Search

Electrical power is essential to modern society, and is necessary for innumerable applications from lighting, heating, household appliances, to large-scale machinery, communication and transportation. Ensuring a reliable, efficient and sustainable electrical power system is therefore crucial. At present, the generation, transmission and distribution of electrical power is being rapidly transformed by advances in technology and sustainability initiatives. Some of the most important new trends include: an increasing amount of distributed generation; widespread use of power electronic converters for generation, energy storage, and loads; and increasing use of DC transmission, forming hybrid AC/DC networks. This raises new challenges and opportunities for the power system, particularly in terms of its control and optimization. In this thesis, we will consider some of these challenges for the control and optimization of contemporary and future power systems, focussing especially on hybrid AC/DC networks and converter-based microgrids. Hybrid AC/DC networks are an effective solution for future power systems, due to the increasing number of converter-based loads and distributed energy resources. The hybrid AC/DC network allows the advantages of both AC and DC networks to be combined. By using high-efficiency converters to connect the AC and DC grids, the overall efficiency of the network can be improved. However, the interconnection of AC and DC networks via interlinking converters (ILCs) to form a hybrid network brings new technological challenges, one key area being the control of such a network. The network, and especially the interlinking converter, must be controlled to ensure that the DC and AC subsystems coordinate to stabilize the network and allocate power appropriately. This is an area which has attracted considerable recent interest due to the non-triviality of the control design. In this thesis, we discuss two main contributions to the literature on hybrid AC/DC networks: the first is to present primary and distributed secondary control schemes for hybrid AC/DC networks. Using a simple model for the interlinking converter, valid for slower timescales, we prove stability and optimality. We demonstrate our proposed controllers with realistic simulations, and show improved transient performance and more accurate power allocation compared to the traditional dual-droop approach. The second part of our work considers dynamic models of the ILCs for slightly faster timescales, and proposes a passivity framework for hybrid AC/DC grids. The use of passivity allows the derivation of decentralized conditions through which the stability of the network can be guaranteed. We discuss how an appropriate ILC control design can facilitate an appropriate power allocation in the network and demonstrate the proposed design with advanced simulations. The second major area of interest is converter-based AC microgrids. Distributed generation and battery storage are being integrated into the power network at an increasing rate, and these are connected to the network via power converters. To enhance efficiency and reliability, it is envisioned that distributed generation, storage and loads will form microgrids which are able to operate autonomously if necessary. Autonomous operation poses challenges in terms of frequency and voltage regulation, requiring grid-forming inverters to coordinate the regulation of the frequency and voltage. A further challenge is to achieve an appropriate steady-state power allocation between multiple sources in the network. This thesis presents several contributions in this area. Firstly, we present a passivity framework to obtain decentralized conditions for assuring stability, and consider its applicability to grid-forming inverter control schemes in the literature. Taking existing control schemes, we assess their passivity properties numerically and discuss how these can be improved. We then design a state feedback controller which: satisfies the passivity condition for stability; allows disturbances and load changes to be regulated by contributions from multiple inverters; and allows a wide range of performance specifications to be achieved. Again, realistic simulations are used to demonstrate the proposed control approach. We also demonstrate that the passivity framework is applicable for general resistive-inductive-capacitive models of the transmission lines with an arbitrary number of states and length-varying parameters. We then present work on optimizing low voltage distribution networks with controllable battery energy storage. This is a highly relevant problem as distribution networks around the world are experiencing an uptake of energy storage systems. The proposed formulation in this chapter allows the incorporation of various important features that have not been addressed in the literature. Previous literature on convexified optimal power flow had not investigated the case of unbalanced distribution networks, especially regarding unbalance constraints, power loss in the neutral wire, and meshed network configurations. These effects are practically important in many distribution networks but existing methods were not able to incorporate this into a convex formulation. The proposed method is then used to demonstrate optimized dispatch of energy storage systems in a suitable four-wire unbalanced distribution test network. All our analytical results are network independent and have been verified with realistic case studies in MATLAB/Simulink.

In the global effort to reduce carbon emissions, renewables are helping to decarbonise our energy systems. However, their intermittent nature means they must be supported by complementary technology. One such option is water electrolysis, which uses electricity and water to produce hydrogen and oxygen. The hydrogen can then be used for sector coupling, relatively high energy density drive trains, intercontinental energy transport and seasonal storage. In order to assess the viability of hydrogen production from a given site, it is necessary to undertake technoeconomic analysis to calculate production rates and unit costs. In turn, this requires validated models of electrolysers. This work ties detailed electrolyser models into technoeconomic analyses methods, which are then applied to a number of different scenarios. I focus on islands, wind and tidal power, and constrained grids. The main outcome of the modelling aspect is that there is negligible difference (up to 1.6%) between: using a constant value for electrolyser performance (kWh/kg); modelling electrolyser performance as a function of power consumption and temperature. This justifies the use of the simple approach in large scope, low detail studies. The main outcome of the technoeconomic analysis is the production/delivery rates and unit costs in three main studies. The first was a supply driven, wind and/or tidal powered electrolyser. This study included production, storage and transport via ferries. Delivery rates ranged from 53 kg/day at £15.53/kg using curtailed power only to 185 kg/day at £7.93/kg, where non-curtailed renewable power was also used. Using curtailed power exclusively gave a low capacity factor which, because of the relatively great capital cost, resulted in expensive hydrogen. In contrast, harnessing non-curtailed renewable power increased the capacity factor and reduced the unit costs. The second study was a 2,500 kg/day demand driven system, where the resulting hydrogen is used in an essential ferry service. This considered production and storage. Here, the best levelized cost was £4.60/kg using predominantly wind power. Tidal and grid power were used as a back up, and the grid was not constrained. The optimal plant capacities obtained were a 7.85 MW electrolyser, a 4,500 kg store, a 50 MW wind farm and a 50 MW tidal farm. The third study was a supply driven system, were only production costs were considered. This used record low prices for solar power and electrolysis capex to estimate hydrogen production rates and levelised cost when an electrolyser has priority dispatch from a 2 GW solar plant. The maximum production rate was 224,000 kg/day. Levelised cost varied between £0.86 - 1.91/kg, depending on electrolyser capacity and capex. This thesis confirms that using only curtailed power results in expensive hydrogen. However, it also found competitive scenarios, e.g. priority dispatch to an electrolyser from a relatively large renewable power installation. In these cases, renewable hydrogen is set to imminently match or even undercut the cheapest hydrogen produced from fossil fuels with carbon capture and storage. The modelling work and approach to technoeconomic analysis showcased in this thesis can be taken forward in other scenarios. These studies will reveal the locations and business models through which renewable hydrogen can be produced competitively, thus becoming a viable energy vector for decarbonising energy and industry where electrification is not suitable.

This thesis proposes new techniques to improve the performance of the piezoelectric transformer (PT)-based resonant power supplies. The work is motivated by the increasing demands of smaller and higher energy density electronic circuits. The emergence of PTs provide a great opportunity to replace reactive components completely. This thesis provides improvements in this area. The state of the art of PT-based techniques are reviewed, in terms of piezoelectric material, modelling and analysis. In addition, novel techniques for design, control and voltage regulation of PT-based converters are demonstrated. New analytical modelling methods for predicting the circuit behaviour of a PT-based converter are also described. The design challenges and research gaps form the basis of this thesis. The resonant current estimation techniques are provided initially, to reconstruct resonant and detect zero crossings. Three different implementations are proposed and validated by both simulation and experimental results. These current estimation techniques are applied throughout the thesis to provide zero voltage switching (ZVS) information for the control circuit. Following the current estimation techniques, nine novel controllers based on the phase-locked loop (PLL) are demonstrated, to lock onto the phase and frequency of the resonant current and generate adequate deadtime thereby ensuring ZVS. A comparative analysis of nine variants of PLL controllers are provided in terms of noise immunity and lock-on period. These control techniques are further extended to achieve an output voltage regulation. A 30V input 5V output PT-based converter is implemented. Finally, a novel converter modelling technique with a new control approach are proposed, with simultaneous ZVS and output regulation. Detailed models are provided for the proposed technique which accurately predict circuit behaviour. An implementation which provides a regulated 5V output voltage with 10V-60V input, by varying the deadtime interval, is reported.

The aim of this thesis is to gain a deeper understanding of the anatase TiO2 nanotubes as negative electrodes in lithium-ion and sodium-ion microbatteries by researching their fundamental intercalation processes. The nano-structuring of the anatase TiO2 electrodes allows to improve their electrochemical behaviour, and thus the batteries capacities. To this end, a DFT study of the lithium and sodium insertions into the anatase bulk, slab and nanotubes has been carried out. The study shows that the anatase TiO2 exhibits a higher reactivity towards sodium than towards lithium, resulting in two different insertion mechanisms: Li ions are intercalated via a two-phases equilibrium, while Na ions insertion occurs spontaneously without phase transformation. This high reactivity towards sodium may induce Na trapping, partly explaining the irreversible capacity losses observed during the first cycle. Lithium ions are intercalated spontaneously with higher capacities into the surface than into the bulk and can explain the capacities exceeding the theoretical bulk value reported for the nanostructured electrodes. The lithium intercalation induces a gradual structural reconstruction, forming a rock salt LiTiO2 phase, but its inward growth is limited to a few nanometres. The sodium ions are favourably adsorbed above all the oxygen atoms of the anatase surface and spontaneously inserted into the surface. However, the surface reconstruction is restricted by the electronic repulsion of the Na ions. This limited stability of the sodiated surfaces can be a possible explanation of their transformation into amorphous structure. The most favourable intercalation sites of the nanotubes are located in their external wall, stabilizing the nt-TiO2 structure. Li and Na ions are also favourably adsorbed on-top the outermost oxygen atoms, explaining why more external wall exposure exhibit better capacities. Reactivity with the internal wall is spontaneous for Li ions, while the Na ions require low potentials and are associated with rupturing Ti-O bonds in agreement with the formation of amorphous nt-TiO2 below 0.5 V vs Na/Na+.

