Role of technology learning in the decarbonization of the iron and steel sector : An energy system approach using a global-scale optimization model

Andrea Moglianesi*, Ilkka Keppo, Daniele Lerede, Laura Savoldi

*Corresponding author for this work

Research output: Contribution to journalArticleScientificpeer-review

6 Citations (Scopus)

Abstract

The iron and steel sector, characterized by fossil fuel-driven processes is one of the most difficult to decarbonize and a significant source of greenhouse gas emissions. Various new technologies promise to change this, but their development is highly uncertain. This paper aims to analyze the prospects of key low-carbon technologies in the sector, focusing on the impact of technology learning, in the light of the uncertainty related to the learning rate. An optimization energy system model was used with an iterative learning formulation, adopting different learning assumptions. The results show that learning may have only a minor impact in the short and medium term, reducing global carbon emissions of the sector by 3% (at most) in 2050, compared to a non-learning scenario. In the long term, high learning potentials for novel processes are important, leading to a market share of up to the 80% by the end of the century. The learning potential for Carbon Capture and Storage processes, however, plays no role in the simulations. Early investments and research and development can help unlock the full potential of the technologies, while more detailed studies should be performed to better understand the retrofitting impact in the shorter term.

Original languageEnglish
Article number127339
JournalEnergy
Volume274
DOIs
Publication statusPublished - 1 Jul 2023
MoE publication typeA1 Journal article-refereed

Keywords

  • Decarbonization
  • Energy system optimization modeling
  • Iron and steel
  • Technology learning
  • TIMES model

Fingerprint

Dive into the research topics of 'Role of technology learning in the decarbonization of the iron and steel sector : An energy system approach using a global-scale optimization model'. Together they form a unique fingerprint.

Cite this