Instruments and Systems: Monitoring, Control, and Diagnostics Annotation << Back A Digital Twin of Refinery EnvironmentalRisks based on SCADA and ArtificialIntelligence for Early Detectionof Emission Anomalies K.D. SKOBELEV, V.G. BLAGOVESHCHENSKY This article proposes a digital twin of refi nery environmental risks, combining SCADA data and a process section simulation model in R-PRO with artificial intelligence. The system performs early detection of emission anomalies based on video features and process context, calculates an integrated risk index, and generates warnings and visual indicators in the digital environment. The effectiveness of this approach has been validated by bench tests, demonstrating earlier detection of deviations and a reduction in false alarms compared to threshold alarms. Keywords: digital twin, oil refinery, emission monitoring, artificial intelligence, SCADA, R-PRO. DOI: 10.25791/pribor.4.2026.1664 Pp. 19-25.