15.08.2006 Ref.: EURISOL DS/Task5/TN-06-12
Thermochemical studies on the mercury/iodine system Contribution to WP5.1, Deliverable D1
R.Moormann, FZJ This report submits themochemical examinations with SOLGASMIX on the iodine/Hgsystem. The iodine volatility will be reduced by HgI2-formation One of the conservatisms already detected in PSAR-SNS, the safety report for SNS with its Hg-target [1], is the volatility of iodine nuclides: Although, due to the pronounced chemical reactivity of iodine, the formation of iodine compounds with lower volatility than the element has to be expected, iodine was assumed as an high volatile nuclide as in PSAR as tritium as noble gases. This leads to a pronounced release of iodine nuclides into the environment, in case that there is a pathway between target and environment (even if the vaporisation of mercury remains very limited, as in loss of confinement events). As shown in Vol. lll (ESS technical report, 2002), for design basis accidents in ESS the iodine release has to be limited to < 0.25 % of the total inventory. Accordingly, there is an urgent need for reduction of iodine source terms even for Hg-targets, as used in Eurisol.. SPECIES He I I2 Hg HgI HgI2 FeI2 NiI H2 HI
Standard Enthalpy [J/mol] GAS
Standard Entropy [J/mol/K]
.00000E+00 .10680E+06 .62440E+05 .61380E+05 .13347E+06 -.16150E+05 .85710E+05 .24685E+06 .00000E+00 .26000E+05
.12600E+03 .18069E+03 .26100E+03 .17501E+03 .28072E+03 .33624E+03 .34960E+03 .27022E+03 .13060E+03 .20650E+03
CONDENSED PHASES Hg (liquid) HgI2 (liquid) I2 (s) HgI (s) HgI2 ( -phase, s) HgI2 ( -phase, s) Hg2I2 (s) NiI2 (s) Ni (s) FeI2 (s)
.00000E+00 -.84000E+05 .00000E+00 -.60300E+05 -.10290E+06 -.10540E+06 -.11910E+06 -.78240E+05 .00000E+00 -.10460E+06
.77400E+02 .22340E+03 .11620E+03 .12130E+03 .18760E+03 .18130E+03 .24130E+03 .13870E+03 .29870E+02 .16736E+03
Thermochemical studies on the mercury/iodine system- by R.Moormann/FZJ (15.08.2006)
1
Fe (s)
.00000E+00
.27170E+02
Table I: Thermochemical input data for SOLGASMIX calculations In order to quantify the arguments for a lower volatility of iodine, some thermochemical calculations on the iodine/mercury system were already performed within the JSNS project [2]; these calculations, which concentrate on the condensed state, indicate that iodine exists within the target mainly as mercury iodides. These calculations were extended at FZJ to the gaseous phase, existing in equilibrium with the mercury target [4]: The Code SOLGASMIX [3], developed for examinations of thermochemistry in nuclear fission reactors, was applied to the target; this code minimises the Gibbs free enthalpy, input data are the molar amounts of elements in the systems, and the compounds in gaseous, condensed and dissolved state to be considered with their respective standard enthalpy and entropy data. A list of input data used is given in the following table, data are taken from [4,5] and concerning solid HgI from [2]: 0
629 K
400 K
-8
ln (p[bar])
-16
I I2 Hg HgI HgI2
-24
-32
-40 0,0015
0,00175
0,002
0,00225
0,0025
0,00275
inverse temperature 1/T [1/K]
Fig. 1: Equilibrium vapour pressures over a Hg/I system (surplus of mercury) He is used only as ‘dummy’ inert gas, required for a constant total pressure. Fig. 1 - 3 contain two examples of SOLGASMIX calculations for the before mentioned problem; results in fig. 1 are valid for a Hg/I system with large Hg surplus at temperatures between 400 K and 629 K (mercury boiling at 1 bar); besides gaseous species listed in the figure legend, in the condensed phase the species Hg, I2, HgI, HgI2 (2 different solid state phases and the liquid phase) and Hg2I2 are taken into account as individual, separated phases. Dissolution of the considered species in mercury is not considered, because ionic Thermochemical studies on the mercury/iodine system- by R.Moormann/FZJ (15.08.2006)
2
compounds usually cannot be dissolved in metals to a remarkable extent; further on, dissolution is expected to decrease vapour pressures, so this assumption is conservative regarding release. It should be noted, that the vapour pressure of elemental iodine, not shown in fig. 2, is already at 400 K (1/T=0.0025) at about 0.2 bar, which is about 4 orders of magnitude higher than the total vapour pressure of all iodine compounds in figure 1; the vapour pressure of elemental iodine was assumed in PSAR for release calculations. Comparison of the vapour pressures of the iodine species with the vapour pressure of elemental mercury indicates, that the total volatility of iodine compounds is much smaller than that of mercury. Obviously, HgI2 is the dominating gaseous iodine species; in the condensed state HgI is the main compound, as already found in [2]. Examination of the influence of other elements formed during spallation concentrated on hydrogen (tritium), because hydrogen iodide HI is highly volatile and might reduce the before mentioned trend. Additional SOLGASMIX calculations, taking hydrogen and HI into account (total molar amount of hydrogen in the system = 100 fold that of iodine) do not show a substantial increase the total vapour pressure of iodine compounds, which means, that the iodine volatility is still below that of mercury at T < 590 K. In addition it has to be mentioned, that formation of other, more stable hydrogen compounds (NH3, H2O etc.) might reduce the availability of hydrogen for HI formation. Results of respective calculations are shown in figure 2. 0
