The redox environment is of utmost importance for the removal of organic compounds during artificial recharge. Within the research project OXIRED-2 five laboratory sand column experiments with natural sediments from the Lake Tegel infiltration pond and with microsieved surface water from Lake Tegel (Berlin) were performed to study the possibility to control the redox environment. Special emphasis was given to the sediments, the set-up of the column experiments, and the contact time within the column. The sediment was used either untreated or heated to 200°C or 550°C to study the effect of activation of organic carbon at 200°C and the effect of at least partial removal of natural organic carbon at 550°C. Additionally, an artificially produced iron coated sand was used for a two-layer experiment to increase the residence time of compounds susceptible to sorption within a given redox zone. Results reveal an immediate decrease of oxygen content at the outflow of the column in every experiment. Likewise, the redox potential also dropped significantly and immediately after the experiments started. However, the redox potential was significantly lower (approximately – 200 mV) in the experiments with the untreated or slightly heated sediments, and higher (about + 300 mV) for the experiment with the sediment heated up to 550°C. The redox zones known in natural environments developed also within the experiments even down to sulfate reduction at experiment No. 2. Ozonation of the influent water did not change the redox environment at the outflow of the column indicating a high reduction capacity of the natural sediment in the column within the duration of the experiments of up to 19 days. A constant input of ozone and an extended duration of the experiments might lead to a depletion of organic carbon in the sand column which could increase the redox potential. However, a complete depletion of organic carbon is very unlikely for managed aquifer recharge systems. The two-layer experiment with natural sand and artificially produced iron coated sands revealed that the iron coated sands had no influence on the redox system and only slight effect on the transport of ions. However, combining layers with different functionality might show great opportunities for designing and controlling redox systems especially with specific residence times in different redox zones for certain compounds in mind.

Subsurface passage as utilized during bank filtration and artificial groundwater recharge has shown to be an effective barrier for multiple substances present in surface waters during drinking water production. Additionally it is widely used as polishing step after wastewater treatment. However, there are limitations concerning the removal of DOC and specific trace organics. The project ”OXIRED“ aims at assessing possibilities to overcome these limitations by combining subsurface passage with oxidation by ozone. Results from the first phase of the project have demonstrated that oxidation with ozone is a suitable method to reduce the concentrations of several relevant trace organic compounds (e.g. carbamazepine, sulfamethoxazole) and to significantly enhance biodegradation of DOC during subsequent soil passage. For efficient removal of DOC in the soil columns, specific ozone consumptions of 0.6 to 0.7 mgO3/DOC0 were sufficient. Project objectives in OXIRED-2 were to i) verify results from laboratory scale experiments at a larger scale with longer retention times, ii) study feasibility under field conditions with seasonal variations by operating a pilot unit, iii) evaluate the formation of oxidation by-products and their persistence during subsurface passage and iv) propose a standardized test protocol to analyse benefits of ozonation and artificial groundwater recharge at different sites. To investigate effects of ozonation on groundwater recharge with longer retention times, a technical scale column system with a length of 30 m and a hydraulic retention time of approximately six weeks was operated at the UBA’s experimental site in Berlin Marienfelde. Pilot studies were conducted at Lake Tegel using an ozone unit from ITT-Wedeco with a 4 g/h generator and subsequent slow sand filtration. Reduction of bromate was assessed in laboratory scale soil columns under different redox conditions. In addition, anoxic reduction of bromate was evaluated in a diploma thesis at TU Berlin. To analyse effects of DOC removal after ozonation, a standardized test protocol using recirculating columns was proposed and tested. Results from the different experiments confirmed the conclusions of the first phase of the project. Removal of surface water DOC during infiltration significantly increased with preozonation. In pilot studies, effluent DOC of approximately 4.7 mg/L after 1 d of retention time was measured, which is comparable to residual DOC from