This dissertation investigates the role of structured expert judgment in quantifying uncertainties central to flood risk assessments, particularly for engineered flood defense systems. Flood risk in the Netherlands involves rare events with high consequences that cannot adequately be quantified with empirical data alone. As an alternative to the traditional approach of physics-based modeling, this research explores the use of expert judgment to address these uncertainties. Through the Classical Model (Cooke's Method), expert estimates of uncertainty are evaluated and combined with the aim of improving the credibility of failure probability estimates. Four key research questions are explored: 1) the performance of statistical tests and distributions in the Classical Model, 2) the accuracy of expert estimates for different types of uncertainties related to flood defense safety, 3) the integration of expert judgment into a Bayesian framework to reduce uncertainty in hydrological extremes, and 4) the quantification of statistical dependence through expert judgment. The findings show that while structured expert judgment effectively addresses certain types of uncertainty and dependence, its accuracy depends on the nature of the variables and the methods used to process expert data. This research demonstrates several methods to incorporate expert judgment in flood risk modeling, offering insights, tools, and recommendations for future studies and practitioners.@en

As the Western European building stock ages, attention is increasingly allocated to the maintenance of building components, particularly mechanical, electrical and plumbing (MEP) systems. Although the latter are essential in ensuring the correct operation of a building and the safety of its occupants, they remain the crafts where the most defects are observed, resulting in significant material costs. This phenomenon partly finds explanation in the shortcomings of current condition assessment methods for MEP systems, which often poorly describe the actual state of the components.

As a result, novel approaches to estimate the condition of these building elements are investigated by industry participants. Among them, Bayesian Networks (BNs) are probabilistic models that progressively gain momentum for real-life applications. In the context of the present research, their relevance is twofold: (i) their graphical structure allows to visually model influence between large sets of variables, and (ii) they robustly handle missing data. Unfortunately, like most probabilistic models, their quantification requires extensive amounts of empirical data which is extremely sparse for MEP systems. Therefore, this thesis attempts to answer the following question: "How can Bayesian Networks be applied to estimate the condition of mechanical, electrical, and plumbing systems in the absence of empirical data?"

In their `traditional' discrete form, BNs have a limited range of applications. First, they do not allow the integration of continuous variables, which for numerous physical problems is a major drawback. Second, the number of parameters to quantify discrete networks quickly becomes intractable as the number of states and parents increases, again limiting their implementation for complex systems. Therefore, Non-Parametric Bayesian Networks (NPBNs) are adopted in this research, whose formulation is based on (conditional) rank correlations (dependence) and marginal distributions associated to each of the network's variables.

To overcome the challenge imposed by the limited availability of empirical data, several studies have investigated the use of field experts' judgments for the quantification of BNs. While the elicitation of univariate distributions has been thoroughly studied, the assessment of dependence remains an emerging topic in structured expert judgment (SEJ) literature. Consequently, this thesis focuses on the development of a method for the assessment of rank correlations by field experts, whereas a lesser effort is allocated to the elicitation of the marginal distributions.

Existing research has delved into the use of two approaches for the elicitation of dependence: statistical and conditional fractile estimates. Here, the suitability of probabilities of concordance, a third type of probabilistic assessment, is investigated. Under the normal copula assumption, common in the context of SEJ, unconditional rank correlations can be retrieved from probabilities of concordance using a set of closed-form relations. Then, the individual experts' opinions are aggregated using dependence-calibration, a performance-based aggregation method gaining momentum for NPBNs. The application of these approaches to MEP systems in discussed later in this summary.

The first step in the creation of a BN is the definition of a graph. Therefore, a classification of the MEP systems is developed and constitutes the foundation of the network. Subsequently, the factors influencing the condition of the sub-systems classified previously are identified. The literature reviewed suggests a distinction between two types of relationships: those between exogenous variables (e.g. maintenance or environmental conditions) and building components, and those between components themselves. Following the identification of these relationships, a `global' graph encompassing all MEP systems arose.

Before engaging in the quantification of this network, it is crucial to interrogate its feasibility given the time span of this research and the absence of empirical data. With 23 variables and over 30 edges, the assessment of all correlations (leave alone of the marginal distributions) is practically unrealistic solely based on experts' judgments. Therefore, the remainder of the report presents a case study on air handling units (AHUs), for which the elicitation method is implemented. The graph defined for AHUs is illustrated in the figure below.

Questions for the assessment of probabilities of concordance related to the newly created graph are then formulated, taking a similar form as follows: "Two buildings A and B are randomly selected among all non-residential buildings in the Netherlands. Given that the AHU in building A is maintained more regularly than in building B, what is the probability that the coils are in better condition in building A than building B ?"

