Filters
Availability
Publication Year
Resource Type
Faculty
School/Institute
Publication Search Results
Sort
Results per page
1 - 3 of 3
-
(2022)ThesisThe integration of variable distributed energy resources and vehicle electrification has come to focus over the last few years. While much work has been done to address the challenges that arise in modern distribution system planning and operation, continuous improvement to the models with the change is essential. The objective of this thesis is to improve the distribution network planning and operation models in the presence of distributed generation and electric vehicles. It aims to build stochastic models including the power generation and the charging demand, determine the location and sizing of the energy resources and charging stations in the coupled systems, and evaluate the impacts of the new low-carbon technologies on the network. Using a mixed-integer nonlinear programming framework through an optimal power flow analysis, this thesis presents three major methodological contributions including uncertainty modelling, coordinated mathematical formulation, and conflicting objective solutions. First, a multivariate stochastic process based on the notion of copula is applied to derive probabilistic charging patterns and to obtain the stochastic charging profiles. Second, a two-stage stochastic program based on statistical analysis and numerical simulation is introduced to generate synthetic time series of solar and wind power generation. The continuous distributions are discretized to generate the scenarios and the number of scenarios is reduced using Kantorovich metrics. Third, a two-dimensional Pareto front of dominant solutions is given for the competing objectives using a multiobjective Tchebycheff decomposition-based evolutionary algorithm. Case studies are conducted to evaluate the effectiveness of the proposed methods. An optimal charging scheduling problem is formulated to assess the stochastic charging models. The problem is formulated as a conic quadratic optimal power flow model and solved with a convex optimization algorithm. Network expansion planning problems are presented with carsharing and non-carsharing models, as well as the distributed energy generations. Overall, these problems aim to minimize the planning and operational cost of feeder routing, and substation alterations while maximizing the utilization of charging stations. It is found that an accurate estimation of the randomness intrinsic to the network is critical to ensure the secure and economic operation and planning of the distribution system intertwined with the transport network.
-
(2022)ThesisThe level of electricity consumption usually can reflect the level of development of a society. With the development of social industrial technology, the demand for electricity and other energy sources is increasing. Meanwhile, the control measures for carbon emissions have become an important social issue that needs to be solved urgently cause a series of problems such as environmental pollution has become more and more serious with the large-scale use of traditional fossil energy. Renewable energy provides new ideas for solving this issue. Recently, some renewable energy sources, including solar and wind energy, have been widely used. However, as renewable energy has some shortages such as instability, inflexibility, it can not replace traditional fossil energy completely. Due to consumers usually having different energy demands, the concept of an integrated energy system including traditional fossil energy and renewable energy has been proposed. The efficiency of energy usage can be improved, and energy losses can be reduced through coordinated planning and optimal operation method. Thus, integrated energy systems become an effective method to solve the current social issues. With the increasing penetration of renewable energy injection in the electric power grid, the impact of uncertainty and fluctuation in renewable energy generation on the grid performance cannot be ignored. Integrated energy networks with the Power-to-Gas (PtG) system have been used to maintain the stability of the system by employing the efficient conversion of different energy forms (electricity and gas) and massive energy storage. Meanwhile, the wave energy converter system can make full use of wave energy to improve the use of renewable energy in the integrated energy system. An economic receding horizon control strategy that solves the constrained optimization problem of the wave energy converter system is analyzed in this research. The Model Predictive Control (MPC) method is adopted to focus on tracking system cost function, and the system state is steered to the steady-state at the end of the optimization horizon by using terminal equality constraint, and the non-convex optimization problem is solved by the controller in real-time. As a form of a hybrid multi-energy system, the integrated energy system usually contains different forms of energy such as electric power, thermal, and gas while meeting the load of various energy forms. Therefore, the model building and optimal operation of the integrated energy system are the key points of the research. The appropriate models established for the integrated energy system can not only reduce the calculation and power dispatch times but can help the entire system reach the optimal state faster. The main work of this thesis is as follows： In chapter 2, various energy conversion systems, including power-to-gas systems and combined heat and power (CHP) systems have been analyzed. A method for optimal operation of a microgrid that contains the PtG system with rolling horizon strategy is proposed, and it applied the power and gas cooperative accommodation to solve the deviation between the forecast load and real load as well as reduce the overall microgrid operation cost in a corrective manner. Models of various components in a microgrid are introduced, and a mixed-integer linear program (MILP) mathematical optimization problem is formulated to describe the day-ahead operation problem. The rolling horizon strategy is adopted to reduce the impact of intermittence natural of the renewable energy sources, and the benefits of cooperative accommodation of electric power and gas in a corrective manner are discussed as well. In chapters 3 and 4, the different energy models in the integrated energy system have been analyzed. The dist-flow method is applied to quickly calculate the power flow, and the gas system model is built by the analogy of the power system model. Meanwhile, the piecewise linearization method is applied to solve the quadratic Weymouth gas flow equation, and the Alternating Direction Method of Multipliers(ADMM) method is applied to narrow the optimal results of each subsystem at the coupling point. Simulation-based calculations and comparison of results under different scenarios are proved that the power-thermal-gas integrated energy system enhances the flexibility and stability of the system as well as reduces system operating costs to some extent. In chapter 5, an economic receding horizon control strategy that solves the constrained optimization problem of the wave energy converter system is proposed. Generally, the standard Model Predictive Control focuses on tracking system cost function. However, in economic MPC, a general economic cost function is established, which contains the objective of maximizing the energy extracted from ocean waves and minimizing the operation cost, as well as directly reflecting the economic indicators of the Wave Energy Converter (WEC). Meanwhile, the system state is steered to the steady-state at the end of the optimization horizon by using terminal equality constraint, and the non-convex optimization problem is solved by the controller in real-time. The auxiliary optimization problem is applied to the stability analysis, while the convergence of the system can be proved by the Lyapunov technique. And several numerical simulation results are presented to demonstrate the effectiveness of the proposed algorithm.
