The pressure on water and energy resources, along with the transition towards new infrastructure, requires an integrated approach to achieve future-proof concepts. A nexus approach can contribute to this by including the interaction between water and energy. This is also known as a water-energy nexus (WEN). Furthermore, when considering decentralization as an ‘infrastructure pathway’, new solutions can be considered where more resources are locally used. Therefore, the circular economy principle was used as a baseline for this research, as it becomes possible to include both water and energy as important resources. There are currently various assessment frameworks that facilitate the decision-making process for different infrastructure pathways. However, it was found that these frameworks do generally not evaluation indicator that go in line with a circular economy approach.

This research presents a six-step generic assessment framework that can be used to evaluate different decentralized WEN systems. The first step was formed to give the opportunity in setting up the research scope. It has the possibility to either select one neighborhood as a study case of multiple depending on the research objectives. After that comes a modular step where it is possible to include different innovative technologies that are relevant for a more decentralized WEN system. The water- and energy balance can be modeled in the third step, providing insight into the (re)use of water- and energy sources on different temporal scales. Subsequently, the generic assessment framework contains 13 evaluation indicators that are divided into four themes: (1) water system, (2) value for people, (3) energy system, and (4) general characteristics. The last step includes stakeholder perspectives to prioritize and weigh the indicators.

A modern Dutch neighborhood with a high building density (City Nieuwegein) was used as a case study to demonstrate the generic assessment framework. Four scenarios were designed (reference, improved centralized, hybrid, and almost decentralized) to assess the impact of a neighborhood with more decentralized WEN systems. The case study results showed that more decentralization strategies improved most indicator scores. Using different decentralized WEN systems increases the collection and storage of local water- and energy resources. It was found that neighborhoods with more decentralization strategies have a higher complexity (e.g., monitoring and spatial limitations) in implementation. Moreover, the investment- and maintenance costs can be up to 51% higher compared to a neighborhood with minimal decentralized WEN systems. However, the outcomes of the six stakeholder perspectives, showed that the scenario with the highest level of decentralization was, in all cases, preferred.

The results of the case study showed that the generic assessment framework can be used to evaluate different decentralized WEN systems. The 13 evaluation indicators followed the circular economy principle as this favors future-proof concepts. Besides, the generic assessment framework included stakeholder perspectives so that it can facilitate the decision-making process of stakeholders. This framework can be further improved by including multi-objective optimization resulting in more scenarios that can be simulated. At last, more research is required for qualitative indicators that improves evaluating the different scenarios.

This study aims to establish a framework for the growing number of trend analyses that have been performed on river flows in Europe. Most studies apply statistical trend test to fixed periods with relatively short records. Fixed periods are crucial for trend testing, as they provide an appropriate visualisation in a geographical sense. However, short-term changes can often be inconsistent with long-term trends. This study adopts two methods of trend testing. First, a trend analysis is performed over five fixed periods with an identical end year, on a varying number of stream gauges based on record length availability. Afterwards, a temporal sensitivity analysis is performed, whereby trends are estimated on all combinations of start and end year. This method is performed on stream gauges with at least 60 years of record length. These two methods are performed on an encompassing set of hydrological signatures, with the intent to capture the temporal sensitivity in all aspects of the streamflow regime. The results of the fixed period analysis display distinct but coherent spatial patterns for each signature. Furthermore, the results reveal a considerable amount of temporal variability in all signatures. This temporal sensitivity is elucidated further by the results of the temporal sensitivity analysis, which shows that trend analyses are extremely sensitive to the choice of both the start and end year of the considered period. The study provides a reference for comparison of both past, present and future studies. The temporal sensitivity analysis is shown to be a powerful tool and is recommended for future trend analysis studies to contextualize the short-term trends.

This project explores the potential of a specific type of natural-based decentralised wastewater treatment solution in a fast-growing area: Campbelltown local government area (LGA), Sydney, with a focus on experimental design based on different densification scenarios and centralisation level of treatment scheme. The analysis and design are carried on for three scales: Greater Sydney area, Campbelltown LGA, and two samples sites in the city centre of Campbelltown LGA. for Greater Sydney area, the design is revealed as a long-term and all-rounded proposal; for the main city centre of Campbelltown LGA, the design focuses on the redevelopment and functional division of its main water bone Bowbowing Creek to serve as a treatment media. In order to experiment the schemes in detail, Leumeah centre and Campbelltown centre are designed with 6 scenarios (2 densification scenarios x 3 levels of treatment centrlisation) for each site.

The results are evaluated with the same criteria, which reveals the feasibility, pros and cons of each scenario while confirming the possibility of implementing decentralised wastewater treatment in this area although it does not bring out the same performance for all the scenarios. Further research can be carried out to simulate the long-term performance of the schemes and to test the performance with different technical components of DEWATS for the locations.

