The main purpose of this thesis is to develop rigorous analytical and CFD models followed by their applications to real case studies in order to: i) identify the optimum multistage compression strategies for minimising the compression and intercooler power requirements for real CO2 feed streams containing various types and amounts of impurities associated with the various types of CO2 capture technologies; and ii) investigate the buffering efficacy of realistic CO2 transmission pipelines as a line packing strategy for smoothing out temporal fluctuations in feed loading and maintaining the desired dense-phase flow for both pure CO2 and its various realistic mixtures representative of the most common types of capture technologies. An analytical model based on thermodynamics principles is developed employing Plato Silverfrost FTN95 software and applied to determine the power requirements for various compression strategies and inter-stage cooling duties for typical pre-combustion (98.07 % v/v of CO2) and oxy-fuel CO2 mixtures of 85 and 96.7 % v/v CO2 purity compressed from a gaseous state at 15 bar and 38 oC to the dense-phase fluid at 151 bar. Compression options examined include conventional multistage integrally geared centrifugal compressors, advanced supersonic shockwave compressors and multistage compression combined with subcritical and supercritical liquefaction and pumping. In each case, the compression power requirement is calculated numerically using a 15-point Gauss-Kronrod quadrature rule in QUADPACK library, and employing the Peng-Robinson Equation of State (PR EOS) implemented in REFPROP v.9.1 to predict the pertinent thermodynamic properties of the CO2 and its mixtures. In the case of determining the power demand for inter-stage cooling and liquefaction, a thermodynamic model based on Carnot refrigeration cycle is applied. The study shows that a decrease in the impurity content from 15 to 1.9 % v/v in the CO2 streams reduces the total compression power requirement by ca. 1.5 % to as much as 30 %, while for all cases, inter-stage cooling duty is predicted to be significantly higher than the compression power demand. It is found that multistage compression combined with subcritical liquefaction using utility streams and subsequent pumping can offer a higher efficiency than conventional integrally geared centrifugal compression for high purity (> 96.7 % v/v) CO2 streams. In the case of a raw/dehumidified oxy-fuel mixture, that carries a relatively large amount of impurities (85 % v/v CO2), subcritical liquefaction at 62.53 bar is shown to increase the cooling duty by as much 50 % as compared to that for pure CO2. The second part of this study focuses on the development and testing of a numerical CFD model employing Plato Silverfrost FTN95 software for simulating the transient fluid flow behaviour in CO2 pipelines with line packing. The model is based on the numerical solution of the conservation equations using the Method of Characteristics, incorporating PR EOS to deal with CO2 and its various mixtures. Following its verification, the numerical model is employed to conduct a systematic study on the impact of operational flexibility involving a temporal reduction in the upstream CO2 feed flow rate on the transient flow behaviour in the pipe over a period of 8 hours. A particular focus of attention is determining the optimum pipeline design and operating line packing conditions required in order to maximise the delay in the transition from dense phase flow to the highly undesirable two-phase flow following the ramping down of the CO2 feed flow rate. The investigations were conducted for both pure CO2 and its various realistic mixtures. For the case studies examined, the results show that the efficacy of line packing can be increased by increasing the pipeline length from 50 to 150 km for the same pipe inner diameter of 437 mm. However, as the pipelines length increased to 150 km, the increase in the pipe inner diameter beyond 486 mm was found to have no further impact on the line drafting time. While, in the case of inlet feed temperature, the line drafting time increases following an increase in the inlet feed temperature of transported fluid from 283.15 K up to 303.15 K. Beyond the operating inlet feed temperature of 311.15 K, the line drafting time only marginally increased. It is also shown that the presence of impurities reduces the transition time to two-phase flow following the ramping down of the feed flow rate.

The advances in data acquisition technologies have generated massive amounts of data that present considerable challenge for analysis. How to efficiently and automatically mine through the data and extract the maximum value by identifying the hidden patterns is an active research area, called data mining. This thesis tackles several problems in data mining, including data classification, regression analysis and community detection in complex networks, with considerable applications in various biological systems. First, the problem of data classification is investigated. An existing classifier has been adopted from literature and two novel solution procedures have been proposed, which are shown to improve the predictive accuracy of the original method and significantly reduce the computational time. Disease classification using high throughput genomic data is also addressed. To tackle the problem of analysing large number of genes against small number of samples, a new approach of incorporating extra biological knowledge and constructing higher level composite features for classification has been proposed. A novel model has been introduced to optimise the construction of composite features. Subsequently, regression analysis is considered where two piece-wise linear regression methods have been presented. The first method partitions one feature into multiple complementary intervals and ts each with a distinct linear function. The other method is a more generalised variant of the previous one and performs recursive binary partitioning that permits partitioning of multiple features. Lastly, community detection in complex networks is investigated where a new optimisation framework is introduced to identify the modular structure hidden in directed networks via optimisation of modularity. A non-linear model is firstly proposed before its linearised variant is presented. The optimisation framework consists of two major steps, including solving the non-linear model to identify a coarse initial partition and a second step of solving repeatedly the linearised models to re fine the network partition.

