Charge mix decisions on an electric arc furnace happen fast, often within minutes of scrap yard inventory data and current market pricing, yet those decisions determine metallic cost per liquid ton, tap chemistry compliance, and energy consumption for the entire heat. Process engineers balancing scrap grade availability, DRI and HBI blending ratios, and pig iron additions are effectively solving a constrained optimization problem under time pressure using spreadsheets and experience, which leaves real savings on the table when market conditions shift faster than manual recalculation can follow. AI-driven charge mix optimization from iFactory continuously recalculates the lowest-cost charge combination that still meets chemistry and yield requirements for every heat.
EAF Charge Mix Optimization: AI-Driven Scrap, DRI, and Pig Iron Blending
iFactory analyzes real-time scrap yard inventory, grade quality data, and market pricing to recommend the lowest-cost charge combination for every heat, maintaining chemistry compliance and tap temperature targets while reducing metallic input cost per liquid ton.
The Cost Impact of Charge Mix Decisions Across a Melting Campaign
Metallic input typically represents the single largest cost category in EAF steelmaking, which means even small percentage improvements in charge mix efficiency translate into significant savings when applied consistently across thousands of heats per year. The challenge for process engineers is that the optimal mix changes constantly as scrap grade availability, market pricing, and target grade chemistry shift, making static charge recipes a source of ongoing cost leakage rather than a one-time planning exercise.
Charge Mix Components: What the Model Balances Every Heat
iFactory connects to your scrap yard inventory system, market pricing feeds, and historical heat chemistry data to recommend the lowest-cost charge combination that meets your grade requirements, updated continuously as conditions change.
Manual Charge Planning vs AI Continuous Optimization
| Charge Planning Task | Manual Spreadsheet Planning | iFactory AI Optimization |
|---|---|---|
| Price Response Speed | Charge recipes updated periodically, often lagging behind daily or hourly market price shifts | Recommended mix recalculated continuously as pricing and inventory data update |
| Grade Combination Coverage | Engineers typically evaluate a limited set of familiar combinations due to time constraints | Model evaluates the full range of available grade combinations against cost and chemistry constraints |
| Yield Loss Estimation | Yield assumptions often based on general rules of thumb rather than specific mix history | Yield estimated from historical performance data specific to the actual mix being considered |
| Chemistry Risk Management | Tramp element risk assessed manually, with safety margins sometimes added conservatively | Tramp element dilution calculated precisely against target chemistry, reducing unnecessary safety margin cost |
Before and After AI Charge Mix Optimization
Expert Perspective
Charge mix planning used to be something one of our senior engineers did in a spreadsheet every morning based on the previous day's scrap prices, which meant we were always working with slightly stale numbers by the time the actual heats were charged that afternoon. The AI model recalculates the optimal mix continuously, and what surprised me most was how much of our savings came not from finding some exotic new combination but simply from responding faster to price movements we already knew about. Our metallic cost per liquid ton has come down meaningfully since deployment, and our chemistry compliance rate has actually improved because the model is more precise about tramp element dilution than our old safety-margin approach ever was.
Frequently Asked Questions
iFactory helps process engineers balance scrap quality, DRI and HBI blending, and pig iron additions against real-time cost and chemistry requirements, capturing savings that static charge recipes leave on the table.