Similar questions were created for each of the network's edges and presented to a panel of five experts, resulting in five individual correlation matrices. As mentioned previously, the experts were then evaluated using seed questions in the context of dependence-calibration. Additional questions were thus asked to the participants with regards to precipitation in the Netherlands, a choice motivated by the absence of data related to AHUs and mechanical systems for calibration. The respondents' calibration scores were then calculated using their assessments on the seed questions and the correlation matrix retrieved from empirical data. Finally, a combination of the experts' dependence structures was built using their calibration scores in a weighted average, resulting in a unique set of correlations which were implemented in the NPBN.

Lastly, two of the five experts consulted previously participated in the elicitation of the marginal distributions, either by the direct provision of the distribution or through answers to qualitative statements. The resulting model includes both marginal distributions and (conditional) rank correlations, and is ready to be validated.

To conclude, the NPBN is validated. While the lack of empirical data prevents the quantification of the model's predictive validity, a scenario analysis is performed to observe its output under different input combinations. It reveals that the exclusion of the environmental conditions from the network results in unrealistic outcomes, thus refuting an assumption made earlier in this research. Moreover, a global sensitivity analysis is conducted based on Sobol's method, which demonstrates the high contributions of all inputs to the outputs' variances. Consequently, evidence on any of the inputs substantially reduces the uncertainty in the output distributions, a comforting conclusion on the relevance of the chosen factors.

The final result of this thesis is a flowchart illustrating the construction process of a Non-Parametric Bayesian Network. It provides academics and practitioners with a foundational framework for the creation of Bayesian Networks, irrespective of the quantification method selected. While this thesis proposes the implementation of a particular expert-based elicitation method, the most suitable approach should be chosen with regards to the system modelled.

The uncertainty paired with the effects of climate change impacts the design aspects of hydraulic structures. To comprehend the uncertainty of the evolving conditions, methods and models that differ from the traditional ones could provide a better understanding of the uncertainty and potential risks associated with a design.
Hydraulic structures are designed to withstand loading conditions related to extreme environmental scenarios. Such natural hazards are characterized by the shared dependence of several random variables. Multivariate probabilistic design clears the way for exploring dependencies between design variables. A multivariate approach models the joint occurrence of a combined condition, revealing new insights e.g., the correlation between wave height, wind speed and water level. Next to incorporating new models that estimate the uncertainty and changing conditions related to climate change, there could be a leap forward could in design methodology. Adaptive design strategies can provide a way to mitigate over-designing hydraulic structures. To cope with the uncertainty of sea level rise, this methodology aims at handling the uncertainty in long-term decisions by emphasizing adaptivity in a design.
This thesis aims to develop a multivariate probabilistic tool to provide insight into incorporating adaptive design strategies, and to gain extra information that contributes to assessing the feasibility of the Wide Green Dike concept given the uncertainty of sea level rise.
A multivariate model using a vine copula is created to approximate the joint behaviour of the extreme hydraulic boundary conditions of the Wide Green Dike project. The results showed that by accounting for the joint behaviour of the extremes and the uncertainty and variability of design conditions, significant reductions in the design could be realised compared to designing in a deterministic manner. The deterministic design increased the cross-sectional area of the dike by 150% compared to the multivariate probabilistic design with a sight year of 2150.
To explore the adaptive design strategies, several strategies are created using the multivariate probabilistic design approach. The strategies consist of an initial design with a certain sight year and one or multiple adaptions during its lifetime up to 2150. The strategies are assessed based on the business case, up-scaling of the clay refinery in the Ems-Dollard estuary, the impact on the Natura 2000 area and feasibility. It was found that incorporating an adaptive strategy could promote the business case and help mitigate the risk of over-designing.

