How Dillinger chose where to put their EAF — and avoided double-digit millions in unnecessary CAPEX

The trade-off

Dillinger Hütte, part of Stahl-Holding-Saar, is transitioning from BF-BOF to a DRI-EAF production route as part of their decarbonization master plan. The central question: where in the existing melt shop should the new Electric Arc Furnace go?

Two locations were identified. Both would fundamentally change the transport routes and ladle logistics across the entire plant. And they couldn’t be more different.

Bay 2: the “safe” choice

Bay 2 seemed like the obvious location. Steel would flow from the same direction as the existing BOF. Entry to secondary metallurgy stays consistent. The logistics felt familiar — everyone could imagine how it would work.

One problem: there’s a building in the way. Demolition and rebuild would cost double-digit millions — guaranteed. But at least the logistics were known.

Bay 5: the cheaper unknown

Bay 5 didn’t require relocating any buildings. Lower CAPEX by a significant margin. But the logistics looked complicated. Transport distances seemed too long. Ladle routes would change completely. Nobody was confident it would actually work at the required production targets.

The real question: pay the millions for the “safe” choice, or take a chance on an unproven layout?

What FESIOS did

We built a 3D production and logistics simulation of the entire melt shop and ran both scenarios against three non-negotiable objectives:

1. Maintain current productivity of 46 heats per day
2. Enable hot metal charging at the EAF
3. Keep BOF production undisturbed during the transition

The model covered the complete material flow — ladle transport, crane coordination, transfer cars, secondary metallurgy, and casting. Each layout was iteratively optimized. We didn’t just simulate the initial concepts — we refined each option to its best possible configuration before comparing them.

The surprises

In the initial layouts, Bay 5 couldn’t charge hot metal to the EAF and Bay 2 disrupted BOF production. After iterative optimization — adding a transfer car for Bay 5, adding a new crane for Bay 2 — both concepts met all three objectives.

But the differences in robustness were stark.

Bay 5 works cleanly. All objectives met. The longer transport distances, once optimized with an additional transfer car, didn’t limit throughput.

Bay 2 had hidden risks. Temporary bottlenecks emerged around the transfer car area (TC3/TC4) and crane interactions. Worse: if the new crane goes out of operation, production is blocked — a single point of failure. Long transport times from ladle furnace to vacuum degassing added further strain. And on top of all that: the guaranteed building demolition cost.

The decision

Bay 5. The “risky” option turned out to be the robust one. The “safe” option had hidden logistical risks — plus a guaranteed eight-figure price tag for building demolition.

The simulation didn’t just compare throughput. It exposed failure modes that static planning and engineering judgment alone couldn’t predict.

This study was presented at ESTAD 2023 (METEC, Düsseldorf).