The purpose of this thesis is to model, investigate and where possible validate the impact of Emergency Shutdown Valve (ESDV) closure on mitigating the fugitive releases from failed CO2 pipelines employed as part of the Carbon Capture and Storage (CCS) chain. Additional mathematical modelling work is also presented for simulating steady-state fluid flow and mixing in CO2 pipeline networks containing the various types of impurities representative of the different capture technologies, including pre-combustion, post-combustion and oxyfuel. The pipeline rupture transient flow model, based on the numerical solution of the conservation equations using the Method of Characteristics, incorporates Wu’s Modified Peng-Robinson equation of state to deal with pipelines containing pressurised CO2. It utilises the homogeneous equilibrium flow (HEM) assumption, where the constituent phases in a two-phase mixture are assumed to be in thermal and mechanical equilibrium. The first part of this study focuses on the development and experimental validation of the CFD model for simulating the dynamic response of inline ESDV’s in limiting outflow following the rupture of pressurised pipelines. The model accounts for the pertinent valve characteristics including the activation and closure times as well as its proximity to the rupture location. The validation of the model involves comparison of its predictions against measurements taken following the controlled Full Bore Rupture (FBR) of a 113 m long, 0.15 m i.d. pipeline containing CO2 at 151 bara and 27 oC incorporating a ball valve along its length. The data recorded and simulated include the transient fluid temperatures and pressures immediately upstream and downstream of the closing valve following FBR. Excellent agreement between the two sets of data is obtained throughout the depressurisation process. The above is followed by the linking of the publically available SLAB dispersion model for heavy gas clouds to the validated outflow model. The combined model is then tested against existing experimental data from the CO2Pipetrans research project involving the blowdown of a 30 m long, 0.6 m i.d. of a CO2 pipeline from initial temperatures and pressures ranging 278 to 284 K and 104 to 156 bara respectively. The combined outflow and dispersion model is next used to determine the optimal spacing of ESDVs for CO2 pipelines. This is done by solving an optimisation problem involving trading off the 7 % (vol./vol.) CO2 concentration contour area (concentrations above this are considered fatal) against the cost for valve installation. Level diagrams are then used to determine the optimal separation distance for ESDVs. Finally, the problem of steady-state flow in pipeline networks is analysed. A flow model is developed to determine the required inlet pressure at CO2 source locations to obtain a specific delivery pressure for given source CO2 mixture compositions and flowrates. The model is then used in a realistic case study with two inlet sources and one delivery location. The required inlet pressures at the source locations are determined for given initial feed flowrates and compositions, to attain a desired delivery pressure. In addition, the downstream fluid temperature and fluid compositions are also determined.

The detrimental effect of increasing global emissions of CO2 on the environment has prompted action to be taken to improve the environmental impact of hydrocarbon-based processes and fuel use. Therefore, producing hydrogen as an alternative fuel for vehicles fitted with fuel cells through solid oxide electrolyser cells (SOECs) has been considered. Coal fired power plants are major energy providers and are operational all day. Introducing SOECs into the plant to utilise hot steam and electricity during times of low energy demand may provide a step to large scale hydrogen production. Through modelling and experimentation of power plants and SOECs, this project aims to evaluate the feasibility of an integrated system based on the thermodynamic, techno-economic and SOEC performance analyses. Results show that SOECs, which operate between 600 and 1000 °C, take advantage of the heat of the steam, which increases electrolyser efficiency. Steam from before the intermediate pressure turbine at 560 °C and 46 atm was located from a simulation of a coal fired power plant. The intermediate-temperature steam of the plant was applicable to less used Gd-doped CeO2 (CGO) than yttria stabilised zirconia (YSZ) electrolyte that performs best at 900 °C, as shown experimentally. Modelling showed SOEC efficiency was improved by 25.2 % through an integrated system rather than traditional methods of heating water to steam, due to reduced energy requirements. Furthermore, the thermoneutral point of 4,644 A m -2 (1.31 V) formed a guide for the design and operation of SOECs. Analysis on the integrated system showed that 250 MW (7500 kg hr-1) and 290 MW (8700 kg hr-1) H2 can be produced with SOECs sized at 43,300 and 50,100 m -2, respectively, for scenarios of 7% steam extraction and a purely H2 production plant, at a cost of 3.76 $ kg H2-1. Although an integrated system shows promise for large scale hydrogen production, further development for suitable electrolytes and hydrogen storage and infrastructure is required.

