The GSM-R network is a mission critical 2G telecom network which is tasked with ensuring the safe communication of train drivers with dispatchers across the EU. More specifically, this research explores the application of GSM-R in the Netherlands, and what threatens its future resilience, reliability, and availability. Of course, large scale critical infrastructure such as GSM-R is not one monolithic thing, and so to gain a better understanding of what these future threats are, the research is clearly divided into hardware, software, and human focuses. GSM-R is due to be replaced by FRMCS in the coming years, as the ERA has stated that GSM-R is getting to be too old of a system for the European train network to continue using. FRMCS is a 5G replacement network that aims to provide the same functionality as GSM-R, but with more modern technologies and knowledge. The problem there is that FRMCS is not yet ready for large scale deployment in the EU, so GSM-R must remain in service until then. From the beginning of the research process, it was clear that telecom applications have not previously had a transitional plan designed with a focus on the legacy system that is to be replaced. As such, this research aims to do just that. To do so, this research first looks into the basic infrastructure components that make up the GSM-R network. This is done so that an understanding can be built of what the network is made up of, so that they can be further discussed when it comes to actual failure data. While the information gathered for this part of the research does not have a large bearing on the ultimate final product, it is important to discuss the specifics of systems this large when considering future resilience threats. The research conducted in this thesis has a very large basis in established literature, as the methodologies and frameworks identified from desk research have direct applications to the task at hand in the thesis. Each sub-section of each focus also has its own different method or framework, which provides a robust means of attacking future threats as they are identified. These same frameworks and methodologies find their direct application in the data analysis conducted later on in the research. Utilizing failure and site completion data scaled from January 2020 until December 2026 for hardware, and support case data from January 2021 until December 2026 for software, trendlines were developed which roughly forecast what the future looks like for these two disciplines. These trendlines then are projected further into the future to provide an understanding of what things will look like, with some of the previously mentioned frameworks and methodologies applied to indicate how projections can change, with the proper planning. The human knowledge discipline of this research then looks into what can be expected in terms of retirees from the current workforce, in both 2035 and 2045. This data provides an indication of when the methodologies and frameworks from the literature review should be put into place, as knowledge management and retainment strategies need to be in place for the years before they are needed. To further bolster the findings from the data analysis, interviews were conducted. These interviews had a more future oriented focus than the data analysis, in the hopes of gaining personal insights that would otherwise not be attainable through historical data. The findings from all of this data indicate that remote radio heads, rectifiers, components of flexi multiradio system modules, unit mechatronics, and scrapped units are all future failures to consider which will have an ultimate impact on the resilience of hardware. In software, customer configurations, issues relating to outside interference, and issues in which a restart was necessary all take the cake for largest future considerations, while legitimate software bugs are not as common. The human focus displays how the passage of time will affect the GSM-R team headcount, and indicates that by 2045, only 22% of current staff worldwide will still be around. Of course, all of this information is quite bulky, and in the hands of project or product management, potentially too much to sift through to be quickly useful and usable in a true corporate environment, such as the one in which GSM-R operates in. For this purpose, a decision tree was designed. This decision tree is intended to provide those same project and product managers a quick solution to the problems they may be facing. The tree covers all three focuses of the research, and indicates where GSM-R should be on the trendlines determined in the data analysis. Depending on how close the trendlines are followed, potential solution areas as determined in the literature review are offered to mitigate the trendline discrepancies. Naturally, there are always more steps to take after actually making a decision to solve a problem, and as such, a secondary flow that follows the decision tree was also designed. This secondary flow covers the thought process of weighing the option to not follow the advice of the decision tree versus following it, and also indicates the deep interconnection between the human focus and everything else. No matter how well planned for and maintained the hardware and software focuses are, if there are no people around to carry out the plans, it does not matter how much effort and thought was put into planning everything else.
... Read more

