Abstract
The electricity sector is undergoing a significant transformation, driven by the large-scale integration of Variable Renewable Energy Sources (VRES) and the growing demand-side electrification and flexibility. This shift challenges the existing electricity market, originally designed around dispatchable power plants and a highly inelastic demand. This incongruity between the current market
... read more
design and the evolving electricity landscape could exacerbate existing market inefficiencies. This thesis aims to develop efficient wholesale market designs suitable for the evolving electricity sector, with a particular focus on the short-term operational timeframe in European electricity markets, spanning from day-ahead to real-time operation.
Towards this end, this thesis begins with a thorough evaluation of market obstacles that can hinder the efficient market functionality. It also classifies an extensive range of market modifications that were proposed in the literature. Subsequently, a detailed analysis is carried out to determine their potential to alleviate these barriers and enhance market performance, thus identifying notable areas that require further dedicated analyses. Indeed, one of the main obstacles facing current market designs, is the growing generation stochasticity, which can lead to expensive balancing actions. To minimize this risk, uncertainty-based market clearing mechanisms have been proposed in the literature to replace the current deterministic day-ahead (DA) market formulation, which does not consider the stochastic nature of VRES.
To address practical limitations of existing methods, this thesis then proposes a novel uncertainty-based DA market-clearing approach based on Light Robust (LR) optimization. The proposed LR DA market-clearing integrates the uncertainty of VRES in its market-clearing formulation through a new bid format that allows stochastic producers to provide uncertainty ranges around their offered quantities. The LR formulation also enables the market/system operator to enhance the market's robustness to uncertainty while limiting the potential impacts on the social-economic welfare. Moreover, the thesis advances by proposing a novel market design that co-optimizes the DA market and the procurement and activation of reserves using LR optimization. This LR formulation allows increased robustness against uncertainty while the coordinated design results in a more efficient use of generation capacity as compared with the current setup where energy and reserves markets are disjointly cleared. To account for the stochasticity of VRES, three different formulations for the specification and sizing of the system reserve requirements are also proposed, including fixed reserve requirements, variable requirements based on the available uncertainty, and a hybrid approach.
Finally, this thesis conducts an analysis of the strategic behavior of VRES when participating in a DA LR market model. In this analysis, a stochastic optimization problem (representing the bidding process of stochastic producers) constrained by other optimization problems (representing the LR market clearing and price formation problems) is formulated to determine whether stochastic producers are incentivized to bid their true uncertainty ranges or strategically deviate to maximize their profits. The analysis includes a comparison between the single and dual imbalance pricing schemes to assess potential differences in the stochastic players' bidding behavior and their impact on the overall balancing of the system.
show less
