Considering the rapidly rising cost of primary fuel for electricity generation and the extensive concern of the international community for global warning, electricity generation with renewable sources has been actively developed all over the world. A large number of renewable energy generators, more highly sensitive electronic equipment and more electronics or microprocessor controllers are used in the power system. It has brought new challenges to supply quality, and thus the study of Power Quality (PQ) has become obviously important. Harmonic analysis plays an important role in PQ study because harmonic has great influence on the power system equipment as well as on their operation. Harmonics can lead to operation failure of electrical and electronic components, overheating of neutral wires and transformer, failure of power factor correction capacitors, loss in power generation and transmission, and interference with protection, control and communication networks as well as customer loads. Therefore, developing an advanced PQ disturbances classification system and a more accurate harmonic analysis method is the key of this thesis. It is necessary to determine the sources and causes of such disturbances to solve PQ problems. When the type of disturbance has been classified accurately, PQ engineers can define the major effects at the load and analyse the source of the disturbances. Many approaches based on Fourier Transform (FT) and neural network for the classification of PQ disturbance have been developed in the last few years. The key factor of these methods is that the correct rate for the actual event is not high enough and thus there is still space to improve accuracy. In this thesis, a fuzzy-expert system based on Wavelet Transforms (WT) to classify power supply waveforms into different groups or categories for PQ classification is proposed with the aim which is to classify the disturbance type with higher accuracy. A new approach for the evaluation of harmonic contents of power system waveforms is also proposed in [sic] thesis. The conventional harmonic analysis method is Fourier analysis. However, Fourier analysis provides signals which are mainly localised in the frequency domain and it gives limited information of the signals in the time domain. Furthermore, the FT cannot obtain accurate values of amplitude and phases from harmonics with frequencies different from that of the window function frequency. In order to overcome the limitations of Fourier analysis and obtain better results, wavelet analysis has been proposed. A novel harmonic analysis method using Discrete Wavelet Packet Transform (DWPT) filter bank decomposition and Continuous Wavelet Transform (CWT) identification has been proposed. In order to evaluate the performance and result of the proposed analysis method, another two conventional methods, i.e. Fast Fourier Transform (FFT) and the combination method of Discrete Wavelet Transform (DWT) filter bank and CWT calculation, are compared through a large number of identical applications. Based on the harmonic analysis, the harmonic penetration is considered and its effects to power networks with increasing of renewable power generations are investigated. With increasing of renewable generators in power networks, it creates PQ problems caused by harmonic injections with a large frequency range, such as integer-harmonics, inter-harmonics and sub-harmonics. Therefore, the steady state harmonic power flow in power system with discrete frequencies is calculated with Root Mean Square (RMS) values of bus current and voltage magnitudes and Total Harmonic Voltage Distortion (THDv) values. Variable of tests are designed to investigate the effects to the harmonic penetration with multiple types of harmonic sources in power networks.

Sulphur Hexafluoride (SF6) is one of the most commonly used gases within switching applications for pulsed power applications due to a large number of desirable properties, however, it is a greenhouse gas and global concerns over the emission of this gas into the atmosphere have led to an increase in research into potential environmentally friendly alternatives. This study focused on an experimental investigation into the breakdown characteristics of two commonly used plasma closing switch topologies when filled with different gasses and gas mixtures not previously considered in as much depth for switching purposes (air, N2, 60%/40% N2/O2, 90%/10% Ar/O2, and CO2) as compared to the characteristics of the switches when filled with SF6. A self-breakdown switch and a field-distortion triggered switch topology with varying inter-electrode gap lengths up to 9mm, filled with gasses at pressures in the range 0.1MPa-0.5MPa were studied and some key operational characteristics and switching parameters such as the self-breakdown voltage of the gases, the spread in self-breakdown voltage, time to breakdown and jitter of the switches were investigated and compared. Temperature of the plasma that forms during breakdown and the conductivity of plasma was extracted for each gas and in addition to this, analysis of post-breakdown waveforms allowed for obtaining values of inductance and resistance of the switches. Experimental results have been used in the development of two computational models of Marx Generators which are used in the voltage erection stage of pulsed power systems. The models developed describe the switches as either having constant resistance or taking into account the transient plasma resistance which allows for a more accurate representation of the voltage and current behavior across the output load over the first quarter of the current oscillation after switch closure occurs.