The Dutch offshore wind sector has come to a crossroads as recent tenders, IJmuiden Ver Gamma A and B and Nederwiek, have either attracted no bids or been postponed. This happened due to rising development costs, subsidy-free tender designs by the government and uncertain electricity prices. In the past, the Netherlands utilised a one-sided Contractfor- Difference mechanism before moving to tender without financial support. This led to zero-subsidy bids by offshore wind developers, who gambled on favourable electricity prices. In response to the failure of the recent tenders, a coalition of market parties called for the introduction of two-sided CfDs. This thesis studies which CfD design will actually best enable the Dutch Government to reach its target of 40 GW by 2040, whilst balancing the trade-off between minimising the public support costs and maximising developer participation under uncertain electricity prices and costs. To address this complex issue, this research builds a multi-agent model in Python using the Mesa library, which enables the simulation of Dutch offshore wind tender auctions whilst incorporating bidding behaviour and strategy. The previous literature provides three different CfD designs: the one-sided CfD, the two-sided CfD and the Financial CfD. Each CfD is simulated in the model under a single-unit, pay-as-bid pricing tender format, with auctions being held twice a year until 2040. Each developer calculates a break-even strike price based on an NPV = 0 strategy, using individual beliefs about future electricity prices spread across three price levels and individual cost parameters, adjusting their risk appetite after each round based on the auction outcome. The model is then subjected to Monte Carlo Simulations, a sensitivity analysis and scenario analysis. The performance of each CfD is determined by the following metrics: deployment success rate, total subsidy cost, mean total number of exited agents per run, mean number of participating agents per tender, mean number of distinct winners per run, ceiling price rejection rate, and mean winning strike price. The research yields the following insights. From a developer’s perspective, CAPEX and capacity factor have the greatest impact on the strike price bid, more than OPEX or electricity price expectations. Furthermore, the ceiling price is the primary constraint on successful deployment. Under baseline conditions, the ceiling price rarely constrains developers. From the results of the ceiling price experiment, it is clear that the ceiling price acts as a market participation constraint, and its influence becomes apparent once the ceiling price level is lowered, leading to more unsuccessful simulations that fail to reach the 40GW target. Under high-cost scenarios, the two-sided CfD collapses almost entirely, succeeding in only 1.3% of the simulations. The one-sided CfD meets the target in only 19.2% of runs, whereas the financial CfD is the most robust, achieving a deployment success rate of 61.2%. This resilience is directly caused by the reference-generator bidding approach, in which the developer’s bid is partially decoupled from their costs, allowing lower bids below the ceiling price. The designs also produce very different fiscal effects. The financial CfD swings the most in fiscal results, from roughly €119bn in support costs under a low-price, high-cost scenario to a clawback of approximately €265bn under high realised electricity market prices. The two-sided CfD offers the largest recovery when electricity prices are high and costs are low, but it is easily disrupted when costs rise. The one-sided CfD results confirm the worries of market parties, where market competition forces zero-subsidy bids and the CfD offers no protection against rising costs. It exposes the government to subsidy costs when realised prices are low, whilst generating no revenue when prices rise. This thesis contributes to the CfD literature by developing a model that compares all three designs within a single multi-agent framework with repeated auctions. No single design performs best on both deployment success rate and cost, so the choice depends on the government’s preference, the first important finding of this thesis. A government prioritising the 40 GW target amid cost uncertainty should favour the financial CfD, which offers the most reliable path to the target but carries the risk of high subsidy costs or a substantial clawback. A government prioritising fiscal certainty in a stable cost environment may prefer the two-sided design, which delivers the largest net recovery, but collapses the hardest and provides the least competition. For both cases, the ceiling price policy plays a crucial role, with the current fixed price of 104€/MWh diminishing the viability of deployment under rising costs. This leads to the second significant finding of this thesis: a fixed ceiling price set above expected bids under normal conditions is rarely active but becomes the decisive constraint on achieving the target when costs rise. The ceiling price acts by excluding participants rather than by changing prices, making its calibration as important as the choice of CfD design itself. Offshore wind is the largest component of the Netherlands’ plan to decarbonise its electricity and industrial sector, and the design choices made for the future offshore wind tender auctions will play an important role in whether the target of 40GW is reached. CfDs offer several possible pathways towards the desired target, but the Dutch government must also take into account the potential subsidy costs relative to deployment success and the ceiling-price policy that creates their desired market. Otherwise, it could lead to a repeat of the past failed tenders. Whether the Dutch government heeds that warning and reignites the offshore wind sector will be seen in the upcoming tenders. ... Read more

IT auditing is a specialised discipline that examines whether an organisation’s IT systems, controls, and processes are reliable, secure, and in control. It plays a particularly fundamental role in financial reporting. Since financial statements increasingly rely on automated systems to generate and process data, their reliability cannot be assessed without examining the underlying IT environment, making Financial Audit IT (FAIT) a large service line for most major audit firms. As organisations grow more dependent on complex, interconnected IT infrastructures, the volume of work that auditors must handle rises, while the potential consequences of IT failures (such as data breaches, disrupted services, and unreliable financial reporting) grow alongside it. This places audit teams under increasing pressure to do more without compromising the quality of assurance that clients, regulators, and society rely on.

Artificial intelligence offers clear potential to relieve this pressure. Much of the IT audit process consists of high-volume, repetitive, manual work that is well-suited to AI support. Yet integrating AI into auditing is not a purely technical matter. It raises questions of trust, accountability, auditor skill, data protection, and regulatory compliance that deploying a tool alone cannot resolve. While audit firms are actively investing in AI, no structured, validated method exists for integrating AI into the IT audit workflow. This gap between recognised potential and absent guidance forms the central motivation for this research, leading to the research question:

How can artificial intelligence be integrated into the IT audit process of organisations through targeted process interventions?
... Read more

