G.D. ROSSITER*, P.M.A. COOK, R. WESTON - MOX Research and Technology, BNFL, Sellafield, Seascale, United Kingdom
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
XA0102592
ISOTOPIC MODELLING USING THE
ENIGMA-B FUEL PERFORMANCE CODE
G.D. ROSSITER*, P.M.A. COOK, R. WESTON
MOX Research and Technology,
BNFL, Sellafield, Seascale,
United Kingdom
Abstract
A number of experimental programmes by BNFL and other MOX fabricators have now shown that the
in-pile performance of MOX fuel is generally similar to that of conventional UO2 fuel. Models based
on UO2 fuel experience form a good basis for a description of MOX fuel behaviour. However, an area
where the performance of MOX fuel is sufficiently different from that of UO2 to warrant model
changes is in the radial power and burnup profile. The differences in radial power and burnup profile
arise from the presence of significant concentrations of plutonium in MOX fuel, at beginning of life,
and their subsequent evolution with burnup. Amongst other effects, plutonium has a greater neutron
absorption cross-section than uranium. This paper focuses on the development of a new model for the
radial power and burnup profile within a UO2 or MOX fuel rod, in which the underlying fissile isotope
concentration distributions are tracked during irradiation. The new model has been incorporated into
the ENIGMA-B fuel performance code and has been extended to track the isotopic concentrations of
the fission gases, xenon and krypton. The calculated distributions have been validated against results
from rod puncture measurements and electron probe micro-analysis (EPMA) linescans, performed
during the M501 post irradiation examination (PIE) programme. The predicted gas inventory of the
fuel/clad gap is compared with the isotopic composition measured during rod puncture and the
measured radial distributions of burnup (from neodymium measurements) and plutonium in the fuel
are compared with the calculated distributions. It is shown that there is good agreement between the
code predictions and the measurements.
1. INTRODUCTION
BNFL has undertaken a comprehensive programme of work in order to support the
development and qualification of its short binderless route (SBR) MOX product [1]. The
programme includes characterisation of as-fabricated fuel, test reactor irradiations, post
irradiation examination (PIE) of commercially irradiated fuel and fuel modelling.
Modelling work is primarily carried out using the ENIGMA-B fuel rod performance code.
The ENIGMA code was initially developed by BNFL and British Energy, to perform fuel
safety analyses in support of the Sizewell-B PWR. Since 1991, BNFL has independently
developed the ENIGMA-B version of the code as a versatile tool, to model the in-pile
behaviour of UO2, UO2-Gd, UO2-Nb and MOX fuels. This version of the code has been used
to support fuel licensing in Finland, Switzerland and the UK. Current model developments are
mainly focused on MOX fuel [2].
Experimental programmes performed by MOX fuel fabricators, including BNFL, have shown
that, in general, the in-pile behaviour of MOX fuel is similar to that of conventional UO2 fuel
[3,4,5,6,7]. These programmes have also shown that fuel performance models based on UO2
experience provide a sound basis for predicting the irradiation behaviour of MOX.
* Currently address: OECD Halden Reactor Project, PO Box 173, N-1751 Halden, Norway.
227However, there are at least four areas where the performance of MOX fuel is sufficiently
different from that of UChto warrant model changes. These are:
• radial power and burnup profile
• fission product and helium generation
• thermal conductivity
• fuel creep.
This paper concentrates on modelling the radial power and burnup profile in MOX fuel,
which is one area where UO2 models must be modified to accurately describe the
performance of MOX fuel. ENIGMA-B models the radial power and burnup profile using the
RADAR (Rating Depression Analysis Routine) model [8]. RADAR has been modified to
track the creation and depletion of plutonium and other actinide isotopes during irradiation.
An underlying assumption has been made that the plutonium distribution in the fuel has a
high degree of homogeneity. The model has been further extended to predict the isotopic
composition of the fission gases, xenon and krypton.
The predictions of the new RADAR model are compared with EPMA (electron probe micro-
analysis) linescans for the elements, neodymium and plutonium, taken as part of the M501
PIE programme [3, 4]. The predictions of the fuel/clad gap gas inventory are also compared
with results from rod puncture tests [3,4].
