Landesmessstelle für Radioaktivität Fachbereich 01 Fachbereich Physik/Elektrotechnik Distribution of radionuclides in soils dependence on soil parameters BIOPROTA-Meeting Nancy 24.05.2012
Contents modelling K d model verification sensitivity study for Refesol 2 (variation of soil parameters) redox sensitivity discussion 2of29
Model description Geochemical code: PHREEQC (Parkhurst and Appelo 1999) Complexation models used with PHREEQC: Dzombak and Morel (1990) Bradbury and Baeyens (2009, 2009) Tipping (Model VI, 2002) Model components: soil solution oxalate extractable hydrous ferric oxides (HFO) DM clay (illite as representative material, including frayed edge sites) BB immobile organic matter T dissolved organic matter (DOM) T solid phases 3of29
Study of 18 different U-contaminated soils: Vandenhove et al. (2007) Model Verification Two Cs-contaminated soils: Nisbet (1995) Figure 1. Comparison of measured and calculated U concentrations in soil solution Figure 2. Activity of 134 Cs in soil solution before and after treatment with 11.5 m potassium 4of29
Reference soils (Refesols) Refesol sand silt clay ph C org CEC eff Fe ox Al ox type 01-A 71 24 5 5,7 0,9 37,9 1,57 0,95 Cambisol 02-A 2 83 15 6,6 1,3 133,2 3,54 0,69 Stagnic luvisol 04-A 85 11 4 5,1 2,9 85,7 0,63 1,51 Gleyic podsol 06-A 9 55 36 6,8 2,5 236,6 5,03 1,57 Cambisol-Rendzina 08-A 69 21 10 5,3 1,04 51,2 3,50 0,35 Fluvisol 09-A 62 30 8 5,5 0,86 39,2 2,74 0,73 Luvisol 12-G 32 47 21 5,1 3,75 126,7 7,65 2,28 Cambisol Table 1: characteristics of the reference soils, texture and C org in %, CEC eff (exchangeable Ca, Mg, H and Na) in mmolc/kg, oxalate-extractable oxides in g/kg, source: K.H. Weinfurtner, Fraunhofer Institute for Molecular Biology and Applied Ecology, Schmallenberg, Germany 5of29
Assumptions for the Refesol model The composition of the soil solution of the Refesols is not yet known use of a standard soil solution for modeling: (concentrations are geometric means of ranges of frequent values, Scheffer/Schachtschabel 2010): Na K Mg Ca NH 4 Fe(3) Al Si Cl N P S DOM 6,3 9,5 11 80 0,9 - - 10 24,5 20 0,1 39 54 Table 2: Composition of the standard soil solution (values in mg/l) Fe and Al determined by equilibrium with ferrihydrite and gibbsite no phosphate fertilization important for comparison with literature values DOC 27 mg/l, org. C 50% of org. matter concentration of contaminating nuclides: 1 Bq/kg DW 6of29
Calculated distribution coefficients comparison with literature values calculation using standard soil solution from Table 2 equilibration of initial soil solution with surface assemblage equilibration with contaminated soil solution soil saturated with water RefeSol 2 (loam) RefeSol 4 (sand) 5.0 5.0 4.0 4.0 3.0 3.0 log K d 2.0 1.0 log K d 2.0 1.0 0.0 0.0-1.0 Cs Ni U Se -1.0 Cs Ni U Se Fig. 3: Logarithmic distribution coefficient compared to literature values (IAEA Tecdoc 1616) 7of29
Important soil parameters and processes contaminant concentration dilution/evaporation clay (mineral) content Fe-/Al-oxides (oxalate extractable) immobile organic matter dissolved organic matter ph (redox state) conditions for modelling: saturated soil, no oxygen in solution 8of29
Concentration Concentration dependence of K d 120 100 change in % 80 60 40 Cs Ni U Se 20 0 0,1 1 10 100 1000 10000 concentration in Bq/kg saturation effects 9of29
Dilution Dilution (Mixing with rainwater) 1,0E+04 K d in l/kg 1,0E+03 Cs Ni U Se 1,0E+02 0 20 40 60 80 100 % of original soil solution dilution but constant activity: less competition by major ions 10 of 29
Evaporation Kd in l/kg 1.0E+04 1.0E+03 Evaporation Cs Ni U Se must not be confused with the case of unsaturated soil (relative concentrations constant, same chemistry) 1.0E+02 1.0E+01 0 0.2 0.4 0.6 0.8 1 water fraction evaporation at constant activity: more competition by major ions 11 of 29