The metal halide perovskite group of materials not only have led to great advances in photovoltaics, but also have shown great potential in light emitting applications. Other than energy efficiency and brightness, perovskite emitters have advantages such as high colour saturation and low processing temperatures to compete with present inorganic and organic light emitting diode (LED) technologies. However, poor device stability continues to be the limiting factor for perovskite LEDs (PeLEDs) to be realised towards commercialisa- tion. In this thesis, the stability study was first demonstrated on methylammonium lead bromide (MAPbBr3) PeLED device under continuous operations. It was characterised in the post-biased device that the applied electric field simultaneously induces the chemical decomposition of MAPbBr3 material and the dissociation of the active Ag metal electrode. The formation of silver bromide (AgBr) marks the electrochemical degradation of MAPbBr3 and the corrosion of Ag electrode. These factors explain the short lifetime, irreversible device performance and poor emission stability during device operation. It is worth mentioning that emission stability (i.e. colour shifting) remains an unsolved problem in mixed-halide perovskites, as phase segregation problem occurs when external optical/electrical energy exceeds the threshold for halide demixing. In the following part of this thesis, vapour-assist crystallisation (VAC) technique was developed to stabilise the bromide/chloride (Br/Cl) mixed-halide phase and give spectrally stable emission. It was dis- covered that VAC-treatment slows-down the crystallisation process and promotes a two-step halide redistribution within the perovskite thin-film. The controlled crystallisation results in a mixed-halide phase with lower defects and homogeneously ordered composition. Spectrally stable emission was achieved when segregation threshold is raised well-above the maximum working condition of PeLED device. As a result, spectrally stable electroluminescence is realised at 478 nm wavelength (CIE coordinates (0.104, 0.124)) in VAC-treated rubidium- caesium-formamidinium triple-cation perovskite (RbCsFA)Pb(Br0.6Cl0.4)3. VAC-treated device luminance shows a 4-fold increase to 1638 cd/m2 and external quantum efficiency shows a 10-fold increase to 8.6 %, compared to the respective control counterparts. Further- more, VAC-treatment was applicable for improving spectral stability and device performance in phenethylammonium (PEA) incorporated PEA-CsPb(Br0.7Cl0.3)3 quasi-2d perovskite system, giving 472 nm wavelength pure-blue emission. Hence, there is potential that emitters made of VAC-treated perovskites could be pushed further to the deep-blue region without losing their colour stability and quantum efficiency. Furthermore, the role of Rb+ in Rb/Cs composition stoichiometry of perovskite was further investigated. The incorporation of Rb+ not only induces further blue-shift in emission wavelength, but also enhances photoluminescence quantum efficiency and compositional homogeneity in perovskite film. The addition of Rb+ also promotes a faster transformation of the stable perovskite phase. Finally, the optimised PeLED devices based on VAC-treated Rb0.1Cs1.2FA0.2Pb(Br1−xClx)3 (x = 0.4 - 0.45) give spectrally stable electroluminescence from blue (478 nm) to deep-blue (468 nm) region. The external quantum efficiencies from PeLEDs developed in this work gives 10.4 % at 478 nm - CIE coordinates (0.104, 0.124), and gives 5.36 % at 468 nm - CIE coordinates (0.130, 0.058), which is close to the Rec.2020 specified primary blue (0.131, 0.046). Both devices exhibit a narrow electroluminescence bandwidth, with full-wavelength at half-maximum (FWHM) given by 15.4 and 15.0 nm for blue and deep-blue PeLEDs, respectively. Spectrally stable emission is confirmed up to maximum luminance of 2485 cd/m2 and 1009 cd/m2 for blue and deep-blue PeLEDs, respectively. The performance of PeLEDs developed in this work is one of the highest reported to date in deep-blue colour regions.

In this thesis, we describe our work discovering perovskite nanocrystals with desirable optoelectronic properties for high-efficiency solar cells. Ultrafast carrier-phonon interactions in semiconductors represent a major loss of the excess energy of photoexcitations above the bandgap. We demonstrate approaches to suppress the relaxation process, which may allow the full photon energy to be usefully extracted before carrier relaxation takes place. Two distinct strategies are presented to achieve this: a hot phonon bottleneck effect and a phonon bottleneck effect. First of all, we demonstrate the synthesis of lead-based perovskite nanocrystals, tin-based perovskite nanocrystals, and tin-lead alloy perovskite nanocrystals where the electronic structure and optical properties can be tuned by composition engineering. The replacement of the toxic lead with tin may also provide a less negative impact on the environment and human health in applications. Having synthesised perovskite nanocrystals, we focus on understanding the electronic structure of nanocrystal quantum dots and the interactions between excitons, carriers and phonons. By using femtosecond spectroscopic methods, we follow the dynamics of photoexcitations inside perovskite nanocrystals. A hot phonon bottleneck effect that leads to suppressed cooling at high carrier densities is observed in both hybrid and inorganic tin-lead alloy nanocrystals. Tin addition is confirmed to prolong the carrier cooling time in nanocrystal alloys, which we attribute to suppressed Klemens decay. Since a hot phonon bottleneck only takes place at extremely high fluence which is not achievable under one sun illumination, we further explore a slow relaxation occurring at low carrier densities in tin-based perovskite nanocrystals, making it practically relevant. We attribute the slow relaxation to a phonon bottleneck effect where a decreased density of states due to quantum confinement is realised between the high-energy and the low-energy states. Photodegradation measurements reveal that the high-energy signal in transient absorption spectroscopy results from two spectrally-overlapping energy levels, both of which may contribute to the slow cooling. This makes the process complicated but more attractive. The slow relaxation we observe offers new insights into the intrinsic photophysics of perovskite nanocrystals, with direct implications for photovoltaic applications where suppressed relaxation could lead to high-efficiency solar cells.

Perovskite solar cells (PSC) are devices based on organic-inorganic perovskite-structured semiconductors that convert sunlight directly into electricity. Such solar devices have become a significant competitor in the photovoltaic field and overall in the renewable-energy race due to rapidly reaching a conversion efficiency of 25%, already comparable to the commonly used Silicon technology. Besides the unique and favourable optoelectronic properties of perovskite absorbers, their ease of synthesis and versatility has opened the possibility to produce lightweight, portable and cheap photovoltaic devices. However, most of the research still relies on inherited materials from preceding technologies, such as TiO₂ or PCBM electron transport materials (ETM), which hinders the use of a wide range of substrates and increases the manufacturing costs. Nowadays, much effort is being made by the research community to enhance further the conversion efficiency, stability and manufacturing scalability of PSCs. This work aims to investigate alternative organic and hybrid electron acceptors as ETMs in PSCs. To this purpose, six different Perylene-based organic semiconductors were designed, synthesised, characterised and assessed as ETMs and interlayers in PSCs. In Chapter 3, a series of three perylene derivatives (PDI-1 to -3) with different functional groups attached to the bay-position of the molecule's core was successfully synthesised through a five-step synthesis procedure. The optical, electrochemical and computational characterisation was carried out through different characterisation techniques. The effect that the electronic nature of the substituents has on the optoelectronic properties of the molecules was studied. The results showed that all the PDI derivatives have suitable optical and electrochemical properties, with reversible cathodic redox potential and suitable lowest unoccupied molecular orbital (LUMO) for favourable charge transfer, comparable to the values of PCBM used in an inverted PSC architecture. The PV characterisation showed comparable performance between a PDI-3 device and a PCBM device in a triple cation mixed halide perovskite. Additional photoluminescence quenching measurements showed favourable energy alignment between MAPI and PDI-1 to -3 films. In Chapter 4, the commonly used TiO₂ as ETM in conventional PSCs architecture was substituted for a low-temperature ZnO ETM prepared through a sol-gel route. Due to the chemical instability of the perovskite/ZnO interface, derivatives PDI-4 to -6 were designed as an n-type interlayer to protect the perovskite from degradation by direct contact with ZnO while avoiding detrimental effects in charge extraction due to their semiconductor nature. All molecules were fully characterised through optical, electrochemical, and computational methods. Perovskite films showed excellent stability when using the PDI interlayer deposited on top of the ZnO film. Photoluminescence quenching confirmed the charge extraction and the suitable match of LUMO levels of PDI derivatives. Photovoltaic characteristics of final devices showed that PDI-5 and -6 had comparable performance to those using the wellestablished TiO₂ ETM. Finally, based on the encouraging results obtained from the previous experience with interfacial materials. In Chapter 5, hybrid ZnO: PDI materials were synthesised to explore a "one-step interlayer" approach by incorporating PDI derivatives in the synthetic route of ZnO. The hybrid materials were studied through two different procedures: One, by mixing a PDI solution in different concentrations with a ZnO sol-gel precursor solution. The other by the hydrolysis of PDI derivatives and functionalisation of ZnO nanoparticles. Although stability results in the sol-gel procedure were not encouraging, it opens the possibility of the tailored design of PDI derivatives to dissolve and readily incorporate in the solvent system. On the other hand, the functionalised ZnO nanoparticles showed better stability; nevertheless, optimisation in film deposition and particle size is required for full coverage and proper charge extraction. In total, this project produced six different perylene-based organic semiconductors to be assessed as ETMs and interlayer materials, and also two different approaches to synthesise hybrid ZnO: PDI materials.