Tidal power is a promising source of renewable energy worldwide, more stable and predictable than alternatives such as offshore wind power. However, the harsh operating environment makes maintenance of tidal turbines difficult and costly to perform. Intelligent condition monitoring systems, utilised as part of a condition-based maintenance strategy, can provide operators with timely and accurate indications of faults before serious damage occurs. Nevertheless, tidal technology is a new application, where deployments are currently in its infancy. Therefore, there is limited practical experience of how faults will develop within tidal turbines in operation. This thesis aims to investigate the requirements of condition monitoring methods, from both theory and knowledge of similar fields (such as offshore wind), to develop an approach for applying condition monitoring in new applications. This work first investigates the response of a commercial-scale tidal turbine in operation through a data mining analysis of condition data from the Andritz Hydro Hammerfest HS1000 tidal turbine. Data mining was performed following the CRISP-DM methodology and was used to build models of the normal condition response of turbine components, from which anomalous behaviour indicative of the development of faults can be detected. This approach was then expanded to include both diagnostic and prognostic modelling, where faults can be automatically classified and the remaining useful life of equipment undergoing degradation can be estimated. This has resulted in a generalised framework, based on CRISP-DM, that can be applied to perform condition monitoring in new applications, learning from the response of machinery over time during its operation.

Large-scale smart metering deployments and energy saving targets across the world have ignited renewed interest in residential non-intrusive appliance load monitoring (NALM), that is, disaggregating total household's energy consumption down to individual appliances, using purely analytical tools. Despite increased research efforts, NALM techniques that can disaggregate power loads at low sampling rates are still not accurate and/or practical enough, requiring substantial customer input and long training periods. In this thesis, we address these challenges via a practical low complexitylow-rate NALM, by proposing two approaches based on a combination of the following machine learning techniques: k-means clustering and Support Vector Machine, exploiting their strengths and addressing their individual weaknesses. The first proposed supervised approach is a low-complexity method that requires very short training period and is robust to labelling errors. The second, unsupervised approach relies on a database of appliance signatures that we designed using publicly available datasets. The database compactly represents over 100 appliances using statistical modelling of measured active power. Experimental results on three datasets from US (REDD), Italy and Austria (GREEND) and UK (REFIT), demonstrate the reliability and practicality of the proposed approaches.

High-power energy storage systems (ESS) are being considered for future aerospace platforms and other compact DC power system applications to improve the overall transient performance of electrical power distribution systems. These sources are being integrated with advanced bidirectional power electronic converter interfaces with high bandwidth control systems and current limiting functionality. To date, the literature has primarily focused on the control and behaviour of high-power ESS during normal operating conditions with an emphasis on the systems level benefits they offer. Little consideration has been given to their response during network fault conditions. Through simulation and hardware experimentation, this thesis demonstrates that an ESS, by design, can contribute significant levels of current to a fault as it attempts to sustain the network voltage. This behaviour inadvertently reduces the fault current contribution from the primary source of power on the network, reducing the effectiveness of associated protection devices (protection blinding). The impact of several key DC power system design and operation parameters on the ESS fault response is quantified and a new critical fault impedance term, beyond which protection blinding can be expected to occur, is introduced. Building upon this new knowledge, enhancements to typical compact DC power system protection schemes which more effectively account for the presence of ESS are proposed and evaluated. Differential protection schemes are shown to eliminate protection blinding whilst offering the greatest flexibility in increasing protection speed and fault discrimination, and maximising ESS availability. Adaptive protection schemes are shown to be a reliable backup option where a consistent protection system response can be obtained despite the potentially intermittent nature of the ESS fault current contribution. A novel control strategy that actively modifies the fault response of the ESS to facilitate the use of conventional overcurrent schemes is also proposed and demonstrated for applications where communications-based protection is unfavourable. The thesis concludes by proposing a framework to guide protection engineers in the selection of appropriate protection and control strategies when considering the integration of high-power ESS within compact DC power systems.

This thesis concentrates on direct drive electrical generators for wind energy applications. A variety of wind turbine configurations and generator topologies are reviewed. Direct drive renewable energy converters introduce a low speed, high torque input into the electrical machine. Due to this, these generators have to be larger and more robust than their high speed counterparts. With very large airgap closing forces, a very stiff structure capable of withstanding the stress is necessary. As a result very heavy machines, with structural ('inactive') material dominating the electromagnetically 'active' material are designed. In this thesis a stiffness approach is introduced which combines electromagnetic stiffness and structural stiffness for different modes of deflection. This is used to minimise mass of the generator by trading stiffness of rotor and stator structures. Design tools are presented, validated and utilised to model lightweight supporting structures ('inactive material') for high torque radial flux permanent magnet synchronous generators. Different structural layouts are statically studied, compared and optimised. Making use of low density materials, such as composites, a simplified generator structure is designed and contrasted with its optimised steel counterpart. As a rotating piece machinery forming part of a bigger and more complex machine, electrical generators are subject to dynamic and external forces coming from the wind turbine rotor. The optimised steel design is looked at from a dynamic viewpoint. Discussions and conclusions highlight the potential design solutions that can be adopted to minimise the mass and therefore the cost of these machines.