The North Sea is emerging as a central hub in Europe’s goal for energy transition with offshore wind capacity projected to reach up to 300 GW by the year 2050. Other than electricity generation, this vast renewable resource presents a significant opportunity for large scale green hydrogen production and the development of an integrated offshore hydrogen transport infrastructure. This thesis investigates how an offshore wind based hydrogen transport pipeline network could evolve in the North Sea under varying temporal, spatial and policy driven conditions. In order to address this question, a multi layered modelling framework is developed which integrates offshore wind farm deployment data, hydrogen production potential, transport capacity estimation, infrastructure cost constraints and dynamic network evolution modelling. Within a NetworkX based simulation environment, the offshore wind farms are represented as hydrogen production nodes and demand centers as sinks. Network growth is evaluated across multiple North Sea regions over the period 2030 to 2050 using four key performance indicators which are Total Pipeline Length (TPL), Average Source Sink Distance (ASSD), Fraction of Network Grown (FNG) and Delivered Hydrogen Potential (DHP). The results indicate that the North Sea has tremendous potential for producing hydrogen of approximately 20 to 27 million tonnes per year, depending on electrolyser technology employed and operational assumptions. Solid Oxide electrolysis offered the highest output. This level of production implies a substantial requirement for offshore hydrogen transport infrastructure particularly in the form of pipelines. The capacity of the pipelines ranged from DN200 for individual wind farms to DN600 or larger for larger pipelines for aggregated flows. Different scenarios were studied and across all scenarios, network evolution is found to be incremental and strongly path dependent. Early stage infrastructure is primarily driven by initial offshore wind farm deployment, forming localized clusters that gradually evolve into interconnected regional systems. Regions with early source activation exhibits rapid increases in pipeline length and hydrogen delivery by 2040, while regions with delayed activation showed slower initial growth followed by rapid expansion once sources becomes available. A key finding is that source availability and activation timing are the dominant determinants of network evolution while source selection strategies such as random or geographically close have limited long term impact. Constrained availability significantly slows infrastructure development in early periods. Phased activation results in smoother and more realistic growth trajectories. Project improvement scenario where the availability increases gradually to fully, showed the most balanced network evolution by avoiding both premature overbuilding and underutilization. The study further shows that offshore pipeline infrastructure evolves from fragmented local connections into structured systems with emergent trunk lines. The network formation is influenced by spatial proximity, flow efficiency ad network connectivity. However, early decisions creates path dependency and potential structural lock in effects that shapes long term network topology. Overall, the findings demonstrates that offshore hydrogen pipeline networks are not entirely cost optimised engineering systems instead they are emergent infrastructures shaped by the interplay of resource availability, temporal deployment, spatial constraints and policy coordination. Strong coordination could result in highly integrated system.
... Read more

How can a data-driven risk-based inspection approach be designed and implemented at the ANVS to improve efficiency and accuracy under uncertainty?

Abstract (Platte tekst, max. 500 woorden)
- In nuclear safety and radiation protection, inspection capacity is limited while the consequences of misprioritisation are high. The ANVS applies a risk-based inspection (RBI) approach, but current risk profiles rely partly on unstructured information and a limited set of measurable indicators. This may lead to blind spots, particularly under changing external conditions.

This study examines how a data-driven RBI approach can be designed for the ANVS to improve efficiency and accuracy under uncertainty. Inspectors’ tacit knowledge is translated into a structured set of risk factors combining measurable characteristics, behavioural dynamics, and external developments. A simulation model is used to analyse how risk evolves over time and how uncertainty in risk prediction affects inspection prioritisation and outcomes.

Results show that uncertainty mainly leads to temporary shifts in prioritisation rather than structural failures. Its impact is limited under stable conditions but increases during unfavourable external developments, when overall risk levels rise and missed high-risk inspections become more likely. Improving data quality has the strongest long-term effect on inspection performance, while increasing capacity primarily yields short-term improvements.

The study concludes that a data-driven RBI approach can strengthen inspection planning when uncertainty is explicitly considered and professional judgement remains integrated in final decisions. ... Read more

The electrification of society increases pressure on Dutch distribution grids, causing local congestion that limits renewable integration and economic growth. Distributed energy resource flexibility can mitigate congestion, but its effectiveness varies strongly by location and is poorly understood due to uncertainty in incentives, regulation, user behavior, and technical capability. This research develops a co-simulation framework combining AC power-flow modeling with a stochastic, agent-based flexibility state evolution model using Monte Carlo sampling. Location- and time-specific financial incentives are applied to assess how incentivized distributed flexibility responds to congestion and where congestion robustly remains. Results show that while technical flexibility capacity is often sufficient, congestion can persist due to local disparities in incentives, behavior, capability, and spatial correlation of flexibility providers—especially for highly local congestion low in the distribution hierarchy. The framework identifies where distributed flexibility is inadequate and where additional, strategically placed flexibility or storage is needed, supporting congestion-efficient planning and targeted congestion management in distribution grids.
https://github.com/developerDuncan/Assessing-congestion-patterns-and-flexibility-reactions.git ... Read more

This thesis investigates how organisations can embed environmental sustainability within the operational use of Generative Artificial Intelligence (GenAI) tools through targeted, user-centred interventions. While research has largely focused on reducing emissions in AI model training, the inference phase, where GenAI is integrated into daily workflows, remains a substantial and under-addressed source of environmental impact.