This thesis further improves upon the original 10 MW design of the double claw pole generator, by systematically addressing its shortcomings. The original design is a fully iron cored machine with a stationary superconducting field winding. Its structure allows it to be highly modular, reliable and cost-effective. Its main disadvantages are, when compared to other superconducting generator designs, its weight and efficiency. The double claw pole generator is a large diameter iron-cored axial-flux machine, due to these features a very stiff mechanical structure is required to main the air gap clearances, which leads to a very heavy structural mass. A novel stator design is introduced, which partially deviates the air gap closing forces into the radial direction, reducing the axial component of forces. This enabled the structural mass to be reduced from 126 tonnes to 115 tonnes. Secondly, the field core of the double claw pole machine was replaced by an inner stator. The additional stator increases the electric loading of the machine while also further increasing its modularity and improving the generator efficiency. With a target efficiency of 95 %, the power output of the generator was increased from 10 MW to 11.5 MW, while maintaining the same machine diameter and axial length. To further increase the power density and modularity, the possibility of stacking machine modules was explored. Stacking two standardised modules concentrically was found to result in a smaller and lighter machine than the original design. Additionally, the standardised modules, due to their smaller size, greatly simplify the transportation of the generator. The addition of the inner stator was found to be a very promising design. It improves the original concept of the machine in terms of power density, efficiency and modularity. It is believed that this makes the design even more competitive in the high-temperature superconducting generator market. Finally, detailed electromagnetic modelling of superconductors was performed in the electromagnetic environment relevant to electrical machines. Particular focus was put on the dynamic loss mechanisms in superconducting field windings. Through the new knowledge gained on the loss characteristics, the cooling requirements can be better understood, potentially increasing the reliability of superconducting windings and their associated cooling systems.

This thesis considers the development of one floating wave and one floating wind energy technology for their application in the Western Europe marine Exclusive Economic Zone (EEZ) and the associated port and grid connection infrastructure required to support growth in deployment. In this thesis, the Western European EEZ incorporates the marine zones of Portugal, Spain, France, Ireland, The United Kingdom, Norway, Denmark, Germany, The Netherlands and Germany. The process associated with developing infrastructure can be lengthy and complex, therefore this work presents a series of methods and evidence to expedite assessments. The core areas assessed for the development of floating wave and wind energy at array scale are evaluated as mooring suitability, power production, marine planning sensitivity and port/grid infrastructure capabilities. Three modelling processes were developed across four time frames of 2018, 2020, 2025 and 2030 to identify the following, 1) basic site potential, 2) costed site potential and 3) infrastructure assessment. The first model constrained the marine zones according to assessment criteria and identified that by 2030, an approximate 300,000km2 of floating wave or 260,000km2 floating wind across the marine zones of Western Europe. This could equate to approximately 15TW and 4.2TW of installable capacity. The second modelling approach sought to allocate sites to theoretically suitable infrastructure based on location and fixed hosting values for port and grid capacity outlining a cost of energy. It was found that a total of 38GW and 75GW of respective wave and wind capacity could be considered capable of being cost feasible, defined as less than 100£/MWh, by 2030. The third modelling approach examined the more practical nature of the two infrastructure types. A grid assessment model utilised a genetic algorithm solver to evaluate a market mix assessing the energy penetration in Western Europe. A port operations model assessed the feasible build out time. It was found that after power market and port modelling, only 3.9GW and 13.6GW of wave and wind was practically deployable. Energy market modelling highlighted that 90% of wave and 80% of wind capacity was reduced due to the lack of access to suitable volumes of demand as well as a clash between production variability due to seasonal and diurnal demand profiles. Furthermore, it was found that by 2030 the electrical grid infrastructure was twice as likely to reduce capacity potential, although impacts were location specific. This thesis also highlighted how groups of countries working in an energy partnership could connect greater numbers of potential sites within the combined EEZ. The best performer being a power union of Denmark and Norway, with vast numbers of sites in Norway being connected to the Danish electrical grid system. This highlights that policy makers could utilise this type of understanding to evaluate infrastructure but also expand the use of decision tools in infrastructure planning.

Offshore Wind has a vast potential to reduce carbon emissions and create economic prosperity, as well as increasing energy security of supply. For these potential resources to be exploited, and to build an offshore wind market fit for the future, the sector has to remain competitive, increase energy yield and reduce financing and technology uncertainties. The main focus of the industry is to create a market that drives innovation and competition, supporting growth and keeping costs down for consumers [1]. This research will assess the current market strategies of the offshore wind by organising the work into five parts: The first section will discuss the main drivers of the electricity market globally. The UK offshore wind market will be assessed to understand the main strategic drivers of the market. The third section will explore the innovative approaches used in advanced sectors, such as the automotive and aerospace industries. The review of the innovative approaches, identified the Quality Function Deployment, the Theory of Inventive Problem Solving and the Failure Mode and Effect Analysis methods as the most appropriate integrated methods that can assist the designer to consider all the interactions between technical solutions to a problem, and the necessary compromises that are required to meet the design requirements. The structured innovation approach is presented and tested in Section four, followed by the demonstration of the application of the method on a direct drive generator and floating subsystem. The results of the two case studies are discussed in the final section followed by a conclusion of the report.

Power grids are undergoing a dominant transformation due to the increasing number of power electronic devices installed in generation, transmission, distribution, and demand, etc. Compared with traditional power systems dominated by synchronous generators, the growth of power electronics induces new behaviour of power network dynamics, physical constraints, running costs, etc, which urgently calls for a new framework of modeling, control, and stability analysis. This thesis first focuses on the impedance method for modeling ac-dc power converters in power systems, designing the controllers, and analyzing the stability. A tool called Impedance circuit model is proposed to intuitively derive the impedance model of a grid-connected power converter. Compared with the conventional all-in-one impedance model, the proposed tool maps the control parameters explicitly to discrete virtual impedances in an impedance circuit and meanwhile captures the multi-control-loop interaction and coupling of a converter. Another tool named frame-dynamics-embedding impedance model is also proposed to systematically embed the frame dynamics (grid synchronization) of a power converter into its impedance model. This tool also enables the impedance model of power networks by making the reference frames of different devices consistent. The frame-transformation impedance model is finally proposed to reveal the relationship of impedance models in synchronous dq and stationary alpha/beta frames. These proposed tools make up an impedance modeling framework, which are applied to the most fundamental and widely-used converters in power systems [i.e., the voltage-source inverters (VSIs)] as examples in this thesis. Both simulation and experiment tests are conducted to validate the corresponding theoretical analysis. VSIs are the widely-used ac-dc converters which integrate sources or loads to the utility ac grids. But with the increasing use of dc in the ac-dominated power systems, an ac-dc converter is also required to perform as an interlinking converter, which links dc networks and the utility ac grids. It forms hybrid ac-dc grids and brings more integration and control problems. Hence, in addition to conventional VSIs, this thesis also explores novel topologies and control algorithms of ac-dc converters with advanced features to connect dc networks to ac grids. They are able to improve the grid integration and enhance the grid reliability (e.g., bidirectional voltage stiffness transfer, bipolar-dc-imbalance handling, dc-side fault blocking, high-frequency galvanic isolation, etc). Down-scaled prototypes for them are built and tested to validate the corresponding theoretical analysis and the proposed control algorithms. It is expected that the work in this thesis can inspire not only the fundamental framework of modeling and control analysis, but also the advanced development of converter topologies and control methods of power converters in power system applications.

Fuel cells are a technology that can potentially revolutionise the means by which we store and convert energy. High-temperature solid oxide fuel cells can do so at higher efficiencies than conventional fuel cells and can utilise the waste heat to further boost overall efficiency, while acting on a variety of fuels that could ease the transition to a clean energy infrastructure. Solid oxide fuel cell materials are currently limited by anode degradation and overall lifetime. In order to make the technology viable, these issues need addressing. One means of doing so is investigating alternative anode materials. Perovskite materials, in particular lanthanum doped strontium titanates, have been of interest recently due to their respectable electronic conductivity and stability in sulfur- and carbon-containing fuel sources. Furthermore, doping of this material has led to further functionalisation through use of exsolution: the growth of socketed catalytic nanoparticles to enhance material performance. This thesis aims to investigate the technology, literature and answer the following questions on the lanthanum-doped strontium titanate materials: 1) Is it possible to synthesise the materials by a non-solid-state route? 2) Can the number of processing steps in electrode formation be reduced? 3) Can an understanding of the exsolution process be further developed? 4) Can these materials be further improved through our understanding? The material is shown to be able to form the required phase at temperatures as low as 1000 °C through a modified synthesis route, still exhibiting the exsolution phenomenon that makes this class of materials of interest, and is also shown to produce structures with inherent porosity. This leads to the formation of one-step processed microstructures which perform slightly worse than conventionally manufactured samples of the same material. Through RC-circuit fitting and Gerischer element fitting of electrochemical impedance spectroscopy data, this change in performance is attributed to be due to the difference in porosity between the two microstructures. The formation of nanoparticles on the surface is also shown to improve performance compared to similar materials in similar conditions, thus a sensitivity study into the exsolution behaviour is undertaken. The factors that control exsolution are briefly investigated and an understanding of how these may improve the exsolution profile is developed, leading to the creation of a new composition of doped-titanate. This composition shows a considerably more nanoparticle-dense exsolution profile than the predecessors treated in the same conditions and also shows an approximately 22% improvement in electrochemical performance measured through electrochemical impedance spectroscopy. Thus, this body of work shows that the titanates can indeed be synthesised at lower temperatures, the inherent microstructures formed this way may be functionalised for use as electrodes - where our understanding of the interpretation of electrochemical impedance spectroscopy data is greatly improved - and the conditions that produce consistent, nanoparticle-dense, exsolution arrays are elucidated on, leading to demonstration of our understanding by improvement of the material.