-
(2022)ThesisGlobally, energy systems are expected to undergo a complete transition from fossil- fuelled generation to renewable energy in the coming decades, with a majority of energy supplied by wind and solar in many countries. In much of the developed world, this transition will take place in the context of restructured electricity markets. This thesis examines whether electricity markets, which are intended to be the key drivers of electricity industry operation and investment, are suitably designed and implemented for transitioning to high penetrations of renewable energy. Of particular interest is the role of competition in delivering efficient market outcomes, the potential for exertion of market power in high-penetration renewable energy scenarios, and whether current auction designs to incentivise efficient behaviour will be effective in the context of energy delivered at near-zero marginal cost. Previous work on electricity market competition in Australia has focused on measuring market concentration, a commonly used indicator of competitiveness, on short-term time horizons, based on historical data. However, competitiveness in Australia’s National Electricity Market (NEM) in the long term has not been assessed, nor how it might change as a result of the transition to high penetrations of variable renewable energy (VRE). This may be due in part to lack of suitable measures of competition in markets with multiple interconnected regions, but also the theory and evidence around VRE bidding patterns now and into the future has not yet been confirmed. Assessing competitiveness of future markets requires new methods for modelling and assessing potential market dynamics that affect market power. While capacity expansion modelling has been used for understanding the future technical and economic performance of electricity systems with different generation technologies, there have been very few attempts to relate these models back to the concepts of competition and market concentration. Machine learning techniques may also have the potential to provide new insights into the strategic behaviour of participants in future energy systems and have been used for modelling and solving many other complex multi-agent interactions, but to date a straightforward method for applying modern machine learning techniques to models of competitive electricity markets has not been proposed. Furthermore, significant changes that are under consideration to facilitate the energy transition, such as the introduction of a new two-sided market design in the NEM that would require all demand-side participants to submit bids, have not been considered in modelling to date. This thesis aims to investigate competition and market power in restructured electricity markets as well as their role in the clean energy transition. It investigates whether the Australian NEM has been and will continue to be a competitive market through the transition to renewable energy and how renewable generators participate in electricity auctions now and into the future. Additionally, it examines the way new tools and frameworks might further understandings of incentives and behaviour to enable more efficient and stable market designs. In order to establish a theoretical base and explore what causes market mechanism failure, a literature review and case study are undertaken into episodes of the exercise of market power globally, with a specific focus on the Californian electricity crisis. To establish how well market mechanisms are currently working, a range of competition metrics are applied to historical datasets in order to study the level of competitiveness of the Australian National Electricity Market. This leads to new answers to the question of whether the NEM is currently a competitive market, showing that current market concentration indicators provide conflicting results depending on how they are applied. A new measure of competition is provided which demonstrates that most regions are generally competitive, but some, such as Queensland, have notable periods of constraint. In order to determine how the transition to renewables might impact competition in the NEM, new indicators of competitiveness are also applied to simulations of future high-penetration renewable energy scenarios. These analyses demonstrate that swings between surplus and constraint can lead to an increase in the frequency of opportunities to exercise market power. This is an important result that shows how high-penetration renewables may significantly disrupt the function of wholesale electricity spot markets. To understand both the underlying incentives acting on renewable generators in the NEM and the current bidding strategies of these generators a case study of these generators in the NEM is undertaken. It is seen that these participants generally offer energy at or below $0/MWh, but are occasionally seen to bid at very high prices, possibly in an attempt to push up the spot price. Following this analysis, in order to examine what strategic incentives might be present in future high-penetration renewable energy grids, new equilibria for near-zero marginal cost generators are proposed. Following on from these investigations, the performance of a two-sided market in a 99% renewable energy grid is explored. In a two-sided market, flexible demand-side participants would be required to enter bids into the wholesale market. Based on forecasts of flexible demand response and renewable energy performance in a 99% renewable energy scenario, this modelling showed that demand response was, counterintuitively, less likely to be present in a two-sided market; additionally, the two-sided market was seen to mitigate the impacts of the exercise of market power because the more elastic supply curve placed upper limits on strategic generator offers. In order to develop a new modelling framework for renewable bidding behaviour in recognition of the difficulties in modelling competitive equilibria for future high- penetration renewable electricity market conditions, a market simulator is developed for the OpenAI platform that can be used to train deep learning models of electricity market bidding. Such models may be extremely useful in the context of the transition to high-penetration renewables, because competitive dynamics could be accurately predicted and understood before new capacity is built and operated. There are several key contributions of this work; it presents a new method for calculating and estimating levels of competition in electricity markets such as the NEM, which are comprised of multiple regions with constrained interconnectors, provides and applies a new methodology for exploring thresholds of competitiveness in simulations of future energy systems, develops the first long-term exploration of renewable bidding behaviour in Australia’s NEM, gives a new tool for running market behaviour experiments with emerging AI tools, and provides an early analysis of the impact of implementing a two-sided market mechanism, as proposed by Australia’s Energy Security Board. Together, these contributions may help to significantly enhance current understandings of the opportunities and challenges associated with transitioning to high-penetration renewable energy within a wholesale electricity market.