In today's societies, electricity is considered essential for welfare and economic growth. Due to political and climatological reasons as well as the depletion of fossil fuels, there is a tendency to use more renewable energy sources. For electricity delivery systems, this means that ever more renewable generators are to be integrated, which behave different from conventional generators. Unlike for many other products, electricity is a rather peculiar product since (in the absence of large amounts of storage) it must be consumed almost simultaneously with its production. Therefore, in any electrical power system, always a continuous balance between generation and load of electricity must be maintained to ensure a reliable operation. Any imbalance between the two will, in a synchronous power system, initially be compensated by a change in the kinetic energy of motors and generators. As a result, the grid frequency will start to deviate from its nominal value. To guarantee stability, the network frequency of a power system must always remain within predefined limits. To enable this, different categories of balancing resources can be defined which all must be available within a specified time. Four causes of imbalance can be distinguished. First of all, there can be events, such as failures of generation, load or part of the network. Secondly there can be inaccuracies in the control of generation. Thirdly, there can be noise in the generation or load. Finally there can be forecasting errors for generation or load. With the introduction of more Distributed Generators based on Renewable Energy Sources (DG-RES), the imbalances due to the fourth reason will increase. Maintaining the balance between generation and load, and therefore ensuring the availability of sufficient balancing capacity is initially a task of the Transmission System Operators (TSO) or the Independent System Operators (ISO) of parts of the power system in close cooperation with each other. In many cases the TSOs acquire balancing resources via market mechanisms, thus forwarding part of their responsibility to Balance Responsible Parties (BRP). This is supposed to lead to the optimization of the utilization of resources in the power system. This thesis elaborates on the different balancing mechanisms by analyzing both the technical and market-related aspects. It is shown that existing regulatory aspects do influence the technical performance of the imbalance settlement system, as current regulation enforces and creates incentives for market parties to balance in terms of energy, while power balancing would ideally be desired. Also it is elaborated what stimuli exist for the expansion or shrinkage of control areas for balancing and how this influences the needs for, and availability of balancing capacity. It is shown that balancing in larger areas leads to relatively lower demand for balancing resources. As the needs for balancing resources are influenced by fluctuations in power generation, which are expected to increase in future power systems due to the increased penetration of renewable generation, a method for analyzing these power fluctuations is proposed. This method is based on the analysis of power fluctuations in the frequency domain and allows for the creation of unambiguous characterizations of both provision and demand of balancing resources which are a precondition for the creation of efficient markets to trade these balancing resources. A number of changes, which lead to changing needs of balancing resources, are assumed to take place in future power systems. These are, but are not limited to, the further restructuring of the electricity market, the further integration of DG-RES and the expansion of the power system with either synchronous or DC interconnections. With a simulation model of a synchronously interconnected electrical power system, the impacts of both technical and regulatory aspects on balancing needs are investigated in both time and frequency domains. By modeling the trading and generation behavior of BRPs, which are constrained by regulation, the occurrence of repetitive imbalances is studied as well. These results have been verified using grid frequency measurements. It is found that repetitive imbalances occur due to the optimization by BRPs while complying with the existing regulation. In future power systems, improved operation and deployment of a generation portfolio can lead to a reduction of imbalances for both BRPs and complete control areas. Two cases, being the application of predictive real-time economic dispatch and the provision of ancillary services for balancing by renewable generators, are discussed in detail. In both cases the generation by renewable sources is more actively being taken into account, either via incorporating predictions of the generation, or via direct control of the renewable generators. It is found that considerable advantages can be reached for both BRPs and TSOs. The main conclusions from the research described in this thesis are that sufficient incentives exist to either or both actively and passively contribute to the imbalance settlement system in the Netherlands. Prices of balancing energy are considered profitable for BRPs to bid part of their generation capacity to the imbalance settlement system. Also it is found that the option of balancing on an international level leads to a reduction of the balancing needs. The integration of DG-RES will demand new methods and procedures to evaluate balancing needs. It was found that these balancing needs can be determined in the frequency domain and by applying wavelets quite effective. Some future technological changes, as the integration of DG-RES, the expansion of the grid, and the use of power electronics, will affect the functioning of the imbalance settlement system and therefore they have an impact on the stability of the power system. Also regulation in a liberalized market is found to influence the behavior of BRPs and the effect of this was observed in the grid frequency. Lastly it is demonstrated that by incorporating predictions of DG-RES, future imbalances and future generation schedules, the economic dispatch can be improved, leading to fewer program deviations for BRPs and simultaneously reducing system imbalances.
|Qualification||Doctor of Philosophy|
|Award date||1 Jun 2011|
|Place of Publication||Eindhoven|
|Publication status||Published - 2011|