Rechargeable lithium-ion (Li-ion) cells are widely used within portable electronics and increasingly for electric vehicles. With the increasing performance demand of Li-ion cells for today's applications, the use of organic materials as electrodes in rechargeable Li-ion cells has become increasingly attractive for the following reasons: (1) Organic electrode materials have lower environmental footprints and greater safety as compared to inorganic metal oxide electrodes; (2) organic electrode materials allow eco-efficient production and disposal; (3) organic electrode materials afford high energy storage capabilities as their structures can be engineered at the molecular level to support multiple redox reactions. However, organic small molecules suffer from dissolution in organic electrolytes leading to poor cyclic stability. To suppress the solubility of organic electrode materials, a polyimide material (PI) was synthesised by the polycondensation reaction and investigated the electrochemical performance. One of the problems of polymeric electrode materials is that they often exhibit low utilisation of the redox-active sites. In Chapter 3, to improve the electrochemical performance of PI, a series of PIX ("X" value refers to either 10, 30 and 50 = wt.% of reduced graphene oxide (rGO)) composites were synthesised through in situ polycondensation on rGO. By optimising the content of rGO in the composite, it was found that the composite with 50 wt.% rGO exhibited the largest specific capacity. All the PI and PIX composites showed good cyclic stability and rate performance. The redox mechanism was studied by ex situ FT-IR. In Chapter 4 and 5, the utilisation of redox-active sites in two covalent organic framework materials (COFs), DAPQ-COF and PT-COF, were improved using a similar strategy. The COF composites (DAPQ-COFX and PT-COFX, where X = 10, 30, and 50 wt.% of carbon nanotubes, CNT) were synthesised by in situ polycondensation on CNTs. These COF composites feature an abundance of redox-active units, ß-ketoenamine linkages, and well-defined pores. The optimised composites possess tube-type core-shell structures with intimately grown COF layers on the CNT surface. This synergistic structural design enables superior electrochemical performance. DAPQ-COF50 showed 95% utilisation of redox-active sites, long cycling stability (76% retention after 3000 cycles at 2000 mA g-1 ), and ultra-high rate capability, with 58% capacity retention at 50 A g-1 . This rate translates to charging times of ca. 11 seconds (320 C), implying that DAPQ-COF50 holds great promise for high-power cells. The rate capability outperformed all previous reports for carbonyl-contained organic electrodes by an order of magnitude; indeed, this power density and the rapid charge/discharge time are competitive with electrochemical capacitors. In addition, as the COFs contained an abundance of electronwithdraw groups, the COFs based electrode has relatively high discharge voltage as compared to other carbonyl-functionalised polymer electrode materials. What is more, by designing the electrochemically redox-active building block to increase the number of active redox sites, the capacity of the PT-COF50 composite achieved 280 mAh g-1 at 200 mA g-1 , with an average discharge capacity of 2.55 V.

The increasing adoptions of distributed energy resources (DERs) have shifted the traditional energy system to a decentralised arrangement and passive consumers to active prosumers. Traditional load forecasting that focuses on the system level has to move towards the grid edge. Hierarchical load forecasting (HLF) has been recognised as a promising solution to reveal granular load patterns and customer diversity through a two-stage process: 1) the hierarchical model that organises customers into groups and optimises the group portfolio based on the similarities of their natural attributes or consumption behaviours (i.e. customer grouping/clustering); 2) the most appropriate forecasting model is then assigned to each group to respect their distinctive characteristics. Due to the lack of a comprehensive understanding regarding the impact of customer diversity to the forecasting performance, the traditional methods connect these two stages in a sequential way that could be considered as an open-loop design. The major drawback of this open-loop design is revealed through the investigation into the relationship between customer diversity and the distribution of forecasting errors, demonstrating that the existing hierarchical models to minimise the within cluster variance do not lead to a unique forecasting model that delivers the best forecasting result for the group. This thesis addresses this challenge by aligning the optimisation objectives through a closed-loop design to identify optimal customer portfolios from the perspective of improving forecasting accuracy. The novelty of the proposed methodology lies in the introduction of a feedback mechanism to return the forecasting error as a signal for the hierarchical model to optimise both the hierarchical model and forecasting model. In this way, the hierarchical model will identify the most compatible customer group in terms of reducing the overall errors in HLF to provide additional (or alternative) criterion to cluster customers. This thesis develops two novel models to implement the proposed methodology: i) A Closed-Loop Clustering (CLC) algorithm based on statistical models: it uses the forecasting error signal from the feedback mechanism as the clustering criterion to iteratively update the hierarchical model. In this way, the hierarchical model is enhanced by re-assigning the cluster membership and the forecasting models are updated correspondingly, thereby achieving the co-optimisation of hierarchical model and forecasting model. ii) A Stacked Model Selection Network (SMSN) based on artificial intelligence: a series of forecasting models are stacked as neurons which are updated in a competitive manner and form the space for the model selection of samples. The forecasting error metric from the feedback mechanism not only serves as the loss function for the update of forecasting models but also the criterion for the winning neurons so that each neuron collects its most compatible sample group. The feedback mechanism enables the interaction between the forecasting system and the model selection system and hence can automatically select the best forecasting model for a mixed dataset. The potential benefits of the proposed methodology are demonstrated through a local energy market (LEM), which aims to enhance the local absorption of DERs. A major obstacle to maximise the value of LEM is energy forecasting due to its high level of diversities and granularities. By employing the proposed methods in local energy trading, the forecasting improvement brought by optimal prosumer grouping has increased the local PV absorption by 9.03% (CLC) and 12.35% (SMSN) and the revenue of PV owners by 6.53% (CLC) and 9.48% (SMSN) on average compared with representative hierarchical forecasting methods.

Hybrid organic-inorganic halide perovskites (HOIHPs) are a new class of material that combine molecular and valence solids to form a disordered material that exhibits flexible bonding and is susceptible to structural distortions. With applications in optoelectronics and as light-harvesters in solar cell devices, HOIHPs have been extensively researched. This thesis focuses on understanding the structure and dynamics of the archetypal HOIHP, methylammonium lead iodide (MAPbI₃), developing new methods for tuning the structure of the material and characterising new structural phases. This research started with a detailed study of the structure and dynamics of MAPbI₃ at ambient temperature, the temperature most critical to the operation of HOIHP based solar cell devices. For this purpose, traditional analysis of X-ray single crystal and neutron powder diffraction measurements were combined with maximum entropy method (MEM) analysis. This revealed the structure to be significantly more disordered than previously realised. Following the observation of interstitial iodine sites the PbI6 framework was found to exhibit distortions that when combined with analysis of variable temperature synchrotron powder diffraction data allowed a mechanism for iodine diffusion to be proposed. Bond distance analysis of both migrating and distorted iodines were consistent with the formation of neutral I2 and suggests the redox couple 2I⁻ → I₂ + 2e⁻. The proposed mechanism for iodine diffusion describes the population of interstitial iodide sites only through the collective motion of MA⁺ cations in a gate opening type mechanism. Following the significant structural disorder observed for MAPbI₃ at ambient temperature, a series of experiments aimed at characterising subtle changes to the average structure of MAPbI₃ following a number of post synthesis thermal treatments were conducted. Some of these treatments are consistent with those described in the literature for the processing of HOIHP based solar cell devices. Analysis of variable temperature synchrotron powder diffraction measurements revealed low temperature (< 80 °C) annealing in air to result in subtle distortions to the structure consistent with the incorporation of interstitial oxygen. Low temperature (< 80 °C) vacuum annealing was found to have little effect, however at higher temperatures (=80 °C) subtle changes to the tilting of the PbI6 octahedral framework were observed. This was followed by a series of X-ray single crystal and neutron powder diffraction measurements on MAPbI₃ samples annealed at high temperature and under a vacuum. This research demonstrates precise control of both the framework tilting and bond lengths of the PbI6 octahedral framework as well as disorder that is dependent on annealing temperature. This treatment was found to have no significant effect on the thermal stability of the material. The changes observed are attributed to the loss of MA+ that results in a change in the interaction between the organic A-site cation and inorganic framework, demonstrating the flexible nature of the structure that allows for a range of structures to be stabilised. After demonstrating precise control of principle structural features of MAPbI₃ following high temperature vacuum annealing, a post synthesis thermal iodine treatment was devised that resulted in the stabilisation of a new structural phase MAPbI₃+x. The structural characterisation of this phase was investigated between 100 K and 400 K through a combination of variable temperature X-ray powder and single crystal diffraction measurements. Between the measured temperature range this compound was found to exhibit four structural phases that are stabilised through the incorporation of additional interstitial iodine positioned at the centre face of the perovskite framework. At ambient temperature this phase exhibits ordering of the MA+ cations, a reduction in both PbI6 framework tilting and Pb-I bond lengths, and a decrease in the disorder of the framework that demonstrates increased rigidity of the structure. Below ambient temperature, ordering of the interstitial iodine sites is observed to correlate with tilting of the framework and ordering of the MA+. This research provides experimental evidence for intrinsic doping of MAPbI₃ and demonstrates the porous nature of HOIHPs, whilst presenting a new route for synthesising novel variants of HOIHP photovoltaic materials. This thesis demonstrates HOIHPs to be flexible, highly disordered materials. Most surprisingly, the structure of these materials is shown be highly dependent on the synthesis method and post synthesis thermal treatment undertaken. Through demonstrating precise control of the structure of these compounds, this thesis contributes towards producing more stable, highly efficient photovoltaic devices through the development of new structural phases for MAPbI₃.

