Against the background of resource conservation, many recycled products have been developed in context of phosphorus recovery. Due to their properties some of them can be used directly as fertilizers. Phosphorus is subject to complex sorption and precipitation processes in the soil, which have a significant influence on solubility and plant availability. Therefore, in this work comparative studies on the fixation of phosphorus of recycled and conventional fertilizers in soil and experiments on solution equilibrium and solution kinetics were carried out. For the investigation of the solution equilibrium and the solution kinetics, batch and pH-stat experiments were conducted. The determined concentrations and pH values were compared with model simulations in Visual MINTEQ. To investigate the fixation of phosphorus in the soil, 8-week incubation experiments were carried out, in which soils of different properties were mixed with a respective fertilizer per batch. For the experiments two struvite products (Crystal Green and Berliner Pflanze), a thermally processed ash (Ash Dec), a rock phosphate (Dolophos 26) and the conventional fertilizers TSP and DAP were selected. Solution equilibrium of all fertilizers in water was achieved after seven days of shaking. Conventional fertilizers showed significantly higher P solubilities in water than recycled fertilizers. Because there was a very high level of agreement with the Visual MINTEQ model simulations, the software is a good tool for first orientation of the dissolution behavior of different P fertilizer products. In terms of solution kinetics, the lowest half-life was observed for struvite, with an average of 9.5 minutes. While the half-life of Dolophos increased from pH 5 to 7, no pH dependence was observed in the dissolution kinetics of the recycled fertilizers. In the incubation experiment, a higher decrease in plant-available phosphorus with conventional fertilizers tended to be observed toward the end of the experiment. Although this could be an indication of increased fixation in soil, this cannot be attributed only to chemical sorption and precipitation processes due to the increased microbial activity. Therefore, under these experimental conditions no clear statements can be made on the fixation of phosphorus of conventional and recycled P-fertilizers in the soil. Further investigations are required here.

As part of the international project Digital Water City (DWC), a water quality monitoring campaign was carried out by the research institute KompetenzzentrumWasser Berlin (KWB) in the Spree Canal during the summer of 2020. The Spree Canal is a sidearm of the river Spree and receives discharges of Berlin’s combined sewer system overflows. The association Flussbad e.V. investigates actions to make this water section usable as bathing water since in its current state the Spree Canal is under high hygienic pressure after heavy rainfall. The aim of this study was to monitor the presence of E.coli as indicator organism for faecal contaminations in the urban water. For this purpose a mobile and an in-situ deployed rapid E.coli measuring device by the company Fluidion SAS were tested in the field as well as under controlled laboratory conditions: the ALERT Lab and the ALERT System. The devices were evaluated regarding their detection of faecal contamination and their suitability to be used as a supplement or replacement for the more time-consuming Most Probable Number (MPN) reference method according to DIN EN ISO 9308-3 (1998) which is required by the EU-Directive 2006/7/EC (2006). The evaluation criteria included the agreement of the measurement results with the MPN reference method as well as the measurement precision of the devices. The ALERT Lab was able to achieve precisions comparable to the MPN method, but the results also revealed a systematic overestimation of the MPN reference. The ALERT System on

Vacuum degasification (VD) of ammonia from waste streams and subsequent production of nitrogen fertilizer can be one element of the effort towards closing the nitrogen (N) cycle and thus avoiding emissions harmful to the environment. This master’s thesis comprises a literature research for suitable substrates and laboratory experiments for the optimization of a design and experimental design of a VD pilot plant for ammonia recovery. Eight streams among the top 20 food waste streams in Europe and their associated streams, all digestates from digestion or co-digestion of animal waste streams, agricultural digestates and manures were identified as suitable substrates for N-recovery with VD. During 120 min of VD at pH 9.0, 190 mbar and 60 °C a total ammonia-N (TAN) removal rate of 0.81 ± 0.03 was achieved with an NaOH load of about 80 ml (L substrate)-1 in biogas digestate. The TAN removal rate during VD at 190 mbar and 60°C was dependent on pH with a dose response function. For efficient ammonia removal pH = 9.0 was necessary. Evidence for an ammonia volatilization inhibition at pH = 8.5 not explicable by the ammonia dissociation was found. At 158 % of the boiling pressure, 69 % of the TAN removal rate at boiling pressure was reached. Air stripping the hot substrate for 60 min lowered the NaOH load for maintaining pH 9.0 during VD by 30 %. Electrical conductivity (EC) and pH during the degasification treatment did not correlate. Alkaline hydrolysis could be the reason for pH decrease during VD at pH 9.5. In the pilot plant a pH sensor and a possibility to adjust the pH continuously should be installed. The pH for VD experiments at 60°C should be in the range 8.75–9.5. Pressures below the boiling pressure should not be excluded. Experiments with CO2 stripping should be conducted to exhaust the potential for NaOH saving.

