August 2003:
1. A uniform and constant core flow of 82.1212 kg/sec/assembly is assumed.
2. The assembly-wise kinetic parameters (neutron velocity, delay neutron
fractions and decay constants) are provided in the cross section data files.
3. The pin power is requested for 6 assemblies along the diagonal from
location (A, 1) to (F, 6) in the quarter core where the rod is withdrawn.
4. The KAERI integral transport code DeCART will be used as a reference.
DeCART is capable of whole-core transport (MOC) transient calculations with
explicit representation of fuel/clad/moderator.
5. Note new beginning of transient conditions: HZP, All Control Rods In,
Shutdown Banks SA, SB, SC Out, withdrawal of control rod (E, 5).
6. The same fission spectrum for prompt and delay neutrons should be used.
7. Moderator density extrapolation might be necessary for HFP calculation.
Describe if and what kind of extrapolation was used in assumption part of
Appendix D.
8. The reflector cross section has been calculated from a two-node
fuel/reflector problem with UO2 4.5% assembly at 20.0 GWd/t burnup, HZP
conditions (mod. dens. 752.06 kg/m3, fuel temp. 560 K). Only boron density
feedback has been calculated for the reflector. The reflector ADF has been
calculated from the analytic solution of diffusion equation in reflector
region with homogeneous group constants and reference current boundary
conditions.
9. In order to simplify the problem and provide consistency between cross
section library and number densities, all cross sections and number densities
contain equilibrium Xe/Sm. Fresh assembly is at 0.15 GWd/t, which corresponds
to about 3 days burnup at HFP.
10. "Sb-nat (FP)" isotope in number densities library is a nautral mixture of
Sb isotopes as a fission product.
11. The participants should use their own methods of cross section
interpolation (3D interpolation for T/H state + 1D for CR).
12. The group constants without feedback (fission spectrum, velocity, delay
neutron decay constant and delay neutron fraction) correspond to assembly at
mod. dens. 752.06 kg/m3, fuel temp. 560 K, 1000.0 ppmB, CR out state.
October 2003:
1. The pin-power form functions correspond to assembly at mod. dens.
752.06 kg/m3, fuel temp. 560 K, 1000.0 ppmB, CR out state.
2. List of benchmark participant has been added to the website. If you do
not wish to be listed, please let me know.
3. Benchmark deadline has been extended at least until the next WPPR meeting.
Final deadline will be decided during the meeting (Feb. 26, 2004 in Paris).
4. The 1/4 core loading has reflective symmetry. Full core model is
necessary at least for Part 1a and 1d (single rod worth calculation) and the
Part 4 (transient withdrawal of a single rod).
5. The correct active fuel length is 365.76 cm.
6. The axial mesh is up to participants. Enough axial nodes should be used
to achieve spatial convergence.
7. The compositions (burnup distribution) are uniform in the axial direction.
8. The control rod pattern for Part 3 (HZP state at the beginning of the
transient) corresponds to all control rods fully in, shutdown rod SD fully
in, shutdown rods SA, SB, SC fully out.
9. Fully inserted rod corresponds to the whole active fuel length.
10. There is no direct moderator heating in this benchmark. I run parametric
with 2% and 0% direct moderator heating and the results are almost identical.
The results are slightly different at HFP, but not enough to justify increased
complexity of the problem.
11. The temperatures in Table 2 are idealized core average quantities, not
inlet quantities. PARCS 2G nodal diffusion results:
HZP core avg. mod. temp. = 560K
HZP core avg. fuel temp. = 560K
HFP core avg. mod. temp. = 582K
HFP core avg. fuel temp. = 903K
12. The core inlet moderator temperature is uniform 560K for both HZP and HFP
conditions. Inlet pressure is uniform 15.5 MPa for both HZP and HFP.
