- Email: [email protected]
Derivation of natural background levels and threshold values for groundwater bodies in the Upper Rhine Valley (France, Switzerland and Germany)
Desalination 226 (2008) 160–168
Derivation of natural background levels and threshold values for groundwater bodies in the Upper Rhine Valley (France, Switzerland and Germany) F. Wendlanda*, G. Bertholdb, A. Blumc, P. Elsassc, J.-G. Fritscheb, R. Kunkel a, R. Wolterd a
Agrosphere, ICG-IV, Forschungszentrum Jülich GmbH, Jülich, Germany Tel. +49 2461 613165; Fax +49 2461 612518; email: [email protected] b Hessian Agency for Environment and Geology (HLUG), Wiesbaden, Germany c Bureau de Recherches Géologiques et Minières (BRGM), Orléans, France d Federal Environmental Agency (UBA), Dessau, Germany Received 20 December 2006; revised accepted 30 January 2007
Abstract According to the requirements of the EU Water Framework Directive (EU-WFD, article 17.2a) criteria for the assessment of the chemical status of groundwater have to be developed, which may serve as starting points for a trend reversal (article 17.2b) in case the good status is failed. Against this background an EU-wide applicable approach to assess natural background levels (NBLs) and threshold values (TVs) for the definition of the groundwater chemical status has been developed. NBLs are derived based on a preselection method using aquifer typologies as natural reference systems. Samples with purely anthropogenic substances (e.g. PAC, pesticides) as well as samples, in which indicator substances for anthropogenic inputs (e.g. nitrate) are exceeding a certain value, are excluded. The NBLs are defined for the remaining groundwater samples of the aquifer typology under concern as the concentration range between the 10% and 90% percentiles. The threshold values (TVs) get established with reference to the NBLs and a “not acceptable reference value” (REF), e.g. drinking water standards or toxicological values, guaranteeing that the TVs are higher than the NBLs, but lower than the REF values. The applicability of the methodology for NBL and TV derivation has been applied to the transboundary groundwater body of the Upper Rhine Valley in Germany, Switzerland and France. Keywords: EU Water Framework Directive (article 17); Groundwater quality; Natural background levels; Threshold values, Upper Rhine Valley
*Corresponding author. Presented at the 10 th IWA International Specialized Conference on Diffuse Pollution and Sustainable Basin Management, Istanbul, Turkey, 18–22 September 2006. 0011-9164/06/$– See front matter © 2008 Published by Elsevier B.V. doi:10.1016/j.desal.2007.01.240
F. Wendland et al. / Desalination 226 (2008) 160–168
1. Introduction The Commission proposal of Groundwater Directive COM(2003)550 developed under Article 17 of the Water Framework Directive (WFD) sets out criteria for the assessment of the chemical status of groundwater, which is based on existing Community quality standards (nitrates, pesticides and biocides) and on the requirement for Member States to identify pollutants and threshold values (TVs) that are representative of groundwater bodies found as being at risk, in accordance with the analysis of pressures and impacts carried out under the WFD. In this context groundwater– background values are required as a reference to quantitatively evaluate whether or not groundwater is significantly modified by anthropogenic influences [1]. However, the term “background level” is used in a number of different senses, e.g. in the context of “baseline concentration” [1,2], “geochemical baseline” [3], “background concentration” [4], “ambient background” [5]. All these terms reflect a different understanding or definition of what is regarded as being “natural” and/or “not natural” and are all connected to a specific location and scale. Because of the omnipresence of human impacts in the case study area strictly “natural” groundwater occurs at best at regionally limited locations. Consequently, the present solutions content of groundwater samples from surface-near aquifers do hardly represent strictly “natural” groundwater concentrations. Against this background a more pragmatically understanding of the term “natural background level” needs to be used, which considers the human impacts on groundwater to a certain degree as inevitable. Therefore, we define “natural background level” as “the concentration of a given element, species or chemical substance present in solution which is derived by not significant anthropogenic influenced processes from geological, biological or atmospheric sources”. In the light of the above, the EU-Specific Targeted Research Project BRIDGE (Background
161
cRiteria for the IDentification of Groundwater thrEsholds) was initiated, involving 27 partners from 19 countries and links efficiently stakeholders of the scientific and policy-making communities. A scientifically based and generally applicable approach to derive natural background levels for the groundwater and groundwater threshold values was derived. The applicability and validity of this approach is checked in 14 case study areas at the river basin district or groundwater body level throughout Europe, including the Upper Rhine Valley as a transboundary French–German–Suisse case study.
