Dissemination and use of palaeohydrogeological results for safety assessment

Work Package (WP) 5 of the PADAMOT project aims at introducing the use of palaeohydrogeological information for the identification of scenarios that are to be considered in Performance Assessments (PAs) of future radioactive waste repositories. With this objective, WP5 reviews existing safety cases to find out the extent to which palaeohydrogeological information has been considered in carrying out PAs and to see what sort of evidence was available from geochemical studies and how it was used. The review shows that palaeohydrogeology has been used very variably in PA for which a number of possible reasons are discussed. One of those is that palaeohydrogeological information from geochemistry and mineralogy tends to contain many uncertainties and is rarely translated into qualitative parameters that can be used directly in PA. To improve this translation, it is proposed here that interpretation of palaeohydrogeology should focus on input to the consideration (or ‘screening’) of Features, Events and Processes (FEPs) in the PA process. FEPs provide a context for palaeohydrogeological information that is recognised in PA. However the challenge is to quantify the information which requires interpretative models to be developed (of which there are some examples in WP4). To illustrate the logic, albeit complex, of transforming geochemical and mineralogical measurements (exemplified by analyses carried out in WP2) into information that is usable by PA, a flow chart of data, supporting information, interpretation and calculations is produced. The interpretation methods that are used to get from geochemical and mineralogical data to palaeohydrogeological information that is useful for screening FEPs are described at a reasonable level of detail. This should assist PA groups in understanding the assumptions and uncertainties that are involved in using information obtained fro applications of geochemistry and mineralogy in site characterisation, as exemplified in WP2. Geochemistry and mineralogy were used in WP2 to investigate past groundwater conditions at study sites in Spain, Sweden, UK and Czech Republic. The specific aims, strategies and output data in each case are summarised and the significance of the results in terms of support for FEPs is assessed objectively. Coupled flow-reaction modelling of the Spanish site, Los Ratones, shows that groundwater compositions and secondary mineralisation are potentially sensitive to changes of groundwater recharge rate during fluctuations of climate from warm-arid to cold-humid. The hydrogeochemistry of the part of the groundwater system that is accessible for study is dominated by weathering and dissolution reactions and precipitation predominantly only of clays. Overall, the investigations here have shown geochemical and mineralogical data that reflect present-day groundwater conditions. There are no data that can be clearly associated with the past climate and hydrological variability for which there is a comprehensive archive of microfossil and organic geochemical data in the Cúllar-Baza and Padul sedimentary sequences. Mineralogical studies at the Äspö/Laxemar area in Sweden have aimed to follow up previous studies at Äspö by investigating the evidence for past meteoric and saline groundwater distributions at Laxemar where the present-day penetration of meteoric water is deeper than at Äspö. Although there are only small amounts of late stage calcite, information from its morphology and distribution indicate both fluctuating salinity at shallow depths and deep penetration of freshwater in the past. This information, though having substantial uncertainties in the chronology and continuing of the calcites, is clearly of qualitative value for FEPs concerning likely salinity changes due to climate and land uplift. Additional petrographic analyses of late stage calcite in drillcore samples from past site investigations at Sellafield in northwest England have consolidated the findings from previous work carried out in the EQUIP project. Morphology variations in overgrowths of late stage calcite have suggested that the position of the fresh/brackish-to-saline water transition zone has fluctuated both above and below its present location, but with additional observations the balance of evidence supports the predominance of a slight downward movement, by at most a few tens of metres, of the transition zone over time, i.e. a dominant trend over time of decreasing salinity at any point in this interval. This indicates that a hydrodynamic and hydrochemical response at 300-400 m depth to changes in the surface environment should be considered in FEPs, but also that the distributions of flow directions in the groundwater system has remained fairly stable over the time period represented by the late stage calcites. Data from various instrumental methods for analysing chemical and isotopic compositions of discrete calcite growth zones suggest that the compositions of groundwaters from which they precipitated changed over time. The patterns of variations support the concept that changes in deep saline groundwaters are more attenuated than in fresher