Batch distillation with fractionation processes: Modeling, Optimization, and Software Review

Remat Chemie has been recognized in the industrial market as one of the go-to companies for distillation in the Netherlands. The vision of the company is to re-duce greenhouse gas emissions and support the development of a circular economy through solvent regeneration. The regenerated solvents are returned to the customer (owners of contaminated mixtures) or back to the market. This project aimed to identify the most suitable engineering software for simulating their complex, dynamic distillation processes, and to develop a model of a fractionation column for a specific regeneration process of a valuable solvent, optimization of said process in terms of cost, time, and yield. Three simulation tools -Aspen Plus, Mobatec, and BatchColumn (Fives ProSim)- were evaluated. Based on critical criteria such as thermodynamic modeling, dynamic capabilities, usability, and cost, BatchColumn was selected. A key focus was building an accurate thermodynamic package using the NRTL model, sup-ported by Simulis Thermodynamics. Binary Interaction Parameters (BIPs) for 28 components were completed through software databanks, regression, and estimation methods. The solvent regeneration industrial process operated in September 2024 was modeled as a reference case. The simulation was validated against lab data after adjusting using an improved lab calibration and thermodynamic updates with acceptable agreement between simulation and industrial results. Four industrial optimization trials were conducted, adjusting operational strategies such as reflux ratio, flow rates, and cut periods. The optimized case showed good improvements: process yield increased from 43% to 65%, solvent yield from 70% to 88%, with a longer process time and higher energy use and despite a 2.6% rise in costs, revenue increased by 48%. An increase of profit by 2.9 times in this case showed more profitability than the current process.