Using a Design Science Research approach, the study combines a literature review, interviews with GenAI users and AI experts, and behavioural theories including the COM-B model, Theory of Planned Behaviour, Nudging, and Affordance Theory. Enabling factors for pro-environmental GenAI use were translated into functional and non-functional requirements, guiding the development of three persona-specific interventions: (1) Sustainable by Default for externally motivated users, embedding energy-efficient model settings and a monitoring dashboard; (2) Sustainability Guidance for aware but uncertain users, offering a sustainable prompt builder and impact estimator widget; and (3) Collective Sustainability for unaware users, providing monthly emissions feedback and rotating green tips.

The resulting integration framework and decision-support tool offer practical guidance for embedding sustainability into organisational AI practices, demonstrating that environmental impact reduction in GenAI requires socio-technical, behavioural, and cultural change alongside technical optimisation.

... Read more

The North Sea plays a pivotal role in Europe’s energy transition, offering vast potential for offshore wind energy expansion. One promising approach to harness this potential is the hub and spoke concept, in which offshore wind farms connect to energy hubs that distribute electricity to surrounding countries and other hubs. Despite the strategic importance of this approach, there remain significant uncertainties regarding the system's design, including decisions about installed capacity, hydrogen integration methods, and future electricity demand.
This study addresses these uncertainties by first identifying the offshore wind generation goals of the countries surrounding the North Sea. Next, it evaluates various methods for integrating hydrogen production into the energy system. Based on the findings, multiple scenarios were designed to represent a spectrum of future system arrangements.
A scenario-based energy flow model was created to simulate and compare the behavior of different network designs under varying assumptions about electricity demand and hydrogen inclusion. The analysis reveals that electricity demand has a major impact on system design. Higher demand reduces the availability of surplus electricity, thereby limiting both hydrogen production and the cable capacity needed to transport electricity across the network. The results also show that onshore hydrogen production requires approximately 11% more cable capacity than offshore hydrogen production, although it avoids additional offshore infrastructure costs.

This study concludes that the North Sea hub and spoke energy system could serve as a foundational element of a flexible, robust, and integrated European energy system. However, to support effective planning and implementation, several critical areas require further research. These include the development of a time-based optimization model that incorporates real weather data and demand profiles to reveal seasonal and daily performance patterns. Moreover, there is an urgent need for more accurate projections of future electricity and hydrogen demand, as these will heavily influence infrastructure decisions. Lastly, detailed economic assessments must be conducted to better understand the investment requirements, market dynamics, and potential policy interventions needed to make this vision feasible.
By clarifying key design choices and highlighting critical next steps, this research contributes to a more informed and coordinated development of the North Sea energy system, advancing Europe’s transition toward a sustainable, carbon-neutral future. ... Read more

Seventy percent of the world’s population live in countries where inequalities have increased over the past three decades. There is growing recognition that global understandings of inequality must be complemented by empirically grounded, context-sensitive analyses that incorporate spatial and temporal dimensions. This dissertation advances that agenda by exploring the structural drivers of urban inequalities through a methodological approach that integrates critical theory with spatial data science. Central to this approach is the development of a theoretical framework that synthesises geospatial analysis and complexity science. This framework is operationalised through its iterative application to three empirical case studies drawn from both the Global North and South, enabling a comparative perspective on urban inequalities. By bridging critical theory with novel empirical methods, the research contributes to contemporary debates on urban inequality, offering conceptual and methodological innovations as well as policy-relevant insights. ... Read more

Desert-based solar projects can aid in addressing the global demand for renewable energy due to their high solar irradiance and vast land availability. These projects are increasingly important as the world seeks sustainable energy solutions to combat climate change. However, they face significant risks such as extreme weather conditions, logistical challenges, water scarcity, and socio-political complexities, which must be managed to ensure their success and sustainability. Therefore, susrainable risk management and mitigation strategies need to be employed. ... Read more