This work describes the design, development and characterisation of high efficiency photovoltaics (laser power converters) for the conversion of monochromatic light from a laser source into electrical energy. The technology provides a means of transmitting power wirelessly through free-space, for applications in the remote powering of electrical devices and systems. It also provides a means of efficiently transmitting power though fibre-optic cables, allowing electrical power to be delivered free from electromagnetic interference. The design of the laser power converter is considered for efficient conversion of monochromatic light at a target wavelength of 1550nm. This wavelength was chosen based on its ability to transmit through the atmosphere and silica-based fibre-optics with minimal losses. It also allows for the maximum exposure limit of 1kWm^-2 to be transmitted in free-space, which is eye- and skin-safe. Various semiconductor materials were explored for this design in terms of their maturity, band-gap tunability and lattice matching to common substrates. The laser power converter was then developed based on the material system InGaAsP/InP with a band-gap tuned to match the incident target wavelength. These cells were then characterised using a tunable laser source and the best cell achieved a conversion efficiency (at 20 degrees) of 38.9% at an irradiance of 0.73kWm^-2 at the target wavelength. However, earlier field tests conducted by Dr. Jayanta Mukherjee demonstrated an efficiency of 45% at 1kWm^-2, which is much higher than conventional single-junction solar cells and currently holds the record for monochromatic PVs operating at 1550nm. The various carrier recombination mechanisms that limit the efficiency are then investigated by measuring the cell performance down to temperatures of 100K. In this measurement the efficiency at 39Wm^-2 is shown to increase from 28.6% to 72% over the temperature range 300-100K and approaches the theoretical detailed-balance limit. An advanced temperature-dependent diode and resistance model is then formulated to predict the dominant carrier loss mechanisms at room temperature. It was found that (to a first approximation) defect-related carrier recombination dominates over the temperature range with a lifetime of 5us at room temperature. The model also determined a carrier mobility at room temperature of 12.4 cm^2V^-1s^-1 in the emitter layer, which results in a high sheet resistance and limits carrier transport to the contacts. Finally, the effects of non-uniform illumination (due to the Gaussian laser beam profile) on the device performance is investigated. A detailed carrier transport model is devised to understand the implication of non-uniform illumination on the diffusion and recombination of carriers generated in the top emitter layer. A light-beam-induced-current scan and a carrier-time-of-flight scan across the cell surface is then conducted to determine local changes in the device performance and obtain the carrier transport properties. From this the emitter diffusion coefficient and SRH lifetime (to a first approximation) were found to be 3.96cm^2s^-1 and 5us, which is in good agreement with the temperature-dependent illumination study. This work then proposes a new top contact design, which overcomes the impact of non-uniform illumination and sheet resistance.

Commercially available organic photovoltaics (OPV) are commonly fabricated using printing and coating techniques that allow for low-cost, high throughput processing of large-area OPV devices. However, the power conversion efficiency (PCE) of scaled-up OPVs is often lower than that of small-area ones. This is because the deposition techniques typically being used in industry are different to those used in research laboratories (printing/coating vs. spin-coating). Thus, detailed studies of functional materials are required to tailor the characteristics of photoactive D/A blends of OPVs in order to preserve high PCE values for scaled-up device sizes. Therefore, the aims of this thesis were to enhance the PCEs of OPV cells made using a well-known donor material (P3HT), and to develop a structured approach to fabricating large-area OPVs thus easing the transfer of fabrication procedures from laboratory to industry. To achieve the first goal, indene-C70-bis-adduct (IC70BA) was chosen as an acceptor material for a photoactive blend with P3HT. A review of P3HT:ICBA-based solar cells indicated a significant variation of reported device PCE values (average of 4.66±1.45%). The majority of reported device efficiencies were measured for OPVs with photoactive areas rarely exceeding 0.1 cm2. Therefore, a detailed study of the intrinsic characteristics of the IC70BA molecule and the morphology of the P3HT:IC70BA blends was carried out in order to design the optimal fabrication conditions for achieving higher PCEs and up-scaled device areas. Record PCEs approaching 7% were accomplished in this thesis for OPVs with photoactive areas of 0.43 cm2. This was achieved by understanding the correlation between the isomeric properties of the IC70BA molecule and the resulting D/A blend morphology depending on the fabrication conditions used. The second goal of this thesis was accomplished by designing a slot-die coating equipment that allows for the deposition of functional materials over large-areas. Different solubilised materials were deposited in ambient conditions on glass and plastic substrates in order to fabricate OPV devices. Two different photoactive D/A systems were used: P3HT:IC70BA and PCDTBT:PC70BM. OPV cell and module PCEs approaching 4% were achieved for devices with photoactive areas of about 35 cm2. The quality of the slot-die coated layers was investigated using LBIC, PL, and Raman mapping. This will allow for future improvements in the coating process and, therefore, increased device PCEs and operational lifetimes. In conclusion, the results obtained in this thesis show a way of fabricating efficient large-area OPVs without the use of the spin-coating deposition technique. The study of materials and the development of deposition procedures allows for an accessible transfer of research outcomes from the laboratory to industry.

