Embed Size (px)
Citation preview
STATE CORPORATION ON ATOMIC ENERGY «ROSATOM»
OPEN JOINT STOCK COMPANY
«A C A D E M I C I A N A . A . B O T C H V A R H I G H - T E C H N O L O G Y
SCIENTIFIC AND RESEARCH INSTITUTE OF INORGANIC
MATERIALS»
( JSC VNIINM)
FUMEX-III cases by START-3
Final report
Deputy director Vladimir Novikov
Head of division Sergey Bogatyr
Head of lab Vladimir Kuznetsov
Scientific investigator Dmitriy Chulkin
Moscow, January 2012
1
Introduction ............................................................................................... 2
SUPER-RAMP .......................................................................................... 3
Base irradiation ............................................................................................................................ 3
Ramp details ................................................................................................................................ 4
Pellet and cladding details ........................................................................................................... 7
Calculation results ....................................................................................................................... 8
Versus time and burnup ............................................................................................................... 8
Diameters ................................................................................................................................... 13
Clad Oxidation ....................................................................................................................... 15
Radial gap .............................................................................................................................. 15
Grain size ............................................................................................................................... 16
FGR ....................................................................................................................................... 16
Maximum fuel centerline temperature .................................................................................. 18 Elongations ............................................................................................................................ 18
AREVA high-burnup case ...................................................................... 19
Linear heat rate ...................................................................................................................... 20
Calculated FGR ..................................................................................................................... 20
The changes made into the FGR module .................................................................................. 22 Diffusion coefficients ............................................................................................................ 22
Single atom volume in the gas cluster ................................................................................... 25 Recalculated cases ..................................................................................................................... 27
Simplified cases from FUMEX-II ............................................................................................. 31
Summary and conclusion ........................................................................ 33
FUMEX-III lessons ................................................................................................................... 33
Brief START-3 code description ............................................................ 34
START-3 code designation ....................................................................................................... 34 General description of START-3 code .................................................................................. 34
Radiation Growth ...................................................................................................................... 36
Oxide layer growth model ......................................................................................................... 36
Literature ................................................................................................. 38
2
Introduction
This document covers «SUPER-RAMP» and «AREVA High-Burnup Idealised case»
cases of FUMEX-III project that were calculated by means of START-3 code.
SUPER-RAMP PK-2 rodlets were chosen for calculation because they were subjected to
strong power ramp and did not fail, because the main objective of the present calculations was to
check the models of fission gas release and gaseous swelling of START-3 in the conditions of
the power ramps (we did not intend to get the limit stresses for Zry-4 fuel rods).
AREVA high burnup (PRIORITY CASE) case was chosen to check the applicability of
the START-3 code to high-burnup PWR fuel rods behavior.
Also, in order to check the stability and quality of the changes made into the code during
the course of FUMEX-III participation, several simplified cases of FUMEX-II were recalculated.
3
SUPER-RAMP
In the report some calculations are made on the PK2 group fuel rods. The rods were
standard rods manufactured by Kraftwerk Union AG/Combustion Engineering (KWU/CE). The
rods were ramp tested in the research reactor R2 at Studsvik, Sweden in the framework of
Studsvik SUPER-RAMP Project. All these rods sustained ramping to power levels in the range
41 to 49 kW/m and power changes 16-24 kW/m without failure, in spite of large deformations,
fuel restructuring and fission gas release1.
The Studsvik SUPER-RAMP Project, an internationally sponsored research project,
investigated the failure propensity of typical LWR fuel in the form of test rods when subjected to
power ramps, after base irradiation to high burn-up. The following information summarizing the
project is abstracted from the “Final Report of the Super-Ramp project”, Seved Djurle,
STUDSVK-STSR-32 project (STSR-32).
All rods underwent a thorough examination program, comprising characterization prior to
base irradiation, examination between base and ramp irradiation and examination after ramp
irradiation.
The irradiation history was drawn from the appropriate his files of the Fumex-III compact
disk.
Base irradiation
The KWU/CE test fuel rods were base irradiated in the commercial pressurized water
reactor Obrigheim (KWO) in FR Germany. Reactor characteristics for KWO are given below:
1 Description is taken from file “SUPER-R.SUM” from the FUMEX-III compact-disc provided by IAEA.
4
Obrigheim Power Reactor Characteristics.
Values At 100 % reactor power
Average reactor power (thermal) 1045 MW
Average rod power 171 W/cm*
Coolant temperature at core inlet 283 ºC
Coolant temperature at core outlet 312 ºC
Coolant velocity at 300 C 3.39 m/s
Mass flow rate 6833 kg/s
Average system pressure 14.5 MPa
Active length of core 2650 mm
* 193 W/cm since cycle 10 (August 78)
Ramp details
The power ramp tests were performed according to the following typical scheme,
characterized by the following phases:
1. A conditioning phase, with a rather slow increase of the linear heat rating from an
initial value to a selected value of 250 W/cm (the conditioning level) and holding at the value for
24 hours2.