Urban wastewater systems can impact the urban ecology by untreated wastewater discharges through combined sewer overflow (CSO) events, or by partial treatment of the wastewater at the wastewater treatment plant (WWTP). CSO events can cause oxygen depletion, eutrophication, and the discharge of pathogens. The partial treatment of the wastewater causes an increased concentration of ammonium in the WWTP effluent, which can lead to toxic ammonium levels in the receiving river water. These problems can be (partially) mitigated by optimizing the existing infrastructure. Optimization of the available storage will help handle the increased pressure on the urban drainage system (UDS) and the stricter environmental regulations simultaneously. A method to control the dynamic performance of the combined sewer system is Real-Time Control (RTC). A RTC strategy controls the combined sewer system dynamically based on real-time information about the system state. This research aims to develop a RTC strategy that decreases the negative ecological impact of the combined sewer system on the river by optimizing the available in-sewer volume. By doing so, the objectives to reduce the total amount of spilled CSO volume and to decrease the ammonium peaks towards the WWTP should be met. This research is applied to the case study of Geldrop-Mierlo, this is a municipality located in the UDS of Eindhoven. The trade-off between those two objectives was explored in the Wastewater Process simulator WEST. Rainfall events with a maximum intensity of 3.1 mm/hr and higher or rainfall events with maximum intensity < 3 mm/hr and total rainfall depth of > 4.8 mm, were found to be more likely to cause DO dips. The objective function which is used in the optimization process is dependent on the forecasted rainfall and the trade-off described above. The UDS is modeled in a full-hydrodynamic (FH) model and a simplified conceptual model. The conceptual model is made to reduce the computation time. The catchment of the FH model is split up into 3 different catchments, and each is modeled as a reservoir in the conceptual model. The characteristics of each reservoir are dependent on the characteristics of these catchments in the FH model. The characteristics that are included are storage curve, outflow dynamics, and CSO dynamics. Both models are calibrated and validated. The UDS is controlled with the Model Predictive Control (MPC) methodology using a Genetic Algorithm (GA) to find the optimal solution to minimize the negative ecological impact of the UDS on the river. Both the FH model and simplified model are used in the MPC optimization. Based on the analysis of the case study, the optimization results show that the impact of the MPC procedure on the receiving river is not significant. The reasons for this are location specific, but the main findings are that 1) the hydraulic constraints of the catchments restrict the MPC procedure from working, 2) although the calibration results of the conceptual model indicated accurate results, this does not guarantee that the model is also accurate enough to use in the MPC procedure.

Under the increasing electrification of end uses in the energy transition towards more renewable integration, the electricity price keeps gaining importance on every scale from individual well-being to the competitiveness of an economy. Though scarce in the scientific literature, Long-Term Electricity Price Projection (LEPP) has great potentials in decision-making and planning, as well as complementing the long-term energy scenarios. This study takes features from the Dutch, Spanish and Danish data in five years (2015-2019) to train deep neural networks in the conditional Wasserstein Generative Adversarial Nets with Gradient Penalty (cWGAN-GP) framework, in order to project Day-Ahead Market (DAM) price series under Dutch 2050 energy scenarios.

The LEPP to 2050 is made possible by normalising the selected markets. As a result, the conditions unprecedented in Dutch data are covered in the normalised and combined data set. Generally, under scenarios with high proportions of hydrogen power in the energy portfolio, the cWGAN-GP model projects that DAM price series would have slightly lower mean and daily standard deviation than the 2019 level. Whereas much lower mean and daily standard deviation are projected when natural gas is still the fuel of the most frequent final generating technology. To explore the possible application of the projector model, the German DAM prices series in 2019 have been projected and evaluated, and the projections under Dutch 2050 energy scenarios have been used in calculating the generic profit potential of energy storage.

Five findings can be summarised from the main results. Firstly, from a literature survey and importance analyses, seven features are shown relevant to the DAM price in the combined data set, namely month of the year, day of the week, total hourly load forecast, national daily mean temperature, fuel cost of the most frequent final generating technology, hourly renewable power generation forecast and total installed renewable power capacity. Secondly, it has been found that two of the four proposed market state normalisation solutions, the Renewable Scarcity Factor (RSF) and the Renewable-Load Ratio (RLR) help the cWGAN-GP model strike a balance between price value distribution and hourly inter-dependencies. Thirdly, in this LEPP study, the cWGAN-GP model performs better than the Conditional Variational Auto-Encoder (CVAE) and multivariate Gaussian distribution (mGaus) models. Compared with the two alternatives, the cWGAN-GP model produces samples in better quality while remaining sensitive to temporal conditions. Fourthly, projections by the cWGAN-GP model are more realistic than those made by the Energy Transition Model (ETM), with price values varying continuously in smooth boundaries. Finally, the fuel cost of the most frequent final generating technology is found critical to LEPP. The annual mean and daily standard deviation of the DAM price series are expected to rise significantly when natural gas is mostly replaced by hydrogen power in the national energy portfolio.

Like many cities in sub-Saharan Africa, Kumasi, Ghana is facing greater groundwater demands in part due to rapid urbanization. However, currently Ghana does not systematically monitor groundwater, which poses a challenge in management and implementation of science-based policy. Additionally, coordination among stakeholders in Ghana’s water sector has been described as inadequate by the National Water Policy which results in greater obstacles in water resource management. In order to simultaneously address these issues, value sensitive design is implemented to synergize diverse stakeholder perspectives to develop groundwater management and monitoring strategies. Value sensitive design provides a theoretical basis for explicitly incorporating values into innovations. In order to address the goal of the study, 46 semi-structured interviews with stakeholders in the Kumasi water sector were conducted as well as literature and policy reviews.
The first research question aims to identify key stakeholders and their role in groundwater management. The results identifed over 40 stakeholder groups in the Kumasi water sector. However, the most well-known groups are Ghana Water Company Limited (GWCL) and the Water Resources Commission (WRC). In addition to these government organizations, technical experts, specifically affiliated with Kwame Nkrumah University of Science and Technology (KNUST), were cited as necessary to involve in the development of a groundwater monitoring technology.
The second research question seeks to identify practical considerations for groundwater monitoring. Here, there was a general negative view of groundwater management with challenges including limited regulation, lack of awareness for groundwater issues, and limited collaboration among groundwater stakeholders. To address these challenges, the most cited design requirement mentioned during the interviews was the need for mass education on water related concerns. Lastly, there were significant concerns among many of the interviewees about borehole drilling and the importance of informal communication between drillers and neighbors to ensure safe and sustainable access to groundwater.
The third research question uses value sensitive design protocols to create value profiles for each of the stakeholder groups. The values incorporated in this study are economic efficiency, environmental sustainability, safety, social equity, participation, reliability, and trust. Respondents were asked, through a token allocation activity, to indicate what values are important for groundwater monitoring. Although the value profiles between stakeholder groups were not statistically different, the anecdotal evidence from interviews suggests that participation is connected to other values. This indicates that participation contributes to achieving other values in the implementation of a groundwater monitoring program.
The last research question sought to identify communication tools to incorporate considerations derived from the research questions, case studies, Ghanaian water policy, and a theoretical framework based on participatory design. This resulted in the recommendation of three parallel strategies: a) multi-stakeholder involvement, b) technology development and c) a water education campaign. The programs are designed to operate in a cyclic manner based on a social learning model specific to water management. This will enable a groundwater monitoring technology to be developed (b) alongside a water education campaign (c) in the community where water will be monitored. Implementation of a multistakeholder advisory board to coordinate these efforts and facilitate collaboration will ensure a participatory process. The next steps are to disseminate findings to key stakeholders in the Ghana water sector and continuously adapt the action plans as new information is identified.