Floods are among the most common and devastating natural hazards worldwide. A key challenge is to select and implement efficient measures to reduce the flood risk, within limitations of budget, time, space, and societal acceptance. Interactions between joining rivers in flood plains and cities lead to several challenges in flood risk studies. Typical ‘critical’ events of such river systems may differ as small (large) rivers are affected by short duration-high intensity (long duration large volume) rainfall events. The challenges are 1) to define design flood conditions (design events) for systems on which the flood resistance measures can be based, e.g. dimensions of measures like levees or reservoirs, and 2) to understand and determine the risk and consequences that a flood can cause. A multivariate analysis is needed to account for the interactions and statistical dependencies (Bender et al., 2016). Copula functions have been applied in different fields as a multivariate method. Nevertheless, its application in risk analysis for river confluences still needs further evaluation.
In this thesis, the objective is to develop a methodology to determine design flood events that account for the different statistical dependencies and interactions between joining rivers, and that balances the required simulation time with the required accuracy of the results. For this objective, we evaluated the different statistical dependencies and interactions of joining rivers according to the catchment characteristics for each model domain: meteorological, hydrological, and hydraulic. First, identifying how the extreme precipitation events (meteorological) of neighbouring catchments are correlated, and evaluating the differences whit correlations of extreme river discharge events (hydrological). Second, we performed hydraulic simulations at a confluence to evaluate the hydraulic interactions of the joining rivers and the flood impacts, from which a response function was obtained. Subsequently, we evaluated the flood risk by implementing an approach to sample combinations of discharges of joining rivers, and determining the flood impacts from the response function. The developed approach consists of three main steps: 1) the selection of the extremes sets at the confluence, 2) the estimation of the copula parameters and 3) Monte Carlo simulations where the discharges of joining rivers are sampled from their respective marginal probability distributions, and the flooded area is calculated by using the response function. The dependence between the two joining rivers is taken into consideration by using a copula (Gaussian, Gumbel, or Clayton) to construct the joint distribution of the confluence from the marginal distributions.

In the last decades, renewable energies gained an increasing interest due to the environmental awareness of people, especially in the developed countries. Moreover sustainable resources represent a long-term investment full of possibilities of use.
In the Netherlands a technology which is getting more and more popular to produce green energy is represented by offshore wind turbines (OWT). However, during the installation of the structural elements for these systems, the risk of noise pollution and animal harm is an issue that has to be considered.
Previous works developed models to predict the noise emission and propagation during the construction phase, however the uncertainty related to the environmental properties has not been yet fully investigated.
Since the model characteristics are uncertain, so will be the prediction of the noise.
This thesis aims to fill this gap, investigating in the underwater soil property uncertainties and the resulting variation in sound predictions. The main goal of this work is to settle a sounding methodology to model the soil characteristics and interpret the sound levels.
In the first part of this work, the soil uncertainties are treated.
A framework on how to use measurements from cone penetration tests (CPT) and obtain mechanical and dynamic soil features is presented.
By means of statistical approaches, the procedure to define the optimal depth for different homogeneous layers (given the software used for the noise prediction) is described. Another topic dealt with is the definition of proper characteristic distributions and the choice of the optimal one representing the available empirical measurements.
Finally a procedure to generate random samples for the analyses that will follow is shown.
An important feature presented is the use of the correlation between the properties to define copulas. The samples then are not completely random and independent, but instead combinations that are more likely to appear are obtained.
In the second part, the results of several analyses are presented.
The steps to treat the sound levels and obtain the probability density (and cumulative) distributions are discussed. These results will help in estimating the probability of exceeding a particular defined sound level.
With this information, additional measures and precautions, as noise barriers, may be adopted in the installation of the pile to prevent exceeding the threshold.
The correlation between soil properties and obtained sound levels is investigated, in order to highlight if there are soil properties that greatly affect the outcomes.
The insight obtained may help in determining which particular features need a careful estimation, both by more accurate measurements or new techniques. Another benefit related to the detection of parameters not affecting significantly the sound levels, is the reduction of simulations necessary to cover enough combinations. That is because if a property
can be neglected in the generation of samples and be taken as a fixed value, less combinations are needed to be considered.
Finally, a comparison between the obtained estimations and empirical measurements in the North Sea is made, to test the validity of the framework proposed.

Climate change causes cities to deal with increased temperatures and more frequent weather extremes. Heat waves will occur more often, becoming a more prevalent issue in especially urban areas. The quantification of heat stress is a first step to define mitigation measures. For that purpose, a standardised method to assess the spatial influence of surfaces on the Physiological Equivalent Temperature (PET) was developed. This study aims to reshape this model into a statistical dependence model which is more flexible regarding missing data. To this end, we used a Non-Parametric Bayesian Network (NBPN). We created a model driven by both data and expert knowledge, that is capable of dealing with input data layers with a grid resolution up to 20 m. Results show that training the model with only 20 sample points did not affect the performance considerably, compared to using 2,000 data points. Inclusion of a layer with sky view factor mainly improves the estimation of observations in the tails of the distribution. The model predicts the PET with a Mean Absolute Error (MAE) of 1 to 2 ºC, dealing adequately with missing data layers. With this limited amount of necessary input, the NPBN in our study helps in standardising the assessment of heat stress outside the borders of the Netherlands. Also, our model offers a framework to make a first assessment regarding the effect of NBSs on heat stress.

