A procedure to define the good chemical status of groundwater bodies in Germany PDF
Preview A procedure to define the good chemical status of groundwater bodies in Germany
A PROCEDURE TO DEFINE THE GOOD CHEMICAL STATUS OF GROUNDWATER BODIES IN GERMANY R. Kunkel*, S. Hannappel**, H.J. Voigt****, F. Wendland* & R. Wolter***** * Research Centre Jülich, Systems Analysis and Technology Evaluation (STE), Jülich, Germany ** HYDOR Consult GmbH, Berlin, Germany *** Dept. of Environmental Geology, Univ. of Technology, Cottbus, Germany **** Federal Environmental Agency (UBA), Berlin, Germany ABSTRACT Commissioned by Germany’s Working Group of the Federal States on Water Problems (LAWA) the authors developed a procedure to define natural groundwater conditions from groundwater monitoring data. The distribution pattern of a specific groundwater parameter observed by a number of groundwater monitoring stations within a petrographical comparable groundwater typology is reproduced by two statistical distribution functions, representing the “natural” and “influenced” component. The range of natural groundwater concentrations is characterized by confidence intervals of the distribution function of the natural component. The applicability of the approach was established for 17 hydro chemical different groundwater typologies occurring throughout Germany. Based on groundwater monitoring data from ca. 26000 groundwater-monitoring stations, 40 different hydro chemical parameters were evaluated for each groundwater typology. For all investigated parameters the range of natural groundwater concentrations have been identified. According to the requirements of the EC Water Framework Directive (article 17) (WFD) this study is a basis for the German position to propose criteria for assessing a reference state for a “good groundwater chemical status”. KEYWORDS Good groundwater chemical status, background values, groundwater typologies, EC Water Framework Directive, 2 R. Kunkel, S. Hannappel, H.J. Voigt, F. Wendland, R. Wolter 1 INTRODUCTION The solution content of groundwater is determined by a variety of factors, such as the properties of the vadoze zone and the groundwater bearing rocks as well as the hydrological and hydrodynamic conditions. Additional to these “natural” factors groundwater quality is influenced by anthropogenic influences, e.g. land cover changes, diffuse intakes from agriculture and the atmosphere, point source intakes (see fig.1). local influences regional influences geogenic, hydrologic, geogenic, hydrologic, anthropogenic anthropogenic pedologic, biologic pedologic, biologic influences influences influences influences influenced component natural component groundwater qualitiy Figure 1: Factors influencing groundwater quality (LfU, 1996, modified) Whereas the occurrence some groundwater parameters (e.g. pesticides) are direct indicators of human impact, most inorganic contents may originate both from natural and anthropogenic sources. This makes it difficult to decide whether an observed groundwater condition is reflecting the “good groundwater chemical status” according to the requirements of the WFD or not, which would imply that measures to decontaminate the groundwater have to be drawn. Because of the omnipresence of human impacts strictly „natural“ groundwater occurs at best at regionally limited locations. Especially groundwater from aquifers taking part in the active water cycle (surface- near aquifers), in most cases those aquifers which were used for water supply, are influenced since decades and centuries by anthropogenic activities, e.g. soil cover induced changes of percolation water quality or fertilizer inputs. In Germany this situation is true for more than 99% of the federal territory. Consequently, the present solutions content of groundwater samples from surface-near aquifers do hardly represent A procedure to define the good chemical status of groundwater bodies 3 strictly “natural” groundwater concentrations. Deriving natural groundwater concentrations from samples from unaffected deeper aquifers, however, would be of no relevance for practical water resources management, as it reflects “untypical” groundwater conditions. Against this background a more pragmatically understanding of the term ”natural groundwater condition”, which considers the human impacts on groundwater to a certain degree as inevitable, needs to be developed. This is done by Schenk (2003), who considers natural groundwater concentrations to be present, if “the concentrations of the most important kations and anions originate from not significant anthropogenic influenced soils and the rocks of a watershed, including groundwater from areas under agricultural use or from areas where land cover changes occurred over the last centuries”. Because of the high importance of groundwater protection a complementary Groundwater Directive is currently developed. In this context the EU member states are invited to propose criteria for assessing a reference state for a “good groundwater chemical status”. Commissioned by the Working Group of the Federal States of Germany on water problems (LAWA) the authors developed a procedure to define the “good groundwater chemical status” in Germany based on existing data from groundwater monitoring networks provided by the Federal States. 