Hybrid Machine Learning and Process Simulation Model for Distillation Column Energy Consumption Optimization

Abstract

Distillation columns are among the most energy-intensive units in chemical processing industries, contributing significantly to operational costs and overall energy demand. Improving energy efficiency in distillation operations is therefore essential for sustainable process optimization and industrial decarbonization. This study presents a hybrid machine learning and process simulation model for optimizing energy consumption in distillation column systems under varying operating conditions. The proposed framework integrates rigorous process simulation with data-driven machine learning algorithms to improve prediction accuracy, operational adaptability, and energy optimization performance. A comprehensive investigation was conducted to evaluate the influence of key operational parameters, including reflux ratio, feed composition, feed temperature, column pressure, tray configuration, and reboiler duty on distillation energy consumption and separation efficiency. Process simulation tools were employed to generate detailed operational datasets, while machine learning models were developed to identify nonlinear relationships between process variables and energy performance indicators. The hybrid framework enabled accurate prediction of column behavior and facilitated intelligent optimization of operating conditions to minimize energy demand while maintaining desired product purity. Experimental and simulation analyses demonstrated significant reductions in reboiler energy consumption and operational inefficiencies compared to conventional optimization approaches. Comparative evaluation revealed improved prediction capability, faster optimization response, and enhanced adaptability to fluctuating feed and operating conditions through integration of machine learning with process simulation. Furthermore, the proposed model supported predictive operational control, process intensification, and real-time decision-making for energy-efficient distillation management.