2. Aquifer typologies for hydrogeochemical characterisation 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 hydrodynamical conditions. Apart from these “natural” factors groundwater quality is influenced by anthropogenic inputs mainly from diffuse sources (e.g. agriculture, atmosphere) [6]. Whereas some of these inputs (e.g. pesticides) are a direct indicator of human impacts, most inorganic contents occurring in the groundwater originate both from natural and anthropogenic sources [7]. This makes it difficult to decide whether an observed groundwater concentration pattern in a certain area is influenced by pollution intakes or still represents an (almost) natural state. In particular, groundwater from aquifers taking part in the active water cycle (surface-near aquifers), in most cases aquifers also 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. For a European wide assessment, an integration of aquifer typologies with respect to its typical groundwater conditions is most needed. For this
162
F. Wendland et al. / Desalination 226 (2008) 160–168
purpose the complexity of individual groundwater bearing rocks (aquifers, groundwater bodies) was classified into simplified main aquifer typologies. This is done mainly on the basis of petrography because the groundwater of aquifers with similar petrographic properties shows very similar natural composition in comparable hydrodynamic and hydrologic environments [8]. In order to provide both a practical framework for an overview about major groundwater composition patterns in Europe and for regional referenciation of natural background (NBLs) and threshold (TVs) for the groundwater in Europe, an aquifer typology map for hydrochemical characterisation has been created within the BRIDGE project. This map is presented in Fig. 1. The location of the case study area is indicated on the map.
3. The case study area Upper Rhine basin The case study area Upper Rhine Valley is a transboundary river basin located between France and Germany with smaller parts in Switzerland. It belongs as a whole to the aquifer typology “fluviatile deposits of major streams” (see Fig. 2, left part). The total area comprises 9.290 km2. It represents a very important drinking water reservoir with high importance for supraregional water supply. Due to the high population density a variety of anthropogenic impacts on groundwater quality (intensive agriculture, industry, water withdrawal) are present. With respect to the BRIDGE aquifer typologies, the Upper Rhine Valley is part of the typology “fluviatile deposits of major streams”. Hydrogeologically, the Upper Rhine Valley represents one of the biggest rift structures in Europe filled up with alternating layers of slit, clay, sand and gravel during the Pliocene and Quaternary period (see cross section, Fig. 2, right part). The hydraulic conductivity of the aquifers decreases from south to north and is associated with a decrease in the grain size of the sediments along the flow direction [9].
4. General applicable preselection approach for NBL derivation An evaluation of existing approaches for NBL assessment [10–16] has shown that preselection methods are appropriate to derive natural background levels, especially for large area considerations. Furthermore, they fulfill the requirements to be applicable on a European scale: they require “no deeper knowledge” about statistical analysis, can be applied by non-experts and to groundwater bodies, for which only few samples are available. By giving a general overview of the natural chemical composition of the groundwater bodies, they fit into the WFD requirements and with WFD working scale. The basic idea of preselection method is that there is a correlation between the concentration of certain indicator substances and the presence of anthropogenic influences. The NBLs are derived only from groundwater samples, where the concentrations of these indicator substances are below a defined level, indicating that the samples can be regarded as not being influenced by anthropogenic activities. Although preselection methods have the disadvantage that the relation between high concentration of indicator substances and anthropogenic induced high concentrations of other substances is not true in all cases [11], this kind of approach is particularly representative and interesting for elements for which a considerable number of samples are available. Typical indicator substances are either anthropogenic substances (e.g. PAC, pesticides) or substances occurring naturally only in low concentrations (nitrate, heavy metals, etc.). An ideal indicator substance for groundwater systems should be easily detectable and typical for a certain type of pollution (i.e. diffuse or point source pollution). As inorganic groundwater substances may occur in the groundwater naturally as well as due to anthropogenic inputs, there are only a few substances, whose presence in high concentrations surely indicates anthropogenic inputs. Nitrate and
Fig. 1. BRIDGE aquifer typologies for hydrochemical characterization.
F. Wendland et al. / Desalination 226 (2008) 160–168 163
Fig. 2. Geographical location and cross section of the case study area “Upper Rhine Valley”.
164 F. Wendland et al. / Desalination 226 (2008) 160–168
F. Wendland et al. / Desalination 226 (2008) 160–168
ammonia are examples for such substances. In most cases high nitrate concentrations coincide with low ammonia concentrations and vice versa, so that they are acting like a supplementary indicator set for anthropogenic pressure to groundwater. Whereas nitrate is a good indicator for pollution in oxidized groundwater, it may be absent in reduced groundwater even in case of high inputs due to denitrification processes (e.g. [16]). In that case ammonia and potassium may serve as an indicator for anthropogenic pressure to groundwater in reduced groundwater bodies instead. The following criteria to preselect the groundwater samples are based on the experiences in several French and German studies on NBL assessment [11–15]: • Exclusion of groundwater samples displaying purely anthropogenic substances, • Exclusion of samples displaying concentrations of indicator substances exceeding a certain value: – NO3 > 10 mg/L in oxidized aquifers (O2 > 2 mg/L and Fe(II) < 0.2 mg/L) or – NH4 > 0.5 mg/L in reduced aquifers (O2 < 2 mg/L and Fe(II) > 0.2 mg/L). The natural background values are defined for all groundwater parameters as the 90th percentiles (P90) of the concentration distributions from the remaining samples. For substances that are purely synthetic with no natural sources (for example, TCE), NBLs are set to zero. 5. Derivation of threshold values for the groundwater The threshold values (TVs) get established with reference to the NBLs and a chosen reference standard (REF). The latter may either be a drinking water standard value (DWS), an environmental quality standard or ecotoxicological value (EQS
165
Fig. 3. Illustration of the approach to derive threshold values (TV).