2. BACKGROUND TO RADAR MODEL CHANGES
As a result of fission and neutron absorption, the isotopic content of the fuel constantly
changes during irradiation. The fission cross section is different for each of the primary fissile
isotopes in thermal reactors, 235U, 239Pu and 241Pu, in particular for the resonances in the epi-
thermal energy range. The neutron absorption cross-section, the flux and the isotopic
concentration determine the extent to which absorption occurs. All the minor actinide nuclei,
such as curium, are generated in this way. The presence of fertile species in the fuel, such as
238
U and 240Pu, allows the creation of new fissile nuclei to partially replace those consumed
during earlier fission events. It is therefore necessary to track the isotopic content of fissile
and fertile species within the fuel, to be able to accurately predict the local fission rate within
the fuel.
From the number of fissions in the fissile isotopes, the known isotopic yields can then be used
to determine the concentration of fission products in the fuel. The yields for the lower mass
fission products, such as krypton, are significantly different in uranium and plutonium
fissions, so calculating the inventory of certain key fission products provides useful additional
fuel performance information. It also means that the amount of experimental data that can be
used to validate models of the radial power, burnup and fissile isotope distributions used in
fuel performance codes is increased.
3. ENIGMA-B MODELLING
3.1. Earlier RADAR models
Many fuel performance phenomena, including fission gas release and densification, are
strongly temperature dependent. An accurate prediction of the intra-pellet radial power
distribution and its evolution with burnup, together with the fuel thermal conductivity, is
therefore essential in any good fuel performance code. The fuel thermal conductivity
modelling in ENIGMA-B is described in [9, 10] and shows good agreement with experiment.
228The RADAR model is implemented to calculate the intra-pellet radial power distribution for
both UO2 and MOX fuel. The fuel pellet is divided into a number of radial annuli for the
calculations.
In earlier versions of the model simplified calculations of the depletion of the 235U, 23 Pu and
241
Pu isotopes were performed, using one-group fission and neutron capture cross-sections
tuned to match predictions from the WIMS-E neutronics code [11] for a typical PWR reactor
environment. The thermal neutron fluxes, in each of the annuli, used in evaluating the
depletions were calculated by solving an idealised form of the neutron diffusion equation,
based on an approximate inverse diffusion length in each annulus. The start-of-life 239Pu and
241
Pu contents were combined into an "effective plutonium content," loosely intended to
represent the fissile plutonium content, which was depleted as 239Pu. Plutonium is generated
by the capture of both thermal and epithermal neutrons. Both of these mechanisms were
modelled. The total amount of Pu generated by resonant capture in U was calculated
from nuclear physics parameters and the radial distribution of the captured material was
f (r) = c(l + 3 exp(- 9 . 7 ^ ^ ) ) (3-1)
defined by an empirical function j(f) of the form [8], where ro is the pellet outer radius and C
is a constant, whose value is determined from normalisation considerations.
The radial power distributions determined in this way were satisfactory for MOX fuel at low
burnups and with low plutonium contents. However, recent studies have shown some
deficiency at higher burnups and/or high plutonium contents, in particular in the soft spectrum
of the Halden Boiling Water Reactor (HBWR). Hence, an improved version of the RADAR
model has been developed.
3.2. Current RADAR model
The improved RADAR model is significantly more sophisticated than the earlier versions.
The key modifications are as follows:
(a) the heavy metal isotope tracking has been expanded to include a more extensive number
of actinides. A simplified schematic is illustrated in Figure 1,
(b) thermal and resonant neutron capture in both U and Pu and fast fission in U are
modelled explicitly. The radial distributions of the material undergoing resonant capture
are described by Equation (3.1), for both 238U and 240Pu, following the approach in [12],
(c) reactor type and 235U or plutonium enrichment dependent fission and neutron capture
cross-sections are modelled for each of the heavy metal isotopes,
(d) the use of reactor type and 5U or plutonium enrichment dependent transport cross-
sections for the heavy metal isotopes and for oxygen in the calculation of inverse diffusion
lengths,
(e) the cross-sections modelled are obtained from correlations fitted to predictions from the
CASMO-4 [13,14] nuclear physics code, using the 70 group ENDF-B nuclear data library.
229d ?4I
Cr i
16r
2
-'Am
14 4y
2-'«pu
' Pu — « - - " P u — * — = Pu
2J s 3 40
2.. 55d
S33
Np - * - n e u t r o n cap ture
+ b e t a decay
I 23.5m
|-«- clpho d e c a y
23S 233
U U
. i. Simplified schematic of heavy metal isotope modelling.
