EAGER: Topology Control for Enhancing the Reliability of the National Power Grid

点击翻译按钮获取中文摘要

Nowadays, fulfilling the demand for electric power is of utmost importance from economic, security, and societal standpoints. Paradoxically, there are very few control actions available to guarantee the power grid's reliability (i.e., to prevent unexpected power outages). This is due in part to the energy industry's conception of the electric transmission network as a static entity. In most situations, the corrective actions in response to a contingency event (i.e., a component failure) consist only of dispatching unused generator capacity and of starting up additional generation units. Conversely, topology control is an innovative approach that equips grid operators with the ability to switch each transmission line's on/off status. Previous research has shown that finding and implementing the best grid configuration (i.e., the best grid topology), in conjunction with the best respective generator dispatch, can decrease energy production/delivery costs. This EArly-Grant for Exploratory Research (EAGER) award supports fundamental research to enhance both the proactive and reactive reliable operation of the power grid without costly infrastructure investments. Specifically, this research will show that controlling the grid's topology can enhance the grid's reliability. In addition, this research will also develop the procedures required to find the most reliable grid topology in response to changes in energy demand. Thus, the primary societal impact of this research is to increase the capability to prevent and resolve unexpected blackouts, which account for approximately $80 billion in losses each year for U.S. businesses and consumers. This research involves several disciplines including power systems, parallel computing and optimization.The specific aim of this award is to enhance the transmission grid reliability. This will be accomplished by incorporating the innovative concept of topology control into representative mathematical programs to optimize the minute-by-minute operations of large-scale power systems. Specifically, the problem that needs to be solved every few minutes is a mixed integer nonlinear problem (MINLP). Three directions are planned to solve the MINLPs that arise when including topology control into the energy system's operations: (1) Combine alternative expressions for modeling alternate current power flows; (2) Devise new decomposition algorithms and combine them with parallel computing; (3) Exploit the fact that, to improve the reliability of the power grid, one needs to solve a MINLP every few minutes; however, since consecutive problem instances are very similar to each other, one can reuse several properties of previous optimal solutions (for example, binding constraints and other properties such as transmission-line switches that have been recurrently effective).