, obtainable with the eigenvalue method were estimated to be 1.139 and 1.132 for the HEU and LEU equilibrium core configurat ~ ons, respectively. Based on the calculations g performed to date, the effective delayed neutron fraction g- for the LEU core is expected to be slightly smaller than the
!. corresponding value for the HEU core.
- 1,
;f-ti ? -
51 1 2
- n. ,
$ L d,
- ~. aa.. . - . . ..-.- - -.
. =
3 , l.j .
'e ,
.. 52 b'J Table 5., Effective Delayed Neutron Fraction "il for the FNa %j Core Calculational Number of '~ f Configuration Method Groups 8,gg/S T perturbation 4 1.136 e.i 25-element eigenvalue 2 1.164 -il HEU batch core ,i 9 non-leakage 2 1.154 l
; probability N
{ (Q perturbation 4 1.123 k! 31 element perturbation 6 1.136
*1 HEU batch core
,, eigenvalue 2 1.153 LJ ~*
? non-leakage 2 1.144
,7 probability
, perturbation 4 1.114 '~
39 element eigenvalue, 2 1.139 J
- HEU equilibrium core estimated
;h perturbation 4 1.107 M -
g
- Y eigenvalue, 2 1.132 estimated ~
- h. -
a g)b s -f 3 s 7 v; . t e,;4 l,
?g '1 d _
.) - g _. D-fi
, .*" i
^^-aum --
w - , ==ma- - , --
. y:: a. , 9. -
. . . a, .= _ __ w - -n. .
- 4 ._ y y .{.,. m . 4 ... . n :A
) r 4.
53 ! a i Iit V.
SUMMARY
AND RECOMMENDATIONS FOR FUTURE WORK
~
The preceeding sections of this report have discussed the work performed as a part of the FNR LEU project during
;j' s
the 1981-82 report period. This section summarizes the cur-y' rent status of the project including recommendations for fu-t l si . - ture effort. ' j The experimental portion of the project has resulted in j significant advances in our ability to' m'easure the neutronic
.i _. behavior of the FNR (HEU and LEU) core. In particular, our j
development and testing (currently in progress) of the ! 3 multiple threshold foil technique will enable accurate l y measurements of in-core and ex-core flux spectra, including l thermal spectra as well as epithermal and fast' spectra. The 3 development of the ra'pid SPND measurement technique has al-lowed for more convenient full-core flux measurements and the determination of spatial sensitivity factors has in-creased the accuracy of the SPND measurements. In addition, y i_ the beam port thermal spectrum and leakage intensity 3 measurements have contributed additional experimental data
~
to assess the impact of the LEU fuel as well as to verify our calculational methods. Finally, of course, the success-2 ful installation and testing of the LEU core has verified
}'
j-that the LEU core is performing as expected, although ques-tions remain with respect to the measurement of the thermal [ ' flux distribution, especially in the D 0 tank. 2 In addition, there remain uncertainties in the measured thermal flux i spectra and intensities at the beam ports. In the analytical area, significant achievements of the current report period include the implementation and testing g of the ENDF/B-IV data base for the LEOPARD code and the suc-j' cessful prediction of the initial LEU configurations. Other h~ tasks which are nearly complete include the development of it'
/,1 the fission product correlation for MTR-type fuels for the p- LEOPARD code, the modification of the 2DB-UM code to incor- "f porate a three-dimensional capability, and the determination p
O \ t-l l
, n -- , - . _ , , . - - -. . , - - , , - - - - - - , , - - , - ,
_. w. u, l ph.h.Oht . sl '
- . e. ,
.x . -- . . . ...- . . . .... - .
4-
- e t g
- ) .
.1 sj 54 ~ ?
of the effective delayed neutron fraction for the LEU fuel. Nj The major unresolved issues, which have already consumed a
] significant amount of effort during the 1981-82 report I/,'
u period, include the prediction of in-core and ex-core spa-fj tial flux distributions and spectra, and the prediction of, j control rod worths. Other items that are not resolved in- - w
; clude the use of partially burned LEU fuel in the HEU core w (or vice versa) and the optimal approach to an equilibrium .
LEU core. lj Additional tasks in the analytical area that might be , ,1 9 addressed in the near future include the calculation of the IAEA benchmark configuration for MTR-type research reactors E and the implementation of our overall neutronic code package on a state-of-the-art microcomputer for use at other I 3 4 research reactor facilities. i 9-9 . i
- . ; s G
q - 41 m E ei4
'i n --
. s' i .]
.1
'i *
. ~
y$ _
'TM ha s"d.f csh n 4,p-aNh
- - ^
r . ___ - e+ =e . ~ M
~
m...___ _f _. ,
, _, q f.- . . .. ,, , y ,, . , . . . _s . . .
j , ... I . . . m-55 b-i , REFERENCES q
- 1. W. Kerr, et. al., " Low Enrichment Fuel Evaluation and
- Q]~ Analysis Program, Summary Report for the Period January, 1979 - December, 1979", Department of N Nuclear Engineering and Michigan-Memorial Phoenix
%- Project Report, The University of Michigan (January j 1980).
n d_ 2. W. Kerr, et. al. , " Low Enrichment Fuel Evaluation and y Analysis Program, Summary Report for the Period
- j. January, 1980 - December, 1980", Department of i Nuclear Engineering and Michigan-Memorial Phoenix 3~ Project Report, The University of Michigan (March 1981).
d 's 3. D. K. Wehe and J. S. King, "FNR Demonstration Experi-7 D ments Part I: 3eam Port Leakage Currents and j Spectra", Presented at the International Meeting on {. Research and Test Reactor Core Conversions from HEU 4 to LEU Fuels, Argonne National Laboratory (November 1t 8-10, 1982).
- .4 j '-
- 4. D. K. Wehe and J. S. King, "FNR Demonstration Experi-a ments Part II: Subcadmium Neutron Flux Measure-6 ments", Presented at the International Meeting on L-.
j' Research and Test Reactor Core Conversions from HEU to LED Fuels, Argonne National Laboratory (November 8-10, 1982). 4 j 5. J. A. Rathkopf, C. R. Drumm, W. R. Martin, and J. C. fq Lee, " Analysis of the Ford Nuclear Reactor LEU
$ Core", Presented at the International Meeting on . i' Research and Test Reactor Core Conversions from HEU j to LEU Fuels, Argonne National Laboratory (November J 8-10, 1982).
4
- 6. R. W. Hardie and W. W. Little, Jr., "PERTV-A Two-
.I Dimensional Code for Fast Reactor Analysis",
' BNWL-1162, Battelle Pacific Northwest Laboratory
., (September 1969).
fj 7. W. W. Little, Jr. and R. W. Hardie, "2DB User's Manual-
'.- Revision 1", BNWL-831 REV1, Battelle Pacific 1; Northwest Laboratory (February 1969). ,p_ 8. R. F. Barry, " LEOPARD-A Spectrum Dependent Non-Spatial Depletion Code", WCAP-3269-26, Westinghouse Electric Corporation (September 1963).
q" 9. J. Rathkopf, " Development of ENDF/B-IV Cross Section Li-A brary for the LEOPARD Code", M.S. Project Report, 1 Department of Nuclear Engineering, The University of k Michigan (November 1981). s l$ ! i.
w _ ._.,_.. _. % _. . L h f 'm - ._ _ _ .. . . . _ . . _ _ _ - _ J - i . ., . a . 1
.J 56 rt p'i 4 10. J. Ba,rhen, W. Rothenstein, E. Taviv, "The HAMMER Code 4 System", NP-565, Electric Power Research Institute j (October 1978). ?j 11. T. R. England, W. B. Wilson, and M. B. Stamatelatos, ,
~d " Fission Product Data for Thermal Reactors. Part 2: Users Manual for EPRI-CINDER Code and Data", yfd) LA-6746-MS, Los Alamos Scientific Laboratory (Decem-ber 1976). ' jj 12. G. J. Stankiewicz Time Dependent Fission Product Cross 4 Sections for 235 U Fuels", M.S. Project Report, _ Mj Department of Nuclear Engineering, The University of
- a. Michigan (July 1981).
h! 13. T. R. England, " CINDER--A One-Point Depletion and Fis-i Sion Product Program", WAPD-TM-334, Bettis Atomic
;.h Power Laboratory (August 1962).
~t i 14. S. B. Gunst, J. C. Connor, and D. E. Conway, " Measured aj
.: and Calculated Fission-Product Poisoning in Neutron-Irradiated Uranium-233", Nucl. Sci. Eng., 58, 387 _
p (1975). .
- 15. W. W. Engle , Jr. , " ANISN , A One-Dimensional Discrete Or-
/.
el dinates Transport Code with Anisotropic Scattering", " i K-1693, Oak Ridge National Laboratory (March 1966) . > .l '
- 16. N. M. Greene and C W. Crave, Jr. , "1SDRN: A Discrete p Ordinates Spectral Averaging Code", ORNL-TM-2500,
~~~
- ] Oak Ridge National Laboratory (July 1969).
I:j 17. R. M. Westfall et al., " SCALE: A Modular Code System for ~ 'l Performing Standardized Computer Analysis for ;
'd j Licensing Evaluation," NUREG CR-0200, Radiation Shielding Information Center, Oak Ridge National '
j Laboratory (1980). e [- 18. D. R. Harris, "ANDTMG3, The Basic Program of a Series of - (a Monte Carlo Programs for Time-Dependent Transport of Particles and Photons", LA-4539, Los Alamos Scien-M tific Laboratory (1970). I 19. K. D. Lathrop and F. W. Brinkley, "TWOTRAN-II - An
! Interfaced Exportable Version of the TWOTRAN Code nJ for Two-Dimensional Transport", LA-4848-MS, Los gj Alamos Scientific Laboratory (1973).
id
- 20. R. W. Hardie and W. W. Little, Jr., "3DB, A Three-l j' l
Dimensional Diffusion Theory Burnup Code", j BNWL-1264, Battelle Northwest Laboratory (1970). [J LJ 21. J. Moreir,t- "The Ef fective Delayed Neutron Fraction for ~' %j y th: FER", M. S. Project Report, Department of f1 7 (i - . . ym ..w w,. 6 .a=,- - - .~ ~~-. - ~ ~ -- -
. m ..: - . .. J- .3 ..-
+- .. -. - -- , , ,
.il q _
li - a ,. 57
~. j . )), "'
Nuclear Engineering, The University of Michigan (July 1981). m; -
.~;. 22. S. Kaplan and A. F. Henry, "An Experiment to Measure Ef- ;j , fective Delayed Neutron Fraction", WAPD-TM-209, Wes- .: tinghouse Electric Corporation (1960).
7" f 23. P. N. Cooper, K. Tirth, M. .<erridge, R. F. Mathams, A. J. Salmon and K. G. Stephens, "Some Measurements i].3.
-.. of Reactivity in a Light-Water Moderated Highly En- 'i .
riched Uranium Assembly", React. Sci. Tech., 16, 65 (1962). ,t. $~ If.tc ~ i*
.. .I .
, -~
~
lA
,)- ,
0
$ aus a i':*
4 I W
*5' t 's.
r." k , E l s M k,g fi t-
! ;}
g.. n-I h,. M
}
r. 4? !?
. g ___?;;}( %, ~a-
~
. _ , ' ,L' .
_ - k. .. a ., -s__
; % 97 . _ . . . _ , . - .. .__
4 ,. - ...
- i -
r3
;)
s., ij , APPENDIX A a, F1 . ,n 4 - y j' FNR Demonstration Experiments J L Part I
.i. ,
j . Beam Port Leakage Currents and Spectra
'l ti , ;,1 '.t 4_
il W / 1_
- s .
I w
-{ ~
1.' j -
- .i-i 34
,a',
.g LI
^$ ;
.l jis I
m
.a .s f
c'
.d 0:.
4 1 1, I. 2
)A.;f j.; y,-. .. , y e m 3 --- _
~ =- -~ ;' '-
W., *
,, . )
N .}; ' - l
, l
$ s
- FNR DEMONSTRATION EXPERIMENTS PART I: BEMI PORT LEAKAGE CURRENTS AND SPECTRA a :-
by i F {g ' D. K. Wehe and J. S. King Phoenix Memorial Laboratory ( j W University of Michigan l (. , Ann Arbor, Michigan 48109 l {* _ Overview The goal of the DIR-LEU experimental, program has been to .
- k. ~
measure the changes in numerous reactor characteristics when the l [ conventional HEU core is replaced by a complete LEU fueled core 7 a or by a single LEU element in the normal HEU core. We have ob-( ._ served comparisons in a) thermal flux intensity, spatial distri-6e : bution and cadmium ratios, both in the core and in the light and heavy water reflectors, b) fast flux intensity and spectral
; shape at a special element within the core, c) the thermal leak-
; age flux intensity at the exit positions of several beam ports g and its spe.ctral shape at one beam port, d)- shim and control rod A(
. worths, e) temperature coefficient of reactivity, and f) xenon poison worth, f..
t - y u The IWR is a 2MN light water pool reactor, reflected on
- three faces by light water and on one face by D 0, 2 composed of
) '_ MTR plate fuel elements. Figure 1 shows a plan view of the core y grid, D 20 reflector tank, and beam ports. The conventional HEU
' fuel element contains eighteen MTR Al plates 3.0" x 24" x 0.06".
p The center 0.02" of each plate is 93% U.235 enriched UA1 . x A j.* normal equilibrium HEU core loading is outlined in Figure 2.
; Each new HEU element contains a140 grams of U 235. The LEU low
{ enrichment fuel retains the same plate and element geometry but
- p. the fuel is contained in a central 0.03" thick UA1x matrix with j ,,~ 19.5% U 235 enrichment. Each new LEU element contains a 167.3 4 grams U 235.
In-core neutron fluxes were routinely mapped by a s-rhodium SPND and by many wire and foil activations. The same (, data, but in more restricted positions, were obtained through the light water reflector (south) - and D2 0 reflector tank (north) .
' [ ,..
Beam port leakage currents were measured during all power cycles, by transmission fission chambers at the exits of ports G,I, and - S J, by a BF3 detector at A-port, and by a prompt y detector at
," the F-port exit. Thermal neutron spectra for both HEU and LEU j cores were measured at I port using a single crystal silicon t.
diffractometer. These measurements p'rovide direct observation of the degree to which the LEU conversion alters the flux conditions for fixed 4,. h *
]L n
N =-- L = ~ -- - '$ 4 . J k - i i n .
! HEAVY WATER
.:! \ TANX -
- i @ @./
/
s N., ' 4/
~
N O O o'O O /' -
~- u O O O
/ O O O O O \
J
' / -
MMMS5MG! ' MMMetMSS O BSMMME 1
" ~
MSSM#M
!e SSM88BM e! ,
l! Fe!RGSMSSiei@! !
!i
!i WATER
~
d4 's, g 't . Fig. 1. FNR core geometry, showing fuel support
. matrix, heavy water tank, and beamports -
~
gius >t _.m__________-
.J
, ,. .. . ,., p.y . ; _
. s
,1 . e
+
;( '
'4
- 1,
,2 (i
g 48
'l \
4 li iu (a f. .d 'M 1 l d f} '.. N ':j ta: s. [. M .
,- 4 N i
1
<I 4
e- u y ,~ LC i Ltl i N L. { r Q - a l
$ I I
7; gu 4t i e! d y-ti Fig. 2. HEU normal core pattern. Soecial I elements noted by E -
- e. .
_ .P _ . _ . ._ . _ . _ - - , _ _ _ . , . , - - - - , . . . . , - - - - . _ - _ . - - .
~.
i . 4
.]
i g power, and at the same time provide a data base against which
'jy model predictions can be compared.
rj , These purposes are best fulfilled if the comparisons can be
] made when both HEU and LEU have approached equilibrium burn-up. } Unfortunately this has not yet been possible, since only HEU con-V ditions near equilibrium (1979-1981) have been available, while ~
1 only a nearly clean, unburned (N 2.0%) LEU core has been a-chieved. This means that until equilibrium a relatively small S (29-31 element) LEU core must be compared with the larger 38-39 _ element equilibrium HEU core. There is, of course, no reason in principal why this initial substantial geometry difference
.] cannot be included in the computer modeling program. However, Ji the difference in buckling, for the HEU and LEU cores measured, i must be kept in mind in evaluating LEU /HEU changes. To show d the importance of these buckling effects extensive data were -
obtained on special cores (which will be designated the "high leakage" HEU) which mocked up the small LEU cores as closely j as possible.* This was done by reproducing the LEU 5-row load-
) ing in the north-south direction and also the east-west dimen- .l sion adjacent to the D 0 2 tank along the north. face of the core. ?
There were several experimental difficulties encountered in obtaining reliable data. First, the conversion from HEU to LEU
- and back again extended over many months. Changes of only a per-
[ . cent or so in count rates or detector currents become Laportant 7 in such a time period. During this interval in the normal life Q of the operating FNR, changes in core and beam port instrumenta-c tion had to be minimired and monitored. Control over beam port changes was particularly difficult. As a result of pool water
- d. leakage, G-port must be pressurized and small changes in pres-J sure require occasional repressurization, resulting in a vari-l~. able water vapor density in the port collimation. Again, un-fi expected changes in A-port beam geometry occurred. As a con-
]
l] sequence G-port. data is somewhat suspect, and A-port data was (y available for only part of the experiments. 8 ,
.J Second, the l'itial n conversion to the LEU core was suffi-d ciently unpredictable in reactivity that fuel element additions d were necessary during the initial LEU experiments. As a conse-fj1 quence three LEU cores of slightly different loading geometry are reported herein; these we designate LEU (1 82 ) , LEU (1- 21-32 ) ,
{ Jj and LEU (4 82 ) . The core geometries for these are shown in $1 *This experiment was suggested by Mr. Gary Cook, to whom we are J indebted. M w
- _ . _ - - . = - - - - _ . _ - - .__ - .-_- . . .. .. . - . _ - -
l
..g ..n, .
y,. y . , ..