In this master thesis the redissolution and recovery of phosphorus (P) and other valuable materials from Ruhleben sewage sludge ash (R-SSA) with HCl and H2SO4 was investigated using experimental laboratory tests. The parameters acid amount, solid liquid ratio (s/l ratio) and reaction time were varied and their influence on the redissolution of Ca, P, Al, Fe and SO4 was measured. Results showed that HCl(37%) resolved 91 ± 4 % P with an acid amount of 4 mL on 5 g R-SSA, a s/l ratio of 1:20 (acid concentration 0.46 mol/L) and a reaction time of 60 min. H2SO4(30%) on the other hand resolved 89 ± 3 % P at 6 mL on 5 g R-SSA, a s/l ratio of 1:10 (acid concentration 0.6 mol/L) and a reaction time of 60 min. Ca and SO4 showed very good redissolution of up to 100 % with HCl. The Ca redissolution with H2SO4 is clearly below that of HCl and is 63%. Al resolved to a lesser extent and reaches 65% with H2SO4 and 53% with HCl. Fe shows the lowest redissolution of 14% with HCl and 12% with H2SO4. The H2SO4 leaching experiments also showed that it is possible to separate the gypsum from the R-SSA if the leaching liquid is separated from the R-SSA after short contact times with H2SO4. The contact time, the acid amount and the s/l ratio have a big impact on the precipitation. It was possible to recover 75% of the total amount of gypsum that can be precipitated from R-SSA. In addition to this, the gypsum-free leaching liquid was then added to the R-SSA again which had already been leached, in order to dissolve P. There was no loss of P redissolution when the gypsum was extracted. This approach could not be observed in any other study. Based on the results, a recovery of 2,114 ± 130 t P/a with HCl and 2062 ± 130 t P/a with H2SO4 are possible if an amount of 30,000 t R-SSA/a is treated, which are forecasted for Berlin in the future. Since the redissolution of 91 ± 4% P requires an acid amount of 946 kg HCl(37%)/t R-SSA, this would result in an absolute annual acid requirement of 28,380 t HCl(37%) for 30,000 t R-SSA/a. H2SO4 experiments showed that 725 kg H2SO4(90%)/t R-SSA would be required for 89 ± 3% P redissolution, resulting in an absolute acid requirement of 21,750 t H2SO4(90%). Since H2SO4 has a lower consumption due to the higher concentration, is the cheaper acid of the two and has the possibility of recovering gypsum, H2SO4 would be preferred for leaching R-SSA from an economic point of view. In addition, about 75% of the gypsum can be precipitated with H2SO4, which corresponds to a quantity of 255 kg of gypsum(dry)/t R-SSA. At 30,000 t R-SSA/a, this results in an annual amount of 6630 ± 51 t gypsum that can be recovered.

