Integrating STAR-CCM+ with a Systems Analysis Code for Nuclear - - PowerPoint PPT Presentation

Integrating STAR-CCM+ with a Systems Analysis Code for Nuclear Reactor Safety Simulations

Justin W. Thomas Nuclear Engineering Division Argonne National Laboratory STAR-American Conference Chicago, Illinois June 28, 2011

Why Sodium-cooled Fast Reactors (SFRs)?

§ All nuclear power plants currently operating the U.S. use water as their coolant – But the first reactor to generate electricity was a fast reactor § Fast reactors get their name because, on average, neutrons are moving faster than in water reactors – Changes the likelihood of the occurrence of various nuclear reactions § Fast reactors can be designed for: – Actinide burning: Continue to produce energy from “used” nuclear fuel from water reactors – Breeding: Produce more fissile fuel than what consumed in the core § As a part of a strategy to recycle used nuclear fuel from water reactors, SFRs help to:

– Extract more energy from uranium – Reduce reliance on uranium enrichment – Reduce the amount of used fuel

STAR-­‑American ¡Conference, ¡Chicago ¡Illinois, ¡June ¡28, ¡2011 ¡ 2 ¡

Passive safety of SFRs

§ Nuclear power plant operators must convince regulators that their reactors will remain safe, even under accident scenarios and off- normal events § Even after the nuclear fission reactions have stopped in the reactor, a small amount of heat is still heat being generated – decay heat – which needs to be removed for an extended period of time § If the coolant pumps fail, SFRs can rely on natural circulation to drive coolant through the reactor’s core and remove heat § The potential for SFRs to survive severe accident initiators with no damage was demonstrated in a series of tests at the Experimental Breeder Reactor-II facility in the 1980s

– Complete loss-of-flow and loss-of-heat-sink tests were performed – Experimental results from this program will be used to validate the work described here

STAR-­‑American ¡Conference, ¡Chicago ¡Illinois, ¡June ¡28, ¡2011 ¡ 3 ¡

Modeling transients in SFRs

§ Argonne’s safety systems code SAS4A/SASSYS-1 models the dynamic response of a reactor during a posited transient scenario § Physics include:

– The core’s response to changes in its environment – Structural mechanics – Fuel performance – Decay heat generation – Fluid mechanics and heat transfer

• Natural circulation, buoyancy-driven flow, thermal stratification

§ By including STAR-CCM+ in SAS4A/SASSYS-1 transient analyses, the goal is to improve the fluid mechanics/heat transfer solution while still maintaining the sophistication of the other models available in SAS4A/ SASSYS-1

– Specific cases where 3-D effects are important – E.g., thermal stratification in large plena

STAR-­‑American ¡Conference, ¡Chicago ¡Illinois, ¡June ¡28, ¡2011 ¡ 4 ¡

5 ¡

Example: Loss-of-flow in Toshiba’s 4S reactor

§ Argonne supported safety analysis for a small SFR concept developed by Toshiba § In a hypothetical loss-of-flow scenario:

• 1. The pumps stop, reducing the flow rate

through the core

• 2. The reactor scrams, stopping the nuclear

fission reactions but decay heat remains

§ Because of #2, the sodium entering the

• utlet plenum is now cooler than the bulk

sodium in the plenum § Because of #1, the time for the cooler sodium to reach the Intermediate Heat Exchanger (IHX) can be significant

à Thermal stratification

§ This effects the natural circulation head in the system 4S Schematic (Not to scale)

STAR-­‑American ¡Conference, ¡Chicago ¡Illinois, ¡June ¡28, ¡2011 ¡

6 ¡

Example: Loss-of-flow in Toshiba’s 4S reactor

§ A model of the 4S outlet plenum was built with STAR-CCM+

– 2-D axisymetric for demonstration purposes

§ Remainder of reactor system modeled with the system code SAS4A/SASSYS-1 § STAR-CCM+ and SAS4A/SASSYS-1 communicate at the flow boundaries § For each core channel, SAS4A/SASSYS-1 sends the outlet temperature and mass flow rate

– Temperature and fluid velocity distributed uniformly along STAR-CCM+ boundary

§ At the IHX inlet, STAR-CCM+ provides the average pressure and temperature 4S Schematic (Not to Scale)

STAR-­‑American ¡Conference, ¡Chicago ¡Illinois, ¡June ¡28, ¡2011 ¡

Temperature predictions in the outlet plenum

STAR-­‑American ¡Conference, ¡Chicago ¡Illinois, ¡June ¡28, ¡2011 ¡ 7 ¡

§ SAS4A/SASSYS-1 predicts the low flow rates and cooler temperatures from the core when the transient starts § Cool sodium slowly progresses upward through the plenum towards the heat exchanger § Important to predict the time delay required for cooler sodium to reach the heat exchanger § Note: These results are preliminary and should not be considered to represent the actual performance of the 4S reactor From Core To Heat Exchanger