or EToxV). The calculation of threshold values is presented for 3 cases (see Fig. 3). • Case 1: NBL ≤ REF: If the NBL of a certain substance is below the REF value the TV is set as the concentration in the middle between the NBL and the REF values: TV = (REF + NBL)/2. • Case 2: NBL < one third of REF: In case the NBL is considerably below the REF-value, the TV is limited to twice the NBL: TV = 2 NBL. • Case 3: NBL ≥ REF: In case the NBL is larger than the REF value, the TV is set equal to REF itself: TV = NBL. The TV for purely synthetic substances with no natural sources (e.g. TCE) is defined as the detection limit.
166
F. Wendland et al. / Desalination 226 (2008) 160–168
6. Application to the case study area Upper Rhine Valley The data base for evaluating NBL and threshold values consists of almost 1700 groundwater samples for the years 2002 and 2003 from French, German and Suisse monitoring networks (see Fig. 4). For each of the monitoring stations one sample containing the solution contents for integral and chemical environment parameters, characteristic major and minor parameters and WFD pollutants was available. However, no information on heavy metals was contained in the data base. The preselection approach described above was applied to the data base for the Upper Rhine Valley. Because of missing information concerning purely anthropogenic substances, no groundwater samples were excluded due to this condition.
Preselection according to Nitrate (NO3 < 10 mg/L) has lead to the exclusion of 1094 samples (64%). Evaluation of the redox status of groundwater has shown that 188 samples indicate reduced aquifer conditions. 35 of those samples were excluded from the NBL derivation because of their high NH4-contents (NH4 > 0.5 mg/L). In the end 594 groundwater samples (see figure, left side) remained. For those samples it was assumed that they are to a large extent free from human impact. Hence, they are regarded as being appropriate for the NBL derivation in the Upper Rhine Valley. The right part of Fig. 4 shows the derived NBL and TV for 14 substances, for which NBLs have been defined and REF values were available. Depending on the NBL very different TV are derived. As the last column in the table in Fig. 4 indicates, that case 1 (see Fig. 3) is typical
Parameter
Unit
TV1
Case
mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L µS/cm
P90 0.1 84 3.6 7.2 25 0.82 41 173 951
Ref
B Cl Fe(II) K Mg Mn(II) Na SO4 LF
1 250 0.2 10 50 0.05 200 250 2500
0.2 167 3.6 8.6 37 0.8 83 211 1726
2 1 3 1 1 3 2 1 1
As NH4 NO2 NO3 PO4
µg/L mg/L mg/L mg/L mg/L
4 0.39 0.04 8.2 0.17
10 0.5 0.5 50 6.7
7 0.45 0.08 16.4 0.34
1 1 2 2 2
Fig. 4. Geographical location of sampling sites in the Upper Rhine Valley (left) and NBL- as well as TV-values (right).
F. Wendland et al. / Desalination 226 (2008) 160–168
for TV derivation for most of the investigated parameters (Cl, K, Mg, SO4, LF, As, NH4), i.e. the TV was set to half of the difference between NBL and the REF. For B, Na, NO2, NO3 and PO4, however, parameters, for which the NBL is significantly below the REF value, case 2 was applied. Only for two parameters Mn(II) and Fe(II), the NBL is above the REF value, so that the NBL was regarded to be equal to the TV. After data preselection 594 samples have been used to derive the NBL and TV. However, it would be interesting to compare these values to the situation assessed on the basis of the total data set containing about 1700 groundwater samples. An analysis shows that about 90–95% of the 1700 available samples display concentrations below the derived threshold values. In these cases a good status of groundwater can be postulated and no further measures to improve groundwater quality are required. However, 5–10% of the samples exceed the TV. For these cases the reasons for this exceedence need to be checked. The impacts of natural influences, like salinisation, redox-conditions, hydrodynamics, which may lead to natural (geogenic) anomalies, as well as of anthropogenic influences like diffuse and point source pollution must be assessed. If the reasons of exceeding the TV are due to natural influences, the “good status of groundwater” is achieved even in case the TV is exceeded as it can be explained by natural processes. Only in case the reasons are caused by anthropogenic influences, the “good status of groundwater” is failed and measures to improve the status have to be implemented. Hence, the local information about the origin of exceeding the TVs is an important prerequisite before the status of groundwater is defined and measures to improve the quality status are implemented.
