Our paper “Technology selection for green transition using cap-and-trade in supply chains: A bilevel programming approach” (with M. Caramia and A. L. Pala) has been published in Omega — The International Journal of Management Science.
The problem
Decarbonising a supply chain is not a one-shot decision: firms must choose which technologies to adopt, when to adopt them, and how to do so under a regulatory framework — such as the EU Emissions Trading System — that puts a price on every tonne of CO₂. These choices are inherently hierarchical: a regulator sets the cap and the trading rules, while each firm optimises its own investment and operating plan in response.
Our approach
We model this interaction as a bilevel programme: the upper level represents the regulator’s cap-and-trade mechanism, while the lower level captures the supply chain’s cost-minimising technology adoption over a multi-period horizon. Because the resulting problem is non-convex and combinatorial, we design a dedicated solution algorithm that scales to realistic instance sizes.
Highlights
- A novel bilevel formulation that minimises supply-chain emissions over time while respecting a cap-and-trade mechanism.
- A custom algorithm with strong performance on synthetic test instances.
- A case study on the adoption of the hydrogen energy vector in the steel industry, showing how transition technologies can bridge the gap to long-term emission targets.
Why it matters
Heavy industry accounts for roughly a quarter of global CO₂ emissions, and steel alone contributes about 7–9 %. Hydrogen-based direct reduction is widely seen as a key pathway to decarbonise the sector, but the transition requires large, phased investments under uncertain policy signals. Our model offers decision-makers a quantitative tool to evaluate when hydrogen (or other low-carbon technologies) becomes cost-effective under different carbon price trajectories — and how early adoption of intermediate technologies can reduce both cumulative emissions and total transition cost.
Read the full paper via its DOI, or browse all our work on the Publications page.