The generation and transmission maintenance scheduling (GTMS) problem, in a competitive electricity market environment, presents electricity utilities scheduling their facilities for maintenance to improve productivity and maximize profits, and an independent system operator (ISO) pushing for maintenance schedules (MS) of generators and transmission facilities that keep the system reliability and minimizes operation cost. Thus, the GTMS is inherently a high-dimensional, non-linear, non-convex, and multi-objective optimization problem that contains mixed integer-real variables and conflicting objectives related to the goals of the different parties in the market. The GTMS problem is crucial in power systems operation and planning due to the increasing complexity of today's power grid, the aging of current operating electricity facilities, and the increasing share of renewable energy in the network and the market. In that sense, this thesis proposes to solve the GTMS problem using hybrid models that combine in a novel way multi-objective evolutionary algorithms (MOEAs) and classical optimization techniques to obtain a set of feasible non-dominated MS solutions. These hybrid models solve the GTMS problem in systems with thermal, hydro, and wind generation, handling maintenance and operation variables separately and sequentially, considering transmission congestion and losses, the opportunity cost in the future of water stored in reservoirs, the stochastic nature of wind generation and the impact of MS in electricity prices in the market. The models used match accepted industry maintenance practices with cutting-edge optimization techniques developed in the academia. The models are evaluated in the IEEE-RTS 24 test system, complemented with hydro units and wind farms belonging to two Bolivian electricity utilities. GENCO's profits, system adequacy, and operation costs are used as objective functions, and their conflicting relationships are evaluated in the obtained set of MS solutions. Finally, the models allow the ISO to use this set to identify the best MS solution using the technique for ordering preferences according to similarity to an ideal solution (TOPSIS) decision-making tool.

Wound field and hybrid excited synchronous machines (WFSMs/HEMs) are good candidates for applications with variable-speed requirements due to controllable field excitation. The performances of WFSMs and HEMs are significantly influenced by not only machine topologies but also parasitic effects. This thesis comprehensively investigates the DC winding induced voltage pulsation, a significant parasitic effect, in stator WFSMs and HEMs, with particular emphasis on hybrid excited switched flux machines (HESFMs). In this thesis, the investigations of the DC winding voltage pulsation initiate from the stator WFSMs and then extend to the HESFMs. Firstly, the phenomena and production mechanisms of the voltage pulsations under different conditions are investigated. It reveals that the on-load voltage pulsation contains two components, i.e., the open-circuit induced voltage and the armature current induced voltage, respectively. In the stator WFSMs, the open-circuit induced voltage is caused by the self-inductance harmonics of DC winding. While in HESFMs, due to the presence of PMs, the open-circuit induced voltage is caused by the self-inductance harmonics of DC winding and the equivalent mutual-inductance harmonics between DC winding and PMs. The armature current induced voltage is caused by the mutual-inductance harmonics between AC and DC windings in both stator WFSMs and HESFMs. The self- and mutual-inductance harmonics are introduced by the modulation effect of salient-pole rotors. Further, the influences of stator slot and rotor pole number combinations and machine topologies on the DC winding voltage pulsations are investigated. It reveals that the voltage pulsations can be actively suppressed by selecting appropriate slot/pole number combinations and machine topologies to eliminate low-order induced voltage harmonics. Finally, various rotor-based passive reduction methods are proposed and employed, including unequal rotor teeth, unevenly distributed rotor teeth, rotor step skewing, rotor pole arc pairing, rotor pole notching, and rotor pole shaping. Overall, all these methods can effectively reduce the DC winding voltage pulsation, cogging torque, back-EMF harmonics, and torque ripple, albeit with slight reduction of average on-load torque. All theoretical and finite element analyses (FEA) have been validated by experiments.

This thesis focuses on the development of the next generation of solar panels. The motivation behind this work is explained in Chapter 1. Chapter 2 gives a background to solar panels, semiconductors and thesis relevant materials while main experimental and theoretical tools used are introduced in Chapter 3. Spectroscopic measurements are used to better understand recombination and passivation in halide perovskites in Chapters 4 and 5. Specifically, low-bandgap halide perovskites are studied in Chapter 4. They are demonstrated to have significantly longer lifetimes and higher photoluminescence quantum efficiencies (PLQEs) than has previously been observed. Furthermore, zinc iodide is shown to increase these materials' tolerance to oxygen substantially. It is shown that zinc iodide increases the mixing of lead and tin, removing tin rich (oxygen sensitive) clumps from the surface of the material. Chapter 5 introduces a method to obtain ratios between recombination rates in luminescent semiconductors rapidly using PLQE measurements. Extracted rates agree well with those from transient absorption spectroscopy. It is also demonstrated that non-radiative second order processes exist in halide perovskites and cannot be explained by parasitic absorption alone. This approach will allow for faster screening of solar cell absorber materials. Chapters 6 and 7 shed new light on the role of re-emitted photons in solar cells. Photon recycling, especially its relationship to controllable parameters, is quantified in single junction solar cells and light emitting diodes (LEDs) in Chapter 6. Photon recycling and device performance are both shown to improve for increased absorber thickness, better back reflection and reduced charge trapping. However, photon recycling reduces and device performance increases for better light management. Photon recycling is also found to be significantly more important in LEDs than solar cells at operating voltages. Chapter 7 calculates the limiting efficiency of two-absorber layer all-halide perovskite and halide perovskite-silicon tandem solar cells using measured recombination rates as 40.8 % and 42.0 %. Luminescence coupling (the emission of light from the high-bandgap layer and its re-absorption in the lowbandgap layer) is found to be important in both these devices at experimentally achievable charge trapping rates. This process relaxes current matching requirements, giving tandems better spectral tolerance and allowing for lower bandgap, more stable halide perovskites to be used as the high-bandgap absorber in tandems. Chapter 8 explores a halide perovskite/singlet fission material interface. Experiments screening for triplet transfer from singlet fission materials to halide perovskites are presented. Triplet transfer was not observed in any experiment. The interface was modelled using density functional theory; its formation energy is found to be weak and triplets are shown to remain strongly localised on tetracene, even at a clean interface optimal for triplet transfer. This goes some way to explain the experimental lack of triplet transfer. Finally, Chapter 9 summarises all findings and suggests future research directions. All the work presented herein helps to pave the way towards a future with cheap, high efficiency solar panels.