values (TVs) for the definition of the groundwater chemical status has been tested with consistent results for the case study area “Upper Rhine Valley”. The evaluation has been done for 14 groundwater parameters using a large number of groundwater monitory data provided by French, German and Suisse water authorities. The results indicate that the proposed methodology gives reasonable results with regard to the requirements of the EU Water Framework Directive (EU-WFD, article 17.2a). In this way the derived NBLs and TVs may serve as starting points for the assessment of the chemical status of groundwater, e.g. for a trend reversal (article 17.2b) in case the good status is failed. The general applicability of the NBL and TV derivation approach on a European scale, i.e. its use for the overall purpose of the BRIDGE-project, is presently proofed in 16 other BRIDGE case study areas throughout Europe. Acknowledgements The groundwater monitoring data sets, provided by French, German and Suisse water authorities are greatly appreciated. The derivation of natural background levels and threshold values for groundwater bodies in the Upper Rhine Valley has been compiled in the framework of the EU-Specific Targeted Research Project “BRIDGE” (Background cRiteria for the IDentification of Groundwater thrEsholds) – Contract No. 006538 (SSPI). The views expressed in the paper are purely those of the writer(s) and may not in any circumstances be regarded as stating an official position of the European Commission. References [1]
7. Conclusion An EU-wide applicable approach to assess natural background levels (NBLs) and threshold
167
[2]
P. Nieto, E. Custodio and M. Manzano, Baseline groundwater quality: a European approach, Environ. Sci. Policy, 8 (2005) 399–409. W.M. Edmunds, P. Shand, P. Hart and R.S. Ward, The natural (baseline) quality of groundwater: a
168
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
F. Wendland et al. / Desalination 226 (2008) 160–168 UK pilot study, The Sci. Total Environ., 310 (2003) 25–35. R. Salminen and T. Tarvainen, The problem of defining geochemical baselines: a study of selected elements and geological materials in Finland, J. Geochem. Explor., 60 (1997) 91–98. S.V. Panno, W.R. Kelly, A.T. Martinsek and K.C. Hackley, Estimating background and threshold nitrate concentrations using probability graphs, Ground Water, 44 (5) (2006) 697–709. C. Reimann and R.G. Garrett, Geochemical background: concept and reality, Sci. Total Environ., 350 (2005) 12–27. N. Campbell, B. D’Arcy, A. Frost, V. Novotny and A. Sansom, Diffuse Pollution: An Introduction to the Problems and Pollutions, IWA Publishing, London, GB, 2004. J. Plant, D. Smith, B. Smith and L. Williams, Environmental geochemistry at the global scale, Appl. Geochem., 16 (2001) 1291–1308. C.A.J. Appelo and D. Postma, Geochemistry, in: A.A. Balkema (Ed.), Groundwater and Pollution – 3, Rotterdam, Brookfield, 1996, p. 536 [corrected print]. L. Chery and O. Lapin, Contribution à la caractérisation des états de référence géochimique des eux souterraines, Application de la méthodologie à la nappe du Rhin, Rapport BRGM/RP-52163-FR, 2003, p 127. A. Blum, R. Kunkel and F. Wendland, Natural background levels, State of the art and review of existing methodologies, in: BRIDGE – Background Criteria for the Identification of Groundwater
[11]
[12]
[13]
[14]
[15]
[16]
thresholds, D 10: Impact of hydrogeological conditions on pollutant behaviour in groundwater and related ecosystems, pp. 85–158. R. Kunkel, H.-J. Voigt, F. Wendland and S. Hannappel, Die natürliche, ubiquitär überprägte Grundwasserbeschaffenheit in Deutschland, Schriften des Forschungszentrums Jülich, Reihe Umwelt/Environment, Band 47, Forschungszentrum Jülich GmbH, Jülich, Germany, 2004, p. 204 S. L. Chery (coord.), Qualité naturelle des eaux souterraines – Méthode de caractérisation des états référence des aquifères francais, Rapport BRGMeditions, Orlèans, 2006, p. 238. HLf U, Grundwasserbeschaffenheit in Hessen, Auswertung von Grund- und Rohwasseranalysen bis 1997, Hessische Landesanstalt für Umwelt, Umweltplanung, Arbeits- und Umweltschutz, Heft 250, 1998, p. 102. LfU BW, Grundwasserüberwachungsprogramm – Geogen geprägte Hintergrundbeschaffenheit. Landesanstalt f. Umweltschutz Baden, Württemberg, 1994, p. 94. F. Wendland, S. Hannappel, R. Kunkel, R. Schenk, H.-J. Voigt and R. Wolter, A procedure to define natural groundwater conditions of groundwater bodies in Germany, Water Sci. Technol., 51 (3–4) (2005) 249–257. T.H. Christensen, P.L. Bjerg, S.A. Banwart, R. Jakobsen, G. Heron and H.-J. Albertsen, Characterization of redox conditions in groundwater contaminant plumes, J. Contam. Hydrol., 45 (2000) 165–241.