The objective of the conditioning was to adjust the rod conditions to the same
conditioning level for all rods, thus equalizing the start-point of the ramp tests.
2. A ramping phase with a rapid increase of about 100 W/cm/min from the conditioning
heat rating to a pre-selected ramp terminal level.
3. A holding period at the ramp terminal level of normally 12 hours.
2 We have to add that there is no further clarification of the words “rather slow increase of LHR” in the documents,
provided by IAEA. This leads to uncertainty in the irradiation history and, hence, is a possible cause of errors and unsatisfactory
agreement of calculated and measured results.
5
Exceptions to this normal ramping scheme were the following tests:
- Rod PK2-S was deliberately tested with a coolant inlet temperature 50 C below normal.
- For rod PK2-4 the hold time at the ramp terminal level was intentionally interrupted
after 1 min.
Table 1. Overview of PRW test matrix
Rod ID PK2-1 PK2-2 PK2-3 PK2-4 PK2-S
Burnup
MWd/kgU 45.2 45.1 44.6 41.4
43.4
Fluence
*E25 /sq m 8.1 8.1 8.1 8.1 8.1
Pre-
condition
power kW/m
25 25 25 25 25
Hold time
hrs 24 24 24 24 24
Ramp power
kW/m 41.0 46.0 49.0 44.0 44.0
Ramp rate
kW/m/min 8.5 9.5 8.5 8.5 8.5
Hold time
mins. 720 720 720 1 720
Failed (Y/N) NF NF NF NF NF
Table 1Table 1. Overview of PRW test matrix shows test matrix for PWR type rodlets
used for calculations.
Figure 1 shows history of base irradiation and ramp. The history of the base irradiation
and ramp was extracted from the ASCII PK2-X.HIS.
6
Figure 1 - Linear power history.
7
Table 2 shows pellet and cladding characteristics used in calculations.
Pellet and cladding details Table 2 - Fuel rod characteristics3
Variable
Rod PK2-1 PK2-2 PK2-3 PK2-4 PK2-S
Cladding material Zry-4 Zry-4 Zry-4 Zry-4 Zry-4
Overall rod length, mm 390.26 390.15 390.16 390.12 390.34
Fuel column length, mm 317.4 317.8 319.0 317.2 317.4
Plenum length, mm 32.6 33.0 32.5 32.6 32.8
Cladding OD, mm 10.753 10.752 10.752 10.751 10.754
Cladding ID, mm 9.283 9.283 9.283 9.283 9.283
He fill pressure, bar 22.5 22.5 22.5 22.5 22.5
Fuel weight, g 210 210 210 209 210
Pellet
Enrichment, % 3.21
Pellet density (g/cm3 10.36
Outer pellet diameter, mm 9.14
Inner pellet diameter, mm -
Average grain size, mkm 5.5
O/U ratio 2.00± 0.01
Portion of open porosity of pellet volume, % 3.16± 0.27
3 We have to mention that inner diameter of cladding (9.38 mm) given in the file PRECHAR.PWR contradicts with the
corresponding value of 9.28 given in the PDF file STSR-32, which seems to be correct. The other important parameter that the
SUPER-RAMP dataset lacks is the fuel pellet and cladding surface roughness.
8
Calculation results
The calculation results are given bellow. Figures 2-6 show various parameters
calculated using START-3 code for rodlets PK2-1, PK2-2, PK2-3, PK2-4 and PK2-s. All figures
show calculation results with burnup up to 45 MWd/kgU and ramp time beginning from 28090
up to 28180 hours.
Versus time and burnup
Figure 2 - Calculated fuel centerline temperature versus burnup and time.
Figure 2 shows fuel centerline temperature versus burnup and ramp time. Linear power
was stepwise raised to 25 kW/m and kept at that value for 24 hours when it was raised to values
of 40 to 50 for different rodlets. This resulted in steep peaks of temperature with maximum at
~2200 °C for rodlet PK2-3 and minimum at 1800°C for rodlet PK2-1.
9
Figure 3 - Calculated radial gap versus burnup and time.
Figure 3 shows radial gap versus burnup and ramp time. As shown, gap closes off at ~15
MWd/kgU and 28105 hours respectively. The gap opens up again with the LHR dropdown for
all rodlets except for PK2-4 at ~45 MWd/kgU burnup and ~28150 hours of ramp time. For PK2-
4 gap opens at ~40 MWd/kgU burnup and ~28140 hours of ramp time. After opening gap
remains unchanged for a short period of time and then quickly widens.
10
Figure 4 - Calculated gas pressure versus burnup and time.
