Title: Overview of Experimental Programs on Core Melt Progression and Fission Product Release Behaviour
1Overview of Experimental Programs on Core Melt
Progression and Fission Product Release Behaviour
- B.J. Lewis, Royal Military College of Canada
- R. Dickson, Atomic Energy of Canada Limited
- F.C. Iglesias, Candesco Corporation
- International VERCORS Seminar
- Gréoux les bains, France
- October 15-16, 2007
2Outline
- Experiment Review
- Integral Severe Accident and Single Effect Tests
- Degraded Core Accident Phenomena
- Fission Product Release (FPR) Behavior
3In-Pile Tests
- Source Term Experiments Project (STEP 1,2,3,4)
- Fission product release (FPR) and aerosol
chemistry - Source Term Tests (ST 1,2)
- FPR aerosols from highly-irradiated fuel
(reducing conditions) - Damaged Fuel (DF 1,2,3,4) Relocation Experiment
- Coolant flow rate, system/fuel-rod pressure,
degree of initial clad oxidation - Severe Fuel Damage Tests (SFD ST, 1-1, 1-3, 1-4)
- Fuel bundle FPR (transport/deposition)
behavior, H2 generation
4In-Pile Tests Contd
- Full Length High Temperature Tests (FLHT 1,2,4,5)
- Oxidation H2 generation in full-length rods
- Loss-of-Fluid Test Facility Fission Product Test
(LOFT FP-2) - Large-scale test on FPR, steam supply/reflood
- Melt Progression (MP 1,2)
- Ceramic pool behavior in blocked-core accidents
- Blowdown Test Facility (BTF-104, -105A, -105B,
-107) - CANDU fuel and FPR behaviour
- Phebus SFD ? Phebus Fission Product Tests
(FPT-0,-1,-2,-3-4) - Core, cooling system containment response
including FPR transport/deposition - Semi-volatile actinide release from UO2/ZrO2
rubble bed
5Out-of-Pile Tests (Electrically Heated)
- CORA (19 tests) ? Quench
- Temporal behavior of core melt progression
reflood - PARAMETER (UO2 pellets and VVER cladding (1 Nb))
6Bundle Configurations
7Out-of-Pile Annealing Tests
- Single Effects Fission Product Release (FPR)
Experiments - FPR from spent fuel (hydrogen, steam, air)
- ORNL Horizontal Induction (HI 1-6), Vertical
Induction (VI 1-7) - CEA-CENG HEVA 1-8, Vercors 1-6, High Temperature
(HT 1-3), Release of Transuranics (RT 1-8) - AECL-CRL gt300 tests, e.g., MCE1-1,-6,-7,
MCE2-13,-19, HCE2-BM3,-CM4, UCE12-8 - JAERI Verification Experiments of radionuclides
Gas/Aerosol release (VEGA 1-10)
8ORNL Experiments
- Test conditions
- Highly-irradiated Zircaloy-clad UO2 fuel samples
15-20 cm long (100-200 g) - Atmospheric pressure up to 1700-2700 K (time at
temperature 2-60 min) - Test atmosphere steam in VI-3, hydrogen in VI-5,
hydrogen followed by steam in VI-6, and air and
steam in VI-7 (investigate atmospheric effect on
FPR) - Major differences between VI and HI tests
- VI tests (vertical) vs HI tests (horizontal)
- VI test conducted at higher burnup and
temperature (2300-2700 K) - Measurements obtained
- Sample temperature vs time measured by optical
pyrometry - Thermal gradient tube (TGT) downstream to collect
condensing vapors - Graduated filters impregnated charcoal
cartridges to collect particulates and volatile I
species - Charcoal cold trap to measure fission gases
- On-line measurements of fuel location and Cs-137
in TGT and Kr-85 in gas traps - Post-test analysis of all components by gamma-ray
spectrometry, NAA, spark-source mass spec,
emission spectrometry
9ORNL Experiments Contd
- Major test results
- Similar release rates for noble gases, Cs and I
- Difference Cs transport behavior in steam
relative to hydrogen. - Reactive vapor forms of Cs predominate in
hydrogen conditions and transportable aerosols in
steam - Similar Te and Sb release from UO2 as for
volatile FPs but retained by metallic Zircaloy
until nearly complete clad oxidation - Both Eu and Sb showed sensitivity to oxygen
potential at high temperature. - Limitations
- Segmented furnace tube did not provide good
containment of test environment (oxidation of
graphite susceptor) - Samples at temperature for short period of time
(20 min) which may not be long enough for
oxidative release
10CEA-CENG Experiments
- HEVA program (1983-1989)
