Sizing of Hybrid Power Plants Considering Operational Strategies

The global energy transition toward renewable energy sources is accelerating, driven by ambitious climate targets and supportive policy frameworks. However, renewable energy generation’s inherent intermittency and unpredictability present significant challenges for cost-effective grid integration. Hybrid Power Plants (HPPs), combining complementary renewable energy sources such as wind and solar photovoltaic (PV) with battery storage, offer a promising solution by providing more predictable, stable, and economically viable energy outputs.

Optimal sizing of HPPs is required to evaluate the techno-economic potential of HPPs. Effective sizing involves careful capacity allocation among wind, solar, and storage assets, directly impacting revenue generation, grid integration efficiency, and system reliability. Inaccurate sizing decisions, such as underestimating storage or overestimating generation capacity, can lead to energy curtailment, elevated operational costs, and reduced profitability. Conversely, overly conservative sizing may result in under-utilization of valuable grid infrastructure. Capturing these dynamic and non-linear interactions, influenced by weather variability, market price fluctuations, and battery operation, requires high-fidelity Energy Management System (EMS) models. Using simplified models risks misrepresenting operational strategies and undervaluing revenue optimization opportunities, thus leading to suboptimal design outcomes. Therefore, embedding detailed EMS simulations into the sizing process ensures that theoretical designs translate into practical, cost-effective solutions in real-world conditions.

This thesis addresses these complexities by developing advanced methodologies for optimally sizing utility-scale HPPs. It examines the detailed interplay among renewable resource variability, electricity market dynamics, operational uncertainties, and technology cost fluctuations. A significant emphasis is placed on integrating high-fidelity EMS models into sizing optimization without imposing prohibitive computational burdens, thereby effectively bridging the gap between modeling accuracy and practical feasibility in HPP design.

Three core research questions guide this work:

1. How can comprehensive energy management system (EMS) models be effectively integrated into the HPP sizing optimization framework?
2. How do uncertainties, including fluctuating technology costs and operational variability (weather and electricity price forecasts), impact optimal HPP design decisions?
3. To what extent should forecast uncertainties be considered beyond operational EMS, and how can they be systematically incorporated into the sizing process?

The thesis progresses through three interconnected studies:

• The first study finds that computational speed and model fidelity can be effectively reconciled by creating a data-driven surrogate of a high-fidelity EMS. This surrogate accurately emulates detailed market participation and battery dispatch behavior, with minimal error, yet delivers answers in a fraction of the time. While sizing optimization was not conducted in this initial stage, this advancement lays a critical foundation for practical and efficient sizing assessments.
• The second study leverages this surrogate-based approach to directly integrate wind and storage technology costs uncertainties into the sizing optimization. By sampling plausible cost distributions for wind turbines and batteries, the optimizer identifies configurations that not only increase expected returns but also manage sensitivity to cost swings. The key insight is that tolerance for cost variability steers the design: when risk tolerance is high, developers gravitate toward designs with a heavier wind focus, whereas those seeking steadier returns allocate more capacity to battery power. In this way, sizing that explicitly incorporates cost uncertainty produces HPPs designs that are both more resilient and more profitable than traditional, single-scenario approaches.
• The third study then turns to another crucial source of uncertainty—short-term forecast errors in wind, solar and electricity prices—and quantifies its impact on both operation and sizing decisions. It reveals that purely deterministic EMS models significantly overestimate profitability when confronted with realistic forecast errors. Introducing a stochastic EMS, which accounts for multiple forecast scenarios, substantially mitigates these losses and reduces operational penalties. Moreover, extending this approach with a surrogate model allows for the comprehensive integration of forecast uncertainties into sizing optimization. This results in new optimal configurations that strike a balance between maximizing mean profitability and limiting profit variability. Altogether, this work underscores that hybrid plant sizing must embrace short-term forecast uncertainty at both the dispatch and design stages to realize robust, profitable outcomes

Ultimately, this thesis contributes advanced methodologies for HPP sizing, highlighting the crucial role of uncertainty management in enhancing the economic and operational feasibility of hybrid renewable energy solutions. These methodological advancements are essential for facilitating the large-scale deployment of HPPs, contributing to the broader global energy transition.