Fuel qualification: Thermodynamic modeling and simulation Molten Salt Reactor Workshop 2017 – Key Technology and Safety Issues for MSRs JW McMurray Oak Ridge National Laboratory TM Besmann University of South Carolina ORNL is managed by UT-Battelle for the US Department of Energy
Thermochemical and thermophysical models are necessary for reactor design and fuel salt behavior predictions • Models are developed from thermochemical and physical properties databases provided by measurements or computational representations • A modeling database is a practical way to retrieve the thermochemical information and provides: – Liquid-liquid immiscibilty – Melting points/precipitation/volatilization – allows for tracking of radionuclides which is key for safety and safegaurds – Chemical potential (corrosion, vapor pressure...) – Inputs for viscosity models (e.g., Quasichemical model gives NNN pairs in multicomponent silicate melts – Grundy et al.) – Heat capacity • Kinetic phenomena can be simulated by coupling with time dependent behavior for representatig reactor/fuel performance • Can be incorporated in reactor simulation codes/real-time reactor control • CALPHAD modeling basis for more reliable prediction of behavior outside empirical data envelope 2 Thermodynamic modeling and simulation
Development of thermodynamic models • The CALPHAD (CALculation of UF 4 - LiF - PuF 3 PHase Diagram) method Projection (J-Salt-liquid), 1 atm PuF 3 – Not just phase diagrams but a T(K) complete thermodynamic picture Four-Phase Intersection Points with J-Salt-liquid T(min) = 762.36 K, T(max) = 1699.95 K 1700 1650 9 1: (Th,U,Pu)F4(ss) / (U,Pu)F3(ss) / Li(Th,U,Pu)4F17(ss)#1 0 . 0 . 1 2: (U,Pu)F3(ss) / Li(Th,U,Pu)4F17(ss)#1 / Li7(Th,U,Pu)6F31(ss)#1 3: (U,Pu)F3(ss) / Li4(U,Pu)F8(ss)#1 / LiF_Griceite_(NaCl_ro(s) – Developed with validated data from 1550 4: (U,Pu)F3(ss) / Li4(U,Pu)F8(ss)#1 / Li7(Th,U,Pu)6F31(ss)#1 8 0 . 0 . 2 A = PuF3, B = LiF, C = UF4 measurements/calculations 1450 X(A) X(B) X(C) K 1: 0.05606 0.39097 0.55297 1040.18 2: 0.01059 0.58458 0.40484 883.66 7 0 . 0 . 3: 0.00506 0.74137 0.25357 770.91 3 1350 4: 0.00423 0.73454 0.26123 762.36 • Semi-empirical physics based Gibbs 1250 6 0 . . 0 4 energy models developed using 1150 5 0 . . 0 5 1050 – Quasi-chemical for molten salts 4 0 . 950 0 . 6 – Compound energy formalism for 850 3 0 . 0 . 7 crystalline phases 750 2 0 . 0 . 8 – Gas law governs vapor phase 1 0 . . 0 9 – Phase equilibria and other relevant thermodynamic values mined from LiF UF 4 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 the open literature or generated with mole fraction experiments/calculations Example of Thermochemical State Calculation in Pseudo-Ternary: Liquidus Projection UF 4 -LiF-PuF 3 • Base models for the fundamental unary and binary subsystems can be integrated to generate many- element simulations. 3 Thermodynamic modeling and simulation
Creating practical, qualified many element thermochemical databases is tractable • For thermochemistry, not necessary to generate cross interactions between all possible constituents – Interaction of minor elements with major constituents sufficient • Experimental efforts would be restricted to – Particularly critical systems Validated – Where uncertainties are outside acceptable bounds databases may – Benchmarking/validation/refining models developed from shorten subsystems development & • Can build on existing empirical data and already speed regulatory assessed databases processes – Data mining to collect and curate information on targeted systems – International partners can provide already developed data from their efforts – NEA Thermodynamics of Advanced Fuels-International Database program can add molten salts and be invaluable asset 4 Thermodynamic modeling and simulation
TAF-ID is an international effort to understand fuel and cladding chemistry with fission products • NEA is developing the Thermodynamics of Advanced Fuels-International Database (TAF-ID) • Current effort objective: Use the CALPHAD method to develop a unified thermodynamic description of fuel/cladding with fission products from key binary/ternary subsystems • Current TAF-ID Systems include: – Fuel: Oxides, nitrides, carbides, metals, minor actinides and fission products add molten salt fuel with fission and corrosion products as well as – Cladding: Zircaloy, SiC, ODS steels, MSR structural materials ferritic steels like FeCrAl, and other advanced materials Application: The database is to be used to aid design of advanced fuel/cladding systems and for physics based fuel performance simulations. 5 Thermodynamic modeling and simulation
