A Steady-State Risk Analysis and Mitigation Framework for Power Systems

Uncertainty in electricity systems is rising partly due to the electricity generation resource mix transforming from primarily fossil-fuel dependent to increasingly reliant on variable renewable energy (VRE) resources. The increasing uncertainty in electricity systems due to the changing resource mix could introduce risks to the reliable planning and operation of electricity systems without new methods and tools. 

One area of concern is day-ahead operations. During day-ahead operations, operators schedule generation, procure ancillary services, and set reserve requirements 24 hours before operation to meet the next day’s load demand and ensure grid performance metrics remain within system operating limits. Commercial systems-level steady-state analysis and optimization tools used in day-ahead operations do not use uncertainty analysis and optimization methods. Uncertainty analysis and optimization methods could help assess and mitigate reliability risks to electricity systems introduced by VRE resources. 

This thesis develops a steady-state uncertainty analysis method, an optimization formulation, and a framework for deploying the approaches. The method, formulation, and framework evaluate the risk of grid performance metrics violating system operating limits in day-ahead electricity systems operations caused by VRE resources. 

The steady-state uncertainty analysis method, known as Risk-Managed Steady-State Analysis (RMSA), can assess system operating limit violation risks to power grids due to the worst-case power uncertainty of VRE generators. Results for hundreds of scenarios show that RMSA estimates worst-case bus voltage magnitude and line flows without significant loss of probabilistic accuracy and provides runtime speedups of up to 21x when compared to parallelized Monte Carlo analyses using 32 CPU cores. 

The steady-state uncertainty optimization method, Risk-Managed Steady-State Optimization (RMSO), is a nonlinear optimization that uses outputs provided by RMSA to change the electricity generation dispatched on a grid and prevent system operating limit violations. Results for hundreds of scenarios demonstrate that an RMSO implementation found a feasible solution for 100% of scenarios assessed without significantly changing the generation dispatched on the grid. 

The framework and software package for deploying both methods is the Solver for Uncertainty in Power and Energy Resources (SUPER). An analysis of a 2030 synthetic New York transmission system model shows that by using SUPER, it is possible to identify and mitigate the system operating limit violation risks to the New York system. In sum, the methods and framework created could empower grid operators to perform system-level steady-state analyses and optimizations that mitigate uncertainty.