Geochemical characterization of subsurface sediments in the Netherlands

Traditionally, the Netherlands' subsurface is mainly used to obtain good quality drinking and industrial waters from the different aquifers. Due to the lack of space on the surface, increasing environmental problems and demand for energy, the subsurface will be used increasingly for other activities, including large underground infrastructural projects, underground storage of waste and greenhousegasses and underground storage capacity for the energy sector.In order to evaluate the effects of the underground activities, detailed knowledge about the subsurface sediments is required. The geochemical composition of the subsurface sediments and the associated mineralogy forms an important part of the information needed to make decisions where and under what restrictions the different activities in the subsurface can best be planned. This thesis is a result of the GEOBON-project, which was started in order to meet this information need. The project studied the chemical composition of subsurface sediments by sampling and analyzing cores that were made by the Geological Survey of the Netherlands as is described concisely in Chapter 2.Chapter 3 present results from two study areas, Brabant and Limburg, which show how sediment source, sorting processes, depositional environment, grain size, weathering and syn- and post-sedimentary diagenesis affects the geochemical composition of Miocene, Pliocene and Lower Pleistocene sediments in the Southern Netherlands:Chapter 3.1 contains the results of a sediment geochemical study that was performed in unconsolidated Upper Cenozoic sediments from the South of the Netherlands. Glauconite-rich sediments (Breda Formation) show anomalously high K contents and low Ba/K ratios. Major shifts in sediment composition as a result of changes in the Rijn system and shifts between Rijn and Schelde provenance as known from heavy-mineral studies are recorded in changes in the grain size dependent variations between Al, Na and K: Pleistocene Rijn sediments (Tegelen Formation) show higher Na contents than Pliocene Rijn sediments (Oosterhout and Kiezeloöliet Formations) and Schelde-derived material (Kedichem Formation), probably as a result of larger contents of sodic plagioclase. Schelde-derived sediments show low K/Al ratios as result of a smectite-dominated clay mineralogy and low contents of micas, whereas Rijn-derived sediments have high K/Al ratios which reflect an illite-kaolinite dominated clay mineralogy and higher contents of muscovite.The presence of siderite causes high Fe-contents in the Tegelen formation in the east of the study area. Increased Mg contents in the siderite-bearing sections from the Tegelen formation and in parts of the Oosterhout and Kiezeloöliet Formation are probably caused by the presence of minor amounts of dolomite. Localized high (pyrite-) S-concentrations are not only found in the marine-estuarine Oosterhout and Tegelen Formations but also in the fluviatile Kiezeloöliet and Kedichem Formations, which indicates at least minor marine transgressions during their deposition.Chapter 3.2 studies the change from a stable to an unstable heavy mineralogy in the composition of Rijn-derived sediments at the Pliocene-Pleistocene transition. This change has previously been attributed to a decrease in weathering intensity due to climatic cooling, and to a change in the Rijn sediment provenance from local to Alpine-derived. We studied the geochemistry of several sections with Pliocene and Early Pleistocene Rijn deposits, and one section (BTAB) in more detail using clay mineralogical and micromorphological techniques to study the exact nature and the cause of this change, and associated changes in sedimentary setting. We found a general increase in Na 2 O- contents at the local to Alpine provenance shift, which can be attributed to the Alpine source supplying fresh, sodic plagioclase-rich material instead of the local, strongly weathered sediments. There is a general trend of increasing K 2 O/Al 2 O 3 from the Pliocene to the Early Pleistocene that can be attributed to a similar decrease in degree of weathering.However, this trend is disturbed by the loss of K from clay minerals during post-depositional in situ weathering in organic-rich layers. In the Upper Pliocene BTAB section, we found a clear transition from kaolinite and high TiO 2 /Al 2 O 3 -ratios to smectite-rich material with lower TiO 2 /Al 2 O 3 that coincides with the local to Alpine provenance shift. However, Early Pleistocene sediments have TiO 2 /Al 2 O 3 -ratios that are similar to the ones before the transition so this effect is not consistent. Local high TiO 2 -anomalies, caused by preferential sorting and concentration of especially rutile in placer-like deposits occur in most Pliocene sections, but they are absent in the Upper Pliocene and Lower Pleistocene Alpine-derived deposits. Overall, the detrital geochemical variation in these deposits are primarily controlled by the source of the sediment, and hence the large-scale tectonic setting, whereas climatic control is limited.The Pliocene organic-rich layers were originally formed in a fresh-water fluviatile environment. Nevertheless they show high concentrations of S due the presence of abundant pyrite as a result of inundation by saline water during short-termed flooding events like spring tides or storm floods, after or alternating with one or more desiccation phases.Chapter 4 focuses on the effects of specific diagenetic geochemical processes that cause enrichments or depletions of certain chemical elements:Chapter 4.1 describes the formation of siderite (FeCO 3 ), which is often used as an indicator for the depositional environment of sediments, as it can only be formed under restricted geochemical conditions. By doing so, the possibility that siderite was formed later, under altered circumstances is often neglected. In this study, siderite