Optimization-Based Active Power Management for Operational Planning of a Multi-Energy System

Abstract

The ongoing transition toward low-carbon energy systems is marked by a rapid expansion of renewable energy sources, particularly offshore wind. However, the variability and limited predictability of wind generation introduce significant operational challenges, especially when conventional fossil-based units that historically provide stability and dispatchability are reduced. Ensuring continuous supply, while maintaining system efficiency and minimizing curtailment, requires new forms of flexibility that can balance supply and demand over different time scales. Hydrogen technologies have emerged as a promising candidate for this purpose due to their ability to store large amounts of energy and reconvert it when required. Yet, their effective deployment depends on coordinated planning and operational strategies, which are often treated separately or simplified in existing studies.

This thesis proposes a bilevel optimization framework that jointly considers long-term planning and short-term system operation for a wind-integrated power system employing hydrogen fuel cells and a hydrogen-fired gas turbine. The planning stage determines the siting and sizing of these units using Particle Swarm Optimization (PSO), while the operation stage employs a Mixed-Integer Linear Programming (MILP) model to dispatch generation and assess feasibility under network constraints. To represent realistic renewable variability, a stacked Long Short-Term Memory (LSTM) model is trained to generate a one-year offshore wind profile that captures daily and seasonal fluctuations. The interaction between the planning and operation layers ensures that the final configuration is both economical and operationally robust.

Application of the framework to a modified IEEE RTS-24 system with offshore wind from Ijmuiden Ver Alpha demonstrates clear benefits. The coordinated hydrogen-to-power configuration reduces annual wind curtailment from 4.42 TWh to 1.65 TWh, decreases load shedding from 452.9 GWh to 18.0 GWh, and lowers total system operating costs from \$313.88 million to \$239.29 million. Additionally, the system maintains adequate supply across seasonal conditions, confirming the value of hydrogen-based flexibility in high-renewable environments.

Overall, this research contributes a structured planning–operation methodology that integrates realistic wind modeling, hydrogen conversion dynamics, and power system constraints. The proposed framework provides a scalable and practical approach for supporting the reliable and cost-effective transition toward future renewable and hydrogen-integrated energy systems.