<_ .a i. - . i i
l l 1 l, -t 1- !i .,.\ d
.l . '
3 ,
- 1 i' e
y?- j
- 1
..g g h
j M CO i1
~
l - i w
]
E N . LLI ff ;
- i l i
1- . a 3
'g.
*i 1,g J r 1.. -
1 e
.t s3 f ej L- '1 q:
- ) t
.,3 . Fig. 3. LEU (1-0-02) core pattern 1 4 e
, , . - , - , , _ , , , , _ , . . , - - - - ,,n,-,,. ,- , , , , - - - ,,---
, - - - - . - - - , - _ , , . , , , , . - - - , - - , - - -.-n,,- -e.n.- v,---n-
- \
-- . . -b w wm - .i a ,_ . .% . _ , _ . _ . , _ _ , _ , , , _; g g' , i l
- )
l
- i. 1 l
*1 i
- $) . ,
-'4 l se
- 'a 4
I
'.1 9 .w
'l J g . 7 Ll l . z h g [ w _ s I] ,.
~
N -
'I 1 Q __I t '
. I, .,
-) i
-1 {
a
'i ;
3 __
+ .
i i
. ]-
- 9 j f 'e
.,t .,, i 1
1 , u .
" .t -;f Fig. 4. LEU (1-21-82) core cattern
'J
,J l
n ;
- i
-$ i i ll i '
, il ,. _e-. --e-- .-
..._.,,,a s- ;. _ _ _ _ . . , . - . - . ___ __
--- -- ww _- - . . . _a__ ,. .- w a __a -. - _a--_a a m ,___aa-m* 4 - .. ., aoum__-A-_Aau--a -m4-- -___,._;_..a.__%.%
4 1 , m ,
+, '
, -~~, ;. - m . . .a w , . . . , . _ _ ,
i
*O . >-4 g e't m
ia l l '. j '. '~ \ i . I j 8 k
- .; n.
1
- ) l
'l ' i o
,o 4;
- 4 k' 1 -
a
~.j
.i
- y g a
(,0 " lj. - a ~ if I l
- s
't 2 Q-L*
a O m 3 , I4 +- [
.' (.
~
,1 . i y L. n 4 5 e. II , 5-f), Fig. 5. LEU (4-16-82) core pattern . .i A 4 kk-0
- - . _ . , . _.-,, .- ,w__ ,.,r.. ... , . , _ , _,_,,,,,-__..n,____,.s-_,,,.,,,._,,
2._ -. . ._ _ a _._ _ - . . . - :c. __ _ _ . _ ~ _ _ . . _ _ , _ _ . . . _ . I
*a .- .
1
.t .. i
( \ j Figures 3,4, and S. The major difference in these cores, as is 5 evident in' the figures,'was a shif t from east to west in the -
) core loading pattern. This had a considerable effect, appa rent-q ly, in shifting the beam port leakage pattern, as will be dis-
;- cussed below.
n:. j Third, the D2 0 tank has presented several special problems. b* Access to the volume of the tank is very limited; it is possible k.) to reach only the upper region of the tank, the deepest penetra-li tion being 5" below the core fuel top level for the SPND detec-
% tor.and, 8" for wire activations. This requires a large extra- ]<
j polation from a position of maximum flux gradient to predict d (j data equivalent to core midplane. In addition there is now evi- _ l dance that the rhodium SPND response in the D 2 0 tank does not
] agree with either Fe or Rh wire activations when all three meas-d urements are normalized to measurements at the same point at the .
y center of the core. 1 . naam Port Leakage Currents ., l Accurate count rates were observed at the exit positions _ for G ,I , and J-ports during the period 8-19-81 through 8-1-82, l for A-port from 8-10-81 through 1-15-82, and for F-port from
') 4-14-82 through 5-19-82. Care was taken to wait for equilibrium
;d' xenon, and data were invariably used only when shim rods were ;
,' within 2"-4" of full withdrawal. All of these ports have source q planes either at the outer north face of the of the D 0 tank 2 (F,G,I) or are tangentially oriented 1050 from north and look q through reentrant voids to the central volume of the D 0 tank 2
j (A , J) . That is, the effective source planes for the latter are
,].
within 8" of the north-south tank center line, and within approx- ., 3 imately 4" of the D2 0 south face. Port axes for A,F, and J are i at core midplane, port axes for G, and I are 6" above midplane. j The currents monitored at A and G are seen only after Bragg re- _ q flection at nominal specific neutron energies of 0.06 and 0.072 ev h re,spectively, while those from J and F correspond to the full y leakage spectrum. I-port currents were also monochromatic, but ., 3 the intensities of all energies within the thermal leakage Max-
] wellian were probed. There was no indication that the effect: of d monochromatization influenced any of the intensity results. -y S ;j The comparison of leakage currents are recorded in Tables I .f and II for three core loading changes. The first two are given m fj in Table I and compare the effects of transforming from a repre-sentative HEU geometry to an LEU geometry, and then beck again P,2
?i in reverse order. In reading Table I it is not meaningful to
- tj compare the two HEU cores or the two LEU cores because for the .
h l1 J l i l l w- q - .. .; . - y . A, . . : ~ 9 ': .T ; ~ - a __.: .-- - ._ , , , _
- i - . : caw. . .. 3 gan g .%;i :9. [w s~ z-a e.t.uw , a : . .n.:a-n.;; m .. .w rs. . 2. u .
._. n; i '
F ! I g..s
. i
~'
TABLE I .- +
~
PORT llEU FALL 1981 LEU JAN.1982 AVG
- LEU '
RATIC OVERALL RATIC AVG .IIEU l AUG OCT NOV AVG 7s 1/8 1/21 AVG 7s AVG. 1 A 51.611.5 51.6fl.5 48.511.5 48.5fl.5 0.94f0.04 P . G 310.516 29814 28986 299fl0 296f4 298f4 297!4 0.99f0.04 0.98f.02 I I J 1260fl0 1210110 1242110 1237110 1223fl0 L250fl0 1237fl0 1.00f0.01 ;; HEU MAY 1982 LEU APRIL 1982 5/15, % c c AVG. LEU PORT 5/14 5/16 5/19 5/28 AVG /s 4/14 4/20 4/21 AVG /s RATIC . AVG. 4[i AVQ .IIEU '; F 8.1 8.0 7.95 8.0110 .lt ').5 910.7 9.59:0.2 1.208.03 G 275.3 275.5 276.6 277.5 276.2f4 324.0 326.3 322.C 324. lT5 1.17f.03 1.111.04 UI-I 77.4 77.5 77.0 77.3!2 87.5 87.512 1.13f.04 '). J 1249 1269 1263 1258 1260Il0 1181 1196 1206 1194Y10 0.95 f.Ol ! N
.h 3
,'t .
, N s
I
.h
- 9 *
.f,.j
't
'f5 I
-._ _wi;;.c. n . '. . . _'.MM. G . ;.21.ke '! 9 *>- 4 ' ;" A An '_ ra. va;&> *!" - E'4;hau.a.2N1.ibs'u;t. 4D A: _ ;,n/s.a ..
'd
+
s i TABLE II 1
~
"HIGH LEAKAGE'"
HEU MAY 1982 HEU JULY'82 , PORT 5/14 5/16 5/19 5/28 AVGfs 6/26 A N/s RATIC AVG.HEU OVERALL AVG. j G 275.3 275.5 276.6 277.5 276.214 290.2 290.214 1.050f0.021 1.0628.024 J 1249 1269 1263 1258 1260410 1353 1353-10 1.074#0.012
. i;. -
~ ~
e o I': I' t
;4 b 3
[ t e l-l } r. I I , 9
- _. - . ; _ . . . . ,. 3 .gg9, ,, 4 .,
. , 3 s ,{ .; .. t , f .. m a. .c. .
3 . ...
- t -
3, . J 4
~
- the G-port detector a recalibration was made. Care was taken, '
however, to leave all detector stations untouched between the
; .. HEU to LEU conversion in both comparisons. The April to May
]j LEU to HEU change shottid be most reliable since the data sets are separated by only one month in time. The first change com-f pares data for the HEU loading taken in August, October and No-j vember, 1981, with that for the initial two loadings of LEU, j shown in Figures 3 and 4, taken in January, 1982. The reverse q change compares data from the LEU core of April, 1982 (Figure 5) and the reinstalled HEU core of May,1982 (Fig. 2) . The two HEU 0 cores are closely equivalent to each other and to earlier HEU j- data reported from September,1979; they differ only in the sub-l stitution of a "special". element in place of a regular element 3 on the south face, as shown present in Figure 2. The LEU cores
' ,- , were somewhat different as needed to meet reactivity require-
~
'] ,
ments, as noted earlier. As may be seen by comparing Figures 3, 4, and 5 there is a. gradual shift toward the west face, partic-j- ularly between January and April, 1982 and an increase in total 4 number of fuel elements. 1 [-
- The data of Table I for the January LEU cores show that within counting statistics and reproducibility in time, there
.i a was no change in leakage intensities from all three monitor h- stations, the average ratios of LEU to HEU levels being 0.94 10.04, 0.99 20.04, and 1.0010.01, giving an overall ratio of 0. 9810.02.
1 By contrast, the reverse transformation of LEU to EEU which }1 , occurred between April and May,1982 shows a significant increase in leakage for the LEU over the HEU loadings, for three of the j four bean ports. These show LEU /HEU ratios of 1.20 f.03,1.17 f.03, l, and 1.13 i.04 for F ,G , and I-ports. At the same time J-port f shows a loss in leakage ratio of 0.95 1.01. A crude estimate of ?j the change in thermal flux per unit volume of the D 20 tank can be 4, obtained from the average of these form stations. This gives
-1.11 .04. While there has been a possible question about the Y p.tise channel electronics at J-port in April, we are inclined' to (s believe all the data are reliable. We observe two effects (a) p a possible shift in leakage from east to west, consistent with h the LEU loading pattern of April,1982, and (b) an average in '
1 crease in leakage flux from the LEU core of at least 11%. The
- loss of A-port data to verify the shift is particularly unfor-
,4 } tunate.
q To estimate the effect of the higher north-south buckling for f the LEU cores a 5 row "high leakage" HEU, Figure 6, was compared
?
h
_ - . . -- _ -- - . _. _. - - -- . ._. =-. _ ,_ .....-.-: .---w~~~~- -- - t. .0
- r,
.e 1 5 1 l
, i i
!] .
'j j
- .! l :
l .. N
;4 5j 7 e LLl i <g* L
.. Z W
- O i-N -
E N/
- /; Q A
( Ji W
~
I, d - s 1 0 +: .. i 4
] -s 9 -
i E. (
,j Fig. 6. "High Leakage" HEU core pattern .- !
,. 4 (7-7-02) !? w
,t s.
1 - I l e
- ..c- h'~ . - _ . , . , . . . _ - . . . " - -- +----~ - - - - - - -
- amm..wiI.m m_.. j., g , .. g,g, , ,, ,, ., m , . , ; s,,
..-.Sh
. . s 1,
a '
}a U-9 -
j- with the HEU loading of Figure 2. In addition to extensive wire j activation and SEND data taken in this core, beam port leakage currents are shown in Table II. As expected an increase was j- found and was about 6%. . The data of Tables I and II are not simple to interpret, and j~ the correlation with in-core data, to be discussed below, is only partially satisfactory. We draw the following tentative conclu-
}n sions:
3-i a) Installation of a full, unburned LEU core in place of an e equilibrium HEU core will change beam port leakage by a factor between 0.95 and 1.15. I b) A significant fraction of this gain, perhaps the major i fraction, occurs because of the enhanced' north-south leakage
,- associated with the clean LEU geometry. This gain would be ex-pected to be reduced as the larger LEU equilibrium core is reached.
1 c) Interpretation of leakage currents from a D2 0 reflector with
,- reentrant oeam port voids is dependent both on the position and
, angle of beam departure from the heavy water reflector. There is i
evidence, for example, that J-port leakage currents closely fol-low in-core changes as measured in the outermost lattice po'sition 3, L-35 by SPND. (L-35 is the element adjacent to the D 20 tank and
.j midway, east to west, between control rod special elements). If 9
both sets of data from the January, 1982 LEU core are normalized to unity the changes in J-port current and L-35 SPND response for i' successive LEU and HEU cores, as shown in Table III, are remark-ably cloce. At the same time, the G-port current appears to com-( pare not with in-core results, but .with SEND data taken' in the 3 outer (northern) volume of the D 20 tank (position D 0-S) 2 . . This
~
" correlation", though less impressive is also shown in Table III.
" It is evident, finally, that changes in the inner D O 2 position, D2 0-X, do not correlate well with'either'J- or G-port data. -
l , - Beam Port Thermal !?eutron Spectrum Changes q The leakage spectrum at beamport I was measured.for both the q equilibrium HEU core (November, 1981) and the LEU core of i April, 1982. This was undertaken to determine whether any signi-
, ficant thermal neutron spectrum hardening of the leakage current y could be detected for the low enrichment design. Measurements d. ' were made at I-port by use of a single silicon crystal diffrac-e l tometer. Flux intensity as a function of energy, from E = 0.02 ev to 0.140 ev, was obtained by a conventional 6-26 Bragg scattering j)q i survey. A very narrow mosaic silicon crystal was used in a trans-mission moder the [111} planes were used to remove second order contamination. Considerable attention was given to the crystal 1 ,i .
m
-- ;L -- .u ks a .. : a :.s. L.:L . . ;, a%.O .i.,1.a: .w. , ; .._;_., . . w:s:
.._:;c.. .= .2,; . . ..; x;: .. G.
j t TABLE III -
'g CORE CONDITION J-PORT SPND G-PORT SPND SPND L-35 D 0-S -
D 0-X-2 2 l LEU Jan 1982 1.0 1.0* ' l.0 1.0 1.0 - LEU Apr 1982 0.965 0.970 1.091 1.097. 0.987 HEU May 1982 1.019 1.004 0.93 - 0.876 ,
"H.L." HEU July 1982 1.'094 1.108 0.98 -
0.934 i f
- LEU Jan 1982 Current for SPND in I.-35 taken as the average '
of values measured for LEU-1-8-82 a.nd LEU-1-21-82 k
- . t.g
. f .' .
i j j * {
') i
$ e l t e e e
- n-p.m --
A _,. 3 _ .. :._ -
, y, j p, g ,
- g. .g,_..-.c:w,e . . c
)
y-1
=
A
)-
1 collimation allignment to guarantee that all of the Bragg beam, in both the vertical and horizontal dimensions, was detected at j each energy. Integral Bragg peak counts were obtained first by j" exact centering of the Bragg beam through very narrow vertical, then narrow horizontal slits, then opening the apertures to in-j] tegrate the total Bragg beam. The detector was a transmission A~ fission chamber of efficiency 6-10-3 Such a detector allows 2g accurate energy dependent efficiency corrections because these ji are simply proportional to the known fission cross nection for 4 U 7.35. Background was determined as the count rate when the y crystal was rotated 20 from the Bragg condition. The flux as a i ~ function of angle, and therefore energy, was determined from the
'. net count rate,according to C.R. (G) = }C ' E(G) *S(GM)
- hl6] (1) s" where K is a geometry constant,6(6)is the detector efficiency, j
and R(G,h) is the integrated reflectivity of silicon in trans-mission. This last term is a well known function from kinematic l crystal theory which depends on crystal thickness t, silicon cell j geometry, and a mosaic width parameter %, according to the rela-l R (e,9r) = [ 9t e ~ N*f (-/)'" \J l: i =i 5 nrae.e ) '
\ ... Q = lF(LLt)l2 A,-
R .es 2 O f,. j F (hk1) is the crystal structure factor and Ng silicon cubic unit cells per unit volume. ia the number ofc [:, Although the leakage spectrum is well characterized by a Maxwellian of neutron temperature T, the weakness'of the method j lies in the need to calculate R( @/rt) . The absolute temperatures j.- , obtained is sensitive to the choice of the mosaic parameter M. j Enile % was included as a parameter in the least square fitting 3 of the data, it was foun'd to have a shallow minimun and is there-j - fore somewhat suspect. This in turn makes the absohte tempera-
- b. ture also suspect. However, it is believed that sms.11' tempera-f;l t ture changes are reliably detected by the method.
Data points were obtained throughout the energy interval
;4 noted above and the neutron temperature obtained' by a fit to the Maxwellian function
-$r \
&lE) = A 'E 6 z3; e,
1 s-I k.
. l 'l _
~ .
.."s...
...e...
j .'. " tt
- .1 4
0 N Q where A is a constant and k is the Boltzmann constant. The data s'i is plotted as A Od- r.f E , from which kT is determined, through [.j a least square fitting procedure, as the inverse of the slope of y the plot. W -
!';l Figures 7 and 8 show the resulting plots for the HEU and LEU
} respectively. The data are extremely well fit by straight l'ines between 0.02 ev and 0.13 ev. The temperatures obtained are 373.0 2.40 K and 370.811.90 K respectively. The difference is well within error limits, so that we are 1.ed to conclude that no apparent spectrum change is evident at I-port. The temperatures measured are not to be considered accurate on an absolute basis, but the relative values should be reliable. .