2 DATABASE The definition of natural groundwater quality by Schenk was the starting point for this research project, which has been commissioned by the Working Group of the Federal States of Germany on Water Problems (LAWA). The aim of this project was to provide data for the definition of “good groundwater chemical status” from existing data from groundwater monitoring networks provided by the Federal States. As a reference17 lithogenetical different groundwater bodies occurring throughout Germany with high relevance for water supply (e.g. limestones, sandstones, Loose-rock sediments) were investigated Before assessing natural groundwater concentrations the individual heterogeneous data sets provided by the Federal States needed to be joined together in one database with unified structure and referenced to the groundwater typologies. In addition, a number of consistency checks, e.g. the elimination of analyses with incorrect ion balances, salt effected stations and elimination of time series by median averaging had to be performed. At the end, a total data from ca. 25000 monitoring stations with one representative groundwater analysis each were used for the analysis. For each of the different groundwater bodies 30 - 40 hydro chemical (inorganic) parameters were evaluated including red ox parameters, summary parameters, main substances, adjoining substances and trace 4 R. Kunkel, S. Hannappel, H.J. Voigt, F. Wendland, R. Wolter elements. Figure 2 shows the total number of observations available for each parameter in the individual groundwater typologies. Figure 2: Regional distribution of evaluated groundwater-monitoring stations The number of evaluated observations for each parameter in each typology was in the range between 5000 and about 25000, allowing a statically relevant analysis (see figure 3). A procedure to define the good chemical status of groundwater bodies 5 Figure 3: Evaluated groundwater parameters for the assessment of natural groundwater concentrations in Germany 3 METHOD It can be expected that the solute content of a typical groundwater is affected both by natural and anthropogenic factors. Anthropogenic inputs of a certain substance of content into groundwater, will lead to an increase of groundwater concentration of this substance. Because of the variability both of natural and anthropogenic influences it is not possible to separate these two influencing factors by just looking at one single groundwater sample (Schleyer & Kerndorff, 1992). Therefore it is necessary to analyze a large number of groundwater samples statistically, with the condition that these samples were taken from aquifers, which can be regarded as being almost hydro chemically homogenous. This is the case for the groundwater typologies investigated here. For each of the investigated parameters the concentration distribution in the individual groundwater bodies is separated into a natural and an influenced component. This is done by representation of the observed concentration distribution by the sum of two statistical distribution functions, describing the natural and the influenced component. The good groundwater chemical status is characterized by confidence intervals of the distribution function of the natural component. Starting point of the analysis is the frequency distribution of the observed concentrations of a groundwater parameter, illustrated as black dots in figure 4. 6 R. Kunkel, S. Hannappel, H.J. Voigt, F. Wendland, R. Wolter observed concentrations (classified) sum of calculated natural and influenced components y c n e u q natural component e r f influenced component concentration Figure 4: Basic approach of separating the natural and influenced component from an observed groundwater concentration pattern. If the database contains both natural and influenced samples, the concentration profile can be expressed by the superimposition of these two components. In that case the observed concentration distribution (f ) obs may be described by the sum of two statistical distribution functions, representing the natural (f ) and the influenced (f ) component: nat infl ( ) ( ) ( ) f c = f c +f c (1) obs nat infl The two distribution functions are not known from a priori. An analysis of the observed concentration patterns using statistical tests