The offshore wind market is developing fast due to climate change. To ful fil in the growing demand for offshore wind market, one has to look for floating offshore wind solutions as nearshore shallow waters are depleting. Several types of floating wind structures can be distinguished in the following categories; Spar, semi-submersible and Tension Leg Platform (TLP) structures. Blue water developed a floating wind TLP, the BLUE-STAR, which is still under development and has not yet been applied in offshore wind projects.At the moment there are many factors unknown about the concept. This in comparison to semi-submersible structures of which more knowledge is available. The aim of the thesis is to describe if the TLP concept has an advantage over the semi-submersible structure and if the newly developed TLP concept can be a viable solution. For this, a simulation model to simulate the logistics of both structures, is developed. First, different types of offshore wind turbines are classified, including TLP and semi-submersible structures. This is followed by challenges the logistics of offshore wind farms are currently confronted with. This points out that the most important challenges are due to substructures, environmental conditions and T&I. Next, the logistic process of both the TLP and SSB structure are described. This is followed by a literature study which is divided into literature on weather conditions and literature on operations and logistics, in which the analytical approach and simulation-based methods are described.Discrete Event Simulation is used for the simulation model in Matlab. This is followed by an extensive description of the logistic process in general. The second part of the chapter elaborates on implementation of both structures into the described model. The fourth part of this thesis, elaborates on implementation of the simulation model by explaining the decisions and assumptions made for the simulation model. Furthermore, this part also discusses the inputs of the logistic process simulation.  The weather data provided by Blue water for the use of the simulation model is presented and this part of the thesis gives an evaluation of the simulation model. Based on this evaluation, is it concluded the simulation model functions correctly. In the final part of this thesis, the results of the logistic simulations of both structures are compared and a  sensitivity analysis is performed. For the sensitivity analysis, 4 cases are studied, these are: varying wind speed and wave height, variation of transport duration, seasonality and varying team performance. The aim of this sensitivity analysis is to study the influences of different input conditions.  Finally, results of the simulations and performed sensitivity analysis indicate that the newly developed TLP concept is not a realistic alternative compared to the SSB structure.    

Offshore and coastal infrastructure must be designed to withstand loading conditions that among others arise from extreme environmental conditions. Physical processes such as storm surges, tides, currents, and waves play an important role in the design of these structures. Several variables characterize the relevant physical processes (e.g. significant wave height, mean wave period, water level, wind speed) and a thorough analysis of these variables is required when dealing with offshore and coastal dynamics, durability, and reliability assessment.

Traditionally, a critical loading condition is defined by characteristic values of environmental variables that are determined based on the highest loads previously experienced. Modern design methods seek to derive loads that correspond to specified reliability by considering the frequency of a specific loading magnitude.Traditional design approaches do not take into account the interrelations and dependencies among the variables of interest. Hence, wrong representations of the physical processes and unnecessary conservative representations of the design loads might occur. This may severely limit their effectiveness and can lead to expensive and inappropriate decisions. Multivariate frequency analysis approaches currently receive much attention within the academic community, however, advanced statistical concepts such as regular vine copula are slow in being taken up by engineering practice.
This thesis presents a practical assessment and further development of a vine-based methodology, used for the derivation of design values, in continuation of the work performed by Sell´es Valls (2019). Regular vine copulae are advanced statistical models for high dimensional distributions using (conditional) bivariate copulae as building blocks. This study contributes to bridging the gap between the academic community and engineering practice on one hand, and on the other hand, contributes to a better understanding of the potential added value of incorporating dependence information in the design process of coastal and offshore infrastructure. It has a conceptual point of view where the concept of using dependence information by applying advanced statistical techniques is explored and the required adaptations throughout the entire design process are evaluated.
In this research, it is found that the multivariate vine-based methodology can be successfully incorporated in the design process of a breakwater structure, and on average results in minimal required dimensions of elements of the cross-sectional design that turn out to be smaller and the corresponding costs up to 25% lower compared to the univariate traditional approach. This is realized by adapting the framework enabling an offshore-nearshore transformation of the wave conditions using SWAN software. Furthermore, the theoretical framework is extended by introducing Kendall’s measure providing a suitable definition of the critical region from which the critical loading conditions can be obtained. It is concluded that the vine-based approach could act as a tool providing extra information about the behavior of the system and insights on the degree of conservatism of the traditional approach. The considered role of the vine-based methodology in the design process of a breakwater structure (or coastal infrastructure in general) is to provide the practitioner with additional insights supporting the traditional design approach and possibly optimizing the design.