The power grid is becoming more intertwined with communication technologies, forming what is known as the smart grid. This integration allows for more efficient management of the power grid, which can help reduce emissions. However, the increasing connectivity with communication technology also expands the attack surface for cybercriminals or cybercriminal organisations. In 2015, a Ukrainian power station was attacked via a cyberattack, leading to an outage that affected almost 1.4 million people. The growing interconnection between the power grid and communication technologies can thus have severe impacts. Therefore, this research introduces an interdependent communication network and power grid to investigate how failures in the communication network affect the power grid's robustness. For generating communication networks, we use Python’s NetworkX package, and for the power grid, we use pandapower’s IEEE 118-bus test system. We analyse the robustness of the smart grid by examining two different communication networks: a mesh network and a double-star network. We simulate attacks on the communication network by initiating failures based on random selection, degree, betweenness, and closeness centrality, with an increasing number of node failures. We also distinguish between two failure behaviours: one that includes communication network failure propagation (e.g., the spread of malware or the failure of dependency nodes within the communication network) and one without this failure propagation behaviour. In general, the smart grid shows higher robustness with a mesh communication network. However, under different failure behaviours, attack strategies, and numbers of nodes attacked, this can vary. Prioritisation should focus on reducing communication network failure propagation, as this significantly impacts the robustness of the smart grid. Even though degree centrality and betweenness centrality have the most impact on double-star networks, for mesh networks, closeness centrality should also be considered, especially in scenarios without network failure propagation. ... Read more

This research explores the role and challenges of openness between digital healthcare platforms through the use of boundary resources in the Dutch healthcare system, focusing on factors influencing openness decisions. Traditionally, boundary resources facilitate arm’s-length relationships between platforms and their periphery. This study examines their specific contribution to platform-to-platform openness in digital healthcare. To properly define openness between platforms, a distinction was made between interoperability and platform-to-platform openness: interoperability focuses on the compatibility of platform resources, while openness emphasizes the provision or accessibility of these resources. Boundary resources are further divided into technical and social aspects, shaped by laws and regulations. A document analysis of key policy documents maps how these relevant boundary resources function. The importance of APIs is emphasized as a means to progress toward an interconnected digital healthcare system, however, the practical implications and impact of openness decisions leading to the formulation of API’s, particularly between platforms, remain somewhat unclear. Interviews with the most relevant stakeholders in the field provide further insight into the drivers and barriers affecting boundary resources adoption. Most emphasis is placed data exchange and transactional based platforms by the stakeholders. The distinction between interoperability and platform-to-platform openness in healthcare can help avoid conflating compatibility requirements with broader governance and strategic considerations, enabling more focused discussions on factors essential for platform-to-platform openness. Boundary resources, while holding significant potential to facilitate openness, are not clearly distinguished by platform owners in their application for inter-platform openness compared to openness toward complementors, while these two forms of openness can serve a distinct purpose. Openness concerns are further composed of involved technical costs, security and data liability issues, as well as competitive pressures, with the potential loss of operating models and market position adding additional complexity to these openness decisions. Regulatory direction stimulates the development of boundary resources to promote inter-platform openness by embedding platform connections within the broader market ecosystem. Boundary resources could be further utilized further to enable platforms to xtend their periphery and for addressing deeper infrastructural and collaborative challenges within the healthcare sector. Platform-to-platform openness is influenced not only by decisions of the platform owners themselves, but also by broader market dynamics, regulatory frameworks, and the platform environment. This research highlights the potential of boundary resources in tackling infrastructure, innovation, and data-sharing challenges in healthcare. ... Read more

The industrial sector is one of the most energy consuming and CO2-emitting end-use sectors. To reach climate goals, decarbonization of these industrial sectors is imminent, however, not so apparent. Especially the hard-to-abate industry sectors, such as the chemical, mineral processing, iron, and steel sectors, are difficult to decarbonize as they require either high temperature heat or use fossil fuels as feedstock.

Hydrogen has the potential to reduce carbon emissions in industries such as chemicals, glass, iron and steel, as well as to serve as a cleaner heat source. To reach net-zero emissions by 2050, sectors that currently use fossil fuels for high-temperature processes and as feedstock will likely need to shift towards blue or green hydrogen. Currently, some industrial hydrogen use relies on gray hydrogen, produced from fossil fuels and contributing to emissions. In contrast, blue hydrogen captures and stores CO2 produced from fossil sources, while green hydrogen is entirely emissions-free, generated from renewable energy. In other words, some processes need to change from gray to green/blue but most of them need to change from other fossil fuel based processes to hydrogen processes.

Hydrogen offers a way to cut carbon emissions in industries like chemicals, glass, iron, and steel, and can also act as a cleaner heat source. Achieving net-zero emissions by 2050 will likely require sectors that currently depend on fossil fuels for high-temperature applications and feedstocks to adopt blue or green hydrogen instead. Today, certain industrial applications still use gray hydrogen, derived from fossil fuels and contributing to carbon emissions. However, blue hydrogen captures and stores the CO2 generated, while green hydrogen is emissions-free, produced using renewable energy. In essence, some processes will need to transition from gray to blue or green hydrogen, while many others will shift from fossil-fuel-based processes to hydrogen-based alternatives.

However, the timing and extent of the hydrogen transition are uncertain as they are heavily influenced by external factors such as hydrogen prices, available subsidies, and alternative decarbonization options. Additionally, industrial plant owners may be reluctant to disclose decarbonization plans due to competitive pressures, adding another layer of demand and participant uncertainty that complicates infrastructure planning.

