The Geochemical Composition of Dutch Subsurface Sediments

G. Th. Klaver State Geological Survey, PO BOX 157 Haarlem, The Netherlands

D. J. Huisman Wageningen Univ., Dept. of Soil Science & Geology PO Box 37 6700AA,

Wageningen, The Netherlands

Knowledge about the chemical properties of subsurface sediments is becoming more and more important, because of the increasing use of the subsurface as source for drinking water, for waste disposal and for large infrastructural projects. The purpose of our study is to characterise the geochemical variation of the Dutch subsurface. Usually, geochemical characterisation are done by taking and analyzing samples from stream sediments or overbank deposits, and regard those as represenatative for a catchment area. In our situation this is not possible, as there is no network of eroding streams, and because in this way only information about the surface sediments can be obtained. For this study, we took samples from borings up to 100 meters depth, and anlyzed them for major- and trace elements on XRF.

As a pilot project, we made a characterisation of the geochemical variation of the subsurface in an area in the South of the Netherlands up to a depth of approximately
100 meters, using samples from a restricted number of borings. The sediments studied comprise marine, estuarine and fluviatile deposits from Miocene to Early-Pleistocene age. We found that the geochemical variation could be described as a function of sediment provenance, grain-size, and diagenetic processes.

In Pleistocene sediments with Rhine-provenance, the K/Al ratios of sand equals those of mica (muscovite) and illite, while clays display lower ratios due to mixing with other clay minerals. In sand there is a positve linear correlation between Na and Al as a result of the presence of plagioclase feldspars. This correlation is negative in clay because of the admixture of (Al-rich and Na-poor) clay minerals.

In sediments with Campine provenance two groups can be distinguished. One group shows similar relations between Al, K and Na contents as Rhine material. Another one has lower and more variable K/Al-ratios due to lower and more variable mica contents, and lower Na contents without correlation with Al due to the absence of plagioclase feldspar. Comparison of the heavy mineral compositions of both groups with source area sediments suggests that the former group consists of reworked Oligocene, the latter of reworked Eocene material from the Belgian Scheldt drainage area.

Reworked Tertiary weathering products in Pliocene sediments cause higher Ti/Al and Ti/Nb, and lower K/Al ratios when compared to Pleistocene Rhine sediments. In all
sediments, except glauconiferous Miocene deposits, the
relation between Al-content and median of the grain-size is log-negative. This holds for most of the other elements, except for Si, Na, As and S. In the Miocene deposits, the presence of coarse glauconite grains causes a positive linear correlation between K, Fe, As and most trace elements with the median of the grain-size.

As a result of diagenetic processes, local accumulations of Fe, Mn, S, As, Co, Cr, Ni, U, V and Zn occur, generally in relation to organic rich layers.

As a follow-up project, the spatial variation of geochemical properties in a single formation, the Kedichem formation, is being studied. The Kedichem formation is a series of fluviatile deposits, possibly with local marine influences, proceeding from Rhine, Meuse, and the former Campine and Baltic river systems. It is present in roughly half of the Netherlands, and serves a.o. as a source for drinking water. The variations in sediment provenance affect the detrital mineralogy and associated geochemical properties of the sediment, while variations in depositional environment affect the presence of diagenetic minerals and accumulations of specific elements.

Using fuzzy clustering, we made a spatial model of the sediment provenance, based on heavy mineral counts and lithological data from the Geological Survey and literature. It is possible to discriminate sediments from Rhine, Baltic and Campine origin on the basis of heavy mineral couts. Early Pleistocene sediments with Meuse provenance however, prove to be difficult to distinguish from Campine material. On the basis of this model, sample sites are chosen such that each provenance class is sampled sufficiently. The results of the geochemical analyses will be used to translate the provenance model into a 3-dimensional model of the geochemical variation in the Kedichem formation. Apart from that, paleoenvironmental reconstructions will be made which can be used to predict the presence and nature of diagenetic accumulations in these subsurface sediments.