Approximate computing has been widely used in error resilient design for improving the energy performance by reducing circuit complexity and allowing circuits to produce acceptable error results (approximation). Generally, the approximate computing techniques have been developed and implemented either at algorithmic level or logic level or circuit level and with no feasibility of on-the-fly or runtime change of approximation. Thus, different from the existing methods, this thesis presents novel energy-efficient integrated approach of implementing approximate computing techniques from circuit level to the algorithmic level that incorporate the change of approximation at runtime without incurring any extra hardware requirement. The two new techniques proposed are namely frequency upscaling (FUS) technique and voltage overscaling (VOS) technique. These two new techniques developed for the logic/circuit level abstract are integrated into two new proposed algorithmic level approximate computing techniques known as zigzag low-complexity approximate DCT (ZLCADCT) technique for image compression and approximate Newton method for image denoising. Thus, developing an integrated approach of implementing runtime based approximate computing techniques from circuit level abstract to algorithmic level abstract for image processing applications. In proposed FUS method, the frequency of the input values applied to an exact and approximate (AMA1) full adder cell is increased (upscaled) beyond its maximum operating value thereby generating errors in the addition operation and at the same time increasing the computational throughput. As for VOS technique, the supply voltage of exact and approximate adder cells is scaled down below the nominal voltage such that the delay in the output increases beyond the worst-case delay thereby generating errors for addition results while reducing energy dissipation. The approximation of a given circuit is realized in runtime through controlling the operating frequency and supply voltage on the circuit without the need to modify or include additional circuits. Also, FUS and VOS techniques optimise the approximation of the approximate adder cell while increasing the processing speed and decreasing the energy dissipation of the cell. It is observed that operating frequency of approximate adder cell using FUS technique can be increased to 1.4 times (11.49GHz to 16.6GHz) for the minimum approximation (2 errors) and the maximum approximation (7 errors) is achieved by increasing the operating frequency by 2.5 times (11.49GHz to 29GHz). Thus, the approximation can be varied at the runtime without the need for any additional hardware. Moreover, the processing speed of the approximate adder cell is increased as well. Also, on applying FUS technique to an exact adder cell shows that the approximate adder cell sustains a higher (around 1.3 times) frequency operation for the same approximation and results in 50% reduction in energy dissipation when compared to the exact adder cell. By applying VOS on both exact and approximate adder cells, it is observed that the approximate adder cell when compared with exact adder cell, reduces 30% energy dissipation for the maximum approximation. In addition, with VOS, the approximation of the approximate adder cell can be varied from minimum value to maximum value at the runtime without incurring any additional hardware. At the same time saving the energy dissipation from 31.1% to 87% when compared with the exact adder cell. The proposed techniques are further validated by analyzing the effect of process variations (such as gate length, supply voltage, input frequency) on applying both proposed techniques to adder cell. It is observed from the FUS technique, the decrease (increase) of both gate length variation and supply voltage variation results in reduced (increased) frequency variations for the same number of errors (approximation). Similar trend is observed in the absolute energy variation as well. When applying VOS, it is observed that the energy variation due to change of gate length for approximate full adder is significantly lower than exact full adder. In addition, the mathematical models as applicable to both exact and approximate full adders are presented for FUS and VOS techniques respectively. The results of the developed mathematical models are validated with the simulation results and it is observed that the results are in close agreement. Further using exhaustive simulations, the proposed techniques are validated by applying it on 4-bits,8-bits RCAs (Ripple Carry Adders) and subtractors followed by addition of two images using exact and approximate adder cells have been presented. From the results it is observed that when the frequency is upscaled, approximate (AMA1-based) RCA can sustain 1.18 to 1.37 times higher frequency than exact full adder-based RCA for having the maximum ER (Error rate) by keeping lower NMED (normalized mean error distance) and MRED (mean relative error distance) values. The error rate (ER) and NMED for approximate adder cell (AMA1)-based RCA is significantly lower than exact full adder-based RCA by using VOS technique; 62% of energy saving is achieved by using approximate adder-based RCA when compared with exact full adder at the maximum ER level. The PSNR results for addition of two images show that approximate circuit achieves a higher output image quality than the exact circuit by using FUS and VOS techniques. Next, at the algorithmic level of approximate computing, two new techniques at known as Zigzag low-complexity approximate DCT (ZLCADCT) technique for image compression and an approximate Newton method using approximate additions for image denoising are proposed. Furthermore, FUS technique and VOS technique are applied into proposed ZLCADCT and approximate newton method using approximate additions to show the performance evaluation across circuit level to algorithmic level. The proposed ZLCADCT is a deterministic technique that accurately configures the size of the transform matrix (T) according to the number of retained coefficients in the zigzag scanning process. This is achieved by establishing the relationship between the number of retained coefficients and the number of rows of the 'T' matrix. When compared with approximate DCT (ADCT), ZLCADCT decreases the number of addition operations and the energy consumption while retaining the PSNR of the compressed image. In addition, ZLCADCT eliminates the zigzag scanning process used in ADCT. Moreover, to characterize the deterministic operation of ZLCADCT, a detailed mathematical model is provided. A hardware platform based on FPGAs is then utilized to experimentally evaluate and compare the proposed technique; as modular, deterministic, low latency and scalable, the proposed techniques can be implemented upon any change in the number of retaining coefficients by realizing only a partial reconfiguration of the FPGA resources for the additional required hardware. Extensive simulation and experimental results show the superior performance compared with previous ADCT techniques under different metrics. Besides, when FUS and VOS are implemented respectively for ZLCADCT, approximate full adder can sustain significant higher input frequency (around 19.2GHZ by 32nm adders) and lower supply voltage (around 0.77v) when compared with an exact full adder (around 15.4GHZ and 0.83v) without having significant decreases in PSNR value. The number of completed DCT operations for ZLCADCT (2.95 to 3.53 for approximate full adder at 16.6GHZ) is higher than 2.3 to 2.77 for ADCT by using FUS technique. Total energy dissipation for voltage overscaled ZLCADCT (9.54E-10J to 6.52E-10J for approximate full adder at 0.76v) is lower than 12.1E-10J to 8.08E-10J for voltage overscaled ADCT. In the proposed approximate Newton method using approximate addition, an additional step length parameter (α) for the approximate newton method using conjugate gradient is initially introduced, such that the number of iterations and total processing time decreases for the total variation-based image denoising. Then, a floating-point adder (32-bits) made of approximate or truncated cells is applied to reduce the processing time in each iteration. The proposed technique is tested on a set of images taken from a public domain library and is found that when 1.39 < α < 1.45, the number of iterations tend to be the lowest. Moreover, the processing time of an iteration decreases significantly by applying an approximate adder at usually a very small loss of accuracy and quality of the output image; the number of iterations remains constant when the number of approximate or truncated cells in the least significant positions (given by so-called NAB) is below 10. Irrespective of the noise level and adder cell type, the quality of output images does not incur in a significant degradation when NAB < 18. Besides, by using FUS and VOS techniques, the PSNR of output images keeps nearly unchanged when NAB < 15 for both adders. At high NAB value level (NAB≥18), the frequency of AMA1(22.64GHZ) can be scaled up to be a little higher than exact full adder (21.04GHZ) for keeping the low number of iterations. By using VOS technique at high NAB value level (e.g. NAB=17 at 50% noise), the number of iterations (12.4 to 16.2) and energy consumption (8.35nJ to 14.33nJ) for AMA1 can be higher than for exact full adder (6.8 to 8.6 for number of iterations and 7.26nJ to 9.61nJ for energy consumption) at the low approximation (less than 5 output errors).

Global warming has been one of the major environmental problems of our planet since mid-20th century and it is attributed to the excessive emissions of greenhouse gases, such as CO2, by human activity. This, coupled with the finite nature of fossil carbonaceous fuels, has generated the need for alternative fuels, such as renewable hydrogen, bioalcohols etc., as well as, more efficient energy conversion devices, such as fuel cells. In this project several M/Pd and Pd/polyaniline (PANI)-based catalysts were investigated for the ethanol electro-oxidation reaction (EOR) in alkaline medium for the development of anode electrodes for alkaline direct ethanol fuel cells (ADEFCs). Specifically, PANI was tested both as a promoter for Pd and as a support material, while several transition metals were also investigated as additional promoters. Under this framework, a series of novel polyacrylonitrile (PAN) and ionic liquid (IL) core/Pd and PANI shell fibrous mat electrodes were developed and investigated in several variations and modifications, aiding to the development of polymeric electrodes for ADEFCs, that could exploit the advantages of a polymer in terms of mechanical properties, cost etc., as well as its promoting effects. All experimental samples of this project were tested mainly via cyclic voltammetry (CV) over a temperature range 25 - 60 oC. The main performance indices used was the forward peak current density at 25 oC, jP, and activations energy, Eα, as calculated by the Arrhenius plots. Also, several samples were characterised via SEM, EDX and XPS. Initially, several added metals were investigated under the same experimental conditions comparatively for simple Pd/glassy carbon (GC) electrocatalyst for the EOR in alkaline medium. Ag primarily and Bi secondarily performed the best and were distinguished as candidate promoters for polyaniline-based Pd electrocatalysts developed, tested and discussed in the following parts. Subsequently, the promoting effects of polyaniline were investigated via electrodeposited Pd/PANI/GC catalysts, that showed better catalytic performance compared to electrodeposited Pd/GC catalyst. Specifically, some of those catalysts were prepared in two steps (electropolymerisation/electrodeposition of PANI followed by electrodeposition of Pd - "2-Step"), some in one step (co-electrodeposition of PANI and Pd - "1-Step") and some in one co-electrodeposition step followed by a second Pd electrodeposition step ("2-1-Step"). Most catalysts prepared with the 2-Step method demonstrated better performance than simple Pd/GC in terms of Eα. Catalysts prepared with the 1-Step method demonstrated better electrocatalytic activity in average in terms of both performance indices than the 2-Step catalysts. Some Pd/PANI catalysts prepared via the 2-1-Step method showed the highest values of peak current density. However, in terms of activation energy, there was almost no improvement compared to the control. In terms of stability though, a better catalytic stability was achieved for a co-electrodeposited Pd/PANI catalyst via this method. Also, a novel type of PAN/IL core/PANI shell fibrous mats was developed and tested as support electrode for Pd-electrodeposited catalyst for EOR in alkaline medium. Those electrodes demonstrated the promoting effects of PANI in terms Eα, but did not achieve significant values of jP compared to Pd/GC. However, just the fact that those mats were the first independent fibrous polymer electrodes-support for Pd that were active for EOR in alkaline medium, combined with their excellent robustness and morphology, set a mark for further development upon this path. Consequently, a series of modified mats were prepared in order to investigate ways of performance improvement. Three distinct modification paths were followed. The first involved the added promoting metals distinguished in the first part (Ag and Bi), which showed encouraging results in terms of activation energy and current density up to half of that of plain Pd/GC. The second-modification-path samples consisted of phytic acid (PA)-doped PANI shell instead of HCl-doped PANI, deposited with Pd. The aim of that was to shield PANI from deprotonation in alkaline environment and enhance the catalytic performance in the long run, which was achieved in both terms of Eα and jP. Finally, the third modification path involved co-polymerisation of Pd and ANI for fibre shell preparation in one step, in a strategy similar to that followed for "1-Step" coelectrodeposided samples. Those samples were characterised by fine catalyst dispersion and showed significant promoting enhancement in terms of Eα.

