UVM Theses and Dissertations
Format:
Print
Author:
Casari, Amanada Marie
Dept./Program:
Electrical Engineering
Year:
2011
Degree:
MS
Abstract:
The rapid deployment of smart grid technology is coupling the physical infrastructure of power grids with increasingly complex information.technology networks. Smart grid has numerous potential benefits, but this coupling could expose new vulnerabilities. This thesis describes the design and analysis of supervisory control and data acquisition (SCADA) networks for two different power grids, using both traditional engineering methods and tools from complex systems methodology, to examine the effect of interactions between information networks and the power grid on electricity reliability.
The integration of microgrids as networks of distributed energy generation sources, data acquisition and control systems, and loads connected into the current utility distribution system provides customers both benefits and challenges. Microgrids are more common at larger facilities seeking increased control over the source of their power, such as industrial parks and university campuses, than at the scale of a smaller facility, such as a small national park. With this size change comes changes in customer design preferences. The first part of this thesis describes the design goals for a smaller microgrid system, which facilitates the use of local resources and allows for the addition of new elements without major changes to the existing infrastructure. A preliminary system at Marsh-Billings-Rockefeller National Historical Park in Woodstock, VT is used as an example of practical implementation of developing local energy resources within a small-scale, agent-based microgrid.
On a larger scale, this thesis examines the impact of coupling large-scale power grids with information technology networks in a simple model of a power generation and transmission system connected to a supervisory control and data acquisition (SCADA) network. The power system is modeled as a network with linearized power flow relationships. The SCADA network is modeled as an information network that monitors and controls components of the power network. Failures are injected into the networks, simulating both random failures and directed attacks on either network. Specifically, the proposed model measures the sizes of the blackouts that result from attacks and failures under two control schemes: one that represents historical standard practice and a second that simulates an interactive smart grid. When comparing the resultant blackouts from both schemes, the interactive smart grid model produces a higher risk for blackouts when failures occur in both networks as opposed to individual network failures. Specifically, in the coupled networks case we find that when failures occur in the power system the risk of small blackouts decrease, medium size blackouts increase, and large blackouts decrease. These results are directly relevant for the planning and design of smarter power grid models and real-world systems.
The integration of microgrids as networks of distributed energy generation sources, data acquisition and control systems, and loads connected into the current utility distribution system provides customers both benefits and challenges. Microgrids are more common at larger facilities seeking increased control over the source of their power, such as industrial parks and university campuses, than at the scale of a smaller facility, such as a small national park. With this size change comes changes in customer design preferences. The first part of this thesis describes the design goals for a smaller microgrid system, which facilitates the use of local resources and allows for the addition of new elements without major changes to the existing infrastructure. A preliminary system at Marsh-Billings-Rockefeller National Historical Park in Woodstock, VT is used as an example of practical implementation of developing local energy resources within a small-scale, agent-based microgrid.
On a larger scale, this thesis examines the impact of coupling large-scale power grids with information technology networks in a simple model of a power generation and transmission system connected to a supervisory control and data acquisition (SCADA) network. The power system is modeled as a network with linearized power flow relationships. The SCADA network is modeled as an information network that monitors and controls components of the power network. Failures are injected into the networks, simulating both random failures and directed attacks on either network. Specifically, the proposed model measures the sizes of the blackouts that result from attacks and failures under two control schemes: one that represents historical standard practice and a second that simulates an interactive smart grid. When comparing the resultant blackouts from both schemes, the interactive smart grid model produces a higher risk for blackouts when failures occur in both networks as opposed to individual network failures. Specifically, in the coupled networks case we find that when failures occur in the power system the risk of small blackouts decrease, medium size blackouts increase, and large blackouts decrease. These results are directly relevant for the planning and design of smarter power grid models and real-world systems.