It is evident that the temperature observed will depend on - t the effective source volume in the D 20 tank seen by the port col- _ limation. I-port views the D2 0 tank from the NE corner face. The I-port axial line extends almost diagonally from NE to SW
;~ across the tank, making an angle of 630 with the north axis. . ,' I-port does not have a reentrant void.
It is expected that the spectrum will soften as the port . - PJ axis moves away from the normal (north) . In July,1982 (Figure 6), , 7 the diffractometer was moved to J-port where the angle of attack is 1050 A scan was repeated using the same geometry and sys- - q tematics. The temperature observed was indeed lower and was . O found to be 330*SOK. Unfortunately, this measurement was not u repeated for the standard HEU and LEU cores, so that a comparison "
!j.
with I-port has not yet been possible at J-port. Si b d a' 4 i
.i.
1 . ~1 4 n w I a O [:, - iQ . LV >s
., ~ '
q ; - -
ii -
- - _ HEU SPECTRUM
'i 24'0 I PORT ' -
SILICON t'ili} DIFFRACTION
- m 7/=1.80 10-4 j,
T=373.0 2.4" K -
. i. - _
q - i- 23.0 - ,q .
. i .-- Lt! -
v b! -
\ --
Q . - i Z - ~ g _; - 22.0 i - l 1
- - -l a.. -
y - r -
! 21.0 -
,:1. - g - 3- - a
- j -
20.0 ' ' ' ' O .02 .04 .06 .08 .10 .12 .14 .16 .1 f. o E(ev) . I ', [j.. Fig. 7. HEU thermal neutron leakage spectrun il .m _ _ . - . _
> . - .. = . .. a - .. . . - . ..
3 q .. . 1 . c.,J a i , i i . . . i i i . . i i i i i B - LEU SPECTRUM I : I PORT : - 240 - -- f SILICON {iti} DIFFRACTION . g : 'r/=1.80 10-4 :
- T=370.8t1.9 K : -
[
~
9-. [ f g .
=
23.0 [1] _ . _ o LLI _ _
.i Z - -
- ] J c _ i 5 22.0 - -
s 4 ~ - .;J
- j - -
.1 t .: _ 9 : : il 21.0 - - 3
~
. 0 -
n! y . _ _ '. a ! $ 200 ! g O .02 .04 .06 .08 .10 .12 .14 .16 .18 -[ ; R E(ev). ~
- -, Fig. 8. LEU thermal neutron leakage spectrum !
!.i ' ^*
L
~_. ,
g.. ,3.;. 1 et.
.. i...s -
.-5 ~ ,.-:....- ..-u- . ~ ': -
ng_.m g. .,-.m. --y.,_: ,,. -? ~ : .. '- ' ~ ~~~~ ~ <
- m. . _._s-. _ ,
,3 . -
- h a
J a
'2 ;4 (1 APPENDIX B n 'l -
g i FNR Demonstration Experiments
~1 L L Part II i Subcadmium Neutron Flux Measurements n
c4
- 4-p.-f 6 ., . e I.)
.a .1 '1 ') 6 .,j - .?
- i.
N
-t . .
4
.k '
i 1
~ * .
s s , 3 t b
'1.
4
' 4I.
d .. i4
.{
T.
'g .
d 1
'~
- . j 4
.d, l- _
? i
- . 4, we l Ng l i, f
!.i EJ.
'J Jj
' f. lI
- ? -
l _i . ;
& i. EbbC.S. W n: '.e - ---,Y '
-- - - - - - ~ ~ '~~ - - ' "
,a ._ .
. m 1 . .
')
p
- i m
]n FNR DEMONSTRATION EXPERIMDITS O PART II: SUSCADMIUM NEUTRON FLUX MEASUREMCITS d by .j - D. K. Wehe and J. S. King ,Ay Phoenix Memorial Laboratory 4 University of Michigan $ Ann Arbor, Michigan 48109 M %! Introduction m -
- ] The FNR HEU-LEU Demonstration Experiments include a compre-f hensive set of' experiments to identify and quantify significant
( operational differences between two nuclear fuel enrichments. 1
- ,, One aspect of these measurements, the subcadmium flux profiling,
- lj is the subject of this paper. The flux profiling effort has been
,i accomplished through foil and wire activations, and by rhodium I self-powered neutron detector (SPND) mappings. I [ - J Activation Data
- Techniques - -
J - . _ . - - The irradiation of wires and foils 'in and around the FNR * !b core provides information on the reactor flux. Irradiations in ' the core are made by taping the probe material to a thin (R/.010") aluminum paddle approximately 30" long. In some cases, samples are enclosed by .020* cadmium capsules or tubing. The bare and 4 () ~ cadmium covered probe materials are irradiated simultaneously, - mounted at the same core height, and separated in the horizontal 4.j plane by about an inch. The paddles are curved to facilitate in-Q sortion between two fuel plates (separation distance is. 117") . A f' paddle stop rests on the top of the fuel plate and provides the ax-1 ial reference point for the samples.
~
q,
?.,
The heavy water tank on the north side of the core contains . % a dozen 1" diameter, vertical tubes which penetrate into the tank. d The majority of these penetrate to 8" below the top of the fuel plate. While these tubes are filled with H20, calculations indi-hl - .j cate that the measurements are representative, within a few percent,
- l of an unperturbed D20 environment. Samples are activated in these '
d tank penetrations by first securing the material to the outside of Al y a 5/8" diameter aluminum rod or' tube, and then lowering the holder to the bottom of the tank's vertical penetration. The samples are ~ f rotated during the irradiation to ensure uniform activations. In d all cases, the reactor must be suberitical during both sample in-k.f sortion and removal. - 0 4' l. g e i h:,w . +.,m~-- = - - ~ - - - ~ ~ -- -
, y s i I
~, a . u.. .L - ,: -
_ y . !.L +. . y' . . . .,.u . ; q . -. . - t _. ... O n_ ' - 0 3 5 d-N Post-irradiation Handling and Counting d
;! After the irradiations, the handling of the samples depends s~ upon the activity and half-life of the activated material. For fi long-lived isotopes, such as Mn-54, the material is normally < ]_ stored in the pool until ready for counting. For short-l'ived fj isotopes, such as Rh-104m, the samples must be expeditiously q and remotely prepared for counting.
9 _.
.i . The counting of the activated samples is performed using j GeLi detectors. Wire samples 'are' counted 'between twc oppositely 9 ._. facing detectors multiplexed together. The sample is positioned by an automatic sample changer into a rotating, cylindrical plexi-i glass holder. Pulse pileup losses are accounted for with a pre-M- cision pulser fed into the GeLi preamplifier. The amplified and d "' multiplexed signals are counted using an ND 570 ADC and fed into an ND 6620 analyzer / computer for analysis. Absolute efficiencies can be determined with NBS and Amersham mixed point source stan-
- dards through a cross calibration technique at a separate GeLi.
detector station. Background interference is made negligible for 1 most gn=== ray energies with 2-6" lead shielding around all-de-tectors. y ,
}-
The counting data is processed to give a saturated activity
; per unit nucleus, Asatn, through the relationship:
3 3 3 1 - 9^ a
- A ACot(t.J(W""de*d3h t c= %l m
- . sain = g_3-atp\_ gw; e s-**R BR'm-% N ,,/ /&7 A L i o
i where: A = decay constant
- q Cnet = net counts observed '
j t1 = detector live time
= detector real time tr
}ii - ti = irradiation time E = absolute detector efficiency BR = branching ratio tw = time between irradiation and counting h.i 'AW = atomic weight of element ?jj Navo = Avogadro's number hi a/o = atom percent of parent isotope
- i m y; = sample weight ,
1 {y Once the data is converted to saturated activity per nucleus, it i s can be further processed to give flux data. The difficulty is that there is no unique method for translating activities into i
- / 4 l
l I' i L _
~ ^
;. l. , l_ i-
..m zu.w w2 s *.c ;w> ,
, .e s.
i
.1 l
a
-9
- t flux. Different approaches yield fluxes which can differ sub-
$j .stantially in their magnitudes. Two separate techniques are sj presented below. j 1) . Beckurts and Wirtz Approach. (Reference 1) . --
'j The thermal flux can be determined from bare and cadmium covered -
1 activations according to:
;j % enJew .
A # - L A +n 3 i 4_O & s (2)
?
- 1 6* T r .
1 where Fed is a material dependent correction to account for the kj opithermal activity produced below the cadmium cutoff energy. ,
?
(Formal definitions of the thermal cutoff energy {Etc,%0.1ev , :
' and the cadmium cutoff energy (Ecc $30.55 ev} , can be found in !
references 1 and 2). To and T are t}ie temperatures corresponding - L4 to neutron velocities at 2200 m/sec and at the most probable j, Maxwellian energy. The flux spectrum is assumed to change smooth-l 1 ly from Maxwellian to 1/E through the use of a. joining function. -- l The constant flux per unit lethargy expected in the epithermal
"} region is determined from the cadmium covered irradiation, and ~
A 1 this is used to infer the flux between Etc and Ecc. If this flux - is termed h t, i.e. " intermediate thermal", then one can define ' the sum =d=4um flux to be: - ksc." kn
- kit:
.; 2). Effective Cross Section Approach. - . ,
N By defining effective group cross sections, one can deter- l
,' mine group fluxes in a conventional multi-group spectrum calcu-
- lation. Define an effective activation cross section <Qby
- r p.
=
g 0 tr(E)$EN ~ f0 e Q(E)dE Et -
.) where d (E) is a computed spectrum. Then the group flux between 3 Ea and Eb is: -
0
~
gb Q tAJew um g f Q '~ &
.i -
4(E)dE * - c)E*- s M Al fj j
- er .
q _
~
)
m .> ; jiR?; ; . , m a w .t :.=. -.: . . . - . w-
., - .- f.-. - 4 9 ., ,
ha n 1-S The subcadmium flux is then determined by choosing the limits of
.la integration to be O to Ecc.
f]s. ' - 3 In both approaches, some knowledge of the spectrum must be ! d known a priori. The Beckurts and Wirtz approach was used through- !j -: out reference 2. The present paper, however, makes use of the l second approach, which is believed to be an improved treatment. d]e pj The spectral calculations needed to determine the effective cross sections are discussed in a separate paper presented in this con-A ference. - 4 . 7 Iron Wire Activation Data k- Irradiation of iron wires in the FNR yields two useful re-3i actions: Fe-58 (n,I)Fe-59, and Fe-54 (n,p)Mn-54. Activation data fy from the first reaction are used to measure flux in the thermal 4 regime, while data from the second (threshold) reaction respond to fast flux near the fission regime. Pertinent material data j ' are summarized in Table I. The long half lives of these daugh-ters obviates the necessity for rapid handling. The irradiations ; 1 are performed at equilibrium xenon, with the three FNR shim, rods typically 85% or more withdrawn from the core. Typically, 29" lengths of bare and cadmium covered wire ^ are irradiated at full ' Power for an' hour, then cleaned, cut into one inch segments, and j coiled to simulate point sources. j To illustrate the quality of typical iron wire results, Figure 1 presents axial flux data at the center (L-37) of three ] different FNR cores. The three cases have been normalized to 4- unity to emphasize the close similarity in axial profile typify- ] ' ing all our HEU-LEU comparisons. Error limits, both vertical " and horizontal, as well as profile smoothness are also typical. ! Figure 2 compares some of the same iron wire results with rhodium - SPND measurements to be described below. Again all profiles have ' ]1- been normalized to unity at the core center. The close agreement between the two techniques when normalized together 'is quite sat-isfactory,.we believe.
}- Rhodium Wire Activation Data
] {' The activation and counting of rhodium wire is quite differ- };, i ont from that of iron wire, as can be inferred from Table I. l j' Rhodium, because of its large cross section and short half life, must be handled carefully, and yet, expeditiously. Because of J, . lack of accurate beta counting equipment available near the re-J actor, and the difficulty in beta counting multiple samples J, quickly,'the weak 555 kev gamma was used to measure the reaction. ) 2 r \ 'k h
s .::w :.a .
... . . . .. &a . . . . . , : . . :, >
~
m - :. ,. . .w.u. .: L a w .;w m s .w u.a.[.. _ sa.hL.:2 2.. I.
. E t
I e t TABLE I Activation Material Data , Ef Ecctive Subcadmium Cross Section Gamma 40"> ba rns
- Atomic Isotpic Ene rgy Brancb IIalf Core D02 1I2 0 Normal IRRAD. IRRAD.
Reaction Weight Abund. (kev) Ratio Life Center (X) Reflect. Diameter Time Power 44.56 Pn5"M Fe 59 55.85 0.29% 1099 56 . 5'X days .911 .976 -
.020" I hr. 2MW ,
54 54 312.2 :' Fa Mg 55.85 0.29% 835 100% days - - -
.020" I hr. 2MW
- Rb --+ Rk 102.91 100% 555 2.0% 43 sec. 104.5 109.4 112.5 .020" 30 min. 20kW 103 104m 4 RFi --+ Rh 102.91 100% - - 4.28 min. 8.58 8.98 9.23 .020" 30 min. 20kW cThora values are based on preliminary spectral calculations and are subject .to refinement.
f.
- t e
e D a O . w .. % $
4=o . i O 0 0
. . .O .O 0 0 0 0 0 1 l .;
0 1 2 3 4 5 6 7 8 9 0 l F 2 i - - - - - - - - - - q. g _ u 8o[og . r 3 e 3 ' _l :_ , EA 4 ' l x ~ ei : ma el n 5 ' . tF l Lu . 3 ,x 6 ' i 7 h D a ,c s 7 ' t . aa . N t 8 oC . ro . mr ae l 9 ' i,C . ze en i . dt 1 e 0 r I I
- - a i i
I
- a 1
'1l1i I ,i 1 i l i I ,iiI 2 II11, !l Ell 1
o0Ew wzr .n_tZw J 3 l 1 4 l i 1 5 l m i 1
' l 6 .
XA0 1 7 LHH l i .
.a EEE -
1 UUU .
' (((
8 ia l 495
/// a 1
' 887 l 9 229
))) i
~
2 ' O 0 l 2 1
' l i
2 O .n , n. 2 > . k ' >t. i j. ) .: t, 4 .,
~
tf,- -
. : # e.
[.p. xm-u:an :... . ... ~ n .. . x . u.a:. . .e. 2 .= ;a 2a;...xama.azra . a. z:- .
) ,
i i i i i g a i i i i i i i i i i i i i i i i i 3 O o
" b . O HEU SPND 5/79 F a -
A HEU FE 5/79 o- . X LEU SPND 1/82 m s,'
.!j -
')
~
i 1.0 - - - -
~~
.9 -
; .8 -
s ! A 2ss ' 1 .7 - O N i m < s s
& .6 X -
\
s' p- - A I .5 A '- F
.' \ -
.4
\
/ ,.
_ e M
.3 - -
y _ e
.2 -
T p.
.j
- 1. .
-l O h 1 2 34 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 ; .4 Figure 2. Axial Flux Profil'es in L-37 Measured by SIND and Pm MOCOE * ..; ' .!
Wire Activation. Data
8 3
. tin li? -
* ' [
L _ _ _ _ _
e ' ,e + .
~'Q.
. ($g'M % :- -.
. m .- .y - - ,--w~ -~ -~ - ~ ~ - - -- - ~'
4 y .c - I g.
- i-.
N~ j._ The rhodium wire dimensions were chosen to match the emitter dimensions of Rh SIMD. Typically, one inch lengths of 99.99% pure, bare and Cd-cov-ered wires were irradiated at 20 kW for 30 minutes, in and around
~
the core. Power normalizations were determined by monitoring t the leakage flux at beam port J. (The beam port geometry is de-( a scribed in Part I of this paper.) In several separate experiments, the observed leakage intensity at J-port (and G-port as well) showed remarkable agreement with reactor power (as determined by j the FNR operational fission chambers) from 20 kw to 2 MW. Post 0 irradiation handling involved removing and cutting the wires into y - 1/4" segments remotely, drying the samples, and transporting ?. hem j to the counting facility. clock. All times were calibrated to a sin >J1e M-N e 1). Activation Kine' tics. b The rhodium activation and simplified decap. scheme is sum-p marized in figure 3. y.] Defining: C""= effective production cross section, - N = isotopic number density per Rh-103 nucleus, y j
,_ A = activity, A = decay constant ] and using subscripts.g and m to refer to the ground and meta-9- stable states, one finds that:
3
, a) during the irradiation:
g, g -2mt
.) < ,i a
4/t)= - Am I- 6 (3)
-2 t O
>mt -
T. AyId = (1-ED+ %k 1.- b b J (4) p$ .- 3 Equations 3 and 4 show that for an irradiation of 30 minutes,
~
both the ground and metastable states are saturated. b) after the irradiation: i._ . q h< n t)3(t) = g #Y g # % L y -a-x [
@EwJ -
- s. .
E i 4 '4. 1 -
c.. . - -a=u w.u:.t-a _..m.
.. r . :. . w.w.. ._;: a .:._.
+w.= .. ., . . . a . . w , .-
- u. w.antu.
. ,x:a.,a.,,a
. . ,., ..aw a s. .a c u.a w r
l Rh-iO3 11 barns
?
Rh-lO4m (4.3-4.4 min) 134 barns i Rh-lO4 (43 sec.) .- : 555 kev (2%) ( Pd-lO4 t,
.y Ci
- c. ,
[J.
'E Figure 3. Rhodium Decay Scheme. l
- :. .- y .'
... u.
)
e * .O 4,_ . - f J I *
! . 3 !
