Load flexibility profiling and power quality assessment in industrial substations: A case study from silicon smelting operations

This study presents a comprehensive analysis of real-time electrical data from a silicon metal smelting enterprise in Yongdeng County, Lanzhou City, China. Over a continuous 7-day period (October 10–16, 2022), high-resolution 15-minute interval data were collected from four key substations, capturing detailed operational characteristics across 15 power-related parameters. The analysis focuses on two critical aspects of industrial energy systems: load flexibility and power quality. Flexibility is quantified through a novel Load Flexibility Index (LFI), integrating duty cycle, ramp rate, variability, energy usage, and peak demand. Power quality is assessed using real-site indicators such as the Voltage Unbalance Factor (VUF), Current Distortion Percent (CDP), and Instantaneous Power Quality Score (IPQS). Unlike prior literature relying on aggregated data or simulation-based assumptions, this study leverages high-granularity data tied to actual equipment infrastructure, including 30 reactive compensation circuits totaling 750 kVAr . Quantitative assessment of capacitor switching events reveals an average 8.6% reduction in voltage unbalance and a 5.3% improvement in power factor stability, highlighting the stabilizing impact of reactive compensation on real-world industrial grids. The findings provide actionable insights for power-aware operational improvements and demand-side strategies in heavy industrial environments. By linking data-driven analysis with practical load management metrics, this study provides a transferable framework for improving energy efficiency and grid interaction in other heavy industrial settings. • Load flexibility profiling using 15-min data from a silicon metal smelting plant. • Multi-dimensional Load Flexibility Index (LFI) covering duty cycle, ramping, and demand. • Real-site power quality assessment tied to 750 , kVar compensation across 30 circuits. • Operational insights linking load flexibility to power quality performance in smelting substations. • Foundation for forecasting and model predictive control in iron and steel enterprises.