Predicting Power Requirement In Steel Manufacturing (EAF)
Problem
In the steel manufacturing industry, Electric Arc Furnace (EAF) and Ladle Refining Furnace (LRF) processes require substantial electrical power, supplied by captive power plants (CPPs). Striking a balance between power supply and the manufacturing demand is a complex task due to the variable nature of manufacturing processes and environmental factors. This often leads to surplus power generation, also known as “infirm power”.
The challenge lies in aligning the power production of CPP with the fluctuating power demand of EAF and LRF. Factors such as the mix of raw materials, atmospheric conditions, and availability of material and assets for subsequent processes complicate this task. Additionally, the limited ramp-up or ramp-down time for power production in CPP can be longer than the 5-minute lead signal provided by the steel plant.
Approach
- Creating a 2-D digital map of the production process
- Developing data connectors and pipelines for real-time and historical data integration
- Creating a real-time dashboard for monitoring KPIs & material and equipment availability
- Developing an AI based dynamic scheduling tool and demand forecasting
- Real time alerts by predicting power consumption and risks of schedule delays
Benefits
- Improved Forecasting: Predict heat start and end times with a lead time of up to 10 minutes
- Reduced Surplus Power: Reduced surplus infirm power production from 15 MW to 5 MW, leading towards zero
- Cost Savings: Savings of approx. $2 Million per year due to reduced infirm power
- Power Production Alignment: With improved forecasting, power production aligns with power demand
- Operational Efficiency: Predictive insights reduced reactive decision-making causing inefficiencies