_j ! i
- I ' '
. , _ . e '
.. -- , + - -- "--
- . ,, . n. , ny y,. nm.,gm3 , ,3 y ; .,y 3
m a, ,_ )
- i
$' Since the minimum wait time is 10 minutes, the contribution from the ground state term is negligible, so that Au 1 - ] k(f)= 3 h-} G k
- 6 ht s.2 F
Thus, although the ground state decay is detected, the decay is governed by the half life of the metastable state. Furthermore', the effective cross section must be defined as: 1 sees a F Ecc, i (qW _ %(EN(EldE . yew - f i 4 whichwehaveassumedtobeequivalent,,to:f(.h , , %lEh qtr9 [ E)J e 19 c, (F,') J (Tg ) , ,
'< s (5)
? _
}$- h Cm !M+ 7g (E ) [ ,
h WWE j where Eo is a fixed energy in the thermal spectrum. This cross a .1 section is shown in Table I for different media.
- 2) . Epithermal Correction Factors for the Rhodium SPND
," As is discussed later, the rhodium SPND responds to neutrons j of all energies. If the subcadmium flux is desired, then it is necessary to know the detector current corresponding to subcad-
]_ mium neutrons. This is accomplished by measuring the rhodium ] subcadmium fractions,. fth, for rhodium wire with the same diameter i as that in the rhodium SPND. The results are shown in Table II
- 1. for locations in and around the equilibrium HEU and fresh LEU d cores. The data indicate tliat the flux is harder for the LEU f't.
fue1. 3 a 3.) Rhodium Subcadmium Flux Measurements el Rhodium activation data can also be used to determine the [~ subcadmium flux intensity and thus provide an independent check !! on other profile methods. This is more difficult because of the y~ problems of self-shielding (estimated to be 20-30%) and flux de-prossion (calculated to be of order 5%), as discussed in reference { 3. Even for relative fluxes, differences in the self-shielding g are medium dependent and must be considered. However, between 9 t the D20 reflector and the central core regicm, this difference is calculated to be only about 5%. Relative subcadmium fluxes from k R rhodium wire activations are presented below. .1 1 - 3 'l k
+
, - .-.a__._.
i.2 _ . . . . . _ . _ , . . . _ . _-. au. . .2.._3u 4 4 y .. TABLE II- .j MEASURED SUBCADMIUM CORRECTION FACTORS, f th
- d 4
- .4 Position fg HEU f th b
d 1 b35 Regular Fuel Element .786 ' [:) (Core Boundary, North Face) . t1
.791
[,-] b37 Regular Fuel Element .749 l-] (Core Center) y
; b39 Regular Fuel Element
~
.795 (LEU South Face) b40 Regular Fuel Element .'830 (HEU South Face) ,
b67 Regular Fuel Element . 830 - [ (Second Column from West Face) ,
~
~.1 c!
b57 Regular Fuel Element .860 ' (Third Column from West Face) - 1 HO 2 Water Reflector, Second Channel . 930 .914 (Center Plane, South Face) _
,q*. o q DO 2 Heavy Water Reflector (Position X) . 895 .892 (Center Plane, North Face) ,,
1
, b39 Special Fuel Element (waterhole) . 913 d
h n A
.:4 d.
n
;7.
.i
- ' il 9 W.
1 4 9
~""'
9' w uuMM Ms er , em M. a . sea.., -es% M M-.*M*'- "
" ' " ' ^
. . .. . .- .. .- --- ~ . . - _ _ . . - - - ._
.; _, 3 7 . ,,, .,. m ,,,g . .g .
,,,. +- & a -
7 - e. a .
- g
- j ,
l R
~ ,_
j Rhodium Self-Powered Neutron Detector Data Rhodium SPND Characteristics
.11 The rhodium SPND has served as the primary flux profiling l L. tool at the FNR. The detector probe, shown in Figure 4 uses a ,Q 20 mil diameter, one inch rhodium emitter insulated from the 1/16" 4 outer diameter inconal sheath with aluminum oxide. A parallel .
H background lead, not shown in the figure, is used to correct for
;l background effects. The probe is mounted on an inconel 600 paddle
(.093" x .625" x 36") with a 1/4" x 1.5" liole around the emitter
$]- 'e section to minimize flux perturbations. The probe can be posi-tioned at any height in either of two water channels in any fuel assembly. Special adapters have been fabricated to permit men- -
- .i.
surements in H2O and D20 reflectors.
~
i. q While rhodium has the largest sensitivity to thermal neu-trons of any commercially available SPND, it also possesses two
]- , principal disadvantages:
i " (i) it responds significantly to epicadmium neutrons -- because of its 5000 barn resonance at 1.25 ev, and , (ii) the presence of a antastable state (4.3-4.4 minute half life) requires waiting several minutes before an equili-
.~ brium signal can be measured.
The first of these problems is overcome by measuring the subcadmium contribution to the detector reaction rate, as de-
] <- scribed above and evaluated in Table II. As shown in reference 4, j l the subcadmium flux can be calculated from: ? [
ka Inet j , IO) !
] wherer f th = subcadmium fraction of the detector current, '
q I nst = not detector current 1
"g < _
S = detector sensitivity to subcadmium neutrons. 4
- The detector sensitivity can be calculated using the methods de- !
4 scribed in reference 5, or calibrated through the use of the ac- ' tivation data described earlier. The sensitivity is related to j the emitter reaction rate per unit incident flux, so that if k is a: a proportionality constant, we may define a sensitivity factor N'~ 9 1l~
= .=-...-- . . . . . . . . .
. _ .-.-x-w w.a i.,O.
~ .
1,q . . d'l 5, s. N
.4'r.] ,
..I J
,.9
.j
-u -
- a. .
.a, .
A ..
* ~
.M.
. Inconel *
.4 7 -
Core :
.'j Aluminum 0xide Inconel 'l Insulator Collector
, e .
- . v n.. . . m . nrL
,, ,,,, ,w.a
.1
/ ,
- 3 Baitter Insulation Inconel
- n .
Sheath .
/2
,1 i Figure k. SPND Construction
- n . .
id.. .,. . . Nj . b, s-. a . Jq e.s g; . .
- J '
.e -
.a-e
.s .
..a . 4 . g . . .
~ . . ; -. g -
e . . g . . ~., . . o
~
. . ... a
.t 1 . .
' =
.- ,= - -
_ = . ._
.k5 ,
- 7. g . _.. . , ,
-----.),g.pp ; pgy ,smj ~ph/gp.4(q p-m ,
b f) which is applicable to any iven spatial volume (core, D20 or j H2O reflector) y p ar. 0- d((oE)C'lE) d5 d5 ?y a Il b* k V O 47) [ y . y , 4.Mo(0dEfd C J j' where fg is the depressed and shielded flux in the detector, J i. is the flux present without paddle or. detector, at the vol-jt une position being measured,and the integration is over the emit-ter volume. A term similar to the one in brackets is calculated v
- h. in a manner described in a later paper presented at this conference.
9 (t It should be noted that there are many physical factors, such 6 as @ and f behavior in the detector and detector leads, included in the constant k. These factors may not be identical for core, D20, and reflector regions. We assume, for the present, these y differences are small. Regardless of such possible variations, 4 as well as the differences in flux depression defined by equa- ' b, tion (7), we have elected in this paper to pr !1 suits using the constant value of 5 = 3.0 10" gentamps all' SPND ~re-
/ny, recom- .
y , mended by the manufacturer, reference 7., The second obstacle, the delayed response of the detec or, O . can be handled through the use of analytic techniques described it in reference 6. All of the data presented in this paper were P .< obtained from an equilibrium detector signal. l 1,[ The quality of SPND profiles is illustrated by two of the ih curves of Figure 2. There is generally greater smoothness than h q for the activation data, but nearly the same axial resolution. -
- g. u -
The latter is fixed by the 1.0" emitter length. ).l ,,'
'1 Experimental Results and Interpretations a0
] single LEU Element Replacement s J In a single element replacement experiment fresh HEU and LEU
- d. elements are alternately placed at the core center in an equilib-
? rium HEU core. Iron wire activations were made along the full Q a.xial length of the elements, but for this experiment only bare
- /
wire activations were made. Figure 5 shows the core geometry I' used. For operational convenience the fresh elements were simply
- j. interchanged between core center (L-37) and core edge (I,.40) .
t e Ia , , p. [I e
! * , , . i f ! ' , ! ;
J
.-t iI't' : j.
.rtff; .! i5!
l s .' . 4 4 c
=
4 1 1 t "s. , yn ko re t mt e een n
- ml e oE m 3'A e e
** Gec
. l a egl s,
rnp oie CSR . _ .J. J . 4 e r
'e. u g . d. .
i 2 3 F - a J 1 8 K
*g -
a. e 5 e
=-
N 7t N\
\
\
1 A N, [y y n \\ h
/ gN'g\ 0 .~-
N . , 1 U. 0
' i, .
E
. 2 F.
D
. ~
L. . b.. - o.. , 4 4 _ a. M ~ . M h h - 4 . 4 4 [ I i! ; .i!i. , fi l! . l' ' f . ;1 .l ' i , i " .
- ~ . - - . . - . .- - -
gy . n.x _ :.m : =~.c.pz.;.:pppym qvz yp.&r;y,y N3-i+ y ? I . l q -
- a. -
j^ The reactivity change associated with this interchange is dis- , l cussed in a separate paper in this conference. An average was i made of the saturated activities of the six one inch wire seg-j~ ments symmetric about the core midplane, for both the HEU and 1 and LEU elements. Table III gives the ratio of these averages 3 ,~ for both iron wire reactions. The (n, p) threshold reaction
$ responds to fast neutrons and suggests very little change in j fission rate. The (n,1) ratio is consistent with the degra- ;
m dation in low energy neutron intensity expected for the higher ' U 235 loading in the LEU element. This reaction has a small l
) epicadmium contribution, not removed by ca'dmium covered acti- ;
l,_ vations. Since it is anticipated that the LEU spectrum is somewhat harder than that of HEU, it may be observed that, if y anything, the HEU/ LEU thermal flux ratio is larger than 1.19. l L . However, it is to be noted that the measured ratio is almost !
;f ,
exactly equal to the U 235 ratio. Rhodium SPND Flux Profiles j, Many SPND maps have been obtained during the EEU/ LEU cos - parison program. These have included partial or full profiles , 4] ' ^ in the core, in the south B20 reflector and in' accessible po-e sitions of the D20 north reflector. Activation results from Fe -
" and Rh wires have been used to verify SPND profiles at specific i points in both core and reflector. This verification has been Y
particularly significant in a) comparing profile intensities - from one core type to another, and b) comparing , the core ver-sus D20 peak intensities in the profile of any given core. );_ Although absolute intensities based on wire activation cal-
' ibration could have been presented, we have elected to present the extensive SRID results as they were measuregaccording to equation (6) with fixed sensitivity, of 3.0 10~ amps /nv through-4 out all regions measured. Improvements can be applied by inter-A 9
ested readers position. for the absolute value of S and its variation with
' The reasons for this choice are several: the SPND data is most easily reproducible, smooth without activation error fluctuations. Perhaps most important is the existence of a sig-2 nificant discrepancy in the ratio of D20 peak flux to core ' peak i flux, as measured on the one hand by the SPND and on the other by "
Fe wire activation. When this measured discrepancy is resolved kJ S will be calibrated by both Fe and Rh wire activities. For the present paper all wire activations will be normalized to the core
]
center SPND value for each core investigated. 4.. a N!< w
+ -
n,-.,-----.,,,,,,n.,,-,.,,-,-,
,.,,.,,,,,.-,..,,_,,._,,,,,,..,,,.,,.--,,,-.,.,.,,-n..-nn.._.-,,,,r.,_, ,,-,,,.,,,-_.,,,,,-w.,,.--
1 c._m___ _ .~_ _ . . . . _ - - . . .. . . I 1
.J *. .,
d> .
,-l u
.f
- )
a d
- i. '
~,'
TABLE III
' n,.
L ll SINGLE ELEMENT REPLACEMENT 3 g[.a , IN EQUILIBRIUM HEU CORE S
,4 p5 . .a i .
Reaction Ratio ' HEU Activity [.] .
?' LEU Activity u,a .
y<. Fe-58 (n, Y) Fe-59 1.19I0.036 . f.ij Fe-54 (n, p) Mn-54 1.02 0.031 . . . 3 - s.*. l
.1
- ,1
'4
.. a
'"l 1 .. J n
~; . - -
1 e g-H .. e 7 (
?:
O r k a e l I [. i ,I t;J J - 4 '* J. 5
.] ,
*h _
>s
'1 P h 1
~~
' ,2
*d I.l !
.i '
A; ,
' i >A e ,
q . e!! 1 f *- T. ..;-. ,.
. , . . . _ c. - , _z. . ..,,,_.; ,. ,. _ .
. . . _ , ,z . , ._7 ..m _ ., p . . . , .
4 A . I 1
~
As reviewed in Part I of this paper there are three sets of cores that have been mapped; a) the large equilibrium HEU cores I of dates 5/79, 9/79, 8/10,11/81, and 5/82. These cores should be closely equivalent in flux profile since loading patterns ], (Figure 2, Part I) 'are essentially identical; b) three small LEU y cores identified acccrding to loading dates 1-8-82, 1-21-82, and 4-16-82 (Figures 3,4, and 5, Part I). The April LEU core is some-4], what larger than the January LEU and shifted in East to West in 4,- loading geometry; c) the so-called "high leakage" HEU cores of
, 7,8,9,10/.82 (Figure 6, Part I) . These have the narrow five row h North-South loading of the LEU cores and are' intended to mock-up 4
~'
the LEU leakage conditions as nearly as possible. These will be designated simply as HL-EEU cores in this paper, i As described earlier, the geometry of the D20 tank precludes
$~ making measurements beyond 8" below the top of the fuel plates.
e Furthermore, the SPND can extend down to only 5" below the top of 4 , the fuel plates because of its design. For mapping in a given
, horizontal plane fluxes must be extrapolated 1" to yield a 1/4-height value, and 7" for the quoted core midplane value. While .
j there is a significant uncertainty in the value of the. D20 tank flux extrapolated to core midplane, the ratio of the core to D20 tank fluxes s,hould be reliable for the core 1/4-plane height. ' _ Since the back row of tank penetrations (i . e . , farthest north, M-Q) do not extend as far into the tank, this data should be viewed with some suspicion. The D20 tank penetrations are shown
. in Figure 6.
f 4 In the light water reflectors, special assemblies were de-signed to allow SPND measurements at four fixed distances radially away from the core. These adapters were designed to fit snugly against adjacent fuel assemblies. But because the fuel plates are u curved away from the SPND adapters, the radial position of the de-N tector channels must be defined carefully. The geometry of the f adapter.is shown as an ins,ert in Figure 7. The 1/8" aluminum face g] plate and 1/16" slot position separators have been treated as voids, and the H2O thickness dimensions are consistent with this assump- 1 i' tion. No adjustments to the data are made for the effect of the 3 - aluminum sideplates. I f'q- In the core, the detector probe is designed to fit into water channels 11/32" on either side of the center bail. The data in 3 plotted to reflect the actual radial position whenever it is known. j '! - otherwise, the detector is assumed to be positioned radially at the j- center of the fuel assembly. 1 S L [
/ A
r7x m m._=. ,u...m w w - .- . _.m : e: z..=macom._cm.sz u , .s... u __ _ _ _ i i
,i o
l, P
@Q @ M --
P 1 I N , ! v
) 3.2 5..
1 i .,
, t
~
@ @ R 6.25" .. ;
h Z Y @ IW U 3,
. i__ : ;
x 25" . M m
. i 1
k n ,w. s. o,o a nt eenetration o.o ery i- ii
- *r
- i. .
,d I I _? I f a *
; t- ie -t _
, _},
. . . =. w .c. =.:......=....:.;.-
=a = . . .. - - - . - w-; a _ w:. :. . . .. ~ n .......~.u-W t C i n s
W *
>f
<t ',
a --
~
I'4 - Z 2 1 3 x % O f.a I'2 ~ 3 A' - g HL-HEU'82 x l . o
. o 1.0 -
^2' -
1 3' t M LEU
-g. 8-
= 4, 2.625' { + l c m_.
.6
.5 ~
~ - - " ---
2-3.18-
.6875 i i
i g -- ( ,y .254 o l
.4 .
[+] t . f.75 7
, l s
.2 - - -
-v- i i
j s l I I' I I 1 i I 4 - 6 3.08" 2.45"212* t.95" i.49' l.33" .990" .3655" - Of ,
\A HO
~
Figure 7. H2O Reflector Flux LOCATION Profiles and Adapter 2 -
/j
- 4
'h
c.- a n --- - - ". t
] The SFND subcadmium flux maps from six core loadings are j given in Tables IV-IX. The tables are arranged so that the 1/4, 1/2, 3/4 height fluxes are arranged in vertical descending order j in each lattice pos4 tion. (Figure 8 gives the grid plate lattice
! designations). All flux values are in units of 1013 Table X lists the values of fth used for each lattice or reflector posi-tion, as interpolated from the measurements in Table II. The for-1 mat of Table X gives a single value of fth for each lattice po-1 sition; the upper number applies to HEU cores, the lower to LEU j cores. No axial variation is assumed. In the D20 tank, extra-
; polated data are presented with the 1/4-height flux above that
- for the core midplane. The positions of the typed data on the 9
l figure approximate the actual D20 positions measured.