Non-linear finite element analysis (NLFEA) is a powerful numerical solution method that can enhance accurate determination of the structural resistance for a more efficient design. However, the implementation of NLFEA for the design of reinforced concrete structures is lagging behind as related uncertain- ties have not been quantified adequately yet. Multiple studies have been conducted to evaluate the effect of modelling choices, which has led to the RTD1016 Dutch Guideline. This guideline enables a better quantification of the uncertainties related to NLFEA for the proposed solution strategy. In this research, a full probabilistic approach is applied to improve the quantification of uncertainties related to NLFEA, and thereby enhance its application for the design of reinforced concrete structures. In particular, to the ultimate limit state (ULS) of simply supported beams subjected to both ductile and brittle failure modes. To achieve this goal, 48 benchmark beams were selected from literature for calibration purposes. Material induced uncertainties of concrete and reinforcement were incorporated through an optimized Latin hypercube sampling strategy. The beams were modelled in a 2D plane in software program Diana based on a total strain crack model and Von Mises plasticity. A displacement-controlled analysis was performed to determine the numerical ultimate resistance. In total, 1104 analysis were performed of which the model uncertainty was quantified and the failure mode was determined by the ductility index. Based on this, a global reliability method was defined as a function of the failure mode. A comparison with existing reliability methods was made in terms of accuracy and robustness. Furthermore, a standalone multivariate non-parametric Bayesian network (NPBN) was developed that allows for extensive reliability assessment possibilities. The research has shown how a full probabilistic approach with benchmarking can be applied. A reliability method as a function of the failure mode was proposed that showed improved efficiency compared to existing reliability methods (GRF,PRF,ECOV). For a 50 year design lifetime, a mean unity check of 77% was attained for the ductile failure mode and 66% for the brittle failure mode. For specific types of concrete and reinforcement, even higher efficiency can be obtained by reduced coefficients of variation of the material parameters. Furthermore, a NPBN was constructed which describes the NLFEA behaviour of reinforced concrete beams. Additional research is necessary to improve the model application, but it has been demonstrated how such a model can be established and used for reliability assessment of reinforced concrete beams. The findings of this study suggest that the design of reinforced concrete beams by NLFEA can be applied for efficient design, while respecting safety standards. The full probabilistic approach enabled an improved quantification of the design resistance. Thereby, this research contributes to the implementation of NLFEA for the design of reinforced concrete structures.

Implementation of nature-based solutions (NBS) in flood defenses is hindered by a lack of probabilistic tools and design guidelines that can be used to assess spatial and temporal variability in these biophysical systems. It is well established that nature-based elements, such as vegetation, attenuate waves, capture sediment, strengthen the subsoil and invoke numerous ecological benefits. This thesis proposes a conceptual framework to design and assess hybrid flood defense (HFD) systems within the context of current Dutch design guidelines: `wettelijk beoordelings instrumentarium' (WBI). The framework builds a dynamic probabilistic tool (DPT) to assess the effect of temporal variability of NBS elements on wave loading. The framework was applied in a case study at Hellegatpolder to investigate temporal effects of nature-based elements on nearshore wave heights during storm conditions. Data was collected specifically for the idealized HFD system at Hellegatpolder. Vegetation and nearshore wave height data was unavailable and therefore, an autocorrelation function was developed to sample temporal vegetation data. A numerical XBeach model was constructed to model nearshore wave heights for a bare and vegetated foreshore, resulting in a uniform database. Development of a static and dynamic Bayesian network allowed dynamic probabilistic modelling of nearshore wave heights. Application-specific model settings combine normative hydraulic storm data with the database and probabilistic models. The developed DPT was applied to model nearshore wave heights for a bare and vegetated transect for each vegetative season.Numerical modelling using XBeach was applied to model 7 years of data with a temporal resolution of 30 minutes for both bare and vegetated foreshore scenarios. The numerical results conclude that vegetation at Hellegatpolder attenuates waves with an average of 54%, where wave attenuation was 1.7 times greater in July compared to December. The effect of bathymetry morphology on nearshore wave height was found insignificant for short time scales.A static Bayesian network (SBN) was built to model nearshore wave heights in a fixed point in time. The validated SBN was able to model nearshore wave heights with 90% accuracy. A dynamic Bayesian network (DBN), was created to model offshore hydraulic parameters time series. DBN achieved an accuracy of >85% for short time scales (<25 hours). Utilizing the DBN for long term modelling resulted in progression towards the mean value of the marginal distributions. Statistical validation of both models rejected the representation of the dependence structure using only a Gaussian copula.Application settings were defined in MATLAB to manipulate the database, probabilistic models and normative hydraulic storm conditions. The configured DPT was run to model nearshore wave heights for a bare and vegetated foreshore specifically at Hellegatpolder. During storm conditions average wave attenuation due to vegetation was 45%. Average wave attenuation results were thus lower than those available in the numerical dataset (54%). Moreover, wave attenuation was 3 times greater in summer compared to winter months. Furthermore, the DPT resulted in dynamic wave loading (i.e. varying through time) compared to the static WBI loading conditions. Resulting DPT wave heights were 82% lower during winter months than the normative loading defined by the WBI.These results show the effect of variability on wave attenuation. The percentages and factors are expected to differ for other locations. Nevertheless, the obtained results clearly illustrate the significance of temporal modelling of HFD systems. The method presents a novel conceptual framework to include the effects of NBS elements in the design and assessment of flood defenses. The conceptual framework, numerical and probabilistic models can be applied for other HFD systems, enabling engineers to assess flood defenses more realistically - a critical step in the implementation of NBS in design guidelines.

