Das Projekt soll mögliche Entlastungseffekte für die Berliner Gewässer durch Einsatz geeigneter Maßnahmen zur Reinigung von Straßenabläufen aufzeigen.

The presented work studies the influence of the sampling strategy on the quality of locally calibrated UV-VIS probe measurements in combined sewer overflows (CSO) and the receiving river. Results indicate that UV-VIS spectrometers are not able to provide reliable measurements of water quality in urban stormwater without being calibrated to local conditions with laboratory analyses of water samples. The use of the global calibration (supplied by the manufacturer) led to errors of at least 30% and 45% for CSO load and river concentration of chemical oxygen demand (COD), respectively. Even with reliable local calibration, COD loads contained significant uncertainties close to 20%. Uncertainties in COD load and concentration decrease below 30% if more than 15-20 samples (i.e. 3-4 stormwater events) are sampled for local calibration. The effort and associated sampling costs to gain more than 15-20 samples are much less effective, since load and concentration uncertainties remain relatively stable with an increasing number of samples used for the calibration. The presented analysis aims at supporting practitioners in the planning, operation and calibration of UV-VIS spectrometer probes.

The present study aims to explore the relationship between rainfall variables and water quality/quantity characteristics of combined sewer overflows (CSO), by the use of multivariate statistical methods and online measurements at a principal CSO outlet in Berlin (Germany). Canonical correlation results showed that the maximum and average rainfall intensities are the most influential variables to describe CSO water quantity and pollutant loads whereas the duration of the rainfall event and the rain depth seem to be the most influential variables to describe CSO pollutant concentrations. The analysis of Partial Least Squares (PLS) regression models confirms the findings of the canonical correlation and highlights three main influences of rainfall on CSO characteristics: (i) CSO water quantity characteristics are mainly influenced by the maximal rainfall intensities, (ii) CSO pollutants concentrations were found to be mostly associated with duration of the rainfall and (iii) pollutants loads seemed to be principally influenced by dry weather duration before the rainfall event. The prediction quality of PLS models is rather low (R² < 0.6) but results can be useful to explore qualitatively the influence of rainfall on CSO characteristics.

Within the project OXERAM state of the art membrane filtration was applied as a tertiary treatment step for advanced phosphorus removal in a municipal wastewater treatment plant. Two membrane types, ceramic and polymeric, were tested in pilot scale, using commercial membrane modules. Due to the drawback of membrane fouling, leading to comparably high investment and operating costs, pre-treatment with ozone was tested. Ozonation was expected to increase the sustainable flux for both membrane types. For both membranes types high filtrate quality was achieved. A mean total phosphorus concentration below 25 µg/L was achieved over two years. Additionally disinfection is reached and therefore the European bathing water standards were met. The effect of ozonation and coagulation on various water quality parameters were evaluated and are presented in this report. Ultrafiltration modules (0.02 µm) made of polyether sulfone (PES) were tested comparing different capillary diameters (0.9 vs. 1.5 mm) leading to different package densities (respectively 40 and 60 m2 per module). Both types were operated in parallel and the experience showed a more robust operation with 1.5 mm capillaries when applying high fluxes targeting high recoveries. Both evaluation parameters, total fouling rate and membrane regeneration by cleaning in place, suggested the 1.5 mm module for the application at the WWTP Ruhleben. Optimizing the operation set up and cleaning strategy proved that recoveries = 95 % could be achieved and therefore a second filtration unit treating the backwash water is obsolete. The design with max 75 L/(m2h), 60 minutes of filtration, and a backwash duration of 40 s is the proposed set up for WWTP Ruhleben. A daily acidic chemical enhanced backwash combined with a weekly caustic cleaning step proved to manage the fouling affinity and a cleaning in place interval of 1 – 3 months was demonstrated in a long term run. The usage of ozone did not improve the overall filtration performance, because the benefit of a higher filterability is compensated by a higher additional fouling resistance after each backwash. Therefore the mean trans-membrane pressure remains in the same range. These results were only collected with the combination of ozonation and PES ultrafiltration membranes. Lab scale tests conducted at the Chair of Water Quality, TU Berlin, confirm this outcome but showed different results for other membrane materials and pore sizes. The potential to reduce the total fouling rate combining ozonation with coagulation prior ceramic membrane filtration was shown. A microfiltration membrane (0.1 µm) consisting of Al2O3 and a surface of 25 m2 was tested in pilot scale. Applying a dose of 15 mgO3/L (z = 1.18 mgO3/mgDOC) could reduce the total fouling rate by half even when doubling the flux from 60 L/(m2h) to 120 L/(m2h). Critical flux experiments showed that the application of 7.5 mgO3/L (z = 0.7 mgO3/mgDOC) was sufficient to recognize the beneficial effect of pre-ozonation. Treating the secondary effluent of WWTP Ruhleben a sustainable flux around 130 – 140 L/(m2h) was identified when applying pre-ozonation of 7.5 mgO3/L (z = 0.7 mgO3/mgDOC) and 8 mgFe/L for coagulation. It was not possible to demonstrate this process set up in a long term run, due to technical malfunctions. An economic evaluation showed however that for the case of WWTP Ruhleben a sustainable flux > 500 L/(m2h) is required to be competitive against tertiary treatment with polymeric membranes without ozone. This high value can be explained by the high module cost for ceramic membranes and the high DOC content of the secondary effluent, leading to increased effort for ozonation.

