Skip to main content

SCALE v6.2.3

This release of SCALE is now unsupported. Please click here for the current production release.


April 2018

The SCALE 6.2.3 update is available for SCALE 6.2 users as of April 2018, providing enhanced features and performance in the areas detailed below. This update is provided as a download and is recommended for all SCALE 6.2.0, 6.2.1, and 6.2.2 users. The 6.2.3 update includes all previous updates and can be applied directly on any SCALE 6.2 release.

Polaris Enhancements

SCALE 6.2.3 introduces several enhancements to the Polaris lattice physics code.

  • Users will notice the ~2× speedup in run time for depletion calculations.
  • A new restart feature reuses the geometry data structures and flux solution from previous calculations.
  • Polaris has a new detector edit capability in SCALE 6.2.3. Input examples of the detector card for a PWR 17 × 17 lattice model and for a BWR 7 × 7 lattice model are shown in Figure 2, and an example of detector response output is shown in Figure 3. The key inputs and their descriptions are given here.
    • Detector material (Lines 24, 25). In this input, a trace amount of 235U is used to define material DET.1.
    • Detector geometry (Line 26). The detector geometry is provided through the Polaris pin card. In this input, the detector geometry is a simple zone of coolant with pin ID "D."
    • Detector definition (Line 27). In this input, the detector "d_235" is defined as pin D inserted into pin IT. The detector response is the neutron fission rate "E(n, FIS)." The detector cross section is from DET.1 (i.e., 235U), and the detector flux is the COOL.2 flux inside the detector geometry.
    • Opt FG input (Line 48). The detector option on the opt FG card designates the detector edit to be included on the t16 file. 
Example detector card for a Westinghouse 17 x 17 model and a BWR 7 x 7 model
Example detector card for a Westinghouse 17 x 17 model (left) and a BWR 7 x 7 model (right).
Example of detector response output
Example of detector response output.


Self-shielding method selection logic: SCALE 6.2.3 includes improved consistency for self-shielding logic across all sequences and now enables the user to use the BONAMI-only methodology for faster calculations or in cases where the CENTRM methodology is not suitable. In SCALE 6.2.3, the logic has been streamlined as follows:

  • 1. CENTRM for double-heterogeneous cells,
  • 2. BONAMI for infinite homogeneous (inf. hom.) cells without fissionable nuclides,
  • 3. For inf. hom. cells with fissionable nuclides, choose
    • method specified by user in parm data (CENTRM, BONAMI, XSLEVEL, 2REGION) or
    • sequence defaults
      • CENTRM for t-newt, t-xsdrn, t-depl-1d, t-depl, csas1x, csas5, csas6, t keno, t-depl-3d, tsunami-1d, tsunami-3d
      • BONAMI for Mavric, XSProc sequence  

Numerical stability in CENTRM MoC solver: An issue was identified with the default option for lattice cell self-shielding in cases with very small macroscopic cross section < 10-7 cm-1. During the
calculations, CENTRM produced nonphysical fluxes, shown as not a number, or NaN. Analyses of typical LWR configurations are not expected to be affected. The issue was resolved in SCALE 6.2.3 by introducing a more robust way to handle very small macroscopic cross sections, with negligible impact on memory and runtime.

Coupled neutron/gamma: An issue was discovered in CENTRM in which enabling upscattering in a coupled neutron/gamma problem produced erroneous results. This was recognized due to no gammas leaving the problem domain. As a temporary stop-gap for SCALE 6.2.3, upscattering will not be allowed in coupled neutron/gamma calculations.

Minor miscellaneous issues resolved in XSPROC 
In the output of XSProc with lattice cell self-shielding, the printout was corrected where a radius quantity was incorrectly labeled as a diameter.

The XSDRN Balance table file (ft76.btf) was missing its lambda value. The lambda value is now included in the balance table file.

The XSPROC section of the manual has been updated to include a description of the double-heterogeneous self-shielding treatment (e.g., for TRISO particle models) for SLAB geometry. Previously, only a description of the capability for SPHERE geometry was present.

There has been some confusion over the filenames used by the various SCALE sequences/modules that write a cross section library to disk. In SCALE 6.2.3, the various filenames used by each sequence/module have been clearly documented in the SCALE manual. For SCALE 6.3, a more consistent naming scheme will be implemented, perhaps with the additional capability for the user to choose the name.


KENO-V.a boundary condition: An issue was identified in SCALE 6.1–6.2.2. An unexpected behavior can occur when a user generates an input that is inconsistent with the documentation and training materials and the code does not detect the input error before executing. SCALE behaves as intended if users generate models consistent with the documentation and training materials. Details and corrective actions are presented
in the section below entitled "SCALE 6.1–6.2.2 KENO-V.a boundary condition issue.”