show that the component distribution functions can be expressed by lognormal distributions. The explicit shape of both distribution functions is determined by three independent parameters each (amplitude, median and variance), which have to be fitted to the observed frequency distribution using standard algorithms. As a result, the observed distribution pattern is represented by two distribution functions of known shape, which can be assigned to the natural and anthropogenic component. Specification of natural groundwater conditions can be done by any two independent parameters characterizing the distribution function of the natural component. For the lognormal distribution, the median and the variance is the most common parameters given. However, these parameters will give not a very transparent measure to specify typical groundwater conditions. Therefore, natural groundwater concentrations are characterized by a concentration range defined by the 10% and the A procedure to define the good chemical status of groundwater bodies 7 90% percentiles of the concentration distribution of the natural component. 4 SELECTED RESULTS According to the procedure described above, the ranges of natural groundwater concentrations in the groundwater typologies occurring in Germany were identified for all 30 (-40) parameters. In the following sections selected results for different parameters and different groundwater typologies are presented and discussed. 4.1 Potassium In The Upper Rhine Valley As an example, figure 5 shows the potassium concentrations in the Upper Rhine Valley. The observed distribution shows of one peak at a concentration level of ~2 mg K/l. The width of the distributions, however, indicates that also potassium concentrations above 15 mg K/l can occur. As can be seen from the correlation coefficients (r2 ~0.99), the observed concentration distributions can be separated very satisfying into a natural and influenced component. As a typical behavior, obtained also for most other evaluated hydro chemical parameters (e.g. Na, Cl, SO4, HCO3, DOC), the natural component is dominating while the influenced component contributes as a broad background. The natural groundwater component of the potassium concentration distribution in the Upper Rhine Valley shows a relatively small scatter in the range between 0.6 and 5 mg K/l. This can be addressed to the natural background of dissolved potassium in this hydrogeological unit from containing minerals. An influenced component is obvious in a broad, high concentration range. The portion of the influenced component is relatively high. Here two natural factors, i.e. the rise of the high-mineralized waters from deeper aquifers and the influence of potassium containing fertilizers are determinative. In the case of the Upper Rhine regional water authorities have confirmed both influences. 8 R. Kunkel, S. Hannappel, H.J. Voigt, F. Wendland, R. Wolter Figure 5: Observed frequency distributions as well as the natural and influenced components of the potassium concentrations in Upper Rhine Valley . 4.2 Magnesium concentrations in Igneous / Metamorphic rocks and Jurassic limestone The observed magnesium concentrations in the hydrogeological units Igneous / Metamorphic rocks show a different behavior compared to the observed magnesium concentrations in the unit Jurassic limestone (see figure 6). Figure 6: Observed frequency distributions as well as the natural and influenced components of the magnesium concentrations in the Magma tic and metamorphic rocks (left) and the Jurassic limestones (right). The observed magnesium contents in the Igneous / Metamorphic rocks show the same type of distribution which had already been discussed for the potassium concentrations, i.e. the natural component dominates, while the influenced component contributes as a broad background. Thus the Mg-concentrations do not exceed 14 mg Mg/l and show a predominance of values below 5 mg Mg/l. This reflects the lack of magnesium- containing minerals and/or short residence times of the groundwater in the aquifers. The second component can be regarded as the impact of an anthropogenic pressure, i.e. magnesium-containing fertilizers or liming of buffer poor mountain soils in order to avoid acidification. The Mg-concentrations in the Jurassic limestone are higher than the one shown for the Ingenious / Metamorphic rocks and reflect the occurrence of Mg–rich minerals (Dolomites). The frequency distribution shows a bimodality, which is in general typical for inhomogeneous data sets, i.e. groundwater analysis from two spectrographic and/or hydrodynamic different hydrogeological units. In the case of the Jurassic limestone however this represents a facies sequence, i.e. a rising degree of dolomitisation in the Jurassic limestone from SW to NE. A procedure