This thesis addresses the planning of hydrogen infrastructure within an industrial port cluster (IPC). IPCs are defined by their proximity to water and concentration of industrial activities related to a specific sector. In order to effectively address spatial constraints, this thesis will plan the hydrogen networks along the current road network within IPCs. Current infrastructure planning methods have a time horizon of ten years. However, as the expectation is that the hydrogen demand will increase towards 2055, a time horizon of ten years can increase the total costs of the network when the network is implemented over time between 2025-2055. This introduces the following research question:

“How can a cost-efficient, robust pipeline network for an industrial port cluster be developed over time under uncertainty?”

To answer this question, the robust backtracking planning method (RBPM) is developed. This method aims to minimize costs over the 2025-2055 time frame while facilitating the hydrogen to the demanding plants. Because the demand for hydrogen is likely to grow over time, this method finds a robust network that is able to facilitate the demand in many possible future demand scenarios of 2055.

The robust network is then implemented incrementally for 2035, 2045, and 2055 using a backtracking approach. In this context, backtracking means that when an industrial plant transitions to hydrogen in one stage, pipelines are installed with the robust networks’ capacity, rather than just the minimum required to meet that plant’s immediate needs. This extra capacity ensures that if other plants transition in later years, the existing network can accommodate the increased demand without needing costly pipeline extensions. By preemptively building capacity, this approach reduces future installation costs and enhances the network’s ability to adapt to evolving demand patterns.

The RBPM is tested on simulations of multiple simplified IPCs. By testing different IPC simulations, it is studied how the difference in industrial plants determines the development of the network. The RBPM is compared to the results of a traditional planning approach which only plans the networks with a time horizon of ten years.

The results show that the RBPM incurs lower costs over 30 years, but it requires a higher investment in 2035 due to the greater capacity installed at that time. This thesis finds that the total potential hydrogen demand and the physical size of an IPC significantly affect the performance of the RBPM compared to the traditional planning approach. Additionally, the projected installation and operational costs over time also impact the RBPM’s performance relative to the traditional planning approach.

For IPCs with comparatively low hydrogen demand — typically clusters with fewer iron and steel facilities, chemical plants, or refineries — the RBPM emerges as the most economical approach. This method requires only slightly higher investment by 2035 but ultimately generates substantial savings by 2055. By installing sufficient pipeline capacity upfront, the RBPM avoids the need for additional pipelines every ten years, leading to long-term cost efficiency through 2055.

For IPCs with high hydrogen demand—typically found in iron and steel plants, basic chemical plants, or refineries—the initial installation costs and ongoing operational expenses of RBPM make it less advantageous. While RBPM may offer slightly better economic profitability over a 30-year period, the substantial investment required in 2035 compared to traditional planning makes implementation challenging due to budget constraints. In these high-demand clusters, the decision between RBPM and the traditional approach for developing a hydrogen pipeline network depends on the cluster’s budget, anticipated future installation costs, and projected operational expenses over time.

Opportunities for further research include the application of the RBPM to a real case study to validate the result, increasing the amount of possible future scenarios by incorporating uncertainty in installation and operating costs and increasing the demand and participant uncertainty range. Lastly, another research direction to explore is the generation of different robust network methods and their performance.
... Read more

Society is facing a huge challenge in switching the energy sectors dependence on fossil fuels into an energy sector using mostly renewable energy sources. The switch towards using more sustainable energy sources is known as the energy transition. The goal of the energy transition is to lower the greenhouse gas (GHG) emissions emitted by the energy sector. Lowering the GHG emissions helps society limit the global warming caused by GHG [3]. 17.5 % of the global energy usage comes from the energy use in buildings [50]. It is thus very important that the energy use in buildings transitions towards using more sustainable energy sources. One of the renewable energy sources that is ought promising in the energy transition for energy use in buildings is geothermal energy [3]. Geothermal energy is energy that is captured in reservoirs of hot water in the earth’s crust. The hot water captured in the hot water pockets is pumped to the surface, to use it in spatial heating. The return pipe returns the cooled water to the geothermal well, where it can heat up again over a certain period of time [63] [23].

In some cases, geothermal energy is applied using a district heating network. A district heating network is an example of a system that provides heating and/or cooling capacities to a group of buildings [65]. A district heating network is a network of pipelines that transport the hot water from the geothermal well to the buildings in the district. A geothermal well in combination with a district heating network is developed in Delft [27]. The district heating network will deliver energy to the TU Delft campus, two neighborhoods in Delft and industry at the Schieweg in Delft [28].

Besides the district heating network in Delft, it is expected that district heating networks will be applied more often to accelerate the energy transition. Yun-Chao and Chen (2012) concluded that most optimization techniques optimize the whole system with its components. Less optimization techniques are applied to the sole components. Besides the fact that most optimization methods optimize the system as a whole, most optimization objectives only include optimizing the cost of the system. Also, effective optimization techniques are required as optimizing large graphs may be computationally time consuming [36]. In literature there are also clear signals that state that the trade-off between thermal comfort, and efficiency with respect to cost has to be tackled [53]. In this research, optimizing district heating networks for cost is compared to optimizing district heating to maximize thermal comfort or efficiency.