artificial groundwater recharge in Berlin Tegel after 30 days retention time [1]. In addition, strong effects of temperature on DOC removal were observed. During experiments with ozonation, overall DOC reduction decreased from approximately 40% in October to about 30% in the end of November. Biological testing of slow sand filter effluent revealed no genotoxic or cytotoxic effects in the water prior to further infiltration into the aquifer. Many persistent trace compounds were efficiently transformed during ozonation with specific ozone doses of 0.8 mg O3/mg DOC0. For example, realistic surface water concentrations of carbamazepine, sulfamethoxazole, phenazone and bentazone were reduced below the limits of quantification (LOQ). Primidone was only partly transformed during ozonation (70%). Since primidone is persistent during infiltration, a breakthrough in combined ozonation and artificial recharge can be expected. Also the substances MTBE and ETBE, the pesticide atrazine and some metabolites detected in Lake Tegel persist partially during treatment with ozone and subsequent groundwater recharge. For efficient transformation of these substances, higher ozone doses or an optimisation of the oxidation process, for example as advanced oxidation process (AOP), should be considered. Efficient reduction of the concentration of adsorbable organic iodine (AOI), an indicator for x-ray contrast media, during ozonation or infiltration was not observed. In contrast, adsorbable organic bromine decreased by 70 - 80 % during ozonation. Formation of the oxidation by-product bromate during ozonation of Lake Tegel water with a specific ozone consumption of up to 1.0 mg O3/mg DOC0 was below the limit of the German drinking water directive. Removal during subsurface passage was observed under anoxic conditions in presence of biodegradable organic carbon. Since artificial recharge after ozonation is likely aerobic, no significant reduction of bromate can be expected. Thus, formation of bromate needs to be controlled during surface water ozonation. Formation of nitrosamines was monitored in batch experiments with a specific ozone consumption of up to 1.15 mg O3/mg DOC0. No formation of nitrosamines including NDMA (LOQ: 5 ng/L) was observed. Operating a preceding bank filtration step will reduce ozone demand for efficient DOC removal. In addition, problems with particles from source water can be minimised. However, additional energy consumption for operation of extraction wells has to be taken into account. Overall, the presented results confirm that the objectives of enhanced removal of trace organics and DOC by combining ozonation and subsurface passage are well met. Further investigations need to focus on seasonal variations in long-term pilot studies and the formation, retention and toxicity of transformation products.

During WELLMA-DNA, 13 diploma and bachelor theses along with several internships have been completed. A sampling system for biofilm samples as well as a sampling device for water samples have been designed and tested. More than 400 DNA samples of different well sites have been collected and analyzed. Microbiological and molecular methods have been combined to gain a better understanding of the community composition of the ochre forming biofilms inside the wells. Molecular methods included PCR, DGGE, cloning and sequencing. During the project, the bacterial populations of an unprecedented number of wells have been analyzed and several indicator bacteria for iron-related well clogging have been identified. Alongside iron-oxidizing bacteria, iron-reducing bacteria have been found in the wells and their potential for ochre-solubilization was confirmed. Alongside the molecular experiments, microbiological trials included the isolation of pure cultures, microscopic analysis and physiological tests. The morphology of the encountered iron bacteria could be classified into four different groups, which may have an impact on the rigidity of the biofilms on a macroscopic level. We were able to cultivate several of these indicator organisms, which could play an important role in the formation of ochreous deposits in the Berlin wells. During experiments utilizing microscopic flow cells, differences in growth rate and patterns of these ochre-forming bacteria have been observed. For several of the identified indicator bacteria, primers have been calculated. These primers will allow for the first time to quantify the amount of indicator bacteria in a water sample and to derive operational pointers. In addition, several experiments regarding the effect of hydrogen peroxide on ochre forming biofilms have been conducted and the effect of an additional electron donor (ethanol) on the communities has been tested. For future data acquisition and documentation, a guideline for classifying the degree of pump clogging has been developed.