Clay content Clay content note linear scale! 2.5E+04 2.0E+04 K d in l/kg 1.5E+04 1.0E+04 Cs K d Ni: 5.0E+03 150 l/kg (0 % clay) 200 l/kg (55 % clay) 0.0E+00 0 10 20 30 40 50 60 clay content in % complexation on clay highly significant for Cs, moderate for Ni, negligible for U and Se (no clay sorption model for Se as yet) 12 of 29
Content of Fe/Al oxides 1,0E+04 Fe/Al hydroxides (oxalate extractable) red: values for Luvisol (Refesol 2) K d in l/kg 1,0E+03 Ni Cs U Se 1,0E+02 1,0E+01 0 2 4 6 8 Fe in g/kg dry soil mass strong influence of Fe/Al oxides for U and Se 13 of 29
Immobile organic matter Organic matter content K d in l/kg 1,0E+04 1,0E+03 Cs Ni U Se red: values for Luvisol (Refesol 2) Soil density effects and blocking of mineral surface sites by org. matter not included 1,0E+02 1,0E+01 0,00 1,00 2,00 3,00 4,00 5,00 6,00 org C in % strong influence on Ni, moderate on U, no model for Se binding as yet 14 of 29
Dissolved organic matter 1.0E+03 Dissolved organic matter 65 % of DOM is active average content of active DOM: 35.6 mg/l K d in l/kg 1.0E+02 Ni U 1.0E+01 0 50 100 150 200 active DOM in mg/l moderate to strong influence on Ni and U, none on Cs, no model for Se binding as yet 15 of 29
CO 2 pressure (organic activity) 1,0E+04 CO 2 - pressure red: values for Luvisol (Refesol 2) K d in l/kg 1,0E+03 1,0E+02 Cs Ni U Se formation of carbonate complexes that keep U in solution, competition effects by carbonate sorption on Fe/Al oxides 1,0E+01 1,00 1,50 2,00 2,50 3,00 3,50 4,00 -log CO 2 16 of 29
Variation of ph ph (equilibrium with Calcite) SI Gibbsite (ph >6) = 2 no precipitation reactions 1,0E+04 pe = 13.5.4.5 1,0E+03 K d in l/kg 1,0E+02 Cs Ni U Se pe 1,0E+01 1,0E+00 4 5 6 7 8 ph ph dependence of K d (U) in batch experiments (Vandenhove et al. 2007) 17 of 29
Comparison of ranges K d ranges 1,0E+05 K d in l/kg 1,0E+04 1,0E+03 1,0E+02 concentration evaporation dilution clay Feox pco2 ph/pe DOM org. C 1,0E+01 Cs Ni U Se 18 of 29
Conclusions sensitivity study in many cases the distribution of radionuclides strongly depends on soil parameters the variation of a single parameter may change the K d by more than an order of magnitude the K d variations can reasonably be modelled by PHREEQC K d variability is important for predicting the influence of environmental conditions on radionuclide distributions in soils 19 of 29
Redox states of soils In soil solutions, a couple of elements (O, C, N, S, Fe.) may occur in various oxidation states e.g. iron as Fe 2+ or Fe 3+ electron transfer example: reduction of ferric hydroxide 4 Fe(OH) 3 + 12 H + + 4e - 4 Fe 2+ + 12 H 2 O the redox potential Eh is given by the Nernst equation: Eh pe = 0 0.059 {Red} = Eh n log { Ox (at 25 C) Eh 0 : standard potential } Eh 0.059 Against hydrogen electrode n: number of transferred electrons { }: chemical activity 20 of 29
Redox states of soils lower pe: reducing conditions, high electron activity higher pe: oxidizing conditions, low electron activity from Sposito 1989 from Appelo/Postma 2005 21 of 29
Modelling redox zonation decomposition of organic matter soil: RefeSol 2 (Luvisol) Assumptions: Decomposition causes no significant changes in the number of organic binding sites Stepwise decomposition of org.matter starting at pe ~ 13 14 (soil solution saturated with oxygen) Reductive decomposition of hydrous ferric oxide (decreasing number of Hfo binding sites) Se precipitated as elemental Se or as FeSe, Ni as Ni(OH) 2, U as Uraninite (UO 2 ) In the K d calculations the precipitated material is assigned to the solid phase (may still be present in the solution as a colloid, e.g. Se!) 22 of 29
Redox behaviour of characteristic solutes concentrations in mol/l 1.0E-03 8.0E-04 6.0E-04 4.0E-04 Fe(2+) sulphate nitrate oxygen Ammonium behaviour as described in the literature (e.g. Appelo & Postma 2005) 2.0E-04 0.0E+00-6 -1 4 9 14 pe 23 of 29
