The drive to lower the environmental impact of power generation has underlined the importance of distributed generation (DG). DG allows a multitude of dispersed renewable technologies to be included within the energy supply network. The energy generation of a DG installation doesn’t necessarily coincide with local power consumption; grid connec- tion allows surplus local power to be distributed using the wider power network. This results in a variety of DG units requiring grid connection. A power electronics interface is commonly needed to achieve connection between the DG unit and the distribution network. Whilst DG units are available in a multitude of sizes, the focus of this work is domestic scale DG. Single phase power inverters are commonly used to connect DG units to the utility. An issue associated with the interconnection of generators within the distribution network is the formation of power islands. A power island is defined as a section of the power network, consisting of generators and loads, which becomes electrically isolated from the wider power network. The majority of grid connection standards stipulate that the grid con- nection power electronics interface must include a robust loss of mains (LOM) detection routine. Once a LOM event has been detected the output power of the DG unit must be reduced to zero to guarantee no power island exists. This thesis details the work carried out during the completion of an En- gineering Doctorate (EngD) Degree in Power Electronics, Machines and Drives. A low voltage laboratory test bench and associated simulation model have been designed and constructed to allow multiple in-the-loop based LOM detection methods to be presented, analysed and compared. A new LOM detection technique has been created, referred to as the proposed technique. The proposed technique is a hybrid LOM detection technique which uses a passive routine to signal when a LOM event may have occurred and an active technique to confirm the LOM event. The passive routine uses Fourier analysis to constantly monitor the magni- tude and spread of high frequency voltage components present at the DG unit connection point. The active confirmation routine is an active power shift function. A fully rated 500W laboratory test bench was created which allows the proposed technique to be verified at power levels more realistic for a standard DG unit installation.