Derivation of natural background levels and threshold values for groundwater bodies in the Upper Rhine Valley (France, Switzerland and Germany) F. Wendlanda*, G. Bertholdb, A. Blumc, P. Elsassc, J.-G. Fritscheb, R. Kunkel a, R. Wolterd a
Agrosphere, ICG-IV, Forschungszentrum Jülich GmbH, Jülich, Germany Tel. +49 2461 613165; Fax +49 2461 612518; email: [email protected] b Hessian Agency for Environment and Geology (HLUG), Wiesbaden, Germany c Bureau de Recherches Géologiques et Minières (BRGM), Orléans, France d Federal Environmental Agency (UBA), Dessau, Germany Received 20 December 2006; revised accepted 30 January 2007
Abstract According to the requirements of the EU Water Framework Directive (EU-WFD, article 17.2a) criteria for the assessment of the chemical status of groundwater have to be developed, which may serve as starting points for a trend reversal (article 17.2b) in case the good status is failed. Against this background an EU-wide applicable approach to assess natural background levels (NBLs) and threshold values (TVs) for the definition of the groundwater chemical status has been developed. NBLs are derived based on a preselection method using aquifer typologies as natural reference systems. Samples with purely anthropogenic substances (e.g. PAC, pesticides) as well as samples, in which indicator substances for anthropogenic inputs (e.g. nitrate) are exceeding a certain value, are excluded. The NBLs are defined for the remaining groundwater samples of the aquifer typology under concern as the concentration range between the 10% and 90% percentiles. The threshold values (TVs) get established with reference to the NBLs and a “not acceptable reference value” (REF), e.g. drinking water standards or toxicological values, guaranteeing that the TVs are higher than the NBLs, but lower than the REF values. The applicability of the methodology for NBL and TV derivation has been applied to the transboundary groundwater body of the Upper Rhine Valley in Germany, Switzerland and France. Keywords: EU Water Framework Directive (article 17); Groundwater quality; Natural background levels; Threshold values, Upper Rhine Valley
*Corresponding author. Presented at the 10 th IWA International Specialized Conference on Diffuse Pollution and Sustainable Basin Management, Istanbul, Turkey, 18–22 September 2006. 0011-9164/06/$– See front matter © 2008 Published by Elsevier B.V. doi:10.1016/j.desal.2007.01.240
F. Wendland et al. / Desalination 226 (2008) 160–168
1. Introduction The Commission proposal of Groundwater Directive COM(2003)550 developed under Article 17 of the Water Framework Directive (WFD) sets out criteria for the assessment of the chemical status of groundwater, which is based on existing Community quality standards (nitrates, pesticides and biocides) and on the requirement for Member States to identify pollutants and threshold values (TVs) that are representative of groundwater bodies found as being at risk, in accordance with the analysis of pressures and impacts carried out under the WFD. In this context groundwater– background values are required as a reference to quantitatively evaluate whether or not groundwater is significantly modified by anthropogenic influences [1]. However, the term “background level” is used in a number of different senses, e.g. in the context of “baseline concentration” [1,2], “geochemical baseline” [3], “background concentration” [4], “ambient background” [5]. All these terms reflect a different understanding or definition of what is regarded as being “natural” and/or “not natural” and are all connected to a specific location and scale. Because of the omnipresence of human impacts in the case study area strictly “natural” groundwater occurs at best at regionally limited locations. Consequently, the present solutions content of groundwater samples from surface-near aquifers do hardly represent strictly “natural” groundwater concentrations. Against this background a more pragmatically understanding of the term “natural background level” needs to be used, which considers the human impacts on groundwater to a certain degree as inevitable. Therefore, we define “natural background level” as “the concentration of a given element, species or chemical substance present in solution which is derived by not significant anthropogenic influenced processes from geological, biological or atmospheric sources”. In the light of the above, the EU-Specific Targeted Research Project BRIDGE (Background
161
cRiteria for the IDentification of Groundwater thrEsholds) was initiated, involving 27 partners from 19 countries and links efficiently stakeholders of the scientific and policy-making communities. A scientifically based and generally applicable approach to derive natural background levels for the groundwater and groundwater threshold values was derived. The applicability and validity of this approach is checked in 14 case study areas at the river basin district or groundwater body level throughout Europe, including the Upper Rhine Valley as a transboundary French–German–Suisse case study.
