Title: Ruthenium behaviour under air ingress conditions : main achievements in the SARNET project
1Ruthenium behaviour under air ingress conditions
main achievements in the SARNET project
P. Giordano (IRSN), A. Auvinen (VTT) , G.
Brillant (IRSN), J. Colombani (IRSN), N.
Davidovich (ENEA), R. Dickson (AECL), T. Haste
(PSI), T. Kärkelä (VTT), J.S. Lamy (EDF), C. Mun
(IRSN), D. Ohai (INR), Y. Pontillon (CEA), M.
Steinbrück (FzK), and N. Vér (AEKI)
- ERMSAR 2008
- Nesseber, Bulgaria, September 23-25, 2008
2CONTENTS
1. Context 2. Main issues to be
addressed 3. What was known/ unknown at
SARNET start 4. What has been achieved
within SARNET 5. What remains to be
addressed 6. Conclusions
3CONTEXT
Test HCE3-H01 HCE3-H02
Max. temperature (K) 2200 2160
Oxidation temperature (K) 1770 1790
Oxidation duration (s) 8500 8740
Gas phase 90 H2O 10 Ar 0.2 Ar Air
4CONTEXT
- Gaseous ruthenium oxides release may lead to
significant impact on radiological consequences - Example for a LOCA scenario vessel failure 6h
after the reactor scram filtered releases from
reactor containment after 36h (procedure of
containment venting)
5MAIN ISSUES TO BE ADDRESSED
Ru release ?
Ru transport ?
Consequences on fuel degradation (e.g. Zr/air) ?
air flows?
6WHAT WAS KNOWN/UNKNOW AT SARNET START
Ru transport ? no experimental data, no model
Ru release ? exp. data (AECL mainly), simple model
Zr-UO2/air ? few exp. data NUREG correlation
Air flows ? NRC/SNL studies documented in NUREG
reports
7WHAT HAS BEEN ACHIEVED WITHIN SARNET
Air flows ?
- Independent calculations have been performed on
a series of scenario (breaks size, location,) - by EDF with MAAP/SATURNE (CFD)
- and by IRSN with ASTEC
- Consensus on the range of air mass flows coming
in the RPV (10 mol/s, lasting several hours)
8WHAT HAS BEEN ACHIEVED WITHIN SARNET
Zr-UO2/air ?
- Zr/air interaction (WP9.3) extensive
experimental work performed under prototypical
conditions (IRSN/MOZART, Fzk analytical and
QUENCH tests, INR tests..), considering mixed
atmospheres and pre-oxidation, combined with an
extensive modelling work (PSI, IRSN, FzK,
EdF,...) - Strong degradation of cladding material
(formation of ZrN and subsequent re-oxidation) - Parabolic correlations may be applied only for
high temperatures (gt1400C) and for pre-oxidized
cladding (gt1100C). For all other conditions,
faster ( more linear) reaction kinetics should
be applied - see outcomes from WP9.3 , and papers 2.2, 2.4
9WHAT HAS BEEN ACHIEVED WITHIN SARNET
Ru release ?
- Experimental database mainly from AECL
- AECL / MCE1 series 8 tests
- Fragments of CANDU fuel 10,7 GWd/tU (without
clad) Tmax between 1973K and 2250K air / mixed
steam-argon-H2 - AECL / HCE 3 series 6 tests
- cladded CANDU fuel 9,2 GWd/tU Tmax between
1800K and 2200K air / mixed steam-argon-H2 - 4 additionnal data selected by partners
- HCE1-M17 cladded CANDU fuel 19,1 GWd/tU Tmax
1770K Ar/2H2 and then air - HCE2-LM4 cladded PWR fuel 57,3 GWd/tU Tmax
1670K Ar/2H2 and then air - UCE12-TO2 fragment of CANDU fuel 18,4 GWd/tU
Tmax 1670K steam then air - UCE12-T15 fragment of CANDU fuel 15,4 GWd/tU
Tmax 1900K Ar/2H2 and then air
- CEA data MERARG-Ru test, defined with partners
- PWR UO2 fuel 72 GWd/t, already used for one of
the VERCORS test (well characterised fuel )
heated at 1350C under air
10WHAT HAS BEEN ACHIEVED WITHIN SARNET
Ru release ?
- Model improvements account for the high
dependance of the oxygen potential on the Ru
release (Mansouri Olander, 1998, Minato et al,
1997) gt ruthenium release model based on a
thermodynamic approach of Ru volatilization from
Ru metal /oxides precipitates
11WHAT HAS BEEN ACHIEVED WITHIN SARNET
Ru release ?
RuO2
12WHAT HAS BEEN ACHIEVED WITHIN SARNET
Ru release ?
- Model improvements (continued) account for all
steps of fuel oxidation, i.e. - Mass transfer in the gaseous phase
- Oxygen exchange at gas/solid interface new
correlation for fuel oxygen potential in bulk
UO2x, based on Labroche data (2003)
13WHAT HAS BEEN ACHIEVED WITHIN SARNET
Ru release ?
- New model assessment on AECL data
- ruthenium kinetic release from both de-cladded
(MCE-1 series) and cladded fuel (HCE1, HCE3
series) well fitted (old model no Ru release
predicted)
14WHAT HAS BEEN ACHIEVED WITHIN SARNET
Ru transport ?
15WHAT HAS BEEN ACHIEVED WITHIN SARNET
Ru transport ?
16WHAT HAS BEEN ACHIEVED WITHIN SARNET
Ru transport ?