- b The H2O reflector data is presented in the same format as for the core. The primed (e .g. , l' ,2 ' ,3 ' ,4 ' ) H2O re flector chan-j nels shown in Table VI refer to measurements made at radial po-1
' sitions 1,2,3,4).
further into the reflector than the normal positions (e.g., The relative distances are shown in the H2O profiles given in Figure 7. In this figure the two sets of points were _
' plotted from the two different LEU cores of Tables VI and VII but appear to fit smoothly together. The H2O peaks for equilibrium 3
HEU cores fall considerably below the LEU peak, but in Figure 7 ' , , , 8 it is evident that the peak is slightly higher for the HL - HEU d.' loading. 1 , -. Figures 9 and 10 are plots of the SPND data given in Tables IV and VII, comparing the normal equilibrium HEU core with the j nearly fresh LEU core of 4/82. The plots are North-South and j East-West profiles both passing through the core center element L-37. Differences in core size and burnup make easy assessment of the effect of LEU fuel replacement difficult. Reflector peak-
]3 ing in both H2O and D20 are clearly greater in the smaller North-South geometry of LEU. The westerly shift in the LEU loading is (a ;
also evident in Figure 10. The LEU flux in L-37 is 13.5% lower j than for the same position in HEU. The large peaking at the spe-cial element, L-57, is nearly the same for both cores. Because -
.? of differences in East-West loading symmetry the maximum core flux, h' seen in Figure 10, is 10.7% lower for the LEU core, rather than g the 13.5% for L-37. -<
q
'q Figure 11 provides a comparison between the normal HEU core and the high leakage cores typified by HL HEU (7/02) . The latter -
I shows the effect on the North-South leakage pattern associated with the smaller 5-row core.
, the same for both cores, Although the core center flux is nearly the considerably larger H2O and D20 re-flector peaking of the smaller core is quite similar to the effect
# 'e e .m Pua a
.. ., .. __ . .. s . , ... .... _ . _
.n_. , . , , ,...,-.y.. .
. ,7 , . , , ,..).,'i. %
3 . . . .. 1 - 1 . j D0 2 o
.635 n
, _.644-
- j.
.813 ~.824 l' S r
. 16 .05 '
.48 ..35 2 X W ,
F 1:" .
+]-
. .933 l
.I -
.821 1.01 1.22 1.20 1.12 .915 I i .948
~~
A C 1
, 1.45 k, '
\
I 1.17 l . l
! . 683 .950 2.70 1.49 1.48 1.45 1.29 .944 1 j j 1.18 i ,
l J . l B CR '
- 1. n I .
m .
- I
.928 l j" .837I .984 1.07 1.02 i .004l
- l' I
.8291 I I .
\
.717 .6781
, \
b i a special . j- HO 2 n Channel j .859 , ,;I ' Tablo IV. IIEU SPMD U Data from 9/27/79 j Channe 2 coro.
.843 1, .
qt ';L Channel 3 .727 H
"hannel
]p k .612 , f E .
. .- ,- -- . - . - . . . . .a .......L.
.. ,an.
, .. 4 Jl D,0 o n iN*
.- L lll t- s r .
.: X W ,
- 3 ,
1 1
.579 l
t { i .896 1.15 l .749
'l l
j 1 .972l l .647 . . 1.08 1 A 1,38 C l
'j M 1.15 l M ,
l I
.627
- 1.07 1.13 .939 .717 I
.789 1.38 1.43 1.26 .926 --
.657 l 1.15 1.17 1.10 .774 ,
'48 B 1.24 ,
C7 .
-i ,
M 1,04 l M
.554 .697 .708 .639 .480 .
~
j .684 .861 .880' .800 .604 - . j .575 .731 .755! .673 .504 j . ,
~
I . 1 =special . j HO 2 p . Table V. LEU SPND d 4 Data from 1/8 core. . i$
*) . .
'.! 3 a ~
~
! .1 ': . g 4-a il , , l
-"& es _ emer_ _,
-;.7.__ -._.-- m. _. .
, . .. mm ._,,____....._..x.2,.,.:,,.gg,_,_,.g... .; . .
- .v, . _.
~
3 '
.I 'l j_
D0 o ' n
- 3 2 ,
t
'4 . E '
d t i 5 __- r 1.21 1 13 -l - 1.55 . 1.45 4 o X W * ' fi 1.77 1.70 M^ 2.26 2.18
.770!
~
.574 l' .781 l' .935 .878 l .542 0' .713 1.00 1.20 1.16 1.00 .707 ,
h . .605 .873 1.02 .993 .870 .618 d- .908 .653
.643 A C
- - .820 1.15 ' .843
.705 i .945: M l M .720 I
* .659
.' j ;. .846
!gL. 1 i .717 .
.513! .718! B 1 CR .585
.645 .930' .758 -
.530 .780.
- M .665
^
n J
~
1 .. ! . 2 l i l 11 . a i ..
?
l 1
'j -
M Spe ciQl' , I j - HO 2
, 1.04 l s -.
Channel ] 1.19 Table M. m SM 1 10 Data from 1/21 core 1 :
.926 Channel 2 1.12
].
. 1.02
~ ~
+-I .792 l.. . Channe'l 3' .941
.850
~ .667
.. Channel 4' .807
.737i m
i d; r i I J y l j D0 2 o n l
}1 Q'ci [
B t s r i d 1.33 _1.19 l
- 1.70 1.51_ I d . -
i
;11 M -- X V/ - I q4 1.74 1.65 j d 2.23 2.12 i d
7; L l' l l _R?O ! I l
$] .
.788 !l 1.04 l 1.16 l 1.12 i .954! .679 ' -
[j l l l .946 l l -f 6
'j ,
- i f {
l A .932 c j 1 1.23 I M i 1.02 ! M l i
;} . >
.684
- i 1.02 1.08 i .842 .642 1 l
.861; 2.63 1.33 1.28 1.13 .824, . l
- l .717l 1.11 ! 1.05 i 1.01 .699' ~~
8 .ssol ER 1 '
- 1.10 .
I
^
M M ' l .923: l A l l . 1
.628l .753I .595- .660 .498i I I .783I $
? .
4 l mspecial 1, H'20 1.09 l Channel ] 1.32 Table VII . LEU SPND -> Data from 4/16/82 y 1.08 core. 1 Channel 2
?.32 .
. ~
2
- ~ l.00
.; Channel ] 1.16 'J t q -
" '.)
; .861 N Channel h 1.01
' ~
- g. --;3--MJ_1 ^,."S_ rah , - mu m -. - T- ~
hw em mm- d==---
..,.....-.............=.a.,.u.+.
4 . 3 i 1 D0 2 o n I s-
.i I 5 r i]
! X W j- 1.55 1,4g R 1.98 1.90 5
~
i i I
.907 l ' ] '
1.16 I s
.947 ] .i l
-t I. 8 A 1 16 C I 1
. 1.45 l l i! ,
l , M 1.20 ; M l l .
<.g
, .553 i .858 ; 2.21 I 1.16 1.24 1.12 .960 l . 719 I
- j .692 1.09 2.80* 1.47 1.56 1.45 1.24 . 913l l
l'- .526 j .844 2.25 1.19 , 1.27 1.19 .987 ! . 747 B i is CR i .
~
- _ l 1.48 '
l 3 1.21 1 M ] l .960I
- l. .-
1.24 i
, i 1.00 l I
e
- )-.
!, ; ) i y
- special q .
1- HO 2 f s- .. 1 raste vzzz.nzu svuo Data from 5/29/82
-' , Core. i 2 l
.:q . d e-3
'5 4, 1 _
--+--,,--,-,,---,e ,--w,,---g ,mm wn-.,w n ---, er - ,w----a n-, - ~ ,--. -
-,---e
. .a .: ~ - . . - . - .. ..
.3 . . . . i . q .. u [ .] 3't .- il ,:3 D0 2 o n H... 3 ? f- s r : g . . X W ._, p 1.65 : C 2.11 t >
,4' $
[
- I l l l I (u .
j l~1.23 1 g .. I I I I g ~ L A l C l l' l.53 I i ; W i
* ! l - !,
1 t
;
- l l l N 1.57
' l I 1
1 F l B 1.43 CR ' w i l M l N iN . j U; - - 1
- 1.01 .
1 >1, e . y . 01 . a ,1 . ..
^
N MSpeClGl , A -s H ,L0 d' 1.14 - 'l M Channel j 1.43 4 1.15 Table IX. HL - HEU SPND 73 1.15 Data from July 7, 1982 - IN Channel 2 1.4s cora . ,-1 g i1 1.21 4 1.03 W Channel D <-, .) 1.29 . . <! 1.04 a ~ M .893 Channel ( i. 1,14
.939 '
r. L j ' _ J. , . - n .- -.+ .
-.,,a_n.., _ , , ,,,... _ _ - __ _
ga + s g <.3 :,,..' j . r._m.a. .. .- .j w
, i . .
.. . _ . _ . . - . . . . . _ .. . . . _ . . ~ ._, ., ..,,,,. s.
1 . .
,? e 'i: -
l L'l -
;y a- a d .I A
b~ d (q -) Beavy water Tank ,
~
[.i !
-i l.
. g. -} 1 N -J ~ ',1, L-65 L-55 L-45 Ir-35 L-25 L-15 e .h I . A C 7, - Rod Rod L-76 L-66 L-56 L-36 L-16 L-6 ] - j .~ . . .
"l L-77 L-67 L-57 L-47 L-37 L-27 L-17 L-7
- Rod Control L-78 L-68 L-58 L-38 L-18 L-8 i .,
. L-69 L-59 L-49 L-39 L-29 L-19 i
( 4
\
- A
'.t } ;J L-60 L-50 L-40 L-30 m
c.'. . ,
, I Y
' l Regular !,Special Empty 4~ Element IElement Location
^?,
.s 4
A .. .
- Figure 8. Key to Lattice Positions ,
.!aI-I -
e k. 4: 't J .
.s .
. s
. . . u. . _ . . . . . z_ .
, .- +
t y . 00 o n (t 4 2 . 895 .895 1 lt t s r ! d .j .895. . 895 .895 ~ .j . 892 .892 l J1 X W q . 895 .895 .-
- -] . 892 .892 ,
.820 $ .820 I .820 I .800 l .820 ! .R1n l L-65 i L-55 L-45 l t-19 I L-2 5 ' L-15 l
-:{ .786 I .786 I .786 l .786l .786l .786 .!. l A 80o i C ' l
'L-76 : L-66 L-56 t-16 T-16
- 7.- 6 !
l .750 I: 790 '
- 750
- I 750 I 1 .
I .830 i .830 l
- I .800 I .791 R00 Ann l .810 1 j L-77 L-67 L-57 L-47 l - t-17 7.-? 7 L-17 l t-7 ~
-( .749 l * '
.749 l .749 I __
.749 .749 l I l l I
@ .800 {p l --
t-78 I r-6R i L-58 L-38 n_ig l t_g i , _, 0 .750 I 750 750 I
* .790 l :
J, i i .830 i .800 .800 .800 .830 l j,~ L-59 I L-49 < L-39i L-29 L-19 l
.795I .795 I .795l .795'
.795 i -
.910l . 91n i .
L-50 i L-4 0 ', L-30I
-7 4 HEU = .913
- SpeCiQl LEU = .860 .
HO 2 i ['
] Table X values of f g
@ values are j presented as: . 930 used in calculation -- HEU[ of 4 sc, j L-g . 914 l-LEU -
; 3 u
lC 4 - h 1 . I !1 's - L1 -
- )s i
* " *' ~ ~ < ' ~ ~ ' * *
~
~
+- s , ,2; . l",l, ; .fA~=~~ 3
~
rn ur -,.-p , ,,, ,,
- .t i
t sc 1
. . . . . .N N
_4 N 4 m co oN 4 m co o N l
. . . i , . . .
. . . _i
- g m
F c
- CORE BOUNDARY HEU --
2 - ; w -
?,
- x%)
=
CORE BOUNDARY LEU .' ns : = G, 8 r - -i mn i O cn O u s w ft G Q \ i .
,-d
- r s
\
xI ,-
+3 - I \ ' .4
- i,
;% u r1 gg mm m
- \, cc m - }.
. r- g 3e EI ss -
(h ~ mN cu $, yg _? n
% ~~ '
,I -:,- 4 H "" '.
l Y (I - X h g w- I - r,
-m f
/ .
[! v "I r
-[
N l2: I 4 E 8j b__ D 0 TANK 2 BOUNDARY } g - < L Z ,' N .;
\
f x, O 9 w - m -<
**$ H
/ ; [-
3' W
/ s',/ ~..
{ i - L (f) i: o - Z 1 I f f f f I I f f f t. h .A-..~ .,,.a...,r,.;..-..-,
..,,-.r, ,. 1.e ..., .
w- .~. -
.;.-- -t..,iv-- -w-.--m.
s , .
_- - - = mm _ m =. 2.--_ - --
- w -i l
f-i :s
^ il.:
l Li ! _ T h SC $ l N 1 1- . .. . . +@ ? p @ . .N N. .N ::w N A 03 CD O N A cn CD ON A CD . L.
.c t.
-- i i o - 1 I I i I l l l l l l J.
m. c g u 2 N S E
- N \
o
- N -
t- \ r l
.s.
m b N kN 5 1 -
's- - -
l 9 r l~ *
=
h N , Y : {Q p-E _ s _ 3o gi \ :: t. g 4 1 > . j -
. . ,Q N I J g -
1 I - ,. c r I . rI di o mm
? N- -
CC I n n sw T I u e r- s s s :. N co m .c
< N ,
.N W h.
- m j .
E 'r / - 1 e - / t- .; i
?N m / LEU CORE BOUNDARY j) i y / t 8-a c
rI / : t; !i T N HEU CORE BOUNDARY ? ' l.. d l to :, i W t< r
- 8) '
! ? - x c i O l-3 i.
- g. I.I.
\ s
- 8 n Ee
+
m .
,o '.
v. g r .' O I i I i l i I 1 I i 1 I p' e
. . .. Ii
, ( .*
_e.== ,. e
'~s-
, . . . - + . . - - -- es v
/****" ,"""**T*'**""" % ~m ' ~ - ~ d *
- r
.;:. .m_ _;
.,..y.. .
g._ _ , _. ......_. 4 .
; i .**
ic . . n_. . A b _ hSC r.i. N N
] N 'A '3 0 CD O N A OS Co O N 9- -
1 l l 1 1 I I I I I I 3 , - I4 tg C
\
A- E CD _ 9-79 CORE BOUNDARY - i p,
.g _ _
s . 8, n x _ m - 4y -e.
. .s 7-82 CORE BOUNDARY ') -
C ,f" 1 t n r-o a
.. .; oa t. .1 e s N
d E - - q c. r- 8
, *o g g o J i
m 5 ()D C \ 11 r -
\ TR' _ ; g r- ) c 5 --
M ti L oa i;i N "4 me C %. c
~ l -N -
Gi r- , l ::d W co 1 ~ Y (}J 'N J.4 5 03
/
- ]l ^ -
/ _
iL 87 /
- n -
4 5 'c.n oa .
..,....... D2O TAN K BOUNDARY
,_ p ~ ........,
M. "y X \ u i - Sc - [' k i-
- e. y
- s _ -
5:{
, . U l< i O M i -
.' .l ' ~ 2 ?-i l l l 1 I I I I I I I t
s.:,,, A& ;>.~ N d' l:-:a w a . - .2. - : . . - 3' "
.~ = u...-+-n ,
l1 n .
.i ,
- ) l d from the LEU comparison of Figure 9. The change in both cases is N ug dominantly due to the larger buckling of the two small cores.
- .JJ
%: Comparison of Rhodium SPND and Wire Activation Data }9 The HL - HEU cores were deliberately designed to have the
~
same small North-South core dimension and core D20 interface tl geometry as for the LEU (4/82). The purpose for these cores was ,j l N~ w two-fold; first, to compare fluxes for equal geometries, and a second, to compare experimental results when all measurement meth-5 ods are made at the same time and same detector position. Figure N 12 reveals several important answers to these questions. First, Q the similarity in reflector peaking for equal core dimensions is gj dramatically demonstrated. The dominating effect of large buck-
- .9 ling, inferred from Figure 11, is clearly evident. The LEU re-j flector peaks differ -from the same peaks in the HL - HEU only by 9 +6% in D20 and -7.7% in H20. Second, despite this similarity, Mi
" the central LEU flux is depressed by 18.5%. This difference, in fact, easily accounts for the already small difference in H2O re-4 flector peaking. Third, it is clearly evident that a gross -dis- - ,j agreement exists in the D20 tank between SPND and wire activatioh Q data, when both sets of data are normalized ati the reactor core ~
1 center. , r This disagreement remains a source of serious concern and q until resolved provides no useful benchmark for the proper cross j,.. sections to be modeled in the D20 tank. Small improvements of -
'i. order 15% can be anticipated from several sources such as mor'e d rigorous evaluation o'f the sensitivity factor in equation (6) but ~'
j the disagreement in Figure 12 is much greater than this. Peaking l
'} ratios of D20 flux to core center flux are given in Table XI for _, j
_ all detectors in each of several core types. The ratios from
' SPND data compara very poorly with those from both 'Fe and Rh ac-tivations. Unfortunately no wire data was obtained from the LEU ,
loadings. ! 4 Particular care was exercised in determining the actual _ l 4- axial depths at which measurements were made simultaneously in both core and D20 tank. It is _ worth noting that the D20 depths (the )
}; lowest possible) are not the same in wire and SPND cases. The flux a ratios given in the last column are extrapolated from the measured q positions to correspond to the 6" or 1/4 core depth. This extra-y polation is believed to be reliable since it is easily seen from I
9 the axial data points 'very near 6". The ratios correspond to the jj midcore flux values given in Tables IV, VII, VIII, and IX and in _' l
.j Figures 9,11 and 12, since the D20 midcore values are just those 4 at 6" multiplied by a constant determined by an in-core axial pro-file such as shown in Figure 1 or 2. -
h 1
-j. -
3 f$
. - _ _ _ _ _ _ _ .. - - . . . - ~l
ya.+s. ,x a. v.. . . . .. ... .- .= a .. .... .. . a n ...m.a = a s.. c n .x w w . a . . a i f i . i i i i i ,i i .-- i i ; 3 , , i.
