How ArcelorMittal Belval validated a 15% production increase before investing

The challenge

ArcelorMittal Long Products Luxembourg was planning the biggest investment at its Belval site since switching from blast furnaces to electric arc furnaces in the 1990s. The goal: supply the rolling mill in Rodange with CO2-reduced steel produced in Luxembourg instead of sourcing it from Poland and Germany.

That meant a 15% productivity increase, new steel grades, a completely revamped EAF with higher tapping weight, a shift from two-bucket to one-bucket scrap charging, and a brand-new vacuum degassing plant. Every part of the plant would be affected — from scrap delivery to casting.

The question was straightforward: will this actually work?

What FESIOS did

We built a full discrete event simulation covering the entire production chain:

• Scrap yard — truck arrivals (one every 8 minutes), 20-wagon train deliveries, two portal cranes on a shared rail, mobile material handlers, height-sensitive scrap piles
• Melt shop — revamped EAF, new VD, deslagging on the crane hook, two overhead cranes with full route planning and inter-crane coordination
• Casting and run-out — multi-format strands, cooling beds, pusher tables with measured mechanical speeds
• Production planning — three distinct production plans covering different format combinations, scaled to a full year

The simulation ran against real production data from the highest-output period and was continuously validated throughout the project.

Did the cranes keep up?

One of the biggest concerns: adding VD transports, deslagging, and 15% more heats — would the two cranes become the bottleneck?

We analysed crane utilization not as a yearly average but during the specific periods of peak demand:

• Phase 1 (fastest production, shortest casting times): crane utilization increased by 12% — easily manageable
• Phase 2 (VD and deslagging every heat): crane utilization increased by 23% — close to maximum, but still no production delays

The cranes could handle it. No additional crane investment needed.

How much faster can you load scrap?

Switching from two-bucket to one-bucket charging changed the entire scrap yard dynamic. Initially, we ran the future production with the existing scrap pile layout. The result: the eastern crane’s bucket had to travel much further than the western one, causing late arrivals at the EAF.

We developed an optimized layout that repositioned scrap types within the pit. The impact:

• West crane bucket: 3 minutes faster on average
• East crane bucket: 7 minutes faster on average

We also discovered a critical threshold: loading time increases dramatically once the scrap pile drops below bucket height (~8 meters). Above that level, no further improvement. This insight alone changes how you manage pile replenishment.

1.3 million tonnes per year — confirmed

The simulation confirmed that the production target of 1.3 MTPY is achievable across all three production scenarios. Better still: depending on the scenario, there were 4 to 13 unscheduled buffer days remaining at year-end — valuable flexibility for maintenance or unplanned downtime.

The result

The simulation validated the entire modernization concept before a single Euro of CAPEX was committed. ArcelorMittal Luxembourg proceeded with the investment, with the first heat expected in 2025.

This study was presented at the European Electric Steelmaking Conference (EEC) 2024 and ABM Week 2024 in Brazil.