Redox behaviour of uranium 1.0E+07 1.4E-01 Below pe 3: Decomposition of Hfo K d in l/kg 1.0E+06 1.0E+05 1.0E+04 1.0E+03 1.0E+02 pco 2 = 3.5-6 -1 4 9 14 1.2E-01 1.0E-01 8.0E-02 6.0E-02 4.0E-02 2.0E-02 0.0E+00 mol Hfo per l soil solution Release of ions bound by complexation Below pe 0: Precipitation of uraninite (UO 2 ) 1.0E+07 1.0E+06 1.0E+05 pco 2 = 1.75 White circles: Hfo content Black diamonds: distribution coefficient pe Kd in l/kg 1.0E+04 1.0E+03 More carbonate complexes 1.0E+02-6 -1 4 9 14 pe 24 of 29
Redox behaviour of nickel K D in l/kg 1.0E+04 1.0E+03 Precipitation of Ni(OH) 2 at low pe Formation of mixed Al-hydroxides discussed in the literature 1.0E+02 1.0E+01-6 -1 4 9 14 pe 25 of 29
Redox behaviour of Caesium 1.0E+05 1.4E-02 1.2E-02 Ionic strength effect due to ion release (Hfo decomposition!) K d in l/kg 1.0E+04 1.0E+03 1.0E-02 8.0E-03 6.0E-03 4.0E-03 ionic strength in mol/l Precipitation of minerals at low pe, e.g. FeS 2.0E-03 1.0E+02-6 -1 4 9 14 pe 0.0E+00 black diamonds: distribution coefficient white circles: ionic strength 26 of 29
Redox behaviour of selenium K d in l/kg 1.0E+12 1.0E+11 1.0E+10 1.0E+09 1.0E+08 1.0E+07 1.0E+06 1.0E+05 1.0E+04 1.0E+03 1.0E+02 1.0E+01 1.0E+00 1.0E-01-6 -1 4 9 14 pe from Ashworth et al. 2008 (sandy loam, column experiment, Hfo content not specified) pe 9 (531 mv): transformation of selenate to selenite in solution pe 2.5 (150 mv): precipitation of elemental Se pe -0.4 (-25 mv): precipitation of FeSe No modelling of Se binding to org. matter! Colloidal Se in solution? 27 of 29
Conclusions redox study Redox behaviour dominated by Hfo dissolution and precipitation of nuclide K d for U und Se degrees of magnitude higher in the anoxic region K d definition: how to deal with finely dispersed or colloid material? Need for studies that provide all parameters necessary for modelling 28 of 29
Appelo, C.A.J. und Postma, D. 2005. Geochemistry, Groundwater and Pollution (2nd ed.). CRC Press, Boca Raton, Florida. Ashworth D.J., Moore J. und Shaw G. 2008. Effects of soil type, moisture content, redox potential and methyl bromide fumigation on Kd values of radio-selenium in soil. Journal of Environmental Radioactivity 99, pp.1136-1142. Bradbury, M.H. und Baeyens B. 2000. A generalised sorption model for the concentration dependent uptake of caesium by argillaceous rocks. Journal of Contaminant Hydrology 42, pp. 141-163. Bradbury, M.H. und Baeyens B. 2009. Sorption modelling on illite Part I: Titration measurements and the sorption of Ni, Co, Eu and Sn. Geochimica et Cosmochimica Acta 73, pp. 99-1003. Bradbury, M.H. und Baeyens B. 2009. Sorption modelling on illite. Part II: Actinide sorption and linear free energy relationships. Geochimica et Cosmochimica Acta 73, pp. 1004-1013. Dzombak, D.A. und Morel F.M.M. 1990. Surface Complexation Modeling: Hydrous Ferric Oxide. Wiley-Interscience, New York. IAEA 2010. Technical Reports Series No. 472. Handbook of parameter values for the prediction of radionuclide transfer in terrestrial and freshwater environments. Vienna : International Atomic Energy Agency Nisbet A.F. 1995. Effectiveness of soil-based countermeasures six months and one year after contamination of five diverse soil types with caesium-134 and strontium-90. Contract Report NRPB-M546. Chilton: National Radiation Protection Board. Scheffer/Schachtschabel 2010. Lehrbuch der Bodenkunde (16. Auflage), Spektrum Akademischer Verlag Heidelberg. Tipping, E. 2002. Cation Binding by Humic Substances. Cambridge University Press, Cambridge, UK. Vandenhove H., Van Hees M., Wouters K. und Wannijn J. 2007. Can we predict uranium bioavailability based on soil parameters? Part 1: Effect of soil parameters on soil solution uraniumconcentration. Environmental Pollution 145, pp. 587-595. 29 of 29