3. ENRICO Runtime Settings

Parameters and settings for each individual physics code are set normally in their respective input files. Parameters related to the coupled simulation are listed in a special enrico.xml file. This file accepts the following elements:

3.1. <heat_fluids>

Base element for heat/fluids driver parameters

3.1.1. <driver>

The physics driver for solving fluid and heat transfer equations. Valid options are “nek5000”, “nekrs”, and “surrogate”.

3.1.2. <pressure_bc>

The pressure of the outlet boundary condition in units of [MPa].

3.1.3. Nek5000- and nekRS-specific Parameters

Under the <heat_fluids> element, these sub-elements are available Nek5000 and nekRS

  • <casename>: Required. The casename for the problem (i.e, the basename for .par or .rea files).
  • <output_heat_source>: Optional. Can be true or false (default false). If true, output the heat source to a Nek field file in units of [W/cm^3]. In an ENRICO simulation, the field file <casename>#.f##### always contains temperature. The heat source is additionally output as follows: - For Nek5000 runs, the heat source is output as the first passive scalar in <casename>#.f######. - For nekRS runs, the heat source is output as the temperature field in a second field file, qsc<casename>#.f#####.

3.1.4. Surrogate-specific Parameters

Under the <heat_fluids> element, these surrogate-specific sub-elements are available:

  • <clad_inner_radius>: The cladding inner radius in units of [cm].
  • <clad_outer_radius>: The cladding outer radius in units of [cm].
  • <pellet_radius>: The fuel pellet radius in units of [cm].
  • <fuel_rings>: The number of rings the fuel pellet should be subdivided into when solving the heat equation.
  • <clad_rings>: The number of rings in the cladding should be subdivided into when solving the heat equation.
  • <n_pins_x>: Number of pins in the assembly in the x-direction.
  • <n_pins_y>: Number of pins in the assembly in the y-direction.
  • <pin_pitch>: Pitch, or distance between centers along the x- and y-axes, between pins. The pitch must be greater than the outer diameter of the pins, which would correspond to touching pins. This pitch is used to determine the pin-pin spacing and the pin- to assembly-edge spacing, which is taken to be half a pitch.
  • <z>: Values along the z-axis that subdivide the fuel region in units of [cm].
  • <inlet_temperature>: Fluid inlet temperature in [K].
  • <mass_flowrate>: Fluid mass flowrate in [kg/s].
  • <max_subchannel_its> * Maximum number of iterations to perform in the solution of the subchannel equations. Convergence is based on the relative change measured in the 1-norm in enthalpy and pressure between two successive iterations. This defaults to 100.
  • <subchannel_tol_h>: Convergence tolerance to use for enthalpy between two successive iterations of the subchannel solver. This defaults to a value of 1e-2.
  • <subchannel_tol_p>: Convergence tolerance to use for pressure between two successive iterations of the subchannel solver. This defaults to a value of 1e-2.
  • <heat_tol>: Tolerance on the heat equation solver. This defaults to a value of 1e-4.
  • <verbosity>: Degree of output printing for diagnostic checking. This defaults to none, but may be set to low and high. Both low and high perform error checks such as ensuring conservation of mass and energy, while high prints some subchannel solution metrics for each channel.
  • <viz>: This element indicates visualization settings for the heat solver.
    • filename (attribute): File prefix for output VTK files
    • <iterations>: what iterations to write output at
    • <resolution>: resolution of the VTK objects. When fluid regions are included, the resolution must be divisible by the number of channels per rod (typically 4)
    • <data>: what data to write. Either “all”, “source”, “temperature”, or “density”.
    • <regions>: what regions to write output for. Either “all”, “solid”, or “fluid”.

3.2. <neutronics>

Base element for neutron transport driver parameters

3.2.1. <driver>

The physics driver for solving particle transport. Valid options are “openmc”, “shift”, and “surrogate”.

3.2.3. Shift-specific Parameters

Under the <neutronics> element, these Shift-specific sub-elements are available:

  • <filename>: Path to the Shift XML input file

3.2.4. Boron search-specific Parameters

Under the <neutronics> element, these boron search-specific sub-elements are available:

  • <initial_boron_ppm>: The first guess of the fluid’s boron concentration in units of ppm to use when performing the critical search. If not provided, this defaults to the average boron ppm across all fluid-bearing cells of the initial the neutron transport model.

  • <target_keff>: The k-eigenvalue to search for by varing the boron ppm. This defaults to a value of 1.0.

  • <target_keff_tolerance>: The tolerance on the k-eigenvalue, to a 95% confidence interval based on the stochastic variation of k-eff, that will be used to evaluate convergence of the boron search. This defaults to a value of 1.0e-3.

  • <B10_enrichment>: The enrichment of B-10 in the boron in terms of an atom fraction. This defaults to a value of 0.1982.