Movable bridge decks experience critical expansion in summer, leading to uncertainty and unpredictability in its availability doe to improper docking and safety hazard. If the bridges are not cooled soon, the inertia of expansion stays, causing prolongation of availability problems. Structural health monitoring of such bridges with a predictive maintenance approach can help plan remedial measures on the exact day and time. For the efficient design of such a structural health monitoring system, a combination of sensor system data and weather API data is tested. A machine learning approach of Gaussian process regression which can give the results on the prediction of critical expansion of bridge deck has been evaluated in this research project. Finally, a check on the transferability of the prediction model is conducted by application on another bridge data and its performance is discussed. Scenario analysis with savings in cost per scenario is also conducted with varying levels of potential unavailability penalty costs, which could be levied on an asset manager of a bridge if the prediction models of sensor system data set or/and weather API data set gives incorrect estimation. Such an analysis is done to justify the use of the prediction models in assumed scenarios to predict expansion of movable bridge deck

The past decades, the increasing availability of data has paved the way for a new, data-driven generation of models. This research proposes a non-parametric Bayesian network (NPBN) to model hydrologic processes. The Bayesian network (BN) is a directed, acyclic graph in which the variables are represented by the nodes, and the conditional probability distribution between variable pairs is represented by the arcs. NPBNs are computationally less expensive than many conceptual hydrologic models and are sufficiently flexible to be able to handle different continuous data sources. The goal of this thesis is to make an NPBN for a lowland catchment and test its performance. The case study concerns the catchment of the Vledder, Wapserveense and Steenwijker Aa. This catchment makes this research the first one in which an NPBN is comprehensively implemented for (1.) a single catchment in which the catchment processes are modelled, (2.) a Dutch catchment, and (3.) a lowland, partially managed, catchment. For the BN model, seven hydro-meteorological variables have been selected for the model, complemented by the target variable, which is the monthly maximum daily average discharge (MMDAD). The aim of the BN is to be able to accurately predict the MMDAD, and the Kling-Gupta efficiency (KGE) acts as a performance indicator by which to optimize the BN’s parameters. For this thesis, the Gaussian copula was selected to be implemented for all variable combinations in the BN, because this type allows for the use of the multivariate normal distribution to calculate a conditioned network. The fit of the Gaussian copula to the data is tested in this thesis. This method is far more convenient than the alternative called the vine-copula method and most likely gives a better fit than the other alternatives. Three distributions are compared to model the marginal distributions, of which the Gaussian mixture model has been selected. This function extrapolated too little, so a novel alteration function has been proposed to shift the predictions. Several other parameters in the BN have been analysed as well. A sensitivity analysis has been performed to understand what influences of artificial errors would be. In general, random errors have a low influence on the prediction of the model, whereas new systematic errors have a larger influence. Criteria for a practical, well-performing BN have been presented and a strategy to create such a model that satisfies these criteria has been assembled. This strategy left the selection of some connection implementations up for interpretation. The chosen implementation has been decided based on which implementation produced the best predictions of the relevant variable within the network. The final model gave a median, k-fold tested KGE of 0.73 when predicting the MMDAD. It is also analysed how well the MMDAD is predicted if not all other variables are fixed. Another novelty is that a BN model is benchmarked against a SOBEK model, a neural network, and a multiple linear regression model. Compared to these models, the BN performs well. Moreover, all these other models lack some advantages that the unsaturated BN has.

The design of offshore and coastal hydraulic structures is very much dependent on the hydraulic boundary conditions, such as significant wave height, mean wave period and wave direction, among other parameters. A proper design value of these parameters is required during the design process, based on the corresponding safety philosophy and the lifetime of the structure. In this context, an extreme event is characterised by a combination of unfavourable parameters. However, the interdependencies between the parameters are not always accounted for in the design process, despite the fact that some parameters are clearly related. This potentially leads to an overly conservative or optimistic design.