Proton exchange membrane fuel cells (PEMFCs) are promising electrochemical energy ® conversion devices, which are based on high cost materials such as Nafion membranes. The high cost and limited availability of noble metals such as Pt hinder the commercialisation of PEMFCs. The research described in this thesis focused on the development of composite materials and functionalised polymer membranes for intermediate temperature PEMFCs that o operate in the temperature range of 120 to 200 C. A higher operating temperature would enhance the kinetics of the cell compared to a perfluorinated polymer membrane based cell and provide a greater opportunity to use non-noble metal electrocatalysts. Inorganic–organic composite electrolyte membranes were fabricated from Cs substituted heteropolyacids (CsHPAs) and polybenzimidazole (PBI) for application in intermediate temperature hydrogen fuel cells. Four caesium salts of heteropolyacid, (CsHPMoO X3-X1240 (CsPOMo), CsHPWO(CsPOW), CsHSiMoO(CsSiOMo) and CsHX3-X1240 X4-X1240 X4- SiWO(CsSiOW)) and an ionic liquid heteropolyacid were used to form composite X1240 , membranes with PBI. The membranes were characterised by using SEM, FTIR and XRD. The CsHPA powders were nano-size as shown in the XRD and SEM data. The CsHPA/PBI composite membranes, loaded with HPO had high conductivity, greater than that of a 34 phosphoric acid loaded PBI membrane. Cs substituted heteropolyacid salt showed better enhancement of conductivity than that provided from ionic liquid heteropolyacid salt. The conductivity increased with an increase in the percentage of powder in the composite. The 30% -1 CsPOMo/PBI/HPO exhibited a conductivity of 0.12 S cm under anhydrous conditions 34 although its mechanical strength was the poorest, but still promising with a value of 40 MPa. The performance of the hydrogen fuel cell with composite membranes was better than that with a phosphoric acid-doped PBI membrane under the same conditions. The CsPOMo gave -2 the best power density, of around 0.6 W cm with oxygen at atmospheric pressure. A novel method was used to prepare poly (ethylene oxide)/graphite oxide (PEO/GO) composite membrane aimed for low temperature polymer electrolyte membrane fuel cells without any chemical modification. The membrane thickness was 80 µm with the GO content was 0.5 wt. %. SEM images showed that the PEO/GO membrane was a condensed composite material without structure defects. Small angle XRD for the resultant membrane results showed that the d-spacing reflection (001) of GO in PEO matrix was shifted from2θ=11º to 4.5 º as the PEO molecules intercalated into the GO layers during the membrane -1 preparation process. FTIR tests showed that the vibration near 1700 cm was attributed to the -COOH groups. The ionic conductivity of this PEO/GO membrane increased from 0.086 S -1 -1 cm at 25 ºC to 0.134 S cm at 60 ºC and 100% relative humidity. The DC electrical resistance of this membrane was higher than 20 MΩ at room temperature and 100% relative humidity. Polarisation curves in a single cell with this membrane gave a maximum power -2 density of 53 mW cm at temperature around 60 ºC, although an optimised catalyst layer composition was not used. Polybenzimidazole/graphite oxide (GO /PBI), sulphonated graphite oxide/PBI and ionic liquid GO/PBI composite membranes were prepared for high temperature polymer electrolyte membrane fuel cells. The membranes were loaded with phosphoric acid to provide suitable proton conductivity. The PBI/GO and PBI/SGO membranes were characterised by XRD which showed that the d-spacing reflection (001) of SGO in PBI matrix was shifted from 2θ=11º, meaning that the PBI molecules were intercalated into the SGO layers during the membrane preparation. A low acid loading reduced the free acid in the membranes which avoided water loss and thus conductivity loss. The ionic conductivities of the GO /PBI and -1 -1 SGO/PBI and ILGO/PBI membranes, with low acid loading, were 0.027 S cm , 0.052 S cm -1 and 0.025 S cm at 175 ºC and 0% humidity. Fuel cell performance with SGO/PBI -2 membranes gave a maximum power density of 600 mW cm at 175 ºC. A quaternary ammonium PBI was synthesised as a membrane for applications in intermediate temperature (100-200°C) hydrogen fuel cells. The QPBI membrane was loaded with phosphoric acid (PA) to provide suitable proton conductivity and compared to that of a similar PA loading of the pristine PBI membrane. The resulting membrane material was characterised in terms of composition, structure and morphology by NMR, FTIR, SEM, and −1 EDX. The proton conductivity of the membrane was 0.051 S cm at 150 °C and a PA acid loading of 3.5 PRU (amount of HPO per repeat unit of polymer QPBI). The fuel cell 34 -2 performance with the membrane gave a peak power density of 440 mW cm and 240 mW −2 cm at 175 °C using oxygen and air, respectively. Inorganic–organic composite electrolyte membranes were fabricated from CsHPMoO X3-X1240 CsPOMo and quaternary diazabicyclo-octane polysulfone (QDPSU using a polytetrafluoroethylene (PTFE) porous polymer matrix for applications in intermediate temperature (100-200°C) hydrogen fuel cells. The CsPOMo/QDPSU/PTFE composite membrane was made proton conducting using a relatively low phosphoric acid loading to provide the membrane conductivity without compromising the mechanical strength to a great extent. A casting method was used to build a thin and robust composite membrane. The resulting membrane materials were characterised in terms of composition, structure and morphology by EDX, FTIR and SEM. The proton conductivity of the membrane was 0.04 S -1 cm with a PA loading of 1.8 PRU (amount of HPO per repeat unit of polymer QDPSU). 34 -2 The fuel cell performance with the membrane gave a peak power density of 240 mW cm , at 150 °C and atmospheric pressure. A composite material for phosphoric acid (PA) loaded membrane was prepared using a porous polytetrafluoroethylene (PTFE) thin film. N, N-Dimethylhexadecylamine partially - quaternised poly (vinyl benzyl chloride) (qPVBzCl ) was synthesised as the substrate for the - phosphoric acid loaded polymer membrane. The qPVBzCl was filled into the interconnected - pores of a PTFE thin film to prepare the PTFE/qPVBzCl membrane. A SEM data indicated - that the pores were filled with the qPVBzCl . The PA loading was calculated to be on average 4.67~5.12 per repeat unit. TGA results showed that the composite membrane’s was stable at intermediate temperatures of 100°C to 200 °C. The composite membrane’s tensile stress was 56.23 MPa, and Young’s Modulus was 0.25GPa. The fractured elongation was 23%. The - conductivity of the composite membrane after PA addition (PTFE/qPVBzCl /HPO) 34 -1 -1 increased from 0.085 S cm to 0.1 S cm from 105°C to 180 °C. The peak power density of the H/O fuel cell, at 175 °C under low humidity conditions (<1%), with the22 PTFE/qPVBzCl /HPO 34 membranes was 360 mW cm.

