A novel assessment of unintentional islanding operations in distribution networks
This thesis aims to investigate an unexpected islanding operation (IO) which has been identified in a real distribution network. The process of recording and processing the data obtained from the field measurements in the distribution network (DN) has been the starting point of this research. It has...
| Autor: | |
|---|---|
| Tipo de recurso: | tesis doctoral |
| Estado: | Versión publicada |
| Fecha de publicación: | 2020 |
| País: | España |
| Institución: | CBUC, CESCA |
| Repositorio: | TDR. Tesis Doctorales en Red |
| OAI Identifier: | oai:www.tdx.cat:10803/669187 |
| Acceso en línea: | http://hdl.handle.net/10803/669187 https://dx.doi.org/10.5821/dissertation-2117-191259 |
| Access Level: | acceso abierto |
| Palabra clave: | Àrees temàtiques de la UPC::Enginyeria elèctrica 621.3 |
| Sumario: | This thesis aims to investigate an unexpected islanding operation (IO) which has been identified in a real distribution network. The process of recording and processing the data obtained from the field measurements in the distribution network (DN) has been the starting point of this research. It has to be underlined that this IO raised a problem and became a major challenge for the distribution operator. Therefore, the aim of this thesis is twofold; solving a real problem as well as further enhance the current research studies about IOs in DNs. IOs have been object of study during the recent years due to the rapid proliferation of the distributed generation (DG) within the so-called smart grids (SGs). Commonly, the power of these DG resources ranges between hundreds of kW and few MW and are allocated at either low voltage or medium voltage levels. One of the significant issues that these resources are raising is, undoubtedly, the IOs. These situations occur when a portion of the grid operates in parallel with the main grid following a disconnection. Thereby IOs, where the DG is energising the grid after a CB opening, must be identified and tripped in the minimum time possible. Failure to do so, the list of hurdles may include; power quality (PQ) disturbances (e.g., frequency and voltage out of range), a safety hazard for the network personnel or out-of-phase reclosings. That is the reason why the research towards the anti-islanding protection methods has elicited great interest. Fundamentally, the substantial improvement of this thesis lies in the fact that, in this IO, there are no DG resources, but large induction motors. In fact, the grid remains energised after the CB disconnection due to the induction motors (IMs) which transiently, act as generators. The island begins with the CB operation and ends when the CB recloses the circuit to restore the electrical supply. This rapid reclosing operation is widely adopted in DNs to avoid manual operations in self-extinguished faults and typically ranges between 0.5 and 1s. Given the fact that usually IOs are originated in the presence of DG, indeed, this IO is utterly unexpectedly for the DSO. Due to the phenomenon mentioned above, the specific goals of this thesis are described down below: 1. The first goal of this thesis focuses on developing a model suitable for validation purposes. To make a proper model validation, the simulations results obtained with this model will be compared with those obtained from field measurements. Thus, once the model has been validated, a thorough investigation regarding the most influential factors will be carried out. 2. The second goal of this thesis falls within the scope of the PQ. During the IO mentioned above, a new voltage sag topology is observed. Consequently, the efforts will be focused on modelling this new type of sag. 3. The third goal of this thesis emerges from the protective point of view. Once the IO has been defined and characterised, the need for identifying and preventing it becomes the main concern. In such a way, the third pillar of the thesis is targeted at implementing a suitable tool to prevent this particular IO. Besides, this new tool will be compared with the currently available methods for ID developed for scenarios with DG. |
|---|