Long-term metal retention processes in a peat bog : Field studies, data and modelling

The study was inspired by the need to assess long-term metal retention in municipal solid waste (MSW) landfills. The long-term processes in landfills are poorly known due to the relatively short time that such landfills have been in existence. Natural analogues where similar metal binding processes could be expected were therefore sought for. The work described in this thesis aims to elucidate the long-term transport and attenuation processes involved in the retention of heavy metals in a peat bog, through field studies and modelling. The Oostriku peat bog (central Estonia) has been exposed to metal-rich groundwater discharge over a long period of time and was found to have accumulated high concentrations of Fe, other heavy metals (e.g. Pb, Cu, Zn, Mn), and As. It was characterised in detail with respect to metal depth distribution and main metal binding mechanisms (using an optimised Tessier extraction scheme). The oxidation of metal sulphides in the surrounding carbonate bedrock was proposed to be a possible long-term source of heavy metals in the water emerging in a spring at the peat site. The water in the spring and peat pore-water was sampled and analysed. The dissolution sequence of the sulphide minerals and evolution of the water composition along a flowpath in the carbonate rock were modelled. Resulting aqueous phase concentration of major and minor elements are discussed in relation to governing geochemical processes. The simulated water composition was compared with that observed. Retention of metals transported with water through the peat was assessed through modelling equilibrium sorption on solid organic matter and amorphous ferric oxyhydroxide by using a simplified quantitative modelling approach and independently obtained data. Dynamic evolution of metal sorption fronts along a peat profile over time was modelled to test metal-metal competition effects. A possible formation of ferric oxyhydroxide in the peat bog was also assessed with the model.QC 2010100