This work aims to develop an economic optimization and experimental validation of ethanol separation by extractive distillation using an ethylene glycol - glycerol mixture as separating agent with and without thermal integration. An anhydrous ethanol production plant with a capacity of 300 000 L/day was simulated through a rate-based approach in Aspen Plus V10®. The optimization problem was solved, firstly, through a rate-based simulation, using a Sequential Quadratic Programming (SQP) algorithm available in Aspen Plus® and secondly, through a stochastic algorithm (genetic) available in Matlab R2019a®. US$ 0.1034 was the optimized TAC/kg of anhydrous ethanol, this parameter reduced by US$ 0.0004 to US$ 0.0006 when thermal integration was included and differed by less than US$ 0.0038 between both methodologies, where the lowest value was found using the stochastic algorithm. Experimental validation was conducted on a pilot scale extractive distillation to verify the results from the process optimization. The use of the 60–40 mol% ethylene glycol - glycerol mixture showed a 40.1% reduction in the total energy consumption of the process compared to the case where only ethylene glycol is used as entrainer. This allowed to confirm the synergistic effect of the solvents mixture and the possibility of employing them.