The DAP well is a project by the TU Delft to be able to provide heat to a part of Delft. However, presently the geothermal well is not economically feasible. The well could be a potential source of lithium, which is mostly used in batteries. The demand for lithium is estimated to rise, thus this could potentially be a promising investment. In this research, the feasibility of extracting lithium from the clay and mica faces of the Delft Sandstone is estimated. The Delft Sandstone was formed in the Valanginian era and consists of sandstone and siltstone with clay interfaces. Potential minerals that could occur in this layer that can contain lithium are smectites, kaolinites, micas and illites, fibrous clays or chlorites. These minerals either contain lithium in their naturally occurring mineral structure or can be formed when in a mineral an element is substituted by lithium. The minerals that will be analysed in this research are hectorite, montmorillonite, bentonite, kaolinite, lepidolite, zinnwaldite, illite, sepiolite, palygorskite and cookeite.
The samples will be analysed and interpreted using Gamma-Ray Spectroscopy and Infrared spectroscopy. Gamma Ray Spectroscopy detects the naturally occurring gamma rays emitted from a rock sample. These are quantified in a general signal as well as concentrations of Potassium, Thorium and Uranium. High clay content will relate to a high K value thus a high GR (Gamma Ray) signal. In IR (Infrared) Spectroscopy a beam of infrared light is passed through a sample, which leads to the absorption of specific wavelengths by the molecular bonds present. These spectra are then compared to library spectra and minerals can be identified. In this thesis, the GR signal could not be related to the well logs as the sample quantity was too low. From the IR Spectra kaolinite and illite were defined in this layer. These clays can contain lithium when elements are substituted in their molecular structure.

This research evaluates the applicability of Multivariate Imputation by Chained Equations (MICE) for estimating missing well-log data across different sedimentary basis. Utilizing various machine learning techniques including XGBoost (XGB), Random Forest (RF), K-Nearest Neighbors (KNR), and Bayesian Ridge (BR), the performance of MICE was tested on three different data sets from distinct geological contexts and preprocessing conditions with minimal user input.
The main results indicate that the performance of MICE varied across different data sets and well-logs, highlighting the complexity of imputing missing data in heterogeneous sedimentary basins. The number of iterations in MICE did not significantly impact the performance of the models, while data quality, pre-processing, and geological complexities played crucial roles. The Force-200 data set, which underwent extensive preprocessing, demonstrated better imputation performance compared to the Montney and Beetaloo data sets. Additionally, XGB often outperformed other algorithms, predicting missing values with different number of iterations.
The main conclusions drawn from this study emphasize the need for more research to minimize user input and to develop more robust and flexible approaches to imputing missing data in well-logs. The study highlights the challenge of determining a single set of hyperparameters optimal for all the well-logs, suggesting the need for more adaptable models or even advanced techniques like deep learning techniques. The research also suggests the importance of refining pre-processing techniques, exploring further combinations of well-logs, and developing cross-validation approaches that effectively replicates real-world scenarios to advance the application and reliability of MICE in data imputation of subsurface data with missing values.

In the geothermal exploration well, NLW-GT-01, in the West Netherlands Basin, a very tight reservoir with low porosity (≤5.0%) and matrix permeability (≤ 0.1mD) has been encountered in the Main Buntsandstein Subgroup. The impact of natural fractures, which can significantly impact fluid flow in tight reservoirs, has been studied. However, there is a lack of consensus in different studies on the fracture classification.

In this study, a re-evaluation is performed of the well data of the NLW-GT-01 and VAL-01 wells. An analysis comparing core, geophysical, and image logs is performed to document the drivers and characteristics of natural fracture distribution in the WNB and investigate the geological and geomechanical parameters influencing the development of fractures in the Main Buntsandstein Subgroup.

On the core of NLW-GT-01 and VAL-01, five types of fractures are classified veins, joints, Mode II fractures, stylolites and drilling-induced fractures. The natural fractures have a dominant NW-SE strike orientation. The drilling-induced fractures and the borehole breakouts indicate a NE-SW oriented in-situ minimum stress and an NW-SE oriented maximum stress. The natural fractures are favourably oriented to be open based on their orientation compared to the in-situ stress orientation.