'r
- i. v. .
i l 13 'p i 1 I I p l l 'I x
. x n\
g -X-LEU (4/82) SPND 8 i i z -o-HL-HEU (7/82) SPND z ' i 8 o HL-HEU (7/82)FE WIRE o
- 2.2 -
m o HL-HEU (7/82)RH WIRE m ! tu x /,, s\ :~ 2.0 - m Z I - s
~
o 4 1
\
' 0 F \s l.8 - - O r/ y 1.6 - 8j h - o s L i ' e"-14 j2 - p' -w ~~ oA' % / o
, g - -
Agy X %'~K 1.0 - d \ s' - V,'
\ ,
s
.8
.6 - -
.4 - -
NORTH = i
.2 -
. a IIIl l l 1 I l l l 2 - H 02 ---- L-39 L-38 L-37 L-36 L-35 X-W R-S l REFLECTOR Figure 12. hsc Core Midplane, LEU (4-19-82) vs IIL-IIEU (7 82 ) ;-
... .- .- .. . - . :.5 - .
,. Li L= :--..a
. aN -
, , . ..j,m.-
t
- 1 LJ B
x 1 5 . TABLE XI Vjj RATIO OF D 2O TO CORE CENTER FLUXES [k ~ (4 Actual Flux Ratio ^ Core Date Detector Depth measured * (D2 0-X/L-3 7
$I
?; HL - HEU 10/1/82 S PND 5.0" l.23
.d HL - HEU 10/1/82 Fe 7.5" .82 HL - HEU 10/1/82 Rh 7.5" .88
~} .
- ,y HL - HEU 8/25/82 Fe 7.5" .83 f4 HEU 5/ /79 Fe 7.5" .73 --
HEU 9/ /79 SPND 50" 1.25
,h HEU 5/ /82 SPND 5.0" 1.25 j) LEU 4/ /82 SPND 5. 0 " 1.61 t
i nj J: j . v,
,s J'
- The flu:e ratios shown in the last column' result from axially .
;l extrapolating the measurements at these depths, to the 6" or p .i 1/4 core depth. *
*I 3
').-
- i,a _
9 s F J
~
a O e. uf
.4 -
fb $
?.
~
L,!
'd 4.,
H: . eMuq 'l l*
., ,. .L '
......,a.,... '
-r.. n,. -.-- . - - ---
Q2.5d32ir. ..':MML q_ ., w y p % ,g y g , ,s. .f y, .. .-
=. . .. ~.. .
A somewhat different summary of core and reflector peaking .M] f d at midplane is presented in Table XII for all representative cores. Relative SPND numbers from core to core should be reli- @- able and reproducible to perhaps 2-3%. Since the method of cal- . orimetric power calibration. was the same in all details for all l yl cores, except HL - HEU (7/82), it is believed that no significant l 1.g error exists from this source [ calibration thermocouples were re-
- g. placed just prior to HL- HEU (7/82) and their calibration and po- ,
q sitioning may have introduced some minor but unknown systematic d~ , error for HL - HEU (7/82)]. It is to be noted that the first two d cores should closely agree since both are. equilibrium HEU load-M ings. The 5/82 core appears to show 2.0% to 5% higher levels h^6 than the equivalent 7/79 numbers. We have r'o ready explanation for this difference other than possible burn-up differences. T However, the five element " cross" at core center presents the N~ most reliable comparison and for this the disagreement is 2.0% - f.] m within experimental uncertainty. fq.$
'~ ~
Summary conclusions Within the experimental limitations discussed in the sec - tions above, the program to measure subcadmium flux profiles
- 0) leads to the following conclusions: I U (1) Replacement of a single fresh HEU element by a fresh LEU element at the center of an equilibrium HEU core produces a
.D local flux depression. The ratio of HEU to LEU local flux is i '- 1.19 .036, which is, well wi' thin experimental uncertainty, equal i
Cj to the inverse of the U-235 masses for the two elements. M (2) Whole core replacement of a large 38 element equilibrium '
~
EEU core by a fresh or nearly unburned LEU core reduces the core fj rp flux and raises the flux in both D20 and H2O reflectors. The re-
^ duction in the central core region is 4.0% to 10.0% for the small fresh 29 element LEU core, and 16% to 18% for a 31 element LEU q core (4/82) with low average burnup ( < 3%) . These changes are
'l consistent with the total core U-235 inventory. The increases (i in reflector peaking are dominated by the reduced core dimension d ,^ d in cores. the North-South direction assiociated with the smaller LEU ( d (3) Special "high leakage" HEU cores, which reproduce the smaller LEU North-South geometry exhibit reflector flux inc nases Q 9, similar to the LEU cores, and at the same time show a core center (L-37) flux 18.5% to 20.5% greater than the LEU (4/82) core. 9 (4) There is no observable change in axial profiles between
; HEU and LEU cores.
(5) Magnitudes of the D20 reflector fluxes relative to core y ' fluxes as measured by SPND with a fixed value of sensitivity, S, ] are in gross disagreement with the same flux ratios measured by 3
, - - , - , - , - , . - - . - . . , - -, -a - +. . . . , , , , . . , . - - -,.n ,es. ~,
. ., , . 4 .
- t. . , i : 6 i r~' [~~ f'~ I I
.) 1 1 .
- !] i' TABLE XII h*
Ti HEU-LEU MIDPLANE FLUX
SUMMARY
.-], ,
3-! 'IOTAL # 112 0' L-37 D20-X CENTRAL SPND FLUX . ij ' FUEL DATE ELEM'T SPEC'L SPHD GPND SPND WIRE 5 AVE. 3. AVE. MAX. COMMENTS U . i ' IIEU 9/79 38 5 .859 1.h8 1.87 1.08 1.45 1.h3 1.h9 Slight flux tilt vest. 11EU 5/82 38 6 ---- 1,56 1.98 ---- 1.48 1.50 1.56 Flux center Ir-37. I-IIL-IIEU T/82 35 5 1 . 18 3 1.57 2.11 1.28 --- 1.51 ---- Five tiered core.
, Not standard equilib. HEU.
LEU If8/82 29 6 ---- 1.h3 -- --- 1 34 1 35 1.1 ' Cores differ: Special 43 'h in L 10 moved to regular LEU 1/21/82 29 5 2.26 -- --- -
--) 4 in L-65. Flux moves vest. Batch core.
( [. LEU h/82 .31 6 1 32 1.28 2.23 1.21 1.20 1 33 Fuel added to vest face. f Flux shifts vest.
.,. . Y
!- NOTES: I (1). Wire data in D20 normalized to SPND L-37 flux. Data taken on a similar core. _
F t (2). "5 AVE" is average of L-36, L-37, L-38, L-27, and I,-hT. [. r. (3). "3 AVE" is average of L-36, L-37, and L-38. 1 (Is ) . " MAX" is the maximum value of the core subcadmium SPND flux (excluding flux traps in special fuel elements ). [ r i 1 e
.... ..-,, ., s ._.
,.,...,,..:,~-.- m , ; , c y . - w...
--m- --
.r,-:- . - . - -
wv - -- E
. , - . _ . . _. _ c . .
.___....,_ ._ . ,y ,...._. ,_. ,,. ,, .g. .......- ....;- -a.____.-
)
- ]
~
y ,
~
I I A [j Fe and Rh wire activations. Space dependent refinements of S are
;j calculated to give some improvement in the discrepancy but the j_ major part of the correction remains to be resolved.
1 Acknowledgements The authors would like to acknowledge the work of Gerald
., Munyan, Sheila Melton, Keith Flint, and Michael Bacovcin who 9 have diligently acquired and reduced much of the data presented !-8 in these two papers. Their dedicated efforts have contributed significantly to the experimental program. Special thanks ; should also be given to Mr. Gary Cook, Assistant Reactor Manager, m -
and Mr. Frank Bernal, Senior Reactor Operator, for their advice, 5 A cooperation, and skilled handling of the reactor and experiments. t,s ' - 3
>) ..
I I 7; _. 1 -
'i
._e h
4 3 l.
- y :
( l 1-e t9
;- d k .$ -'
il 9, c_ n'*,
. .)
- .,j .
h'~ A
~ -
R. a
.tL
- c;- ..
n - . . . . _- a
.. , . . . . . . . - . . . w. _ w- ...u ~ ~ a.r.w. .
- ) .._ .. .
,1 , .. . . , '1 1
L! y
!.1
> ;j id . t . I
'J i:' -,. ,tl M ~'
l y References (1) . K. H. Beckurts and K. Wirtz, Neutron Physics, . 1 Springer-Verlag (1964). _ U (2). ~ "The RERTR Demonstration Experiments Program at the Ford
. Nuclear Reactor," Proceedings of the International Reduced , .-l Enrichment for Research and Test Reactor Program Symposium, November 1980. (unpublished) '
d
.i (3). W. Jaschik and W. Seifritz, Nuc. Sci. Eng., 53, p. 61-78,.
1, (1974). - -
~
I (4) .
^!
' Low Enrichment Fuel Evaluation and Analysis Program, University of Michigan Department of Nuclear Engineering '
Technical Report, January 1980. - (5). H. D . Wa rren, Nuc . Sci . Eng . , 48, p. 331 - (1972) .
. :i A (6). " Inverse. Kinetic Analysis of Rhodium SPND for Thermal 4
Flux Mapping," Trans. Am. Nucl. Soc., 35, p. 571 (1980).
; (7). Neutron Detector Data Sheet, Reuter-Stokes Canada LTD. ,
T{ RSN-202, Serial number LC-0511. Also, Reuter-Stokes publica-j tion, "Self-Powered, In-Core Neutron Detector, " RSN-202-M2. '
.J y .
ci 1 m
~j d
[3 - 6-di - 65 . y% 1 .
- > , .= - -
.: . L - .
-- , - -; . y ., ., g g. .., 3 y . ~_. m , , .. ,.9
~
i .~ .
- d. ,
FE r H t APPENDIX C a.
- b v Analysis of the Ford Nuclear Reactor 3 '- LEU Core h.t r
e
,I a) b5 ~
_~
@G ,'f **
lk . 1 4 u _ __ _ (# *
- s
!. t 1 ~4 -
- 1
't 1
a
's .
- s .
9'
- e.
O
,a .
I:4
- t
.3 t.
." l
*l 4
i
~..
.6
..t ..
I'+a j.* J;. 6 . e
?*i
- s
".}
IS
.W. '!r
.*j s, r r.I go i, h k ts.. u "9 h,: i eg
,i e
a.- : ,. p . s _; w '.:. e . w -
. .:- n -= ----=m-p ., s . 5
~-
,y 1 .
4 . 3 Qj ANALYSIS OF THE FORD NUCLEAR REACTOR LEU CORE y ej J . A. Rathkopf, C. R. Drumm, W. R. Martin, and J. C. Lee 7 Department of Nuclear Engineering s The University of Michigan Ann Arbor, Michigan 48109 W 9 . Introduction - d 4.i The University of Michigan Department cf Nuclear Engineering
% ; and the Michigan-Memorial Phoenix Project have been engaged in a i cooperative effort with'Argonne National Laboratory to test and
- 2.i analyze low enrichment fuel in the Ford Nuclear Reactor (FNR).
yj The effort was begun in 1979, as part of the Reduced Enrichment , by., Research and Test Reactor (RERTR) Program, to demonstrate on a whole-core basis, the feasibility of enrichment reduction from ij. 93% to below 20% in MTR-type fuel designs.
,3 .
M The low enrichment uranium (LEU) core wa,s loaded into the . i , FNR and criticality was achieved on December 8, 1981. .The criti-a cal loading followed one-for-one replacements of high enrichment f uranium (EEU) fuel elements with LEU fuel elements in the center-and periphery of the FNR core. Following the critical loading, J M; approximately six weeks of low power testing of the LEU core was performed including measurement of control rod worths, full core -
% flux maps, and spectral measurements in-core and ex-core. This g was then followed by 2 months of high power testing (2MW), during f which similar measurements were taken. These measurements were -
8 . , performed as part of the demonstration experiments portion of the - Q overall FNR LEU testing and are described in two companion $4 papers *-* to be presented at this meeting. The focus of this _ d paper is the analysis of the LEU core and prediction or simula-D tien of the various measurements, such as critical mass, control -
$ rod worths, relative worths of LEU.vs. HEU elements, and relative
( flux profiles in-core and ex-core. and calculated values are included wherever possible. Comparisons between measured - b1 _ f* Previous reports *-8 have described the demonstration experi- - ?: ments program and the analytical effort to develop and verify the Q calculational methods used for analyzing.the FNR HEU and LEU con-M figurations. In particular, Section VI of Reference 5 compares - the FNR LEU and HEU cores with respect to relative flux / power 4 distributions, control rod worths, various reactivity coeffi-a cients, and fuel cycle parameters. Noting that these comparisons 3 are strictly valid only for the FNR HEU/ LEU comparison, we sum- ' s marize the important effects below. (These comparisons are for ji fresh LEU versus fresh-HEU cores.) l (1) The in-core thermal flux level in the fuel is expected to decrease 15-20% due to the increase in fissile loading.
, , (2) The D,0 tank thermal flux is expected to decrease ap- '_ '
50 proxihtely 4-8%. This decrease is less than that in the m
~
y _ p .- u - . , 3.+;, m :7.33:,ygy3..,9y y mpm:q g.y g-i3 p.
- j. '. .
l . x w 2 in-core flux *due to'the fact that the slowing-down source is g nearly constant j (3) The control rod worths are predicted to decrease ap-proximately 4-5% (relative). i (4) Th'e cycle length increases approximately 50%.
)a j (5) The shutdown margin decreases approximately 4-5% (relative).
4
'i The validity of these predictions can only be inferred from 3~
comparisons of predicted versus measured values on the equi-librium (depleted) HEU core and the fresh LEU core since it will d take a considerable time to reach an equilibrium LEU configura-11 tion (at least 2 years) and a fresh HEU core was not available. 1 The purpose of this report is to give the status of such com-O parisons and indicate the areas of uncertainty which we are in - j' _. vestigating at the present time. 3 Calculational Methods
} -
j; - In this section we will describe very briefly the cal-culational methods used to perform the FNR HEU/ LEU analyses.
] References 4 and 5 provide additional detail if needed.
Cross Section Generation N. The cross sections that are used in the global diffusion d_ theory analyses are generated via an extensively modified version 1 of the LEOPARD code
- suitable for slab lattices and enrichments i characteristic of HEU and LEU MTR-type fuel. The LEOPARD code is
'[7 a zero-dimensional spectrum code employing the MUFT code' in 54 fast groups and the SOFOCATE code
- in 172 thermal groups. A critical buckling search is included to maintain the lattice cell critical throughout the core lifetime and depletion is accounted
; for. Within the past year we have incorporated an ENDF/B-IV data j base into the LEOPARD code, as discussed below. 1 i
The EPRI-HAMMER code' is also used to generate cross sec- l tions, but primarily for control rod calculations. The EPRI-
~'
HAMMER code is a one-dimensional integral transport code which includes a 30 group thermal calculation and a 54 group fast cal-culation with an ENDF/B-IV data base. Although EPRI-HAMMER does not include depletion, a link between the LEOPARD and EPRI-HAMMER
,- codes does allow for HAMMER generated control rod cross sections as a function of depletion (of the fuel in the rodded element).
1 Global Calculations
$r An extensively modified version of the 2DB code, 2DB-UM, b! [- is utilized for all global calculations for flux and power dis-tributions in the FNR. The 2DB code is a standard finite dif-3 ference code for solving the neutron diffusion equation. The di 2DB-UM code accounts for burnup via a macroscopic burnup method n
du
- R:
! ',..)
----.,,n . , - - . , , . . -
- ~ , - - . , -. . - - - - . ~ , - - - - - -
- u. . .a ,{, ,
. - . a s.--_.a-- -.e w ---. . ~. --
.a Si ,
.1 3 -
~
- that is based on the interpolation of macroscopic cross sections
.;) for a particular mesh as a function of the local depletion. The . h burnup library is constructed from a LEOPARD depletion run for Jj
~
the particular fuel type. The key to this method is efficiency and ease of use--it is quite easy to simulate several years of 1.1 FNR operation, accounting for the bi-weekly startups and shut-4 downs and accompanying fuel shuffles. We have also used the VEN-TURE code for 3-D calculations to obtain space and group 1 j , dependent bucklings for the 2DB-UM code. The PERTV code'8 is -. di also employed for perturbation calculations for various reac-jj
.~
tivity calculation's. It interfaces with the 2DB-UM code.