The production and consumption of electricity need to be balanced at all times. Due to the ever-growing shift towards renewable energy generation, this poses an increasingly difficult challenge. Currently, supply is regulated to maintain balance. However, there is potential to improve reliability and save costs by shifting the balancing to the demand side, known as demand response. The flexibility of water systems can play a role in this, thereby benefiting from cheaper price fluctuations and reducing operating costs.

This research investigates the IJmuiden pumping station, which drains water from the Noordzeekanaal-Amsterdam-Rijnkanaal system into The North Sea. The primary focus of the control of this system is ensuring safe water levels as it runs through areas of high economic value. The flexibility of the range of safe water levels allows costs to be minimized by selecting favourable moments to consume electricity. This simultaneously contributes to the stability of the electrical grid. This research explores the potential for a Reinforcement Learning controller for such an optimization problem, as there are some drawbacks to the Model Predictive Control methods that are currently widely used. The research objective is formulated as follows:

To optimize the control of the IJmuiden pumping station using Reinforcement Learning while complying with local water level restrictions and compare it to the state-of-the-art Model Predictive Control methods in terms of constraint violation, energy costs, and computational speed.

The Reinforcement Learning controller will use a deep Q-learning algorithm that chooses the most cost efficient control in IJmuiden while respecting the water level restrictions. To do so, the model makes decisions based on electricity prices and details about the state of the water system for the current time step as well as a forecast of 48 hours ahead. This data is provided as an input to the model.

The inputs of the model consist of historical data, meaning that the associated uncertainties are not included. The water system that the model can interact with is represented by a linear reservoir model. Therefore, the water system is influenced dynamically by the actions taken by the model. The possible actions are determined by the state of the water system.

The trained model was tested on 2 years of unseen data (data that was not used during training). Using the same test data, control plans were generated using Model Predictive Control. The Reinforcement Learning model was very successful in ensuring safe water levels. However, this did result in approximately 50\% higher energy costs. The use of the gate was close to optimal but the pumping was not clearly correlated with favourable prices and power consumption. The trained model was robust, with consistently accurate results with regards to respecting the water level constraints.

The most significant difference with the Model Predictive Control was the computation time. The Reinforcement Learning model was able to create a control plan approximately 300 times faster. This opens doors for further development of the model and increased complexity. A more accurate model of the water system can be used to take into account temporal and spatial effects and individually representing the six pumps in IJmuiden.

There are still many steps before such a model can be used for operational control, but the method has potential for such an application. Many aspects of the model can be improved as well as making adjustments to increase the usability for control operators.

From three coherent SAR images it is possible to estimate three interferograms. Combined in a circular way, the sum of the three interferometric phases is called the closure phase which necessarily adds up to zero on a single pixel level. However, if the interferograms are spatially averaged, phase consistency is not guaranteed. In most of the interferometric studies, those mismatches were assumed to be caused by decorrelation noise alone, and were either not considered or deemed negligible, eluding further investigations of its origin. However, recent publications have confirmed that inconsistent phase closures are systematic and not the exception, pointing to an underlying geophysical cause. Comparisons of the spatial signatures of phase closures with land cover maps suggest a spatial and temporal correlation that is related to the characteristics of different land cover types. Since interferometric measurements are sensitive to variations of the dielectric constant, those similarities have been attributed to dynamics in vegetation and soil moisture. A closure phase significance test developed at the Geoscience and Remote Sensing department at TU Delft aimed to increase the signal-to-noise ratio of this geophysical signal component by providing a significance ratio for phase closures. However, the sensitivity of (significant) phase closures to dynamics in vegetation and soil over different land cover types has not been assessed yet. Here we show that with enough averaging of the interferometric phase, the spatial and temporal characteristics of closure phase can be used to distinguish between different land cover types. We found that the degree of spatial averaging has a significant impact on both the phase closure values and its spatial and temporal consistency. The magnitudes of significant phase closures generally increased over low-vegetated land covers, suggesting that closure phases are most sensitive to soil moisture dynamics, whereas vegetation cover was associated with decreasing phase closure magnitudes and spatial inconsistency. Besides spatial averaging, significant differences were observed between closure phases from different polarizations. Furthermore, we found that amplitude backscatter and closure phase are spatially and temporally correlated, pointing to similar influencing mechanisms. Our results demonstrate the importance of applying a closure phase significance test and describe the effect of spatial averaging on the characteristics of phase closures with respect to different land cover types. We anticipate this study to provide useful steps towards using the closure phase for soil and vegetation monitoring in the future. For example, the findings could be used to further exploit potential synergies with amplitude backscatter for soil moisture retrieval from closure phase or develop more sophisticated methods for land cover mapping using InSAR. If not used for applications linked to land cover, vegetation or soil, being able to better predict the effect of those parameters on the interferometric phase and coherence, eventually enables to separate their contribution from other signals, such as deformation estimates. Additional research is needed to relate significant phase closures to moisture changes in vegetation.

