Year

2015

Degree Name

Doctor of Philosophy

Department

School of Electrical, Computer and Telecommunications Engineering

Abstract

Increasing levels of penetration of distributed energy resources (DERs) have trans- formed distribution networks from passive to active networks and introduced the concept of microgrids. Dynamic characteristics of microgrids operating either in grid connected or islanded modes can be different from the traditional distribution networks due to the combination of different DERs. In order to make microgrid operation attractive, the issues associated with microgrids need to be properly anal- ysed. This thesis examines the modelling of microgrids and investigates different aspects of their operation.

In the first phase of the work presented in this thesis, dynamic characteristics of microgrids comprising different distributed generators are investigated. The importance of understanding the dynamic behaviour of microgrids is highlighted through a comparative analysis carried out on a hybrid microgrid. A simulation model of a hybrid microgrid comprising a PV system, a doubly-fed induction generator (DFIG) based wind power plant, a mini hydro power plant, and loads is developed for the analysis. This study revealed that the dynamic characteristics of the microgrid are significantly influenced by the characteristics of individual DERs and their control systems. It has been noted that during grid connected mode, features of the external grid also have an impact on microgrid behaviour.

The second phase of this thesis is focused on aggregated modelling of grid connected microgrids comprising both inverter interfaced and non-inverter interfaced DERs. For stability analysis, the common practice is to separate the power sys- tem into a study area of interest and external areas. In general, the study area is represented in a detailed manner while external areas are represented by dynamic equivalents. This thesis investigates the applicability of modal analysis as a tool for dynamic model equivalencing of grid connected hybrid microgrids while introducing a new index to identify the dominant modes of the system. The grid connected microgrid is represented as a single dynamic device while retaining the important dynamics. Linearised models of different DERs with control systems and loads are developed for this study. Several case studies are carried out to validate the reduced order dynamic model of the microgrid by testing under different operating conditions. Furthermore, the model equivalencing is applied on microgrids in a multi-microgrid environment to validate the methodology.

Similar to the large generators in conventional power systems, grid connected microgrids have the potential to participate in energy markets to achieve technical, financial and environmental benefits. In order to enable such operation, a systematic approach in developing a capability tool for a grid connected microgrid is presented in the next phase of this thesis. A grid connected microgrid can be viewed as a single generator or a load depending on power import or export at the grid supply point. However, unlike in a single generator with simple machine limitations, active and reactive power transfer limits of a grid connected microgrid depend on many factors, including different and multiple machine capability limits, local load demands, and distribution line capacities. A mathematical model is developed to establish the active and reactive power transfer capability at the point of common coupling, considering all aspects of grid connected microgrids. Capability diagrams for different microgrid scenarios are derived using the mathematical model, and the applicability of microgrid capability diagram as a tool in the energy market operation is also presented.

The low voltage ride through (LVRT) capability of grid connected microgrids and the potential to provide voltage support as an ancillary service for the main grid are investigated in the final phase of the thesis. Two approaches are followed to investigate the LVRT capability of a microgrid as a single entity. In the first approach, dynamic voltage support at the microgrid point of common coupling is improved by using a distribution static synchronous compensator (DSTATCOM) connected to the low voltage side of the distribution transformer of the microgrid. The collective effect of the LVRT capabilities of the distributed generators in the microgrid is used to provide voltage and reactive power support to the external grid in the second approach. Furthermore, operation of the DSTATCOM in multi-microgrid environment and islanded mode are also investigated under different operating conditions. Impact of the DSTATCOM location in the microgrid is also analysed by installing it at the low voltage side of the microgrid distribution transformer, at distributed gen- erator terminals and at the bus bar with lowest reactive power margin. Variations of the microgrid system parameters during the fault and after fault clearance are analysed to identify the most appropriate location for DSTATCOM operation. It was identified that having the DSTATCOM at the low voltage side of the microgrid distribution transformer is far more beneficial in situations of microgrid transition from grid connected to islanded mode of operation, which would improve the microgrid voltage profile. DSTATCOM operation would reduce the reactive power demand from the external grid which arises due to faults in microgrids containing mains connected induction motor loads.

Based on the studies presented in this thesis, it can be identified that integration of multiple microgrids into the utility grid will allow the microgrids to provide ancil- lary services to the main grid during grid connected mode, and provide emergency services to adjacent microgrids during a utility grid outage. The work presented in this thesis provides the groundwork which will enable microgrids to perform such ancillary services.

Download

Share

COinS
 

Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.