~
Di Control Rod Analysis l-UC .1 tj Reference 5 should be referred to for a detailed description M of our overall method for performing control rod worth calcula-d tion. Basically, the EPRI-HAMMER code is used to generate cross :l,
?.',
~
sections for the TWOTRAN code'* which is a two-dimensional J discrete-ordinates transport theory code. The TWOTRAN code - generates reaction rate ratios which are then matched with the . 1; -- 2DB-UM code by adjusting the fast and thermal absorption / removal - j cross sections for the control rod region. We have found that ^ q - the adjusted control rod cross sections are independent of the . fuel environment; therefore, only one set is.needed f.or the.HEU .s ': fuel and one set for the LEU fuel. H' Ex-Core Calculations . a 4 The ANISN'* and ANDT'* codes have been used to calculate - fluxes in the D2 0 and H O reflectors.and the beam ports. The [, ANISN code is a one-dim 3nsional discrete ordinates code and ANDr - ng is a general purpose multi group Monte Carlo code. Cross sec- ,a tions for these codes are generated via the AMPX system. ' 4 ENDF/B-IV LEOPARD Library 1 -
] Disagreement between macroscopic constants generated by the HAMMER and LEOPARD codes have been attributed to differences in .
O their respective cross section data base'. The HAMMER code uses _ M END'F/B-IV data while the LEOPARD code used an early industrial 1 cross section library. Inspection of microscopic cross sections-
; generated by the two codes shows serious disagreement for several A important isotopes including oxygen, aluminum, and ***U. In or- -
Q der to remedy this discrepancy a library for LEOPARD was as-3 sembled from MUFT and SOFOCATE compatible libraries obtained from' [I the Westinghouse Electric Corporation. Implementation of the - p new library required modification of the LEOPARD code to accom-p modate the additional data contained in the ENDF/B-IV library.
'I Verification of the ENDF/B-IV library included simulations
; of critical experiments, comparisons with established benchmark JJ codes such as the HAMMER code, and modeling of the depletion of g fissile fuel in pressurized water reactor fuel. Table 1 shows -
pj i some important microscopic cross sections obt'ained by the LEOPARD 1
~
h' l' . o __ u. m m. - m% _ _m
~ .
L ,
- m. ; - c +;--,w ; m c :r- y _ :~m &m,7-9. ny.v: =4 a ~ ~ :"%:-+;"" l 4d 4
1 ' code with its two libraries and the HAMMER code. The disagree- ) i ment between HAMMER and LEOPARD resulth with the old library is i not present for the ENDF/B-IV library. This is expected because, j as mentioned earlier, the HAMMER code also uses ENDF/B-IV as its
.: data base. The disagreement on the macroscopic level between d< MAMMER and LEOPARD results has not been completely eliminated, as seen in Table 2. In fact, for some parameters the old library's w '
d ,~ values are closer to those of HAMMER than are the new library's, 1 3 The serious discrepancies, such as the fast fission cross sec- , k{, tion, have been reduced by the use of the LEOPARD ENDF/B-IV li- ' brary.
)! ' . .
Ij j Table 1. Comparison of Some HAMMER and LEOPARD j - Microscopic Cross Sections for MTR-Type Fuel j LEOPARD h.~
! Ele- Cross HAMMER old library ENDF/B-IV library J ment Section i value % diff. value % diff.
16 g C a1 1.16-2* 34.4-2 +3124. ' 1.14-2 -1.72 27 ' Al O a3 1.08-2 2.80-2 +159. 0.997-2 -7.69 27 Al C a1 6.80-3 9.74-3 +47.24 6.36-3 -6.18 .f. l j.. 27 Al c a2 3.06-3 3.50-3 +14.38 3.02-3 -1.31 235 jg U vofj 3.07 3.12 +1.63 3.39 +10.42 235 g a3 39.3 37.0 -5.85 39.7 ~ +1.02 35 g
; "O f3 63.9 59.3 -7.20 63.6 -0.47 l' .
j- 1.16-2 represents 1.16x10 -2 ,4, - - 4 Although the.ENDF/B-IV LEOPARD library does not provide per- l J.- feet agreement with benchmark codes, it can be used with more { 9 confidence than the old data set. Differences between LEOPARD $} calculated results and those of either benchmark codes or exoeri-JL ment can now be attributed primarily to LEOPARD's methocolocy rather
- than its data base.
k y, Analysis and Comparison with Exceriment j HEU/ LEU Single Element Exchange 3
I
. -.o ~,*--,...e' . ....- . , MJ_Q ._ . ;e . mM . .. 2 . - - . . - am.e M ,
l th; q ,
"i
'? 5
- i-!
O Table 2. Comparison of HAMMER and LEOPARD P Two-Group Macroscopic Constants for 'f, MTR-Type Fuel 7, ti
; LEOPARD 0
Q m Parameter HAMMER old library ENDF/B-IV library , ij
~
value % diff. value % diff. d ti HEU
- A N
k, 1.76447 1.76302 -0.03 1.76610 +0.00 ij $ /6 2 3 3.6736 3.8063 +3.61 3.8311 +4.20 L I,j 3.8131-3* 3.8860-3 +1.04 3.7620-3 -1.34 it (* I gi 2.1050-3 2.0138-3 -8.20 2.1608-3 -1.10 I a2 0.10832 0.11312 +4.43 0.11370 +5.05 I f2 8.1130-2 8.4828-2 +4.56 8.5249-2 +5.08 d ' " LEU ?;
,'l k, 1.65980 1,66196 +0.13 1.66368 +0.23 J $j/62 4.6196 4.7911 +3.71 4.8317 +4.59 I,j 6.9685-3 6.9712-3 +0.04 6.8765-3 -1.32
>? m I.,1 2.7718-3 2.5600-3 -7.64 2.7390-3 -1.18 - 95 0.12643 0.~3252 I a2 +4.82 0.13327 +5.41 b fj I f2 0.0958 0.10063 +5.04 0.10111 +5.54 - A
- 3.8131-3 represents 3.8131x10 -3 3
s .A $ In order to examine some of the performance differences of - 7; the LEU and HEU elements as well as to provide additional oppor-et. tunity for analytical methods verification, an experiment was t performed on September 15, 1981. The experiment consisted of the !2 substitution of a fresh LEU fuel element for a fresh HEU element fj in an equilibrium HEU FNR core. The substitution was first made it at the center of the core. Then, after returning the core to its 9 original configuration, the exchange was repeated at the' edge of r 1
- w. .
. . - - . -.. . . . - - . .. . , , , e y p.g, gp~ m_ ,,. .~~? n m. .
*'~~~
.i. ' 6 i_ the core. After each substitution the relative reactivity of the , j LEU element was. measured. ' a f The reactivity of the exchange at the center of the core was d- determined from the resulting positive period. At the edge, the j effect of the change was so small, however, that the period was , G too long to accurately measure. Thus, in'this case, the reac-Q '-- tivity effect was deduced from the change in the critical posi-tion of the regulating rod. The configuration of the HEU core is ! shown in Figure 1. The center position is marked 37; the edge j_ position is 40. The results of the experiment are summarized in ;
; Table 3. ,
- t " l j- Table 3. Reactivity Effect of 1, HEU/ LEU Exchange : ,i 'l - 7 ok/k (%) ; Location of Exchange . Analytic d_ Experiment 2DB-UM PERTV l Center (37) -0.1176 -0.1301 -0.1105 . / 1. . Edge (40) +0.011 +0.0036 +0.010 h + f ! The HEU core and its three variations (LEU in the center, ; HEU on the edge, and LEU on the edge) were simulated by the 2DB- ;
, UM code. In the calculations each fuel element was approximated t 1 by an 8x8 mesh. From the calculated eigenvalues, reactivity ef-l fects of the two exchanges were determined. These values are ,
presented in Table 3 as are those calculated by the PERTV code '
., which uses forward and adjoint fluxes from the 2DB-UM code to 9
@qj calculate. changes in eigenvalue. For the center exchange, where ,
'- the reactivity effect is large, both analytical methods provide U satisfactory results. On the other hand, the perturbation tech- i 9 nique simulated the edge exchange much better than did the 2DB-UM :
- code. This is because the small reactivity effect strains the ,
j- eigenvalue convergence criteria in the 2DB-UM code. i The differences between the two elements are less apparent ~ o
}_ on the edge of the core than at' the center simply because of the lower flux in that region. The reason why the LEU element is * !< 1ess reactive than the HEU at the center of the core but more 3} reactive at the edge is more subtle. Infinite medium calcula- i l' tions shed some light on this phenomenon. In particular, the LEU l H, element is less reactive in an infinite medium (k,(LEU) <
j k,(HEU)) but for a finite core, the LEU fuel is more reactive j 0' 4 7
- , ; i I
4
.....~....n.--~ '
~ ~ G d
, ...v.
--~=. - .-. o :.. -
~ - -- -
~~*".'*
., G ?.
y d 7
'-i .
s q 4
.' t
, ~.
N l
] Heavy Watee Tank .? * , .a -t '1 t
. .;4
\\ >
\\
j \\ 2.:t \\
\\ \\ -;i ~
7s \\ es ss 45 35 2s
.4 15 s \\ l 1
a l el -1
.) 7s as ss 44 as as is s
,} centec 77 57 57 47 37 27 17 7
^
".}
c.y,] - e; - -- - . -- 7s - se as 4s as 2s is a
]
's \\. \\ \\ ._
\\ \\
') \\ \\
\\
.. j 7s \\ se \\ ss
\\ -
y 4e as as ts s \\ 9
\\ \\ \\ \\ \\
..] \\ \\ --
\\ \\ \\ \\ \\
"Ja \\ \\
\\ \\ \\ \\ ecce \\ \\
} so \\ 70 \\ ;
\\ s0 \\ 50 \\ 40 30 \\ 20
,' \\ 10 \\
4' - ni i
- \\
neouise \\ anoty ------ specisi d Element \\ Location Element -i -{
.s' ei
+1 4 i i - A\ se (,* Figure 1. Core Configuration for Single-Element ^ Exchange Experiments , -($ I n i i
~
Y . 2, .ji d$ g f 4 ' # N ._3 f e g
./.4.. _h_
. _ , , , *g ~ . .
< . , . , . .v.,
n . . .. ; n. h . .m , , w m .g ;.u n e., _, w . g,; u n. ~ ~, -. n ~ ~ >-~ .s . o ., a '
- ] . ' -
., v. , a 4_' 1 u. (k This } is'$$s(LEU) oseenbfdo(HEU)),aspredictedbythe2DB-UMcode.
> kf mparing the migration area, M8, for the two 4
fuels from the LEOPARD code, in that M8 (LEU) < M8 (HEU), in-dicating the LEU fuel is less " leaky". Therefore, inserting the d~ LEU fuel in the center of the core, where the leakage is low, is , similar to an infinite medium and hence results in a lower reac- l i tivity. On the other hand, inserting it on the edge of the core, i i" where leakage dominates, the smaller M8 f or the LEU fuel causes a l ] slight reactivity increase. '
. LEU Critical Loading The December 1981 loading of the first LEU FNR core provided another opportunity to verify the ability of the computational 4- methods to simulate an LEU experiment. The initial critical ii loading took two days and was completed at about noon on December 8 after the placement of the 23rd LEU element. Figure 2 shows 1- the critical configuration, the positions of the three fission J chambers, and, in the upper right corner of each element loca-3' tion, the order of loading. Upon withdrawal of all control rods t i execpt the regulating rod, the reactor experienced a period of 113 seconds, which corresponds to an excess reactivity of 0.067%. ;
If the wetth of the regulating rod (measured to be 0.3831) is a considered, the cold, clean LEU core had an excess reactivity of' ! 1~ 0.451. 4
The critical loading sequence was simulated twice by the '
2DB-24 code using a structure of 36 (6x6) mesh per f uel element. l For each of the two simulations different LEOPARD generated cross
; sections were used: one using the old library, the other ENDF/B-i IV. Table 4 presents the eigenvalues obtained from the two cal- ! culations. The nominal masses of 888U were those considered in 3; the calculations. The masses labeled " actual" were :alculated by
{}- summing up the mass of each loaded element as reported by the manufseturer. l! The 23 element core was found to be slightly super-critical 4 1 in the 2DB-UM calculation using the ENDF/B-IV cross sections but
$ slightly sub critical with the old library. The core simulated d'
by the 2DB-UM code is one with all rods withdrawn. The measured [f- critical mass was for the 23-element LEU core with the regulating
]i rod fully inserted. The analytical and experimental results can be compared by examining (1) the multiplication constant and (2) the estimated critical mass. The LEU core with all rods j" withdrawn was estimated from experimental results to have a
, multiplication constant of 1.0045. This compares with the 2DB-UM Cj values of 1.0025 (ENDF/B-IV) and 0.9985 (old library). From the
' mass of-3512.82 g for the super-critical core and the excess reactivity a critical mass of 3436 g 888U is estimated. The i-values estimated from the 2DB-UM code are 3471 g and 3545 g for
!! the ENDF/B-IV and old libraries, respectively. The better agree- i ment provided by the ENDF/B-IV library adds further support for
. the new LEOPARD library.
k-
- i 4
!};.
. . . . . . - . e ._. . w . . , . . . . , . . . ..a. s_ : ' .. p s a k na 3 .m. . . . 1.
4 9
,.i *)
n .
.,3 .
1
...]
'?
g __ Heavy Water Tar * . a
^
. \\ \\ 23 12 to ei 22 \\
! \\ \\ \\
\\ \\ \\
, 7s \\ as \\ ss 4s as as is s \\ i i .
\\ \\ 21 3 7 1 ts \\
'I \\ \\ .
\\
\\ \\ \\ -'
76 \\ 64 \\ ss 44 34 26 16 6 \\
\\ \\ to e a s 13 \\ //
! \\ \\ \\//
\\ \\ //\\ '
77 \\ s7 \\ 57 47 37 g
, 27 17 7 \\
j 1 \\ \\ no 4 a 2 14 \\
; \\ \\ \\ .
1 \\ \\ \\ i 7s \\ sa \\ se 4s as as is a \\
\\ // \\ - \\ ~
is to 17 \\ \\ // ..'
\\// \\ \\ \\ \\//
//\\ \\ \\ \\ //\\
7s \\ se \\ se \\ 4e as
] as is \\ s \\ - .1 \\ \\ \\ \\ \\ \\ \\ \\ . -) \\ \\ \\ \\ \\ \\ \\ \\ .
i \\ \\ \\ \\ \\ \\ \\ \\ so \\ To \\ so \\ so \\ 4o \\ so \\ 2o \\ to \\ 1 d 1 \\ \\ // neoular \\ Emoty - - - - - spec at \/ P1eston , Element \\ Locatton Element // \\ chemoor t l 3: i L; L( 1.-
. Figure 2. LEU Critical Loading Configuration'at y the FNR, December 8, 1981 r
b) n U i3 - t l'- __ .=n_.= .. _ ..
.- . - . . _ _ - - - - - . a
.i - -
. w . .,.,. ,.c j.e , c. . ,_.7 _ , , 7_ . ._. , . . . 3,,. p .,. .
7, ,... .
,.,..,,y Z 1 i '
]
- 10 j
; Table 4. LEU Critical Loading
-l '- 2MU' Mass (g) 2DB-UM calculated k h: Number of LEU Elements ENDF/B-IV j{ ._ actual nominal, old library library i I ti fi g.. _ 21 3117.20 3178.7 0.9792 0.9835 22 334 85 3346.0 ,,0.9888 0.9926 )- 23 3512.82 3513.3 0.9985 1.0025 i l :. nominal elemental 235 g ,,,,,, 3 L. ( regular: 167.3 g special: 83.65 g [ Control Rod Worth Calculations .
- Among the many reactivity-related paramet'ers measured f'or the LEU configurations at the FNR during the past year, our -
~
simulation effort to date has been lir.ited to the shim rod worth data. The rod worth calculations used a combination of transport and diffusion theory codes as outlined above. In Table 5, the j calculated reactivity worths of shim rods are compared with the j" full-rod worth measurements taken for the December 1981 LEU core. g. In a similar comparison for the' February 1982 LEU core presented in Table 6, full-rod worths are estimated from the measured half-1_ rod worths. Based on the full- and half-rod worths data obtained i for the December 1981 LEU configurations, a multiplication factor
.h g,
of 1.93 worths. is chosen to convert the half-rod worths to full-rod y Table 5. Shim Rod Worth' Comparison 7 December, 1981 LEU Core with 27 Fuel Elements l;j'. l Reactivity worth (%Ak/k) i Rod Relative error (%) ,j . Measured Calculated h}- 4
- A 2.22 2.28 2.7 B 2.32 2.65 14.1 l- C 2.28 - 2.25 -1.6 L:
r i f.-
-_ .,. - .. c . ., - -- - - - - - - - - - - - - - - -
~
A . q 11 -
.a v.i -
1 M Table 6. Shim Rod Worth Comparison M February, 1982 LEU Core with 30 Fuel Elements S Qd Reactivity worth (%ak/k) dt Rod Measured Relative error (%)
~
Calculated j Half-rod Full-rod
- 3 .