The reactor types, which can be modelled, include commercial PWRs and BWRs and the
Halden reactor. The modelling of the americium and 242Cm isotopes allows accurate
calculations of helium generation due to a-decay to be performed, which has been described
previously [2]. The calculations performed by RADAR are necessarily highly simplified
when compared to those performed by neutronics codes. The applicability of the modelling
was therefore assessed by comparing the isotopic number density and radial power profile
predictions of the improved RADAR model with those of CASMO. Excellent agreement was
observed. The results for MOX fuel of 5 wt% plutonium in a typical PWR environment and
with a standard plutonium vector are illustrated in Figure 2.
3.3. Fission gas generation model
A new fission gas generation model has been integrated with the existing fission gas release
model, to complement the improved RADAR model. The generation model calculates the
through-life radial distributions of each of the krypton and xenon isotopes 83Kr to 86Kr and
131
Xe to 136Xe, plus the important precursors 131I and 132Te, in each of the fuel annuli.
Generation, neutron capture and decay are computed based on the power history, the
distribution of fission events of each of the fissile species (as calculated by RADAR) and the
fission gas isotopic distributions. The new model allows a more accurate calculation of the net
fission gas generation rate, which takes account of the different krypton and xenon isotopic
yields of the various fissile isotopes. In addition, the new generation model allows the isotopic
composition of the fission gas in the fuel-clad gap to be evaluated, as the decay and neutron
capture in the gap is also modelled.
4. VALIDATION
All the validation data shown in this paper has been taken from the M501 PIE programme.
M501 was one of the first four assemblies of SBR MOX fuel to be fabricated in MDF1 for a
commercial reactor. The assemblies were irradiated for three cycles in NOK's Beznau-1 PWR
to an average burnup of 33 GWd/tHM. After discharge, seven rods were withdrawn and sent
for PIE at ITU (Trans-Uranium Institute, Karlsruhe). Results from this programme have
already been published in the open literature [3,4]. Experimental data from the four high
enrichment rods are used in this paper.
1
MDF, MOX Demonstration Facility, Sellafield, UK.
230Vertica! scale: number density or relative mass rating CASMO RADAR
: U235, O.I MWd/kgHM T U235, 10 MWd/kgHM T U235, 30 MWd/kgHM T U235, 60 MWd/kgHM
U238,0.1 MWd/kgHM U238,10 MWd/kgHM T U238, 30 MWd/kgHM U238, 60 MWd/kgHM
_j—M501 : Power Histories
-Rod1
-Rod 2
Rod 3
Rod 4
-Rod 5
-Rod 6
-Rod 7
5 10 15 20 25 30 35
A s s e m b l y average burnup (MWd/kgHM)
FIG. 3. M501 power histories for all PIE rods.
Water rod Low enrichment Med. enrichment High enrichment
FIG. 4. Assembly M501, showing positions of the PIE rods.
4.1. Isotopic fission gas inventory
The isotopic composition of the fission gases was measured for all of the M501 rods. Figure 5
shows the comparison of the code's predictions of the isotopic content of the gas in the fuel-
clad gap with the PIE measurements, for the four high enrichment rods.
Figure 5 shows good agreement between the modelled and experimental results. The gas
volume is least well predicted for 131Xe. The over-prediction is due to an under-prediction of
the amount of the isotope undergoing neutron capture to 132Xe. The results for rods 4, 5 and 6
232— P/M=1
* Rod 4
x
Rod 5
* Rod 6
• Rod 7 Xe132
£
a
0.01 -^
0.01 0.1 1 10
measured value (cc@STP)
FIG. 5. Predicted versus measured volumes for fission gas isotopes.
lie very close to the P/M = 1 line. There is a slight under-prediction in general for rod 7,
which has the highest buraup and for which the centre temperature in a section of the rod just
exceeded the Vitanza threshold for fission gas release [15] at the end of its third cycle of
irradiation [4].
4.2. EPMA radial linescans
EPMA radial linescans for plutonium and neodymium were performed on two samples from a
M501 high enrichment rod. Table 2 lists the samples examined by EPMA and their locations
relative to the bottom end cap. For each sample, measurements were performed at 45 points
across the fuel radius. The measurements correspond to the amounts of plutonium and
neodymium in the fuel matrix.