Der Einfluss organischer Spurenstoffe (z.B. Arzneimittel, Hormone, Kosmetikprodukte und Pestizide) in Oberflächengewässer sorgt für eine starke Belastung der Gewässerqualität und deren Lebewesen. Ebenso werden Spurenstoffe in der Trinkwasseraufbereitung gefunden. Kläranlagen sind der Haupteintragspfad der organischen Spurenstoffe in die Oberflächengewässer. Die Elimination von Spurenstoffen kann durch Adsorption an Pulveraktivkohle (PAK) oder granulierter Aktivkohle (GAK) erfolgen. Das andere Verfahren ist die Ozonung, bei der Ozon (O3) mit der Organik reagiert und die Spurenstoffe eliminiert oder reduziert. Die Spurenstoffelimination kann online nicht gemessen werden. Weil sich die Spurenstoffe in der organischen Wassermatrix befinden, bedient man sich an Ersatzparametern (Surrogatparameter), welche diese messen. Ein bewährter Surrogatparameter ist der spektrale Absorptionskoeffizient bei 254nm (SAK254). Eine andere Onlineüberwachung ist die Messung der Fluoreszenz der organischen Wassermatrix. Es wurde gezeigt, dass es eine Korrelation zwischen UV und der Fluoreszenz in Bezug auf die Spurenstoffelimination gibt. In dieser Arbeit wird die Praxistauglichkeit der Fluoreszenzmessung untersucht und mit der bewährten UV-Messung verglichen. Die praktische Durchführung dieser Arbeit wurde an einer Ozon-Pilotanlage in der Kläranlage Schönerlinde durchgeführt. Für die SAK254-Messung werden I::SCAN- Sonden der Firma S::SCAN genutzt. Diese nutzen eine photometrische Einzelwellenlängenmessung bei 254nm. Die genutzte Fluoreszenzsonde ist die YSI-EXO2 der Firma Xylem. Sie misst die Fluoreszenz über einen fDOM-Sensor (fDOM = fluorescent Dissolved Organic Matter). Die Extinktion dieser Sonde beträgt 365 ± 5 nm und die Emission erfolgt bei 480 ± 40 nm. Für den praktischen Vergleich zwischen SAK254 und fDOM dienen Onlinedaten die in einem Zeitraum von einem Jahr aufgenommen wurden. Die Onlinedaten wurden unter verschiedenen Randbedingungen, gleichbleibende Wasserqualität-variierender Ozoneintrag und variierende Wasserqualitätkonstanter Ozoneintrag, verglichen und ausgewertet. Es konnte gezeigt werden, dass die Fluoreszenzsonde alle Werte unter den Randbedingungen analog zur SAK254-Sonde abbilden konnte. Die gebildete relative Abnahme delta fDOM war nahezu gleich des delta SAK254. Dies bestätigte die Korrelation von Fluoreszenz und UV auch im praktischen Betrieb. Zudem wurden die Auswirkungen des Foulings auf die Fluoreszenzsonde betrachtet. Fouling sind organische und mineralische Ablagerungen auf den Sensoroberflächen. Das Fouling wurde bei den SAK254-Sonden und der Fluoreszenzsonde festgestellt. Bei der SAK254-Sonde hat das Fouling Einfluss auf die Zu- und Ablaufwerte. Dadurch ändert sich die gebildete Abnahme und es kommt zu Überdosierungen. Auch bei der Fluoreszenzsonde sind Veränderungen der Zuund Ablaufdaten zu erkennen. Dennoch wird die gebildete Abnahme von fDOM durch das Fouling nicht beeinflusst und es kommt nicht zu Überdosierungen. In allen Abschnitten wurde die Elimination der Spurenstoff Benzotriazol, Diclofenac, Metoprolol und Iopromid gegenüber delta SAK254 und delta fDOM gestellt. In allen Auswertungspunkten war die Elimination über die Fluoreszenz tendenziell höher als die der UV-Messung. Weil die Spurenstoffabnahme mit der spezifischen Ozondosis (EDOC,korr) korreliert und diese konstant ist, kann die minimale höhere Elimination nur mit der „Unschärfe“ der Korrelation zwischen SAK254 und fDOM begründet werden. Anhand der Ergebnisse dieser Arbeit konnte die Korrelation zwischen Fluoreszenz und UV im praktischen Betrieb bestätigt werden. Die Fluoreszenz könnte als Monitoring-Parameter genutzt werden. Auch zur Regelung könnte die Fluoreszenz eine Alternative sein. Dennoch besteht dort noch Forschungsbedarf, weil der Surrogatparameter fDOM in diesem Forschungsprojekt nicht als Regelparameter eingesetzt wurde.