Molecular-level understanding of properties of hydrogen-bonding liquids and their mixtures at solid-liquid interfaces plays a significant role in several applications including membrane-based separations, shale gas production, etc. Liquid water and ethanol are common hydrogen-bonding fluids. All-atom equilibrium molecular dynamics simulations were employed to gain insights regarding the structure and dynamics of these hydrogen-bonding liquids on various free-standing solid surfaces. Models for silica, alumina, and magnesium oxide were used in these works. The results show a highly well-ordered layer of the hydrogen-bonding liquids near solid substrates and a pronounced dipolar orientation of the hydrogen-bonding molecules found in this layer, which is dependent on the surface chemistry of the substrate. Our simulated results are in good agreement with the experimental data. Many studies have paid attention to mixtures of hydrogen-bonding fluids such as liquid water-ethanol mixtures due to their critical roles in industrial applications. We have conducted simulations to examine the sorptivity, structure and dynamics of liquid water-ethanol mixtures confined in alumina pores. Analysis of the structure and dynamics suggests the possibility of using alumina as perm-selective membranes to produce anhydrous ethanol from liquid water-ethanol solutions. In addition, it is important to understand properties of mixtures of water and volatile hydrocarbons under confinement as recently water is used as fracturing-fluid to stimulate subsurface formations in the practice of hydraulic fracturing. We have investigated the behaviour of aqueous methane confined in 1 nm-wide pores obtained from materials such as silica, alumina, and magnesium oxide. Our results show that methane solubility in confined water strongly depends on the confining material, with silica yielding the highest solubility. Studying dynamical properties of confined aqueous methane suggests a direct proportional coupling between methane and water dynamics. These results help refer to multiple possible applications for fluid transport.

Lithium–sulphur batteries are considered as a promising battery system due to its high theoretical capacity (1675 A h kg−1), high energy density (~2500 W h kg−1) and the natural abundance of sulphur. However, despite intensive research there are certain limitations to be overcome to bring Li/S to practical application; these limitations stem from the multiple reactions and phase changes in the sulphur cathode. Herein, for the first time to author’s knowledge, the effect of the cathode morphology as a function of charge cycles was studied by a multi-scale 3D in-situ X-ray tomography approach. The microstructural evolution within the same Li/S cell is studied without disrupting the contents and revealing significant changes to the electrode morphology. The uneven distribution of the sulphur phase fraction within the electrode thickness and sulphur agglomeration upon cycling were shown. The advantages of in-situ X-ray tomography are compelling, enabling a non-destructive imaging of battery. Furthermore, the strategies for Li/S optimisation were reflected. A comparative study of the effect of widely available conducting polymers: polyacrylonitrile and polyaniline; and metal oxide additives: Mg0.6Ni0.4O and Al2O3; on the Li/S performance, both capacity and cycle life was conducted. Commercially viable cell configurations were developed by a simple ball milling followed by a heat treatment; the best performance was by S/PANI/ Mg0.6Ni0.4O composite with an initial discharge capacity of 1500 mA h g-1. Many problems arise due to polysulphides solubility; therefore, the optimised cathode was tested with electrolytes to investigate the effect of high concentration and viscosity, as well as LiNO3 addition. It was shown that although increasing the electrolyte concentration leads to the higher battery performance and stability, the similar results could be achieved with the addition of LiNO3. Generally, it was shown the tailoring electrolyte and electrodes parameters for Li/S cell is as important as development of efficient and easy scale-up S electrodes.