up-gradient groundwaters. In these freshwater calcites, contents of the redox-sensitive trace elements Fe, Mn and Ce are correlated which indicates that the fluctuations in compositions are related to changes in palaeoredox conditions. Thus palaeohydrogeological information suggests that FEPs for PA should consider long-term changes of redox, although it is also evident that the scale of change is attenuated in deeper saline groundwaters. The quantitative significance of the observed Fe and Mn variations has been studied in WP4 by geochemical modelling, which shows that absolute and relative changes in Fe and Mn concentrations have non-unique interpretations in terms of redox (Eh) values. Samples of secondary calcite from the other UK study sites at Dounreay and Cloud Hill pose an analytical and interpretative problem because the amounts of late stage calcite are low. In both cases the calcite morphologies tend to be dominated by that of older secondary calcite on which late stage calcite has precipitated as a veneer. Uncertainty in calcite characterisation means that the significance of the substantial discordance in the depth locations of the transitions of morphology of late stage calcite and of salinity in the present-day groundwater profile at Dounreay is interpreted with much less confidence than for samples from Sellafield. Variations of redox-sensitive trace elements (Fe, Mn, Ce) are less systematic than at Sellafield although there is a general contrast between trace element contents of shallow and deep calcites with more variation at depth. However there is too much uncertainty and lack of reproducibility in these sparse localised data to interpret reliable palaeohydrogeological information. The few stable isotope data from Dounreay calcites tend to repeat the inference from isotope analyses at Sellafield that glacial or cold-climate water volumetrically replaced pre-existing water down to about 450-500 m depth since late stage calcites below this do not have such light δ18O values. In summary, Work Package 5 has found that there is a need to incorporate more palaeohydrogeological information into PA in order to improve the credibility of assumptions about stability of deep groundwaters and of estimates of the likely magnitudes of impacts of external changes in scenarios. The most appropriate way to do that is to use palaeohydrogeology to screen and quantify FEPs that are the basis for developing scenarios to be used in PA. Logical approaches to doing this have been illustrated, showing the considerable steps of data acquisition, interpretation and expert judgement that are involved in attempting to quantify information for transfer into FEPs and scenarios. This process of interpretation and expert judgement is usually carried out by means of a narrative assessment of the evidence. A more rigourous approach with ESL has been exemplified here but it is likely that the conventional narrative approach will continue to be used until much more comprehensive palaeohydrogeological evidence than that from geochemistry alone is available. Narrative interpretations have been abstracted from the geochemical and mineralogical investigations carried out on samples from the various study sites in WP2 of PADAMOT. The principal methods that have been deployed here, i.e. analyses of mineralogical, geochemical, isotopic and fluid inclusion characteristics of late stage secondary calcite, have been found to have a somewhat limited applicability that clearly depends primarily on the occurrence of late stage calcite in adequate abundance for the characterisation to be analytically feasible and reasonably reliable. Where these conditions obtain, as in the rocks at Sellafield and to a lesser degree at Äspö/Laxemar and Dounreay, a remarkable amount and diversity of data is obtainable. In other rock formations and for other groundwater conditions, e.g. in shallow groundwater environments, and at earlier stages of site investigations before comprehensive drillcore samples are available from deep boreholes, different sampling and analytical strategies are appropriate. Developments of appropriate strategies have been illustrated with the studies at Los Ratones and Melechov. Modelling - geochemical, hydrodynamic and coupled reaction-transport - is a valuable and sometimes essential tool for quantitative interpretation and for quantifying the associated uncertainties, as illustrated in WP4. The full significance of the geochemical and mineralogical indicators of palaeohydrogeology, i.e. of past physical and chemical groundwater conditions has not yet been fully realised even in the most advantageous studies. These studies have shown that uncertainties, assumptions involved in ‘expert judgement’, and calibrations of process models, remain as substantial sources of uncertainties. Nevertheless for some sites, calcite morphologies, redox-sensitive trace elements and stable isotope ratios can provide qualitative evidence of greater or lesser degrees of stability in past groundwater conditions. This evidence should be taken into account in considering whether deep groundwater conditions will be more or less stable in scenarios of future climate changes in safety cases.