to define the good chemical status of groundwater bodies 9 4.3 Iron (II) in Triassic Sandstones and Pleistocene Gravel and Sands Figure 7 shows the iron (II) concentrations in the Triassic sandstones and loose rock Sediments of the North German plane. Compared to the potassium concentrations shown in figure 4 the iron concentrations show a very different behavior. Most of the samples display very small iron concentrations in the lowest category. This distribution pattern is typical for parameters, which were almost completely removed from groundwater e.g. by red ox-reactions. The same distribution type is also observed for other hydro chemical parameters like Mn and NH4 as well as NO3 in the loose rock sediments of the North German plane (see chapter 4.4). Figure 7: Observed frequency distributions as well as the natural and influenced components of the iron (II) concentrations in the Triassic sandstones (left) and the loose rock sediments of the North German plane (right). The iron concentrations in the Triassic sandstones (figure 7 left) show a monotonous decay of the frequencies towards higher concentrations. This behavior is typical for an oxidized groundwater type, which is almost free (P <0.1 mg Fe/l) of solved iron (II). Although this observation can be 90 reproduced by a lognormal distribution, a second component representing an influenced component cannot be identified. This indicates that the iron concentrations in the groundwater are hardly influenced by anthropogenic inputs and/or geogen anomalies. Although the iron concentrations in the loose rock Sediments of the North German plane (figure 7 right) also show a monotonous decay, the observed iron concentrations are much larger than those in the Triassic sandstones. This is due to different chemical situations. In these loose rock sediments two components; both following a lognormal distribution can be identified. However, none of the two components can be addressed directly to anthropogenic influences. The first component, having its maximum in the lowest concentration class, arises from oxidized groundwater zones in preferably shallow and unconfined aquifers, where almost no solved iron (II) is present. The second component, which has its maximum around 1 mg Fe/l characterizes the situation typical for reduced groundwater conditions well known occurring for loose rock sediments. It is not allowed to classify one of these two components as being influenced 10 R. Kunkel, S. Hannappel, H.J. Voigt, F. Wendland, R. Wolter and exclude it from the further assessment of natural iron concentrations. Instead, two groundwater types occur in parallel in the loose rock sediments of the North German plane, resulting in a broad concentration range for the natural groundwater concentration between 0.5 and 10 mg Fe/l. This example shows, in a more general sense, that it is urgently necessary to assess natural groundwater concentrations according to the petrography and the hydrodynamics of the groundwater typologies. Consequently, iron concentrations of 3 mg Fe/l can be considered as being “natural” in the case of the loose rock aquifers in the North German Plane aquifers. In the case of the hard rock groundwater units 3 mg Fe/l would have to be considered as “influenced”. 4.4 Nitrate in Jurassic Limestones and Pleistocene Gravels and Sands Compared to the other groundwater parameters evaluated in this study, nitrate is an exception. The geogen fraction of nitrate in groundwater is low. The reason for this is that all nitrate minerals are very water-soluble, so that in the course of geological history no nitrate rocks were formed from which nitrate circulating water could solve. The most important source of nitrate in groundwater are fertilizers and to a minor extent atmospheric nitrate inputs. Many studies have shown, that nitrate concentrations above 5 mg NO /l are a reliable indication of 3 anthropogenic influences. Figure 8 shows the observed nitrate concentrations distributions in the Jurassic limestone and the loose rock sediments of the North German plane. The nitrate concentrations in the Jurassic limestone cover a wide range between 0 and about 80 mg NO /l. Two widespread components 3 may be identified. Even the natural display relatively large concentrations, which is due to the presence of typically oxidized groundwater conditions, The second, influenced, component dominates the observed concentration pattern and reaches to concentrations up to 75 mg NO /l. This behavior 3 can be addressed to the presence of groundwater conditions, which are mostly influenced be nitrate inputs from agriculture. Nevertheless, it can be concluded that natural nitrate concentration below about 20 mg NO /l 3 can be regarded as being natural for this groundwater typology.
The list of books you might like