Table II. EPMA SAMPLES
Sample Rod Initial Pu / U+Pu (%) Rod Bumup (GWd/tHM) Location (cm)
A 7 5.54 35.6 60
B 7 5.54 35.6 259
4.2.1. Radial burnup profile
The local buraup is determined by measurements of the neodymium concentration [16]. The
total amount of neodymium present in the sample is increased by decay of 144Ce and therefore
the cooling period between discharge and measurement is taken into account. The fission
yield used to derive the buraups below has been calculated using the FISPIN fission product
inventory code [17] and has taken neutron absorption and radioactive decay into account.
Figures 6 and 7 compare the local burnup profile calculated by ENIGMA-B with the bumup
profile inferred from EPMA neodymium measurements. The pellet was divided into 70 radial
annul i for the ENIGMA-B calculation.
23380 !
ENIGMA (Sample A)
< EPMA (Sample A)
70 -
f
5 60 -
a 50-
a
40
I
n
1
•
8 30 -*
x v x x x x
x x x X * X X X
o
_i
20 -
10 -
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
R/RO
F/G. 5. Local burnup profile for sample A.
90 -
80 -
K
ENIGMA (Sample B)
70 - * EPMA (Sample B)
S
5 60
a so
40 -
20 -
10 -
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
R/RO
FIG. 7. Local burnup profile for sample B.
In general, there is good agreement between the calculated and experimental data for both
samples.
4.2.2. Radial plutonium profile
Figures 8 and 9 compare the prediction of the radial plutonium profile with the EPMA
measurements of the plutonium distribution at end of life. The total initial plutonium content
is also plotted.
2346 l
5 "
X*5
x x * * x * x x * x * «x*
X * X *
Ie 3
0
o
21
— Unirradiated Pu content
* Irradiated EPMA (Sample A)
K
1- ENIGMA (Sample A)
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
R/RO
. 5. Radial plutonium distribution for sample A.
61
5 '
- 4 -
* * * * ******
C
0
o
1
2\
"•"• Unirradiated Pu content
* Irradiated EPMA (Sample B)
* ENIGMA (Sample B)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
R/RO
FIG. 9. Radial plutonium distribution for sample B.
hi general, there is good agreement between the modelled and experimental data for both
samples, though there is a slight over-prediction of the plutonium content for Sample B
(Fig. 9). It should be noted that the decay and capture of 238Pu and 242Pu are not tracked in the
new RADAR model. Thus the predicted plutonium content, which is plotted, is 239Pu + 240Pu
+241Pu. However, the initial percentages of 238Pu and 242Pu are 0.013 wt% and 0.065 wt%
respectively, so the above is a good approximation to the total plutonium content.
The predicted radial plutonium distributions, for the isotopes 239Pu, 240Pu and 241Pu, are
illustrated in Fig. 10. They show the expected form of the radial distribution of each of these
plutonium isotopes. There is an increase in the concentration of 239Pu and 241Pu, and a
:-240 238 T
decrease in the concentration of Pu, at the edge of the pellet, due to resonant capture in U
and 240Pu.
2355 " * Irradiated EPMA (Sample A)
• ENIGMA (Sample A)
4.5 " x Pu239
a
Pu240
4" ;
PU241
3.5 "
3"
S 2.5 -
c
o
1 2-
1.5 " :x x x x x
1
0.5 "
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
R/RO
FIG. 10. Radial distribution ofplutonium isotopes for sample A.
hi Figures 6, 7, 8 and 9, it can be seen that there is some discrepancy between the measured
and predicted radial distribution of burnup and plutonium at the pellet rim. This is due to the
use of Equation (3.1) in the model to determine the radial distributions of both 238U and 240Pu
undergoing resonant capture. Previous work [18] has indicated that this equation is not fully
optimised against available EPMA data. This appears to be confirmed by the comparisons of
measurements and predictions described in this paper. Hence, further development of the
RADAR model is anticipated in the area of pellet rim modelling, using EPMA and SIMS
data, when available, to obtain an optimised function.
5. CONCLUSIONS
This paper presented a new version of the RADAR model for predicting the radial power and
burnup profile during irradiation, based on the tracking of the underlying fissile isotope
distributions and their evolution with burnup. The new model has been incorporated into the
ENIGMA-B fuel performance code, together with calculations of the fission gas generation
isotopics.
The ENIGMA-B code predictions were compared with experimental data from the M501
programme, hi general, the measurements and predictions of the radial profiles of burnup and
plutonium distribution were in good agreement. The model also predicted well the xenon and
krypton isotopic content of the fission gases in the fuel-clad gap.
