Approximately 70% of the drinking water in Germany (BGR) and about 50% worldwide (IGREC 2011) are abstracted from groundwater using filter wells. Their implementation, operation and maintenance are major factors contributing to the costs of drinking water production. According to an international survey (Howsam, Misstear & Jones 1995 ), 40% of worldwide used water abstraction wells work inefficiently in terms of well performance or water quality. This implies high costs and a great potential for improvement, both for the (re-) construction of new wells (capital investment) and well operation (energy consumption, maintenance needs). The main reason for inefficient well performance is so-called well ageing, i.e. the decrease in performance due to biological, chemical and / or physical processes in and around the well. Dominant factors determining type, extension and location of deposits are the geology of the exploited aquifer together with the qualitative properties of the abstracted water, the well design (dimensions and materials), construction (drilling method) and operation. The project WellMa, initiated and financed by the Berliner Wasserbetriebe (BWB) and Veolia Eau, and coordinated at the Berlin Centre of Competence for Water (KWB), aimed at improving the efficiency of drinking water abstraction wells by identifying, evaluating and testing methods of well management including design, operation and maintenance to slow down well ageing. Set into relation to ranges, in which ageing processes are known to occur, the initial data of well sites were used to differentiate a low, medium or high potential for the occurrence of well ageing and to define the monitoring needs accordingly. Well ageing processes were distinguished into six types, each of them implying different pre-requisites and site conditions and leading to different monitoring and/ or maintenance requirements. For carbonate scaling, iron ochre formation, biofouling, corrosion, colmation and sand intake pre-requisites, triggers and threshold conditions were identified and implemented in a decision support system enabling well operators to prioritize the needs for monitoring, diagnosis or maintenance action taking into account the specific well and site characteristics.

Approximately 70% of the drinking water in Germany (BGR) and about 50% worldwide (IGREC 2011) are abstracted 2 from groundwater using filter wells. Their implementation and operation are major factors contributing to the costs of drinking water production. Within the joint research project ANTIOCKER , funded by the German Ministry of Research and Education, and coordinated at the Dept. of Applied Microbiology of the Technical University Berlin, the partners Berliner Wasserbetriebe (BWB) and the Berlin Centre of Competence for Water (KWB) focus on the efficient operation of drinking water abstraction wells. One major reason for inefficient wells is so-called well ageing, i.e. the increase in drawdown at constant discharge rate due to biological, chemical and / or physical processes in and around the well. In Berlin, approximately 80% of clogging deposits are described to be of biochemical nature involving iron-related bacteria. Previous studies, i.e. in the scope of the KWB research project WELLMA have revealed that such well ageing phenomena are determined by multiple correlated biological and chemical processes. For this reason, it is the sound understanding of the main processes and key parameters that will provide the basis for the systematic control of iron bacteria occurrence by an optimized well operation. A new approach to a large variety of data from well construction and maintenance of the Berlin drinking water wells focused on the determination of key parameters for monitoring and the identification of hidden variables for ageing by means of probabilistic statistics. Cumulative distribution plots are used to visualize large data amounts and frequency distribution plots filter correlations between e.g. maintenance events in the lifetime of a well and monitoring data. First results indicate that small changes in the discharge rate Q on a daily basis could be used to monitor the well performance on a much higher frequency than the currently used evaluation of the specific capacity. In addition, the electric conductivity proved to be a key variable for clogging. Both parameters are now being verified in field investigations and further data analyses within the research project ANTIOCKER and about 50% worldwide are abstracted 2 from groundwater using filter wells. Their implementation and operation are major factors contributing to the costs of drinking water production.