Figure 4 shows gas pressure versus burnup and ramp time. During the preconditioning
pressure remains almost constant at the value of ~6 MPa up to the point of raising linear power.
Then it smoothly grows up to the maximum value of 15 MPa for PK2-3 rodlet and minimum
value of 7 MPa for the rodlet PK2-4.
11
Figure 5 - Calculated fission gas release versus burnup and time.
Figure 5 shows fuel gas release (FGR) versus burnup and ramp time. For all rodlets the
FGR values stay constantly lower than 3% during the base irradiation period. During the ramp,
after the raising of linear power FGR grows quickly up to maximum value of 45% for PK2-3
rodlet and minimum value of 4% for PK2-4.
12
Figure 6 - Calculated axial elongations versus burnup and time.
Figure 6 shows cladding elongation versus burnup and ramp time. It shows quick growth
of the elongation at the beginning of the preconditioning to the value of ~2800-3000 mkm and
another one to ~4000 mkm at ~28110 hours of ramp time. Then the elongation gradually drops
until it increases even further during the ramp to values ranging from ~4250 mkm to ~4500 mkm
for different rodlets. Then the elongation slightly decreases until the power dropdown.
13
Diameters
Table 3 lists initial cladding outer diameters. Table 4 lists cladding diameters after ramp
test with oxide layer thickness included.
Table 3 - Initial clad outer diameters.
Rod Initial diameter
(mm)
PK2-1 10.754
PK2-2 10.752
PK2-3 10.752
PK2-4 10.750
PK2-s 10.754
Table 4 - Calculated clad diameters after ramp (including oxide layer thickness), mm
Section 1 Section 2 Section 3
PK2-1 10.799 10.821 10.789
PK2-2 10.827 10.848 10.809
PK2-3 10.836 10.864 10.818
PK2-4 10.763 10.784 10.750
PK2-s 10.806 10.828 10.789
Table 5 and Figure 7 list diameter values after base irradiation and after ramp test.
14
Table 5 - Calculated diameters change (after base irradiation and after ramp), mkm
Rodlet
Section 1 Section 2 Section 3 Experiment (average)
Diameter
change
during base
irradiation
Diameter
change
during ramp
Diameter
change
during base
irradiation
Diameter
change
during ramp
Diameter
change
during base
irradiation
Diameter
change
during ramp
Diameter
change
during
base
irradiation
Diameter
change
during
ramp
PK2-1 -11 56 -11 78 -11 35 -75 139
PK2-2 -9 83 -9 105 -9 58
PK2-3 -12 95 -12 124 -12 66 -85 193
PK2-4 -28 41 -28 62 -28 27
PK2-s -13 64 -13 87 -13 48 -90 111
Figure 7 - Calculated diameters change (after ramp), mkm
15
Clad Oxidation
Table 6 lists cladding oxidation thickness values for different sections of rodlets.
Table 6 - Calculated clad oxidation thickness, mkm
Section 1 Section 2 Section 3
PK2-1 41 41 41
PK2-2 41 41 41
PK2-3 40 40 40
PK2-4 33 33 33
PK2-s 36 36 36
According to STSR-32 the average oxidation thickness for PK2/2 and PK2/4 is
Table 7 - Oxidation thickness according to STSR-32, mkm
PK2/2 19-38
PK2/4 60-72
As we can see from the tables above, the coincidence is good for PK2/2 but it is also bad
for PK2/4. The measured value of 60-72 mkm for PK2/4 seems very strange judging from the
point of view of the adopted by the current version of START-3 oxide layer growth model
[Error! Reference source not found.] since there is no obvious specific reason for this rodlet to
have such a thick (relative to other rodlets) oxide film.
Radial gap
Table 8 lists values of calculated radial gap of different sections after ramp test.
Table 8 - Calculated radial gap after ramp, mm
Section 1 Section 2 Section 3
PK2-1 0.026 0.031 0.022
PK2-2 0.029 0.043 0.029
PK2-3 0.040 0.048 0.036
PK2-4 0.024 0.031 0.021
PK2-s 0.027 0.032 0.026
16
Grain size
The grain growth due to the elevated temperatures can be seen on the
Figure 8. The maximum grain diameter in the center of the pellet is ~42 mkm (PK2-3
rodlet); the minimum is ~11 mkm (PK2-4).
0 1000 2000 3000 4000 5000Distance from pellet edge, mkm
0
10
20
30
40
50
Gra
in d
iam
eter
, m
km
Grain size distributionPK2-1PK2-2PK2-3PK2-4PK2-s
Grain size distribution
Figure 8 - Grain diameter vs distance from fuel pellet edge.
FGR
Fission gas release for PK2 rodlets is demonstrated on the Figure 9 and in the Table 9.