The fractures are associated with large-scale tectonic events. During the burial of the formation related to the extensional phase of the WNB during the Mesozoic, veins, joints, conjugate fractures, and stylolites could be formed. During the inversion of the basin in the Late Mesozoic-Cenozoic, tectonic stylolites could have been formed, and the previously formed conjugate fractures could have been reactivated.

Natural fractures are concentrated in more heterolithic intervals of the VAL-01 and NLW-GT-01 wells in both image logs and cores. These heterolithic sections display alternations of medium sandstones with silt- and claystones. The identification of more fractures in VAL-01 compared to NLW-GT-01 can be explained by the difference in basin location. VAL-01 is located in the centre of the basin where distal playa environments produced more fine-grained material alternating with coarser sands. The more proximal NLW-GT-01 is dominated by fluvial sands. The lithological variability produces Young’s and Shear modulus variability that seems to be driving increased fracture density rather than the
absolute value of the Young’s modulus causes this.

Evaluating an adequate computer model representing the characteristics of a subsurface reservoir is of great importance in the oil and gas industry. With recent data acquisition developments, high-resolution geological models can be generated for the subsurface reservoirs; however, uncertainties of geological data and too many details in the model have brought up remarkable limitations for application of high-fidelity geo-models in reservoir management. To generate efficient low-resolution models, which are relatively representative of the reservoir dynamic behaviour, data-driven models are proposed. The data-driven models are trained principally based on the production history of the reservoir with primary knowledge of structural information. The Data-driven Discrete Well Affinity (DiWA) model is designed to generate efficient proxy models that require basic geological information and full production history of the reservoir. Since it benefits the computational power of the Delft Advanced Terra Simulator (DARTS) and adjoint gradient-based optimization, the DiWA model has become an efficient framework in terms of accuracy and computational costs. This research shows that the DiWA model can relatively reconstruct the reference reservoir parameters out of an ensemble of priors, and the final results can be improved utilizing closer prior knowledge to the true response. Additionally, it is demonstrated that the performance of the DiWA model in a complex, realistic oil field can be enhanced using a comprehensive objective function, applying appropriate well controls, understanding the underlying driving mechanisms, using sufficient degrees of freedom, and applying consistent boundary conditions.

Several waterway protection structures in the Netherlands dealt with rapid degrading armourstone after construction. This suggests that the current selection procedure of armourstones is not flawless to all degradation mechanisms, especially in dynamic environments. In this work armourstone is meant to include all rock that is used in river, coastal and offshore constructions, including rip rap and rock fill. Armourstones are in the Netherlands selected by their CE marking, which lists the description of the armourstone product and information on regulated characteristics. These characteristics are tested according to the European standard EN 13381-1&2:2002, which specifies the properties of aggregate acquired by processing natural materials for use as armourstone. This research is executed to provide a contribution in the quality control and assessment of armourstone and rock fill to ensure proper handling and installation in coastal and waterway protection structures. The research started with a durability investigation according to the standard EN 133831&2:2002 on sampled sandstones and limestones that will be used in a submarine trench backfill, to check the regulated armourstone characteristics. Next, a petrographic analysis under the microscope was performed on methylene blue (MB) stained thin rock sections to investigate the presence of deleterious constituents and structures. This analysis was followed by extensive index testing to investigate the applicability of simple testing tools in a durability investigation. These tools include the Brazilian tensile strength (BTS) test, an indirect tensile strength test, the Equotip test, a surface hardness tester that records the surface rebound of an impact body, and the MB adsorption and staining test. MB is a dye that colours constituents with an excess in negative electric charges. The tests according to EN 13383-1&2:2002 suggested that the sampled armourstone bulk satisfies the required durability parameters and no evidence was found that indicates rapid degradation during the engineering lifetime. Nevertheless, variability in the armourstone pieces was spotted and some individual rock pieces approached or exceeded minimum durability requirements. The BTS test revealed considerable variation in tensile strength, a key parameter to assess degradation, between armourstone blocks and within a single block when a similar orientation of the bedding was maintained. Moreover, the orientation of the bedding in the sandstone had large influence on the tensile strength, where the values obtained perpendicular on the bedding were twice as high than parallel to the bedding. Single impact method (SIM) Equotip measurements on rough, untreated aggregate surfaces were consistent. The rebounds correlated to visual features like grain size, cracks, surface roughness and degree of weathering within handheld specimens. Equotip measurements on the rock cores and sawn surfaces obtained higher mean rebound values and smaller standard deviations. The mean and standard deviation are considered most suitable to be used in a durability investigation. The repeated impact method (RIM) by the Equotip was not successful to indicate the degree of weathering of single hand specimens. The mean Equotip values on the aggregate correlated well to the water absorption when divided into proper density and size classes. The mean rebounds on the cores distinguished well between weathered and intact cores when saturated, and correlate well to the unconfined compressive strength (UCS) values. Furthermore, the Equotip mean rebound value related to the BTS value when performed on isotropic rock disks. The MB adsorption test and MB staining of the thin sections indicated the presence of localised spots and laminae rich in clay or organic matter. The staining of sawn aggregate surfaces agreed with some of the deleterious structures in the thin sections, yet was not consistent throughout the tested rock pieces and varied between the dark coloured limestone and light coloured sandstone. The Equotip test, BTS test and MB adsorption test are quick, easy and cheap methods to obtain more understanding in variability and rock behaviour which are not necessarily captured by the standard laboratory tests according to EN 13383-1&2:2002. The simple index tests should always be accompanied by the standard laboratory tests to provide a proper reference and understanding. Detailed mechanical durability tests, like the slaking test or wet-dry cyclic tests, should be performed when the durability assessment indicates a high amount of deleterious minerals and structures, to identify the amount of degradation caused by swelling behaviour of these minerals.

