Cape Town is a city with over four million people and a growing population. Due to three consecutive dry summers as a result of climate change, a growing population and an increased per capita water demand, the city’s main water supply was nearly depleted. Cape Town depends for 98% on surface water stored in dammed reservoirs, which is replenished by rainfall. There is a temporal mismatch between water availability and peak demand, and thus harvesting rainfall can be potentially become an additional source of water. Cape Town’s urban drainage system has 737 detention ponds, which are used to attenuate flooding in case of heavy rain events. These ponds can be used to harvest the stormwater and store it in the Cape Flats aquifer using managed aquifer recharge for seasonal availability. The complexity of retrofitting stormwater ponds into infiltration ponds calls for a systematic approach. This research offers a retrofitting framework for the context of Cape Town. The framework can be used to determine suitable detention ponds to allow managed aquifer recharge via infiltrating stormwater, and to retrofit these detention ponds into infiltration ponds. The framework consists of three phases; spatial assessment, physical assessment and conceptual design. It is highly flexible in usage due to the fact that every phase can be used separately. Additionally, the framework can be extended to include important socio-economic aspects. Following the framework standardizes the procedure of obtaining data on individual ponds, which allows for objective comparison in assessing their suitability for infiltration.

The hydrolysis of sludge solids especially for difficultly degradable sludges such as WAS is not fully understood, yet. The first-order hydrolysis rate was shown to function well for most easily degradable sludges and soluble substrates. This description for substrate hydrolysis in the context of anaerobic digestion has the benefit of being very simple and therefore applicable for many engineering applications where little data is available. On the other hand, in the last decades many studies reported that the first-order hydrolysis would need a modification to better describe the degradation of difficulty degradable solids. Guo et al. (2021) developed a cascade system for anaerobic digestion of WAS that does not seem to follow first-order hydrolysis kinetics when lowering the applied SRTs from 22 to 15 and 12 days, respectively. Based on observations by Guo et al. (2021) and a statistical analysis of the cascade system performed in the study at hand it seems that the first order hydrolysis rate constant is in fact a coefficient and that the first-order hydrolysis rate is not solely dependent on sludge characteristics and substrate concentrations. This hypothesis is tested in the thesis at hand. In Guo et al.’s study the cascade system was always compared to a reference system. To test this hypothesis and understand the kinetics of the cascade system in more detail a statistical analysis was performed for both systems from which an empirical hydrolysis model was derived. This model was implemented in ADM1 to replace the existing hydrolysis rate expression and was tested for the mentioned cascade system and the reference system. The empirical model was compared to the results of the standard ADM1 which uses a first-order hydrolysis expression. The empirical model assumed a dependency of the hydrolysis rate based on load and residence time along the cascade system to achieve a change in hydrolysis rate coefficients along the cascade system. The models were compared based on visual inspection and quantitative analysis of the simulated results. Both models showed low R² values which is likely due to the high level of detail implemented in ADM1 that does not fit to the resolution of the experimental data. However, calculated RMSE values agreed with the standard deviations of the experimental results. Therefore, the overall predictive capability for both models is given. The ADM1 managed to model the reference system with reasonable agreement to the experimental data. The performance of the empirical model for the reference was comparable. For the cascade system however the ADM1 could not fully describe the experimental at the applied low SRTs of 15 and 12 days. The empirical model in this case showed better predictive capabilities. This is an indication that a hydrolysis rate which is made dependent on system characteristics such as load and residence time might indeed have its justification and be better applicable to anaerobic digestion systems that show a concentration profile along the reactor as it is in the case with plug-flow and cascade systems.