The coincidence is reasonable for all rodlets except PK2-s (relative error 171%). The minimum
relative error is for PK2-3 rod (1%)
17
Figure 9 - Fission gas release after ramp for PK2 rodlets. Calculated value vs measured value.
Table 9 - Fission gas release after ramp for PK2 rodlets. Calculated and measured values.
Rodlet Calculated FGR,% Experiment Relative error,%
PK2-1 27.09 28 -3
PK2-2 38.60 32 20
PK2-3 45.39 44.9 1
PK2-4 3.99 9.5 -58
PK2-s 28.27 10.4 171
18
Maximum fuel centerline temperature
Table 10 lists values of maximum fuel centerline temperature of different rodlets.
Table 10 - Maximum fuel centerline temperature.
Rod Temperature, °C
PK2-1 1873
PK2-2 2083
PK2-3 2202
PK2-4 1918
PK2-s 1923
Elongations
According to STSR-32 report, the measured elongation for PK2-1 during the ramp equals
approximately 400 mkm. As we can see from the Figure 6, this value is smaller than START-3
prediction of 821.8 mkm. The residual elongation for PK2-1 according to STSR-32 is 250 mkm
and the START-3 prediction is 211.1 mkm. The calculations were performed on the fuel stack
length only.
19
AREVA high-burnup case
The data of new idealized AREVA case was provided by Martin Sperlich from AREVA.
Table 11 - Fuel rod characteristics
Length of active zone 3650.0 mm
Plenum volume 8.04 cm**3
Fill gas He
Fill gas pressure 1.6 MPa
Flow area of rod 87.8 mm**2
Cladding
Material Zy4 (stress-relieved)
Inner diameter 8.25 mm
Outer diameter 9.5 mm
Pellet
Material UO2
Enrichment 4.5% U-235
Density 95.0% of theoretical density
Diameter 8.085 mm
Length 13.25 mm
Chamfer height 0.27 mm
Chamfer width 0.5425 mm
Dishing radius 3.0 mm
Dishing depth 0.31 mm
Dishing Volume 8.8 mm**3 (both pellet sides)
Expected FGR values:
End of cycle Insertion time Burnup Expected FGR value
[-] [d] [MWd/kg(HM)] [%]
3 916.4 36.6 0.5 +0.5/-0.2
4 1239.1 49.7 1.9 +1.0/-0.7
7 2141.9 81.5 9.0 +2.5/-2.0
20
Linear heat rate
Figure 10 - Linear heat rate for all 14 sections in AREVA High Burnup case.
Calculated FGR
As one can see from the Figure 11, in the new AREVA HB case the START-3 prediction
of FGR is very conservative. START-3 calculation show approximately doubled FGR.
21
0 20 40 60 80Burnup, MWd/kgU
0
4
8
12
16
20
FG
R(%
)
START-3
Experimental results
Figure 11 - Preliminary FGR result for AREVA High Burnup case.
22
The changes made into the FGR module
Diffusion coefficients
The overprediction of the AREVA High-Burnup (referred to as AREVA HB case in the
following text) case made us to reevaluate the START-3 FGR model parameters. The first place
to look was the diffusion parameters in UO2.
The model of intra-granular processes is based on the description of FGR mono-atomic
diffusion.
For the description of FGR mono-atomic concentration C1(ρ,t) in a grain the below non-
stationary problem of atomic diffusion is solved
21 12
1g eff
C CD G
tρ
ρ ρ ρ ∂ ∂ ∂= − + ∂ ∂ ∂
,
2/)(0 td sx≤≤ ρ
0)2/(1 == sxdC ρ
0)0(1 ==∂∂ ρ
ρC
The following symbols are used in equations:
ρ - current grain radius, m
t - time, sec
Geff - effective source, m-3s-1
Dg - diffusion coefficient, m2s-1
dsx - equivalent grain diameter (governing equations for dsx – see paper [2)]), m
The effective source Geff in the equation of diffusion takes into account FG initiation
during the nuclear fission of heavy atoms, initiation of gas-filled bubbles, capture of FG atoms
by various traps (gas-filled bubbles and initial intra-granular pores), irradiation re-dissolution of
FG atoms from these traps into the fuel matrix.
23
The total coefficient of atomic diffusion of fission gas is accepted as Dg=D1+D2, where
D1 and D2 – thermal and irradiation-stimulated components, respectively [3), 4)].
Diffusion coefficients of FP in UO2 in the literature vary significantly [5)], and the
difference of ~104 between authors is usual thing even for the same temperatures. So is it a
principal question how sensitive the calculation result is to the thermal diffusion coefficient?
Turnbull’s expression for diffusion coefficient for Xe differs from START-3 more than
any other, so we use it to check how the results of START-3 calculations for AREVA High-
Burnup case. Figure 12 demonstrates START-3 calculations with Turnbull’s diffusion coefficient
and START-3 original expression.