Microgrids are local energy providers which reduce energy expense and gas emissions by utilising distributed energy resources (DERs) and are considered to be promising alternatives to existing centralised systems. However, currently, problems exist concerning their design and utilisation. This thesis investigates the optimal design and planning of microgrids using mathematical programming methods. First, a fair economic settlement scheme is considered for the participants of a microgrid. A mathematical programming formulation is proposed involving the fair electricity transfer price and unit capacity selection based on the Game-theory Nash bargaining approach. The problem is first formulated as a mixed integer non-linear programming (MINLP) model, and is then reformulated as a mixed integer linear programming (MILP) model. Second, an MILP model is formulated for the optimal scheduling of energy consumption of smart homes. DER operation and electricity consumption tasks are scheduled based on real-time electricity pricing, electricity task time windows and forecasted renewable energy output. A peak charge scheme is also adopted to reduce the peak demand from the grid. Next, an MILP model is proposed to optimise the respective costs among multiple customers in a smart building. It is based on the minimisation/maximisation optimisation approach for the lexicographic minimax/maximin method, which guarantees a Pareto-optimal solution. Consequently each customer will pay a fair energy cost based on their respective energy consumption. Finally, optimum electric vehicle (EV) battery operation scheduling and its related degradation are addressed within smart homes. EV batteries can be used as electricity storage for domestic appliances and provide vehicle to grid (V2G) services. However, they increase the battery degradation and decrease the battery performance. Therefore the objective is to minimise the total electricity cost and degradation cost while maintaining the demand under the agreed threshold by scheduling the operation of EV batteries.

In this thesis two approximation techniques are proposed: Artificial Neural Networks (ANN) and Multi – Parametric Programming. The usefulness of these techniques is demonstrated through process optimisation case studies. The oil refining industry mainly uses Linear Programming (LP) for refinery optimization and planning purposes, on a daily basis. LPs are attractive from the computational time point of view; however it has limitations such as the nonlinearity of the refinery processes is not taken into account. The main aim of this work is to develop approximate models to replace the rigorous ones providing a good accuracy without compromising the computational time, for refinery optimization. The data for deriving approximate models is generated from rigorous process models from a commercial software, which is extensively used in the refining industry. In this work we present three model reduction techniques. The first approach is based upon deriving an optimal configuration of artificial neural networks (ANN) for approximating the refinery models. The basic idea is to formulate the existence or not of the nodes and interconnections in the network using binary variables. This results in a Mixed Integer Nonlinear Programming formulation for Artificial Neural Networks (MIPANN). The second approach is concerned with dealing with complexity associated with large amounts of data that is usually available in the refineries; a disagg regation¬aggregation based approach is presented to address the complexity. The data is split (disagg reg ation) into smaller subsets and reduced ANN models are obtained for each of the subset. These ANN models are then combined (aggregation) to obtain an ANN model which represents the whole of the original data. The disagg reg ation step can be carried out within a parallel computing platform. The third approach consists of combining the MIPA NN and the disagg reg ation¬aggregation reduction methods to handle medium and large scale training data using a neural network that has already been reduced through nodes and interconnections optimization. Refinery optimization studies are carried out to demonstrate the applicability and the usefulness of these proposed model reduction approaches. Process synthesis and MIPANN problems are usually formulated as Mixed Integer Nonlinear programming (MINLP) problems requiring efficient algorithm for their solution. An approximate multi-parametric programming Branch and Bound (mpBB) algorithm is proposed. An approximate parametric solution at the root node and other fractional nodes of the Branch and Bound (BB) tree are obtained and used to estimate the solution at the terminal nodes in different sections of the tree. These estimates are then used to guide the search in the BB tree, resulting in fewer nodes being evaluated and reduction in the computational effort. Problems from the literature are solved using the proposed algorithm and compared with the other currently available algorithms for solving MINLP problems.

Polymer Electrolyte Membrane fuel cells (PEMFC) are clean and efficient electrochemical energy converters that can be adapted to a wide range of domestic and automotive applications. However, large-scale commercialisation is hindered by issues of cost and durability relating to the catalyst layer. This work aims to address the need for cheaper and durable catalysts through the development of novel catalyst and catalyst support. The initial aim of this work is to investigate the potential application of Pd-based alloy catalyst in PEMFC. Pd is about 42% cheaper than Pt and 50 times more abundant on earth. Previous studies have shown that there is a correlation between electronic structure and catalytic activity of Pd binary alloys, and therefore it is possible to design a highly efficient Pd-based alloy catalyst. In this work, Pd-based catalyst was synthesised and characterized electrochemically in ex-situ and in-situ configurations to determine their activity and durability. It was found that Pd-based catalyst could potentially replace Pt as a low-cost anode catalyst. The second part of this work investigated the potential application of graphitic carbon nitride materials as catalyst support. Carbon black is the most widely used catalyst support for state-of-the-art PEMFCs even though it is known to undergo carbon corrosion during operation. Graphitic carbon nitride could offer enhanced durability and activity due to their graphitic structure and intrinsic catalytic properties. In addition, graphitic carbon nitride is low-cost, fairly simple to synthesise and highly tunable. In this work, various graphitic carbon nitride materials were prepared and characterised using accelerated carbon corrosion protocol. They were found to be more electrochemically stable compared to conventional carbon black. Superior methanol oxidation activity is also observed for graphitic carbon nitride supported Pt catalysts on the basis of the catalyst electrochemical surface area. However further work is needed to optimise the deposition and utilisation of metal catalyst on graphitic carbon nitride materials.