The operating temperature of the components within an electronic device has a significant impact on the reliability of a product. In a variable speed drive the power semiconductors in the inverter stage are often operated close to their maximum temperature when the inverter is operating at a low output frequency or during an overload. The temperature of these components must be continuously monitored to prevent them from overheating, but direct measurement of the temperature is only possible if a special test configuration can be used. This is not practicable in a commercial drive and to protect the inverter the temperature of the power semiconductors must be estimated by an on-line thermal model. The work presented in this thesis describes the development of a novel thermal model that can be implemented using the existing computational resources available in a commercial variable speed drive. The thermal model is based on the transient thermal impedance measured between each device and the internal thermistor in a power module. These form a thermal impedance matrix which can be used to calculate the instantaneous temperature of every device in the inverter. However, with the existing computational resources it is not possible to implement the complete matrix without aliasing. To reduce the risk of aliasing the number of calculations performed during each sample period must be reduced. This is achieved by using a frequency domain model that has been developed to calculate the peak temperature of the hottest devices. To validate the thermal model it has been implemented in a commercial drive. The drive has been modified to allow the temperature of the power semiconductors in the inverter to be measured using a high speed thermal camera. This allows the temperature estimated by the on-line thermal model to be compared directly with the temperatures measured when the inverter is operating under typical load conditions. Comparisons of the measured and estimated temperatures in several operating conditions are presented. These conditions were chosen to highlight the advantages and disadvantages of the frequency domain model.

For systems depending on methods, a fundamental contradiction in the power delivery chain has existed between conventional to supply it. DC/DC conversion (e.g.) has therefore been an integral part of such systems to resolve this contradiction. be made tolerant to a much wider range of Vdd variance. This may open up opportunities for much more energy efficient methods of power delivery. performance of different power delivery mechanisms driving both asynchronous and synchronous loads directly from a harvester source bypassing bulky energy method, which employs a energy from a EH circuit depending on load and source conditions, is developed. through comprehensive comparative analysis. Based on the novel CBB power delivery method, an asynchronous controller is circuits to work with tasks. The successful asynchronous control design drives a case study that is meant to explore relations between power path and task path. To deal with different tasks with variable harvested power, systems may have a range of operation conditions and thus dynamically call for CBB or SCC type power set of capacitors to form CBB or SCC is implemented with economic system size. This work presents an unconventional way of designing a compact-size, quick- circuit overcome large voltage variation in EH systems and implement smart power management for harsh EH environment. The power delivery mechanisms (SCC) employed to help asynchronous- logic-based chip testing and micro-scale EH system demonstrations.

Throughout the years numerous studies for non-aqueous Li-air or Li-oxygen batteries have been investigated to elucidate their reactions and mechanisms. However, there have been only a few models developed for Li-air batteries. Therefore, the main objective of this work was to develop mathematical models for non-aqueous Li-air battery to increase understanding of the air cathode behaviour as well as predict the battery performance during cycling. A micro-macro homogeneous mathematical model was developed for a rechargeable Li-air battery using a concentrated binary electrolyte theory, and validated against experimental data. The dynamic behaviour of the porous cathode was determined by a numerical solution of the combined continuity, transport and kinetics equations. The microscopic behaviour included the local mass transfer between lithium peroxide (Li2O2) layer inside the cathode and active surface morphology changing with the Li2O2 solid precipitate growth. The model predicted that the capacity and discharge potential were sensitive to the solubility of oxygen and also the cathode porosity, the cathode structure and kinetic parameters. In addition, the charging behaviour was simulated by modelling the dissolution of solid Li2O2 product. The model suggested that the charging voltage can be decreased depending on capability of electrolyte to dissolve the Li2O2 discharge products. To improve the battery performance, the promising structure of a Li-air flow battery system with a electrolyte recycling unit continuously delivered the discharge capacity and provided high power density.