In this research two models are developed: a model that calculates the cost of the district heating network, and a model that calculates the thermal losses of the district heating network. Both models are applied to a district heating networks that is developed in a street network. Furthermore, multiple heuristics are applied to come up with better district heating networks. The optimization technique is tested on 100 small, randomly generated district heating networks. After that, the district heating network in Delft is optimized. The differences in cost, efficiency, etc. will be evaluated. Besides, the performances of the district heating networks are evaluated by introducing energy deficits under different conditions.

Optimizing the district heating networks for cost led to a very consistent result: When compared to their individual starting point, the district heating networks became cheaper and more efficient. A moderate-strong correlation is found between the the increase in efficiency and the decrease in cost while optimizing the district heating networks. In contrast to that, the networks that maximize efficiency are much more expensive than their cost optimized alternative, while the increase in efficiency is in most cases moderate. However, there are rare cases where the efficiency is increased much at a moderate increase in cost. This phenomenon is also found in Delft. Given the result that the efficient district heating network also performed much better than the cheapest alternative during energy deficits, in this research it is shown that choosing an objective function has a very large impact on the characteristics of the network. Therefore it is shown that for future district heating network optimization, it is important to trade off cost against efficiency. ... Read more

Explainable AI (XAI) systems are rapidly gaining significance. While frameworks for XAI interpretability for experts abound, metrics for laypeople’s comprehension are absent. This study addresses this gap by investigating interpretability factors from both developer and layperson perspectives. The core research question is: "How can XAI developers assess to what extent XAI is interpretable for laypeople?".

Applying the Design Science Research Methodology, findings from multiple literature reviews are combined to construct a preliminary XAI interpretability framework for laypeople, featuring crucial factors and their relationships, as well as associated principles. The proposed framework underwent validation through semi-structured interviews with 12 XAI experts, informing revisions and refinement of our key principles. Subsequent layperson surveys, considering a specific use case, offered insights into preferences about interpretability factors, informing further refinement.

The final theoretical framework highlights pivotal factors including simplicity, transparency, comprehensiveness, complexity, clarity, generalizability, trustworthiness, explanation fidelity, model fidelity, intentionality, relevance, affordance, coherence with prior beliefs, and actionability. Surrounding the framework are key principles emphasizing trustworthiness, relevance, simplicity, clarity, coherence, intentionality, actionability, fidelity, contextualization, and ethical considerations, serving as actionable guidelines for XAI developers and researchers.

The implications of the study are profound, offering valuable insights for advancing XAI research and system design. The refined framework and principles act as a foundation for both novice and experienced XAI developers, fostering interdisciplinary research among AI, human-computer interaction, psychology, and philosophy experts. These findings can drive the responsible adoption of AI systems across sectors like healthcare, finance, and transportation, while informing policies and regulations governing AI technologies. Our study promotes responsible AI practices, enhancing user trust and understanding, while facilitating the creation of more effective guidelines and standards. ... Read more

Freshwater is becoming an increasingly scarce resource globally. The chemical industry is the second largest freshwater consuming industry and the largest wastewater producer. To reduce the freshwater consumption sustainable measures are introduced within the industrial sector, of which Industrial Symbiosis (IS), has gained prominence. Specifically for water, this involves one company utilizing the residual wastewater of another company based on the differences in water quality requirements However, the chemical industry lags behind in embracing IS practices. Two important reasons cause this trend. Firstly, it is technically challenging to recycle water because of all the different water qualities used and produced in chemical parks. And secondly, the industrial symbiosis initiatives have to start with the stakeholders involved in the chemical park themselves in bottom-up regulated countries, while currently, many stakeholders decide not to join an IS water network. This research focuses on the feasibility of implementing IS water networks within chemical parks, considering stakeholder behaviour (wanting to join or not join an IS water network). A model is developed to measure the optimal performance of an IS water network, encompassing technical optimization and stakeholder decisions. The study reveals that the potential for industrial symbiosis in chemical parks relies heavily on existing treatment facilities and stakeholder participation. The type of treatment facility structure, with in-house or separate treatment facilities, impacts costs and water recycling capabilities, with chemical parks with separate treatment facilities exhibiting higher costs and higher recycling potential. Key decision criteria for stakeholders include revenue growth, compliance to regulations, trust among stakeholders, and awareness of water recycling possibilities. Recommendations to enhance industrial symbiosis of water include technical adjustments and construction efforts to the IS water network within the chemical park, and policies addressing stakeholder concerns. The contributions of this research lie in integrating technical and social aspects in the design of optimal IS water networks, providing insights into decision criteria that facilitate or hinder industrial symbiosis in specific chemical parks and how to address them to recycle most water as possible with least amount of costs. ... Read more

This thesis aims to answer the question "What are the properties of a resilient road network?" The resilience of a road network indicates the magnitude and consequences of a disruption, which can be events such as a traffic accident or a flood. The literature is not in agreement about the exact definition of road network resilience, and the way to quantify it. A literature search is done to look for definitions and resilience metrics. Eleven metrics are found that quantify resilience in road networks to accidents. These eleven metrics are compared based on whether they are on a network level, and how many variables are used. Two metrics are chosen, one based on travel time, and one based on space-mean flow. Another metric, based on the outflow of the network is added.