The new RADAR model employs the same function, as earlier versions of RADAR, for
radially distributing the U (and Pu) undergoing resonant capture. Thus, little
improvement was expected in the modelling of the rim region of the fuel. It is predicted that
there will be future development of the RADAR model, using experimental data, to optimise
the function used by RADAR, and so improve ENIGMA-B's performance in the rim region.
236REFERENCES
[I] BROWN C, COOK P, EDWARDS J, FISHER S, GATES G, PALMER I & WHITE R,
"Performance of BNFL MOX Fuel for the Next Decade," TopFuel99, Avignon,
September 1999.
[2] PALMER I, ROSSITER G, & WHITE R, "Development and Validation of the
ENIGMA Code for MOX Fuel Performance Modelling," IAEA MOX Fuels
Symposium, Vienna, May 1999 IAEA-SM-358/20.
[3] COOK P, PALMER I, STRATTON R & WALKER C, "PIE of BNFL's First
Commercially Irradiated SBR MOX Fuel," IAEA MOX Fuels Symposium, Vienna,
May 1999IAEA-SM-358/16.
[4] COOK P, STRATTON R & WALKER C, "PIE of BNFL MOX Fuel," ANS
International Topical Meeting on LWR Fuel Performance, 9-13th April 2000, Park City
Utah USA.
[5] BLANPAIN P, TFUBAULT X & TROTABAS M, "MOX Fuel Experience in French
Power Plants," International Topical Meeting on LWR Fuel Performance, West Palm
Beach USA, April 1994.
[6] GOLL W, FUCHS H, MANZEL R & SCHLEMMER F, "Irradiation Behaviour of
UO2/PuO2 Fuel in Light Water Reactors," 1993, Nuclear Technology 102 pp.29.
[7] LIPPENS M, MALDAGUE T, BASSELIER J, BOULANGER D & MERTENS L,
"Highlights on R&D Work Related to the Achievement of High Burnup with MOX
Fuel in Commercial Reactors," IAEA MOX Fuels Symposium, Vienna, May 1999
IAEA-SM-358/15.
[8] PALMER I, HESKETH K & JACKSON P, "A Model for Predicting the Radial Power
Profile in a Fuel Pin," IAEA Fuel Modelling Meeting, Preston, March 1982.
[9] GATES G, COOK P, DEKLERK P, MORRIS P & PALMER I, "Thermal Performance
Modelling with the ENIGMA Code," OECD/IAEA Meeting in Cadarache, March 1998.
[10] ROSSITER G, "Modelling of MOX Fuel Performance Using the ENIGMA Code,"
from "Minutes of the Workshop on Fabrication, Performance and Modelling of MOX
Fuel, Halden 28-29 September 1999," HWR-623, November 1999.
[II] ASKEW J R & ROTH M J, "WIMS-E: A Scheme for Neutronics Calculations,"
AEEW-R1315, January 1982.
[12] O'CARROLL C, VAN DE LAAR J, WALKER C T, OTT C & RESTANI R,
"Validation of the TUBRNP Model With The Radial Distribution of Plutonium in MOX
Fuel Measured by SMS and EPMA," IAEA Technical Committee Meeting on Water
Reactor Fuel Element Modelling at High Burnup and Experimental Support,
Windermere, England, 19-23 September 1994.
[13] KNOTT D & EDENIUS M, "The Two-Dimensional Transport Solution Within
CASMO-4," Trans. Am. Nucl. Soc, Vol. 68, pp. 457, June 1993.
[14] EDENIUS M, KNOTT D & SMITH K, "CASMO-SBVIULATE on MOX Fuel,"
International Conference on the Physics of Nuclear Science and Technology, Long
Island, October 1998.
[15] VITANZA C, KOLSTAD E & GRAZIANI V, ANS Topical Meeting on LWR Fuel
Performance, Portland, Oregon, USA, May 1979.
237[16] WALKER C, GOLL W & MATSUMURA T, "Effect of Inhomogeneity on the Level of
Fission Gas and Caesium Release from OCOM MOX Fuel During Irradiation," Journal
of Nuclear Materials 228 (1996), pp. 8-17.
[17] BURSTALL R F, "FISPIN - A computer code for nuclide inventory calculations," ND-
R-328(R), (October 1979).
[18] LASSMANN K, O'CARROLL C, VAN DE LAAR J & WALKER C T, "Radial
Distribution of Pu in High Burnup UO2 Fuels," Journal of Nuclear Materials 208 (1994)
223-231.
238You can also read