- Heated Zircaloy-clad specimens of irradiated PWR
fuel in mixtures of steam/H2 and pure H2 from
1800-2370 K (8 tests) - Gamma spectrometry measured FPR from fuel and
transport - Aerosols collected in heated (temperature varied)
cascade impactor and filters - Control rod materials used in HEVA-07 (Ag-In-Cd
exclusively) and HEVA-08 (control rod and fuel
material)
11CEA-CENG Experiments Contd
- VERCORS program (1989 to 1994)
- Spent fuel samples heated to maximum temperature
of 2620 K (6 tests) - 3 PWR pellets with 2 half pellets (depleted U at
ends) in unsealed Zrly clad - Re-irradiated in SILOE reactor to restore
short-lived FPs (I, Te, Mo, Ba, La) - Post-test gamma scanning (including gamma
tomography) (complete FP mass balance) - Results of FP behaviour
- Nearly complete release of volatiles (Cs, I, Te
and Sb) - Te and Sb initially trapped in unoxidized
cladding - Semi-volatiles (Mo, Rh and Ba) (1/2 that of
volatile release depending on atmospheric
conditions) - Increased Mo release in oxiding conditions ? 92
release (VERCORS 5) vs 47 (VERCORS 4) - Increased Ba and Rh release in reducing
conditions ? 45 and 80 of Rh and Ba, (VERCORS 4)
vs 20 and 55 (VERCORS 5) - Low-volatile FPs and actinides between 3 to 10
(Ru, Ce, Np, Sr and Eu) - Increased Np and Ce release under reducing
conditions (VERCORS 4) - No release of non-volatile FPs (Zr, Nb La and Nd)
- No significant enhancement in release in VERCORS
6 (early fuel collapse and partial liquid corium) - Similar problems due to flow bypass (as for ORNL
tests)
12CEA-CENG Experiments Contd
- VERCORS HT and RT program (1996 to 2002)
- Study FP and actinide release during later phase
of accident with fuel liquefaction - Study FPR behaviour as influenced by
- Fuel type (UO2 versus MOX)
- Fuel morphology (intact pellets versus debris
fragments) - Fuel burnup
- Presence of control materials (Ag, In, Cd and
boric acid) - Environmental conditions (oxidizing or reducing
conditions) - Nb and La release in severe VERCORS HT and RT
tests - Fuel collapse temperature (2400 to 2600 K for
fuel burnups of 47-70 GWd/tU) 500 K below UO2
melting temperature
13AECL-CRL Experiments
- CRL Program (gt 300 annealing tests)
- FPR from clad unclad spent fuel samples (800 to
2350 K in Ar/H2, steam and air atmospheres) - Bare UO2 fragments (0.2-1.5 g each) and cladded
specimens (Zircaloy foil bags and short segments
of Zircaloy-clad fuel with end caps) - Presence of Zircaloy can inhibit/delay release of
volatile FPs - Associated with time to oxidize Zircaloy
cladding. - Volatile FP release rates almost independent of
temperature from 1670 to 2140 K after complete
clad oxidation - FPR Behaviour
- Deposition and transport of FPs studied.
- Volatiles release (Kr, Xe, I, Cs and Te) low in
inert/reducing atmospheres but increase
significantly after clad oxidation in oxidizing
atmospheres - Large fuel volatilization in high temperature
tests with bare fuel (highly oxidizing
environments) ? low-volatile release (Zr, La,
Ba, Ce, Pr, Eu) via matrix stripping process - Eu, Ba released in hydrogen-rich atmospheres vs
Mo, Ru, Nb in steam - Oxygen potential of environment well
characterized - Models developed for steam and air oxidation of
UO2 - Significant release of fission products in air
(Ru, Nb)
14Degraded Core Accident Phenomena
15Degraded Core Accident Phenomena Pressure
- Comparable behavior
- Phebus FP (0.2 MPa), LOFT FP-2 (1 MPa), CORA
(0.2 to 1 MPa), TMI-2 (5 to 15 MPa) - Enhanced clad ballooning failure (low pressure)
- FPT-0 (trace-irradiated) at 735?C
- Gap release measured (SFD, LOFT FP-2, Phebus FP)
- Aerosol composition
- Phebus FPT-0 and FPT-1 control rod (Ag,In,Cd),
thermocouple (Re), fuel rod (Sn,U) materials
(0.2 MPa) - PBF SFD 1-4 FPs more important role (7 MPa)
16Degraded Core Accident Phenomena Control Rod
Effects
- Pressure-Dependent Phenomena
- (i) Low Pressure
- SS clad/Zry guide tube contact with ballooning
(high Cd vapour pressure) ? liquid phase 1150?C - Ejection of molten control rod material ?