0 20 40 60 80Burnup, MWd/kgU
0
4
8
12
16
20
FG
R(%
)
START-3Turnbull D for Xe
AREVA experimental results
Figure 12 - START-3 calculations with Turnbull’s diffusion coefficient and START-3
original expression for the diffusion coefficient.
24
This picture shows no evidence of significant influence of thermal diffusion coefficient
variations. So this mechanism (simple thermal diffusion of monoatoms) is not the main
mechanism for FGR in this case.
Next, we considered the directed motion of the gas bubbles, implemented in START-3.
The bubble mobility in UO2 is given by the expression [6), 7)]:
( ) 3( ) (3/ 4 ) /vb vD r Pi D r= Ω
here Dv – is a self-diffusion coefficient of UO2:
0.3exp( / )
4.6 0.1v m
m
D E kT
E
= −= ±
here Dv – is a self-diffusion coefficient of UO2.
As one can see from the Figure 13, slight variations of the preexponential factor and
activation energy Em leads to the significant change of the calculated FGR. Taking into account
possible dependence of self-diffusion coefficient on the local burnup [8)], this strong sensitivity
of the result to self-diffusion coefficient is not very pleasant.
So one of the possible causes of over-prediction of FGR for AREVA HB is an
overestimation of the intragranular bubble mobility at high burnups.
25
0 20 40 60 80 100Burnup, MWd/kgU
0
4
8
12
16
20
FG
R,%
PressureSTART-3
D0=D0/2Em=Em*1.05
Figure 13 - Slight variations of the preexponential factor and activation energy Em leads
to the significant change of the calculated FGR.
Single atom volume in the gas cluster
Another possible cause of overprediction for AREVA HB case is an underestimation of
HBS ability to hold FGP in START-3 model, this ability is proportional to the probability for gas
atoms to get into the closed pore. This probability increases with the pore size and the pore
number, which results in the increase of the local fuel swelling.
The local fuel swelling due to the fission products in START-3 is described as:
ragfabricatedasFFfuel nBPPV
Vint00.
0
)( Ε++−≈Ε=∆− ν
where the first term corresponds to the swelling due to grain boundary pores; the second term
represents the gas clusters on the boundaries; the third one is the intra-granular swelling. The
intra-granular swelling is
26
∑∑+
+++=ΕbN
sii
s
iigsolidra BrBNCvb1
3
11int 3
4)(
πα .
The first term describes the swelling due the solid fission products (αsolid= 0,32 vol.% per
1% h.a. of burn up b). The second term represents the monoatoms and intragranular gas clusters,
treated as the solid spheres (Ni – number of gas atoms; νg - single gas atom volume). The third
term accounts the most intragranular gas bubbles).
This single gas atom volume is thought to be a constant in the present version of
START-3, but the recent researches have shown that it can vary in relatively broad range at least
of 30-85 Å [9), 10)]. Several calculations with different values of this parameter were performed.
Figure 14 shows the impact of different constant values of Vg.
Fig. 15 - Impact of different constant values of Vg.
As expected, fission gas is lower for higher values of Vg, But, since this value
significantly affect not only fission gas release, but the dimensional changes of the pellet also,
that are quite well verified for START-3 for lower burnups, we can not simply set this value as
high as we need (staying in the range of the experimental data).
0 20 4 0 6 0 8 0 100 Burnup, MWd/kgU
0
4
8
12
16
20
F GR, %
FGRS TART- 3: VG AT=4 5V g = 100
Figure 14 - Impact of different constant values of Vg.
27
Recalculated cases
After all we have tried to implement simple local burnup dependence into our code to
keep the lower Vg for low burnup and high for high burnup. This act can be justified by the
degradation of the elastic properties of fuel pellet with burnup [11)], but further investigations
are needed. Still, the calculation results coincide with the experimental data much better for all
considered cases. The recalculated results are given bellow.
Figure 15 shows the FGR calculations for AREVA HB case with improved START-3
version. New results are situated much closer to the experimental points than the results of the
standard version, but the prediction still is a little bit conservative.
0 20 40 60 80Burnup, MWd/kgU
0
5
10
15
20
FG
R(%
)
START-3 AREVA HB case
AREVA experimental results
Figure 15 - START-3 calculations for AREVA HB case.
28
Figure 16 - START-3 calculation for FUMEX-III PK2 rods.
Fission gas release for all for the considered SUPER-RAMP rodlets is demonstrated on
the Figure 16 and in the Table 12. The coincidence is reasonable for all rodlets except PK2-s
(relative error 136%). The minimum relative error is for PK2-3 rod (-2%)
Table 12 – Fission gas release after ramp for PK2 rodlets. Calculated and measured values.