The last years have seen a swift increase in the proportion of connected renewable energy sources (RESs) in modern power systems to lessen unsafe levels of CO2 in the atmosphere and cut fossil fuels dependence. Despite the environmental gains of a decarbonised power system, a higher proportion of RESs could affect the grid's safe operation in at least two ways, i.e. reduce the amount of rotational inertia and increase the uncertainty of the energy balance. As RESs typically lack storage capabilities and are connected to the grid through power electronic converters (PECs), they are decoupled from the mass of synchronously rotating generators that provide inertia and counteract frequency changes, leading to quicker frequency dynamics. Furthermore, the intermittent nature of wind and solar PV makes energy balance requirements less predictable. To assist with RES integration and provide various grid support functions, energy storage systems (ESSs) are being included in power grids worldwide. Consequently, the modern power system is becoming a more decentralised entity, with large numbers of electric vehicles (EVs) and ESSs that interact with the transmission and distribution operators and offer various auxiliary functions, e.g., reducing maximum demand and integration of renewable energy. In parallel with the development of RES, since 2010, there has been a significant increase in EVs' uptake worldwide. Currently, in the UK, the number of EVs is not substantial enough to produce noticeable effects on the power system; however, extensive penetration of EVs can have adverse if not appropriately managed. Moreover, it has long been documented that considering EVs as distributed energy resources (DERs) or equivalently as mobile ESSs, can enhance power grid stability by providing valuable services such as fast frequency response (FFR). Indeed, the decrease in system inertia induced by the expansion of RESs and the extensive deployment of grid-connected EVs are two of the most critical challenges facing the operation of present-day power systems. A reduction in the system rotational inertia induces high rates of change of frequency (ROCOF) which could activate the loss of mains (LOM) protections of DERs, causing system-wide instability. Fast-acting frequency response assets are required to cope with the rapid frequency dynamics. One of the main obstacles for the power system's secure operation under increased RES penetration is the increase in the volume of response needed for frequency containment. In this aspect, EVs and electrical ESSs (EESSs) are uniquely positioned to deliver FFR; however, these assets need an accurate estimation of their charging demand and charging infrastructure characteristics to deliver an adequate service. Furthermore, the EVs and charging equipment scattered throughout the country has different charging characteristics. Therefore, an aggregate model that considers intra-cluster variations of the EV population is needed. Of particular concern is the management of the state of charge (SOC) of these EESSs. Indeed, the asset should be capable of injecting and absorbing energy depending on the charging requirements of the transmission system operator (TSO). For frequencies above the nominal, the asset should absorb energy (increasing its SOC); conversely, for frequencies below the nominal, it should be capable of injecting energy (depleting its SOC). The thesis focuses on the impact that the FFR from EESSs and large groups of EVs have on the system frequency response (SFR). To this aim, simulations are carried out on models with various degrees of complexity to validate proposed methodologies. Equivalent models of single and multiple control area power systems are used in the development stages—the results from these simulations allowed to establish the critical timescales for the provision of FFR. The more detailed representations of the power system and the proposed aggregated EV cluster models are developed in the DIgSILENT PowerFactory simulation software through Dynamic Simulation Language (DSL). Two prominent control techniques, namely Fuzzy Logic (FL) and Reinforcement Learning (RL), are used to manage the SOC of EESSs that provide FFR. These results contribute substantially to SOC management techniques for the growing number of assets distributed throughout the electrical network, which increases the power system flexibility – especially of systems with large proportions of RES. Through detailed models and simulations, it was highlighted that by supplying FFR to the power system, groups of EVs could assist the TSO and alleviate the effect of decreased system inertia on the frequency after a disturbance. This work has also shown that increased electromobility has significant implications for operators of power distribution and transmission systems. The role of EESSs as power system stabilisers by allowing greater penetration of RESs was explored. The simulation results also supported the resonant interaction between aspects such as EV penetration and time of charging on the SFR. It was found that, even at low EV penetrations, the frequency after a disturbance improves when EVs can regulate their charging rate dynamically. Also, it was evidenced through the simulation results that faster frequency stabilisation is attained when bidirectional power flow capabilities are available. Furthermore, by developing control strategies to regulate the SOC of an M-EESS, it was possible to extract more value from the energy assets. Indeed, the control schemes developed are independent of the specific components of the M-EESS and base their control actions solely on the measured SOC. Therefore, they are relatively immune to environment perturbations. Moreover, the proposed DRL controller was compared and found to outperform conventional control techniques such as PID in keeping the asset available while delivering FFR.

Solar photovoltaics (PV) holds great promise to change the way that electricity is produced and used globally. As it stands, electricity is generated mainly by large coal and gas-fired power stations, which are expensive to build and rely on a fuel supply that becomes more expensive over time. By contrast, the costs of solar PV are falling rapidly, and solar is already producing electricity at lower levelised costs than coal and gas power stations. Moreover, it can do so with a very low environmental impact, and since it is a 'distributed' power source that does not require a fuel supply, also improves access to electricity and the overall security of supply. However, if these benefits are to be realised, deployment of solar PV needs to continue to scale significantly. Solar PV currently supplies only ~3% of worldwide electricity demand, and demand for electricity is set to nearly double by 2050 as a result of the electrification of heating and transport and rising living standards. In order to continue its rapid growth and help to meet a significant portion of future electricity demand, solar module efficiencies need to continue to rise and production costs need to decrease further. Fast-deposited thin-film PV technologies like cadmium telluride (CdTe) offer a promising route to achieve the necessary price decreases and industry scale-up because they are intrinsically less expensive to produce than the incumbent silicon PV modules, which require careful crystal growth and individual processing of each wafer, cell and module. The downside of fast thin-film deposition however is that the devices invariably have polycrystalline absorber layers with small crystal 'grains', and high defect densities. This not only limits power conversion efficiency compared to single crystal counterparts like silicon, but also makes the devices much more microstructurally and compositionally complicated, and hence more difficult to characterise and control. In particular, device-level characterisation techniques that were developed for homogeneous single crystal absorber layers are not sufficient to resolve the complexities of thin-film cells and high-resolution characterisation techniques have been under-used, slowing device development. Here we use high-resolution correlative characterisation techniques to investigate the effects of two elements that are vital to producing high efficiency cadmium telluride solar cells - chlorine and selenium. Using 3-dimensional NanoSIMS compositional mapping we find that following the essential cadmium chloride heat treatment, chlorine is not just present in grain boundaries - where it is known to have a passivation effect - but permeates every region of the CdTe absorber layer. It is found segregated at the front interface between the CdTe and the buffer layer, at incoherent twin boundaries that span grain interiors, and at dopant concentrations in grain interiors. In selenium alloyed CdTe devices we use high resolution NanoSIMS and SEM-based cathodoluminescence, on the same area of the absorber, to reveal that selenium alloying lowers non-radiative recombination levels in CdSeTe grain interiors, helping to explain the record performance of selenium-graded devices. We then use TEM-based cathodoluminescence to show that selenium also has a passivation effect on grain boundaries in CdSeTe, which is the first time that high-resolution TEM-CL mapping has been achieved on a solar cell. Together, these results help to explain how cadmium telluride devices have achieved efficiencies of over 22%, despite their fast absorber layer deposition and small grain sizes. The results suggest new routes for further efficiency improvement of CdTe solar cells, including by increasing selenium concentrations at grain boundaries and in the bulk material at the back of the absorber layer. This can reduce costs further for what is currently the lowest cost of all solar and fossil fuel electricity generation technologies, and hence help to spread the cost, security, and environmental benefits of solar photovoltaics. It is also intended that the work will encourage more high resolution, correlative characterisation of thin-film PV technologies in general.

To avoid driving climate change on a dangerous path, a substantial reduction in greenhouse gases emissions is required. Hence, a high penetration of renewable energy technologies is essential, but most renewables are either affordable or dispatchable but not both. Energy storage systems integrated into concentrating solar power (CSP) plants can enhance dispatchability and solar-to-electricity efficiency. Besides, the combination of dispatchable CSP plants with lower cost photovoltaic (PV) plants exploits synergies between the reliability of CSP with energy storage and cost of PV. However, this integration leads to complex interactions between the different technologies and requires sophisticated design guidelines to achieve low costs and high dispatchability simultaneously. In this thesis, a two-stage multi-objective optimisation framework for the design and operation of hybrid CSP-PV plants with energy storage is developed. The two-stage optimisation simultaneously optimises the design and operation of a hybrid solar power plant with respect to competing technical and financial performances. The multi-objective operational optimisation stage finds the best operational strategy of a hybrid power plant with energy storage systems. The model, written in Python, uses a typical meteorological year to optimise one-year hourly operation. The results demonstrate that the integration of an energy storage system in a concentrating solar power plant provides dispatchability and, when hybridised with photovoltaic, enhances its competitiveness with current electricity prices. The low mismatch between supply and demand, even when a fixed commitment is required throughout the year, together with high overall efficiency, indicates that the integration of energy storage in hybrid solar power plants is an opportunity to increase the penetration of solar energy in the power sector. The design of reliable and cost-competitive hybrid solar power plants requires the careful balancing of trade-offs between financial and technical performance. Hence, the design optimisation stage optimises the capacities of the main components of the hybrid power plant and handles financial and technical objectives. Different configurations are analysed as case studies throughout the thesis to analyse the impacts, interactions, and synergies of technology integration. Three locations are investigated, which present different solar resource profiles: Seville (Spain), Tonopah (USA), and the Atacama Desert (Chile). The optimisation results are used to develop some guidelines for the optimal design of dispatchable hybrid solar power plants with energy storage based on the given solar resource and required dispatchability. These guidelines provide an initial design for affordable and dispatchable hybrid solar power plants and can enable their widespread deployment. The model developed can be applied to other locations under different input parameters and demand profiles. Besides, the flexibility of the model allows it to be extended in order to evaluate different energy conversion and storage technologies to design hybrid power plants with energy storage under different configurations and requirements. Thus, the optimisation framework can provide valuable information for the integration of different technologies to support the affordable and sustainable transition to a clean energy system.