chemically dissolves guide tube/ clad of
surrounding rods well below Zry MP (1760?C) (CORA
tests) - (ii) High Pressure
- Failure of SS clad at MP (1450?C)
- Phebus FP tests consistent with low pressure
scenario - Control rod failure at 1120C (FPT-0) and 1350C
(FPT-1)
17Degraded Core Accident Phenomena Metallic Melt
Formation
- Interaction of spacer grids/Zry cladding/control
materials - Relocation
- PBF SFD TMI-2 accident (below coolant level)
- LOFT FP2, Phebus FP CORA (cooler bundle
regions) - Freezing temperature of melt
- 1070 K (Ag-In-Cd alloy) to 1220 K (Zr-Fe
eutectic), 1230 K (Zr-Ni eutectic elemental
silver) and 1460 K (Zr-Ag eutectic) - Metallic blockages in integral tests similar to
TMI-2 but not as extensive (shorter duration) - FPT-0 metallographic examination
- Demonstrates role control rod plays in early melt
formation - Attack of Zry clad by molten Ag-In-Cd alloy
- Zr (20-40 wt), Ag (10-50wt), In (10-40wt), U
(lt15wt), O (lt10wt), SS (lt5wt)
18Degraded Core Accident Phenomena Zircaloy
Oxidation
- Exothermic reaction (6.5 kJ/g-Zr oxidized)
- Accelerated heatup rates (?10 K/s) at T gt 1500 to
1700 K - In-pile tests (PBF SFD, LOFT FP-2 and Phebus FP)
and out-of-pile tests (CORA)
19Degraded Core Accident Phenomena Hydrogen
Generation
- Related to steam availability
20Degraded Core Accident Phenomena Fuel
Liquefaction
- UO2 ZrO2 dissolved by metallic Zircaloy/?-Zr(O)
(1760 to 2000C) - Fuel liquefaction in integral tests
- PBF SFD tests (15-18), LOFT FP-2 (15)
- TMI-2 core (45)
- Phebus FPT-1 (20), FPT-0 (50)
- Separation between ceramic and metallic blockage
- (U,Zr)O2 freezes at higher temperature
21Degraded Core Accident Phenomena Molten Pool
Formation
- Ceramic heatup
- Steam diversion around blockage poor thermal
conductivity - FP decay heat (TMI-2) or increased
fission/electrical heating (integral tests) - Molten pool (surrounded by U-rich crust) under a
cavity - TMI-2 (U,Zr)O2 ceramic with transition metal
oxides (Cr2O3, Fe3O4) 2700 K - Phebus FPT-0 (U,Zr)O2 lattice with U (62wt),
Zr (22 wt) O(14wt), Fe(0.6wt) 2720 K - Fuel movement
- TMI-2 Thermo-mechanical failure of crust ? 20 t
lower plenum - FPT-0 Downward motion of pool (from lower grid
spacer) (18100 s) - Comparable to MP tests
22Degraded Core Accident Phenomena Debris Bed
Formation
- TMI-2
- Top of molten pool lower plenum region
- Integral tests
- Upper debris bed formed by coolant injection
(fragmentation) (SFD-ST LOFT FP-2) - Less steam-rich transients - decladded
fuel/fragments in upper part of bundle due to
melting/relocation of clad (SFD 1-4, Phebus FP)
23FPR Behavior
- Comparison
- PBF SFD-ST (steam-rich)/SFD (steam-starved),
Phebus FPT-1 (steam-rich), TMI-2 - Ce actinides (typically lt 0.01)
- Ru,Sr,Sb (lt1)
- Ba (few )
- Mo (up to 50)
- Te (between 1 to 83)
- I,Cs, Noble Gas (up to 90)
- Comparable to annealing tests (ORNL, CEA, CRL)
24FPR Phenomena
- FP trapping
- Te release (clad-oxidation state) Sn segregation
? SnTe release - Sb sequestered in metallic melts (Ni, Ag alloys)
- Burnup
- Enhanced volatile FPR in SFD 1-4 (high-burnup) vs
SFD 1-1 (trace-irradiated) - Swelling (irradiated rods) in Phebus FPT-1
- Oxygen potential (H2/H2O ratio)
- Low Ba, Sr (Eu) release in steam tests (low-
volatile oxides/hydroxides) vs higher release in
reducing tests (ST, VI, HEVA, VERCORS, CRL) - Enhanced Ba release in FPT-0 during escalation
phase (H2 generation) - Low Ru release (PO2 too low to form high-volatile