Work package WP 5.2 “Combination of Managed Aquifer Recharge (MAR) and adjusted conventional treatment processes for an Integrated Water Resources Management“ within the European Project TECHNEAU (“Technology enabled universal access to safe water”) investigates bank filtration (BF) + post-treatment as a MAR technique to provide sustainable and safe drinking water supply to developing and newly industrialised countries. One of the tasks within the project is the development of a Decision Support System (DSS) to assess the feasibility of BF systems under varying boundary conditions such as: (i) quality of surface and ambient groundwater, (ii) local hydrological and hydrogeological properties (e.g. clogging layer) and (iii) well field design (distance to bank) and operation (pumping rates). Since the successful, cost-effective implementation of BF systems requires the optimization of different objectives such as (i) optimizing the BF share in order to maintain a predefined raw water quality or (ii) maintaining a predefined minimum travel time between bank and production well, both aspects are addressed within the DSS. As an example for a practical application the DSS is tested with data from the Palla well field in Delhi/India. As a result optimal shares of bank filtrate were calculated for the monsoon and non-monsoon season. By simulating different pumping and clogging scenarios with the BF Simulator optimal pumping rates were derived. The DSS proved to be a good qualitative tool to identify and learn about the trade-offs a decision maker has to make due to the (i) inherently competing nature of different objectives (e.g. high BF share and minimum travel time > 50 d) and the (ii) inherent uncertainty due to the large natural variability of boundary conditions (e.g. clogging layer). Since both characteristics can be addressed within the DSS it helps to add transparency and reproducibility to the decision making process. An additional advantage is that its application requires only low effort concerning time, money, and manpower. Thus the application of the DSS is recommended to accompany decision making processes especially in developing and newly industrialised countries where data availability and low financial budgets are usually the major burden for the application of more complex, data-demanding decision support tools. However, it needs to be considered that in practice additional parameters like water availability, energy efficiency and cost-benefit need to be taken into account.

Work package WP 5.2 “Combination of Managed Aquifer Recharge (MAR) and adjusted conventional treatment processes for an Integrated Water Resources Management“ within the European Project TECHNEAU (“Technology enabled universal access to safe water”) investigates bank filtration (BF) + post-treatment as a MAR technique to provide sustainable and safe drinking water supply. One of the tasks within the project is the testing of a data-driven approach for the identification (pattern recognition) and quantification of the key processes that drive the groundwater (GW) dynamics in observation wells (OW) near well fields of a BF waterworks. For this BUSSE (2010) used a multivariate statistical method (principal component analysis - PCA) with daily GW level time series of 41 OWs and was able to identify four processes that explained 95% of the total variance in the data set. On the one hand GW recharge (58.9%) and its temporal delay (3.3%) explain 62% of the GW level fluctuations within the study period. On the other hand any discernible impact of waterworks abstractions is limited to one of the three well fields with the highest production rate (29.8% of explained variance). In addition the infiltration of a marshy ditch into the GW accounts for another 2.9% of the GW level fluctuations. Regarding the ability to identify driving forces for GW level fluctuations the main advantage for using PCA compared to process-driven GW flow modelling is that the driving forces for GW level fluctuations can be identified and quantified without requiring exact knowledge about the structural properties of the subsurface (e.g. aquifer transmissivities) and its input parameters (e.g. GW recharge, production rates). Note that the latter do not enter the PCA directly but are used for spatiotemporal interpretation of the results, which also requires some expertise. In addition, it is recommended to perform a sensitivity analysis of the PCA results in a next step, so that it can be tested whether the processes identified above are robust in case of changing input parameters such as: - Reduced spatiotemporal resolution - Study period with different boundary conditions (e.g. pumping regime). The contents of this report were presented to the involved experts from the Berliner Wasserbetriebe (BWB). In agreement with their recommendations it was decided to focus further research within follow-up projects on the (i) sensitivity analysis of the PCA results and (ii) to apply nonlinear approaches for identification and quantification of processes that drive GW quality dynamics within the study area.