2. Aquifer typologies for hydrogeochemical characterisation 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 hydrodynamical conditions. Apart from these “natural” factors groundwater quality is influenced by anthropogenic inputs mainly from diffuse sources (e.g. agriculture, atmosphere) [6]. Whereas some of these inputs (e.g. pesticides) are a direct indicator of human impacts, most inorganic contents occurring in the groundwater originate both from natural and anthropogenic sources [7]. This makes it difficult to decide whether an observed groundwater concentration pattern in a certain area is influenced by pollution intakes or still represents an (almost) natural state. In particular, groundwater from aquifers taking part in the active water cycle (surface-near aquifers), in most cases aquifers also 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. For a European wide assessment, an integration of aquifer typologies with respect to its typical groundwater conditions is most needed. For this
162
F. Wendland et al. / Desalination 226 (2008) 160–168
purpose the complexity of individual groundwater bearing rocks (aquifers, groundwater bodies) was classified into simplified main aquifer typologies. This is done mainly on the basis of petrography because the groundwater of aquifers with similar petrographic properties shows very similar natural composition in comparable hydrodynamic and hydrologic environments [8]. In order to provide both a practical framework for an overview about major groundwater composition patterns in Europe and for regional referenciation of natural background (NBLs) and threshold (TVs) for the groundwater in Europe, an aquifer typology map for hydrochemical characterisation has been created within the BRIDGE project. This map is presented in Fig. 1. The location of the case study area is indicated on the map.
3. The case study area Upper Rhine basin The case study area Upper Rhine Valley is a transboundary river basin located between France and Germany with smaller parts in Switzerland. It belongs as a whole to the aquifer typology “fluviatile deposits of major streams” (see Fig. 2, left part). The total area comprises 9.290 km2. It represents a very important drinking water reservoir with high importance for supraregional water supply. Due to the high population density a variety of anthropogenic impacts on groundwater quality (intensive agriculture, industry, water withdrawal) are present. With respect to the BRIDGE aquifer typologies, the Upper Rhine Valley is part of the typology “fluviatile deposits of major streams”. Hydrogeologically, the Upper Rhine Valley represents one of the biggest rift structures in Europe filled up with alternating layers of slit, clay, sand and gravel during the Pliocene and Quaternary period (see cross section, Fig. 2, right part). The hydraulic conductivity of the aquifers decreases from south to north and is associated with a decrease in the grain size of the sediments along the flow direction [9].
4. General applicable preselection approach for NBL derivation An evaluation of existing approaches for NBL assessment [10–16] has shown that preselection methods are appropriate to derive natural background levels, especially for large area considerations. Furthermore, they fulfill the requirements to be applicable on a European scale: they require “no deeper knowledge” about statistical analysis, can be applied by non-experts and to groundwater bodies, for which only few samples are available. By giving a general overview of the natural chemical composition of the groundwater bodies, they fit into the WFD requirements and with WFD working scale. The basic idea of preselection method is that there is a correlation between the concentration of certain indicator substances and the presence of anthropogenic influences. The NBLs are derived only from groundwater samples, where the concentrations of these indicator substances are below a defined level, indicating that the samples can be regarded as not being influenced by anthropogenic activities. Although preselection methods have the disadvantage that the relation between high concentration of indicator substances and anthropogenic induced high concentrations of other substances is not true in all cases [11], this kind of approach is particularly representative and interesting for elements for which a considerable number of samples are available. Typical indicator substances are either anthropogenic substances (e.g. PAC, pesticides) or substances occurring naturally only in low concentrations (nitrate, heavy metals, etc.). An ideal indicator substance for groundwater systems should be easily detectable and typical for a certain type of pollution (i.e. diffuse or point source pollution). As inorganic groundwater substances may occur in the groundwater naturally as well as due to anthropogenic inputs, there are only a few substances, whose presence in high concentrations surely indicates anthropogenic inputs. Nitrate and
Fig. 1. BRIDGE aquifer typologies for hydrochemical characterization.
F. Wendland et al. / Desalination 226 (2008) 160–168 163
Fig. 2. Geographical location and cross section of the case study area “Upper Rhine Valley”.
164 F. Wendland et al. / Desalination 226 (2008) 160–168
F. Wendland et al. / Desalination 226 (2008) 160–168
ammonia are examples for such substances. In most cases high nitrate concentrations coincide with low ammonia concentrations and vice versa, so that they are acting like a supplementary indicator set for anthropogenic pressure to groundwater. Whereas nitrate is a good indicator for pollution in oxidized groundwater, it may be absent in reduced groundwater even in case of high inputs due to denitrification processes (e.g. [16]). In that case ammonia and potassium may serve as an indicator for anthropogenic pressure to groundwater in reduced groundwater bodies instead. The following criteria to preselect the groundwater samples are based on the experiences in several French and German studies on NBL assessment [11–15]: • Exclusion of groundwater samples displaying purely anthropogenic substances, • Exclusion of samples displaying concentrations of indicator substances exceeding a certain value: – NO3 > 10 mg/L in oxidized aquifers (O2 > 2 mg/L and Fe(II) < 0.2 mg/L) or – NH4 > 0.5 mg/L in reduced aquifers (O2 < 2 mg/L and Fe(II) > 0.2 mg/L). The natural background values are defined for all groundwater parameters as the 90th percentiles (P90) of the concentration distributions from the remaining samples. For substances that are purely synthetic with no natural sources (for example, TCE), NBLs are set to zero. 5. Derivation of threshold values for the groundwater The threshold values (TVs) get established with reference to the NBLs and a chosen reference standard (REF). The latter may either be a drinking water standard value (DWS), an environmental quality standard or ecotoxicological value (EQS
165
Fig. 3. Illustration of the approach to derive threshold values (TV).