The total phosphorus concentration (TP) in Berlin’s surface waters has to be decreased to 6090 µg/L to achieve good ecological conditions. Furthermore, securing the bathing water quality of surface waters gains more attention. The wastewater treatment plant (WWTP) in Ruhleben should be upgraded by tertiary treatment to reduce the TP concentration (< 80 µg/L) and the pathogen germ concentration in the WWTP effluent. The project OXERAM compares different filtration technologies with regard to their efficiency and applicability as tertiary treatment. In this thesis the combination of microsieve filtration and UV-disinfection has been investigated. Before this trial the microsieve was rebuilt after a one year operation period. The volume of the coagulation tank was reduced to 26 % (0.56 m³) of its initial volume. The coagulation stirrer was exchanged by a Turbomix TM, applying a constant G·t-value of 25 000. Additionally, the angle of the blades of the flocculation stirrer was changed, almost doubling the G-value (154 1/s.) Long term trials have shown that an average TP effluent concentration of 60 µg/L can be achieved through microsieve filtration (10 µm) with chemical pretreatment (2.0 mg Al/L PACl and 0.61 mg/L cationic polymer). 80 % of the grab samples had a TP concentration = 73 µg/L. An average effluent suspended solid concentration of 2.2 mg/L was achieved. The average residual Aluminum concentration was 0.35 mg/L. The increase of the G-value during coagulation and flocculation after the rebuild led to an increased energy demand of both stirrers. However, through the new hydraulic conditions, the average polymer dose was reduced by 65 % in comparison to the dynamic operation in summer 2011 and the hydraulic retention time during coagulation could be reduced to 1 minute at peak flow. Furthermore, the impact of applied energy during coagulation and flocculation was investigated. A reduction of the G·t-value during coagulation led to a higher SS effluent concentration of 25 % (2.8 mg/L). During flocculation a high G-value (153 1/s) was favorable for the microsieve performance. A lower backwash time and effluent turbidity were observed. As a result, the increased energy demand of the stirrers can be justified, alongside to the polymer and hydraulic retention time reduction, with an improved effluent water quality and a lower energy demand for the backwash. The UV-disinfection operated reliably after the microsieve filtration. The UV effluent concentration of Enterococci and E.Coli were always under the limit of quantification (15-38 MPN/100 mL), even at a fluence of 361 J/m². Coliphages (= 7 PFU/100 mL) were detected in the effluent of the UVdisinfection, when the calculated fluence was lower than 549 J/m². In this study it was demonstrated that the microsieve filtration with chemical pretreatment and a subsequent UV-disinfection represents an alternative as tertiary treatment. TP effluent values lower than 80 µg/L were reliably achieved. An excellent water quality accordingly to the EG Bathing Water Quality Framework Directive was attained through the subsequent UV-disinfection.

In order to meet the requirements of the European framework directive on water, authored by the European parliament, measures have to be taken to enhance the situation of the bodies of water in Berlin. The discharge of treated wastewater is an important source of phosphor into the bodies of water of Berlin. The OXERAM project investigates different possible technologies to decrease these entries of phosphor and the conditional growth of algae. Pilot plants for these technologies are being operated at Sewage Treatment Plant Ruhleben (STP Ruhleben). Installed are a microsieve plant and two membrane plants. In previous tests, it was possible to proof the functionality of the microsieve plant under static influent flow and volume proportional chemical dosing. Effluent concentrations of < 80 µg/L were reliably achieved. The goal of this thesis is to evaluate the functionality of the pilot plant under dynamic operating conditions and load proportional chemical dosing. Therefore several tests were conducted. In a first step the pilot plant was operated with an artificial daily flow pattern, which was calculated regarding the daily flow variations in the influent of the STP Ruhleben. Tests under volume proportional dosing, and under load proportional chemical dosing, were conducted. The dosing of the coagulant was thereby varied according to the ortho phosphate concentration in the influent of the pilot. In a third step the influent flow was directly correlated to the influent flow of STP Ruhleben, in combination with load proportional dosing. During all conducted tests, sufficiently low total phosphorus concentrations in the effluent of the pilot plant could be achieved. By using load proportional dosing, savings of 11 % for the coagulant and 14 % for the polymer could be achieved. In addition several tests concerning the optimization of the whole process were conducted e.g. different reaction times for the flocculation and different polymer concentrations were tested. At the end an estimation of the operating costs for a microsieve plant of the size that would be required for the STP Ruhleben was made. Thereby the costs for Energy, Coagulant and Polymer and repair/ maintenance were discovered as main parts of the operating costs. The specific operating costs would be ca. 3.1 Cent/m³. To be able to compare the investigated technology with the other processes further tests of the pilot plant in combination with a UV plant as disinfection step of the effluent water from the microsieve have to be conducted.