Artificial Recharge (AR) is a method to replenish groundwater in case of insufficient water availability or poor quality. For drinking water production, AR is often used as water purification step to avoid direct surface water abstraction. Besides physical filtration, purification is achieved through chemical processes like precipitation, sorption and (bio-) degradation. These are usually closely linked to redox conditions. It is the activity of micro-organisms and related chemical reactions that change the redox conditions, which in turn control the presence of substances and therefore the water quality. Typical pollutants in surface water that need to be addressed are organic compounds (e.g. pharmaceutical residues or pesticides), pathogens and heavy metals. The purpose of this report is to introduce the theoretical background on redox zoning in infiltration ponds and to review publications in the search for applicable methods capable of controlling redox conditions. This shall serve as basis for further laboratory and technical scale experiments in the course of the OXIRED project. The “optimal redox zonation” for maximum removal of redox-dependent substances is a concept with the aim of defining optimum residence times based on the degradation kinetics of contaminants in the source water: If substances or substance groups that show enhanced removal under anoxic to anaerobic conditions are not present in the source water at drinking water relevant concentrations, anoxic to anaerobic conditions should be avoided in order not to mobilize iron and other inorganic trace elements. Maximum benefit for aerobic subsurface passage is reached after 30 d, for anoxic / anaerobic subsurface passage after 100 d. However, already 15 d of aerobic and 2 d of anoxic / anaerobic passage lead to substantial removal or redox-sensitive substances or substance groups. The main drivers for redox zonation in AR systems are the availability of oxidizing agents (oxygen, nitrate), of reducing agents (organic matter, reduced mineral phases), of nutrients, the biological activity (in infiltration pond and subsurface), and the residence time. These drivers are in turn controlled by many natural, site-specific (exogenous) and design & operation-related (decision) variables. Exogenous variables are e.g. aquifer geochemistry, temperature or natural groundwater recharge whereas the decision variables comprise factors such as pond geometry, distance between pond and well, well depth, pumping rate etc. Theoretically, a wide range of possibilities could be applied to adjust the infiltration pond, the hyporheic zone and the subsurface passage, but only few seem technically feasible. These are e.g. the control of sunlight and temperature in the infiltration pond and upper sediment, the control of water movement in the pond to avoid excessive algal growth while enriching the water with oxygen. For the same reason nutrients could be added or avoided, influencing biomass production. Specific filter material could be used with defined content and characteristic of organic carbon to serve as electron acceptors. Infiltration rates could be controlled by adjusting the hydraulic head in order to enhance the formation of an unsaturated zone. Further downstream the application of redox controlling substances via injection wells could be possible, as well as controlling the residence times by adjusting pumping rates or creating hydraulic barrier wells at different distances from infiltration pond. For newly constructed AR systems the well field design (pond geometry, distance between pond and well, well depth) could be optimized with respect to redox zonation, as long as the other requirements (mainly sufficient production rates) are met. No examples for redox control in infiltration ponds were identified. Therefore, two examples of redox control measures are described: the first serves an artificial reoxidation of a polluted aquifer “BIOXWAND®” and the second provides injection of treated water to influence the redox conditions in the aquifer “Vyridox” and “Nitridox”.

Combined sewer overflows can lead to acute, critical conditions for aquatic organisms in receiving surface waters (Borchardt et al. 2007; FWR 1998; Harremoes et al. 1996; Krejci et al. 2004; Lammersen 1997). Based on the river type of the River Spree, CSO impacts of possible concern were identified to be high ammonia (NH3) and low dissolved oxygen concentrations (DO) (Senatsverwaltung für Stadtentwicklung 2001; Leszinski et al. 2007). For DO, existing continuous measurements from the River Spree from 2000 to 2007 were assessed in detail in the KWB report by Riechel (2009). However, Riechel (2009) neglected NH3 toxicity assessment, since no continuous NH3 measurements were available. The present report aims at filling this gap by estimating the potential for toxic NH3 concentrations in the River Spree with recent data. Based on stormwater impact guidelines for ammonia, critical total ammonium concentrations ([NH4,tot] = [NH4+] + [NH3]) were calculated and compared to continuous NH4,tot measurements in the Berlin River Spree. NH4,tot was measured i) at a heavily CSO impacted river stretch (year 2011) and ii) at a monitoring station several kilometres downstream of the combined sewer area (years 2010 and 2011). The analysis led to the following results: (i) Two years of continuous NH4,tot measurements showed clear increases in NH4,tot due to CSO but no occurrence of critical toxicity levels for cyprinid fish, according to Lammersen (1997) (ii) Maximal observed concentration of ~1.3 mg-N-NH4,tot l-1 was ~5 times smaller than the lowest existing threshold, which would need to be exceeded for 24 h to be considered as critical. The observed maximal concentration peak had a duration of only 3 h. The threshold, corresponding to the 3 h-duration would be even ~8 times higher than the observed ~1.3 mg-N-NH4,tot l-1. (iii) Ammonia toxicity would only be possible if maximal NH4,tot occurred during highest sensitivity of the river due to very high pH > 9. However, it was observed that pH drops significantly during CSO impacts due to low pH in rain water, which makes pH > 9 during CSO very unlikely. Given the results, the risk for ammonia toxicity due to CSO is judged as very low, particularly in comparison with regular problematic DO conditions after CSO events in summer.