ITR - (ASE+NHS) - (dmc+int): Toward a Multi-Layered Architecture for Reliable and Secure Large-Scale Networks:
This project develops an information technology (IT) framework that addresses basic problems in critical national infrastructures that can lead to catastrophic failures, especially when the system operates near full capacity. These systems use large stand-by reserves to manage the variations in operating conditions and equipment status. When the system is under stress, reserves make up for lack of flexible decision making. Current modeling, monitoring, and decision tools assume that system response to disturbances is local, and that it is also locally manageable without on-line coordination. While these tools generally result in reliable operations, the occasional costly and life-threatening blackouts, explosions, and water main breaks are proof that the tools and "conservative" operating practices can fail. Continued growth, technology change, aging equipment, and regulatory changes alter the system, making the old models and practices less reliable. Operating these infrastructure systems to achieve higher performance along with greater reliability and security requires advances in modeling, simulating, and operating complicated systems. This project focuses on the electricity grid, using the widely accepted representations. While simplifications of reality, these representations of the infrastructure grid are highly complex models. The project research treats the following important issues: (1) Handling Complexity. The project develops layers of simpler models that run in real time and capture different time and space resolution scales, with a "systems manager" layer that corrects for the "externalities" caused by approximate decomposition. (2) Monitoring. A network using sparse sensors strategically placed measures the most relevant data at the location of impact. (3) Characterization of the System Dynamics. Unlike the state-of-the-art static state estimators used in current power grids, this project develops a "quasi-dynamic state estimator" (QDSE) that uses the current knowledge (estimate) of the state to update the current estimate of the network state. (4) Institutional Environment. This project seeks to integrate the regulatory restructuring and operations models have not been integrated for the electricity infrastructure.
The project (a) develops an IT framework for smooth, reliable, and efficient operation of a complex network infrastructure over a broad range of inputs/disturbances and topological changes; (b) creates a family of models that allow informed decisions through better representations of the network state and dynamics; (c) identifies the information and models required to run the network reliably and efficiently, and (d) seeks the most cost-effective way to gather and process the required data, and build a better model of system dynamics. To simplify computation and allow the model to run in near real time, models are decomposed using a "systems manager" to correct for the externalities arising from inexact decomposition. An Interplay of physical, financial/economic, regulatory, and IT signals drives the complex infrastructure system. These general concepts are applied to the electric power infrastructure, which has recently been challenged for its lack of reliability and inability to support efficient markets. This project demonstrates the outlined techniques by assessing the performance of several candidate architectures of the evolving electric power industry.