The three chosen metrics are then compared in small networks to see if they give the same results. The small network consist of five nodes with different link configurations. The conclusion about which network is the most resilient changes with the resilience metrics. The metric based on travel time is deemed the best for this type of research. The metric based on network outflow can have distorted results due to high peaks, and the metric based on space-mean flow is better suited for a different experiment design.

To see how network parameters such as link density influence the resilience, a simulation with random networks is done. 266 networks with nine nodes in random locations, and random links between them are simulated. It is concluded that networks with a lower link density have a higher resilience. This is not only illustrated by the relation between resilience and link density, but also by several other network parameters which are related to link density such as average number of lanes and connectivity. A reason that these networks are more resilient could be that there is less spillback, the links in the high density networks are longer and have more lanes, so congestion will not spread to other links as fast. ... Read more

Power systems are large-scale, complex socio-technical systems that provide modern society with one of its most indispensible assets: electricity. Electricity supply is not only an irreplaceable asset in daily activities, it is also vital for operation of other critical infrastructures of the technological age. Crucial socio-economic systems depend on electricity supply to support infrastructures such as telecommunications, transportation, water and natural gas supply, as well as financial and healthcare services. Therefore, ensuring secure and reliable operation of power systems as an enabling infrastructure is crucial.... ... Read more

Ransomware has evolved over the years, shifting from widespread attacks targeting individuals to focused attacks on businesses and agencies. These attacks are performed by ransomware gangs while establishing interaction within the ransomware ecosystem. In this thesis, the ransomware ecosystem is posited as being constructed of three separate sub-ecosystems: the attacker sub-system, the defender sub-system, and the governance sub-system. Since ransomware gangs put in an effort to hide their internal communication and operation from the outer world, difficulties arise in correctly understanding the ransomware ecosystem and a ransomware gang’s establishment of interactions within this ecosystem. As a result, current interventions are ineffective.

While earlier research has been conducted on ransomware, we observe two knowledge gaps: 1) there is a lack of understanding of how ransomware gangs establish interactions with actors in the ransomware ecosystem, and 2) There has been a lack of research that uses ground truth data due to ransomware gangs keeping their internal communication and operations hidden. This thesis uses the leaked internal communication data of the Conti ransomware gang to fill these knowledge gaps and answer the research question: To which extent can the ransomware ecosystem be reconstructed using ground truth communication data of the Conti ransomware gang?”.

To answer this question, a novel methodology is proposed that uses Latent Dirichlet Allocation (LDA) topic modeling to empirically determine overarching topics in Conti's internal communication. It is then researched how these overarching topics map to Conti’s tactics, techniques, and procedures (TTP) which is a commonly used methodology to better understand how ransomware gangs operate. Subsequently, these TTP are leveraged to reconstruct the ransomware ecosystem while taking the perspective of how the Conti ransomware gang establishes interactions within the ransomware ecosystem.

The findings of this thesis indicate that Conti is a large and professional organization that incorporates and adjusts services of service-providing cybercriminals in the attacker ecosystem rather than developing their ransomware themselves using scarce IT talent. In addition, reconnaissance is one of the most critical activities that ransomware gangs perform to get to a successful ransomware attack. While researching Conti's TTP, this thesis identifies novel TTP of ransomware gangs, such as Conti's attack chain, reconnaissance procedure, and money laundering procedure.

We conclude that the ransomware ecosystem can be reconstructed from the attacker ecosystem, the defender ecosystem, and the governance ecosystem, in which ransomware gangs establish interactions within each sub-ecosystem while operating from the attacker ecosystem. In the attacker ecosystem, ransomware gangs establish interactions with service-providing cybercriminals to outsource sub-commodities of their ransomware value chain. This allows them to strengthen their attack vectors by relying on the expertise of others and have a more varied set of attacks. The defender ecosystem is comprised of defenders that defend themselves against ransomware. Ransomware gangs establish interactions by performing extensive reconnaissance on defender territories and valuable information and open-source tools that strengthen their attack vectors. The governance ecosystem comprises governance actors that create and maintain the governance framework that influences the attacker ecosystem and defender ecosystem. Ransomware gangs establish interactions with actors in the governance ecosystem to observe the regulatory frameworks in place and adjust their TTP based on the involved risks of getting caught. ... Read more