Reactor system response

§ The natural circulation driving head depends on the temperature difference between the heat exchanger (heat sink) and the core (heat source) § Some interesting phenomena were predicted during the coupled simulations

• f SAS4A/SASSYS-1 and STAR-CCM+

STAR-­‑American ¡Conference, ¡Chicago ¡Illinois, ¡June ¡28, ¡2011 ¡ 8 ¡

Secondary-side becomes a heat source rather than a heat sink due to flow stagnation Long delay before IHX senses cooler core temperatures Temperature

Implementation of STAR-CCM+ coupling

§ Coupling with the SAS4A/SASSYS-1 code is implemented through the STAR-CCM+ client via a Java macro

– Portions of Fortran routines developed for STAR-CD coupling preserved – Java calls the Fortran functions via Java Native Access (JNA)

§ Communication between SAS4A/SASSYS-1 and STAR-CCM+ via file I/O § Synchronize each SAS4A/SASSYS-1 time step

– SAS4A/SASSYS-1 determines its time step size using its normal approach

• Monitors temperature changes and other conditions, user-input tolerances

– STAR-CCM+ time step is the smaller of:

• ½ the SAS4A/SASSYS-1 time step
• The user-input value in the simulation

– Linear interpolation performed as needed

STAR-­‑American ¡Conference, ¡Chicago ¡Illinois, ¡June ¡28, ¡2011 ¡ 9 ¡

Implementation of STAR-CCM+ coupling (cont)

§ At the end of its time step, SAS4A/SASSYS-1 prints for each inlet flow boundary

– Mass flow rate – Temperature

§ STAR-CCM+ assumes a uniform velocity and temperature profile at each flow boundary, computed from the SAS4A/SASSYS-1 data § Just before the next SAS4A/SASSYS-1 time step, STAR-CCM+ prints for all flow boundaries

– Area-averaged absolute pressure – Mass-flow averaged temperature

§ STAR-CCM+ annotates a plot with the current time (from SAS4A/ SASSYS-1) and prints for the animation § Heat transfer at boundaries to be implemented soon

– Exchange heat flux or temperature at wall boundaries

STAR-­‑American ¡Conference, ¡Chicago ¡Illinois, ¡June ¡28, ¡2011 ¡ 10 ¡

Reports

11 ¡

Future Work: EBR-II Analysis

§ But are these predictions accurate? § Measured data from the EBR-II tests provides a validation exercise of whole-plant response to a loss-of-flow scenario

– Cold pool tank modeled with STAR-CCM+ – Remainder of the cooling system modeled with SAS4A/SASSYS-1

§ Seven flow boundaries that connect to SAS4A/SASSYS-1

STAR-­‑American ¡Conference, ¡Chicago ¡Illinois, ¡June ¡28, ¡2011 ¡

EBR-II Initialization

STAR-­‑American ¡Conference, ¡Chicago ¡Illinois, ¡June ¡28, ¡2011 ¡ 12 ¡

RGG: Reactor Geometry and Mesh Generator

• A set of tools to generate reactor assembly, core geometry and mesh

models.

• Fuels and other rods are grouped in to form assemblies and lattice of

assemblies are grouped in to form a core.

• RGG takes advantage of information about repeated structures in both

assembly and core lattices.

• Provides a balance between lattice-guided automation and opportunities

for user interaction at key points of the process.

• Supports rectangular and hexagonal lattices.
• Operates in 3 stages:
• 1. AssyGen
• 2. Meshing
• 3. CoreGen

13 ¡

• In this step assembly

model and mesh script are created.

• Keyword based input file

is used to define assembly geometry.

• AssyGen supports

rectangular and hexagonal assembly types.

• AssyGen created mesh

script, MeshKit algorithms

• r user defined mesh

script can be used to for meshing the assembly geometry.

• Side skin surface of all

the assemblies forming the core must have matching nodes.

• CoreGen copy-move-

merges assemblies to form the core.

• Metadata propagation from

individual assembly meshes to the core

• Core geometry/mesh can be

exported into several file formats

• Several symmetry options

available STAGE ¡1: ¡ASSYGEN ¡ STAGE ¡2: ¡MESHING ¡ ¡ STAGE ¡3: ¡COREGEN ¡

MONJU reactor, full core model: 9.7M hexes, 99k vols takes 4.3GB RAM and 176

• mins. 715

assemblies.

RGG: Reactor Geometry and Mesh Generator

Thank you

STAR-­‑American ¡Conference, ¡Chicago ¡Illinois, ¡June ¡28, ¡2011 ¡ 16 ¡