Many industrial heterogeneous catalysts often use precious metals such as Pt and Pd thanks to their ability to catalyse a vast array of chemical reactions with exceptional activity. Unfortunately, the excellent reactivity of these metals results in poor selectivity, high susceptibility to poisoning and catalyst deactivation. One strategy that has been fruitful in overcoming these shortcomings is to alloy the catalytically active metals with those that are more selective, for example the coinage metals. A special class of these bimetallic surfaces may be formed by doping the inert host metal with a sufficiently low concentration of the catalytically active metal such that these dopant atoms isolate as individual, atomic dispersed ensembles in the surface layer of the host metal; such a material is known as a Single Atom Alloy (SAA). In this thesis, we use a dual-scale theoretical approach to develop a fundamental understanding of SAAs and their behaviour in catalytic systems. On the atomistic level, we make use of density functional theory (DFT) to investigate the electronic structure of SAAs, evaluating their thermodynamic stability and quantifying their surface interactions with various chemical species. Combining data acquired from DFT with kinetic Monte Carlo (KMC) simulation, we perform dynamic studies on length scales that are more relevant to real catalysis, allowing for the prediction of catalytic metrics. In particular, we show that the surface chemical heterogeneity of a SAAs results in novel catalytic properties, arising from combined weak adsorption and low activation energies for several bond dissociation reactions; that Pt/Cu SAAs can perform low temperature C-H bond without carbon deposition; and that SAAs offer strong resistivity to catalytic poisoning. Our findings will facilitate the discovery of new alloy catalysts that exhibit novel catalytic behaviour that can be fine-tuned in terms of activity, selectivity and stability.

Narrow residence time distributions (RTDs) are extremely desirable in many chemical engineering processes where plug flow behaviour is requested. However, at low Reynolds numbers the flow is laminar resulting in strong radial velocity gradients. This in turn causes spreading of fluid particles, usually referred to as hydrodynamic dispersion. Such problem is particularly relevant to microfluidic devices operated in laminar regime due to the reduced dimension and low operating flow rates. Many solutions have been proposed to reduce the hydrodynamic dispersion: static mixers, segmented flow, secondary flow, etc. The latter relies on the action of centrifugal force inducing transversal mixing in helically coiled tubes. Further mixing and therefore reduced dispersion can be achieved by introducing geometrical disturbances, generating chaotic advection. Coiled flow inverters (CFI) exploit the beneficial effects of secondary flow and chaotic advection. They consist of sections of helically coiled tubes with 90-degree bends placed at regular intervals along a cylindrical support. Despite being a very promising solution, they have not been extensively adopted. This is due to the lack of experimental data and correlations relating the design parameters and operating conditions to the reduction of hydrodynamic dispersion. In this thesis, a flexible and reliable experimental procedure was developed to investigate RTD in microfluidic devices. It resorts to step input injections and UV-vis inline spectroscopy for detecting the concentration of tracer. The procedure was validated using Taylor’s dispersion for straight tubes. The platform was then employed to perform experiments on CFIs, constructed with microfluidic capillaries, varying operating conditions and a geometrical parameter. A similar characterization was carried out on helically coiled tubes. A significant reduction of axial dispersion was observed as compared to straight pipes, confirming the available data in the literature. It was also demonstrated that the curvature ratio primarily defines the strength of radial mixing in CFIs and therefore represents a crucial design parameter.

Fuel cells represent a promising technology for alternative electricity generation in both automotive and stationary applications. However, at present, cost and durability of the materials employed in fuel cells are barriers to commercial ubiquity. One of the main sources of cost in fuel cells is the platinum or platinum based catalysts used in the electrodes, particularly at the cathode where the sluggish oxygen reduction reaction (ORR) kinetics require high loading of precious metals. An alternative to the more widely studied polymer electrolyte membrane (PEM) acidic fuel cell is the alkaline anion exchange membrane (AAEM) fuel cell. Though the alkaline membranes are less developed than the acidic membranes used in PEM fuel cells, AAEMs are seen as a promising route to cost reduction due to the more facile ORR kinetics in alkaline media. This allows the employment of non-noble metals at the cathode, significantly reducing the amount of precious metals required in the fuel cell. However, the hydrogen oxidation reaction (HOR) kinetics (an often neglected area of study in acidic PEM fuel cells due to the negligible activation losses on the anode) in alkaline are an order of magnitude slower and thus, in order to unlock the potential of cheaper cathode catalysts, more active anode catalysts must be developed before AAEMs can be seen as a true alternative to the more established PEM technology. This thesis describes the synthesis, characterisation and electrochemical activity of a novel carbon-supported PdIr catalyst for the HOR in alkaline media. Initial synthesis methods showed the catalyst to have comparable activity with platinum through electrochemical testing, and on increasing the surface area with improved synthesis a two-fold increase in exchange current density was achieved The catalyst has been characterised with a variety of methods including SEM, HR-TEM, XRD, EXAFS and LEIS, and initial in-situ fuel cell polarisation curves are also presented.