or EToxV). The calculation of threshold values is presented for 3 cases (see Fig. 3). • Case 1: NBL ≤ REF: If the NBL of a certain substance is below the REF value the TV is set as the concentration in the middle between the NBL and the REF values: TV = (REF + NBL)/2. • Case 2: NBL < one third of REF: In case the NBL is considerably below the REF-value, the TV is limited to twice the NBL: TV = 2 NBL. • Case 3: NBL ≥ REF: In case the NBL is larger than the REF value, the TV is set equal to REF itself: TV = NBL. The TV for purely synthetic substances with no natural sources (e.g. TCE) is defined as the detection limit.
166
F. Wendland et al. / Desalination 226 (2008) 160–168
6. Application to the case study area Upper Rhine Valley The data base for evaluating NBL and threshold values consists of almost 1700 groundwater samples for the years 2002 and 2003 from French, German and Suisse monitoring networks (see Fig. 4). For each of the monitoring stations one sample containing the solution contents for integral and chemical environment parameters, characteristic major and minor parameters and WFD pollutants was available. However, no information on heavy metals was contained in the data base. The preselection approach described above was applied to the data base for the Upper Rhine Valley. Because of missing information concerning purely anthropogenic substances, no groundwater samples were excluded due to this condition.
Preselection according to Nitrate (NO3 < 10 mg/L) has lead to the exclusion of 1094 samples (64%). Evaluation of the redox status of groundwater has shown that 188 samples indicate reduced aquifer conditions. 35 of those samples were excluded from the NBL derivation because of their high NH4-contents (NH4 > 0.5 mg/L). In the end 594 groundwater samples (see figure, left side) remained. For those samples it was assumed that they are to a large extent free from human impact. Hence, they are regarded as being appropriate for the NBL derivation in the Upper Rhine Valley. The right part of Fig. 4 shows the derived NBL and TV for 14 substances, for which NBLs have been defined and REF values were available. Depending on the NBL very different TV are derived. As the last column in the table in Fig. 4 indicates, that case 1 (see Fig. 3) is typical
Parameter
Unit
TV1
Case
mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L µS/cm
P90 0.1 84 3.6 7.2 25 0.82 41 173 951
Ref
B Cl Fe(II) K Mg Mn(II) Na SO4 LF
1 250 0.2 10 50 0.05 200 250 2500
0.2 167 3.6 8.6 37 0.8 83 211 1726
2 1 3 1 1 3 2 1 1
As NH4 NO2 NO3 PO4
µg/L mg/L mg/L mg/L mg/L
4 0.39 0.04 8.2 0.17
10 0.5 0.5 50 6.7
7 0.45 0.08 16.4 0.34
1 1 2 2 2
Fig. 4. Geographical location of sampling sites in the Upper Rhine Valley (left) and NBL- as well as TV-values (right).
F. Wendland et al. / Desalination 226 (2008) 160–168
for TV derivation for most of the investigated parameters (Cl, K, Mg, SO4, LF, As, NH4), i.e. the TV was set to half of the difference between NBL and the REF. For B, Na, NO2, NO3 and PO4, however, parameters, for which the NBL is significantly below the REF value, case 2 was applied. Only for two parameters Mn(II) and Fe(II), the NBL is above the REF value, so that the NBL was regarded to be equal to the TV. After data preselection 594 samples have been used to derive the NBL and TV. However, it would be interesting to compare these values to the situation assessed on the basis of the total data set containing about 1700 groundwater samples. An analysis shows that about 90–95% of the 1700 available samples display concentrations below the derived threshold values. In these cases a good status of groundwater can be postulated and no further measures to improve groundwater quality are required. However, 5–10% of the samples exceed the TV. For these cases the reasons for this exceedence need to be checked. The impacts of natural influences, like salinisation, redox-conditions, hydrodynamics, which may lead to natural (geogenic) anomalies, as well as of anthropogenic influences like diffuse and point source pollution must be assessed. If the reasons of exceeding the TV are due to natural influences, the “good status of groundwater” is achieved even in case the TV is exceeded as it can be explained by natural processes. Only in case the reasons are caused by anthropogenic influences, the “good status of groundwater” is failed and measures to improve the status have to be implemented. Hence, the local information about the origin of exceeding the TVs is an important prerequisite before the status of groundwater is defined and measures to improve the quality status are implemented.