  • <boron_epsilon>: The target boron search convergence criterion. If \(ppm_i\) and \(ppm_{i+1}\) are the set of total boron number density concentrations (on a number density basis) at iterations \(i\) and \(i+1\), convergence is reached if both of the following conditions are achieved:

    \[\lvert ppm_{i+1} - ppm_i \rvert < \epsilon_{boron} \]
    \[\lvert k_{eff,i+1} - k_{eff,i} \rvert < \epsilon_{keff} \]

    This defaults to a value of 1.0e-3.

Note

In ENRICO, the boron parts-per-million (ppm) is defined as the ppm boron on a number-density basis.

Note

ENRICO assumes that the reactivity effect of the hydrogen and oxygen in boric acid is small and on the order of their number density variations in water. Therefore, the ENRICO’s boron search only modifies the boron concentrations and not the hydrogen and oxygen as the boron ppm varies.

3.3. <coupling>

Base node for coupling parameters

3.3.1. <verbose>

A boolean (“true” or “false”, default “false”). If true, print detailed info, including MPI communicator layouts for each rank and volume comparisons for each cell (summarized volume statistics are always printed).

3.3.2. <power>

The power of the reactor in units of [W].

3.3.3. <max_timesteps>

The maximum number of timesteps.

3.3.4. <max_picard_iter>

The maximum number of Picard iterations within a timestep.

3.3.5. <epsilon>

Convergence criterion, \(\epsilon\). If \(T_i\) and \(T_{i+1}\) are the set of temperatures at iterations \(i\) and \(i+1\), convergence is reached if

\[\lvert T_{i+1} - T_i \rvert < \epsilon \]

Default: 1.0e-3

3.3.6. <alpha>

Underrelaxation parameter used on a heat source update. Let \(q_i\) be the heat source at iteration \(i\) and \(\tilde{q}_{i+1}\) be the next estimate of the heat source as determined by the neutronics solver. Then, the heat source for iteration \(i + 1\) is:

\[q_{i+1} = (1 - \alpha) q_i + \alpha \tilde{q}_{i+1} \]

Choosing \(\alpha = 1\) corresponds to no underrelaxation. A special value of “robbins-monro” indicates that Robbins-Monro relaxation is to be used:

\[q_{i+1} = \frac{1}{i} q_i + \left (1 - \frac{1}{i} \right) \tilde{q}_{i+1} \]

Default: 1.0

3.3.7. <alpha_T>

Underrelaxation parameter used on a temperature update. Let \(T_i\) be the temperature at iteration \(i\) and \(\tilde{T}_{i+1}\) be the next estimate of the temperature as determined by the thermal-fluids solver. Then, the temperature for iteration \(i + 1\) is:

\[T_{i+1} = (1 - \alpha_T) T_i + \alpha_T \tilde{T}_{i+1} \]

Choosing \(\alpha_T = 1\) corresponds to no underrelaxation. A special value of “robbins-monro” indicates that Robbins-Monro relaxation is to be used:

\[T_{i+1} = \frac{1}{i} T_i + \left (1 - \frac{1}{i} \right) \tilde{T}_{i+1} \]

Default: 1.0

3.3.8. <alpha_rho>

Underrelaxation parameter used on a density update update. Let \(\rho_i\) be the density at iteration \(i\) and \(\tilde{\rho}_{i+1}\) be the next estimate of the density as determined by the thermal-fluids solver. Then, the density for iteration \(i + 1\) is:

\[\rho_{i+1} = (1 - \alpha_\rho) \rho_i + \alpha_\rho \tilde{\rho}_{i+1} \]

Choosing \(\alpha_\rho = 1\) corresponds to no underrelaxation. A special value of “robbins-monro” indicates that Robbins-Monro relaxation is to be used:

\[\rho_{i+1} = \frac{1}{i} \rho_i + \left (1 - \frac{1}{i} \right) \tilde{\rho}_{i+1} \]

Default: 1.0

3.3.9. <temperature_ic>

The initial temperature distribution can be determined either from the neutronics solver or the heat-fluids solver. A value of “neutronics” will use the temperatures specified in the model for the neutronics solver whereas a value of “heat_fluids” will use the temperatures specified in the model for the heat-fluids solver.

Default: neutronics

3.3.10. <density_ic>

The initial density distribution can be determined either from the neutronics solver or the heat-fluids solver. A value of “neutronics” will use the densities specified in the model for the neutronics solver whereas a value of “heat_fluids” will use the densities specified in the model for the heat-fluids solver. Note that this density initial condition strictly refers to the fluid density - the solid density is constant throughout the simulation, and is unchanged from the value used in the neutronics input.

Default: neutronics

3.3.11. <convergence_norm>

This element indicates the type of norm to use for convergence checks. At each Picard iteration, the norm of the difference between the temperature at the previous and current iterations is compared to the value of <epsilon> in order to determine convergence. Valid values for this element are “L1”, “L2”, and “Linf”.

Default: Linf