To complicate matters further, a given sea state might consist of a combination of wind wave and swell systems, sometimes coming from different directions and with different spectral shapes. Different combinations of crossing wave systems might lead to the same total significant wave height, mean wave period and mean wave direction. Only analysing the total wave parameters might oversimplify the situation in the presence of combined wave systems. In this thesis a methodology has been developed to establish extreme offshore wave conditions given the presence of these combined wave systems.

A time series that partitions the total wave into a wind wave- and swell component is used as input for the analysis. The location of interest being off the coast of southern Brazil, where combined sea states are observed regularly. The main objective is to compute design values for all wave parameters of interest. With these design values a number of extreme offshore sea states are described in terms of a single total wave system and equivalent combinations of two wave systems. The former resulting in a single-peaked wave spectrum and the latter in an equivalent double-peaked wave spectrum. The extreme offshore sea states are transformed to the nearshore and compared. The single-peaked and equivalent double-peaked wave spectra may result in very similar values for the wave energy nearshore, albeit with different spectral shapes and directions. For the investigated directional combination, this means that the more elaborate approach with two wave systems potentially affects the design of coastal infrastructure if it is sensitive to spectral shape and direction, although the uncertainty of the result is not quantified. Equivalent wave systems could be compared for other directional combinations in future research to investigate if the more elaborate approach results in a more cost-effective design of coastal infrastructure.

The quality of the multivariate vine copula model, used to compute the set of design values for the wave parameters of interest, is assessed in multiple ways. It is recommended not to pick a single set of design values at a point of high joint probability density. Instead it is suggested to use conditionalised samples from the vine copula model to determine the most unfavourable combination of load parameters, which has to be evaluated case-by-case.

PIANC has published several working group reports related to the design of fender systems. The work of PIANC WG33 is widely accepted by the industry and has been used to design marine structures worldwide. However, the existing design approach does not distinguish uncertainties in fender engineering, e.g. uncertainties related to vessel sizes, berthing velocities, and berthing angles. This paper aims to show how to take into account some of these uncertainties into fender design using a reliability-based approach. The influence of multiple fenders contact and multivariate dependence between vessel size, berthing velocity, and berthing angle on the failure probability of a fender system was analysed. These correlations were modelled using a Vine-Copula, while the contribution of multiple fenders contact was investigated by performing simulation. Furthermore, the failure probability of the fender system was determined using the First Order Reliability Method and Monte Carlo simulation. The results show that uncertainty in berthing velocity, the effect of multiple fenders contact, and dependence between design variables largely influence the reliability of a fender system. It is highly recommended to incorporate all these aspects into the design approach to accomplish a cost-effective design solution. The key findings of this study can be used to update the existing design approach of fender systems and help to interpret the berthing records collected by Port Authorities.

New and existing bridges in the Netherlands must abide by structural safety codes, such as the Eurocode. In this code, structural safety is expressed through the reliability index 훽. For certain reference periods a threshold value for 훽 exists. When applying prescribed load models given in the Eurocode, the structure is guaranteed to at least fulfil to this threshold value. However, these prescribed load models are deterministic in nature and can be rather conservative for bridges in urban areas. This thesis focusses on creating a probabilistic load model based on actual traffic loading by making use of a camera system that registers license plates to check whether the vehicle is allowed to enter the inner city of Rotterdam due to environmental zones. From this camera system data, technical information such as wheelbase, legally allowed axle loads, gross vehicle weight and such can be extracted since they are coupled to license plates. This technical information is then used to create load models based on actual registered traffic. This load model represents trucks as point loads with interspatial axle distances. In total, one year of collected data by the camera system is stored, called the LP data. This load model is then used in a probabilistic reliability analysis as a load variable input. When comparing the LP data with available weigh-in-motion (WIM) data from two measurement locations in Rotterdam, it turned out that the LP data does not incorporate under- and overloaded axles and was overestimating the accompanying reliability index. Hence to account for this, an axle load factor 휂௜ is introduced to simulate under- and overloaded axles. This factor 휂௜ is based on the WIM data and is different for each vehicle type. With the use of this factor, a second, improved load model is constructed. This is referred to as the modified LP data. A third and final load model was constructed from the available WIM data, called the WIM model. For each of these three load models the load effects were calculated, and distributions were fitted accordingly for simulating several 25 year periods of traffic. The output of these load models is a loadeffect maxima distribution that can serve as a direct input in a probabilistic reliability analysis. With these three load models, a hypothetical slab with a span length of 10 m was probabilistically analysed where the LP model, the modified LP model and the WIM model resulted in reliability indexes of 4.8, 4.1 and 3.7 respectively. When compared to the requirement in the Eurocode, all load models comply. Concluding from this, the modified LP model suggested in this thesis can be used as a load input in a probabilistic verification for this very considered bridge location. For this load model to be applicable to multiple bridges, more research must be done since only one location was considered in this thesis. However, the suggested approach to construct load models based on license plates can be used verify the applicability to multiple bridges.