Mechanical dredging of a stiff clay results in clay lumps. Those clay lumps can be reused for the construction of land reclamation projects, and hereby economic and environmental value is generated. Placement of the clay lumps
in water result in a matrix of clay lumps and an interlump void space. The collapse of the interlump void space will cause large settlements and needs to be overcome before the site can be used for construction. Preloading is
an effective method to close the interlump voids. Stiff clays soften over time due to unloading and swelling. As a result, the strength and stiffness of the clay lumps decrease over time. The presence of discontinuities accelerates
the softening process. It was proposed by Leung et al. (2001) that the interlump void space closes under a reduced
preload of 25 kPa. The interlump void space closes under a reduced preload because the lumps soften over time.
The question rises if closure occurs within a normal construction timespan, i.e., 1 - 2 years, under this reduced
preload. In this study, a combined experimental and numerical approach is applied to determine the influence of soil
characteristics, softening and the presence of discontinuities on interlump void closure. The influence of softening due to chemical and hydro-mechanical swelling is tested by experimental swell-load tests on stiff overconsolidated
Boom clay samples. Additionally, the presence of discontinuities is studied by CT images, and a miniature clay fill test is performed to study the softening time and the rearrangement effect. Furthermore, a numerical study is performed in which the influence of specific soil characteristics on void closure was researched by a sensitivity analysis. By the experimental tests, it is shown that a pore water chemistry change alters the degree of swelling and its compressibility. Furthermore, fissures were identified in the sample material by CT images. The smallest microfissures could not be identified due to the resolution of the images. Consequently, it is impossible to estimate the effect of fissures on the hydraulic conductivity. Therefore, literature data was used to estimate the hydraulic conductivity and this was used as the input in the calculations. In the numerical study, it was shown that MPM could model the softening behaviour over time. The sensitivity analysis showed that the MPM model responds consistently to parameter sensitivity analyses. This leads to the conclusion that MPM can be used as a investigation tool to increase
the understanding of the influence of parameter variability on the final interlump void closure problem. The resulting strains of the numerical model comply with the theoretically calculated strains. It was expected
that the numerical strains were smaller than the experimentally determined strains, due to the presence of interlump voids in the numerical model. But, the numerical strains are larger than the experimental strains. In the sensitivity
analysis, it was found that the final strains and closure time are highly dependant on the soil characteristics. It is likely that the input of the MPM model differs compared to the true sample material and thus different strains result.
The time until the interlump void space closes was determined by a simplified geometry in MPM under a preload of 25 kPa. For an unfissured Boom clay, interlump closure takes place after approximately 16 years while for a
fissured Boom clay it only takes 2 months. Thus, closure of the highly simplified geometry takes place within a normal construction timespan for a fissured Boom clay. It must be kept in mind that the results are highly dependant
on the geometry, lump size and soil characteristics. Therefore, these results can not be generalized into an estimate of the interlump void closure time for any stiff lumpy clay fill. In conclusion, the feasibility of MPM was explored and it turned out to be a promising method to model the interlump void closure problem. Further studies are required to check if the model gives plausible results for more advanced constitutive soil models and geometries. If the results are positive, MPM can be used as a investigation tool to increase the understanding of the interlump void closure time for more refined geometries.