13. Table 6 units are in centimeters.
14. The Figure 1 fresh assemblies should be at 0.15 GWD/MTU. The figure has
been corrected in October 2003 revision of the specifications. Loading
pattern and burnup of other assemblies has not been changed. Other minor
typos not affecting the specifications content have been corrected as well.
15. The rod worth calculation for Part 1b) and Part 1d) should be preformed
only for a single rod insertion/withdrawal, not for symmetrical position in
each quarter of the core.
16. The assembly is 1/8 symmetric, so there is just one ADF for all four
surfaces. The same is true for corner and midpoint discontinuity factors that
are part of the pin-power form functions.
17. All radial and axial powers should be normalized to average value of 1.0.
If they are not normalized properly, I'll normalize it myself, so that's not
a big problem.
18. We will post our PARCS and DeCART results on the benchmark website as they
become available. With permission, we would like to post other results as well.
December 2003:
1. Fully inserted and fully withdrawn control rod positions correspond to the
bottom and top boundaries of the active core, respectively.
2. Radial assembly peaking (fxy) is the maximum value from the normalized
radial assembly power distribution at each time step.
3. Radial pin peaking (fH) is the maximum value from the normalized core-wide
radial pin power distribution at each time step. Normally this requires pin
power reconstruction for the whole core, but for the transient we can safely
assume that it will be in the location of the ejected rod or one of the 8
surrounding assemblies. If using nodal method, please explain in the
assumptions where pin power reconstruction was used.
4. Point-wise pin peaking (fQ) is:
fQ = maximum(normalized core-wide radial pin power * normalized axial pin power).
5. Pin power form functions are (pin flux) * (pin kf). They were generated
for controlled and uncontrolled state at HZP T/H conditions. Per participants'
request, they were not normalized to preserve power ratio between groups.
December 2003 changes to the specifications:
6. Control rod material was replaced with B4C. All cross section libraries
changed, the format is exactly the same as before. Only number densities
for the control rod changed in the number densities files.
7. Because of different CR material, the beginning of transient control rod
position changed to All Control Banks In, All Shutdown Banks Out.
8. Axial reflector with zero flux boundary conditions has been added to
Part 2, 3 and 4. Part 1 of the problem is 2D with no axial reflector and
reflective axial boundary conditions.
9. There are two new edit requests: 'radially averaged axial power' for
Part 3 and 'control rod reactivity' for Part 4.
10. The deadline has been extended to June 1, 2004. Preliminary results are
requested for the next WPPR meeting (Feb. 26-27, 2004 in Paris).
May 2004:
1. Because of cross section library change, the deadline has been extended
until the next WPPR16/TFRPD9 meeting on September 13/14, 2004 in conjunction
with ARWIF-2004 conference.
2. For participants using specified 2G or 4G nodal library, the same reflector
XS data should be used for axial and radial reflector. Also, using provided
library assumes that there is no CR movement through the reflector.
3. Participants are free to use any T/H they feel appropriate, including coupled
codes, as long as it is consistent with the benchmark description. This is very
fast transient and we don't believe T/H solution should significantly affect
transient solution.
4. Eq. 5 on page 11 of the benchmark specifications is assumed to apply to both
UO2 and MOX fuel.
5. The assembly burnup is assumed to be uniform. Pin burnup variation within
assembly is ignored, and all pins are at the same burnup with number densities
corresponding to the ones provided in the benchmark.
6. For rod worth calculations in Part 1, a single rod should be
withdrawn/inserted from full core configuration.
July 2004:
1. I was asked to add diagonal of the scattering matrix to make it
possible to calculate total xsec for the transport codes. I added this
data to the xsec library, but the format had to change slightly, now there
is G -> G scattering term included. Note that the diagonal is total
within group scattering, not transport corrected scattering.
2. I posted new 2G nodal, 4G nodal with and without upscattering, 8G nodal
and 8G cell cross section library. May 2004 and July 2004 2G and 4G xsec
library are identical except for the diagonal of the scattering matrix.