Reaction cross sections: In continuous energy (CE) KENO, the method for calculating average reaction cross sections, which are of particular importance in CE TRITON depletion calculations, has been modified
to increase robustness. Previously, multigroup reaction cross sections were calculated at each active generation and were averaged over all active generations to arrive at the best estimate. For SCALE 6.2.3, the methodology has changed so that the multigroup reaction rates and fluxes (instead of reaction cross sections) are accumulated over active generations and after all active generations are complete, the reaction cross sections are calculated as the reaction rate divided by flux. An additional positive effect of the change is that the options for tallying reaction cross sections, cxm=2 (multigroup xs) and cxm=4 (1-group xs), generally show better agreement.

Reaction rate output: An output processing issue was discovered in the KENO reaction rate output: absorption (MT=27) and capture (MT=101) cross sections were not being updated when constituent cross sections for scattering (MT=2), fission (MT=18), inelastic scattering to the first excited state (MT=51), and n,gamma (MT=102) cross sections are sampled from probability tables in the unresolved resonance region (URR). This does not impact the transport calculation or depletion calculations under t5-depl or t6-depl that use the constituent cross sections directly. This issue was discovered by a user attempting to calculate keff from reaction rate output and match the result to the output eigenvalue. Analysis using the reaction rate output where the URR range is important should be repeated with SCALE 6.2.3.

Minor miscellaneous issues resolved in KENO
The output for the Shannon entropy convergence test has been updated in SCALE 6.2.3 to emphasize where active generations are used.

The nuclide identifiers for metastables and bound nuclides were not correctly displayed in the KENO output. These isotope edits are now correct.

Filenames for reaction rates have been tallied. Instead of ${BASENAME}.keno_micro_rr.* in SCALE 6.2.2, KENO in SCALE 6.2.3 writes reaction rate tallies in ${BASENAME}.keno_rr.*.

The Doppler broadening of CE cross sections to user-specified temperatures in KENO now allows for four options for the DBX parameter:

  • 0 = No problem-dependent or on-the-fly Doppler broadening
  • 1 = Perform problem-dependent Doppler Broadening for 1D cross sections only
  • 2 = Perform problem-dependent Doppler Broadening for both 1D and 2D (thermal scattering data) cross sections
  • 3 = Broaden 2D cross sections normally, and broaden 1D cross sections using a less robust but faster interpolation method.

DBX=3 was previously only supported as an option in Monaco.

Occasionally, the Doppler broadening of very small cross sections can result in very small negative values for a few data points. The effect of these values on criticality and reaction rates is undetectable, but these very small negative values could cause CE TSUNAMI sensitivity calculations to fail. In SCALE 6.2.3 these values are set to zero.


Undefined mixtures: For cases that have undefined mixtures referenced in the geometry, NEWT calculations assumed that any mixture that did not include any data in the cross section library was a void material with a cross section of zero. This issue has been fixed in SCALE 6.2.3, and additional checks and error messages have been added to verify that every mixture referenced in the geometry is defined in the composition block. This issue affects all uses of NEWT, including the t-newt and t-depl sequences.


Neutron emission calculations: An issue was found in ORIGEN in which a tolerance for including a nuclide in the emission calculation did not include the initial isotopics. Thus, any nuclide that fell below the threshold over the first timestep would not be included in the neutron emission calculation for the entire case. For common timestep sizes on the order of days, the main effect was that the delayed neutron emitters, which have effective time ranges of much less than a day, are bypassed, and no delayed neutron information is shown. The issue was fixed in SCALE 6.2.3 by including the initial isotopics in the time-averaging to determine the active nuclides for an emission calculation. A workaround for SCALE 6.2.0, 6.2.1, and 6.2.2 is to add a very small, initial timestep (on the order of milliseconds).

Volume input: ORIGEN volume input is only used when the user specifies the isotopic input by number density, and the volume is needed to convert the number density to ORIGEN's internal mole units. However, unlike other SCALE sequences, number density input is not commonly used in ORIGEN. Volume input was not being correctly handled in SCALE 6.2.2. Any calculations using the number density input option should be rerun using SCALE 6.2.3.

Elemental input: A minor change was made regarding how ORIGEN handles elemental input. It is now flagged as an error which specifies any element for which the natural abundances sum to zero (e.g., Tc) according to the data on the specified ORIGEN library. Note that the current ORIGEN library format only allows abundances for light nuclides (hydrogen to lead), so it does not include natural uranium.

ARPLIB utility module: This module, which is used to manipulate the number of burnup-dependent data sets on an ORIGEN library, was not properly processing libraries in the SCALE 6.2 format. In SCALE 6.2.3, ARPLIB properly reads the SCALE 6.2 library format (and SCALE 6.1 format) and has been redesigned to allow the user to easily create an arbitrary library from cross sections using data from a number of other available libraries. Note that this utility module is rarely used for thinning or combining existing libraries. The module does not intervene in any ORIGEN calculation or SCALE sequences involving ORIGEN.


Significant Polaris enhancements are discussed in this newsletter under “Polaris Enhancements.” A new preview capability is available for basic detector modeling and gamma transport. This is fully described in the appendix of the Polaris manual in SCALE 6.2.3. As with earlier previews, =polaris_6.3 must be used as the sequence name. Users may experience a runtime reduction of 10–50% due to optimization of some internal initialization routines.