This report describes the development of an outfall structure in Panama City. At this location a Matasnillo river enters the bay along the Cinta Costera. Since this river is polluted it causes a bad water environment and smell at the river mouth area such that it is unable for recreation. Boskalis wants to improve the area in co-operation with local authorities. To do so a solution had to be found for the problems caused by the river, while the peak river discharge during rainstorms will not cause flooding in the area. Several concept designs were developed and in the end a hybrid design, consisting out of an outfall pipeline and culvert, was selected as most viable solution. This design discharges the polluted water through the outfall pipelines several kilometers offshore. During rainstorm events the surplus in river discharge is discharged through the culvert structure. The design is substantiated by functional and structural design calculations. In the end a cost analysis was made.

The increasing frequency and intensity of extreme events due to global warming and climate change is increasing flood risk. To act, rather than react, nature-based solutions (NBS) involving vegetation and wetlands are being explored on top of conventional solutions like dikes. WHY? There was a dire need for global study quantifying the potential of vegetation in reducing flood risk and eventually make a decision support tool which enables quick assessments about flood risk reduction in a vegetated hydrodynamic system. WHAT? The developed tool can predict flood risk anywhere in the world without rigorous modeling through user defined conditionalization of in-situ hydrodynamic or vegetation characteristics. HOW? Multivariate dependence among parameters of schematized system can exhibit characteristics of vegetated hydrodynamics. To ensure global representation of vegetated hydrodynamics a copula-based multivariate stochastic model has been developed which caters global ranges of each parameter, their probability distributions and the inter-parameter dependencies through ranked correlations. Numerical modeling has been carried out through XBeach non-hydrostatic model by resolving full spectrum of high and low frequency waves. A non-parametric Bayesian network-based flood risk prediction tool has been developed from the synthetic dataset developed from the simulations. SO? Bulk results conclude that saltmarshes attenuates waves by 87% and mangroves by 94% as a mean value. Wave attenuation, flood risk reduction and wave run-up manifests maximum dependence on offshore wave height, water depth, drag coefficient, vegetation height, frontal width, and forest length and least on offshore slope and vegetation density. NOW? The flood risk prediction tool would help decision makers in implementing NBS, in making better informed decisions about early warnings and policy making related to flood risk reduction and climate change adaptation by incorporating vegetation. NOVELTY? To the author's knowledge no such study exists which captures natural variability of hydrodynamics and vegetation together in a probabilistic model over global scales. Additionally, no such study exist which applies non-parametric Bayesian networks to predict flood risk. The dependence modeling of global vegetated hydrodynamic environments is also unique which skims out the most critical parameters.

When designing hydraulic structures, engineers follow a structured and predetermined set of design phases, referred to as the design process. Within this process, the preliminary design phase is a highly iterative process, which requires many calculations. From these calculations many different initial designs of the structure are derived. Generally, these possible configurations are compared to the design requirements. An important requirement is the safety of the structure over its lifetime. This can be reasonably accounted for by applying different reliability methods. Within these methods, different elements of uncertainty are considered. In the preliminary and final design stage, depending on the engineer, the quantification of these uncertainties and the safety of a structure during its service lifetime are often based on expert judgement, partial safety factors or are not considered at all. It is generally perceived that a full probabilistic approach gives more insights into the reliability and safety of the design. Due to the complex nature of this process and the mathematical workload, computer automated designs are becoming more popular. One possible automation method is the parametric design method, which translates different mathematical relations into parameters which can be easily altered. This research aims to develop a model which parametrically determines multiple solutions for a probabilistic design approach of a rubble mound slope protection, which takes into account uncertainty. This leads to an expansion and acceleration of the insight into possible design options in the preliminary design phase.

This research describes a way to determine the reliability of coastal pump stations - from which their environment is influences by climate change – in time. Last decade, the challenge within the hydraulic infrastructure network shifted the focus from expanding the infrastructure inventory towards the maintenance and management of the existing network. The reason for this tendency is that the predominant part of the Dutch infrastructure inventory is approaching its end-of-life while their functions are still required. As a consequence, the asset owner is confronted with the challenge to choose the suitable maintenance action in order to extent the assets life-cycle.