Rodlet Calculated FRG,% Experiment Relative error,%
PK2-1 23.6 28 -15
PK2-2 36.4 32 13
PK2-3 44 44.9 -2
PK2-4 10.2 9.5 7.5
PK2-s 24.6 10.4 136
and Figure 17 list diameter values after base irradiation and after ramp test.
29
Table 13 and Figure 17 list diameter values after base irradiation and after ramp test.
30
Table 13 - Diameter change calculated by Standard and improved version of START-3 code for
SUPERRAMP PK2 rodlets after base irradiation and ramp, mkm.
Rodlet
Section 1 Section 2 Section 3 Experiment (average)
Diameter
change
during base
irradiation
Diameter
change
during ramp
Diameter
change
during base
irradiation
Diameter
change
during ramp
Diameter
change
during base
irradiation
Diameter
change
during ramp
Diameter
change
during base
irradiation
Diameter
change
during
ramp
prev now prev now prev now prev now prev now prev now
PK2-1 -11 -9 56 97.5 -11 -9 78 152 -11 -9 35 69 -75 139
PK2-2 -9 -9 83 153 -9 -9 105 188 -9 -9 58 114
PK2-3 -12 -13 95 176 -12 -13 124 213 -12 -13 66 147 -85 193
PK2-4 -28 -34 41 51 -28 -34 62 90 -28 -34 27 31
PK2-s -13 -22 64 109 -13 -21 87 165 -13 -21 48 81 -90 111
Figure 17 - Calculated diameters change after ramp, mkm
According to STSR-32 report, the measured elongation for PK2-1 during the ramp equals
approximately 400 mkm. Corrected version of START-3 predicts of 429 mkm. The residual
elongation for PK2-1 according to STSR-32 is 250 mkm and the corrected version of START-3
predicts is 324 mkm.
31
Simplified cases from FUMEX-II
To verify the changes made into the code, the corrected version of the code was used to
recalculate the FUMEX-II simplified cases 27.2.
The FUMEX-II* coordinated research program (CRP) was initiated by the IAEA
following a recommendation of the IWGFPT. It was conducted over the period 2002–2006.
Seventeen countries took part.
The FUMEX-II program continued the work of the former CRP on "The development of
computer models for fuel element behavior in water reactors" (D-COM), which started in 1982
and was terminated in 1984, and the FUMEX CRP “Fuel Modelling at Extended Burnup” which
started in 1993 and concluded in 1996.
Following the lead taken in the original FUMEX CRP, a number of simplified cases were
constructed in order to investigate mathematical stability and more easily compare model and
code predictions without the vagaries of real power histories. In this section, each case is
outlined together with the reason for its inclusion before presenting the results and comparing the
predictions.
The second idealized case was to illustrate code predictions of FGR as a function of
burnup up to 100 MWd/kgU. There were four separate idealized cases for this task:
27(2a) a constant power of 15 kW/m from BOL to 100 MWd/kgU,
27(2b) a linearly decreasing power from 20 kW/m at BOL to 10 kW/m at 100 MWd/kgU,
27(2d) idealized ‘real’ history supplied by F Sontheimer of FANP.
*J. Killeen, Fuel Modelling at Extended Burnup (FUMEX-II)
IAEA report of a coordinated research project 2002–2007
32
0 20 40 60 80 100 120Burnup, MWd/kgU
0
10
20
30
40F
GR
(%)
START-3 improved versionSTART-3 standart version
Figure 18a. – Simplified case 27.2a. Calculated values of
fission gas release for an idealized irradiation at a constant
power of 15 kW/m to a burnup of 100 MWd/kgU.
0 20 40 60 80 100 120Burnup, MWd/kgU
0
10
20
30
40
FG
R(%
)
START-3 improved versionSTART-3 standart version
Figure 18b. – Simplified case 27(2b). Fission gas release
calculated for an idealized irradiation history where the
power reduces linearly from 20 kW/m to 10 kW/m to a
burnup of 100 MWd/kgU
Figure 18c. – FUMEX II simplified case 27.2d. Power
history for 5 cycles. New cycle at time slice 12, 22, 32, 42.
Red bottom at axial slice R1.
0 20 40 60 80Burnup, MWd/kgU
0
5
10
15
20
FG
R(%
)
START-3 improved version
Experimental results
START-3 standart version
Figure 18d. – START-3 calculations of FUMEX II
simplified case 27.2d.
Figure 18 – FUMEX II calculation results.
Figure 18 presents the calculation results of Simplified cases 27(2a), 27(2b), 27(2d). As
one can see from the presented data, improved version of START-3 works fine up to burnups of
100 MWd/kgU and calculations are in reasonable agreement with experimental data for
Simplified case 27(2d) from FUMEX-II.
33
Summary and conclusion
FUMEX-III SUPER-RAMP exercise included up to 49 kW/m power ramps of fuel rods
with burnup up to 45 MWd/kgU. START-3 calculations of this exercise are in reasonable
agreement with the experiment.