Drinking water temperature is an essential parameter for water quality related to the physical, chemical and biological processes in water. However, in many countries, the drinking water temperature has not been taken seriously and was excluded from water quality standards. For example, in the Netherlands, the temperature of drinking water should not exceed 25℃ at customers’ taps, which is an advanced guide for ensuring water safety. Moreover, with the tendency of global warming, the drinking water temperature in the distribution system will become higher and higher without sufficient attention and practical solutions. Therefore, understanding the mechanism and cause of water temperature fluctuation is instrumental in finding appropriate measures to cope with it and improve the drinking water quality.

This thesis provides relatively comprehensive and overall ideas to research drinking water temperature. The objective of this thesis consists of three parts: i) Determining impact factors on water temperature; ii) Simulating water temperature in the distribution system; iii) Choosing measures to control water temperature. Firstly, data measurement and analytical methods were applied to determine impact factors on water temperature, and the influence level of each impact factor had been identified. Subsequently, implementing these impact factors to calibrate the water temperature simulation model to verify the model’s feasibility. Finally, the performance on reducing water temperature of three measures, porous asphalt, pervious interlocking concrete pavement, and grass cover, were compared to determine the most effective measure from the standpoint of pipe cover. The results show as following: i) Four impact factors are summarized as surface cover material, district heating pipe, shade effect and groundwater level based on the collected data; ii) It is feasible to simulate the water temperature in the water distribution system. For the model of city Almere, around 88% of simulation values had a difference smaller than 1℃ compared with measurement data; iii) Grass cover has a better performance than the porous asphalt and pervious interlocking cement pavement. Additionally, this thesis discusses limitations during the measurement and simulation process and more relative interventions to reduce water temperature.

In summary, this thesis further summarizes the various impact factors that affect the drinking water temperature and the measure to control drinking water temperature has launched from the point of pipe cover compared with previous references. These results provide guiding advice on the engineering projects of constructing and renovating drinking water distribution systems considering water temperature.

This research is part of the project "Water Efficiency in Sustainable Cotton-based Production Systems” between Solidaridad Asia and TU Delft. The project aims to increase the livelihood of smallholder farmers in the Maharashtra, India through. A socio-hydrological (SH) model is used extensively in this research as an evaluation tool. However, the baseline research indicates that the lack of stress mechanics in the SH model used in the intervention might cause inaccuracies in yield estimation. Furthermore, it has never been validated at a farmer level before. This research aims to implement the stress mechanics in the SH model, evaluate the overall performance of the model in terms of predicting crop yield, identify potential sources of errors, and give recommendations for future studies. The research uses iterative top-down approach due to the large study area and the varied nature of the 308 farmers surveyed. The research implements the water and temperature stress mechanics based on Food and Agriculture Organization's (FAO) AquaCrop framework. For the performance evaluation process, this research uses four model scores namely Nash-Sutcliffe (NS), log of NS, Mean Absolute Error (MAE), and the coefficient of determination. Furthermore, the sources of uncertainties are divided into two categories namely lack of knowledge (i.e. generated by parameter, input, observation, and structural errors) and variability (i.e. generated by climate variations). The lack of knowledge uncertainty is investigated using Monte Carlo Sampling calibration and the Generalized Likelihood Uncertainty Estimation (GLUE) concept is used to obtain the uncertainty intervals of the model. Going further, the errors are divided into residual and structural error. The latter is explained and quantified through a structural error model using a combination of qualitative analysis, Principal Component Analysis, multiple linear regressions, and projection of the data into kernel space. Then residuals between the model yield + structural error vs. the observed yield is thought to be explained by the residual errors. Lastly, the effects of climate variations to the stability of the model are evaluated.
After the initial calibration, the model scores are NS: -0.343 to -0.996, log of NS: -0.655 to -1.91, MAE: 447.1 to 553.2 kg/ha, and r-squared: 0.003 to 0.008. Because of the poor performance of the model, the uncertainty intervals from GLUE are not enough to capture the total errors of the model. However, after adjustment using the structural error model, the model scores become NS: 0.83, log of NS: 0.56, MAE: 149 kg/ha, and r-squared: 0.859. The adjusted yield calculation has a residual error as Gaussian distribution with standard deviation of 150 kg/ha. The qualitative analysis identified several factors that contribute to the errors viz. farmers' capital, irrigation behavior, and crop production process such as canopy cover growth. Lastly, there is no major instability found through the bootstrap analysis. The physical model is not performing well, especially when it is calculating yield for individual farmers over a large study area. However, the structural error model can adjust the yield prediction so that it is close to the observed yields. This indicates the poor performance is likely to be caused by the prevalence of structural errors in the model instead of the uncertainties regarding parameters, input, or observation values. Therefore, it is recommended for future research to address this first. This can be done by further study and incorporation of more crop production processes, soil water simulation, and exploratory interviews to identify patterns and more factors that can influence the errors.

