PhD defence of Abdelrahman Ayad – Advancing Electric Grid Architecture: Flexibility-driven Approaches for Optimizing Power Systems Planning and Management
Abstract
The ongoing electric grid modernization efforts strive to develop grid architecture concepts, power systems tools, and technologies to provide a clean, resilient, affordable, and flexible electricity infrastructure. These developments have been essential to achieving critical milestones of grid decarbonization and reliability targets. However, they have simultaneously added several layers of complexity and uncertainty to the grid infrastructure and have affected the fundamental power systems planning and management models.
This dissertation investigates and addresses the challenges of the electric grid’s added uncertainties and complexities in planning models by exploiting power system flexibility. Here, we leverage the broader definition of flexibility as the system’s ability to react to uncertainty and changes. System flexibility is vital in optimizing these transformations while maintaining reliability and lowest-cost solutions. Four flexibility paradigms are investigated: grid architecture flexibility, planning framework flexibility, demand-side flexibility, and network topology flexibility. We develop quantitative models for each form and present case studies to validate their performance.
The dissertation investigates grid flexibility on the architecture level by leveraging Systems Engineering (SE) tools to develop insights and manage its increasing complexity. A numerical framework based on the design structure matrix (DSM) is developed to improve the coordination across the spatial-temporal scales of the grid functions and evaluate the potential benefits of grid designs and the impacts of infusing emerging technologies. The method proposes a technology infusion index (TII) metric to assess the risk-benefit takeoff of these upgrades on the overall grid structure.
To proactively manage long-term planning uncertainties, the dissertation proposes a novel real options in-projects (ROiP) framework to embed flexibility in the transmission expansion planning (TEP) optimization problem. Specifically, a decision-making framework is developed to exercise investment decisions based on the change in the system’s physical features as the uncertainty factors become known. The constraints are added to the TEP optimization model and compared against the stochastic formulation with and without battery energy storage system (BESS). The results reveal the added value of enhancing the system’s strategic flexibility, which lowers the model’s total cost by getting insights from the system’s technical indicators.
On the demand side, a two-step optimization approach is proposed to mitigate the aggregate peak demand from the increasing electric vehicle (EV) loads. The approach combines advanced retail electricity tariffs with electric vehicle managed charging (EVMC) to limit the aggregate distribution feeder load profile and increase the system’s flexibility in managing peak loads. Additionally, transmission and distribution (T&D) coordination is analyzed to understand the cascading impacts of the distribution systems loading on transmission congestion and planning operations.
Finally, a machine learning (ML) model is proposed to improve the solution of the optimal transmission switching (OTS) problem.
The data-driven approach presents a hybrid two-step framework that combines the ML capabilities to predict the power flow patterns and provide the output as an initial solution for the mixed integer linear programming (MILP) problem. We demonstrate the enhanced performance of the proposed approach in improving the optimal solution and reducing the solving time on multiple large-scale test systems compared to the state-of-the-art MILP solvers.