Organic solar cells offer the prospect of an entire new class of renewable energy resource. This thesis aims at addressing the issue of triplet state formation in these devices, which can contribute to both a reduction in device efficiency and operational lifetime. Furthermore, triplet states can be exploited to enhance device efficiency via photophysical processes such as triplet-triplet annihilation upconversion and thermally activated delayed fluorescence. Transient absorption spectroscopy on a timescale of ps-ms is employed as the primary investigative technique throughout this thesis. In chapter 3 we explore the dynamics of a prototypical small molecule DPP/fullerene system, which was found to elegantly demonstrate the issue of triplet state formation stemming from recombination. It was found that ultrafast spin mixing, followed by subsequent geminate recombination yielded the donor-triplet exciton, the lowest energy state of the system. In chapters 4 and 5 novel polymeric materials were investigated, which were designed with the intention of manipulating/harvesting triplet states such that they no longer constitute a loss mechanism. In chapter 4 a zinc porphyrin-F8BT hybrid copolymer solution, F8BT-HAPAPP, was found to undergo a dual energy-transfer mechanism, whereby the fate of each photogenerated F8BT singlet exciton was found to depend upon its distance from the porphyrin unit. Intriguingly, the F8BT-HAPAPP triplets generated were found to possess a lifetime intermediate between the two pristine materials. In chapter 5 a donor-orthogonal acceptor low band-gap polymer, Thiro, was found to exhibit facile triplet formation on a <200 fs timescale. This rapid triplet population was thought to derive from a spin-orbit charge transfer interaction, afforded by CT states deriving from the perpendicular substituted spirobifluorene groups. Most peculiarly, free charge formation was observed in both the solution and pristine Thiro film, as evidenced via the observation of a radical cation in the μs-TA spectra. This assignment was further validated via comparison with a Thiro:PC60BM μs-TA spectrum, which is the first time a D-oA conjugated polymer has been investigated in a bulk heterojunction.

Wind energy will play an essential role in the fight against climate change. By 2050 it is expected to be about a quarter to one third of the total electricity generation. One of the main disadvantages of wind energy is its high variability and low predictability, influenced by physical phenomena at a wide range of time and length scales. The chaotic nature of wind limits the ability of engineering models to predict the performance of wind farms. Furthermore, as wind turbines and wind farms continuously increase in size, thereby increasing their contribution to the power generation industry, the need to better understand the aerodynamic interaction between wind turbines and the atmospheric boundary layer has also increased. Computational fluid dynamics has become an essential tool to enhance our understanding of wind turbine aerodynamics, however, uncertainties are usually overlooked, due to the high computational cost and the lack of characterisation of the different sources of uncertainties. This thesis presents the development of a new computational framework for uncertainty quantification in offshore wind farms. Uncertainty quantification has been identified as one of the key research challenges in the wind energy industry and this work aims to provide a tool that facilitates the propagation of uncertainties in CFD models of wind farms. It is expected that this tool can help to increase our understanding of the wind energy physical system by increasing the amount of information obtained from CFD models providing greater insights and improving the accuracy and confidence on their predictions. The framework implemented integrates the generalized polynomial chaos method (gPC) with OpenFOAM, where a non-axisymmetric actuator disk model (ADM) has been implemented. The ADM was validated against MEXICO and NASA Ames NREL-Phase-VI experiments, and other state-of-the-art numerical models. The framework has been named gpcADM and it has been tested with relatively simple wind turbine arrays considering inflow parameters as random variables. gpcADM captures the response of the system and provides probability density functions for any quantity of interest with a reduced number of deterministic evaluations compared to other traditional sampling strategies.

The power grid is undergoing a transition process characterised by increased interconnectedness and heterogeneity. This makes maintaining its reliability and quality of supply indicated by a key service known as load frequency control (LFC), a more complex task due to the need for flexible control methods which can effectively multiple energy resources working towards a common objective. Battery energy storage systems (BESS) are key devices that can support LFC. This thesis is concerned with the development of model predictive load frequency control (MPLFC) strategies incorporating battery energy storage systems. A review of the MPLFC applications within legacy and future grids is given including BESS LFC applications. A model consisting of generators and BESS which includes all dynamic subsystem interactions is then developed using the deregulated power system framework. Centralised model predictive load frequency control is applied to this system for the cases of both contracted loads and uncontracted load demands occurring in the network. Limits on subsystem inputs, incremental generator rates, BESS power and energy are considered including BESS state of charge management. The centralised model is a large scale heterogeneous system of coupled subsystems performing a common task of LFC but having different control loops. For independent flexible controller design decentralised control is often desirable. Hence, local decentralised MPC controllers are designed for the BESS and generator subsystems. However, this requires creating suitable decomposed prediction models, from a system of strongly connected subsystems of power generators. In addition, the design process should have stability guarantees despite the dynamics ignored in the decentralised prediction models without compromising control performance. A model decomposition technique that explicitly accounts for all dynamic interactions, while ensuring overlapping information is incorporated in local controllers is adopted. Stability is guaranteed using the inherent robustness property of MPC with the assumption that interacting dynamics ignored by the decentralised controllers are within established interaction bounds linked to inertia of the power system. Simulations show stability is achieved when this approach is applied to LFC. This work is then extended to account for the temporal nature of the future grid characterised by slow and fast states. A multi time scale hierarchical MPC algorithm is developed where the challenge for MPC in multi time scale systems is selecting a suitable sampling time that ensures acceptable dynamics responses at each time scale. Hence, multirate sampling is employed to account for two timescales. The algorithm was then extended the case of decentralised lower controllers for independent systems with strong dynamic coupling. Simulations show that the LFC objective of regulating the frequency deviation to zero is achieved with the proposed methods.

Sodium ion batteries (SIBs) are a promising substitute for lithium ion batteries (LIBs) because of the natural abundance and lower price of sodium. Hard carbon (HC), known as "non-graphitizable" carbon, is the most popular negative electrode material in SIBs. In this thesis, we investigate the development of composite materials in which the hard carbon is combined with a sodium conversion material with the aim of improving the capacity and cycling stability. Hard carbon obtained from cotton wool at 1400 °C shows a best reversible capacity of 319 mA h g-1 at current of 20 mA g-1 . The sodium storage analysis is consistent with the traditional insertion/absorption mechanism in the hard carbon. The slope region is more related with interlayer distance and degree of graphitization while the plateau part is more related to the micropores size and volume. An effective route to synthesis composites of metal nitrides and carbides with carbon was reported. Titanium tetrachloride is reacted with hydroxide groups on cellulose (cotton wool) before firing to convert the cellulose to hard carbon. Hard carbon-nanocrystalline titanium nitride composites with a good distribution of the titanium across the fibrous hard carbon structure were obtained by firing the treated cellulose under nitrogen. Hard carbon-nanocrystalline titanium carbide composites were obtained by carbonized under argon. Both composites show similar first cycle capacities to hard carbon, but the titanium nitride composite delivers a better capacity retention (85.2%) after 50 cycles than that of hard carbon (74.3 %). Ex situ grazing incidence XRD patterns of the HC-TiN composites suggest the reactions occurring only on the surface region of TiN. VN-HC composites have been synthesised using the same pyrolysis process after reacting VOCl3 with cellulose. The introduction of VN produces an increased capacity: with addition of 8.6 wt% VN, the hard carbon-based electrode achieves a first cycle reversible (oxidation, de-sodiation) capacity of 354 mA h g-1 at 50 mA g-1 , while with pure hard carbon it is 302 mA h g-1 . The additional specific capacity achieved upon addition of VN, compared with the pure hard carbon, is 605 mA h g-1 when referred to the mass of VN only, which is the highest capacity of VN materials in sodium-ion batteries reported to date. In addition, VN also improves the capacity retention with cycling: after 50 cycles the reversible capacity of hard carbon electrodes with 8.6 wt% VN is 294 mA h g-1 , while with pure hard carbon it is 239 mA h g-1 . Insights into the reaction mechanism are obtained by ex situ characterization of the discharged and charged electrodes. Amorphous silicon nitride and silicon oxycarbides were obtained at 1200 °C. The as-prepared silicon nitride coated on hard carbon shows a reversible capacity of 351 mA h g-1 at 50 mA g-1 , better than that of 284 mA h g-1 from pure hard carbon. Furthermore, hard carbon coated with silicon nitride delivers a capacity retention of 85.4% in 50 cycles. The surface evolution of electrode before and after reduction cycling has been investigated by XPS measurement.