In this report, the tidal method is used to estimate hydraulic aquifer parameters. The principle of the tidal method is to use head fluctuations in observation wells, caused by tidal motion in a sea or river, to determine regional hydrological characteristics of an aquifer system. As the wave propagates into the aquifer, the amplitude of the signal decreases and the phase increasingly lags behind the tide at sea. The extent of this is determined by hydrological aquifer characteristics, hence aquifer parameters can be estimated, by calibrating a model that reproduces the measured tidal propagation. Knowledge of hydraulic aquifer parameters is important since these are needed in groundwater models. The case study was performed on Schouwen-Duiveland where use was made of hydraulic head fluctuations from three groundwater observation wells and water-level observations in the Oosterschelde. First, these time series were analyzed to estimate the amplitudes and phases of the constituents present in the data. Noise in the data was reduced with the use of Pastas (a model to analyze hydrological time series; Collenteur et al., 2019) and a Butterworth filter, after which both methods were compared. Pastas was used to decompose the fluctuations observed in the groundwater to different contributions of hydrological stresses (e.g. rain and evaporation) and Butterworth was used to flatten the frequency response for frequencies that are not of interest. The use of the Butterworth filter is preferred, partly because it produces the smallest standard deviations for the amplitude and phase estimates and because it is easier to use. Moreover, especially for the wells with a low signal to noise ratio, the Butterworth filter is better at extracting the tidal signal. In addition, a graphical determination of the amplitude and phase was performed to check if this relatively quick and easy analysis gives accurate estimates of the M2 amplitude and phase as well. It was concluded that with a high signal to noise ratio and a dominantly present constituent (M2 in this case), the amplitude can be reliably estimated. Determining the phase with this method did not give satisfactory results. In the inverse modeling part, the hydraulic aquifer parameters are determined, which is done with a least-squares minimization of the observed and modeled amplitudes and phases. For the optimization, both a one-aquifer and a two-aquifer model were used, both based on the one-dimensional, multi-layer solution presented by Bakker (2019). The optimization was performed with a global optimizer. The obtained fit to the observations was somewhat poor, moreover, unrealistic optimal parameter estimates indicate that the model, to some extent, incorrectly represents the real system. This also implies that some of the model simplifications do significantly impact the results. Simplifications presumed to mostly influence the model results include homogeneity, one-dimensional flow and the use of a straight shoreline. The presence of some model error signifies that the resulting parameter estimates should be treated with care. The hydraulic conductivity and the resistance of the one-layer model were consistently estimated at their upper boundary and the storage in the leaky layer was estimated to be negligible. Therefore, only one parameter group (i.e. the diffusivity) could be estimated with the optimization. The fit of the model was not perfect, a compromise must be sought based on what residual the model minimizes. As a result, all diffusivity estimates within and around the range [1.07E6, 1.31E6] m²/day are all considered to be reasonable estimates. For the resistance of the aquitard (CU) and the storage in both the aquifer (S_sU) and the aquitard (σU) it was analyzed which parameter values gave unreasonable results. This resulted in a lower bound for CU (CU=663 days) and an upper bound for σU  (σU =1.22E-4 m^-1). For S_sU reasonable results are obtained between 2.28E-6 m^-1 and 1.87E-5 m^-1. The lower bound for the CU estimate seems to comply with estimates based on two models (REGIS-II and GeoTOP). Finally, the two-aquifer model was considered to be useless for parameter estimation. Presumably, this is predominantly caused by a small amount of amplitude and phase data compared to the number of parameters to be optimized.

This MSc thesis is a contribution to the African Water Corridor (AWC) project that gains insight into the future of water supply systems in the Sub-Saharan African small towns. The future water demand in these areas constitute a great challenge in the effort to provide safe water for everyone. Unfortunately, there are high uncertainties in the future posed by unforeseen changing factors such as global climate change and urbanization which makes it a challenge for decision-makers to develop strategies for these water supply systems in the long-term planning. The concept of resilience is introduced by many studies to address those uncertainties The objective of this thesis is to provide an approach for decision-makers to develop a resilience water supply system in small towns for long-term decision planning to answer the following research question: ‘How can a water supply system in a small African town be resilient and provide sufficient water in the future?’ This thesis presents a methodology that is resilience-based and develops reduced future supply and increased future demand scenarios for the small town in a period of 30 years. To analyze the water supply systems in small towns, Moamba has been used as a case study..

Punchiná reservoir is part of the San Carlos Hydroelectric Power Plant, situated in the Guatapé watershed. The Guatapé river is an affluent of the Samaná Norte river, which in turn is an affluent of the Magdalena river. San Carlos Hydroelectric Project uses the waters from the rivers San Carlos and Guatapé and discharges the turbined flow directly into the Samaná Norte river by a tunnel. Currently, there is flow downstream of the Punchiná dam only on the days where the spillway operates, significantly impacting the riverine ecosystem. Additionally, claims have been made about how hydropeaking causes floods in villages downstream, particularly in La Pesca village. This town is located in the confluence of Samaná Norte and Magdalena river, on the left bank of Samaná Norte river mouth. The present report deals with the multi-objective optimization of the Punchiná reservoir of San Carlos Hydroelectric Project in Colombia by considering the objectives of maximizing hydropower revenues, maximizing the ecological discharge at Guatapé river, downstream of the dam, and reducing the flood risk at La Pesca village. Four numerical methods were coded in Python to solve the reservoir routing. To solve the multi-objective optimization, the non-dominated sorting genetic algorithm II (NSGA-II) using the Pymoo framework in Python was set up, along with the use of an Explicit Euler numerical method for modelling the river routing. The simulation was performed for 3 periods (high, average and low flow conditions) within the years 2010-2017. After multiple optimization scenarios, it can be concluded that the hydropower and environmental flows are competing objectives, i.e., allocating water for environmental flow purposes from the Punchiná reservoir will always result in a reduction of the hydropower revenues. Hence, it is recommended that an incentive system is developed so that the ecosystemic services are compensated to persuade the generating companies into including ecological objectives into their optimal operation curves. In addition, suggestions on considering a bypass tunnel to let the discharge flow into Guatapé river dry trajectory while adding a turbine to take advantage of this flow are also given. The results also show that the flood mitigation objective does not result in a competing objective against the hydropower and environmental flow objectives when there are average flow conditions in the Magdalena river. Floods commonly occur during extreme weather periods whereas the optimization of the Punchiná reservoir is performed for monthly average flow conditions at Magdalena river. Thus, to assess the hydropeaking effect in the water levels at La Pesca site, it is recommended that the reservoir optimization should also include extreme flow conditions at Magdalena river when experienced.