AREVA high burnup priority case (~81.5 MWd/kgU) was calculated and overprediction
of the FGR was found. This overprediction is related with the conservative overestimation of
rim-structure capability to retain fission gas and can be adjusted implementing a simple
correction of this property.
To verify the changes made into the code (oxide layer growth model and Zry-4 radiation
growth), the corrected version of the code was used to recalculate the FUMEX-II simplified
cases 27.2 and satisfactory results were obtained.
FUMEX-III lessons
FUMEX-III activity was very useful for us and has shown that:
• New material properties (Zry-4) coupled with START-3 work fine, but further
calculation experience is required.
• FGR model of START-3 has strong sensitivity on the fuel microstructural parameters
(UO2 self-diffusion coefficient and single gas atom volume).
• START-3 overpredicts FGR at low LHRs + high burnups. Possible reasons and measures
to improve this situation were suggested and tested. The improved version of the code
adequately predicts the fission gas release and dimensional changes of the fuel rods for
all considered cases.
At the current state START-3 can adequately predict PWR fuel rods behavior
during ramps up to ~50 kW/m and burnups up to ~82MWd/kgU.
We would like to thank the participants of the FUMEX-III for sharing their opinions and
results, IAEA for organizing FUMEX-III, and especially we would like to thank Dr. John
Killeen for his assistance and patience. FUMEX-III was extremely useful for START-3 team
and we would like to ask IAEA to organize a new project of this kind (FUMEX-IV for
example).
34
Brief START-3 code description
START-3 code designation
START-3 code is developed at JSC VNIINM and designed for performing strength and
thermal calculations aimed at the study, justification and licensing of fuel elements for nuclear
power reactors on thermal neutrons operated under normal conditions and during anticipated
operational occurrences.
START-3 code is used to perform the computational justification of fuel rod operability
by the below criteria:
• SC1 – Limit stress in the fuel rod cladding;
• SC5 – Limit residual strain of the fuel rod cladding;
• DC1 – Limit value of the fuel rod cladding diameter change;
• DC2 – Limit value of the fuel rod elongation;
• TC1 – Limit fuel temperature;
• TC2 – Limit value of the inner gas pressure.
START-3 code is validated for WWER, PWR and BWR fuel rods up to 78 MWd/kgU,
and the calculations in this report allow us to extend the verification base up to 82 MWd/kgU.
General description of START-3 code
The mathematical formulation of the mechanical state problem means a joint use of
equations of state, conditions of strain compatibility, equations of radial and axial equilibrium
and boundary and initial conditions. The above relations make a closed system for the
unambiguous solution of the problem. The problem is solved by parameter increments according
to the finite-difference scheme.
Temperature fields in the fuel are determined based on the solution of a non-stationary
heat equation for a cylinder with internal heat sources assuming there is no axial heat transfer
(such approximation is quite permissible as the radial temperature gradient significantly exceeds
the axial one).
Fuel-Cladding heat conductance coefficient is calculated using the Ross-Stout gap
conductivity model and is supposed to comprise three components:
• Gas conductivity in the gap;
• Contact component;
• Radiation
35
To take into consideration the irradiation induced densification of the fuel the empiric
relation obtained on the experimental data massive of in-pile tests and PIE performed on
industrially-fabricated fuel pellets is used.
The computational model of FGR from the fuel to the free volume of the fuel rod during
stationary modes and transients includes:
• description of athermic mechanisms, to which FGR refer due to the «direct output» and
«knockout» and the peculiarities of fission gas behavior in the rim-layer;
• description of thermal release mechanisms conditioned, first of all, by the direct diffusion
of fission gas mono-atoms and gas bubbles at grain boundaries and by further diffusion
percolation of the fission gas through the stochastically opening system of tunnels on the
grain boundaries.
START-3 code permits to evaluate the impact of processes related to the rim-effect on
the fuel rod behavior. These include:
• edge effects of Pu accumulation and increase of volume power rate and burnup in a
particular area of the fuel pellet
• athermic and thermal gas release;
• swelling due to solid fission products;
• swelling due to fission gas, which, besides the thermo-mechanics, conditions the fuel-
clad interaction under the modes involving fast power rise;
• growth and distribution of minor pores in the rim-layer under irradiation;
• growth of open porosity;
• generation of minor sub-grains (< 1 µm) on the pellet periphery due to the accumulation
of radiation damages of the crystal structure under irradiation;
• reduction of the fuel heat conductivity related to the reduction of its density and burnup
growth;
• positive feedback between the gas release and fuel temperature due to the dilution of
inner gas with fission gas.