-8
ln(p[bar))
-16
I I2 Hg HgI HgI2 HI
-24
-32
-40 0,0015
0,00175
0,002
0,00225
0,0025
0,00275
inverse temperature 1/T [1/K]
Fig. 2: Equilibrium vapour pressures over a Hg/I/H system (surplus of mercury, molar ratio H/I = 100) Thermochemical studies on the mercury/iodine system- by R.Moormann/FZJ (15.08.2006)
3
Altogether, this means, that the volatility classification of iodine within the schema used in PSAR-SNS may be changed from high volatile class to at least the mercury (medium volatility) class for loss of confinement events (where boiling of mercury has not to be considered). This will substantially reduce the iodine source terms for this accident: The immediate fractional iodine release of PSAR-SNS was assumed to 0.05, whereas under consideration of the above analysis the source term is calculated to 0.0005, assuming mercury = iodine volatility. For accident with mercury volume boiling the iodine source term is probably not changed: The reasons are, that the total iodine vapour pressure is slightly higher than that of mercury at 629 K and that the density of mercury iodide HgI is with 7700 kg/m³ [2] substantially smaller than that of mercury; the latter means, that iodides should swim on top of the mercury, which is favourable for vaporisation. So, for accidents with mercury boiling, as in PSAR a complete release of iodine has to be assumed, even if the mercury is not completely evaporated. The same is true for accidents without volume boiling of mercury as for seismic events with following explosion and for hydrogen explosion: Here, by explosion heat the mercury surface becomes evaporated, where - as outlined before - the HgI is probably concentrated. The assumption of iodine = mercury volatility does not yet reflect the complete potential, shown by the thermochemical volatility analysis for loss of confinement accidents; however, an application of this complete potential requires validation experiments as proposed at the end of this subchapter. Calculation results shown in fig. 3 consider - in addition to the assumptions described for fig. 1 - the influence of iron (target hull, assumed as large surplus compared to iodine) on the iodine volatility: FeI2 as solid and as gas is taken into account in the calculations using SOLGASMIX.
Thermochemical studies on the mercury/iodine system- by R.Moormann/FZJ (15.08.2006)
4
0 629 K 400 K -8
-16
ln (p[bar])
I I2 Hg HgI HgI2 FeI2
-24
-32
-40 0,0015
0,00175
0,002
0,00225
0,0025
0,00275
inverse temperature 1/T [1/K]
Fig. 3: Vapour pressures over a Hg/Fe/I system (surplus of Hg and Fe) This consideration of iron leads to an additional decrease of the iodine volatility; main iodine component in the condensed state becomes FeI2. Further calculations on the influence of Ni (Ni, NiI2 in solid, NiI in gas phase) indicated no significant volatility reduction effect as for Fe. If the formation if FeI2 is assumed, the following iodine source term reduction can be taken into account; on the left of the slash the immediate fractional iodine source term of PSAR-SNS / on the right of the slash the expected reduced iodine source term assuming iodine = mercury volatility, based on the volatility analysis above: -
loss of mercury flow with proton beam shut down failure (1.0 / 0.04-0.3)
-
loss of heat sink with proton beam shut down failure (1.0 / 0.3)
For the following explosion events, a substantial retention can only be assumed in the first analysis step, if FeI2 remains adsorbed on the hull surface and is not transported into the mercury; in the latter case, a vaporisation by explosion heat cannot be excluded, although the boiling point of FeI2 is with 1334 K about 700 K higher than that of mercury. Assuming again mercury = iodine volatility leads to the following source terms: -
seismic events with following explosions, but without following fire (1.0 / 3.e-3)
-
hydrogen explosions without fire (1.0 / 0.005-0.015)
Altogether, because the iodine/iron interaction occurs as a surface reaction, the kinetic conditions are not as favourable for a formation of FeI2 as it is the case for mercury iodides. Accordingly, this effect of the target hull cannot be considered within the first Eurisol safety analyses without an experimental validation. Thermochemical studies on the mercury/iodine system- by R.Moormann/FZJ (15.08.2006)
5
Such validation experiments are planned, equilibrating (inactive) mercury with traces of iodine at different temperatures and determination of the gas phase composition over the liquid mercury (sampling by freezing out the gas and measuring the composition by activation analysis). Experiments will be done with and without added target hull material. References [1] SNS Preliminary Safety Analysis Report (PSAR), Feb. 28 (2000), US DOE Contract No. AC05-96OR22464 [2] K.Kobayashi et al.: Physical and thermochemical properties for inorganic mercury compounds; 2nd Int. Workshop on Mercury Target and Cold Moderator Engineering; 13.-15.11.2000, JAERI, Tokai [3] T.M.Besmann: SOLGASMIX-PV, a computer program to calculate equilibrium relationships in complex chemical systems; ORNL/TM-5775 (1975) [4] JANAF Thermochemical Tables, J.Phys.Chem.Ref.Data 14 (1985) Suppl. No 1 [5] O.Knacke et al., Thermochemical properties of inorganic substances, 2nd ed. (1991), Düsseldorf
Thermochemical studies on the mercury/iodine system- by R.Moormann/FZJ (15.08.2006)
6