Various tertiary treatment processes were compared in the OXERAM project, including a polymeric membrane and a microsieve pilot plant which were installed at the Ruhleben WWTP in Berlin and operated for almost two years. To increase the performance of both these processes, pre-treatments with ozonation, coagulation and/or flocculation were tested. In order to optimize the hybrid processes and to develop a control strategy, online monitoring was implemented. After a literature review and lab trials at the Technische Universität Berlin (TUB) during the project preparation phase, two instruments were recommended. An NS500 device by Nanosight was installed in the UF membrane pilot (pore diameter = 20 nm) influent with sampling every 15 minutes before and after the inline coagulation. The particles between 50 and 1000 nm were analysed to evaluate the impact of the ozonation / coagulation or the coagulation alone on the nanoparticles below 500 nm which are most responsible for fouling. For a better reproducibility and quality of the results, samples were pre-filtered by an online metallic 5 µm filter. Particle analysis by Nanoparticle Tracking Analysis (NTA) was obtained to give reliable and reproducible information about the concentration and size distributions of the colloidal fraction in the tested treated domestic wastewater. Correlation between the membrane reversible fouling measured with the help of the trans-membrane pressure (TMP) and the concentration of particles between 100 and 200 nm were detected. Online measurements at the pilot-scale indicate that colloid peak concentrations can be compensated for by coagulation with an optimum dose of 8 mg Fe3+/L. Furthermore, a comparison of FeCl3 and PACl demonstrated that the former is more effective in colloid removal in this treated domestic wastewater. Due to the combination of pre-ozonation and subsequent coagulation, a synergy effect was determined as the combined treatments lead to a better particle removal compared to the effect of the single treatments at same dosages of O3 and Fe3+. A combination of 0.5 mg O3/mg DOC0 and 8 mg Fe3+/L leads to a total reduction down to < 5 % of the initial colloid content1. However a direct prediction of irreversible fouling was not possible. This device should be further optimized for its potential to reduce operational costs and lower solid loads and thus fouling on the membrane. A Pamas particle counter device was installed in the microsieve effluent pipe bypass and this measured the particle size distribution continuously by light extinction at a wavelength of 635 nm at 25 mL/min. No pre-treatment was necessary and it was possible to automatically clean the instrument every hour with distilled water or another cleaning solution. Piping and sensor cell maintenance was crucial to improve the quality of the results due to the high potential of the effluent water to post-flocculate. For optimization of the coagulant and flocculant mixing velocity, the particle counter results were more accurate than the turbidity sensor which did not detect any changes in the effluent water quality. The monitoring tool detected the lowest particle concentration for the optimized mixing velocity. However, the particle counter did not provide better information than an online turbidity sensor for other parameters such as the coagulant types or doses. Therefore, while it is recommended to use an online particle counter during the microsieve plant (10 µm) start-up phase to optimize the coagulation and flocculation, for routine controls an online turbidity sensor is sufficient. Moreover turbidity sensors are less demanding in terms of maintenance effort. The project showed that using the turbidity signal to adapt the coagulant dose was very efficient.

In work package 4 the influence of different treatments (ozonation, coagulation) on macromolecular organic substances (biopolymers) in secondary effluent and the effects on subsequent ultrafiltration were investigated at lab-scale. Furthermore, fouling mechanisms were intensively investigated and an analytical method was developed to observe the formation of ozonation by-products. Analyses with LC-OCD showed a significant reduction of major organic foulants (biopolymers) for coagulation while ozonation appears to transform macromolecules into compounds smaller than approx. 50 nm. With ultrafiltration tests (PES membranes) it could be shown that coagulation is capable to reduce total fouling resistance to some extent and additional ozonation can further enhance the membrane filtration process. However ozonation as a pretreatment step caused more irreversible fouling. The lowest irreversible fouling was achieved with coagulation. LC-OCD analyses showed that the transformation of organic matter by ozonation is mainly responsible for the observed increased irreversible fouling of ultrafiltration membranes. Tests with different membranes showed comparable results for pretreated secondary effluent concerning total fouling resistance. Total fouling resistance was reduced with additional ozonation compared to coagulation without ozonation. In contrast to the observations with all tested UF membranes, for the tested microfiltration membranes irreversible fouling was reduced with additional ozonation. In general, the pore size seems to be strongly influencing irreversible fouling if ozonation is used for pretreatment of membrane filtration. Intensive investigations of fouling mechanisms using filtration laws identified cake filtration as the dominant filtration process for coagulation while additional ozonation leads to increased pore blocking/in pore fouling. Experiments with secondary effluents from different sewage treatment plants in Berlin showed comparable fouling behavior for all observed pretreatments. Thus membrane filtration results generated with samples from WWTP Ruhleben seem to be transferable to other WWTPs in Berlin. MALDI-TOF-MS analyses of secondary effluent were not suitable to identify major organic foulants, neither in solution nor on top of the membrane after filtration. Consequently, MALDI-TOF-MS was primarily used for investigations of theoretical aspects of fouling by using model fouling substances. An analytical procedure for bromate was successfully developed with LC-MS/MS at TUB. With the procedure it was possible to quantify samples up to a limit of quantification of 0.5 µg bromate per liter. Higher concentrations of bromate (> 10 µg/L) were produced only at specific ozone consumptions higher than 0.9 mgO3/mgDOC0.