Bounitsis, Georgios;
(2025)
Advances in data-driven stochastic programming and application in power system planning.
Doctoral thesis (Ph.D), UCL (University College London).
![[thumbnail of PhD_THESIS_BOUNITSIS.pdf]](https://discovery.ucl.ac.uk/style/images/fileicons/text.png) |
Text
PhD_THESIS_BOUNITSIS.pdf - Accepted Version Access restricted to UCL open access staff until 1 February 2026. Download (45MB) |
Abstract
The Net Zero race to avert the worst impacts of climate change necessitates transformation of energy systems. As many challenges and uncertainty exist within this context, state-of-the-art methodologies and complex analysis can fortify the decision making surrounding decarbonised and cost-efficient future energy systems. Thus, this thesis contributes both on uncertainty quantification methodologies for data-driven stochastic programming and on power system planning under uncertainty. A novel optimisation-based Scenario Generation (SG) methodology to create small representative sets of uncertainty from historical data is proposed. The methodology exploits statistical analysis, simulation, copula function sampling, clustering and a novel reformulation of the Distribution and Moment Matching Problem as Mixed-Integer Linear Programming (MILP) problem. Quality and stability assessment against state-of-the-art optimisation-based SG methods highlights its benefits on stochastic programming case studies from the Process Systems Engineering literature. Then, the work focuses on Great Britain’s power system planning with hydrogen and ammonia pathways. Spatially explicit snapshot optimisation models are formulated towards optimal planning and operational optimisation of coupled power and heat system in 2040. Regarding the temporal resolution, time series aggregation methodologies are studied. The proposed priority-based chronological time-period clustering approach is indicated to outperform its counterparts and it is integrated in a multi-chronological approach. Systematic analysis using MILP and LP models under various scenarios highlights the roles of renewable technologies, heat electrification, hydrogen and ammonia’s inter-seasonal long-term energy storage for the optimal planning and the cost-efficient operation of the decarbonised system. Finally, the proposed SG methodology and historical data are employed to generate scenarios regarding future’s uncertain wind availability. Hence, a two-stage stochastic programming model for the power system planning problem under wind uncertainty is formulated. Results and Monte Carlo simulation indicate that a stochastic programming approach can provide optimal planning decisions that lead to lower expected total system’s cost and more flexible system operation compared to the deterministic approach.
| Type: | Thesis (Doctoral) |
|---|---|
| Qualification: | Ph.D |
| Title: | Advances in data-driven stochastic programming and application in power system planning |
| Language: | English |
| Additional information: | Copyright © The Author 2025. Original content in this thesis is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) Licence (https://creativecommons.org/licenses/by-nc/4.0/). Any third-party copyright material present remains the property of its respective owner(s) and is licensed under its existing terms. Access may initially be restricted at the author’s request. |
| UCL classification: | UCL UCL > Provost and Vice Provost Offices > UCL BEAMS UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science > Dept of Chemical Engineering |
| URI: | https://discovery.ucl.ac.uk/id/eprint/10203583 |
Loading...
Loading...Loading...Loading...Archive Staff Only
 |
View Item |
