Optimising Industrial Participation in the Day-Ahead Electricity Market
A Stochastic Bidding Framework with Risk Management
M. Badarinath (TU Delft - Electrical Engineering, Mathematics and Computer Science)
L.J. de Vries – Mentor (TU Delft - Energy and Industry)
M. Cvetkovic – Mentor (TU Delft - Intelligent Electrical Power Grids)
C. Doh Dinga – Mentor (TU Delft - Intelligent Electrical Power Grids)
Seyed Hossein Jamali – Graduation committee member (Tata Steel, 1970 CA IJmuiden, the Netherlands)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
Increasing electricity price volatility in European energy markets presents a significant financial challenge for large industrial consumers. While academic literature has explored demand-side bidding, a knowledge gap persists in applying these strategies to complex, process-driven industries with limited operational flexibility and highly interdependent systems. This thesis aims to bridge that gap by designing, implementing, and evaluating an adaptive stochastic bidding framework to enable industrial consumers to optimise their participation in the day-ahead electricity market, minimising cost while managing risk.
A two-stage stochastic Mixed-Integer Linear Program (MILP) was developed to formulate and compare two distinct, EUPHEMIA-compatible bidding strategies: a granular stochastic hourly bidding strategy and a holistic stochastic exclusive group bids strategy. The framework incorporates Conditional Value-at-Risk (CVaR) for downside risk management and models price uncertainty using a combination of Meta's Prophet forecasting model and a Levy stable distribution to generate realistic, heavy-tailed price scenarios. The model's logic was first verified on a simplified green hydrogen system before being applied to a detailed case study of Tata Steel's IJmuiden plant, analysing both its current rigid blast furnace-basic oxygen furnace (BF-BOF) and future flexible direct reduction plant-electric arc furnace (DRP-EAF) configurations.
The core finding is that the optimal bidding strategy is fundamentally contingent on the industrial asset's specific operational flexibility and economic structure. For the current, inflexible BF-BOF system, a price-insensitive strategy that prioritises material efficiency is superior, as the financial penalties from disrupting the production chain far outweigh potential electricity cost savings. Conversely, for the future, flexible DRP-EAF system, a granular, price-sensitive hourly bids strategy becomes the most profitable approach, creating significant value by leveraging the Electric Arc Furnace for price arbitrage. Furthermore, for flexible assets whose profit margins are primarily defined by electricity costs, such as the green hydrogen system, a conservative exclusive group bids strategy is optimal due to its superior risk hedging, which prioritises capital preservation in volatile markets.
This research concludes that a universally optimal bidding method does not exist. Effective market participation requires that industrial consumers first diagnose their system’s unique techno-economic architecture and then deploy a strategy that aligns with its inherent nature: either insulating rigid processes from market volatility or actively engaging flexible assets with it.