guest ::
%0 Thesis %A Papadimitriou, Chrysanthi %T Dynamic optimization strategies for scheduling of chemical processes under time-varying electricity prices %V 43 %I Rheinisch-Westfälische Technische Hochschule Aachen %V Dissertation %C Aachen %M RWTH-2025-10005 %B Aachener Verfahrenstechnik series - AVT.SVT - Process systems engineering %P 1 Online-Ressource : Illustrationen %D 2025 %Z Veröffentlicht auf dem Publikationsserver der RWTH Aachen University %Z Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2025 %X Demand-side management enhances grid stability by aligning electricity consumption with power availability in markets with a high share of renewables. In this context, process scheduling enables electricity-intensive processes to enhance efficiency and meet production constraints while profiting from spot markets by adjusting production to price fluctuations. Achieving such flexibility requires incorporating process dynamics into scheduling decisions, particularly when system and price time scales align. While energy market volatility suggests online process decision-making, embedded dynamics induce computationally challenging dynamic optimization problems. This dissertation develops solution approaches for dynamic scheduling and methods to exploit electricity spot markets, addressing economic and computational barriers in industrial applications.We first extend a global dynamic optimization approach to nonconvex scheduling with dynamics embedded, using Hammerstein–Wiener models trained on experimental data from an electrochemical process. The method demonstrates strong economic potential in scheduling succinic acid recovery but is unsuitable for real-time use. To address this, we introduce control grid refinement to accelerate computations and reach high-quality solutions, though computational intractability remains.Next, we highlight the impact of electricity price scenario selection in demand-response studies. To avoid arbitrary choices and scenario explosion, we propose a framework for generating day-ahead and intraday price profiles from historical data. The method captures key data features, requires minimal computational effort and yields results comparable to state-of-the-art techniques over a full year of operation.Systematic consideration of day-ahead and intraday prices enables market participation across different time scales, aligning with process dynamics in scheduling and control. We propose a two-economic-layer scheme for scheduling and control of processes participating in both markets. Building on integrated dynamic scheduling and economic model predictive control, we perform scheduling under day-ahead prices followed by intraday adjustments and market trading. Applied to an air separation process, the scheme, unlike common methods, achieves consistently high economic performance.Building on the proposed methodologies, we investigate the trade-off between model fidelity and optimization complexity in dynamic scheduling within integrated scheduling and control. We compare full-order and reduced-order models under local optimization with scale-bridging models versus global optimization under varying formulations. Application to air separation shows local optimization with full-order models offering an improved economic and control performance, while local optimization with nonlinear reduced-order models is suitable when process data are available. %F PUB:(DE-HGF)11 ; PUB:(DE-HGF)3 %9 Dissertation / PhD ThesisBook %R 10.18154/RWTH-2025-10005 %U https://publications.rwth-aachen.de/record/1022412