Status of molten salt fuel thermochemistry: Data exist for multi-element salts, including U, Th, Pu, but more is needed • Fresh fluoride salt fuel (plus some • These non-fuel elements can affect additional elements) currently the properties and chemistry of the represented by existing data/models fuel. For example, they can: – LiF-NaF-BeF 2 -ThF 4 -UF 3 -UF 4 -PuF 3 - – Alter salt melting point PuF 4 -CrF 2 -CrF 3 -NiF 2 -MoF 5 (solidus/liquidus) resulting precipitates – Chloride salt systems have an even more restricted data set – Alter vapor pressures (source term!) – Create liquid-liquid immiscibility • More data is needed because fuel compositions evolve due to build up of: – Produce unexpected corrosion mechanisms – Fission and transmutation products as – Modify thermal conductivity, heat salts or secondary solid or liquid capacity, and viscosity phases – Corrosion products as salts/solids – Air/moisture contaminants – Refueling – Salt conditioning (redox adjustment) – Mechanical filtering, etc. 6 Thermodynamic modeling and simulation
Significant effort on models and data needed for simulating salt thermochemical state with operation • Liquid and crystalline CALPHAD models for major salt elements are needed – This means determining binary through quaternary interaction behavior • For minor constituents (transuranics, FPs, corrosion products...) only need interaction representations with each of the major constituents – Probability of minor constituents interacting and mangitude of the affect is small and can be neglected • Experimental and computational efforts will be required to obtain the needed data and models – collaboration is essential to provide adequate resources to cover all compositional regions of interest • Uncertainties in values/models can and should be included 7 Thermodynamic modeling and simulation
Need for thermochemical measurements of MSR fuel with major fission products • CsI is a stable iodide • CsF can form when Cs is in excess • Need for thermochemical measurements to define key binary systems and validate models for higher order systems Filled symbols from JRC-Delft Univ. collaboration. Capelli et al. Thermochemistry of fuel, fission products and corrosion products in Molten Salt Reactor The international experimental thermodynamic community has the tools and techniques to attack this problem Example DSC data Thermochemical measurement needs Techniques Vapor pressures Knudsen effusion, transpiration Differential Scanning Calorimetry (DSC), Differential Thermal Analysis (DTA) Phase equilibria Mass Spectrometry (MS) Heat capacities DSC Heats of fusion DSC, other calorimetry Corrosion behavior Redox potentials, exposure testing, etc 8 Thermodynamic modeling and simulation
Coupling thermochemistry to kinetic processes: Example for UO 2+x •Thermodynamic inputs were Requires an Efficient Gibbs Energy coupled to a Finite Element (FE) Minimizer: Thermochimica transport code Database that represents •Thermochimica (similar to thermodynamics of fuel ThermoCalc, FactSage, etc.) is an with burnup open-source software library for computing thermodynamic equilibria with the primary purpose of direct integration into multi-physics codes. •The software is written in Fortran and it can be called from a Fortran, C, or C++ Application Programming Interfaces (API) on a desktop workstation or high performance computing environment. •Software development by M.H.A. M.H.A. Piro and S. Simunovic, Piro is currently being maintained CALPHAD, 39 (2012) 104-110. and developed by M.H.A. Piro and S. Simunovic (ORNL). Outputs become Inputs FEM simulations 9 Thermodynamic modeling and simulation
Temperature distribution with time 33s 200s 0s Oxygen concentration distribution with time Simulation conditions • Composition 1 mole U, 2.05 moles O • 1500 and 500 °C boundary conditions • Single phase fluorite urania test case 10 Thermodynamic modeling and simulation
Recommend
More recommend
