Agent-Based Model for Steel Decarbonization

This dataset and accompanying source code provide a comprehensive Agent-Based Model (ABM) designed to simulate the long-term decarbonization transition of the steel industry. The model explores the complex interplay between economic factors, policy interventions, and technological evolution in shaping the industry's path towards carbon neutrality.

At the core of the simulation, individual steel production facilities are represented as autonomous agents. These agents make strategic decisions regarding technology adoption, retrofitting existing equipment (e.g., with Carbon Capture or Hydrogen Injection), or transitioning to entirely new production routes (e.g., from Blast Furnace-Basic Oxygen Furnace to DRI-Electric Arc Furnace). Decisions are driven by a Net Present Value (NPV) analysis that considers investment costs, operational expenses, carbon pricing, and constraints on key resources like green hydrogen and high-quality scrap metal.

The model is built using the Mesa framework in Python and is parameterized with extensive data, including:

• Micro-level data on existing steel facilities.

• Forecasts for technology costs and performance improvements.

• Projections for national production targets.

• Provincial-level environmental policies, such as carbon emission caps and air quality standards.

• Future availability scenarios for critical resources.

This repository includes the full Python source code (Agent-Based Model for Steel Decarbonization.py), all input data in CSV format (derived from steel_industry_data_int_en.xlsx), and supplementary documentation to ensure full reproducibility. The model serves as a powerful tool for policymakers, researchers, and industry stakeholders to test scenarios and gain insights into the most effective strategies for achieving a sustainable and low-carbon steel sector.