- Decomposition of of RuO4 to RuO2s significantly
influenced by reactions on the surface
oxidized/metallic surfaces, w/wo pre-deposits of
RuO2, w/wo presence of other FPs (ex Molybdenum
seems to decrease the surface catalysis effect on
decomposition of RuOx)
17WHAT HAS BEEN ACHIEVED WITHIN SARNET
Ru behaviour in containment building
- Literature review
- revealed a lack of quantified data on RuO4
gaseous phase stability in our containment
temperature range, and on RuO2(c) / RuO4(g)
radiolytic conditions.
- Study of the oxidation of ruthenium species in
the aqueous phase (in EPICUR) - Dissolved powder of KRuO4 (representative of
ruthenium aerosols) - RuO2(s)
- RuO4(aq)
18WHAT HAS BEEN ACHIEVED WITHIN SARNET
Ru behaviour in containment building
- Main results stability of RuO4(g)
- RuO4(g) is not as unstable as indicated by the
literature half-life time was experimentally
evaluated to 5 hours in representative
conditions of a SA - Substrate interaction no special affinity
(steel, epoxy,..) no influence on the
decomposition kinetics - decomposition reaction is accelerated by the
presence of steam and deposits of ruthenium
oxides, which act as catalysts.
- RuO4(g) is not as unstable as indicated by the
literature - Half-life time was experimentally evaluated to
around 5 hours in representative conditions of a
SA (temperature, steam) - RuO4(g) decomposition process into RuO2 deposits
is catalysed by steam and ruthenium dioxide
deposits
- RuO4(g) is not as unstable as indicated by the
literature - Half-life time was experimentally evaluated to
around 5 hours in representative conditions of a
SA (temperature, steam) - RuO4(g) decomposition process into RuO2 deposits
is catalysed by steam and ruthenium dioxide
deposits
- RuO4(g) is not as unstable as indicated by the
literature - Half-life time was experimentally evaluated to
around 5 hours in representative conditions of a
SA (temperature, steam) - RuO4(g) decomposition process into RuO2 deposits
is catalysed by steam and ruthenium dioxide
deposits
- RuO4(g) is not as unstable as indicated by the
literature - Half-life time was experimentally evaluated to
around 5 hours in representative conditions of a
SA (temperature, steam) - RuO4(g) decomposition process into RuO2 deposits
is catalysed by steam and ruthenium dioxide
deposits
19WHAT HAS BEEN ACHIEVED WITHIN SARNET
Ru behaviour in containment building
- Main results (continued) revolatilization from
deposits - Revolatilization from ruthenium oxides deposits
have been experimentally evidenced, induced by
oxidizing effect of ozone producing RuO4(g) - 2 key factors T, humidity their increase
clearly favours the oxidation reaction
20WHAT HAS BEEN ACHIEVED WITHIN SARNET
Ru behaviour in containment building
- Main results (continued) revolatilization from
aqueous phase - Ruthenium aerosols behaviour
- No Ru volatilization occurred from aqueous
solution of RuO4- (KRuO4 representative of Ru
aerosol ), and from aqueous solution of
RuO2,2.6H2O, whatever the sump pH - RuO4(aq) behaviour (made by dissolution of RuO4
vapours ) in the case of a basic sump - 75 of Ru volatilized for a 16 hours irradiation
at 90C - 50 of Ru volatilized for a 31 hours heating at
90C without irradiation strong impact of the
heating phase - gt g radiations act only as a booster agent for
the revolatilisation phenomenon - no formation of new aqueous species from RuO4
(aq) during heating, evidencing that Ru
volatilization occurred directly from RuO4 (aq)
- Conclusion the sump is an efficient trap for
Ru aerosols species, while it is not for RuO4(g)
(at least for basic conditions of pH)
21WHAT HAS BEEN ACHIEVED WITHIN SARNET
Zr-UO2/air ? See papers 2.2, 2.4
Air flows ? Independant studies (EDF, IRSN)
convergence on range flows
22What remains to be addressed
- The issue cannot be considered as completely
solved remaining key questions - FP release from high Burn up and MOX fuels
(adaptation of model) - FP release under mixed steam-air conditions
(adaptation of model) - Thermodynamic and nonequilibrium behaviour of
ruthenium oxides during their transport in RCS,
reactivity with surfaces and other chemical
compounds (model development) - Potential release of pre-deposited FPs (like
Iodine), when submitted to oxygen, inducing a
delayed release to containment (experimental work)
23CONCLUSIONS
- Air ingress can occur after a lower head vessel
failure - Air ingress flows have been assessed by
independent calculations (IRSN, EDF). - Air ingress situations have significant
implication for the source term as ruthenium can
form volatile oxide species when exposed to air
- Predictive model for ruthenium release have been
improved satisfactory assessment on analytical
tests both for UO2 fragments and clad fuel (AECL,
CEA, ENEA, IRSN) - Ruthenium is transported in RCS either as RuO2
particles or gaseous RuO4 this has been
experimentally evidenced (VTT, AEKI) - Predictive model for radio-chemical reactions
inside the reactor containment building have been
developed based on experimental observations
(IRSN et al)
SARNET added value
- The issue cannot be considered as completely
solved remaining questions identified
24WHAT HAS BEEN ACHIEVED WITHIN SARNET
Appendix Ru transport ?
- Interpretation decomposition process of RuO4 to
RuO2 was not fast enough to follow the
equilibrium gt kinetic limitations likely
occurred. Transit times of transported Ru oxides
from the hotter zones (furnace areas) to colder
zones (liquid traps)
- Transit times from 1 to 4 seconds within the
range of those from the core outlet zone to the
reactor containment zone calculated for typical
LOCA with a break located at the hot leg or at
the PORV of a Pressurizer