M.Phil. / The construction of mathematical models for making projections of future energy demand is an important step in the formulation of an energy policy for a country or region. In this dissertation the different approaches to energy modelling are investigated. The history of the modelling effort in the world and South Africa is briefly discussed. The development of a comprehensive database is of vital importance for the success of any modelling effort. An energy database for South Africa was developed. An existing energy forecasting model was updated and extended to cover eleven final consuming sectors. Projections of energy demand were made to illustrate the mechanism of the model.

The suitability of Compressed Air Energy Storage (CAES) as a source of peaking plant capacity in South Africa is examined in this research report. The report examines the current state of CAES technology including examples of operational and planned facilities. It further evaluates the potential challenges and benefits of the use of CAES in South Africa. A high level proposal for plant design capacity is documented, and potential costs for construction thereof are estimated. The cost of a CAES plant is compared to generating options using the Levelised Cost of Energy (LCOE) method. The study proposes that by 2018 additional peaking plant capacity will be required and that a CAES plant able to provide additional capacity up to 3 500MW would help to alleviate the potential shortfall which may be experienced at this time. The report further proposes conversion of underground mines for use as air receivers for high pressure storage of large volumes of compressed air required for CAES. The research report concludes that CAES presents a feasible solution to the potential future shortfall in peaking plant capacity in South Africa, and that site identification and construction of a suitable storage cavern presents the main obstacle to the implementation of this technology.

Digitized by Kansas Correctional Industries

This dissertation studies the science and policy-making of alternative energy using biofuels as a case study, primarily examining the instruments that can be used to alleviate the impacts of climate change and their relative efficacy. Three case studies of policy-making on biofuels in the European Union, United States of America and Brazil are presented and discussed. It is found that these policies have had large unintended negative consequences and that they relied on Lifecycle Analysis studies that had concluded that increased biofuels production can help meet economic, energy and environmental goals. A close examination of these Lifecycle Analysis studies reveals that their results are not conclusive. Instead of continuing to attempt to find answers from Lifecycle Analyses, this study suggests an alternative approach: formulating policy based on recognition of the ignorance of real fuel costs and pollution. Policies to combat climate change are classified into two distinct approaches: policies that place controls on the fuels responsible for emissions and policies that target the pollutants themselves. A mathematical model is constructed to compare these two approaches and address the central question of this study: In light of an ignorance of the cost and pollution impacts of different fuels, are policies targeting the pollutants themselves preferable to policies targeting the fuels? It is concluded that in situations where the cost and pollution functions of a fuel are unknown, subsidies, mandates and caps on the fuel might result in increased or decreased greenhouse gas emissions; on the other hand, a tax or cap on carbon dioxide results in the largest decrease possible of greenhouse gas emissions. Further, controls on greenhouse gases are shown to provide incentives for the development and advancement of cleaner alternative energy options, whereas controls on the fuels are shown to provide equal incentives to the development of cleaner and dirtier alternative fuels. This asymmetry in outcomes--regardless of actual cost functions--is the reason why controls on greenhouse gases are deemed favorable to direct fuel subsidies and mandates.

Reducing the energy consumed in the built environment is a key objective in many sustainability initiatives. Existing energy saving methods have consisted of physical interventions to buildings and/or behavioral modifications of occupants. However, such methods may not only suffer from their own disadvantages, e.g. high cost and transient effect, but also lose aggregate energy saving potential due to the oftentimes-associated single-building-focused view and an isolated examination of occupant behaviors. This dissertation attempts to overcome the limitations of traditional energy saving research and practical approaches, and enhance residential building energy efficiency and sustainability by proposing innovative energy strategies from a holistic perspective of the aggregate human-environment systems. This holistic perspective features: (1) viewing buildings as mutual influences in the built environment, (2) leveraging both the individual and contextualized social aspects of occupant behaviors, and (3) incorporating interactions between the built environment and human behaviors. First, I integrate three interlinked components: buildings, residents, and the surrounding neighborhood, and quantify the potential energy savings to be gained from renovating buildings at the inter-building level and leveraging neighborhood-contextualized occupant social networks. Following the confirmation of both the inter-building effect among buildings and occupants' interpersonal influence on energy conservation, I extend the research further by examining the synergy that may exist at the intersection between these "engineered" building networks and "social" peer networks, focusing specifically on the additional energy saving potential that could result from interactions between the two components. Finally, I seek to reach an alignment of the human and building environment subsystems by matching the thermostat preferences of each household with the thermal conditions within their apartment, and develop the Energy Saving Alignment Strategy to be considered in public housing assignment policy. This strategy and the inter-building level energy management strategies developed in my preceding research possess large-scale cost-effectiveness and may engender long-lasting influence compared with existing energy saving approaches. Building from the holistic framework of coupled human-environment systems, the findings of this research will advance knowledge of energy efficiency in the built environment and lead to the development of novel strategies to conserve energy in residential buildings.