The Doubly-fed Induction Generator (DFIG) is the dominant technology for variable-speed wind power generation due in part to its cost-effective partially-rated power converter. However, the maintenance requirements and potential failure of brushes and slip rings is a significant disadvantage of DFIG. This has led to increased interest in brushless doubly-fed generators. In this thesis a Brushless Doubly-Fed Reluctance Generator (BDFRG) is compared with DFIG from a control performance point of view. To compare the performance of the two generators a flexible 7.5kW test facility has been constructed. Initially, a classical cascade vector controller is applied to both generators. This controller is based on the stator voltage field orientation method with an inner rotor (secondary stator) current control loop and an outer active and reactive power control loop. The dynamic and steady state performance of two generators are examined experimentally. The results confirm that the BDFRG has a slower dynamic response when compared to the DFIG due to the larger and variable inductance. Finally a sensorless Direct Power Control (DPC) scheme is applied to both the DFIG and BDFRG. The performance of this scheme is demonstrated with both simulation and experimental results.

The attenuation and distortion of surges by corona discharge on transmission lines is relevant to high-voltage insulation coordination. The predetermination of these aspects is desirable in terms of surge shape and conductor geometry , so that realistic situations can be accurately reproduced. In this thesis, an experimental investigation of the corona discharge on line conductors is described which, with the help of numerical models of corona and space charge, explains some aspects of the discharge under surge voltages. In simulating the corona characteristics, a numerical model is developed which relies on the physical properties of corona to give the non-linear variations with voltage of corona charge and corona capacitance, at present frequently used to simulate corona losses in surge attenuation calculations. The procedure results in a set of inter-related generalised equations whose solution requires the knowledge of the conductor geometry, the corona inception voltage and the geometrical capacitance of the system, together with the per-unit distributions of potential and electric field in the electrode configuration. These can be obtained numerically where analytical solution is not available. The model predictions are compared with both published test data and present laboratory measurements of charge and voltage on single and bundle conductors. New field filter probes incorporable in conductors of cylindrical cross-section are developed for measuring unipolar and bipolar field changes in impulse corona. The first type includes multiple-orifice filters for fields with radial symmetry, and filters with a single row of circular orifices for fields with no radial symmetry. The other type, which is of different design, is used with impulses of oscillatory shape, which give rise to bipolar variations in the surface field. Electric field measurements o n the surface of a twin-cond uctor assembly confirm the validity of the Deutsch-Popkov approximation and of the field boundary condition on the conductor used in the corona model. Charge and field measurements with oscillatory impulses show that the main corona charge is injected during the first voltage cycle, and that this charge is larger for impulses with higher frequency. On the second cycle, corona takes place at high overvoltages but is significantly weaker than on the first cycle. This has an important implication on the surge per formance of overhead lines. The importance of the surge steepness on corona inception voltage and corona injected charge is investigated by conducting experimental measurements of corona charge flow at various rates of voltage rise. The results show that minimum corona charge is obtained for impulse fronts intermediate between the standard lightning and switching shapes. This result is associated with the interdependence of corona charge and the statistical time lag of corona inception, and with the time required to clear the corona space charge. Time-lag effects are simulated by applying the concept of critical volume to present line conductor geometry.

No description available.