The energy landscape of Great Britain is undergoing substantial changes, not least with rapidly increasing renewable generation and the closure of large fossil fuel plants. Energy storage has been advocated as an essential solution for mitigating the negative impact these changes may bring. Despite the apparent benefits, the uptake of energy storage in GB has been low, partly due to the uncertainty surrounding its economic feasibility. This study investigates the value of electricity storage as traded in currently available markets; three mechanisms are examined - a wholesale market, the Balancing Mechanism and the potential to provide ‘ancillary services’ to the System Operator. Optimisation models were developed to identify the maximum potential value in each market under a perfect foresight assumption, before applying a co-optimisation approach to assess a multi-market strategy. The sensitivity to perfect foresight was evaluated by testing simple storage management strategies. An econometric approach was then applied to examine the impact of increased wind generation on the markets and to extend this to a 20GW assumed scenario. The results showed that substantially larger revenues were generated under co-optimisation compared to single market participation; these were also shown to be more resilient to inter-annual variability and market constraints due to the flexibility of the storage system in adjusting participation accordingly. Cost data from previous studies suggests that such revenues are still insufficient to support the deployment of Lithium-Ion batteries and Vanadium-Redox flow batteries. Pumped Hydro Energy Storage was shown to be the most economically viable, followed by Compressed Air Energy Storage, Advanced Adiabatic Compressed Air Energy Storage and Iron-Chromium Flow batteries. Replacing perfect foresight with alternative strategies caught between 52%-62% of revenues, highlighting the importance of forecasting accuracy when drawing on optimisation models. A 20 GW wind penetration scenario showed a clear depressing effect on prices. However, in most cases examined, additional revenues were generated for storage due to the increased price volatility presenting greater arbitrage opportunities. These results imply that while energy storage can be viable, caution should be made in the choice of technology and operational strategy with a clear preference for the co-optimisation of revenues across the three market mechanisms considered.

This thesis offers a study of the Curry-Howard correspondence for a certain fragment (the canonical fragment) of sequent calculus based on an investigation of the relationship between cut elimination in that fragment and normalisation. The output of this study may be summarised in a new assignment Theta, to proofs in the canonical fragment, of terms from certain conservative extensions of the lambda-calculus. This assignment, in a sense, is an optimal improvement over the traditional assignment phi, in that it is an isomorphism both in the sense of sound bijection of proofs and isomorphism of normalisation procedures. First, a systematic definition of calculi of cut-elimination for the canonical fragment is carried out. We study various right protocols, i.e. cut-elimination procedures which give priority to right permutation. We pay particular attention to the issue of what parts of the procedure are to be implicit, that is, performed by meta-operators in the style of natural deduction. Next, a comprehensive study of the relationship between normalisation and these calculi of cut-elimination is done, producing several new insight of independent interest, particularly concerning a generalisation of Prawitz's mapping of normal natural deduction proofs into sequent calculus. This study suggests the definition of conservative extensions of natural deduction (and lambda-calculus) based on the idea of a built-in distinction between applicative term and application, and also between head and tail application. These extensions offer perfect counterparts to the calculi in the canonical fragment, as established by the mentioned mapping Theta. Conceptual rearrangements in proof-theory deriving from these extensions of natural deduction are discussed. Finally, we argue that, computationally, both the canonical fragment and natural deduction (in the extended sense introduced here) correspond to extensions of the lambda-calculus with applicative terms; and that what distinguishes them is the way applicative terms are structured. In the canonical fragment, the head application of an applicative term is "focused". This, in turn, explains the following observation: some reduction rules of calculi in the canonical fragment may be interpreted as transition rules for abstract call-by-name machines.