This thesis proposes a novel design approach for a monitoring system that can detect hidden leaks in intermittent water supply (IWS) systems. Cities with IWS conditions in their drinking water network, such as Nairobi and Harare, often have a high percentage of non-revenue water (NRW) in their system. Estimations of the amount of NRW in these cities range from 40% to 50%, of which a large part is due to a leaky infrastructure. Intermittency of water supply is usually caused by a shortage of available supply, making it extra poignant to notice that these areas lose significant volumes of water. The leaks are also important locations for contaminant intrusion, which deteriorate the quality of drinking water. Additionally, intermittency of supply results in people using storage to fulfill themselves with their weekly water demand, which provides new challenges when constructing hydraulic models. Hidden leaks, which are leaks that do not appear at the surface, can be noticed in continuously supplied areas through reports of pressure deficiencies or the absence of supply. As these are regular circumstances in IWS areas, these hidden leaks are seldom noticed. Therefore, methods that are applicable in IWS systems need to be developed to detect these hidden leaks. This thesis proposes a new approach to detect hidden leaks in IWS areas with a smart hydraulic monitoring system. The approach optimizes the design of such a system in a district metered area (DMA) with IWS conditions in sub-Saharan Africa, by balancing information density and investment costs. By using as little equipment as possible, this optimization study aims to be not only scientifically and practically relevant, but also cost-effective. The methodology that was used to design the monitoring system makes use of a similar concept as the Dynamical Bandwidth Monitor (DBM), which is a smart hydraulic monitoring system that has been applied regularly in networks with continuous supply. The monitoring system consists of sensors that continuously measure flow or pressure and it compares these measurements to a range of expected values, attributing deviations from these expected values to a potential leak. A case study of Ashdown Park, a DMA with IWS conditions in Harare, was used to assess the performance of the design. The flow into this DMA and the pressure at its inlet had been monitored for one year. Two designs of the monitoring system were made, one which mainly consisted of flow sensors and one with mostly pressure sensors, to showcase which type of sensor could best be used in Ashdown Park. A hydraulic model was constructed for the DMA using pressure dependent outflow modelling. Daily demand patterns were constructed from analyzing the inflow measurements and used to calibrate the hydraulic model. The proposed calibration method assumes linear relationships between the demands and inlet pressure on one side and the pressure at a specific node and flow at a specific pipe on the other side. The range of expected flows and pressures within the DMA was calculated by Monte Carlo analyses, during which demand realizations were modelled by using a novel method which made use of a random weighted choice of demand, based on the outflow from a single tap. The ability of the monitoring system to detect leaks during different demand realizations was stored in a three-dimensional Boolean matrix, which was then used to determine the optimal sensor placement. A social and financial analysis, summarized in a business model canvas, shows more practical challenges and opportunities that could arise from implementing the monitoring system. The lessons learnt from this thesis were used to showcase whether the monitoring system could be applicable for IWS systems around the globe. Several conclusions can be drawn from the results of this thesis. The daily demand patterns in Ashdown Park showed a different pattern than in continuously supplied systems, showing less strong peaks. This could be due to a constant water demand for filling storage, leaks in the system or different consumer behaviour. The calibration method made it possible to model flows and pressures at the DMA inlet which were comparable to the measurements. The novel method to model demand realizations with a random weighted choice and a single tap capacity, showed promising results since the spread of the modelled inflow was well comparable to the spread of the inflow measurements. This standard tap capacity is especially suitable for IWS areas, since most people in IWS areas usually only have one tap directly connected to the water supply system and water end-use devices are not directly connected to the network. Furthermore, it was found that the water use behaviour of inhabitants of Ashdown Park had been more constant than the supply behaviour of the water utility. This irregular supply behaviour of the utility increased the difficulty of designing a pressure monitoring system. Using a flow monitoring system to detect leaks showed a better performance (leaks could be found on a daily basis in 25% of the pipes in the DMA) than using a monitoring system with pressure sensors (leaks could be found in 1% of the pipes). Making the monitoring system with pressure sensors dependent on the inlet pressure increased its performance (from 1% to 8.3%). Branched parts of the system were more favorable locations to place sensors and sensors at the DMA inlet were crucial for calibrating the hydraulic model. Practical barriers that were identified during this thesis were irregular operational schemes, unknown demand patterns and incomplete GIS data. Furthermore, costs can be saved as soon as leaks are detected, making the financial profitability very dependent on the performance of the system and the occurrence of leaks. The applicability of the monitoring system in IWS areas around the globe is determined by the priorities of a local water utility, its network characteristics and the ability of the local utility to overcome implementation barriers. The main limitations in this research are due to making some simplified assumptions, such as assuming a constant flow-rate from the tap in all households in Ashdown Park, and due to a lack of understanding of the local situation, since this research was performed in the Netherlands. To validate assumptions and get better understanding of the local situation, it is advised to conduct follow-up research at the location of interest. Especially a pilot project of the proposed monitoring system would likely find more practical barriers and limitations than could be thought off in this thesis and therefore bring more valuable information for the implementation of a smart hydraulic monitoring system. If prioritized, properly installed and operated, the proposed smart hydraulic monitoring system could generate substantial water savings and provide many social benefits, such as an increased access to clean drinking water and employment opportunities. Above all, it can assist a local utility with fulfilling their responsibility: supplying people with the basic need of drinking water.