3 A 1.37 2.64 2.76 4.5 c' B 1.16 2.24 2.47 . 10.3 - do C 1.08 2,08 2.11 1.4 11
- Estimated as 1.93 times measured half-rod worth l-i t'
i N ,s M The comparisons given in Tables 5 and 6 indicate that the l reactivity worths for shim rod B are significantly overpredicted " j) , by our calculations, while those for rods A and C are in reasonable agreement with the measured worths. This discrepancy-j for rod a may be associated with inaccurate flux distr.ibutions ' t calculated for the LEU configurations, and is under further study
,' in conjunction with the flux distribution calculations discussed -7:
__below. . d.
]
Thermal' Flux Maos
.d Since the thermal neutron fluz both in the core and reflec- 'j tor regions plays an important role in the use of the FNR as a ,,
research reactor, emphasis has been placed on determination of - the thermal flux distribution both for LEU and HEU core con-figurations. Based on the favorable comparisons noted in early 4 simulations
- of the rhodium self powered neutron detector (SPND) '
.j and iron wire activation data, all of the thermal flux mappings j for the LEU configurations were performed with the SPND. Subse- .: quent analysis and simulation of the SPND data, however, indi-9 cated. considerable discrepancy between the calculated flux dis- q i tributions and<the.SPWD data. This is in part due to resetting ; of the reference vertical position for the SPND. To resolve this 4
discrepancy between the calculated results and SPND maps, both g iron and rhodium wire activations'were performed for an HEU core j configuration in October, 1982. We performed simulation of the recent HEU data as well as that of an LEU configuration of April ^ 2 16, 1982 and of the HEU configuration of September 27, 1979 q analyzed in Ref. 5. m
- p For the purpose of our comparison here, the thermal flux j calculated through the 2DB-UM code, with a thermal cutoff of
'i 0.625 ev, is assumed to correspond to the integrated neutron flux below the cadmium cutoff. The 2DB-UM calculations were performed a
1 I ,t 1: o , t . xgm= m _ -
- .. . - - - ~ _ - = - - - --
m
~
gN , L.._ .:.a % . -
.. . , .g, . . 3. ,, .. .. A ._,_, 49 $_
1.
);. .
h L . 12 q-u , 1 s r through the steps outlined above, with 6x6 meshes per fuel ele-S'-- ment. Comparisons between the calculated thermal flux distribu-3 tions and the SPND data are shown for the September 1979 HEU and t April 1982 LEU cores in Figure 3 and 4, respectively. Data are Li j _ given for all of the core locations where measurements were made, p and also for four locations in the D 0 tank on the north side of the core and four locations in the H O reflector on the south side of the core.
]4 ;- The fluxes are no malized so that the sum of j, the measured and calculated fluxes within the core are the same.
The agreement between calculation and measurement'is generally y . good in the core, with some underprediction of the thermal flux
?w peaking in the water hole in the special fuel elements. A large f underprediction is noted, however, in the 2DB-UM simulation of , the SPND data in the heavy water region. '
h~ Figure 5 compares the thermal flux distributions on a north-
,' south traverse through the center of the 1979 HEU core. In addi-tion to the rhodium SPND data, iron wire activation data are also included in Figure 5, with the fluxes normalized at L-37.
e n Similarly, Figure 6 compares the SPND data and 2DB-UM results f or
;t . an east-west traverse through the center of the 1979 HEU core.
l<- Figure 5 indicates a good agreement between the iron wire data
.]
11 and the 2DB-UM results, both of which are substantially lower than the SPND measurements in the heavy water. Flux traverses- . for the April 1982 LEU core are compared in Figures 7 and 8 7 with a similar discrepancy between the 2DB-UM results and SPND data in j, the heavy water region. To understand this discrepancy, we com- . J - pare in Figure 9 the 2DB-UM results with the iron and rhodium wire activation data and the rhodium SPND data obtained for the October 1982 NEU core. Figure 9 indicates that the 2DB-UM cal-a culations are in reasonable agreement with the wire activation ? data, while a substantially higher thermal flux i.s obtained from the SPND data. The.large discrepancy between the thermal flux distributions s' obtained with the SPND, and the corresponding wire activation h data and 2DB-UM calculations is currently under investigation. 4 It appears that several f actors may have to be accounted for f- before this discrepancy can be resolved. One factor that has not
- j. been considered explicitly in the 2DB.-UM calculation is the ef-E .
i.'. feet of the' aluminum and H O that surround the detector when measurements are made in t$e D o tank. .The measurements are made 3 in tank. aluminum tubes that are filled with H O that penetrate the D 0
- f. The water in the tubes tends to i$ crease the thermal flud K
d~ that the detector would see. Another factor that needs to be j, taken into in spectrum account the coreis the anddifference that in thebetween D3 0 tank. the thermal neutron 4 The spectrum is H- -softer in the D o tank, which increases the effective absorption cross section of the detector, yielding _a larger detector current there. i(, The thermal flux depression is expected to be large in the rhodium SPND'because of the large resonance in rhodium at 1.2 eV and also because of the Inconel paddle that the detector is .i
- [.
a ~' , .
, saw . . . . . . .A_. '
. ..u _ : .._~
,j . 1 . !j - - .6,1 .
,3 13 l 3 !
c l DD p1 f 2 i 4 Exp 1.48 1.35 [h
,:2 2DB I 1.00 .995 I s.
l.k
- ft Heavy Water Tank 3 1.87 1.83 'l .
I d X 1,yg W 1,17 M -
. l
[c p t ,n 6 s i
i h .821 1.01 1.22 1.20 1.12 .915 '
L:
- c .
.755 .936 1.12 1.14 1.07 .895
( < 5 l ,. A 1.45 C ,ll J- 1.57
; .683 .950 2.70 1.49 1.48 1.45 1.29 .944 l !) . .634 .949 2.17 1.53 1.57 1.48 1.19 .831 ,
1.37
,, 1.47 3 . ~'
2
,, g .837 .984 1.07 1.02 .884 l'4
.837 1.02 1.07 .995 .855 'l
\
N .717 .678 (j.
.705 .652 {
. r, 2
ii ,
- n 1
- t
- t
-?i Regular Special Empty ( '9 .d - Ele:nent Element Location i EI 0 2
.k
- p Figure 3. Thermal y .859
.810 ]
[.} .843 ^ Mi Flux Maps for the- 2
}j September 27, 1979 *838 q HEU Core ,727 j ,3
.751 F .612 4
l
.598 , 7
[. .
- l. ". . . - m.,m w i . . . - . - . . . . . - . . . ~ - - - - - - . - - -
~
> a_ _ c _ _ _ . ..- ,
-s .,qy.p. 7. _.g ;_.g - . 7y, ,.. . ...-. . . -- a U M t . .
R ' . ., , i: :: . 1, y : 1 ,
.i, , I n D0 2 a
1 9- EXP 1.70 1.51 d ' 9.- 2DB S R 1.02 .975
-l ' ' .i -) n 3
e Heavy Water Tank g- 2.23 2.12 i X W
;. .1 18 1.12 1 ,~
t' i 9 ~
.788 1.04 1.16 1.12 .954 .679
' .727 .879 .977 .957 .827 .637 Il "
1.23 a i 1.29 l j- .861 2.63 1.33 1.28 1.13. .824.
?! -
2.17 1.39 1.31 1.14 .852 ,
.l -
1.10 l 1.26
.1. . -
f}
.628 .753 .783 .660 .498
?.
.784 .887 .887 .774 .573 b
t tj - d-I! 1 ij) ' Regular Special Empty
- L Element Element Location i H,30
- 9
<1 1.32 gL 1 1.33
- d. 1.32
@ 2 jI . 1.50
'J Figure 4. Thermal Flux $ Maps for the April 16, 3 1.16
- , 1982 LEU Core. 1.40 p.
i 1.01 4 1.17 9
T& ::hclkhSDi.iL;l,iiLii%A'.: sLa.r a-
> ; I- - * ^-' ^- I A ~~~ A '
~^~"*
'" ~ ~
i
?!
cT
.h
, ~ t;
. ~.
1 f, l' n. t.' l-9/27/79 2.5 , , HEU CORE WORTH-SOUTH TRAVERSE ! e HO2 CORE DO HO 2 2 ; h d x 2.0
--)
L-37 - i J e M Iron Wire ! L'- e Rhodium - SPND f
- _.J l 4 *5 -
[ 2 - - 'J g at L-37 m Ld r 1 i W e' I td 1.0 - l 2 - p H .i,
}
4 . -
.J -
Ld
- 0.5 - -
0.0 ' ' ' ' ' ' '
!. 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 .
t POSITION (CM) ' Figure 5. Thermal Flux Profiles along the North-South Traverse, . September 27, 1979 IIEU Core. 7 g . ., . 4 t .- ._
. . :: m c.c. . .. x.. .. ww. . . . . ..,,......~.,..a.;.. _ a.~.-. ., n = - a :. .. :.w ..a x.'.z: m ,.:;..ss.
.-....x. : n:..
. I~' ' -
( , i F- t. e i . g (J--- p g ,. g. ,.'
.. i
~
. j 9/27/79 3.5 ,
HEU CORE EAST-WEST TRAVERSE i , ,, , , , i , i i i
~
11 O COR.9 3.0 _ 2 't O 2 _ X D
- Rhodium SPt'D
" 2.5 La_
2o' - ' d ;.. e 4 Normalized at L-37 $ 22.0 m w j - m ' x : F-w 1.5
- - :6
$ 1.0 - a w
0.5 - -
~
0.0 ' ' ' ' ' ' ' ' ' ' 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 .jk Figure 6. POSITION (CM)' @ Thermal Flux Profiles along the East-West Traverse, u. Cepte:nber 27, 1979 IIEU Core y;,
.33
.p-i
.:..a. - = =
x . = . = . . . . =:..
. = .. .:. = . . = . = w....a.
. a =.c.. .
z.
.. , , ;,, , , i e .
't s
c t n t p t 4/16/82 2.5 i i LEU CORE NORTH-SOUTH TRAVERSE i i i i i i i i i i i . p" 3 DO 2 Hjo CORE , HO 2 ,; {? t jr i ' 4- 2.0 - - 6 { a -2 DB '
' J f , e Rhodium ,
SPND f
; 4 y 1.5 i,
- L-37 _
g , , Normalized ' ! Ld at L-37 ' I - y. l Ld 1.0 - -
> a p l
<J .
Ld cr. 0.5 - -
~
0.0 l 0 10 20 30 40 50 60 70. 80 90 100 110 120 130 140 -
-~
POSITION (CM) Figure 7. Thermal flux Profiles along the North-South Traverse, . April 16, 1982 LEU Core .
.. j i .
. . < 4
- . x. L s,a, ~121L2. . ir.huinM:J ._. t. Jan .D.Juz a n. v
.... . .L J.~iz r u~.+.:1:.as u : J ~.L".. L r.: ;;c:. r. c :a , . . . .:r. J - s . . . . s .w I F i - i i i F i l [r-- e ' I y ; . I ; , g g
, 7 4
4/16/82 3.5 LEU CORE EAST-WEST TRAVERSE h" 1 I I 3 I I I I i i i I I'30
- COPR :: o 3.0 -
7 X ~ 3 c' 2.5 - PMium SPPD _
- 2Drt 4
y , Ibrmilized at L-37 5 W I l-- g 1.5 -
> L p~ -
- y 4 1.0
._J F
Ltl , E h. n' O.5 - . _ E w Q,Q l i I I I I
-5 I I i O.
o n 10 20 30 40 50 so 70 80 90 100 110 120 130 140 d;; POSITION (CM) Figure D. TI ermal Plux Profiles alpg the PJtst-t%st 7 averse, F April 16, 1982 Core . I'
' K - u i..' w :)La: L ' ..:!:.., .9i :: G '.3
. . La.L ': -
G.,_av ..
. ...i u R am :n s L M & f.r;i.:. k :u b a % . W ...,
p o ..' f lC
-t h [.
h-l' - t . T 10/6/82 HEU CORE NORTH-SOUTH TRAVERSE 2.5 i i i i i i i i i i i i i
., a .. = w . a . :.
I. 4 a I { ; f; i 1982 LEU AND HEU NORTH-SOUTH TRAVERSE 2.5 , , , , , ,- , , , , , , , HO CORE DO HO '. 2 y 2
. i xD 2.0 - -
i' . J , i k i LEU HEU I: J ' j . j 1.5 - LEU - # 'i r x , W LEU T.
, W w 1.0 - -
'l.-
l 2 i 4 HEU V J - y p ..-
" 0.5 -
i .: Q 0.0 'b 0 10 20 30 40 50 60 70 80?90 100 110 120 130 140 POSITION (CM)- - Figure 10. Comparison of Thermal Flux Profiles between the ' j LEU and lieu Cores . p.
, i -
V
, m:,pI(2,y. cp . ,Nily y'i. i , '
r ww, . _~ 3 ~Z.1 i AE lhNh ,
.s..
i
~' '" ^
'7.-. . . '
)4 i
, a c: 32
~
s , by the discrepancies mentioned above, we consider our present ]- y calculational package to be a useful, reasonably accurate, and 6 h efficient system for performing analyses of MTR LEU /HEU core con-y_ figurations. M References N h- 1. D. K. Wehe and J. S. King, " Ford Nuclear Reactor High j--. - Enrichment / Low Enrichment Demonstration Experiments--Part
- s I", presented at the International Meeting on Research and 2
Test Reactor Core Conversion from HEU to LEU Fuels, Argonne i National Laboratory (November 8-10, 1982). a . A_ 2. D. K. Wehe and J. S. King, " Ford Nuclear Reactor High j Enrichment / Low Enrichment Demonstration Experiments--Part p4 _ II", presented at the International Meeting on Research and Test Reactor Core Conversion from HEU to LEU Fuels, Argonne S~ National Laboratory (November 8-10, 1982). ! n
- 3. D. K. Wehe and J. 5. King, "The RERTP Demonstration Experi-ments Program at the Ford Nuclear Reactor", presented at the International Meeting on Research and Test Reactor Core Con-
, version from HEU to LEU Fuels, Argonne National Laboratory -
4 ,, (November 12-14, 1980). , d 4. D. C. Losey et al., " Core Physics Analysis in Support of the FNR HEU-LEU Demonstration Experiment", presented at the h - 1 International Meeting on Research and Test Reactor Core Con-p version from HEU to LEU Fuels, Argonne National Laboratory (November 12-14, 1980). il 5. W. Kerr et al., " Low En'richment Fuel Evaluation a'nd Analysis h Program", Department of Nuclear Engineering and the
~ Michigan-Memorial Phoenix Project, The University of Michigan, Ann Arbor, Michigan (January 1980). *
- 6. R. F. Barry, " LEOPARD--A Spectrum Dependent Non-Spatial 3- Depletion Code", WCAP-3269-26, Westinghouse Electric Cor-M poration (September 1963).
- O. 7. H. Bohl, E. M. Gelbard, and G. H. Ryan, "MUFT-4--Fast Neutron Spectrum Code for the IBM-704", WAPD-TM-72, Wes-j tinghouse Bettis Atomic Power Laboratory (July 1957).
Y j~ 8. H. Amster and R. Suarez, "The Calculation of Thermal Con-
- d. stants Averaged over a Wigner-Wilkins Flux Spectrum; y Description of the SOFOCATE Code", WAPD-TM-39, Westinghouse 3-,
Bettis Atomic Power Laboratory (January 1957). Yq 9. J. Barhen, W. Rothenstein, and E. Taviv, "The HAMMER Code 9 ... System", NP-565, Electric Power Research Institute (October p 1978), a fL 10. W. W. Little, Jr. and R. W. Hardie, "2DB User's Manual--
- L -
s h.: f
- - y a;
" ~7
-,.., w +. ; - a; w. s..
. . . 2. . .~ . . - . ~ - - - . . -a s -. p. - . . . - -
8 - 4
!~ P 5 ] 23 i
I..' fp Revision 1", BNWL-831 REV1, Battelle Pacific Northwest
? Laboratory (February 1969). ~
b
;; 11. D. R. Vondy et al., " VENTURE: A Code Block for Solving the ] Finite-Difference Diffusion Approximation to Neutron ij Transport", ORNL-5062, Oak Ridge National Laboratory (1975).
J 3 12. R. W..Hardie and W. W. Little, Jr., "PERTV--A Two-gj D'mensional Code for Fast Reactor Analysis", 3NWL-1162, Bat-9 telle Pacific Northwest Laboratory (September 1969). a
- 13. K. D. Lathrop and F. W. Brinkley, "TWOTRAN-II--An Interfaced L,k i Exportable Version of the TWOTRAN Code for Two-Dimensional i Transp rt", LA-4848-MS, Los Alamos Scientific Laboratory
;.! (1973).
- 14. W. W. Engle, Jr., "A User's Manual for ANISN, a One-
' Dimensional Discrete Ordinates Transport Code with c Anisotropic Scattering", K-1693, Oak Ridge Gaseous Diffusion d Plant (March 1967).
1 i' 15. D. R. Harris, "ANDYMG3, The Basic Program of a Series of D Monte Carlo Programs for Time-Dependent Transport of Par-ticles and Photons", LA-4339, Los Alamos Scientific Laboratory (1970). -- i' 16. Private Communication, W. D. Henderson, Westinghouse Electric Corporation, to J. C. Lee, University of Michigan (August 1, 1980). .
- j. 17. W. Jaschik and W. Seifritz, "Model for Calculating Prompt-1 Response Self-Powered Neutron Detectors", Nucl. Sci. Eng.,
j 53, 61 (1974). - si l st A *
.i Vk H * .; i k;5 '
i, n - N . fc) . _ a .
.- p
. . ..a...-_ .. . . . .
.. . _ . -- .- ._. :_..._: .~ .. . . . _ . . . . . . . . . _
-- + . . .
I I ct
!J (i
- 4 -
4
- j 1
s I.i 6 9 4 N. 4 0 g 4
-t ,
t 4 e a 1 i d O 9 I4 d e 0
'J e
'd J
.' l-e
>^
t 9 1
'h
'5 1 p I ,]
. 1 i
1
^.e v+ v n . s a m.. x - gg , .m,. ,
, _ y ='y p g ,
e__ _1 , g ,p j , , ,r}}