oxides) - Low actinide release (fuel volatilization (UO3)
)
25FPR Phenomena Contd
- FP behaviour in molten pool
- No enhancement with fuel liquefaction
(non-coherent process) - Volatile FPs bubble nucleation,
coalescence/growth release via buoyancy - Bubble trapping at pool/surrounding crust
- Volatile FPs (I,Cs) in previously molten ceramics
in PBF SFD, Phebus FP TMI-2 reactor - TMI-2 Iron oxides in melt (lower limit of -120
kJ/mol) - La, Ce, Sr as oxide (soluble in (U,Zr)O2)
- Ru, Sb metal immiscible in ceramic melt
- Cooldown/reflood
- Small release (fuel relocation) in Phebus FPT-0
vs large release in LOFT FP-2 (12 volatile
release) (local heating)
26Ba Release in ORNL, CEA and Phebus Tests
Test Temp. (K) Duration (min) Atmosphere Ba Release ()
HI-4 HI-5 VI-2 VI-3 VI-4 VI-5 HEVA-4 HEVA-6 VERCORS-1 VERCORS-4 VERCORS-5 VERCORS HT-1 Phebus FPT-0 PEBUS FPT-1 2200 2025 2300 2700 2440 2720 2270 2370 2130 2570 2570 3070 2700 2500 20 23 60 20 20 20 7 30 17 30 30 7 - - H2O H2O H2O H2O H2 H2 H2O H2 H2 H2O H2 H2 H2O H2 H2O/H2 H2O/H2 lt1 lt1 19 30 27 76 6 27 4 80 55 49 1 1
27Difference in Ba Release from Out-Pile and Phebus
Tests
- Short duration of temperature escalation in
in-pile tests - No high temperature plateau as in annealing
tests but rather temperature escalation (with
formation of molten pool) in Phebus test - Ba volatility is reduced with significant amount
of ZrO2 in fuel melt (47 mol) and small amounts
of iron oxide in in-reactor test - Thermochemical analysis with GEMINI2
- Reduced Ba vapor pressure in solidus-liquidus
transition zone in the U-Ba-O phase diagram
(2400-3100 K)
28Concluding Remarks
- In- and out-pile experiments on severe accident
melt progression FPR behaviour reviewed - Melt progression non-coherent process
- Phebus FPT-0 and -1 tests performed for longer
high-temperature period than earlier in-pile
experiments ? information on late-phase behaviour - Local propagation of core melt due to control rod
failure (at lower temperature) - Metallic blockages result from interactions of
spacer grids, fuel rod cladding material and
control rod materials that flow down bundle and
solidify at lower (cooler) position - Separation between metallic and ceramic blockages
with freezing of (U,Zr)O2 melt at higher
temperature - Observed melting temperature of ceramic blockage
in Phebus FPT-0 test (2720 K) slightly lower
than pure ceramic (2800 K) due to eutectic
interaction (consistent with TMI-2 examination) - Molten pool formed due to increased fission heat
generation and reduced heat transfer in several
in-pile experiments (held in place by ceramic
crust)
29Concluding Remarks Contd
- Consistent release behaviour of volatile (Xe, Kr,
I, Cs, Te and Sb), semi-volatile (Mo, Rh, Ba),
low volatile (Ru, Ce, Np, Sr and Eu) and
non-volatile (Zr, Nb, La and Ne) FPs observed in
annealing (ORNL, CEA, CRL) and in-pile tests - Reduced volatility of Ba for in-reactor
experiments (thermochemical effects with presence
of iron and zirconium oxides) - Local atmospheric condition/oxygen potential
influences low-volatile fission product release
behaviour - Non-coherent melt progression masks individual
release mechanisms identified in out-of-pile
experiments - Enhanced release due to fuel liquefaction not
typically observed in separate effects experiments
30Acknowledgements
- Discussions with P. Elder, L. Dickson, M.
Schwarz, R. Zeyen, B. Clement, M. Kissane and G.
Ducros - Funded by NSERC/COG Collaborative Research and
Development (CRD) Grant