Urban Drainage Systems (UDS) are one of the most vital yet, complex infrastructures that support people's livelihood in urban areas. However, due to their mainly underground infrastructure and complexity, the planning and management of UDS are usually associated with high investment, which stakeholders sometimes overlook. As long-lived infrastructures, UDS’s limited capability is being put under constantly increasing pressures. Amongst the pressures, the global effects of climate change on rainfall extremes is the most important. As climate change affects the rainfall extremes and the overall hourly and daily rainfall events, urban flooding issues are becoming more costly to manage. Several rehabilitation efforts have been made to address this issue with minimum cost and optimal performance in flood reduction by increasing the resiliency of UDS in order to minimise the duration and magnitude of urban flooding.

Rehabilitation of UDS can be done in several ways, including implementing Green-Blue-Grey Infrastructures (G-B-G measures). The combination of G-B-G measures can increase the resiliency of the UDS to withstand higher intensity rainfall by reducing both the peak flow and enlarging the capacity of the UDS system. Therefore, this thesis aims to develop a method to find the optimal way to rehabilitate an existing UDS to reduce the risk of flooding under the climate change rainfall scenarios.

The method developed coupled a hydrodynamic model, Storm Water Management Model (SWMM), and Genetic Algorithm (GA) to find the optimal solution to rehabilitate UDS. The effect of climate change was incorporated by simulating the solutions using composite design storms that represent the increase in hourly and daily rainfall extremes for 2030, 2050, and 2085. The objective function of this optimisation problem becomes the minimisation of the total cost to implement the measures for the rehabilitation of UDS, under the constraint that no flooding can happen on the system when tested against the climate change rainfall scenarios. Therefore, the decision variables of this optimisation are the size and location for each implemented measure, while the penalty cost is associated with the cost of each m3 of flooding.

Based on the analysis of the case study, the most appropriate Green-Blue measures to be implemented is Rain Barrels, Infiltration Trenches, and Pervious Pavements. Meanwhile, for grey measures, it is best to consider pipe and pump replacements and increasing the CSOs’ weirs. The optimisation was done using the developed formal method and manual trial-and-error. The results of the formal optimisation have been confirmed to outperform the result from manual optimisation using the traditional trial-and-error method. The optimal solutions proved that a combination of both grey and G-B measures produced the lowest cost to reduce flooding. Although the solutions can be adapted over time from 2030 until 2085, the results show that adaptive solutions might not be needed when the solution for 2085 is better implemented from the year 2030. Overall, it can be projected that in the future, the combination of G-B-G measures can produce an economically optimal solution to be implemented in order to achieve zero floodings in the case study location.

Transfer function noise (TFN) modelling is a form of time series analysis which regularly uses the recharge as a stress to explain the groundwater table fluctuations. Often the recharge flux is estimated as a linear combination of the precipitation and the (potential) evaporation. However, this is a simplification of the actual hydrological processes in the unsaturated zone. This is tried to be overcome by implementing a nonlinear recharge model in TFN time series models. Additionally, TFN models can use different impulse response functions, where some of them account for dispersion and retardation due to the unsaturated zone.

In this report the performance of a linear and nonlinear recharge model, inside the TFN model, are tested against synthetic time series of the groundwater table. These time series for the groundwater table are created with the unsaturated/saturated zone model HYDRUS-1D. With HYDRUS-1D, thirty-five synthetic time series are created for five different soil types and seven different unsaturated zone thicknesses (up to 5 m). The three most commonly used response functions, exponential, gamma and four-parameters are also tested for these thirty-five time series.

The results show that TFN models using the nonlinear recharge model are almost always better in estimating the groundwater table time series than the linear recharge model. This is confirmed in both the calibration and validation period. The common disadvantage of the linear recharge model, undershooting the groundwater table in (dry) summers, is not observed for the nonlinear recharge model. This can improve the forecasting abilities of TFN models during droughts.

Additionally, the nonlinear recharge model gives a more realistic representation of the fluxes in the root zone. This is confirmed goodness-of-fit parameters when comparing of the recharge flux and evaporation reduction calculated by HYDRUS-1D and the nonlinear recharge model. Especially when using the exponential response function, the recharge flux can be estimated quite well by the nonlinear recharge model. However, the nonlinear recharge model is currently not able to estimate groundwater uptake (upwards recharge) while it is observed in the HYDRUS-1D simulations.

The linear model does perform decently for shallow groundwater tables down to a depth of 150 cm since that is where large groundwater fluctuations and more days with groundwater uptake (upward recharge) are observed. The use of the gamma and four-parameter response functions significantly improves the performance of the linear recharge model. This can be explained by the compensation of these response functions for dispersion and retardation in the root zone. Nevertheless, when performing groundwater table time series analysis on synthetic time series created with HYDRUS-1D, the nonlinear recharge model is preferred to simulate the groundwater table.

This research provides benchmark accuracies for forecasting of an aggregated price of the Dutch intraday market. While point forecasts in a single-step-ahead horizon for that unresearched market provide novel insights already, the scope of this research also includes interval forecasts in a multi-step-ahead horizon. A forecasting procedure is established that organizes several stages of in-sample and out-of-sample testing so that the number of arbitrary choices regarding features and hyperparameters is kept as low as possible. It is concluded on the basis of accuracies attained by naive, regression, and artificial neural network models that the machine learning models that are capable to incorporate linear and nonlinear relationships are able to infer to a varying degree what drives intraday from day-ahead prices. Furthermore, it is addressed whether superiority in terms of accuracy coincides with what is deemed as superior in practice. A simulation of a generic system, which consists of a battery and a wind turbine located in the Netherlands, smartly dispatches stored energy according to a schedule optimized with model predictive control based on point forecasts of intraday price. It is concluded that, in general, slightly higher profits are obtained with more accurate point forecasts and that different point forecasts lead to very different dispatch schedules that vary more than 10% in terms of dispatch frequency.