The main objective of this study was to identify design and operational differences between the Danish and the German waterworks in order to learn if the groundwater treatment in the Danish and German water supplies is comparable. Furthermore, the aim is to compile existing national designs to create an overview of Danish and German designs, respectively. The work was carried out with an emphasis on iron, manganese, and ammonium removal, and thus the aeration and filtration method were evaluated. Design differences in groundwater treatment between 19 Danish waterworks, 9 waterworks from Berlin and 3 from Hamburg were compiled. The design differences were revealed by questionnaires, review of standards, and an expert interview. The most significant difference was the design of the filters. In Denmark most of the waterworks apply two-stage filtration, where both pre- and afterfilters consist of quartz. A total of 69 % of the Danish filters used in the project had a supporting layer. The average height of the Danish filters was 1.05 meter and 0.66 meter for prefilter and afterfilter, respectively. In Berlin and Hamburg the most applied filtration type was a single filtration through a mono-media (quartz) filter with an average filter height of 2 meters. Half of them (47%) were constructed with a supporting layer. In all filtration types identified in this study the average German filter height exceeded the average Danish. While the grain size distribution in the effective filter layer was found to be relatively alike and within the range of 1 to 3 mm, with a slightly larger grain size in the Danish filters. The grain size found in the Danish supporting layer was much larger than in the German supporting layers. Furthermore, a special design for prefilters, so-called roughing filters, was identified in Copenhagen waterworks, where the grain size distribution was between 24 and 150 mm. Despite the generally smaller grain size, the average filter velocity was found to be slightly larger in the German filters. The application of aerators and the backwashing procedure in Denmark and Germany were found be different. In Denmark the most frequently found aerators are the stairs aerator and the multiple slat tray aerator, whilst in Germany the most applied forms of aerators are spray aerators and waterfall aerators. The backwashing procedure in Denmark was mainly found (37 %) as a combination of first an air flush, and then water, whilst the most frequently found (50 %) backwashing procedure in Germany was a combination with an additional simultaneous air and water flush in between the separate air and water flushes. The more extensive backwashing procedure in the German filter is believed to be connected to the higher height of the filters. Empty-bed contact time was calculated, and was found to be significantly shorter in the Danish filters, where the average empty-bed contact time was 23 minutes, compared to an average of 38 minutes in the German filters. The water chemistry suggested that the groundwater types found in Denmark, Berlin and Hamburg are similar, but this was not investigated further. The conclusion was that the filters in Denmark are not comparable to the filters found in Berlin and Hamburg, based on the fundamental differences in the construction concerning the filter height, grain size, and numbers of filter steps and the resulting diverse EBCT. The difference in filter designs needs to be considered when applying future research to both supplies.

The usage of membranes in wastewater treatment often leads to problems with scaling and fouling, which results in an irreversible loss of membrane permeability. Various pretreatments as well as mechanical and chemical cleaning possibilities are described and evaluated in order to ensure continuous operation. The cleaning has been performed by a sequential backwash with filtrate. In addition, the membrane was cleaned chemically to restore the flux to an acceptable level. The effect of chemically enhanced backwash (CEB) and cleaning in place (CIP) has been considered in more detail. The UF membranes are typically cleaned by soaking in alkali, acids and/or oxidizing solutions. Sodium hydroxide (NaOH), sodium hypochlorite (NaOCl) and acid sulphur (H2SO4) were used as cleaning agents. Furthermore, the impact of pre-treatment by pre-ozonation and subsequent coagulation on the performance of a polyether sulphone ultrafiltration membrane has been investigated in a pilot plant. Ozone is used in water treatment for the oxidation of organic substances, which leads to a reduction of organic fouling. A subsequent coagulation is applied to form stable aggregates out of biopolymers, which are most relevant for membrane fouling in order to backwash them easily from the surface and the membrane pores. Both pre-treatments have an influence on an improved filtration performance. The scope of the current paper is to critically evaluate the impact on the hydraulic and chemical treatment of an organic membrane and to find out which cleaning strategy is the best against membrane fouling.

This paper is one of the results, developed within the project OXERAM II by the Kompetenzzentrum Wasser Berlin. A pilot plant, equipped with a full (microfiltration) Berlin/Ruhleben. During the is operated for advanced wastewater project, different scale monolithic ceramic membrane treatment at WWTP pre-treatments, namely coagulation and the combination of coagulation and ozonation are applied. Multi-filtration trials in dead-end mode and constant flux operation are performed for over 6 months, with varying operational parameters like flux, time of filtration, dosages of coagulant and ozone. Operational behaviour is evaluated through the evolution of trans-membrane pressure via time. Also total and irreversible fouling rates are calculated showing benefits within the combined pre-treatment, regarding membrane fouling and stable operation at high recoveries (98 %). The application of 15 mgO3 · L-1, respectively a specific ozone dosage from 1.0 to 1.4 mgO3 · (mgDOC)-1 leads to a total fouling rate reduction of 75 %. LC-OCD analysis is furthermore used for a more detailed view on changes in DOC, especially biopolymers. Sampling of the pilot plants influent and effluent water is additionally used for the evaluation of treatment efficiency, e.g. disinfection and in particular phosphorous, where emissions are reduced to 20 ± 5 µg P · L-1 in accordance with the European Water Framework Directive (Directive 2000/60/EC).