Energy optimization in aluminium melting operations

Extreme swings in regional electricity prices weighed on the operation. The energy-intensive melting runs were to be shifted into low-tariff hours — but while preserving rigid shift patterns and firmly agreed working hours.

We integrated an AI forecasting service that evaluates real-time grid prices and spot-market developments and proposes optimal four-hour melting windows. The optimization models treat the inherited, untouchable shift windows as a non-negotiable constraint and calculate the efficiency gains strictly around people's availability.

Monthly electricity costs fell noticeably, and did so within the first few months — entirely without touching the workforce's established shift patterns.