Intensive urbanisation enhances warming of cities’ ambient and subsoil environment. The local drinking water distribution system (DWDS) is likewise affected. Hotspots of anthropogenic heating were perceived to influence drinking water temperature and pose a threat for microbial drinking water quality. Information on temporal and spatial relation between drinking water temperature and microbial indicating parameters was however scarce. This is especially the case for a full scale unchlorinated DWDS in a metropolitan area. Therefore, this research aimed to explore the spatial and temporal relation between drinking water temperature and microbial indicating parameters Aeromonas and Heterotrophic Plate count (HPC). 11 Years of sample data from the DWDS in the metropole of Amsterdam was explored, with the objective to possibly draw conclusions beyond this DWDS. This DWDS consists of two interconnected subsystems. Each subsystem is fed by its own treatment facility. Areas with repetitive exceedance of threshold temperature (14 oC) for accelerating Aeromonas growth, were not inextricably tied to repetitive exceedance of the Aeromonas standard for safe drinking water. Areas with repetitive exceedance of the Aeromonas standard were often linked to prolonged residence time. For regions on the outskirt of observed DWDS it was suggested that the influence of residence time was more important than the absolute water temperature to explain Aeromonas concentrations.

Globally, water demand is rising and resources are diminishing. In the context of climate change and a growing world population, a further increase in water scarcity seems inevitable. Aiming towards a sustainable future, water should be used as efficiently as possible by minimizing water losses, which can be higher than 50% in some drinking water networks. To minimize water losses it is crucial to detect, localize and repair leaks as soon as possible. Leaks can be efficiently and automatically tracked down in the early stages by using leak detection and localization techniques. These techniques are based on coupling information from flow and pressure measurements with the hydraulic model of the drinking water network. The success of such methods depends to a great extend on the estimation of the uncertain water demand in the area. The water demand within a hydraulic model is usually estimated in a deterministic fashion, which lacks the ability to realistically describe the fluctuations in water demand. To include realistic fluctuations in water demand, this study proposes a novel approach by estimating the water demand in a stochastic way by simulating it with SIMDEUM. This stochastic demand model uses information of water users and water-use appliances from Dutch statistics to simulate realistic domestic drinking water demands and its stochastic variations. With this approach, this study aims to assess the influence of a realistic stochastic demand model on the robustness of leak detection and localization algorithms. The applied case study is a residential area in the Netherlands, consisting of a drinking water network with an inflow and six pressure sensors. The corresponding hydraulic model uses stochastic demand loading conditions as inputs. The model represents the network reliably when the SIMDEUM software with local statistics is used instead of average Dutch household statistics. By conducting a Monte Carlo analysis, the influence of stochastic demand on the variability of simulated flow and pressures is determined. Artificial leaks are simulated and the influence of stochastic demand on the performance of leak detection and leak localization is analyzed. The leak detection method consists of setting up confidence intervals per sensor from Monte Carlo simulations and checking whether data falls within these intervals. The leak localization method is based on simulating artificial leaks on all possible locations in the model and comparing the resulting simulated pressures for each simulation to the observed data by using Pearson’s correlation coefficient. The position of the leak in the simulation most similar to the observations is most likely to be near the real leak position. The results show that the stochastic demand variability is strongly linked to the performance of leak detection and localization. The leak detection and localization is most sensitive during the night due to low nocturnal demand fluctuations and least sensitive during the morning peak, when diurnal demand fluctuations are highest. Moreover, the results show that the position and size of the leak significantly influence the performance of leak detection and localization. This study shows that stochastic water demand can be used to quantify the influence of realistic demand variations on the performance of leak detection and localization. Hence, it is recommended to assess the robustness of more leak detection and localization techniques by using stochastic water demand. Furthermore, it gives insight into the expected variability in pressure throughout the network, hence, can prove to be useful in optimal sensor placement.

The Climate Risk Informed Decision Analysis (CRIDA) framework incorporates the uncertainties of climate change that impact project planning, socioeconomic justification, and engineering design into a step-wise and collaborative planning process to guide a technical analyst to low-regret risk- and cost-effective solutions; Research has been carried out to demonstrate and improve, through additional guidelines, the usability of CRIDA, in a pilot for the Limari basin in Chile. The added guidelines (1) offer the analyst numerically based justifications for analytical decisions to ensure a more structured application of CRIDA and (2) improves on co-design aspects by incorporating stakeholder risk perceptions and opinions explicitly in the process. The Limari Basin has experienced an increase in drought frequency and severity over the last decades. A strategic approach for adaptation is recommended through CRIDA based on an evaluation of the future risk to climate change and the confidence in this analysis and a subsequent systematic assessments of adaptation options. The resulting strategy requires the increase of water supply robustness by adding new water sources that can be implemented in combination with flexible measures for managing demand (i.e. implementing agricultural meshes and improving irrigation efficiency) in parallel or in series to create adaptation pathways. The study demonstrated the functionality of CRIDA. While the added guidelines required more processing time, subjectivity in the method is reduced thus also reducing possible bias introduced by the analyst. In addition, overall acceptability of the proposed strategies is improved by incorporating stakeholder risk perceptions and opinions explicitly in the process.