Wastewater treatment plants (WWTPs) are evolving towards a more sustainable manner, by which not only the effluent quality and operational costs but also the greenhouse gases (GHG) potential is incorporated into the assessment inventory. GHG emissions from the WWTPs include CH4, CO2 and N2O, of which the N2O is of special interest due to 265-fold CO2-equivalent. Thus, even a low amount of N2O is undesired. Aerobic granular sludge (AGS) is a promising biological nutrient removal technology due to considerable structural and microbiological distinctions compared with conventional activated sludge (CAS) flocs, leading to huge improvements of carbon footprint saving. Nevertheless, the N2O formation from the AGS reactor is likely higher than that from the CAS, in terms of sequence batch reactor (SBR) configuration and inherent complex mechanism. In addition, there wasn’t any long-term monitoring campaign on the AGS reactor focusing on N2O emissions so far.This study focusses on a N2O emission online monitoring campaign, which was carried out in a Nereda® AGS reactor treating domestic wastewater from the Berlin region, lasted more than 6 months, including two different phases, namely feeding with pre-treated and raw wastewater after aerated sand trap and 2mm sieve box. The off-gas was sucked from the top of the SBR reactor and measured with online gas analyzer. Then the emission factor (EF) was calculated based on the correlated influent nitrogen load, which was converted from the influent NH4-N concentration by fixed ratio of 0.8. During the first phase, the EF was equal to 2.97%, while during the second phase, the EF was equal to 4.52%. Generally, the EF calculated in terms of both phases was 3.71%. Compared with other long-term campaigns based on CAS and SBR processes, higher GHG potentials could be induced, which also challenges the predominance of the AGS reactors from the perspective of minimizing GHG when only considering the energy consumption into scope. In-depth analysis indicated that the hydroxylamine oxidation pathway was the most likely over the monitoring course and EF calculated during main aeration incorporate negligible fraction of N2O produced from the pre-denitrification phase. Correlation test combining two specific time frames showed the moderate positive correlation between temperature and EF, which was in contrast to what has been assumed before but coincided with the inference from the micro-level analysis of our study. The weak negative correlation ship of COD/N ratio and EF was reported for each individual phase. Due to the insignificant impact from pre-denitrification and exclusion of the post-denitrification phase, it could not be considered as reliable. In terms of narrow range of DO and no accumulation of nitrite, the weak negative correlation ship of DO and EF could not infer to any further conclusion. In addition, it should be noted that some uncertainties may distract the reliability of our results. High resolution online measurement should be applied for the determination of off-gas flow, COD and TNb concentration, instead of correlation method or infrequent laboratory analysis. The detection of dissolved N2O along the course are needed to provide more insights about the N2O formation during the process and to distinguish the contribution between aerated phase and non-aeration phase. At last, more frequent monitoring of the significant precursor nitrite and hydroxylamine is demanded to figure out the dominant pathway for AGS reactors.

The present work deals with the possibility of using capillary nanofiltration (NF) in combination with bank filtration as a pretreatment step for drinking water production. Main goal is to show the removal of selected compounds including sulphate and organic micropollutants (OMP) during the NF. Also demonstrating the benefits of bank filtration as a pretreatment step in comparison to the direct treatment of surface water using capillary NF. For that purpose, a demonstration site in pilot scale was operated as a part of the European project AquaNES at the waterworks Tiefwerder in Berlin. Since 2017 several studies (Hoff, 2017; Rohde, 2018; Jährig et al., 2018) were conducted with drinking water and anoxic well water as feed sources for the NF pilot plant. Building on these the main objective of the present thesis is to optimize the NF pilot plant operation using suboxic well water and surface water as a feed source. Operating data was analyzed and sampling campaigns were carried out to determine the retention capacity of the capillary NF membranes for selected compounds. High retention was achieved for sulphate (45 - 67 %), hardness (55 %), iron (80 %) and selected OMPs further with removal of DOC (56 – 88 %). Microbiological sampling showed a higher impact of microbial growth during the operation of the pilot plant with surface water compared to well water. Furthermore different cleaning agents were tested concerning their ability to restore the membrane performance and their costs. Regarding the specific energy consumption, different measures showed a reduction of plant energy need can be achieved. In conclusion the experiments show a higher impact of biofouling during the direct surface water treatment in comparison to the pretreatment with bank filtration. Long term operation of the tested capillary NF is possible. Under the influence of oxygen and microbial growth there is a high demand of chemical cleaning.

Thermal-pressure hydrolysis and thermal-alkaline hydrolysis of secondary sludge, from the wastewater treatment plant Waßmannsdorf, were compared based on the physical changes of the treated sludges. For this purpose, seven parameters were determined for investigation. These were: Viscosity, particle size distribution, microscopic images, capillary suction time (CST), TR after the laboratory centrifuge test, zeta potential and foaming potential. To measure these parameters, methods were developed and then applied respectively to sludges from both treatments. The thermal-pressure hydrolysis performed better than the thermal-alkaline hydrolysis in each parameter investigation. In particular, the dewaterability of the sludges after digestion, which represents an important cost factor in sewage plant operation, could be improved by thermal-pressure hydrolysis, but not by thermal-alkaline hydrolysis.