Lithium-ion batteries are becoming increasingly energy and power dense, and are required to operate in demanding applications and under challenging conditions. Both safety and performance of lithium-ion batteries need to be improved to meet the needs of the current demand, and are inextricably linked to their microstructure and mechanical design. However, there is little understanding of the complex, multi-length scale, structural dynamics that occur inside cells during operation and failure. From the evolving particle microstructure during operation to the rapid breakdown of active materials during failure, the plethora of dynamic phenomena is not well understood. In this thesis, both ex-situ and operando X-ray imaging, and computed tomography, in combination with image-based modelling and quantification are used to characterise battery materials and components in 3D. Degradation mechanisms are investigated across multiple length-scales, from the electrode particle to the full cell architecture, and direct comparisons between materials in their fresh and failed states are made. Rapid structural evolution that occurs during operation and failure is captured using high-speed synchrotron X-ray imaging, and quantified by correlating sequential tomograms. Consistent degradation mechanisms that occur over fractions of a second are identified and are shown to contribute significantly towards uncontrolled and catastrophic failure, and previously unexplored interplay between the mechanical design of cells and their safety and performance is described. The experiments reported here assess the thermal and mechanical responses of cells to extreme operating and environmental conditions. The interaction between the dynamic architecture of active materials and the mechanical designs of commercial cells are revealed, highlighting the importance of the engineering design of commercial lithium-ion batteries and their efficacy to mitigate failure. These insights are expected to influence the future design of safer and more reliable lithium-ion batteries.

The optimisation of dynamic systems is of high relevance in chemical engineering as many practical systems can be described by ordinary differential equations (ODEs) or differential algebraic equations (DAEs). The current techniques for solving these problems rigorously to global optimality rely mainly on sequential approaches in which a branch and bound framework is used for solving the global optimisation part of the problem and a verified simulator (in which rounding errors are accounted for in the computations) is used for solving the dynamic constraints. The verified simulation part is the main bottleneck since tight bounds are difficult to obtain for high dimensional dynamic systems. Additionally, uncertainty in the form of, for example, intervals is introduced in the parameters of the dynamic constraints which are also the decision variables of the optimisation problem. Nevertheless, in the verified simulation the accumulation of trajectories that do not belong to the exact solution (overestimation) makes the state bounds overconservative and in the worst case they blow up and tend towards ±∞. In this thesis, methods for verified simulation in global optimisation for dynamic systems were investigated. A novel algorithm that uses an interval Taylor series (ITS) method with enhanced overestimation reduction capabilities was developed. These enhancements for the reduction of the overestimation rely on interval contractors (Krawczyk, Newton, ForwardBackward) and model reformulation based on pattern substitution and input scaling. The method with interval contractors was also extended to Taylor Models (TM) for comparison purposes. The two algorithms were tested on several case studies to demonstrate the effectiveness of the methods. The case studies have a different number of state variables and system parameters and they use uncertain amounts in some of the system parameters and initial conditions. Both of the methods were also used in a sequential approach to address the global optimisation for dynamic systems problem subject to uncertainty. The simulation results demonstrated that the ITS method with overestimation reduction techniques provided tighter state bounds with less computational expense than the traditional method. In the case of the forward-backward contractor additional constraints can be introduced that can potentially contribute significantly to the reduction of the overestimation. Similarly, the novel TM method with enhanced overestimation reduction capabilities provided tighter bounds than the TM method alone. On the other hand, the optimisation results showed that the global optimisation algorithm with the novel ITS method with overestimation reduction techniques converged faster to a rigorous solution due to the improved state bounds.

This research presents decision-support tools for the assessment of energy systems development at national and regional scales. For this purpose, mathematical frameworks for the design and optimisation of energy systems are developed. A methodology is proposed as a preliminary assessment of shale gas development. For this purpose, economic and environmental metrics are proposed to address different aspects of well-pad designs such as productivity and water intensity. The outcome of this methodology is included in a comprehensive optimisation-based decision-support tool developed to address the design of shale gas supply chains along with water management strategies. In this framework, the optimisation of well-pad designs is regarded as a critical decision variable. Next, implications of water scarcity, the role of economy of scales, and the impact of wastewater quality are addressed through a case study focusing on the development of shale gas supply chains in Colombia. The production of synthetic natural gas is studied as a possible substitute of natural gas. In this case, an optimisation approach is proposed to address decisions such as feedstock procurement, transportation and optimal production schemes of BioSNG and power. The mathematical framework can be implemented to investigate policies that encourage the development of renewable energy sources. The impact of uncertainty in input data is addressed through a global sensitivity analysis (GSA). The implementation of GSA assists not only in the identification of key parameters in the design of BioSNG supply chains, but also in revealing recurring trends in light of uncertainty. Finally, the development of BioSNG supply chains in the UK is investigated through the implementation of the proposed mathematical framework.