Proton exchange membrane water electrolyser (PEMWE) technology can be used to produce hydrogen from renewable energy sources; the technology is therefore a promising component in future national power and transportation fuel systems. The main challenges faced by the technology include prohibitive materials costs, maximising efficiency and ensuring suitable longevity. Therefore, research is needed to understand the internal operation of the systems so that cell design can be optimised to obtain maximum performance and longevity. PEMWE is a low temperature electrolysis system that consists of cell components such as end plates, current collectors, bipolar plates, gas diffusion layers (GDLs) and membrane electrode assemblies (MEAs). Cell performance is strongly reliant on the materials and designs of each of the components. Three cell designs were used to study different aspects of PEMWE operation: commercial cell, optically transparent cell and combined optical and current mapping cell. Polarisation measurements performed on a commercially available lab-scale test cell at ambient conditions illustrated an increase in mass transport limitations with increasing water flow rate which was confirmed using electrochemical impedance spectroscopy (EIS) measurements. A transparent cell was constructed to allow optical access to the flow channels. Measurements made on the cell showed a transition from bubbly to slug flow that affects mass transport limitations and consequently the electrochemical performance. Thermal imaging measurements supported a mass and energy balance of the system. Finally, a combined transparent and current mapping cell was constructed using PCB technology that indicated higher current densities closer to the exit of the channel. Optical measurements showed that this increase in current was associated with larger bubbles and a transition to slug flow which led to enhanced mass transport of water to the electrode surface. A model developed for the system showed that the cell potential is dominated by the anode activation overpotential. Experimental data obtained at similar conditions with the commercially available lab-scale test cell agreed well with the model and the fitted parameters were in close proximity with values published in literature.

Motivated by the fact that many novel fluid transport phenomena have been discovered at nano length scales, in this thesis I use molecular dynamics simulations to investigate how a solid surface affects the fluid properties and fluid transport in nanochannels. My ultimate goal is to search for the molecular signatures of the macroscopic observations. From the understanding of the mutual relation between molecular properties and macroscopic observations, I learn how to tailor the fluid-solid interaction to improve the performance of practical applications including nano-fluidic devices, water desalination, energy storage, and shale gas exploration. For example, in Chapter 3 I find out that liquid water can slip on hydrophilic surfaces, which contradicts conventional knowledge. The responsible molecular signature appears to be the dynamical properties of interfacial water molecules, governed by the strength of water-surface interactions and surface morphology. When water molecules can migrate from one preferential adsorption site to the next without requiring hopping events, hydrodynamic liquid slip occurs. In Chapter 4 I illustrate that the structural and dynamical properties of the electric double layer formed near graphene electrodes are crucial to the performance of supercapacitors and capacitive desalination devices. By tailoring the electrode, thin and mobile electric double layer can be obtained that can tremendously enhance the capacitance of supercapacitors and the manner that capacitive desalination devices is operated. Finally, in the study of two-phase (water and methane) flow through muscovite nanopore reported in Chapter 5 I demonstrate that the flow pattern change not only affects the movement of methane with respect to that of water but also affects the pore structure, in particular its width. As muscovite has similar structure to illite, a clay often found in shale rocks, these results advance my understanding regarding the mechanism of water and gas transport in tight shale gas formations.