A major challenge in integrating high penetrations (>20%) of solar- and wind-energy rests in the grid's ability to cope with the intrinsic variability of these renewable resources. Although such high levels of penetration may be a decade or two away in most operating regions, we must find measures to manage the variability of these sources, especially when conventional market-based approaches are exhausted or ineffective. Furthermore, besides assuring reliability, effective integration of high levels of solar- and wind-power can reduce the 'hidden' environmental costs and emissions associated with larger than necessary backup capacity. With large-scale PV plants (>250 MW) becoming significant generators on the grid in the near future, system operators became concerned about the plants' inherent variability, and questions were raised regarding the predictability and reliability of the output from such PV plants. In the first part of this research, the variability in the power output of six PV plants in the United States and Canada, with a total installed capacity of 195 MW (AC), is characterized. A new metric called the Daily Aggregate Ramp (DAR) is introduced to quantify, categorize, and compare daily variability across multiple sites. With this metric, and by harmonizing for climatic differences across the plants, we quantified the effect of geographic dispersion in reducing the cloud-induced power fluctuations. In addition, the reduction in variability was assessed by simulating a step by step increase of the plant size at the same location, using individual inverter data. Our data analysis showed maximum ramp rates 0.7, 0.58, 0.53, and 0.43 times the plant's capacity for 5, 21, 48, and 80 MW (AC) plants, respectively. After the variability in plant outputs was understood and quantified, we investigated algorithms for operating Energy Storage Units (ESU) to perform ramp rate control at the plant level. This task is designed to support proposed plans of grid balancing authorities to deal with ramps of variable energy resources (i.e., solar and wind). ESUs can be used to mitigate penalty fees caused by sharp ramps and perhaps allow for additional revenue streams by participating in grid balancing markets (e.g. frequency regulation). Consequently, we focused on building and optimizing ESU dispatch models for controlling ramp rates of individual PV plants within predetermined levels. The model comprised dispatch strategies tailored to specific fast response ESU technologies (e.g., flywheels, capacitors, batteries). The optimization involved trial and error testing of different combinations of ESU technologies, power and energy capacities, dispatch strategies and violation reduction requirements. For four PV plants (5, 21, 30.24 and 80 MW) in various North American locations, we found a required ESU power capacity of 2.2, 9, 12 and 22 MW respectively, to mitigate 99% of the violations of a 10%/minute ramp rate limit. These ESU capacities may add capital costs of about $0.35-0.63 per Watt PV for the 80 MW plant and $0.56-0.94 per Watt PV for the 5 MW plant. Lowering the mitigation requirement to 90% reduces the necessary ESU power capacity (and per Watt PV costs) to 1.1 MW ($0.27), 4.4 MW ($0.27), 6.4 MW ($0.27) and 10.8 MW ($0.18), respectively. Curtailment of power at the inverter during upward ramps reduces the number of violations even further and effectively decreases the necessary ESU capacity to approximately: 0.8, 3.1, 4.5, and 7.6 MW (for the 90% violation mitigation). It is noted that the reported ESU capacity additions and associated costs are based on the assumption of no forecasting or only a one-minute ahead forecasting of cloud-induced solar variability. If forward time forecasting is available, the optimization we developed should result in lower ESU capacity requirements as gradual ramp rate controls could be implemented in advance. Another way to reduce the costs associated with ramp-rate controls is to use the ESU for other revenue-generating activities, such as frequency regulation for which markets exist in different operating regions (e.g. the Real-Time Market of the New York Independent System Operator (NYISO)). Since ramp rate violations in the various facilities we studied, occurred in less than 2% of the time during the year, such additional uses of ESUs are possible.

