Fate of organic micropollutants in a karst aquifer system - Implications for sustainable raw water management

Abstract

The fundamental understanding of karst aquifers is vital for the sustainable management of raw water quality and eventually the access to clean drinking water for up to one quarter of the world’s population. In order to improve this understanding the storage and attenuation potential of a karst aquifer was investigated in the presented work, employing organic micropollutants as indicators for transport paths, attenuation and attenuation processes. As a prerequisite for reliable data acquisition, suitable stabilisation and storage strategies for organic micropollutants in water samples have been evaluated: addition of the biocides (i) copper sulphate and (ii) sodium azide to water samples directly after sampling with subsequent sample storage as liquid phase and (iii) direct solid phase extraction (SPE), stabilising the samples by reducing the water content. River water and treated effluent were chosen as commonly investigated matrices with a high potential of biodegradation activity. Analyses were carried out for sample storage temperatures of 4 and 28 °C for water samples stored as liquid phase and for sample storage temperatures of 4, 20 and 40 °C for SPE cartridges. Cooling of water samples alone was not sufficient for longer storage times (> 24 h). While copper sulphate caused detrimental interferences with azole- and imidazole-like compounds, sodium azide proved to be a suitable stabilising agent. The best results could be obtained for SPE cartridges stored coolly. Recommendations for sample preservation are provided. In the following chapter the long-term storage potential of a karst aquifer was investigated. To achieve a sustainable raw water quality for drinking water production, the understanding of this potential is highly essential. The transport dynamics of the two herbicides metazachlor and atrazine as well as a degradation product of the latter (desethylatrazine) were investigated at a karst spring over one year. Even 20 years after its ban in Germany, atrazine and its degradation product were almost always detectable in the spring water in the low ng L–1. Metazachlor could only be detected after precipitation events and the observed concentrations were significantly higher than atrazine or desethylatrazine. Comparing the dynamics of the herbicides with the inorganic ions Ca2+, Mg2+ and the electrical conductivity, a positive correlation of atrazine with these parameters could be observed. From this observation, atrazine is concluded to be located within the aquifer matrix, deteriorating the raw water quality for decades. In order to identify the attenuation potential within the conduits of karst aquifers in-situ and to estimate the risk posed by micropollutants, a dualtracer experiment was conducted to investigate differential transport in the subsurface: the reactive compound caffeine was used as a tracer to indicate the attenuation potential within the aquifer in-situ. Due to the low limit of quantification, only small amounts of caffeine needed to be injected. To calibrate a model and to visualise the attenuation of caffeine a conservative reference tracer (uranine) was injected simultaneously. The methodology was tested in a well characterised karst system in southwest Germany. The results indicate a significantly higher attenuation rate than was expected for karst aquifers. The attenuation was described as a first-order process. The corresponding half-life was 104 h. This low half-life suggests that a generally assumed low natural attenuation potential of karst aquifers is unjustified. The observed mass loss of caffeine illustrates the potential of caffeine to be used as reactive tracer for indicating in-situ attenuation potential within hydraulically highly conductive systems, such as karst aquifers. Due to the high attenuation rate of caffeine it does not pose a threat as a long-time contaminant. In combination with a conservative reference tracer an economical and environmentally benign method is presented in this chapter for the in-situ determination of the attenuation potential of highly conductive aquifer systems. Based on the results of the dualtracer experiment, a multitracer experiment was performed for verifying the results, examining the transferability of the attenuation potential of caffeine to other substances and to specify the attenuation processes responsible for the observed mass loss. Uranine, acesulfame and carbamazepine were injected into a sinkhole as reference tracers together with the reactive compounds atenolol, caffeine, cyclamate, ibuprofen and paracetamol. The breakthrough curves of the reactive compounds were interpreted relative to the reference substances. No significant retardation was observed for any of the investigated micropollutants. The determined half-lives of the reactive compounds range from 38 to 1400 h (i. e. persistent within the investigation period) in the following order (from high to no observed attenuation): paracetamol > atenolol ≈ ibuprofen > caffeine >> cyclamate. The attenuation rates are generally in agreement with studies from other environmental compartments and with the results from the dualtracer experiment. The occurrence of the biotransformation product atenolol acid served as evidence for the occurrence of in situ biodegradation within the aquifer system.