Thesis (M.S.)--Virginia Polytechnic Institute and State University, 1992. / Vita. Abstract. Includes bibliographical references (leaves 87-90). Also available via the Internet.

Thesis (M.E.E.)--University of Delaware, 2008. / Principal faculty advisor: Allen Barnett, Dept. of Electrical & Computer Engineering. Includes bibliographical references.

Thesis (M.S.)--Humboldt State University, 2009. / Includes bibliographical references (leaves 69-70). Also available via Humboldt Digital Scholar.

Thesis (Ph.D.)--University of Delaware, 2006. / Principal faculty advisor: Mary E. Galvin-Donoghue, Dept. of Materials Science & Engineering. Includes bibliographical references.

Prepared under contract no. EX-76-A-01-2295-037.

The purpose of this investigation is to compare high frequency to low frequency (50Hz) technologies. To accomplish this, two highly optimized candidate converters are built to champion each of the inverter classes. Both candidates are constrained to produce the same quasi-sinusoidal output waveforms in identical operating conditions. The low frequency inverter is designed with optimization of its output power quality and accurate loss characterization taken in mind. A new iron core design procedure is proposed that may predict core losses when using non-sinusoidal excitation. Experimental results indicate that it has a promising degree of accuracy. The high frequency candidate is designed using planar integration technology. A topology selection determines that a critical-conduction mode flyback is the most suitable option. Components are added the topology to allow resonant switching for reduced switching losses. Common-mode current arising from good capacitive coupling between windings in the planar coupled inductor is reduced using charge balancing technology. The novelty of the converter lies is in the fact that all passive components, barring the bulk of the input capacitance and the output capacitance, are incorporated into the single planar structure. Finally, a comparison of the inverters is made in the categories of power quality, regulation, efficiency, robustness and applicability to the application. The high frequency inverter is found to promise substantial advantages over the low frequency variety in most of these elements. / Dr. I.W. Hofsajer

A review of the technology of useful conversion of wave power and tidal power is presented. These two power resources are reviewed separately, but on the same basis: principles of operation, existing devices or plants and research and development. Promising wave power devices in Britain, the United States and Europe are discussed. If wave power is to be competitive, one of the first requirements may be energy densification. Proposed energy densification schemes include resonance, high pressure water and wave focussing. Wave focussing is a Norwegian invention, technically feasible, and although more research and development is required, it appears to be more promising than alternative forms of wave power utilisation. According to a preliminary cost analysis, it could be competitive with conventional hydro-electric power. The large scale exploitation of tidal power has been considered seriously for about half a century; the literature on the topic is voluminous. The main limitations of tidal power are its intermittent nature and the high costs involved in the construction of a plant. The existing pilot plants at the Rance and Kislaya Guba have respectively proved that tidal power is technically feasible and that construction costs could be reduced. With the rapid increase in the price of fossil fuels, tidal power plants may be realised at the two best sites in the world, the Bay of Fundy and the Severn Estuary.

The ability of liberalised electricity markets to trigger investment in the generation capacity required to maintain an acceptable level of security of supply risk has been - and will continue to be - a topic of much debate. Like many capital intensive industries, generation investment suffers from long lead and construction times, lumpiness of capacity change and high uncertainty. As a result, the ‘boom-and-bust’ investment cycle phenomenon, characterised by overcapacity and low prices, followed by power shortages and high prices, is a prominent feature in the debate. Modelling the dynamics of generation investment in market environments can provide insights into the complexities involved and address the challenges of market design. Further, many governments who preside over liberalised energy markets are developing policies aimed at promoting investment in renewable generation. Of particular interest is the mix and amount of generation investment over time in response to policies promoting high penetrations of variable output renewable power such as wind. Consequently, improved methods to calculate expected output, costs and revenue of thermal generation subject to varying load and random independent thermal outages in a power system with a high wind penetration are needed. In this interdisciplinary project engineering tools are applied to an economic problem together with knowledge from numerous other disciplines. A dynamic simulation model of the aggregated Great Britain (GB) generation investment market has been developed. Investment is viewed as a negative feedback control mechanism with current and future energy prices acting as the feedback signal. Other disciplines called upon include the use of stochastic processes to address uncertainties such as future fuel prices, and economic theory to gain insights into investor behaviour. An ‘energy-only’ market setting is used where generation companies use a classical NPV approach together with the Value at Risk criterion for investment decisions. Market price mark-ups due to market power are also accounted for. The model’s ability to simulate the market trends witnessed in GB since early 2001 is scrutinised with encouraging findings reported. A reasonably good agreement of the model with reality, gives a degree of confidence in the realism of future projections. An advancement to the dynamic model to account for expected high wind penetrations is also included. Building on the initial model iteration, the short-term energy market is simulated using probabilistic production costing based on theMix of Normals distribution technique with a residual load calculation (load net of wind output). Wind speed measurement data is combined with the outputs of atmospheric models to assess the availability of the GB wind resource and its relationship with aggregate load. Simulation results for 2010-40 suggest that the GB system may experience increased generation adequacy risk during the mid to late the 2020s. In addition, many new investments are unable to recover their fixed costs. This triggered an investigation into the design of a capacity mechanism within the context of the modelling environment. In light of the ongoing GB market electricity market reform debate, two mechanisms are tested; a strategic reserve tender and a marketwide capacity market. The goal of these mechanisms is to mitigate generation adequacy risk concerns by achieving a target winter peak de-rated capacity margin.

A new Markov chain based energy storage model to evaluate the power supply availability of microgrids with renewable energy generation for critical loads is proposed. Since critical loads require above-average availability to ensure reliable operation during extreme events, e.g., natural disasters, using renewable energy generation has been considered to diversify sources. However, the low availability and high variability of renewable energy sources bring a challenge in achieving the required availability for critical loads. Hence, adding energy storage systems to renewable energy generation becomes vital for ensuring the generation of enough power during natural disasters. Although adding energy storage systems would instantaneously increase power supply availability, there is another critical aspect that should be carefully considered; energy storage sizing to meet certain availability must be taken into account in order to avoid oversizing or undersizing capacity, which are two undesirable conditions leading to inadequate availability or increased system cost, respectively. This dissertation proposes to develop a power supply availability framework for renewable energy generation in a given location and to suggest the optimal size of energy storage for the required availability to power critical loads. In particular, a new Markov chain based energy storage model is presented in order to model energy states in energy storage system, which provides an understanding of the nature of charge and discharge rates for energy storage that affect the system's power output. Practical applications of the model are exemplified using electrical vehicles with photovoltaic roofs. Moreover, the minimal cut sets method is used to analyze the effects of microgrid architectures on availability characteristics of the microgrid power supply in the presence of renewable energy sources and energy storage. In addition, design considerations for energy storage power electronics interfaces and a comparison of various energy storage methods are also presented. / text

published_or_final_version / Chemistry / Doctoral / Doctor of Philosophy