in Early and Late Pleistocene deposits is investigated to establish whether it was formed syndepositional, or postdepositional under different circumstances. Within the Early Pleistocene Tegelen Formation, siderite is found as coatings around detrital dolomite grains, together with partially dissolved detrital calcite grains. Siderite also occurs as single nodules and homogeneously mixed through the groundmass. Siderite precipitation at the expense of calcite is proposed, which is likely to be an actual process as the current groundwater composition agrees well with the thermodynamics necessary for this process. Siderite formation in Late Pleistocene river fens is syndepositional and associated to both calcite and vivianite.A comparison between these two settings shows that siderite may be an indicator for the environment during deposition, but the presence of siderite in sediments may also be the result of interaction with groundwater under diagenetically altered circumstances.In chapter 4.2, we focus on the effects of multiple cycles of interchanging of fresh and saline water and oxic and reduced diagenetic environments that occur during the formation of low-gradient deltas of large river systems, driven by the glacial- to interglacial climatic cycle. Because of the reactivity of organic matter, the geochemical characteristics of organic-rich sediments can be strongly affected by these changing diagenetic environments.In the geochemical/sedimentary record of the Pliocene and Early Pleistocene Rijn-Maas delta, organic-rich layers display anomalous enrichments of trace elements, heavy metals and rare earth elements (REE). During marine highstands, high groundwater levels cause Fe-hydroxides to become reduced, thus releasing Fe and associated trace elements into the groundwater. At the same time, transgressions of saline water over organic surface layers and saline groundwater intrusions cause the formation of pyrite and other sulfides that may contain elevated levels of As and Mo and, depending on the Fe-source, Co, Ni, Pb and Zn. Associated with the pyrite-forming reduced environment, Y, REE, Cr, V and U are immobilized and accumulate. During lowstands, lower groundwater levels cause part of the sulfides to oxidize, except for that are present in the organic layers, where they are protected from oxidation. Fe-oxides form, and trace elements like As, Co, Ni, Pb, Zn, Y and REE are incorporated, only to form a source for Fe and trace elements during the next reduced phase.Chapter 5 describes the natural variation in heavy metal contents of subsurface sediments in the Southern Netherlands. The detrital heavy metal contents of these sediments show linear correlations with Al as a result of their joint occurrence in phyllosilicates. Anomalous enrichments occur as a result of the presence of glauconite (As, Cr, Ni, Pb, Zn), pyrite (As) or Fe-oxides (As, Ba, Ni, Zn), due to the interaction of organic-rich subsurface material with groundwater (Co, Ni, Zn) or as a result of anthropogenic pollution in topsoils (Cu, Pb, Zn). The contents of Al, Fe, K and S are well suited to determine background values, and to identify the cause for anomalous accumulations of heavy metals.Chapter 6 describes the result of a geochemical mapping campaign in an Early Pleistocene fluviatile formation (Kedichem) in the Netherlands. This is the first step towards a nation-wide geochemical database that can be used to meet the demand for information about the composition of subsurface sediments. We first determined the spatial extension and thickness of the sediment body. Subsequently, we used Fuzzy clustering techniques on approximately 2000 heavy-mineral counts from the NITG-TNO database to map the spatial extension of the Schelde, Rijn and Baltic sediment provenances within the formation.Geochemical data was collected during a sampling campaign in which about 600 samples from the Kedichem formation were analyzed. We used factor analysis to determine the major factors that determine the geochemical composition. These factors include clay content, presence of carbonates, pyrites and Fe,Mn-hydroxides, sodic plagioclase and zircon, and organic-matter-related diagenetic processes. We tested which lithological data from the NITG-TNO boring database is correlated the geochemical composition and therefore can be used to make a geochemical prediction model. We found that the classes Sand, Clay + Gyttja and Peat are significantly different and therefore can be used to predict the contents of Si, Ti, Al, Fe, Mn, Mg, Ca, Cr, Cu, Pb, V, Zn, Ba, Ga, Nb, Rb, Sr and Y. For the elements As, Ni, U and S, the classes Organic-poor Clay, Sand+Peat, Gyttja+Organic-rich Clay are significantly different. For Na and K, a division can be made in to Mica-rich Clay, Mica-poor Clay, Mica-rich Sand, Mica-poor Sand and Peat + Gyttja. By classifying the lithological data from the NITG-TNO core description, we made a geochemical model to predict the geochemical composition in the Kedichem formationWe visualized this model by calculating and interpolating the average composition of 5 m horizontal slices of the Kedichem formation. The model performance is fairly good, although it has a tendency to underestimate extreme values.The results of our study demonstrate that geochemical characterization of sediments can be performed by doing a large number of low-cost XRF analyses, supported with a limited amount of XRD-, ICP-MS, SEM and micromorphological analyses. The geochemical variation can be determined and the speciation and hence the reactivity and conditionally availability of harmful elements can be deduced. The geochemical methods employed in this study can not only be used to study sedimentation history and stratigraphy, but they can also yield important information about the composition and reactivity of subsurface sediments which can be used for ground water quality management and evaluation of underground activities.