START-3 code simulates the generation and healing of cracks in a fuel pellet. The
increase of pellet strain due to cracking leads, first of all, to the change of fuel-clad heat transfer
conditions because of fuel pellet relocation and, in case of contact, to increased load on the fuel
rod cladding after the crack healing.
The cladding oxidation is simulated within the mechanical (fuel rod clad thinning) and
thermal blocks of the code (growth of the thermal resistance of the clad-oxide film system).
36
The following properties of the Zry-4 cladding were added to the START-3 properties
libraries in the framework of FUMEX-III project.
Radiation Growth
In order to provide better applicability of START-3 code to PWR calculations, some
efforts were taken recently. One of the main differences between E110 and Zry-4 is a
significantly greater radiation growth. So we have decided to implement a new radiation growth
model derived from the available data on radiation growth (SKI Report 2005:41 [1)]).
Figure 19 - Axial growth data for PWR fuel rods with M5 cladding, in comparison with
the FRAPCON-3.2 models for stress relieved annealed (SRA) and recrystallized (RXA) clad
materials.
Oxide layer growth model
The second improvement of our code is introduction of oxide layer growth model for
Zry-4 from the same FRAPCON SKI Report 2005:41 [1)]. Under PWR conditions, the clad
oxide layer thickness oxδ [m] is assumed to grow according to
37
1
2
2
/12
/0.242 3 0
/0.242 3 0
, 2
( ( / ) ) , 2 35
1.8( ( / ) ) , 35
co
co
co
Q RTox
ox
Q Roxox
Q RTox
Ce m
dC C e m
dtC C e
δ µδ
δ φ φ δ µφ φ δ
−
−
−
<= + ≤ ≤ + <
where Tco [K] is the clad metal-to-oxide interface temperature, φ [neutrons m-2 s-1] is the
fast neutron flux (E > 1MeV), R is the universal gas constant and 0φ , C1, C2, C3, Q1 and Q2 are
empirical and constant model parameters, as defined in Table 14.
Table 14 - Model parameters of oxide layer growth, implemented into START-3 code.
Model parameter Value
C1, [m3s-1] 7.29 x 10-14
C2, [m s-1] 9.31 x 10-4
C3, [m s-1] 2.75 x 10-3
0φ , [neutrons m-2 s-1] 5.24 x 1018
R, [J mole-1 K-1] 8.3143
Q1, [J mole-1] 135 188
Q2, [J mole-1] 114 526
38
Literature
1) Lars O. Jernkvist Ali R. Massih, “Models for Fuel Rod Behaviour at High Burnup”,
December 2004
2) G. Khvostov, V. Novikov, A. Medvedev, S. Bogatyr “Approaches to Modeling of High
Burn-up Structure and Analysis of its Effects on the Behaviour of Light Water Reactor
Fuels in the START-3 Fuel Performance Code” 2005 LWR Fuel Performance
Meeting,Kyoto, Japan, October 2-6, 2000
3) Zacharie et al., “Thermal treatment of uranium oxide irradiated in pressurized water
reactor: swelling and release of fission gas”, JNM 255(1998) 85-91
4) J.A. Turnbull, E. Kolsdat, “Investigation of radioactive and stable fission gas release
behavior at the HALDEN reactor”, International Seminar, Cadarache, France, 26-29
September, 2000.
5) Paul Blair, “Modelling of fission gas behaviour in high burnup nuclear fuel”, école
polytechnique fédérale de Lausanne pour l’obtention du grade de docteur ès sciences.
6) E.Ya. Mikhlin, “The mobility of intragranular gas bubbles in uranium dioxide”, Journal
of Nuclear Materials, Volume 87, Issues 2–3, 1 December 1979, Pages 405–408
7) J.H. Evans, “The role of directed bubble diffusion to grain boundaries in post-irradiation
fission gas release from UO2: a quantitative assessment”, Journal of Nuclear Materials
238 (1996) 175-182
8) K. Govers, C.L. Bishop b, D.C. Parfitt b, S.E. Lemehov a, M. Verwerft a, R.W. Grimes b,
“Molecular dynamics study of Xe bubble re-solution in UO2”, Journal of Nuclear
Materials 420 (2012) 282–290
9) D.R. Olander, D. Wongsawaeng, “Re-solution of fission gas – A review: Part I.
Intragranular bubbles”, Journal of Nuclear Materials 354 (2006) 94–109
10) K. Govers, S. Lemehov, M. Verwerft, “In-pile Xe diffusion coefficient in UO2
determined from the modeling of intragranular bubble growth and destruction under
irradiation”, Journal of Nuclear Materials 374 (2008) 461–472
11) D. Laux a, D. Baron b, G. Despaux a, A.I. Kellerbauer c, M. Kinoshita, “Determination
of high burn-up nuclear fuel elastic properties with acoustic microscopy”, Journal of
Nuclear Materials 420 (2012) 94–100