Minor issues resolved in Polaris
Mismatches between user-specified symmetry and actual problem symmetry are now better recognized. Defining compositions with names that are also element symbols (e.g., Zr or F) is no longer allowed. Rare convergence issues occurring with CMFD and buckling calculations are now fixed. In the Polaris nodal data file (T16 or X16 file): energy release per fission parameters now properly include energy released from capture, and few-group flux is normalized as per source neutron instead of per source neutron per unit volume.


Perturbed decay data: The uncertainty in the decay energy release (aka Q-value) for primary alpha emitters was found to be at least 50% larger than the actual value due to an issue in the sampling code that is used to generate the 1,000 perturbed decay data files. The code error has been fixed. The updated perturbed decay data files are not available as part of the SCALE 6.2.3 release, but they will be made available as an additional download. This only affects Sampler uncertainty calculations with perturb_decay=”yes” (the default is “no”). Moreover, it only affects the decay heat uncertainty prediction; it does not affect the decay heat prediction itself. Validation of calculated decay heat against measured spent fuel decay heat has shown good agreement, typically within 2%, between predicted and measured values. SCALE developers are committed to making uncertainty calculations a simple, routine part of nuclear engineering analysis. Additional tests will be applied to the current data set, along with new uncertainty data from ENDF/B-VIII.

Minor issues resolved in Sampler
The output to additional auxiliary files (recommended for post-processing of the sampled results with a tool external to SCALE) only contained a single significant digit instead of the desired five significant digits. If the
input only used a single sample, then Sampler would show standard deviations of 0.0 (or for some outputs, NaN) instead of the more appropriate infinity.


Calculations using thorium: TRITON did not correctly include thorium (atomic number 90) in the definition of initial heavy metal (IHM) for the purpose of setting specific power levels (units of MW/MTIHM) and reporting burnup (MWd/MTIHM). The error has been corrected in SCALE 6.2.3, and a verification problem using 232Th was added to the regression suite. The error made it impossible to deplete a thorium-only mixture, and for a mixture that included thorium, the input power was misinterpreted, and burnup was misreported. Any depletion calculations for thorium in SCALE 6.2.0–6.2.2 are incorrect and should be rerun with 6.2.3.

Keep block: In TRITON, keeping the ORIGEN output (origen keyword found in the keep block) would lead to extremely large transition coefficient tables in the TRITON output file, and cases with many burnup points could have output file sizes in gigabytes. In SCALE 6.2.3, requesting TRITON to keep origen will only enable output of detailed depletion number density tables. To recover the previous output, which may be useful in small problems to understand specific depletion pathways, the user must indicate to keep both ORIGEN and COUPLE output in the keep block.

Self-shielding and material swap: In TRITON (MG) within SCALE 6.2, the interaction between self-shielding and the material swap (and control rod branch) has changed. In previous versions, the mixture swap occurred before self-shielding. This required users to take the swap into account when creating the celldata block. In SCALE 6.2 and later versions, self-shielding in the celldata block is based only on the defined mixtures. This difference only affects calculations in which latticecell or multiregion self-shielding is used with swaps or control rod branches. It does not affect the default inf. hom. In SCALE 6.2.3, the user is warned when these conditions are present.


Autocomplete: SCALE 6.2.3’s Fulcrum now has the ability to validate and autocomplete the MAVRIC input.

Visualization: Lines in OPUS plot files now display in different colors instead of only blue. Covariance data color bar limits and axis fonts can now be changed.

Input Editor: Input validation is now faster by a factor of two, is fully case insensitive, accounts for all possible SCALE sequences (including utility modules), and is no longer disrupted by uncommented text preceding
a sequence. New error messages are emitted if a user attempts to create a new file without the proper write permissions. If a user attempts to view a geometry for an input that has no geometry, then appropriate error messages are displayed. The cursor retains its original position when a text file is reloaded instead of moving to the top of the file.

Other Minor Miscellaneous Issues Resolved in SCALE 6.2.3

The legacy addnux value of -2, which was disabled in TRITON in SCALE 6.2.2, has been included again in SCALE 6.2.3.

In SCALE 6.2.3, various updates have been implemented in the STARBUCS burnup credit analysis sequence for consistency with updates in KENO and ARP.

ORIGAMI now allows the simulation of an initial decay interval, for example, to simulate some initial decay time before the first irradiation. This is accomplished by entering a power level of zero for the first cycle, which would previously cause the calculation to fail. Additionally, the format MCNP source output (enabled with mcnp=yes) has been updated for improved compatibility with MCNP.

In TSURFER, a minor issue was corrected in which the printout of cross section adjustment information (print_adjustments) did not work.

An issue was discovered with the SCALE composition block’s atom type input in which the atom fractions that were entered were being treated as weight fractions. This issue has been resolved in SCALE 6.2.3. Note that this only affects atom composition inputs (e.g., “atom 1 19.0 2 92235 0.1 92238 0.9 end”), which is a rarely used form.