values (TVs) for the definition of the groundwater chemical status has been tested with consistent results for the case study area “Upper Rhine Valley”. The evaluation has been done for 14 groundwater parameters using a large number of groundwater monitory data provided by French, German and Suisse water authorities. The results indicate that the proposed methodology gives reasonable results with regard to the requirements of the EU Water Framework Directive (EU-WFD, article 17.2a). In this way the derived NBLs and TVs may serve as starting points for the assessment of the chemical status of groundwater, e.g. for a trend reversal (article 17.2b) in case the good status is failed. The general applicability of the NBL and TV derivation approach on a European scale, i.e. its use for the overall purpose of the BRIDGE-project, is presently proofed in 16 other BRIDGE case study areas throughout Europe. Acknowledgements The groundwater monitoring data sets, provided by French, German and Suisse water authorities are greatly appreciated. The derivation of natural background levels and threshold values for groundwater bodies in the Upper Rhine Valley has been compiled in the framework of the EU-Specific Targeted Research Project “BRIDGE” (Background cRiteria for the IDentification of Groundwater thrEsholds) – Contract No. 006538 (SSPI). The views expressed in the paper are purely those of the writer(s) and may not in any circumstances be regarded as stating an official position of the European Commission. References [1]
7. Conclusion An EU-wide applicable approach to assess natural background levels (NBLs) and threshold
167
[2]
P. Nieto, E. Custodio and M. Manzano, Baseline groundwater quality: a European approach, Environ. Sci. Policy, 8 (2005) 399–409. W.M. Edmunds, P. Shand, P. Hart and R.S. Ward, The natural (baseline) quality of groundwater: a
168
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
F. Wendland et al. / Desalination 226 (2008) 160–168 UK pilot study, The Sci. Total Environ., 310 (2003) 25–35. R. Salminen and T. Tarvainen, The problem of defining geochemical baselines: a study of selected elements and geological materials in Finland, J. Geochem. Explor., 60 (1997) 91–98. S.V. Panno, W.R. Kelly, A.T. Martinsek and K.C. Hackley, Estimating background and threshold nitrate concentrations using probability graphs, Ground Water, 44 (5) (2006) 697–709. C. Reimann and R.G. Garrett, Geochemical background: concept and reality, Sci. Total Environ., 350 (2005) 12–27. N. Campbell, B. D’Arcy, A. Frost, V. Novotny and A. Sansom, Diffuse Pollution: An Introduction to the Problems and Pollutions, IWA Publishing, London, GB, 2004. J. Plant, D. Smith, B. Smith and L. Williams, Environmental geochemistry at the global scale, Appl. Geochem., 16 (2001) 1291–1308. C.A.J. Appelo and D. Postma, Geochemistry, in: A.A. Balkema (Ed.), Groundwater and Pollution – 3, Rotterdam, Brookfield, 1996, p. 536 [corrected print]. L. Chery and O. Lapin, Contribution à la caractérisation des états de référence géochimique des eux souterraines, Application de la méthodologie à la nappe du Rhin, Rapport BRGM/RP-52163-FR, 2003, p 127. A. Blum, R. Kunkel and F. Wendland, Natural background levels, State of the art and review of existing methodologies, in: BRIDGE – Background Criteria for the Identification of Groundwater
[11]
[12]
[13]
[14]
[15]
[16]
thresholds, D 10: Impact of hydrogeological conditions on pollutant behaviour in groundwater and related ecosystems, pp. 85–158. R. Kunkel, H.-J. Voigt, F. Wendland and S. Hannappel, Die natürliche, ubiquitär überprägte Grundwasserbeschaffenheit in Deutschland, Schriften des Forschungszentrums Jülich, Reihe Umwelt/Environment, Band 47, Forschungszentrum Jülich GmbH, Jülich, Germany, 2004, p. 204 S. L. Chery (coord.), Qualité naturelle des eaux souterraines – Méthode de caractérisation des états référence des aquifères francais, Rapport BRGMeditions, Orlèans, 2006, p. 238. HLf U, Grundwasserbeschaffenheit in Hessen, Auswertung von Grund- und Rohwasseranalysen bis 1997, Hessische Landesanstalt für Umwelt, Umweltplanung, Arbeits- und Umweltschutz, Heft 250, 1998, p. 102. LfU BW, Grundwasserüberwachungsprogramm – Geogen geprägte Hintergrundbeschaffenheit. Landesanstalt f. Umweltschutz Baden, Württemberg, 1994, p. 94. F. Wendland, S. Hannappel, R. Kunkel, R. Schenk, H.-J. Voigt and R. Wolter, A procedure to define natural groundwater conditions of groundwater bodies in Germany, Water Sci. Technol., 51 (3–4) (2005) 249–257. T.H. Christensen, P.L. Bjerg, S.A. Banwart, R. Jakobsen, G. Heron and H.-J. Albertsen, Characterization of redox conditions in groundwater contaminant plumes, J. Contam. Hydrol., 45 (2000) 165–241.