The structural evolution, phases and amount of deformation, and the strain distribution of the sedimentary cover of a geologically complex area of 270 km2 in the French subalpine chains (Southeast of France; in the surroundings of the village La Motte-Chalancon) were deduced from the modelling of two folded and faulted competent layers (Tithonian and Barremian formations) both in 2D and 3D. These models were constructed from collected geological data such as BRGM (Bureau de Recherches Geologiques et Minères) geological maps, previous Bachelor’s fieldwork data, new field data and the study of aerial photographs (Google Earth). A grid of 20 vertical 2D cross-sections (10 N-S, and 10 E-W) was generated which was then imported and digitized in the geological modelling software package Move and the 3D model of the competent layers was built, validated and restored to its initial configuration prior to deformation by first removing each fault displacement (Fault Parallel Flow method) and then unfolding each structure (Flexural Slip method) in Move, showing deformation and shortening in two directions: N-S and E-W. The amount of deformation in both directions was estimated from the 3D model. Major deformation took place in the N-S direction with a maximum of 21% shortening for both layers which corresponds to an absolute shortening of 4067m for the Tithonian layer and 4076m for the Barremian layer. Less deformation took place in the E-W direction with a maximum of 9% to 10% for the Tithonian and Barremian horizons respectively which corresponds to an absolute shortening of 1697m and 1976m respectively. This resulted in an area reduction of 16% and 19% of the original area prior to deformation for the Tithonian and Barremian, respectively.
Two phases of deformation over geological time were deduced from the orientations of the folds and faults present in the studied area. The presence of NW-SE trending folds and thrust faults (e.g. Mt D’Angele fault, Pommerol fault, or Chalancon fault) along with a conjugate strike-slip system (of one oriented N-S and one NE-SW) reflects the NE-SW compressional stress regime of the first stage of deformation which is related to the Pyrenean phase of the Alpine Orogeny. The second stage, the Alpine phase, which resulted in E-W contractions, is associated with the formation of domal/basinal structures as well as folds with plunging fold axes within the studied area. The northeastern part of the studied area exhibits different fault trends. Two tear faults (L’Aiguille and Ruelles fault), a dextral N-S strike-slip fault (Establet fault), and two E-W trending reverse faults (Peyssias and Hidden fault) resulted from N-S compression. These are related to the first phase of the Pyrenean stage, and then later partly rotated during the latter phase of the Alpine stage.
The strain maps produced from the 3D model displaying the high-/low-strain zones of the area mainly show E-W and NW-SE trending strain zones which confirm the direction of the main stress regime of the Pyrenean compressional phase oriented NW-SE.

The Late Cenozoic Eridanos River brought vast amount of sediments to the North Sea Basin accumulating more than 1000m of deltaic deposits in less than 3.5 Ma. Since the gas discoveries in the shallow sediments were made, more attention is being paid to these deltaic deposits and that resulted in more data availability.
This study concentrates on the Eridanos Delta and identifies the properties of the delta in terms of its progadation distance and sedimentation rate and compare different time models to the ones provided by Overeem (2001) and Kuhlmann (2008). Furthermore, studies the reservoir rocks in terms of their petrophysical properties and facies distribution. Moreover investigates the reasoning behind the shallow sediments gas accumulations and factors behind the concentration of all the discoveries around the A and B blocks. Lastly, constructs a 3D model of the reservoirs distribution in the study area and their properties by integrating the study results of geological, geophysical, petrophysical and reservoir engineering data.
Such results were achieved by interpreting a high resolution 3D seismic survey and 2D lines in combination with wells’ wireline logs and core data. Seismically, a total of 12 units were mapped to categories the successions (3.6 – 1.8Ma) that deposited post the Mid Miocene Unconformity. In addition, more than 20 wells were used to understand the reservoir properties in which seven wells have cores. Sedimentological aspects were assessed by studying the descriptions of cores and thin sections and the measurements of the core plugs. Gas accumulations were studied by evaluating the production data for gas and linking the gas to its source.