Chiral molecules often show different pharmacological and toxicological properties, making their separation crucial for pharmaceutical companies. The resolution of racemic mixtures is often achieved via crystallisation methods. The lack of experimental data has been a major constraint in validating proposed computational methods for aiding the design of crystallisation processes for chiral resolution. This thesis provides both structural and thermodynamic data, and uses it to assess the limitations of current computer modelling methods. Progress in computational methods might eventually result in the design of resolving agents and hence reduce production costs of drugs and fine chemicals. Previous studies of naproxen have concentrated on the marketed enantiopure form of this anti-inflammatory drug. A crystallisation screen was conducted to identify all possible crystal phases of racemic and enantiopure naproxen. No polymorphs were detected and the crystal structure of the racemic compound was solved from powder X-ray diffraction data. The nature of the racemic species was confirmed with thermal methods, and differential scanning calorimetric and solubility measurements were used to estimate the enthalpy difference between the crystals at 156 °C and in the range of 10 to 40 °C. These data were used to test the different approximations involved in determining the energy differences between the racemic and enantiopure crystals. An extensive crystallisation screen was also performed for (1R,2S)-ephedrine 2-phenylpropionate salts. The crystal structure of the least soluble salt and three polymorphs of the most soluble salt were determined by low temperature single crystal X-ray diffraction or powder X-ray diffraction. Solubility measurements and differential scanning calorimetry were used to determine the relative stability of the salt pairs and polymorphs. These results showed the inadequacies of lattice energy calculations of the diastereomeric salt pair and their polymorphs. Experimental work on related diastereomeric salt pairs emphasised the difficulty in fully structurally and thermodynamically characterising these systems.

With the rising trend in energy demand and consumption, there is greater drive for innovation in the development of energy technologies that are more efficient and reliable. One of the technologies under consideration is the polymer electrolyte membrane fuel cell (PEMFC), which produces electrical power via the conversion of chemical energy in the form of hydrogen and an oxidant, with a by-product of water. In order for PEMFC stacks to become a serious contender in today’s market for energy production, they must compete with, and surpass, incumbent technologies in all aspects of performance, including safety and cost. One of the obstacles facing the widespread adoption of PEMFCs is the ability to manufacture a long-life stack in a cost-effective manner. Planar PEMFCs are a promising solution to these challenges as the cell configuration can be arranged to reduce the size of the stack and consequent materials needed, thereby minimising cost. This work demonstrates the development and design optimisation of a novel planar PEMFC stack with the focus of using printed circuit boards as an integrated current collector and flow field. An iterative optimisation approach of the short-term developmental analysis of the stack architecture using computational fluid dynamics and experimentation in combination with the long-term projections of the cost analysis of a PEMFC system was used to inform the discovery process of the most advantageous configurations and operating practices. Results indicate that optimisation at a single-cell level can translate into successful scalability of a larger system. Moreover, long-term analysis suggests that larger stacks of 10 kW and 80 kW based on this technology can achieve volumetric power densities in excess of 2.5 kW l-1 with costs lower than that of the US Department of Energy’s 2020 targets of 40 $ kW-1 and 450 $ kW-1 for automotive and residential CHP units respectively.

The structure of the oil-water interface of stratified flows in a 38 mm ID pipe is investigated in this Thesis with double-wire conductance probes. The fluids used – tap water and Exxsol™ D140 oil (ρo = 830 kgm-3; μo = 0.0055 kgm-1s-1) – are pumped into the facility and brought together in a “Y” inlet section, designed to minimize the mixing between phases (r = 0.6 – 2.4; Umix = 0.5 – 2.5 ms-1). The piping is made of acrylic and the flow was observed with the aid of high-speed imaging. The waves seen on the oil-water interface further downstream the inlet consists of small 3D fluctuations, rather than 2D structures. Conductance probes are used to investigate the oil-water interface in cases where clear wavy structures cannot be followed or analyzed. The signal of interface height in time is found to be stationary and follow a Gaussian distribution when the signal is collected at 256 Hz during 4 min. Based on these properties, a thorough methodology for analysis is presented, which allows estimating time-average parameters of the flow and the power spectrum of the interface. This analysis reveals that, in fully-developed flow conditions, oil and water phases show very little slip and tend to flow both at roughly the mixture velocity regardless of the flow conditions. The power spectrum detects a unique frequency of 19 Hz, but reveals that mechanic vibrations propagating through the facility are a major contribution to the structure of the interface. The 19 Hz frequency corresponds to clearly identifiable waves that develop at the inlet section only if the oil-to-water input ratio is different from 1. The power spectrum at the inlet tends to be dominated by this frequency. This finding is verified with the information of high-speed images. Wave characteristics and their evolution along the inlet are determined from high-speed images collected with a Phantom Miro 4 camera at 1,000 – 1,200 fps. The theoretical analysis of the stability of inlet waves suggests that their origin is a Kelvin-Helmholtz instability and characterizes the waves as dynamic in nature. The two types of oil-water interfaces seen (i.e. that at the inlet with 2D waves and that downstream the pipe with small 3D contributions) are discussed in this Thesis at length and abundant details are given.