Adaptive Relay Settings for Wide Area Coordination of the Power Grid

The reliable operation of the electrical power grid is foundational for our society, and it requires an active protection scheme to guard against failures. Protective relays play a critical role in grid protection, but their configuration is challenging and is growing more difficult with the grid’s increasing complexity. A well-engineered, predictable relay configuration (i.e., relay settings) must include a coordination of its activities with neighboring relays, but a full coordination analysis strains the limits of what engineers can reasonably do with the resources allocated to such a project. We will develop an automated coordination solution for relay settings in wide area applications, substantially reducing the time and effort requiring for this critical activity. We will also increase the type and scope of studies that can be done, allowing for more realistic simulation of the conditions deployed relays encounter in power systems. We will achieve this with a combination of (a) novel grid topology analysis drawing from graph traversal algorithms in computer science, (b) machine learning algorithms that will be driven by a coordination constraint system, (c) software interaction with established short-circuit models in the field, and (d) a validation process-driven analysis of experimental data of real-world grids and wide area coordination studies performed for multiple utilities. In Phase 1, we will use existing industry leading short circuit programs to implement a prototype software to retrieve fault studies and create auto-coordinated relay settings on real world grids with realistic constraints matching the requirements consistent with utilities’ standards. We will address coordination of heterogenous protection elements (e.g., directional time overcurrent and distance elements), analysis of common contingency cases (e.g., N-1, N-2), and criteria relaxation when full coordination is provably infeasible. The prototype will be evaluated by comparing the quality of produced solutions with that of studies performed by more conventional means. One future use of this advancement is integration into a software product for system protection engineers at both utilities and the engineering consultancies that serve them that would enable them to rapidly create coordinated relay settings. A long-term application is in a more autonomous relay configuration system that could re-coordinate relay systems in the grid. This solution will provide a widespread benefit to the United States, especially when different distributed generation sources are inserted in the system. In addition, this advancement will equip utilities and service providers to use the latest technology on the power grid and cope with the relay coordination complexity without compromising reliability.