Current electricity infrastructure is being stressed from several directions -- high demand, unreliable supply, extreme weather conditions, accidents, among others. Infrastructure planners have, traditionally, focused on only the cost of the system; today, resilience and sustainability are increasingly becoming more important. In this dissertation, we develop computational tools for efficiently managing electricity resources to help create a more reliable and sustainable electrical grid. The tools we present in this work will help electric utilities coordinate demand to allow the smooth and large scale integration of renewable sources of energy into traditional grids, as well as provide infrastructure planners and operators in developing countries a framework for making informed planning and control decisions in the presence of uncertainty. Demand-side management is considered as the most viable solution for maintaining grid stability as generation from intermittent renewable sources increases. Demand-side management, particularly demand response (DR) programs that attempt to alter the energy consumption of customers either by using price-based incentives or up-front power interruption contracts, is more cost-effective and sustainable in addressing short-term supply-demand imbalances when compared with the alternative that involves increasing fossil fuel-based fast spinning reserves. An essential step in compensating participating customers and benchmarking the effectiveness of DR programs is to be able to independently detect the load reduction from observed meter data. Electric utilities implementing automated DR programs through direct load control switches are also interested in detecting the reduction in demand to efficiently pinpoint non-functioning devices to reduce maintenance costs. We develop sparse optimization methods for detecting a small change in the demand for electricity of a customer in response to a price change or signal from the utility, dynamic learning methods for scheduling the maintenance of direct load control switches whose operating state is not directly observable and can only be inferred from the metered electricity consumption, and machine learning methods for accurately forecasting the load of hundreds of thousands of residential, commercial and industrial customers. These algorithms have been implemented in the software system provided by AutoGrid, Inc., and this system has helped several utilities in the Pacific Northwest, Oklahoma, California and Texas, provide more reliable power to their customers at significantly reduced prices. Providing power to widely spread out communities in developing countries using the conventional power grid is not economically feasible. The most attractive alternative source of affordable energy for these communities is solar micro-grids. We discuss risk-aware robust methods to optimally size and operate solar micro-grids in the presence of uncertain demand and uncertain renewable generation. These algorithms help system operators to increase their revenue while making their systems more resilient to inclement weather conditions.

The temporal mismatches in the varying demand and supply pose a major challenge for today’s U.S. electricity grid. Demand response (DR), aiming at reducing demand on the grid during times of electricity generation capacity shortage and very high wholesale prices, is one of many approaches to address this challenge. Unlike the sophisticated automatic controls to operate appliances (such as lights and air-conditioning) on shifted or reduced schedules, which are more common in the commercial sector, the proposed DR scheme discharges storage when demand on the grid is high so as to enable DR without affecting actual appliance usage. As small-scale storage technologies and residential demand response tariffs (e.g., time-of-use tariffs, which charge in differing rates for peak times and off-peak times) become more available, distributed energy storage for the residential sector DR is now technically ready and has the opportunity to generate financial incentives for residential consumers. However, such storage-based DR is still largely underutilized in the residential sector, partly due to consumers’ concerns about cost-effectiveness of storage. Thus, these concerns call for a comprehensive economic analysis to answer the following two questions: 1) Could storage yield actual profits (i.e., electricity cost reduction via arbitrage minus levelized storage cost) for residential consumers? And 2) Which particular combination of storage technology and tariff yields the highest profit? In addition, from the perspective of the grid, a third question is yet to be answered: If a large portion of households were to apply economically optimized storage-based DR systems, what would be the implications and emission impacts (i.e., CO₂, NOₓ, and SO₂ emissions) for the grid? To address the above questions, I 1) develop a levelized storage cost model, based on the simulated storage lifetime — a hybrid of the total-energy-throughput lifetime and the calendar lifetime. Storage technologies included in this dissertation are conventional and flow batteries, flywheel, magnetic storage, pumped hydro, compressed air, and capacitor; 2) devise an agent-based, appliance-level demand model to simulate demand profiles for an average household in the U.S.; 3) dispatch storage via loadshifting (to time-shift energy requirements from peak times to off-peak times) and peak shaving (to reduce peak power, i.e., kW, demands and smooth demand profiles) strategies, under realistic tariffs (Con Edison, New York); and 4) optimize the storage capacity (in kWh) and the demand limit on the grid (in kW; above which the strategy will attempt to use stored electricity in addition to grid electricity to satisfy appliance demand; used for the peak shaving strategy only) and determine the potential profits (or losses). I find that: 1) For economically viable technologies, annual profits range from <1% to 28% of the household’s non-DR electricity bill by utilizing the loadshifting strategy and from <1% to 37% by implementing the peak shaving strategy, depending on the storage technologies; 2) Of the two DR strategies, the peak shaving strategy can render more storage technologies economically viable. To evaluate the potential implications for the New York state grid, the electricity consumption features of households in New York state are then fed into the demand model. A dispatch curve is then developed, based on the marginal generation cost of each power plant, to simulate the dispatch order of the available power plants in New York state. The potential implications and emission impacts are investigated by comparing the statewide demand profiles as well as generation emissions with and without residential sector storage and DR. I find that: 1) Although yielding substantial financial incentives for households, the peak shaving strategy only leads to minor impact on the grid (assuming 15% household participation rate); 2) The loadshifting strategy would cause extra grid stress, and likely lead to brownouts, when all participating households start to re-charge their storage by purchasing inexpensive electricity uncoordinatedly; and 3) The overall emission impacts for both strategies are less than 5% of the total non-DR emissions in the state of New York.