At present silicon wafer technologies dominate the market place with cost driven by the technological requirement for optically thick and electronically pure silicon. A solution to the high cost of wafer based panels is a thin film approach where micron thick layers of silicon replace the ~250 micron thick silicon wafers. Thin film silicon has gone to market in the form of amorphous and microcrystalline Si where performance is an issue as well as stability due to the hydrogenated amorphous Si structure. This project involves the growth of three dimensional wire structures based on crystalline silicon and their integration into solar cell devices. Nanowire solar cells featuring a radial junction have the potential to reduce the required volume and purity of the silicon used in cell fabrication compared to wafer based technologies. The vapour-liquid-solid effect (VLS) has been used to grow Si nanowires using Au and Sn as catalyst metals together with electron cyclotron resonance chemical vapour deposition (ECRCVD) for silicon deposition. Experimental results include the formation of seed particles from both gold and tin catalyst layers leading to the growth of silicon nanowires on both silicon wafer and glass substrates. The study of nanowire growth parameters from both gold and tin catalysts has led to the proposal of a model for growth in a low pressure environment as found in ECRCVD. Structural characterisation of the wires has taken place. Wires grown from gold catalyst layers on silicon are single crystal where the growth direction has a clear dependence on the orientation of the substrate. Those from tin exhibit a nanocrystalline shell over a single crystal core and have a tapered morphology. Furthermore, the growth direction is independent of substrate orientation. Reasons for this are discussed. Optical characterisation of the nanowire arrays has revealed high levels of broadband absorption. These results have been compared with finite difference time domain modelling (FDTD) and conclusions on the effects of wire density and catalyst layer thickness, along with that of the underlying silicon layer, have been drawn. The implication of these results for the design of solar cells based on these wires is discussed. Simple solar cell devices have been fabricated demonstrating photovoltaic response but with limited performance. The reasons for this are discussed with areas for improvement.

Polycrystalline silicon seed layer formation by aluminium-induced crystallisation (AIC) for solar cell applications was investigated. Precursor amorphous and microcrystalline silicon layers were deposited on SiO2 substrates using electron-cyclotron resonance plasma-enhanced chemical vapour deposition (ECR PECVD) and RF sputter deposition followed by thermal evaporation or RF sputter deposition of aluminium layers. Samples were then thermally annealed leading to layer exchange and crystallisation of the silicon layer. Selected samples were used as seed layers for ECR PECVD epitaxial thickening. Processing was usually done at temperatures less than 600°C compatible with glass substrates. Deposition parameters (substrate temperature, deposition time, gas flow rates, chamber pressure and magnetic currents) for ECR PECVD Si growth using SiH4 and H2 on bare and oxidised silicon wafers were varied to examine the effect on the crystallinity of deposited layers. This was related to properties of the plasma using diagnostic measurements. A method of adding argon to the chamber, found to be beneficial to material properties of deposited films, without causing the often observed reduction in growth rates was found. A study was then undertaken to optimise growth conditions of precursor silicon and aluminium layers for suitable seed layers by AIC. Using ECR PECVD deposited silicon and evaporated aluminium was found to lead to continuous layers after short annealing time, sometimes less than 15min, than using sputtered aluminium or silicon precursor layers which did not form continuous layers after several hours of annealing at 500°C. Using thin microcrystalline silicon was also found to lead to better film quality after annealing than thicker amorphous silicon films. Increasing the temperature during AIC from 250°C to 475°C was found to reduce the annealing time to typically 30 minutes at the highest temperature while allowing relatively large grains of several microns to form. Growth of a thick Si-Al interfacial oxide also lead to larger grain sizes. Under optimal preparation and annealing conditions, i.e. ECR PECVD μc-Si growth conditions and an annealing temperature of around 400°C, AIC produced continuous polycrystalline silicon seed layers with large grains up to 100μm, a preferential (001) orientation and a relatively smooth surface morphology with Ra values typically in the range of 20-30nm. Improvements in surface morphology and crystallinity were seen after subjecting AIC polycrystalline silicon to excimer laser irradiation at optimal fluence. Based on the above results and previous published work the process of AIC is discussed. ECR PECVD overgrowth on AIC seed layers was found to be epitaxial but the crystalline quality diminishes with thickness. Adding argon during deposition was found to be beneficial in producing thick, epitaxially grown layers with 2.5μm thick epitaxial layers achieved.