3. I posted 8G cell cross section library per participants' request.
Note that the 8G cell library is not constant with the nodal library. This
is because cell homogenization error is not treated in any way (SPH/discontinuity
factors). Since methods used to deal with pin-cell homogenization error are
unique to each code, it is not possible for me to prepare a library that will
work for everyone.
4. Participants that want to generate their own group constants should use
number densities posted on the website.
5. The order of data in the cross section library is shown in Appendix C of
the specifications. 'D' is mod. density, 'B' is boron ppm, 'F' is fuel temp.
The order of each state variable is such that:
D1 = 661.14, D2 = 711.87, D3 = 752.06
B1 = 0.00, B2 = 1000.00, B3 = 2000.00
F1 = 560.00, F2 = 900.00, F3 = 1320.00
For example, the 2G library MOX 4.0% (2G_XSEC_m40_unrodded) transport XSEC
for fast group at 0.15 GWd/t is:
Tf = 560.0
rho=661.14 rho=711.87 rho=752.06
--------------------------------------------+
2.28636E-01 2.36580E-01 2.42881E-01 | ppm=0.0
2.28152E-01 2.35981E-01 2.42172E-01 | ppm=1000.0
2.27671E-01 2.35403E-01 2.41516E-01 | ppm=2000.0
Tf = 900.0
rho=661.14 rho=711.87 rho=752.06
--------------------------------------------+
2.28447E-01 2.36366E-01 2.42652E-01 | ppm=0.0
2.27958E-01 2.35776E-01 2.41960E-01 | ppm=1000.0
2.27473E-01 2.35195E-01 2.41294E-01 | ppm=2000.0
Tf = 1320.0
rho=661.14 rho=711.87 rho=752.06
--------------------------------------------+
2.28262E-01 2.36174E-01 2.42447E-01 | ppm=0.0
2.27767E-01 2.35581E-01 2.41756E-01 | ppm=1000.0
2.27283E-01 2.34998E-01 2.41086E-01 | ppm=2000.0
6. Please provide preliminary results before WPRS01 (former WPPR16)
meeting (Sept. 13/14, 2004) for me to include in the presentation.
Please finalize all results before end of 2004.
7. Any outstanding problems should be resolved before WPRS01 meeting.
November 2004:
1. Please submit final results by the end of the December 2004.
2. I posted DeCART solution for Part 1 of the benchmark. It is 47G, MOC,
heterogeneous solution and it will be used as a reference. I would like
to thank Seoul National University and KAERI who helped to prepare it.
3. I would like to have Monte Carlo solution for Part 1 of the benchmark
as well. Would someone be willing to prepare that? Even partial data
for ARO and ARI state would be very helpful.
4. To generate own cross-sections, one should use data from Appendix B.
Complete number density data are posted on the benchmark website. Even
though it is possible to start with data from Table 3/4/5, this is not
objective of the benchmark and could be made into a separate benchmark of
its own. Objective is not to deplete the core correctly but rather to
predict core state correctly assuming that compositions (number densities)
are known exactly. That's why all number densities are given for all
fuel/WABA/IFBA at every burnup used in the core. Without this information
the benchmark becomes much more difficult.
May 2006:
1. Fresh fuel is really 0.15 GWd/tHM - first burnup step in HELIOS that
is supposed to simulate FP buildup. Once burned assemblies are at 17.5,
20.0 or 22.5 GWd/tHM, twice burned assemblies are at 32.5, 35.0 or
37.5 GWd/tHM. This is because it is almost impossible to get a core with
reasonable peaking using only three burnups (0, 20, 35).
2. The kinf in the number density spreadsheets correspond to average burnup
conditions (HFP average conditions): Tf = 900K, TM = 580K, rhoM = 711.87 kg/m3,
boron = 1000 ppm, CR out.
---
tomasz@ecn.purdue.edu
tomasz@safety.sci.kth.se