Surfaces and interfaces are a ubiquitous part of nature. They influence the behavior of devices and are essential components in charge transfer and charge trapping. While surfaces and interfaces are important studying them is difficult because they consist of only the first few layers of a material. Therefore, surface-specific techniques are needed to investigate their properties and dynamics. Perhaps the most common surface electronic surface characterization techniques are electron spectroscopies which have become the standard for determining surface electronic band structure. However, these spectroscopies require ultra high vacuum which precludes the study of surfaces at ambient pressures and buried interfaces. Ambient pressures and interfaces are precisely the conditions under which most devices operate. Therefore there is a need for a technique which can reveal information about electronic states and their dynamics of buried interfaces at ambientconditions. This thesis describes the implementation of broadband time-resolved second harmonic generation and the recovery of the time-resolved amplitude and phase by employing spectral interferometry. The even order nonlinear process allows the measurement to be surface specific which the spectral amplitude and phase reveal information about surface state transitions and couplings. The first chapter motivates the study of surface and interfaces while chapters 2 and 3 cover background information about surfaces and nonlinear optics to help understand the experiments presented in the following two chapters. Chapter 4 presents a broadband time resolved spectral SHG technique whose usefulness is demonstrated on gallium phosphide passivated undoped gallium arsenide. In this case the spec-tral features are due to the E 1 resonance in GaAs and the dynamics are assigned to band gap renormalization. Chapter 5 details a method to recover the time resolved amplitude and phase and then demonstrates the recovery of the amplitude and phase from SH emitted from n- and p-type GaAs. The spectra reveal a discreet surface state ascribed to defect formation specific to n-type GaAs. The asymmetric line shape of this state indicates that it is coupled to a continuum; most likely a surface projected bulk band. We found that this coupling can be controlled by changing the azimuthal angle. However, p-type GaAs does not show distinct features in the second harmonic spectrum. Experiments on bilayers consisting of p-type GaAs and copper pthalocyanine (CuPc) are also presented in chapter 5. No changes in the signal are observed for either the constituents alone. However, when CuPc is deposited on GaAs a transient state forms at 200 fs delay between the pump delay which also exhibits an asymmetric line shape. This indicates the formation of a new state at the heterojunction that was not present before and may be evidence for a charge transfer state. Chapter 6 closes the thesis with concluding remarks which suggest improvements in the experimental design and implementation of time-resolved second harmonic spectral interferometry.

Thesis (Ph. D.)--University of Adelaide, 1988. / Includes bibliographical references (leaves 240-248).