Geological surveying is a common practice performed by geologists nowadays. Determining layer orientation parameters and identifying folds are examples of features needed in order to create a geological map or model of the (sub) surface. Surveying is expensive and time consuming, therefore such survey is recently being investigated and implemented in 3D point clouds. This thesis provides a method to quickly obtain geological information such as orientation parameters and fold information in a semi-automated manner. Obtaining this parameters is currently still a difficult problem because of data density and calculation complexity.
Fast and reliable are terms important for the created method. Large point clouds can easily jam a system and reliability is important for usability and compatibility of the created results. A method is designed which can handle bulk data, performing operations such as filtering and noise reduction assisting in data handling. Next to general calculations such as roughness filtering, intensity filtering and directional filtering of generated normals, the use of a Graphical User Interface (GUI) greatly improves the reliability of the feature calculations. The user selects areas which are potentially interesting for acquiring layer orientation information, which are then analyzed by the method itself creating a base for determining the layer orientation parameters. This base is then used to attempt the reconstruction of the bedding layers of the geological outcrop resulting in the desired geological features.
The method has been tested on the “La Charce” dataset and the “Meso-scaled fold” dataset for method validation and on the “Pradelle 2” dataset for compatibility checking. Average calculation time is with approximately 10 minutes reasonable fast. The acquired dip direction of approximately 156 degrees is very similar to that of Prentice (2017), Bisschop (2017) and the fieldwork teams of 2016. The dip angle of 52 degrees is on the other hand off compared to validation results. The layer thickness calculations are comparable, all theses show that the bulk thicknesses between ~30 and ~70 cm. The fold dataset has no validation results but the images look good. It has to be mentioned that all results obtain commission and omission errors but these are in no comparison to the results which are correct.

Process-based method forward stratigraphic modelling provides advantages in reservoir modelling by simulating the geological process mathematically, and the genesis of geologic formations over time (Michael et al., 2010). In spite of its advantages, Miller et al. (2008) have recognised significant challenges in process-based simulation models, one of them is the incapability to condition to subsurface data. The problem in conditioning the data can be addressed by using an alternative method named multiple-point geostatistics (MPS) in modelling the subsurface since its introduction in 1993 by Guardiano and Srivastava. MPS considers the relationship between multiple data points that is different from the conventional geostatistical methods that are commonly limited to using a linear relationship between data (Guardiano and Srivastava, 1993). By using the process-based simulation model as training image (TI) for MPS simulation, MPS should be able to address the conditioning issue in process-based simulation model.

Over the past decade, most research in MPS has emphasized on new algorithms for improving efficiency of MPS (Mariethoz and Caers, 2014; Mariethoz and Lefebvre, 2014), but there are still issues remain for the workflow to be widely used in geosciences. Furthermore, using nonstationary TI such as process-based simulation model in MPS are still an issue because the workflow is always different for specific cases. Until recently, little published works are available in applicating nonstationary TI in MPS.

The principal objective of this project was to determine an optimised methodology that allows the use of nonstationary process-based simulation model for TI input with MPS simulation in the fluvial-dominated delta. The process-based simulation model used in the study is a post-processed data from numerical model done in process-based modelling software Delft3D (Lesser et al., 2004) which is the PhD work of van der Vegt in 2018. There were two different cases utilised in this study that represents the whole delta development: Case A with high repetition in the patterns and Case B with low repetition in the patterns.

In order to achieve the desired outcome, this study links process-based simulation model with MPS using unconditional and conditional 2D MPS simulation with two different approaches: zonation approach and control map approach. The realisations from the unconditional simulation have to be validated until successful unconditional MPS simulation. The conditional MPS simulation were carried out when the unconditional realisations have been acknowledged as the approved results in mimicking the patterns of the Delft3D model. Lastly, the results were evaluated with four methods: connectivity function, E-type models, conditional variance models, and analysis of distance (ANODI).

In all of the MPS simulation results, the use of control map approach with unilateral simulation path proved to deliver better realisations for unconditional and conditional MPS simulations. Also, this study has presented an optimised workflow of 2D MPS simulation on using process-based simulation model in fluvial-dominated delta environment as TI based on different conditions of patterns’ repetition and hard data distribution.